December 29, 2017

In this piece, we assess the state of the diabetes/obesity fields going into 2018. We look back on 2017 in diabetes therapy, diabetes technology, big picture topics, and obesity to note scientific strides and surprises, to celebrate progress and take stock of remaining challenges, and to identify pressing questions for the year ahead. Within each section, you’ll find major themes, a list of highly-read reports (top 10 for the year!), what we got right/expected, what we got wrong/didn’t expect, key questions for 2018, and updates on what’s coming in 2018.

By and large, this Reflections piece is our favorite to write all year – and that’s up against the 945 other reports we’ve published since January 1, 2017! (If you’re interested, this “2017” filtered search on the CC Knowledge Base will let you scroll through every report we published this year.)

It’s an exciting time for the field, and for the first time in quite a long time, we sense real optimism as science advances, business/reimbursement models evolve, and diabetes care becomes more holistic – How do we think beyond glucose? How do we make patients’ lives easier in big and small ways? How can we rethink traditional drug delivery, care delivery, business models, and incentives? What new data streams can unlock significant value, enhance patient safety, make providers’ lives easier, and make care more continuous? This year saw progress and announcements on all those fronts.

We owe a lot of this optimism to CVOTs, which have identified diabetes therapies that make a difference on meaningful hard outcomes, not just A1c. This year, CANVAS showed cardioprotection for a second SGLT-2 inhibitor (Invokana) and DEVOTE revealed a significant hypoglycemia benefit with a newer basal insulin (Tresiba). These clinical findings are actually making their way into patient care: FDA added a CV indication to the Victoza label in August, J&J has filed for Invokana’s CV indication (decision expected late next year), and in 1Q18, FDA could add a hypoglycemia claim to the Tresiba label. This last move would be revolutionary, hopefully setting a precedent wherein FDA acknowledges hypoglycemia as a valuable beyond-A1c outcome.

Powering the future of beyond-A1c outcomes will be continuous glucose data, which also had a pivotal year. This once-nascent technology has matured to a point of national reimbursement in several countries (Australia, France, Germany, UK, Japan), avoiding fingersticks entirely (FreeStyle Libre), reaching people of a wider variety of ages and therapies (G5 Medicare, MDI users), improving very costly outcomes (CONCEPTT, Belgium data), and finding its way onto major smartphones and now smartwatches (iOS, Android, Apple Watch, Fitbit). The data from connected, constantly-streaming CGM – as well as BGM, smart pens, and pumps – will power a revolution in how patients and healthcare providers manage diabetes, especially for those on intensive insulin therapy. There is now an avalanche of momentum to gather data where we haven’t had it – How much insulin are people taking, and how much should they be taking based on their patterns and habits? It’s worth noting how quickly these ideas have taken hold and how fast it has moved. In our 2012 Reflections piece, the top five tech stories of the year were the launch of Dexcom’s G4, Insulet’s second-gen Omnipod cleared by FDA, launch of Tandem’s original t:slim, FDA submission of Medtronic’s MiniMed 530G (Veo), and FDA’s final guidance on the artificial pancreas. Whoa – not a single device with connectivity, and not a single mention in that report of “Smart pens” or injection titration. This year, we saw the first hybrid closed loop system come on the US market (MiniMed 670G), 20+ companies working to build smart pens and/or insulin titration, and a whole new era of remote coaching driven by connected devices.

But with this optimism comes the same slew of challenges we’ve seen in our 2016, 2015, 2014, 2013, and 2012 Reflections pieces: Access/affordability remains a major roadblock in getting advanced diabetes drugs and devices into patient hands. This is something we must address alongside the scientific advances, the industry developments, and the regulatory progress – if patients cannot pay for advances therapies, they won’t make a bit of difference. With all the remarkable data showing cardioprotection, renal protection, glucose-lowering, and weight loss from GLP-1 agonists and SGLT-2 inhibitors, why do these advanced therapies account for only ~14% of second-line diabetes prescriptions in the US? CGM, for all its benefits, still reaches a small minority of people with type 1 diabetes at the very best US centers (~24% penetration in the T1D Exchange).

The field must improve reimbursement for these safer, more effective drugs and devices, and to this end we need a compelling evidence base for short- and long-term cost-savings. Industry has started to come forward with pricing transparency reports, and we’re noticing early signs of a pricing evolution: Novo Nordisk has priced Fiasp on par with NovoLog, for example, and Merck/Pfizer has priced Steglatro much lower than existing SGLT-2 options; notably, the combination SGLT-2/DPP-4 Steglujan has a list price that is actually on par with standalone Jardiance. As well, Abbott’s FreeStyle Libre came to the US with very strong pharmacy pricing, studies of CGM are showing benefits on very costly outcomes, and next-gen CGMs (e.g., Dexcom/Verily) have a design mandate to be less costly. Will this focus trickle down to patients? Some of it already has (see Merck/Pfizer’s Steglujan and Novo Nordisk’s Fiasp) and we believe there is progress on manufacturing leaders working with various stakehoders. Other questions include how can  enhance transparency around PBMs, and also ensure that formularies don’t restrict patient choice? These questions are now inescapable, and the answers will be critical for new products to make a meaningful dent in the ever-growing diabetes epidemic – now 425 million globally as of IDF’s 2017 Atlas.

Here’s to a 2018 that sees tremendous gains in access, alongside continued advances in therapies and technologies.

Diabetes Therapy

Themes

Major Outcomes Studies Point to CV and Renal Class Effects (CANVAS, EXSCEL, CVD-REAL, EASEL)

• CVOTs continued to gain traction in 2017, and with results from CANVAS and EXSCEL, we might call this the “year of class effects.” CANVAS reported full results at ADA 2017, showing a second SGLT-2 inhibitor to be cardioprotective. J&J’s Invokana (canagliflozin) reduced risk for three-point MACE (non-fatal MI, non-fatal stroke, and CV death) by 14% vs. placebo (HR=0.86, 95% CI: 0.75-0.97, p=0.0158 for superiority), which is exactly the relative risk reduction seen with Lilly/BI’s Jardiance (empagliflozin) in EMPA-REG OUTCOME (HR=0.86, 95% CI: 0.74-0.99, p=0.038 for superiority). Heart failure risk was similarly reduced by both canagliflozin and empagliflozin vs. placebo, by 33% in CANVAS and by 35% in EMPA-REG. Back in 2015, the positive data from Lilly/BI’s trial came as a happy surprise – but a surprise nonetheless – and CANVAS now offers compelling evidence for an SGLT-2 cardioprotective class effect and reassurance that EMPA-REG was more than a fluke finding. Real-world data also supports these notions. The AZ-sponsored CVD-REAL trial was first presented at ACC 2017 in March: All SGLT-2 products (Invokana, Jardiance, and Farxiga [dapagliflozin]) together were associated with a 46% risk reduction for heart failure hospitalization/all-cause mortality vs. other glucose-lowering drugs. The J&J-sponsored EASEL study just reported in November, showing 43% relative risk reduction with SGLT-2 inhibitors for the same composite endpoint of hospitalization for heart failure/all-cause death. Again, this consistency across trials suggests a class effect. We’ve heard very little contention from thought leaders that cardioprotection is indeed a class effect for SGLT-2 inhibitors. If anything, experts (like Dr. Silvio Inzucchi) are asking what’s next for this highly efficacious group of drugs, whether that’s demonstrating CV benefit in a lower-risk population or showing efficacy in heart failure prevention independent of diabetes. The DECLARE trial for Farxiga could answer some of these questions when it wraps up in 2H18 – the study features a large primary prevention cohort and includes heart failure in a co-primary endpoint. Based on a recent post-hoc analysis of CANVAS, it seems like the heart failure benefit may extend to lower-risk patients, but the MACE benefit might be exclusive to those at high-risk (drawing on this data, Dr. Itamar Raz has predicted that DECLARE will show superiority for the primary endpoint containing heart failure but only non-inferiority/safety for atherosclerotic events in three-point MACE). Lilly/BI and AZ have initiated dedicated studies of their SGLT-2 agents in people with chronic heart failure, with or without diabetes (EMPEROR HF and Dapa-HF). The fact that AZ launched this trial before seeing final data from DECLARE is another sign of confidence in CV class effects for SGLT-2 inhibitors.
  
  • The renal outcomes in CANVAS also closely aligned with the EMPA-REG data: Canagliflozin reduced risk for the composite endpoint of renal death, renal replacement therapy, or 40% reduction in eGFR by 40% vs. placebo (HR=0.60, 95% CI: 0.47-0.77), on par with empagliflozin’s 39% relative risk reduction for incident or worsening nephropathy (HR=0.61, 95% CI: 0.55-0.69, p<0.001). All three manufacturers in this market have either started or plan to start a dedicated kidney outcomes study: (i) CREDENCE for Invokana is expected to complete in June 2019, (ii) Dapa-CKD for Farxiga is expected to complete in November 2020, and (iii) Lilly/BI’s still-unnamed trial will commence in 2018. This is abundant evidence for a renal protective class effect, with consistent RCT results and confidence from every company that these products could show distinct benefit in chronic kidney disease (CKD). Microvascular outcomes have been equally important findings from most major diabetes CVOTs (even though the CV results make headlines). We’re noticing greater attention paid to the kidneys at diabetes conferences of-late, given the enormous burden of CKD within this patient population and the link between renal impairment and heightened CV risk, and thought leaders are excited about the promise of SGLT-2 inhibitors for renal risk reduction. Dr. David Fitchett put it best at ESC 2017 in Barcelona: He explained that SGLT-2 inhibitors are currently contraindicated in people with low eGFR due to a presumed loss of efficacy (rather than safety concerns), but argued that this particular patient population with comorbid diabetes/kidney disease may actually stand to benefit the most from SGLT-2 therapy.
• Class effects for GLP-1 agonists are more controversial, though clearly EXSCEL – a defining moment for diabetes this year – does more to support the possibility of a CV class effect than it does to refute it. Back in the day, there were warnings not to compare phase 3 trial results on the therapy side and we are surprised we don’t hear more warnings like this today. Several hypotheses have been proposed for why AZ’s Bydureon (exenatide once-weekly) just narrowly missed the statistical threshold for superiority in this CVOT (HR=0.91, 95% CI: 0.83-1.00), which made headlines with topline data in May and full results at EASD 2017 in Lisbon. There was a larger primary prevention cohort in EXSCEL (27%) compared to LEADER (19%) or SUSTAIN 6 (17%), and mean diabetes duration was also slightly shorter at 12 years vs. 13 (in LEADER) or 14 (in SUSTAIN 6). In other words, the EXSCEL population was “less sick” at baseline, which could have muted the effect size for CV risk reduction. After all, CVOTs intentionally enroll high-risk patients to drive up event rate, and exenatide did significantly reduce risk for three-point MACE in participants older than 65 (presumably, longer duration of diabetes and “more sick” overall). Adherence was lower and drop-out rate was higher in EXSCEL vs. LEADER or SUSTAIN 6, and we wonder whether results would have looked different had AZ given all patients the newly-approved Bydureon autoinjector instead of single-dose reconstitution kits. EXSCEL’s pragmatic design featured no run-in period to exclude individuals with low adherence, and allowed a wide range of concomitant medications including DPP-4 inhibitors and SGLT-2 inhibitors. When drop-in SGLT-2 inhibitors were excluded from the placebo arm, EXSCEL did meet statistical significance for CV efficacy. Moreover, component endpoints in EXSCEL (MI, stroke, CV death, all-cause death) trended in the right direction (Bydureon even showed a statistically significant hazard ratio for all-cause mortality at 0.86, 95% CI: 0.77-0.97, p=0.016) and largely aligned with findings from LEADER and SUSTAIN 6. Drs. Rury Holman and Angelyn Bethel both presented meta-analyses for the GLP-1 agonist class this year, which did find significant cardioprotection across the board: ~10% relative risk reduction for three-point MACE (HR=0.90, 95% CI: 0.82-0.99, p=0.033) when considering ELIXA (for Sanofi’s lixisenatide), LEADER, SUSTAIN 6, and EXSCEL all together, ~12% relative risk reduction when considering the three longer-acting agents (HR=0.88, 95% CI: 0.81-0.95, p=0.002). In <four months since EXSCEL reported full results, the trial has already contributed meaningfully toward an important discussion about one of the most advanced diabetes drug classes available today. We’re hearing support from thought leaders that EXSCEL was “probably positive.” We understand that molecular differences may contribute to some variation in efficacy for GLP-1 agonists, and certainly, these products are differentiated on dosing frequency and ease of injection. GLP-1 agonists/class effects were a major topic of conversation and debate within the diabetes community in 2017, and we foresee this continuing into 2018, especially with a topline readout from REWIND for Lilly’s Trulicity.
• All things considered, we want to underscore that CVOTs have delivered amazing news on cardioprotection in 2017 (the amputation signal in CANVAS is of course an exception). A few years ago, we weren’t sure we’d see a single positive CVOT, let alone four and maybe five. This is a huge win for patients, who can now take medicine not just to lower A1c, but to avoid heart attacks, strokes, and other CV complications that are costly from a financial perspective and that take a toll on quality of life. For the diabetes field, these are studies to celebrate, even though they only present more questions to investigate in 2018 and beyond.

A Call for Standardization of CVOT Design

• If the controversies over GLP-1 class effect (EXSCEL vs. LEADER) and SGLT-2 amputation signal (CANVAS vs. EMPA-REG OUTCOME) tell us anything, it’s that the diabetes field desperately needs more standardized outcomes research. EXSCEL’s trial design was markedly different from LEADER’s, and amputations were collected differently in CANVAS vs. EMPA-REG, which makes it hard to compare across studies or to compare agents in the same class. Given the massive investment that goes into CVOTs, including funding and resources from the sponsor as well as millions of patient-hours, we’d hope to gain as much insight as possible from these studies. This includes actionable evidence on the differential advantages/disadvantages of different therapies, which is a critical piece of the personalized medicine movement in diabetes as it could empower providers (and payers, and guideline-writers) to match the right drug to the right patient at the right time. We could even argue that differences between CANVAS and LEADER interfere with comparisons across classes, making it harder for HCPs to determine ideal patients for GLP-1 agonist vs. SGLT-2 inhibitor therapy. This leads to our call-to-action, one we’ve alluded to before, but here we make it official:
  
  • We see a massive need for one large CVOT that randomizes patients to all available drugs, within and across classes. We understand this would be a tremendous undertaking, requiring participation and investment from all major players in diabetes pharma, and/or a generous philanthropic grant. But we also point out that five years ago, we didn’t dare dream of an obesity CVOT because the possibility seemed so far-fetched, and Novo Nordisk is now initiating one for its second-gen GLP-1 agonist semaglutide (branded Ozempic for type 2 diabetes). So, maybe we’re dreaming big, but an outcomes study spanning therapy classes could yield a wealth of insights, moving the field leaps and bounds further along in personalized diabetes medicine. So long as CVOTs lack standardization, we’ll continue to push for this one large study. We also urge manufacturers to reach a consensus on key elements of trial design and data collection, and this very well might require leadership from FDA. In fact, if FDA is going to mandate these studies per its 2008 CVOT guidance, we certainly think the agency could implement standardization procedures so that these trials are more fair and more fruitful.

Two Diabetes Products Win CV Indications, Bringing Endocrinology and Cardiology Together

• With Novo Nordisk’s Victoza joining Lilly/BI’s Jardiance as the second diabetes drug and first GLP-1 agonist with a CV indication on its label, we’re inevitably seeing more interaction between the fields of endocrinology and cardiology. At Keystone 2017, Dr. Steven Nissen (who played a key role in establishing FDA’s 2008 CVOT guidance) pushed for more collaboration between the two: “We have to work together. More than half of my patients in coronary care have diabetes, and we need approaches to take and integrate the best knowledge from the cardio and diabetes worlds to take better care of our patients.” Dr. Nissen argued that it has taken the diabetes field too long to emphasize CV outcomes, and that it’s time to move away from a glucose-centric view of diabetes. This is a message we’ve heard from many diabetes thought leaders over the past year: Dr. Wendy Lane positioned CV risk mitigation at the “heart” of diabetes care in a talk at CMHC 2017. Dr. William Cefalu remarked at Keystone 2017 that Merck/Pfizer’s new SGLT-2 ertugliflozin (branded Steglatro) will have to show CV benefit in order to compete, within its class and also within diabetes therapy more generally, as patients/providers are starting to look for beyond-A1c benefits from their diabetes medicine. Of course, this message goes both ways, and we’ve also heard several calls-to-action this year urging cardiologists to step to the plate in treating diabetes. One of our favorite quotes came from Dr. Michael Farkouh at AHA, who issued a wake-up call to his cardiology colleagues: “Type 2 diabetes is a CV disease. We need to get into the game.” For the first time at ESC 2017 and then at AHA 2017, Novo Nordisk hosted a booth on the exhibit hall floor, promoting a diabetes drug (Victoza) at a cardiology meeting. We noticed we were substantially busier at cardiology conferences this year as opposed to last, simply because of more diabetes content on the agenda as well as broader and more detailed discussion, and we expect this to amplify in the years ahead. Ultimately, as more diabetes CVOTs read out – and as more positive results accumulate – we’re getting the sense that diabetologists are investing more in CV risk reduction strategies, while cardiologists are showing a greater interest in prescribing diabetes drugs. At ESC this year, Drs. Neil Poulter and John Deanfield recognized cardiologists’ hesitation to prescribe an injectable (i.e. a GLP-1 agonist) – so, there is still some resistance. The duo suggested, however, that cardiologists will have to embrace GLP-1 agonists as highly-effective treatments at the intersection of diabetes/CV disease that happen to be injectable, but that are still easier to dose and prescribe vs. insulin (the association of “injectable” with “insulin” may be at the root of this hesitation). 
• That we now have two diabetes drugs with CV indications is certainly something to celebrate. At the very end of 2016, Lilly/BI’s SGLT-2 inhibitor Jardiance became the first-ever diabetes product to gain a CV indication (for reducing CV death), based on results from EMPA-REG OUTCOME. Eight months later, Novo Nordisk’s GLP-1 agonist Victoza joined the elite club with an indication for reducing MACE (MI, stroke, and CV death), based on LEADER results. Most recently, J&J filed an sNDA requesting a CV indication for SGLT-2 inhibitor Invokana – quite quickly, we might add – based on CANVAS results presented at ADA in June. We can’t emphasize enough how valuable these label updates are in spreading awareness among busy HCPs about the CV benefits of diabetes drugs. As CV risk reduction becomes a central component of best practice diabetes management, the regulatory role will be crucial in getting compelling data into product information so that patients/providers have ready access. We also hope that FDA-approved CV indications sway payers toward better coverage of advanced therapies that will be cost-saving in the long run. And most importantly, it’s exciting that people with diabetes in 2017 can finally take medicine not just to lower average blood sugar, but to prevent costly, life-threatening complications. These CV indications are a win for patients, for HCPs, and for outcomes-based medicine.
  
  • While the 2017 ADA Standards of Care encouraged the prescription of Jardiance or Victoza (prior to an indication!) to patients with high CV risk, the 2018 guidelines have formally included these two drugs in the treatment algorithm for people with diabetes/ASCVD. Cardioprotective diabetes agents are now recommended as first-line therapy for people with ASCVD and A1c ≥9% (in combination with lifestyle + metformin), and as second-line for people with ASCVD and A1c <9%. The 2018 Standards of Care feature much more detail than past versions on the CV effects of all available drugs, divided into MACE effects and heart failure effects. Notably, the guidelines recognize the observed heart failure benefit of SGLT-2 inhibitors, which has yet to make it onto an FDA-approved drug label.
• Cardiologists also have reason to be excited for their patients without diabetes, as SGLT-2 inhibitors have shown strong signs of heart failure benefit. Cardiology meetings this year heavily featured SGLT-2 inhibitors, and we’ve gathered that the cardiology community is even more enthusiastic about prescribing SGLT-2s as opposed to GLP-1s (although both are slow in terms of becoming established as CV prevention therapies). At AHA this year, Dr. David Fitchett presented a post-hoc analysis of EMPA-REG OUTCOME suggesting that the mechanism for heart failure benefit may be completely independent of hyperglycemia, which means empagliflozin could be a treatment option for people with heart failure without diabetes. This hypothesis centers around volume status, as indicated by very early increases in hematocrit and hemoglobin, and you can read more about it below or in our AHA coverage. Early in 2017, both Lilly/BI and AZ launched dedicated investigations of their SGLT-2 inhibitors in heart failure. The EMPEROR program (n=7,000), announced in March, is investigating Jardiance (empagliflozin) in patients with chronic heart failure with either preserved or reduced ejection fraction, enrolling those with or without diabetes. Dapa-HF (n=4,500) was posted to ClinicalTrials.gov in February, and is investigating Farxiga (dapagliflozin) in patients with heart failure, with or without type 2 diabetes.

GLP-1 Agonists Continue to Shine with Remarkable Sales Growth, Quick-Moving Innovation; Bar for New Market Entries Will be High Moving Forward

• It was a fantastic year for the GLP-1 agonist class, in steep sales growth and impressive innovation. Pooled class revenue grew 35% YOY in 1Q17 to $1.4 billion, 29% YOY in 2Q17 to $1.6 billion, and 25% YOY in 3Q17 to $1.6 billion. GLP-1 agonists drove 55% of overall diabetes industry growth in 1Q17 and 57% in 2Q17, contributing a larger share of growth than any other drug or device class. In 2016, the GLP-1 market grew 25% YOY to nearly $5 billion, up from an already-high base of $3.9 billion in 2015. In 2017, we expect a similar magnitude of YOY growth so that the class tops $6 billion (sales through 3Q17 total $4.6 billion already). As expected, the market was dominated this year by Novo Nordisk’s Victoza (liraglutide) and Lilly’s Trulicity (dulaglutide). By our calculations, pooled market share by value for Victoza + Trulicity rose from 83% in 1Q17, to 86% in 2Q17, to 88% in 3Q17 – which points to intensifying competition within this class. GSK formally withdrew support for Tanzeum (albiglutide) in July as part of a larger restructuring of the company’s pharmaceutical business, although management also noted subpar efficacy shown in albiglutide clinical trials (“the data told us that this product wasn’t going to be commercially successful in our hands”). We have no doubt that further class growth for GLP-1 agonists is possible, considering the ample head room – these agents currently account for only 7% of second-line diabetes prescriptions in the US, which means far more patients could be reaping the benefits of profound glucose-lowering, weight loss, and anti-atherosclerotic effects. In other words, multiple products in this class can absolutely be commercially successful, but the bar for innovation appears to be rising, and new market entries will likely have to offer some advantage on efficacy, administration, side-effect profile, affordability, etc.
  
  • Perhaps the most relevant of innovations in 2017 was Novo Nordisk’s once-weekly semaglutide, now FDA-approved as Ozempic and slated for a 1Q18 US launch at parity pricing to existing once-weekly GLP-1s. Ozempic also received a CHMP endorsement recently, which bodes well for EU marketing authorization. With demonstrated CV benefit (albeit, in a small trial with wide confidence intervals) and superior efficacy to dulaglutide in a head-to-head study (~0.4% greater A1c drop with Ozempic vs. Trulicity, which Dr. John Buse called a “thrashing in the head-to-head trial business”), we expect Ozempic to soon be a blockbuster addition to the GLP-1 agonist class. While Novo Nordisk’s second-gen doesn’t offer reduced injection burden compared to current products – AZ’s Bydureon (exenatide) and Trulicity are once-weekly and, what’s more, will soon both be available in autoinjectors – it does offer meaningful innovation in A1c-lowering and weight loss. Semaglutide is considered to be a much more potent molecule than any other GLP-1 (and any other anti-hyperglycemic agent, for that matter). Some concern over retinopathy remains: This issue dominated the FDA Advisory Committee meeting convened prior to Ozempic approval, and Novo Nordisk will conduct long-term post-market studies of retinopathy as a condition for EMA-approval. To our understanding, most thought leaders think this is a manageable risk, if it exists at all (it could very well be due to “early worsening phenomenon,” wherein rapid A1c decline leads to a transient increase in retinopathy, as seen in the DCCT and with insulin therapy). Above all, we’ll be interested to see how patients perceive semaglutide going into 2018.
  • With the phase 3 PIONEER program for oral semaglutide fully underway, and with all 10 trials expected to read out in 2018, our anticipation has shifted somewhat from injectable to oral semaglutide (a sign of how quickly the field is innovating in GLP-1). At Novo Nordisk’s recent Capital Markets Day, management explained how oral semaglutide could target an earlier place in treatment algorithms vs. injectable GLP-1 agonists, competing foremost with DPP-4 inhibitors and SGLT-2 inhibitors. As such, we’re particularly looking forward to results from PIONEER 2 and 3, head-to-head trials vs. Jardiance (Lilly/BI’s SGLT-2) and Januvia (Merck’s DPP-4). Novo Nordisk has estimated filing for oral semaglutide in 2019 or 2020. Other oral GLP-1 agonist candidates include vTv’s TTP273, with phase 2 results presented at ADA 2017, and Oramed’s phase 1 ORMD-0901. Oral delivery of a GLP-1 agonist would be revolutionary for patients, and this seems like the next innovative stride for GLP-1 therapy, with Novo Nordisk at the helm.
  • AZ has improved Bydureon by developing an autoinjector, FDA-approved in October as Bydureon BCise. Previously, the only methods for administering Bydureon were single-dose reconstitution kits or a dual chamber pen, which still required a lengthy mixing/tapping process to create an aqueous solution prior to injection. Bydureon BCise is a patient-friendly device not unlike the IDEO-designed Trulicity autoinjector – the latter has been well-received by patients, and we imagine BCise could have a similar favorable for Bydureon. On the other hand, EXSCEL data is highly unlikely to support a CV indication for Bydureon, even though there are several viable explanations for why this was, in reality, a positive trial. This may be a challenge for AZ from a commercial standpoint as Novo Nordisk invests heavily in promoting Victoza’s CV indication.
  • After delaying phase 3 for a year, Sanofi initiated its first phase 3 trial of once-weekly GLP-1 agonist efpeglenatide in early December. Based on modeling presented during the company’s recent Analyst Day, Sanofi believes efpeglenatide could compete with Ozempic in terms of efficacy – a bold claim so early on, but phase 2 data are certainly impressive. Sanofi has also emphasized that efpeglenatide appears to have shorter-lasting GI side-effects than other GLP-1 agonists, which could improve patient adherence. Greater tolerability is another example of potential innovation in this class. The company has already developed a proprietary autoinjector for the drug, in line with the Trulicity pen and Bydureon BCise.
• Our excitement over Ozempic and the other innovations listed above is tempered only by the Complete Response Letter (CRL) for Intarcia’s ITCA 650. The FDA issued this CRL in September, after the company filed a 1.3 million-page NDA for the implantable exenatide mini pump in November 2016. ITCA 650 would provide continuous subcutaneous release of exenatide for three months (initiation stage) or six months (maintenance stage). All signs pointed to Intarcia’s product joining semaglutide as a market disruptor going into 2018. Adding to our enthusiasm, ITCA 650 wouldn’t have to deal with PBMs, which has numerous advantages. In a way, it might be that ITCA 650 is too innovative: those 1.3 million pages covered submission of the pharmaceutical agent (and its continuous subcutaneous release) as well as the delivery device, Intarcia’s Medici platform that includes the implantable pump, components conferring high temperature stability, and placement/removal kits. We don’t know what piece gave the FDA pause, but we expect Intarcia to resubmit as soon as possible, especially because no further pivotal trials will be required. This drug/device combo is about as “new” as it gets, and we wouldn’t be surprised if FDA called for an Advisory Committee after Intarcia re-files.
• There’s no room for inferiority in the GLP-1 agonist market: Sanofi management acknowledged this during the company’s recent Analyst Day. GSK’s decision to withdraw support from Tanzeum also supports this theme, which extends to CV benefit. We’re not ready to give up on the possibility of a cardioprotective class effect for GLP-1 agonists from a scientific/clinical perspective: EXSCEL was a pragmatic trial with low adherence, wide range of concomitant medications allowed, low-risk population at baseline, and other elements of study design that may have muted positive results. That said, the commercial implications of mixed CVOT results may be unavoidable. Victoza is now indicated for CV risk reduction based on the LEADER trial – this is on the label for HCPs to see readily, and the company is able to promote CV benefit. In contrast, it’s extremely unlikely that AZ will get a CV indication approved for the Bydureon label, which presents an uphill battle for the franchise to grow in volume/sales (we do anticipate some boost from the launch of Bydureon BCise, the new patient-friendly autoinjector). Novo Nordisk management has been adamant that “we cannot at this point say that cardioprotection is anything close to being a class effect of the GLP-1 agonists.” To this end, the company’s CSO Dr. Mads Thomsen presented a skeptical outlook on REWIND results for Trulicity (the CVOT is expected to complete in July 2018 and topline results are expected by end of 2018). He emphasized that the study population in REWIND has a relatively low starting A1c (~7.3%) and a relatively short duration of diabetes (~10 years), not to mention only 31% of participants have established CV disease at baseline, which could lead to neutral rather than positive findings. Lilly management was actually very optimistic about the chance for positive results during the company’s 2018 financial guidance call. Only time will tell, and we eagerly await REWIND data. In the meantime, we’ll continue to gauge how diverging CVOT results are translating to volume/sales for these different GLP-1 products. The best case scenario would be increased prescription of all GLP-1 agonists as these agents show cardioprotection, because again, the class still comprises only a small percentage of the overall market for diabetes drugs.

Thought Leaders Advocate for More Emphasis on Heart Failure in Diabetes; SGLT-2 Inhibitors as a Promising Treatment Option + Potential for Dedicated Heart Failure Indication

• The buzz around SGLT-2 inhibitors for heart failure grew more and more apparent throughout 2017, as AZ initiated Dapa-HF in February, Lilly/BI launched the EMPEROR HF program in March, and thought leaders started highlighting the burden of heart failure for people with diabetes – to the point where this CV outcome can no longer be ignored. A JAMA viewpoint published in September by Drs. Subodh Verma, John McMurray, and David Cherney called out this class as a possible “sweet spot” for heart failure, and Dr. Verma picked up on this theme during multiple talks at AHA 2017. Cardiologists have expressed excitement about this diabetes drug class since EMPA-REG results (one of our favorite quotes from an ESC corporate symposium: “this is a CV drug with a glucose-lowering effect”), and this interest has been amplified by the new investigations of Jardiance and Farxiga in chronic heart failure (both programs enroll participants with and without diabetes). It also helps that researchers have identified a likely mechanism for heart failure benefit that is completely independent of hyperglycemia or A1c. Dr. David Fitchett presented this intriguing hypothesis at AHA, showing how volume status (as indicated by very early increases in hematocrit and hemoglobin) was the primary driver of empagliflozin’s heart failure benefit in EMPA-REG OUTCOME: In a univariate analysis, hematocrit was associated with a 51% mediation and hemoglobin was associated with a 54% mediation, while glycemic metrics showed no influence on heart failure outcomes (-9% mediation for A1c, -2% for fasting plasma glucose). We found it noteworthy how MUCH attention SGLT-2 inhibitors/heart failure received at AHA and other cardiology conferences this year.
• While the prospect of a dedicated heart failure indication for SGLT-2 inhibitors is definitely tantalizing, we can’t forget about the unmet need in addressing heart failure within diabetes care. As Dr. Verma outlined at IDF, diastolic dysfunction appears earlier in the course of diabetes progression than does atherosclerosis, and yet the field has biased attention toward atherosclerotic CV events like stroke and MI. He explained that most CVOT participants, even those comprising a primary prevention cohort, probably have subclinical heart failure, which might explain why J&J’s Invokana (canagliflozin) showed consistent heart failure benefit across high-risk and low-risk participants in CANVAS, even though MACE benefit was not significant for the primary prevention subgroup. Most importantly, Dr. Verma argued that we should place as much emphasis on heart failure as we’re starting to do with ischemic MACE events, based on the prevalence, prognosis, and cost burden of these complications. Several other thought leaders have echoed this view (see our coverage of Dr. Jens Øllgaard at EASD, and of Drs. Bertram Pitt and John McMurray at AHA), regretting that heart failure was left out of FDA’s 2008 CVOT guidance (an unfortunate oversight). The good news is that as heart failure management works its way into best practice diabetes care, SGLT-2 inhibitors are rising to the top as a promising treatment option, and some experts are even suggesting that these agents be used as first-line in people with established CV disease or prior heart failure. Moreover, to the delight of these opinion leaders, heart failure is now included as a primary endpoint in DECLARE for AZ’s Farxiga (expected to report in 2H18) and in SCORED for Sanofi/Lexicon’s SGLT-1/2 dual inhibitor sotagliflozin.
• SGLT-2 inhibitors also had a big year in general, with a second positive CVOT, a CV indication on the Jardiance label, and impressive class growth. The potential indications for these agents in heart failure and CKD could be another substantial boost to class sales in the years ahead.

CANVAS Introduces Ongoing Discussion Over SGLT-2s/Amputations; Only Invokana Sales Suffering So Far

• The bad news for SGLT-2 inhibitors this year was the lower limb amputation signal in CANVAS (HR=1.97, 95% CI: 1.41-2.75, p<0.001), although only J&J’s Invokana (canagliflozin) seems to be taking the hit. In May, the FDA issued a black box warning for lower-extremity amputations on all medicines in the canagliflozin family of products. The EMA has extended a similar warning to the entire class, although the US labels for Lilly/BI’s Jardiance (empagliflozin) and AZ’s Farxiga (dapagliflozin) are amputation warning-free. From a commercial standpoint, this might explain why US Invokana sales took a dive in 2017, while Jardiance revenue continued to climb as Lilly/BI rolled out the new CV indication, and Farxiga sales grew to keep pace with overall class growth (~20%). Full results from the integrated CANVAS program were presented at ADA in June, and as the year went on, we hoped that subsequent analyses would show the amputation risk to be explained by something other than the canagliflozin molecule. Dr. Bruce Neal presented post-hoc safety findings at EASD, and despite controlling for other known risk factors, canagliflozin still increased amputations nearly two-fold vs. placebo. That said, Dr. Neal did highlight other variables that also predicted amputations in CANVAS, including infection, gangrene, peripheral arterial disease, ulcers, acute limb ischemia, and neuropathy. This leads to a key takeaway from CANVAS – that amputations are preventable with diligent monitoring of the feet, and that real-world diabetes care needs stronger patient/provider education on this topic. Indeed, many thought leaders including Drs. Anne Peters and Daniel Drucker have suggested that amputation risk should be manageable in the real world. After all, base rate was very low (6.3/1,000 patient-years on canagliflozin vs. 3.4/1,000 patient-years on placebo). Dr. Andrew Boulton has emphasized that amputation is a soft endpoint, since it requires active decision-making from patient and provider. While it’s tempting to compare CANVAS and EMPA-REG OUTCOME, we note that amputations were collected differently in each CVOT; baseline characteristics of the study populations also differed meaningfully. We expect the field to continue grappling with the CANVAS amputation signal in 2018, and data from DECLARE on dapagliflozin (expected in 2H18) will also be telling – is this risk specific to canagliflozin, or is the concern relevant with any SGLT-2 agent? Merck/Pfizer’s Steglatro (ertugliflozin) showed an imbalance in lower limb amputations in the VERTIS clinical program, with more occurring for patients on ertugliflozin vs. placebo, but this was not statistically significant. FDA has included a waring on the Steglatro label, but the language clearly stipulates that this risk was seen with another agent in the class and that no causal link has been established between ertugliflozin and amputations. We’ll be curious to see how this plays out in market dynamics.  Some thought leaders have been more skeptical about Invokana since CANVAS reported – Dr. Jay Skyler shared at CMHC 2017 that he’s switched all his patients on canagliflozin to empagliflozin, because “why deal with this concern if you don’t have to?” Some patients may not have a choice in 2018, because CVS Health has excluded Jardiance in favor of Invokana.
  
  • Above all, we want to emphasize that amputation concern should not undermine the tremendous composite benefit of SGLT-2 inhibitors, including Invokana. These drugs are oral, easy to prescribe and take. They offer profound glucose-lowering, weight loss, and blood pressure reductions (Dr. Peters even told us, anecdotally, that Invokana is more effective in reducing A1c and body weight vs. Jardiance in her clinical experience). They are cardioprotective and renal protective – which is huge for patients, HCPs, and payers – and the safety profile is overall, very manageable.

DEVOTE Leads to Hypoglycemia Claim for Basal Insulin Tresiba – Precedent-Setting in the Regulatory World, a Key Contribution to the Beyond-A1c Movement

• DEVOTE was the other CVOT featured at ADA 2017, and it was a gift that kept on giving, with two post-hoc analyses presented at EASD alongside a new hypoglycemia risk score app. Novo Nordisk’s Tresiba (insulin degludec) showed non-inferior CV effects vs. Sanofi’s Lantus (insulin glargine), but the real headline was a significant hypoglycemia benefit. The next-generation basal was associated with a 40% risk reduction vs. Lantus for severe hypoglycemia (HR=0.60, 95% CI: 0.48-0.76, p<0.001) and with a 53% risk reduction for severe hypoglycemia overnight (HR=0.47, 95% CI: 0.31-0.73, p<0.001). This is now included on the EU product label – EMA approved this update in 3Q17, making Tresiba the first diabetes drug with a hypoglycemia benefit displayed clearly in product information. An FDA decision is expected in 1Q18, and would surely set a precedent for outcomes beyond A1c, with regulatory recognition of hypoglycemia as a valuable clinical endpoint in diabetes care. Drs. Bernard Zinman and Thomas Pieber presented DEVOTE 2 and DEVOTE 3 at EASD 2017, further emphasizing the importance of glycemic outcomes beyond A1c. In DEVOTE 2, investigators found day-to-day glycemic variability to be significantly associated with all-cause mortality (HR=1.58, 95% CI: 1.23-2.03, p=0.0004). And in DEVOTE 3, an episode of severe hypoglycemia significantly increased risk for CV death and all-cause death via a temporal relationship (greatest in the first 15 days, declining as more time passes). If the high cost of a hypoglycemia hospitalization and the adverse impact on patient quality of life weren’t compelling enough, the DEVOTE dataset has revealed a tight link between hypoglycemia and hard outcomes (CV death, all-cause death). This was a very important study in 2017, a milestone for diabetes care and another sign that the bar is rising for diabetes therapies – merely lowering A1c is no longer good enough, and advanced molecules will have to show advantages on outcomes that matter to patients, like hypoglycemia, heart attacks, and death.
  
  • Dr. John Buse discussed development of the DEVOTE hypoglycemia risk score and accompanying app at EASD. He reviewed rigorous analyses to show that the risk score is indeed accurate, predicting risk for severe hypoglycemia or a MACE event based on input variables such as age, A1c, duration of diabetes, gender, and insulin status (naïve, bolus, or basal-bolus). Dr. Buse suggested one potential application of the app in facilitating patient/provider conversations when it comes to insulin initiation – people with type 2 diabetes are often resistant to starting insulin treatment, but he showed how adding bolus insulin on the app only modestly changes hypoglycemia risk for certain other pre-set variables. All in all, though the purpose of this app is not definitely decided, we thought it was a brilliant tool, and we’re excited to see where Novo Nordisk goes with this.

Basaglar Makes Waves as First-to-Market Biosimilar Insulin, Paves the Way for Upcoming Biosimilars; Sanofi’s Admelog Approved as First-to-Market Biosimilar Mealtime Insulin

• As the first-to-market biosimilar insulin, Lilly/BI’s Basaglar (biosimilar glargine) was a significant news item in 2017. Lilly’s reported sales from 1Q17, 2Q17, and 3Q17 totaled $279 million, including $197 million in US revenue (these are underestimates of total franchise revenue, because BI’s portion is not reported to the public). The product showed an encouraging sales trajectory following late 2016 US launch, driving 22% of growth for Lilly Diabetes in 3Q17 (second only to GLP-1 agonist Trulicity) and capturing 6% of pooled class sales (pretty impressive for a drug so early in its launch cycle). Basaglar certainly benefited from its preferred status over Sanofi’s Lantus (the original insulin glargine) on the CVS Health and UnitedHealthcare formularies, plus equal footing to Lantus on the Express Scripts formulary. Indeed, the story of “Lantus Formulary Exclusions” was a defining chapter for the insulin market this past year, explaining Basaglar’s strength and Lantus’ steep decline. This story will likely continue into 2018, as Basaglar remains a preferred drug. Moreover, as Novo Nordisk CEO Mr. Lars Jørgensen pointed out on the company’s 3Q17 earnings call, the competitive threat to other basal insulins will only be amplified next year as HCPs become more familiar with the concept of a biosimilar and with the equivalent safety/efficacy of Basaglar, specifically. Lilly management has cited favorable reimbursement status, high patient switching, and growing share of new-to-brand prescriptions (NBRx) as the main reasons underlying Basaglar’s strong financial performance to-date, and these are all factors that will continue past the New Yea. Against the backdrop of a basal insulin class that is struggling overall (pooled sales fell 4% YOY and 2% sequentially in 3Q17, and this was the third consecutive quarter of whole class decline), Basaglar is one of the only bright spots in this market, alongside next-generation products Tresiba (Novo Nordisk’s insulin degludec) and Toujeo (Sanofi’s insulin glargine U300). Interestingly, biosimilars and next-gen insulins are very different offerings: Basaglar is no more effective than the insulin that came before it, but is priced at a discount – which highlights the need for more affordable insulin – while Tresiba and Toujeo are substantially better than older basal options – highlighting the need for safer, more effective insulin (Tresiba even comes with a significant hypoglycemia benefit vs. Lantus, which is on the EU label and will hopefully soon be added to the US label, pending an FDA decision in 1Q18). Although we’ve heard some criticism that Basaglar is not as cost-saving for patients as initially promised, we note that ≥two generics are often needed within a marketplace to meaningfully drive down cost for patients, and to this end, we look forward to market entries from Merck’s Lusduna Nexvue and Mylan/Biocon’s candidate (both biosimilar formulations of insulin glargine).
• Basaglar is carving a niche for biosimilars within diabetes care, which could help the two biosimilar basals on the horizon as well as Sanofi’s first-to-market biosimilar rapid-acting insulin Admelog (biosimilar lispro). Merck’s Lusduna Nexvue (formerly known as MK-1293) was tentatively approved by the FDA in July, which means the candidate has fulfilled all regulatory requirements but cannot be launched in US pharmacies until all patent disputes are resolved. Mylan surreptitiously filed an NDA for its Biocon-partnered biosimilar glargine this Fall. Sanofi has issued patent infringement lawsuits against both Merck and Mylan/Biocon, so it could be ~two years or more until these second and third biosimilar basal insulins reach patient hands – but in our view, they are somewhat inevitable market entries. Sanofi was also engaged in a patent lawsuit against Lilly/BI, which was eventually settled with a royalty deal and an agreement to delay US launch until December 15, 2016, but now Basaglar has arrived and it’s making waves. To be sure, biosimilars have to demonstrate safety/efficacy that not only meets non-inferiority standards in clinical trials, but that is truly convincing to real-world HCPs (we anticipate that Mylan/Biocon may face an uphill battle because they lack history and expertise as insulin manufacturers, or even as diabetes manufacturers). That said, Basaglar appears to be cultivating familiarity with the concept of biosimilars, which will help smooth the path for Lusduna Nexvue, for Mylan/Biocon’s candidate, and even for Sanofi’s Admelog (a biosimilar version of Lilly’s rapid-acting Humalog), which was just FDA-approved this month. EMA approval came in July, and the product’s EU brand name is Insulin lispro Sanofi. We’re looking at a diabetes market with four biosimilar insulins in the not-too-far future.
• In January, FDA released highly-anticipated draft guidance on the interchangeability of biosimilars. Notably, neither Basaglar nor Lusduna Nexvue is approved with an interchangeability designation, which would allow the pharmacist to switch patients from the originator product to the biosimilar without consulting the prescriber (for that matter, FDA considers these insulins to be “follow-on biologics” rather than true biosimilars). Lilly management expressed no concerns about future interchangeability of Admelog for Humalog during the company’s recent 2018 financial guidance call. On the other hand, Mylan/Biocon are seeking an interchangeability claim for their biosimilar insulin glargine, which according to FDA’s draft guidance requires studies with at least two back-and-forth switches between the originator product (Lantus) and the biosimilar to rule out the possibility of adverse events from abrupt switching. FDA also stipulates that the biosimilar has to be dispensed in a very similar device as the originator drug, so as not to confuse patients or caregivers. Interestingly, Sanofi management suggested on the company’s 3Q17 earnings call that biosimilar insulins could be interchangeable by 2020. We imagine that conversation around interchangeability will increase as FDA evaluates Mylan/Biocon’s candidate, and if a designation is granted, we’ll be curious to see what impact this has on commercial sales and on patient experience. Of course, this is extremely early speculation, since Mylan/Biocon will likely face a long road ahead in getting their product approved (with or without an interchangeability claim), settling the patent lawsuit with Sanofi (we wonder if similarity of device means more chance of patent infringement), launching the product, and stimulating real-world uptake.

Fast-Acting Insulins Get Even Faster: Fiasp Approved, Afrezza Receives Ultra-Rapid-Acting Label Update, Exciting Data on Ultra-Rapid Candidates from Lilly, Adocia

• A new category of rapid-acting insulins emerged in 2017, separating fast from faster. News of these ultra-rapid-acting offerings came in the same week: FDA approved Novo Nordisk’s next-gen Fiasp (faster-acting insulin aspart) and granted a label update to MannKind’s inhaled insulin Afrezza to reflect its faster PK/PD profile. Fiasp is now slated for a 1Q18 US launch, and will be priced on par (!) to NovoLog (insulin aspart), as it is in the UK and other geographies where it’s already available. Some have argued that Fiasp showed only incremental improvement over NovoLog in the phase 3 Onset program (slightly better A1c and postprandial glucose excursions, no increase – but also no decrease – in hypoglycemia), but we continue to believe that what we have right now for mealtime insulin is not good enough, so any improvement – especially with parity pricing – is a win for patients. Moreover, we wonder if real-world data will show a difference in hypoglycemia, since an ultra-rapid-acting therapy will presumably reduce uncertainty around meals, and faster-offset means a shorter tail, or less residual insulin in the bloodstream. Afrezza has faced numerous commercial obstacles to-date, including termination of the Sanofi partnership in January 2016 and relaunch of MannKind-branded Afrezza in August 2016, but an ultra-rapid-acting label claim that differentiates the product from Humalog and NovoLog could serve as an important boost to volume and sales. Indeed, 3Q17 was a relatively strong quarter for MannKind financially: Afrezza revenue grew 28% sequentially to $2 million, though this is still a tiny fraction of the ~$1.6 billion/quarter rapid-acting insulin market, while new-to-brand prescription volume (NBRx) rose 37% sequentially and total prescriptions (TRx) rose 29% sequentially. It will of course take some time for the company to spread awareness of Afrezza’s newfound classification as an ultra-rapid-acting insulin, so 2018 sales/volume numbers will be more telling.
• Faster speed has been a missing ingredient when it comes to mealtime insulin, and we thus pause to celebrate these milestones, but not for long, because there’s still substantial room for improvement in prandial therapy. Which brings us to the late-stage pipeline:
  
  • Lilly presented compelling phase 2 data on its ultra-rapid-acting insulin candidate at ADA this year, and subsequently launched the phase 3 PRONTO program in type 1 and type 2 diabetes. This internally-developed candidate is a modified formulation of insulin lispro, with citrate and trepostinil added to the molecule to accelerate absorption. With a positive phase 3 readout, this agent would certainly be part of the new ultra-rapid-acting category, though we should keep in mind that PRONTO-T1D and PRONTO-T2D aren’t expected to complete until September and February 2019, respectively.
  • Adocia recently announced head-to-head results showing faster-offset with phase 3-ready BioChaperone Lispro vs. Fiasp. This was a topline release, and we look forward to full data at a major scientific meeting next year. From the data we’ve seen so far, it seems like faster-offset may be more of a differentiating factor for insulins in this emerging category rather than faster-onset – the long tail of post-meal insulin is a major cause of hypoglycemia, and shortening this could meaningfully improve outcomes.
  • Under the radar, Sanofi also launched a phase 3 study of its ultra-rapid-acting insulin candidate in August 2017. Management hasn’t commented on this trial explicitly (besides posting to ClinicalTrials.gov and updating the company’s pipeline page), but it’s nonetheless a player to watch on the rapid-acting insulin competitive landscape.

Strategic Shifts for Novo Nordisk, Lilly, and Sanofi Away from Insulin Toward Advanced Therapy Classes + Diabetes-Adjacent Indications

• It became increasingly apparent throughout 2017 that the Big Three insulin companies – Novo Nordisk, Lilly, and Sanofi – are in transition. We can’t reasonably call them “insulin” companies anymore. It was a tough year for Sanofi’s flagship product Lantus (insulin glargine) due to biosimilar competition and formulary exclusions in favor of Lilly/BI’s Basaglar (biosimilar insulin glargine), and this drove down sales for the company’s overall diabetes portfolio. Sanofi recently updated guidance to 6%-8% annual loss for the diabetes business between 2015-2018, which is at the low end of previous guidance of 4%-8% annual loss. While Basaglar was a bright spot for Lilly this year, the company’s rapid-acting insulin Humalog (insulin lispro) didn’t prove to be quite the growth driver as forecasted in 2017 financial guidance. Novo Nordisk has been the most explicit about its broadening focus, which now extends well past insulin therapy, ever since management announced a new R&D strategy in 3Q16 (higher innovation threshold, expansion into diabetes-adjacent indications like obesity and NASH). More recently, the company’s CMO Dr. Mads Thomsen discussed commercial challenges for insulin in an interview with the Wall Street Journal: He pointed to intense, persistent pricing pressure surrounding insulin products, which reduces profit margins for the manufacturer (more money is going toward patient discounts or to PBMs in the form of rebates). Dr. Thomsen plainly stated that companies can no longer rely on rising insulin profits to fuel growth, which has necessitated a strategic shift for major insulin manufacturers. That said, Novo Nordisk, Lilly, and Sanofi have each responded with a slightly different strategy:
  
  • Novo Nordisk is investing heavily in its GLP-1 agonist business and is well-equipped to be the market leader for many years ahead. This was one of our main takeaways from the company’s Capital Markets Day at the end of November. Victoza (liraglutide) currently leads the class by value, capturing 54% of the $1.6 billion market in 3Q17. Lilly’s once-weekly Trulicity (dulaglutide) has been catching up in market share, and Victoza’s patent expiry is anticipated in 2022 or 2023, but Novo Nordisk has planned ahead: Once-weekly Ozempic (semaglutide) will be launched in the US in 1Q18, and marketing to HCPs has already begun. All 10 trials in the phase 3 PIONEER program for oral semaglutide will read out in 2018. If results are positive, the company could file an NDA as early as 2019 or 2020, which means a second semaglutide product would be available before generic liraglutide. Oral delivery is the next big frontier for GLP-1 therapy (we celebrate less frequent injections, but imagine eliminating this injection burden entirely?). Notably, Novo Nordisk is further along with GLP-1 than any other company: Lilly has an oral GLP-1 agonist listed as preclinical, while vTv has an oral GLP-1 candidate in phase 2, and Sanofi just initiated phase 3 for a once-weekly injectable GLP-1, efpeglenatide. In terms of expanding beyond insulin, GLP-1 is an extremely smart investment in our view. The class is growing rapidly, matching pooled sales for rapid-acting insulins in 2Q17 and 3Q17 (~$1.6 billion), and still there is vast head room, as only 7% of second-line diabetes prescriptions in the US go to a GLP-1 agonist. The investment is also great news for patients, as this therapy class offers profound A1c-lowering, weight loss, and possible cardio/renal protection.
  • Lilly is looking at a bright future for its SGLT-2 inhibitor business with Jardiance (empagliflozin). The product leads the class in sales, capturing 41% of the $935 million market in 3Q17, by our calculations. Jardiance was the first diabetes drug to have a CV indication added to its label (which was rolled out in 2017), following landmark positive results from EMPA-REG OUTCOME. Lilly is committed to ongoing clinical development for empagliflozin as well, and the agent is being investigated for type 1 diabetes (EASE-2 and EASE-3), for heart failure (the EMPEROR HF program was launched in March of this year, enrolling patients with and without diabetes), and for chronic kidney disease (an outcomes study will begin next year, also enrolling patients with and without diabetes). We recently learned that Lilly plans to file a triple combination (empagliflozin + DPP-4 inhibitor linagliptin + metformin XR) with the FDA in 2018, which could be another boost to the franchise. Like GLP-1, we see SGLT-2 as an extremely promising investment with ample room for continued commercial growth: The same Diabetes Care paper reported that only 7% of second-line diabetes prescriptions in the US go to an SGLT-2 inhibitor, and new diabetes-adjacent indications for heart failure and/or CKD would expand the market even more. The composite benefit of this class cannot be understated: CV and renal risk reduction, a glycemic-dependent nature (which means no hypoglycemia risk), meaningful weight loss, and ease of use with oral administration, no titration, and manageable side-effects (the amputation signal from CANVAS seems to have only affected Invokana sales thus far). We should note that Lilly is also a major player in the GLP-1 agonist market with Trulicity, and there is more than enough room for multiple GLP-1 products to be commercially successful. Still, it seems like the company’s overarching diabetes strategy might increasingly emphasize SGLT-2 in the years ahead, especially if/when Novo Nordisk comes out with a trifecta of once-daily injectable, once-weekly injectable, and oral GLP-1 (meanwhile, Lilly’s portfolio would feature only once-weekly injectable Trulicity).
  • Sanofi is intent on making basal insulin/GLP-1 fixed-ratio combination Soliqua a commercial success. Although there has been unfortunate resistance to uptake from real-world payers and HCPs, we applaud Sanofi’s commitment to this promising new therapy class, which is more than we can say for Novo Nordisk (the company has de-prioritized Xultophy in favor of component monotherapies Tresiba and Victoza). Basal insulin/GLP-1 combos offer greater efficacy in getting patients to target A1c without weight gain or hypoglycemia than any other diabetes drug in recent history. We’ve definitely been disappointed by sluggish sales for Soliqua and Xultophy since their early 2017 launch in the US, but we hold out hope that this class will become a more central piece of diabetes care in the near-future – Sanofi could emerge as the leader of this charge. The company also has a once-weekly insulin/efpeglenatide combination listed as preclinical within its diabetes pipeline, so innovation continues on the fixed-ratio front. During an Analyst Day earlier this month, Sanofi announced plans to expand into diabetes-adjacent indications with its GLP-1/glucagon dual agonist candidate – the agent will be investigated for obesity and NASH. By contrast, Lilly maintains a sharp focus on type 1 and type 2 diabetes and to date has not entered the obesity market though this could easily change: We thought the company might evaluate high-dose dulaglutide in people with obesity, but the phase 3 program scheduled to start next year will be in type 2 diabetes. We look forward to Sanofi’s programs in obesity and NASH. Both chronic conditions remain therapeutic areas of high unmet need, and a major pharmaceutical company like Sanofi getting involved (given its expertise in clinical as well as commercial development) is almost certainly good news.
• Despite this changing focus for Novo Nordisk, Lilly, and Sanofi, we should note that all three are still committed to innovation in insulin – there’s simply a higher bar for advancing an insulin therapy through clinical development and onto the commercial market. A good illustration of this is that all three companies continue to invest in faster mealtime insulin, even though rapid-acting insulin products are facing intense competition from GLP-1 agonists and SGLT-2 inhibitors, which address postprandial glucose excursions without extra hypoglycemia risk (and these companies also manufacture GLP-1 agonists and/or SGLT-2 inhibitors themselves). This is important, as bolus insulin is still an essential medicine for many people with diabetes, and is still not good enough in terms of dosing ease or speed. Novo Nordisk will launch Fiasp (faster-acting insulin aspart) in US pharmacies come 1Q18, while Lilly and Sanofi both have ultra-rapid-acting insulins in phase 3. Novo Nordisk management has expressed optimism for two phase 1 insulin candidates – a once-weekly basal and a liver-preferential prandial. Both could be truly disruptive: the former would greatly lower injection burden and could improve adherence, while the latter would more closely resemble physiologic post-meal insulin response (>50% of endogenous insulin is typically cleared by the liver). Additionally, we heard confirmation at IDF that Novo Nordisk is continuing to work on oral insulin to bring down cost of goods, despite the 3Q16 discontinuation. Next-gen basal insulin Tresiba is a definite bright spot for Novo Nordisk, and the hypoglycemia risk reduction seen in DEVOTE was an exciting headline in 2017. This data led to a hypoglycemia claim on the EU label for Tresiba, and a parallel FDA decision is expected in 1Q18. The company is not backing off from commercial investment in this next-gen. Sanofi is similarly committed to its next-gen Toujeo (insulin glargine U300) and has an impressive real-world evidence campaign ongoing. Both next-generation basal insulins are a substantial jump from what was available before – in our view, the advantage of Tresiba and Toujeo over earlier options is much more than marginal. Lilly also has a next-generation basal insulin in its preclinical pipeline, although as we understand it, the company’s main priority for this candidate is combination with dulaglutide (Trulicity) for a basal insulin/GLP-1 fixed-ratio offering. Sanofi also has a once-weekly insulin/efpeglenatide combination listed as preclinical within its diabetes pipeline, which is further confirmation of the company’s investment in this fixed-ratio drug class.

Despite Clinical Enthusiasm, Basal Insulin/GLP-1 Agonist Combos Underwhelm on the Market

• Hopes were high at the start of 2017 for Sanofi’s Soliqua (insulin glargine/lixisenatide) and Novo Nordisk’s Xultophy (insulin degludec/liraglutide), launched in the US in January and May. Twelve months later, we note that the commercial performance of this new class has not lived up to the clinical hype. In 3Q17, pooled sales of basal insulin/GLP-1 agonist fixed-ratio combinations totaled only $37 million, up just 12% sequentially from $33 million in 2Q17, which marked a 75% sequential increase from $19 million in 1Q17. This financial performance is incredibly underwhelming for products so early in their launch cycle, especially considering the remarkable clinical utility of these agents. The efficacy in clinical trials of Soliqua (the LixiLan program) and Xultophy (the DUAL program) surpasses what we’ve seen from any other diabetes drug in recent history. These products are particularly powerful in getting patients to target A1c <7% without hypoglycemia or weight gain, and Xultophy in particular has been associated with significant weight loss. Dr. John Buse has called Xultophy “the most effective anti-hyperglycemic agent on the planet” – talk about clinical enthusiasm, and the highest of high praise from diabetes thought leaders. What’s missing is reimbursement, for one, as well as uptake from real-world HCPs. We’ve gathered that providers are unfamiliar with the concept of a fixed-ratio combo, and that they’re worried about side-effects and attributing adverse events to one molecule or another (the irony, of course, is that both combos offer milder side-effect profiles vs. component monotherapies, including fewer GI symptoms and less hypoglycemia). It’s a travesty that these highly-effective, highly-anticipated new drugs are finally out there, but are reaching so few patients in-need. One of our most pressing questions for 2018 is how Sanofi and Novo Nordisk can stimulate greater uptake of their basal/GLP-1 combos, and make no mistake, we believe this is urgent. Moreover, we have said for some time that this class needs a pithier name than “basal insulin/GLP-1 agonist fixed-ratio combinations,” and we’re taking suggestions. For that matter, the brand names – Soliqua and Xultophy – are so obscure, they seem to evoke mortal combatants from the Amazon, or maybe doomed lovers from Greek mythology.
• The two companies have adopted distinct approaches to commercializing their basal insulin/GLP-1 products. Soliqua is clearly a strategic priority for Sanofi, and the company has shown commitment to improving reimbursement and educating more real-world HCPs about the benefits of fixed-ratio injection. Management has previously identified clinical inertia/HCP reluctance as an enormous hurdle to Soliqua’s commercial success, also acknowledging the persistent issue of high drug costs. Sanofi shared in its 3Q17 update that Soliqua is covered for 65% of commercially-insured patients and 25% of Medicare Part D patients, though that does not speak to co-pays. The company’s struggling diabetes business could benefit immensely from Soliqua’s potential success, and the fixed-ratio agent could benefit from the fact that it’s priced on par with standalone GLP-1 agonists, while Xultophy is priced at a premium, to our latest knowledge. It’s not unreasonable that many diabetes care providers may have some hesitation in integrating this new class into their practice, if only because learning how to prescribe, dose, and treat patients with these agents will take some time. To this end, Sanofi has highlighted peer-to-peer and medical education efforts around Soliqua. In contrast, Novo Nordisk has made an explicit strategic decision to de-prioritize Xultophy for now, in favor of its component monotherapies Tresiba (insulin degludec) and Victoza (liraglutide), but we’re eager for the day this changes, because the company could contribute positively in generating greater awareness of this class and its clinical advantages. Xultophy combines a next-generation basal insulin (carrying lower hypoglycemia risk) with a GLP-1 agonist that has proven CV benefit, so our hunch is that this product could distinguish itself further in real-world data – but that of course requires company investment in marketing, education, and collecting RWE. We’d love to eventually see head-to-head data comparing these two agents, but above all, we want to emphasize that both Soliqua and Xultophy represent a major jump ahead of current monotherapy treatment paradigms, regardless of how the agents stack up against one another. In the coming months and years, we see plenty of room for both these advanced therapies to be commercially successful, and to help many people living with diabetes.

Renewed Promise for Adjunct Type 1 Diabetes Therapies

• It was a big year for a small company, with three full readouts from Lexicon’s phase 3 program for Sanofi-partnered SGLT-1/2 dual inhibitor sotagliflozin in type 1 diabetes. At ADA 2017, Dr. John Buse presented inTandem1 results, and a late-breaking poster shared inTandem2 results. Professor Melanie Davies presented primary inTandem3 results during a prominent symposium at EASD 2017. inTandem1 (n=793) and inTandem2 (n=782) both randomized participants to 200 mg sotagliflozin, 400 mg sotagliflozin, or placebo. After insulin optimization, sotagliflozin was associated with a ~0.35% placebo-adjusted A1c decline (p<0.001 for all comparisons). The higher dose of sotagliflozin was also associated with a statistically significant decrease in bolus insulin dose in inTandem1. inTandem3 (n=1,402) was designed to more closely reflect real-world clinical circumstances, with no insulin optimization period. The primary endpoint, defined as A1c <7% with no severe hypoglycemia or DKA, was met by 29% of patients in the 400 mg sotagliflozin arm vs. 15% of those in the placebo arm (p<0.001). Mean A1c reduction was 0.8% with sotagliflozin vs. 0.3% with placebo (p<0.001), and the drug candidate was associated with ~5 lbs weight loss vs. ~2 lbs weight gain for participants on placebo (p<0.001). We’ve been impressed overall by the efficacy data from this Lexicon-led program in type 1 diabetes – additional A1c-lowering, weight loss, and lower bolus insulin dose are all meaningful benefits for patients (mealtime insulin is still extremely challenging, posing burdens on quality of life, exposing individuals to hypoglycemia risk, and preventing people from optimal glycemic control). While there’s some lingering concern over DKA risk (more on this below), thought leaders have been largely optimistic about sotagliflozin’s potential as an adjunct type 1 treatment. Patient feedback has also been pretty positive, and to a certain extent, we see it as inevitable that people with type 1 diabetes will take SGLT-2s and SGLT-1/2s off-label (we very much hope FDA listens to patients in evaluating the sotagliflozin NDA, anticipated in 1H18). To be sure, adjunct therapy for type 1 diabetes remains an area of high unmet need, as emphasized by Drs. Julio Rosenstock and Chantal Mathieu at EASD.
  
  • In September, Lexicon also released compelling pooled CGM data from inTandem1 and inTandem2, reinforcing our view that increased time-in-range/decreased glycemic variability will be among the most appealing benefits of sotagliflozin for patients with type 1. These data showed that patients on 400 mg sotagliflozin spent an additional 12% of the day – or 2.8 hours – in range at the end of 24 weeks (p<0.001 vs. placebo). The significance of this should not be underestimated: Almost three hours of additional time-in-range is three hours when patients don’t have to worry about hyperglycemia or hypoglycemia, and this has enormous implications for quality of life.
  • We look forward to a Sanofi-led filing of sotagliflozin for type 1 diabetes with both the FDA and EMA in 1H18 (as early as 1Q18, according to Lexicon’s latest earnings call).
• EASD also featured positive data from DEPICT 1, a study of AZ’s SGLT-2 inhibitor Farxiga (dapagliflozin) in type 1 diabetes (n=833). From a baseline 8.5%, A1c declined 0.4% with dapagliflozin 5 mg vs. placebo and 0.5% with dapagliflozin 10 mg vs. placebo (both p<0.001). Most exciting, perhaps, was that CGM readings showed 52% (12.5 hours), 55% (13.2 hours), and 44% (10.6 hours) time-in-range for 5 mg, 10 mg, and placebo, respectively. This is two-to-three additional hours in range for people taking adjunct dapagliflozin, in line with the time-in-range increases achieved with sotagliflozin. There was no imbalance in confirmed DKA across DEPICT 1, with event rates of 1% (four patients), 2% (five patients), and 1% (three patients) in the 5 mg, 10 mg, and placebo groups, respectively. It is our understanding that a guideline in DEPICT 1 protocol, stipulating that investigators should reduce a patient’s total daily insulin dose by no more than 20%, contributed to the lower rates of DKA. This decision by AZ was based on a post-hoc of phase 2 data in type 1 diabetes. Dr. Chantal Mathieu elaborated on this rationale, underscoring that insulin has many physiological functions besides lowering blood glucose, and that minimizing total daily insulin dose shouldn’t be the primary goal of adjunct type 1 treatments. Our biggest takeaway from this discussion was that AZ’s and Lexicon’s trials should not be over-compared, due to differences in study protocol, population, etc. We are enthusiastic about both candidates and the prospect of the first indicated oral therapies for type 1 diabetes. Many patients could benefit from the decreased glycemic variability that these agents offer. AZ’s injectable Symlin (pramlintide) is currently the only adjunct to insulin for type 1 diabetes, and though it helps with postprandial glucose spikes, it has not been shown to reduce A1c and is poorly reimbursed.
  
  • We eagerly await full 52-week results from DEPICT 1, including the 28-week follow-up treatment period. AZ’s DEPICT 2 trial (n=815) will also complete in April 2018. One-year data from DEPICT 1 will be important in demonstrating durability of benefits, and DEPICT 2 could lend further evidence to support the positive effects of dapagliflozin in patients with type 1 diabetes. These trials will be important as we move toward instituting SGLT-2 and SGLT-1/2 inhibitors in type 1 diabetes care. We believe the well-established safety/efficacy of dapagliflozin in type 2 diabetes can only help this agent and this class take root in type 1 diabetes. Previously, a smaller (n=352), J&J-sponsored study of Invokana (canagliflozin) in patients with type 1 diabetes found modest improvements in A1c, but a striking ~15% increase in time-in-range. Lilly/BI are also investigating Jardiance (empagliflozin) for type 1 diabetes in EASE-2 and EASE-3, both of which wrapped-up in September 2017, though we have yet to see the data.
• Notably, we identified 2016 as a year of uncertainty (or “mixed feelings”) around type 2 diabetes drugs for type 1. In contrast, we’ve perceived more consensus from thought leaders in 2017 that SGLT inhibitors have applications in type 1 diabetes, and we’ve perceived a much more positive outlook overall that DKA risk will be manageable in the real world with proper patient education and risk minimization strategies. Capturing this perfectly is an editorial authored by Dr. John Petrie, published in the Lancet Diabetes & Endocrinology alongside the DEPICT 1 paper: “SGLT-2 Inhibitors in Type 1 Diabetes: Knocked Down, but Up Again?” On the conference circuit, we heard appreciably fewer talks this year criticizing off-label SGLT-2 use in type 1 diabetes, and an appreciable increase in commentary on how to use these agents optimally in people with type 1 (again, suggesting that this is sort of inevitable). We’d be remiss not to mention Dr. David Nathan’s editorial on inTandem3, but we found this to be overly-critical (letting DKA risk overshadow the tremendous benefits to A1c, body weight, time-in-range, quality of life, etc.) and plainly unfair (he claims that participants in this study were highly-educated around DKA, but Dr. John Buse said that this is false, that the education patients received was probably far from optimal). The dominant characterization that has emerged is that DKA risk should not be understated, but that it also shouldn’t deter patients who could garner tremendous benefit from SGLT inhibitors as adjunct to their insulin regimen, especially since the risk should be manageable with more diligent monitoring, proper patient selection, and strong education. Overall, the field seems less concerned than it did 12 months ago.
  
  • One of the most compelling exchanges we had this year was with Dr. John Buse, citing Lexicon VP Dr. Paul Strumph, on best practices for prescribing these agents to people with type 1. If a patient doesn’t monitor blood glucose, is less engaged in diabetes management, is petite, insulin-sensitive (total daily dose <40 units), avoids carbs, and/or exercises a lot, he/she may be high-risk. Pump therapy also appears to be a risk factor for DKA, and carb avoidance in general is problematic. Dr. Buse recommended daily ketonemia screening at least when starting SGLT inhibitors, and after any signs of fatigue, nausea, malaise, or anorexia. Conceptually, Dr. Buse explained DKA associated with SGLT inhibitors as accelerated starvation ketosis – in other words, feeding the body is one straightforward way to avoid DKA events. For treatment, he recommended fast-acting carbs with bolus insulin every hour or two until ketone levels subside, or correction doses targeting the lower end of the normal target range if the patient cannot eat. Additionally, Dr. Buse has suggested that these drugs be avoided when basal insulin rates will be lowered (i.e. during prolonged exercise, surgery, GI distress, fasting, dieting, and alcohol binges).

A Disappointing Year for Type 1 Cures and Prevention; Does this Signal a Shift of Investment Toward Closed Loop?

• Though 2017 brought good news for adjunct type 1 therapies, the news flow around type 1 cures and prevention was decidedly more discouraging. Results from numerous late-stage studies were disappointing at worst, mixed at best. Data from TrialNet’s Oral Insulin study was highly-anticipated at ADA 2017, and it missed its main primary endpoint of time from intervention to diagnosis. That said, researchers did notice an interesting signal. Study participants in the primary stratum had first phase insulin response above threshold, while those in the secondary stratum had first phase insulin response below threshold. This seemingly small difference in disease state had a major impact on the preventative efficacy of 7.5 mg oral insulin daily. Members of the primary stratum receiving oral insulin progressed to diabetes diagnosis at the same rate as those in the placebo group, hence the negative overall result. However, diabetes diagnosis in the secondary stratum was delayed by an average of 31 months (2.5 years!) with oral insulin vs. placebo. According to Stanford’s Dr. David Maahs (who provided independent commentary at ADA), TrialNet’s readout was not a home run, but a 31-month delay to diagnosis in a specific cohort of patients is surely an incremental victory – for a young child, this delay could mean >6,000 fingersticks not performed, 930 nights during which a parent can sleep soundly without severe hypoglycemia on the mind. There’s a lot of mechanistic research that needs to be done, but we are cautiously optimistic about the potential for oral insulin to delay type 1 diabetes in a specific cohort. To this end, the Helmsley Charitable Trust recently announced a $52 million grant that will support investigations of high-dose oral insulin in infants at-risk for type 1. We also note that Kamada’s alpha-1 antitrypsin (AAT) candidate for newly-diagnosed type 1 diabetes showed efficacy only in the subgroup age 12-18 (topline results were released in November 2017). It seems like prevention/cure candidates may have very distinct applications (e.g. secondary stratum in TrialNet’s study, specific age group in Kamada’s study). This isn’t necessarily bad news, but it does pose an additional obstacle to research, and we understand that this area of research is already particularly challenging, requiring very long study durations and high-cost materials. Dr. Jay Skyler spoke about these challenges to developing an intervention for type 1 diabetes at this year’s Keystone conference: He highlighted the complexity and heterogeneity of type 1 pathogenesis, suggesting that to truly salvage the beta cell, we’ll need an aggressive, multiple target, combination therapy approach.
• Also at ADA, the DiAPREV-IT trial dealt similarly mixed results. This investigation in children with multiple IAAs but no diabetes showed that Alum-GAD was safe but ineffective in delaying or preventing progression to diabetes. Given the storied history of GAD-induced immune tolerance, many had high hopes that this trial would deliver positive results, but the intervention demonstrated no tangible benefit. Regardless of stratum (two antibodies or three-six antibodies) or baseline glucose tolerance, the therapy failed to delay diabetes by any significant margin. That said, sample size was small (n=25 Alum-GAD; n=25 placebo), meaning the study was not properly powered for these sub-analyses. Larger trials could be a worthwhile investment, as they might unearth subgroups of responders, in line with the subgroup-specific benefits of oral insulin and AAT discussed above.
• The JDRF-sponsored REMOVAL trial of metformin in type 1 diabetes failed to meet its primary endpoint of improvement in mean carotid intima media thickness (cIMT), a common surrogate marker for CV risk. The secondary glycemic endpoint and insulin dose findings were also modest at best, leading investigators to conclude that the REMOVAL results do not support the assertion in current treatment guidelines that metformin has any clinically-meaningful benefits for people with type 1 diabetes. Commentary on REMOVAL has been mixed, ever since primary full results read out at ADA 2017. On the one hand, Dr. Julio Rosenstock called this “a negative study from all perspectives,” since metformin didn’t meet its primary endpoint, only demonstrated statistical significance for a tertiary CV endpoint (maximal cIMT), and lowered total daily insulin dose by only ~two units which is “absolutely nothing” in terms of clinical significance. On the other hand, Dr. Partha Kar focused on the ~3 lbs treatment difference in body weight favoring metformin over placebo. He explained that he uses metformin off-label in his type 1 patients who are overweight – REMOVAL supports his clinical practice, even if the trial didn’t find dramatic CV benefits. To be sure, we don’t want to understate the massive effort that went into this important study: This was the single longest (three years) and largest (n=428) RCT to examine a CV parameter of any kind in type 1 diabetes, a patient population that faces high residual CV risk, without many cardioprotective treatment options.
• It’s been a long time since we’ve seen a very positive biological study in type 1 diabetes. Our sense is that at least in the short term, the tide has turned toward more toward treatment-oriented approaches such as closed loop and, in the longer-term, glucose-responsive insulin among other approaches. That said, the competitive landscape for type 1 diabetes prevention/cures remains robust, with potentially-promising early-stage ventures. JDRF continues to invest heavily in prevention and cure research. In 2016, the organization allocated 7% of total grant funding ($197 million) toward in vivo cure research (108 grants) and 6% toward prevention (108 grants) vs. 4% toward artificial pancreas (52 grants) and 3% toward complications (55 grants).

Expanding Possibilities for Diabetes Prevention in Pharmacotherapy; New Data on Remission Approaches

• A number of studies presented in 2017 piqued our interest around possible large-scale interventions for prediabetes. In terms of pharmacotherapy, Bayer’s alpha-glucosidase inhibitor Glucobay (acarbose) showed efficacy in diabetes prevention in the ACE CVOT, presented in full at EASD 2017. The study enrolled 6,526 Chinese patients with coronary heart disease and impaired glucose tolerance. Despite failing to demonstrate superiority on any of its CV endpoints, the trial did find an 18% risk reduction for new-onset type 2 diabetes with acarbose vs. placebo over a median five years (HR=0.82, 95% CI: 0.71-0.94, p=0.005). Small but significant reductions in A1c, two-hour postprandial glucose, triglycerides, and body weight were also observed with acarbose. The diabetes prevention results greatly outweigh any disappointment on the CV front, in our view, as we now have evidence that a well-studied, generically available oral medication reduces progression to diabetes. Admittedly, the ACE population was rather homogeneous, but given the sheer size of the prediabetes epidemic in China, we’ll still take this as a win. Five hundred million adults in China have prediabetes – that’s almost half the population over age 18. Acarbose and other alpha-glucosidase inhibitors are most commonly prescribed in Asia, so despite the known GI side-effects, we see distinct potential for this agent to be used for prediabetes treatment in China. That said, we’d also love to see these results translated in more diverse populations across the world, and we’d like further analyses that parse out which patients stand to benefit the most from acarbose. Like diabetes, prediabetes should be perceived as a more serious condition in both the medical and public arenas: More often than not, prediabetes evolves to diabetes, so earlier intervention could have a meaningful impact on the size of the population with diabetes. Experts have also identified prediabetes itself as a CV risk marker (Drs. Jaakko Tuomilehto, Rury Holman, and John Buse, to name a few), introducing yet another reason that we should be treating hyperglycemia earlier on (especially with a cardioprotective therapy, or with CV risk reduction strategies).
  
  • Throughout the year, we continued to hear support for the use of metformin in diabetes prevention as well. While no drug is indicated for prediabetes – indeed, formal recognition of prediabetes as a disease would be helpful in combating actual diabetes – there is solid evidence from DPPOS that metformin lowers incidence. Dr. David Nathan presented a new subgroup analysis of DPPOS at ADA 2017: He highlighted the strong correlation between gestational diabetes and later-onset type 2, positioning these patients as ideal candidates for metformin. While intensive lifestyle intervention seemed to be more effective in lowering diabetes incidence after 2.8 years, Dr. Nathan emphasized that with longer-term follow-up, the gap closes between new cases of diabetes in the lifestyle group vs. the metformin group (58% risk reduction and 31% risk reduction after 2.8 years vs. 27% and 18% after 15 years). We find the collective evidence to be quite compelling, and we see metformin as an affordable drug with a significant preventative benefit. Moreover, metformin could be more appealing to those unable to commit to serious lifestyle changes. Metformin is the dominant first-line therapy in diabetes right now, but as thought leaders grow more critical of guidelines, we’re hearing more and more that its glycemic and weight loss effects could have a more meaningful impact in prediabetes (and perhaps we’ll move to another oral medicine like DPP-4s or SGLT-2s as first-line for diagnosed diabetes). Recognizing prediabetes as a disease and getting an indication on metformin’s label are critical steps, though FDA has shown reluctance to-date.
  • We’re hopeful that five or ten years down the road, it’ll be common practice to use safe/effective drugs in prediabetes care. Other diabetes and obesity therapies have also been investigated in prediabetes. Dr. Leigh Perreault reviewed several of them at ENDO 2017, also calling out insulin glargine (Sanofi’s Lantus), TZD pioglitazone, GLP-1 agonist liraglutide (Novo Nordisk’s Saxenda), orlistat, lorcaserin (Arena/Eisai’s Belviq), and phentermine/topiramate (Vivus’ Qysmia). What will it take to get regulators, guideline-writers, HCPs, patients, and payers all on board with treatment that aims to prevent type 2 diabetes? Unfortunately, we see this as a key question not for 2018, but for the next several years, because it will certainly take time for prediabetes drugs to gain traction.
• Diabetes remission also emerged as a feasible clinical outcome this year, with the presentation of impressive DiRECT results at IDF. While this study (n=298) was conducted in patients who already had diabetes, shorter duration of diabetes predicted success, and we think this benefit would only be magnified in those with prediabetes. At one year, 24% of the 149 participants randomized to an intensive weight loss program (withdrawal of anti-diabetic and anti-hypertensive drugs, total diet replacement with an ~800 calorie/day meal plan for three-five months, stepped food reintroduction for two-eight weeks, and structured support for long-term weight maintenance) achieved ≥15 kg (~33 lbs) weight loss. Moreover, 46% of the treatment arm (68 individuals) achieved diabetes remission vs. 4% of the control group (six individuals, p<0.0001). Average weight loss was ~30 lbs in the intervention group, with only ~4.4 lbs weight regain at 12 months. In those who lost ≥15 kg (~33 lbs), a striking 86% achieved diabetes remission. While we await longer-term results and cost-savings projections, we do note that even these short-term improvements should have a long-term impact on health, and we’re encouraged that this intervention was delivered by modestly-trained, usual care providers, without pharmacotherapy. As a caveat, many may find the program – at ~800 calories/day – difficult to follow, and indeed 32 individuals dropped out of the intensive arm. Other diabetes remission approaches are also under investigation, including Dr. Hertzel Gerstein’s method of aggressive short-term combination therapy (8-16 weeks of multiple drugs simultaneously, reduced calorie diet, and a personal exercise plan with weekly coaching). And then there’s Virta Health, which launched in March of this year with a goal to reverse type 2 diabetes in 100 million people by 2025. The Virta approach combines nutritional ketosis with digital health coaching, behavioral support, biometric feedback, and online peer support.

NASH Landscape Heats Up: An Expanding Late-Stage Pipeline, More Focus from the Major Diabetes Companies, New Conference Series

• Many heavy-hitters in pharma doubled down on their NASH investments in 2017, recognizing this as an area of high unmet need (with no FDA-approved therapies to-date). In April, Novartis/Allergan launched the phase 3 AURORA trial investigating co-administration of a CCR2/CCR5 inhibitor (cenicriviroc) with an FXR agonist. Gilead’s ASK-1 inhibitor selonsertib also moved into phase 3 early this year, with the STELLAR 3 and STELLAR 4 studies ongoing. These candidates join Genfit’s elafibranor and Intercept’s obeticholic acid in phase 3. Notably, large, traditionally diabetes-focused companies have also thrown their hats into the NASH ring as part of an expanded R&D strategy that encompasses diabetes-adjacent indications (and of course, NASH is a common comorbidity of type 2 diabetes). Sanofi announced during its recent Sustaining Innovation Analyst Day that it will move GLP-1/glucagon dual agonist SAR425899 into phase 2 development for NASH (and into phase 3 development for obesity). Novo Nordisk’s best-in-class GLP-1 agonist semaglutide is currently in phase 2 development for a NASH indication. Lilly has a GLP-1/glucagon dual agonist in phase 1 for NASH. Furthermore, Novartis has several FXR agonists in its NASH pipeline in addition to cenicriviroc, as well as Conatus-partnered emricasan (in phase 2) and SGLT-1/2 dual inhibitor LIK066 (also being developed for obesity). Novartis management has emphasized the company’s commitment to NASH on recent earnings calls, and we heard a similar message from AbbVie during its 1Q17 update (“NASH is on our strategic roadmap”). Gilead remains keen on NASH with earlier-stage candidates as well, including FXR agonist GS-9674 and ACC inhibitor GS-0976: Both are in phase 2, and both could be potentially combined with selonsertib. See our detailed competitive landscape for a full overview of the NASH pipeline.
• To parallel this uptick in investment, a new NASH-focused conference series was introduced in 2017, consisting of NASH Summit Boston and NASH Summit Europe. With a diverse array of participants hailing from the worlds of basic science, biotech, and big pharma, the discussion at these meetings converged on one especially difficult challenge in NASH – anticipating regulatory behavior. As the NASH field heats up and as more drug candidates advance into late-stage clinical development, companies are especially attentive to designing clinical trials with regulatory agencies’ positions in mind. This is a difficult proposition, however, because no real consensus among different agencies exists, and most advice is asset-specific without pertaining to NASH drug development as a whole. Indeed, drug developers and regulatory agencies alike are wrestling with questions such as what stage of fibrosis constitutes a “treatable” NASH patient, whether NASH drugs will be chronic treatments, how to best define the efficacy of a NASH candidate, and the list goes on. This makes clinical trial design an enormous challenge, and, as our understanding of NASH evolves, it is not uncommon for a trial’s primary endpoint to shift from “important to not-important” from a regulatory standpoint before its completion date. Once NASH therapies reach the market, we imagine this uncertainty will likely spill over to the payer landscape. Additional uncertainty comes from the fact that outcomes for NASH patients all play out over the course of decades – indeed, without hard data from long-term outcomes trials in NASH, it may be an uphill battle to convince payers that these likely pricey drugs would be cost-saving for the healthcare system in the long run. The complexity here is great, and we view this as an invitation for greater conversation and collaboration among basic science researchers, drug developers, regulators, payers, and of course, patients, in order to foster more communication and eventually standardization at every level – from identifying gold-standard mouse models for preclinical research, to agreeing on a set of NASH outcomes, to defining which classes of NASH patients these drugs will be geared toward. 

Most Highly-Read Reports

• FDA approves Novo Nordisk’s once-weekly semaglutide under brand name Ozempic – Label mentions retinopathy signal from SUSTAIN 6, acknowledges possibility of early worsening; Parity pricing to other once-weekly GLP-1s; US launch slated for 1Q18
• ADA 2017: CANVAS Full Results – Full CV, renal, and safety results with study design and discussant presentations as well; Our initial themes, questions, and takeaways from the study
• FDA Advisory Committee Meeting: Label Update for Novo Nordisk’s Victoza (liraglutide) – Panel votes 17-2 in favor of CV indication for Victoza; Enthusiasm for LEADER’s consistency across many CV endpoints; Concerns raised over generalizability of the LEADER results to the US population, and whether Victoza’s cardioprotective effects encompass primary prevention
• EASD 2017 EXSCEL First Look – Bydureon narrowly misses superiority in EXSCEL CVOT, HR=0.91, 95% CI: 0.83-1.00
• FDA approves Bydureon BCise, new autoinjector for AZ’s exenatide once-weekly – Patient-friendly device comes with hidden needle, eliminates lengthy mixing process; US launch in 1Q18; Exclusive interview with AZ’s Executive Director for Diabetes Ms. Sarah Walters and VP of US Medical Affairs Dr. Jim McDermott
• FDA grants full approval to Sanofi’s Admelog, first-to-market biosimilar mealtime insulin – No word on US launch timing or pricing; Admelog projected to drive growth for Sanofi Diabetes in 2019; First-to-market biosimilar mealtime insulin; EMA-approved as Insulin lispro Sanofi in July
• J&J submits sNDA for Invokana’s CV indication based on CANVAS results – Filing comes in time with end of Sept. guidance; Company seeking indication for reduced risk of all major CV events (Lilly/BI’s Jardiance indicated for CV death, specifically) extending to entire canagliflozin family of products
• FDA approves Novo Nordisk’s Fiasp (faster-acting aspart) the second time around – Parity pricing to NovoLog!! Launch slated for late 4Q17 or early 1Q18; First next-gen mealtime insulin approved for the US; Great potential to improve patient quality of life with faster onset/offset and less unpredictability around meals; Decision comes in time with 3Q17 expectation, follows CRL in Oct. 2016
• FDA adds warning for increased risk of leg and foot amputations to labels for J&J’s canagliflozin franchise, including Invokana, Invokamet, and Invokamet XR – Leg and foot amputations (most commonly of the toes) occurred about twice as often in patients treated with canagliflozin compared to those treated with placebo in CANVAS and CANVAS-R; Full results will be detailed at ADA 2017
• FDA approves Merck/Pfizer’s SGLT-2 ertugliflozin under brand name Steglatro; Two FDCs also approved – Steglujan (ertugliflozin/sitagliptin) + Segluromet (ertugliflozin/metformin); Pricing questions loom large as do CV questions; 2018 will bring four-product SGLT-2 market

What We Got Right/Expected

• Novo Nordisk’s Victoza is approved for CV risk reduction by FDA and EMA – what a win for patients! This followed a positive 17-2 vote from an FDA Advisory Committee convened in June, and was one of the best pieces of news we received all year in diabetes therapy.
• DEVOTE shows a hypoglycemia benefit for Novo Nordisk’s Tresiba over Sanofi’s Lantus. These results are in line with the SWITCH studies, although DEVOTE contributes even more compelling evidence because of its much larger sample size. As we continue to push for more recognition of outcomes beyond A1c, including hypoglycemia, we can’t emphasize enough the importance of DEVOTE (as well as subsequent post-hoc analyses that linked glycemic variability to all-cause and CV death).
• FDA approves Merck/Pfizer’s SGLT-2 inhibitor ertugliflozin. The decision came on time with 4Q17 expectations, and the product’s brand name is Steglatro. Major pivotal trial data on ertugliflozin read out at EASD 2016, with one-year results from several studies reporting at ADA 2017.
• Sanofi’s Lantus franchise struggled in 2017, which management has attributed, at least in part, to the CVS Health and UnitedHealthcare formulary exclusions. Meanwhile, Lilly/BI’s biosimilar Basaglar – preferred on these major US formularies – showed a strong sales trajectory following late 2016 US launch.
• New biosimilar basals inch toward the commercial market, but not without resistance from Sanofi. Merck’s Lusduna Nexvue received tentative FDA approval, and Mylan/Biocon’s biosimilar insulin glargine was also filed in the US, but Sanofi has issued patent infringement lawsuits against both. This follows the plot of Sanofi suing Lilly/BI for Basaglar, before eventually reaching a resolution.
• Merck expands its diabetes portfolio and is now poised to become an even bigger player in diabetes. In addition to the Januvia franchise, which leads the DPP-4 inhibitor market, Merck will launch the Pfizer-partnered SGLT-2 Steglatro (ertugliflozin) franchise in 2018. The company has tentative approval for its biosimilar glargine Lusduna Nexvue, with full FDA approval contingent on resolution of the patent lawsuit from Sanofi.
• Cardiologists become more active participants in diabetes care, prescribing SGLT-2 inhibitors and GLP-1 agonists to their patients in light of new CV indications for these drugs. Cardiology meetings in 2017 featured more diabetes content than ever before. There’s slightly more hesitation among CV specialists to prescribe injectable drugs, but thought leaders believe this will soon be overcome with both GLP-1 agonists and PCSK9 inhibitors.
• Advocacy continues for metformin as a prediabetes intervention, but there’s no movement from regulators.
• Amgen’s PCSK9 inhibitor Repatha receives a CV indication for FOURIER. CV benefit from a powerful lipid-lowering agent isn’t particularly surprising, but this shouldn’t undermine what it means to have a CV indication added to the label – a key milestone for the PCSK9 class in terms of generating real-world awareness, promoting uptake, and improving reimbursement.
• Investments increase for Dr. Zhen Gu’s glucose-responsive insulin patch. Dr. Gu’s group started the year strong, publishing positive preclinical results in ACS Nano. In June, the group received $5.8 million from MicroPort Scientific Corporation, which follows a $4.6 million grant from JDRF/Sanofi.

What We Got Wrong/Did Not Expect

• EXSCEL presents a complicated story, with a narrow miss for CV superiority. We’d be lying if we said we weren’t hoping for positive results on AZ’s Bydureon (exenatide once-weekly), to support a cardioprotective class effect for GLP-1 agonists, which is now a source of ongoing debate in the field. In any case, few predicted that the hazard ratio would miss significance for superiority by such a razor-thin margin, with an upper bound of the 95% confidence interval at 1.00.
• FDA adds black box warning for lower limb amputations to all Invokana franchise medicines. This follows a safety review initiated back in May 2016, and at the time, we didn’t fathom that a highly-significant amputation signal might emerge from CANVAS data (though we note that base rate was very low). We were also surprised by the EMA’s conservative decision to issue an amputation warning extending to the entire SGLT-2 class.
• SGLT-2 manufacturers launch large, dedicated outcomes trials in heart failure and CKD. Since EMPA-REG reported back in 2015, we’ve known this class holds potential to decrease heart failure hospitalizations and to prevent adverse renal outcomes – what we didn’t realize is just how compelling these data were to Lilly/BI (initiated EMPEROR HF, CKD trial to begin next year) and AZ (initiated Dapa-HF and Dapa-CKD).
• Sluggish uptake of basal insulin/GLP-1 fixed-ratio combos, despite tremendous clinical enthusiasm for these agents (Sanofi’s Soliqua and Novo Nordisk’s Xultophy) going into 2017. We consider this one of the biggest disappointments in diabetes therapy this year. Both products were FDA-approved on the same day in November 2016.
• Intarcia receives an FDA Complete Response Letter for ITCA 650. This was disappointing and unexpected news, announced in September. The company intends to resubmit ITCA 650 as soon as possible, though very limited details were shared on reason for CRL or necessary next steps.
• Pooled GLP-1 revenue matches pooled revenue for the rapid-acting insulin class in 2Q and 3Q17 ($1.6 billion). More broadly speaking, whole class growth for GLP-1s and SGLT-2s was even more impressive than we expected in 2017, while insulin sales fluctuated.
• Merck receives Complete Response Letter for inclusion of TECOS data on the Januvia label. This CRL was announced in April. Adding these CVOT results to product information could have helped assuage concerns over heart failure hospitalization, as TECOS found a resoundingly neutral hazard ratio of 1.00 for this endpoint.
• FDA extends heart failure warning to all DPP-4 products, not just Onglyza. Only the SAVOR-TIMI study for AZ’s Onglyza found a significant signal for heart failure hospitalization, and yet FDA took a conservative approach here in adding a warning to the labels for Januvia (Merck), Tradjenta (Lilly/BI), and Nesina (Takeda) as well. This decision has been harshly criticized by thought leaders.
• Lilly terminates its partnership with Adocia for BioChaperone Lispro. This came as sudden news in January 2017. Lilly is now focused on its internally-developed ultra-rapid-acting insulin (in phase 3), while Adocia continues to seek a new partner for BC Lispro (that said, management told us that a go-alone pathway is not off the table for the phase 3-ready candidate).
• JDRF funds clinical trials of Biocon’s oral Insulin Tregopil. While oral bolus insulin is a fascinating therapeutic proposition, we don’t have a lot of confidence in Biocon and we hesitate to get our hopes up given the technical challenges. That said, an endorsement from JDRF is certainly noteworthy.
• The Helmsley Charitable Trust announced one of its largest-ever grants ($52 million) to support an oral insulin trial for type 1 prevention. HCT has high hopes that earlier intervention (in infancy) with higher doses of oral insulin will effectively delay the progression of autoimmunity, even though TrialNet reported a negative oral insulin study at ADA.

Key Questions for 2018

• Will FDA approve a hypoglycemia claim for Novo Nordisk’s Tresiba? A decision is expected in 1Q18, based on the compilation of SWITCH and DEVOTE data, and our fingers are so very crossed for this one. The EMA approved a similar claim in 3Q17, making Tresiba the first (!) diabetes therapy with a hypoglycemia benefit displayed clearly on its product label. Given the enormous financial cost associated with hypoglycemia each year, not to mention the severe adverse impact on patient quality of life, we cannot emphasize this point enough: Once we have evidence-based knowledge that a drug can reduce hypoglycemia risk, this information must be disseminated swiftly to real-world patients and HCPs.
• Will commercial sales of Ozempic (semaglutide) live up to the clinical hype? Novo Nordisk seems well-prepared to invest in marketing and to follow-through on this tremendous opportunity, but we’ll have to wait-and-see how reimbursement shakes out. We don’t anticipate major concerns over retinopathy risk, but again, time will tell after Ozempic launches in US pharmacies.
• Do REWIND results support or refute notions of a CV class effect around GLP-1 agonists? What other insights will be gleaned from this CVOT for Lilly’s Trulicity? (It certainly seems like every CVOT published to-date has uncovered surprise findings, whether positive like renal protection, or negative like amputation risk.)
• How will Novo Nordisk design CVOTs for semaglutide in type 2 diabetes and obesity? We’ll have our eyes peeled for any innovation or standardization of CVOT design in the year ahead.
• Will Intarcia re-submit ITCA 650 to the FDA? The second time around, will this highly-innovative GLP-1 agonist be approved?
• Could GSK sell Tanzeum to another company? Perhaps, to a company like Walmart that could then offer a much more affordable GLP-1 agonist option?
• Will SGLT-2 inhibitors show CV benefit in primary prevention? Thought leaders are leaning toward a heart failure benefit that extends to lower-risk patients, even though MACE benefit may be limited to those with established CV disease or several risk factors. A post-hoc analysis of CANVAS supports this theory, and DECLARE data come 2H18 will be telling – AZ’s CVOT enrolls a large primary prevention cohort and includes heart failure hospitalization in a co-primary endpoint.
• Will J&J’s Invokana return to growth? Will FDA grant the SGLT-2 inhibitor a CV indication? Will further analyses of CANVAS show that the amputation signal can be attributed to something other than the canagliflozin molecule? Or has the FDA black box warning signed and sealed Invokana’s doom?
• How will the diabetes community respond to amputation risk surrounding SGLT-2s? The FDA has added an amputation warning to the Steglatro label (Merck/Pfizer’s ertugliflozin), but the language specifies that this signal was seen for another agent in the class (J&J’s canagliflozin). Neither Jardiance (Lilly/BI) nor Farxiga (AZ) has this warning on its US label, whereas Invokana has a black box warning. How will this play out in market dynamics? Will 2018 bring new evidence to show amputations as a class effect, or will these concerns slowly fade?
• Will Jardiance pull ahead as the market leader for the SGLT-2 class? Or will the CVS Health formulary exclusion be a substantial headwind (with Lilly/BI yielding prescriptions to J&J)?
• Will Lilly/BI report findings from EASE-2 and EASE-3 (Jardiance in type 1)? If so, how do the results compare to DEPICT data on AZ’s Farxiga and to inTandem data on Sanofi/Lexicon’s sotagliflozin? Will AZ and/or Lilly/BI pursue a type 1 diabetes indication for their SGLT-2 inhibitors? Both EASE-2 and EASE-3 completed in September, according to ClinicalTrials.gov.
• Will basal insulin/GLP-1 agonist combinations take hold in diabetes care? Can Sanofi meaningfully boost volume/sales for Soliqua? Will Novo Nordisk finally prioritize Xultophy within its diabetes portfolio?
• Could SGLT-2/DPP-4 fixed-dose combos finally take off? Merck/Pfizer’s Steglujan (ertugliflozin/sitagliptin) is set to launch in January, and we wonder whether this will be the first real commercial success within this fixed-dose class (as opposed to Lilly/BI’s Glyxambi or AZ’s Qtern).
• Is Fiasp fast enough for payers to cover it? The parity pricing to NovoLog bodes well for reimbursement prospects. We have heard (from Novo Nordisk, Sanofi) that formularies are more exclusive for rapid-acting insulin than any other diabetes drug category because payers view these products as largely interchangeable – could Fiasp change this mentality, with speed as a distinguishing factor?
• Will Afrezza gain commercial traction? The label update advancing the product into a category of ultra-rapid-acting mealtime insulin bodes well for MannKind, but the company will need to see a significant sales boost in 2018.
• What will prescription volume look like for sulfonylureas? Will these harmful agents (known for weight gain, hypoglycemia, beta cell burnout, and possible CV risk) still be the most common option for second-line therapy next year? How can we shift payers and HCPs toward safer alternatives, or at least, to the safest alternative within the SU class (which we’ve heard is glimepiride)?
• Provided ODYSSEY Outcomes is a positive study, will payers respond to having two PCSK9 inhibitors with demonstrated cardioprotection? Can Sanofi/Regeneron and Amgen work collaboratively to move the needle on reimbursement for this class?

What’s Coming in 2018?

The list below covers late-stage pipeline products with expected milestones in 2018. We acknowledge that this list may not be 100% complete, though we’ve tried to be as comprehensive as possible based on the most recent public updates. If you notice anything missing, please write us.

Insulin

• Novo Nordisk’s Tresiba (insulin degludec): An FDA decision on a hypoglycemia claim is expected in 1Q18. The EMA has already approved this label change, on the basis of SWITCH and DEVOTE data (degludec showed significantly lower hypoglycemia risk vs. glargine, or Sanofi’s Lantus).
• Novo Nordisk’s Fiasp (faster-acting insulin aspart): US launch – at parity pricing to NovoLog – is expected in 1Q18. We’ll also see full results from Adocia’s phase 2 head-to-head trial of BioChaperone Lispro vs. Fiasp (topline data suggested faster-offset with BioChaperone Lispro and comparable onset time between the two).
• Sanofi’s Admelog (biosimilar insulin lispro): This first-to-market biosimilar mealtime insulin will launch in the US sometime in 2018, although Sanofi management isn’t necessarily anticipating significant sales until 2019 (due to timing of the payer contracting cycle, and the need to build up strong reimbursement).

GLP-1 Agonists

• Novo Nordisk’s semaglutide (branded Ozempic): Launch in 1Q18, following FDA approval in early December 2017. We also anticipate full approval from EMA following CHMP’s positive opinion in mid-December. Novo Nordisk has agreed to post-marketing studies in pediatrics and investigating long-term retinopathy outcomes. We’ll also have our eyes peeled for details on the post-market CVOT in type 2 diabetes and the one in obesity!
• Lilly’s Trulicity (dulaglutide): The REWIND CVOT is on track to complete in July 2018, and management announced that topline data should be available for the public before year-end. Will results support or refute a cardioprotective class effect? Will the readout be good news or bad news for Lilly, as Trulicity faces an increasingly competitive GLP-1 agonist market?
• AZ’s Bydureon (exenatide once-weekly): The company’s autoinjector, Bydureon BCise, will launch in US pharmacies come 1Q18. An EMA decision is also expected in late 3Q18 or early 4Q18. BCise offers a much more convenient injection process vs. single-dose reconstitution kits or the dual chamber Bydureon pen.
• Novo Nordisk’s oral semaglutide: All 10 trials in the phase 3 PIONEER program will report in 2018.

SGLT-2 Inhibitors

• Merck/Pfizer’s ertugliflozin (branded Steglatro): Launch in January, following FDA approval this month. Steglujan (ertugliflozin/sitagliptin) will also be available in January, while Segluromet (ertugliflozin/metformin) will be available in February. Notably, Steglatro will be priced lower than existing SGLT-2 inhibitors ($8.94/day vs. ~$17/day). Steglujan will be priced at a ~25% discount to existing SGLT-2/DPP-4 fixed-dose combos ($17.54/day vs. ~$22/day).
• J&J’s Invokana (canagliflozin): FDA decision on CV indication (risk reduction for MI, stroke, and CV death) expected between August and October 2018.
• AZ’s Farxiga (dapagliflozin): Results from DECLARE expected in 2H18. This CVOT could have profound implications for CV class effect and the possibility of a heart failure benefit in primary prevention. Turning to a potential type 1 indication for this SGLT-2 inhibitor, we look forward to 52-week data from DEPICT 1, and DEPICT 2 is expected to complete in April 2018.
• Lilly/BI’s Jardiance (empagliflozin): A new outcomes study in chronic kidney disease (CKD) will begin next year, enrolling patients with and without diabetes. AZ also has Dapa-CKD ongoing, while J&J has high hopes for the CREDENCE trial (expected to complete June 2019).
• Sanofi/Lexicon’s sotagliflozin: The SGLT-1/2 dual inhibitor could be filed for a type 1 indication as early as 1Q18, and surely in 1H18. The type 2 diabetes program for this candidate will also forge ahead next year (with the first estimated completion in January 2019).

Glucagon

• Lilly’s nasal glucagon: The company plans to submit an NDA in 2018, which positions this product to be the first-to-market next-gen glucagon option for the treatment of hypoglycemia.
• Zealand’s dasiglucagon: A phase 3 pivotal trial is expected to complete in 2H18, when we’ll also get to see full results from a phase 3 immunogenicity trial of the candidate.
• Xeris’ G-Pen glucagon autoinjector: A phase 3 study wrapped up in September 2017, and we hope to see full results next year, though this is speculation as we haven’t heard any updates from the company.

DPP-4 Inhibitors

• Lilly/BI’s Tradjenta (linagliptin): We’ll see full results from the CARMELINA CVOT (linagliptin vs. placebo) next year, and we might also see topline data from the CAROLINA CVOT (linagliptin vs. SU glimepiride).

PCSK9 Inhibitors

• Sanofi/Regeneron’s Praluent (alirocumab): An announcement of topline results from ODYSSEY Outcomes is expected in 1Q18, while full results will likely be presented at a major scientific meeting in 1H18 (perhaps ACC). Sanofi plans to file for a CV indication in 3Q18, contingent on positive CVOT data.
• Amgen’s Repatha (evolocumab): The company will roll out Repatha’s new CV indication, FDA-approved in early December based on positive FOURIER results.

Competitive Landscapes

Peruse our collection of therapy-related competitive landscapes. On each page, you’ll find a table of all candidates in development for that category, to the best of our knowledge. We frequently update these competitive landscapes as timelines change and as new developments arise.

• Insulin
• Glucagon
• GLP-1/glucagon dual agonists
• GLP-1/GIP dual agonists
• Diabetes-related nephropathy
• Type 1 cures and prevention
• NAFLD/NASH
• Ongoing CVOT timeline

Diabetes Technology

Themes

CGM Momentum: Major Publications, Reimbursement Victories, International Growth, and Expansion to Broader Populations

• 2017 was a major year for CGM, particularly on the publication, reimbursement, international growth, and user base fronts. The table below summarizes some of the key milestones this year, with themes discussed in the bullets directly below.
• We saw several important CGM journal publications this year, including clear benefits of using CGM in MDI (DiaMonD and GOLD in JAMA), in pregnancy (CONCEPTT in The Lancet), and in type 2 diabetes (DiaMonD in Annals of Internal Medicine). Meanwhile, a poster from Belgium at EASD showed that 12 months of CGM use reduced days spent in the hospital for severe hypoglycemia/DKA, plus fewer days lost to work absenteeism. Real-world data is coming out too, reinforcing these RCT outcomes. Abbott showed real-world user data at ATTD in February (50,000+ users) and then again at DTM (237,000+ users) – consistently, those who scan their sensor have a lower A1c with less hypoglycemia. From professional associations, it was excellent to see the positive CGM mentions in the December issue of Diabetes Care, where Drs. Cefalu, Gerstein, and Riddle noted: “Periodically, a new idea, method, or tool leads to a turning point in the management of diabetes. We believe such a moment is now upon us, brought by development of reliable devices for continuous glucose monitoring (CGM).” ADA/EASD also published a statement on improving the clinical utility of CGM, sharing optimism about CGM’s “great progress,” but also honesty regarding the current barriers to adoption. and the technology did get a few positive mentions in the pediatric section of ADA’s Standards of Care (we hope to see more enthusiasm for adults with diabetes in future editions).
• CGM reimbursement made strides in 2017, particularly Abbott’s FreeStyle Libre and Dexcom’s G5. FreeStyle Libre secured national reimbursement in UK’s NHS (a major victory), France, and Japan. Dexcom’s G5 successfully secured Medicare reimbursement in the US at robust pricing (~$250/month) and began launching to thousands of beneficiaries in the second half of the year – this was a tremendous victory for the field. (Progress is still needed to allow Medicare users to access G5’s app and remote monitoring.) The Australian Government also announced in April that it would fully reimburse Dexcom and Medtronic CGM for type 1s under 21 years meeting fairly broad criteria. We hope to see this reimbursement momentum continue in 2018, especially as Dexcom and UnitedHealthcare begin their pilot testing CGM and coaching in ~10,000 people with type 2 diabetes. An unfortunate, unexpected frustration in this area was news that the reimbursement rate for physician interpretation of CGM data with report (95251) will decline 17% to $36.72 in 2018 (DTM 2017) – this is already depressingly low.
• We estimate CGM may now be used by roughly 0.7-1 million people globally, with global sales on pace to exceed ~$1.5 billion in 2017. FreeStyle Libre’s global user base exceeds 400,000 users, doubling the 200,000+ in Europe we noted in last year’s Reflections. As of 3Q17, Dexcom expected 270,000+ global users by the end of 2017, up ~35% from 200,000 at the end of 2016. And based on Medtronic’s ~25% of diabetes sales now from CGM (roughly ~$115 million per quarter), we estimate its global CGM user base is at least 100,000 users and possibly more.
• In the first three quarters of 2017, we estimate CGM sales from outside the US drove an estimated ~74% of the category’s growth (Dexcom, estimated Abbott, estimated Medtronic, Senseonics). US and OUS CGM sales are split almost down the middle, with an estimated ~49% of CGM sales coming from outside the US in the first three quarters of 2017. This looks far, far different from 2016, where we estimated that the US drove more than 75% of the field’s growth and sales. This year, we estimate FreeStyle Libre contributed just over 50% of the CGM category’s 1Q17-3Q17 growth, all on its OUS strength. Of course, the geographic growth split could change in 2018 as FreeStyle Libre’s US launch ramps up, and Dexcom, Medtronic, and Senseonics plan to launch their own updated products in the US (G6, Guardian Connect, and Eversense, respectively) – more on the US CGM market dynamics in the next theme.

US CGM Competition Heating Up On Products and Distribution – What Will The Market Look Like In 2018? Who Has the Edge On Different Features?

• The November US launch of Abbott’s FreeStyle Libre (real-time) has many wondering – how will the CGM market evolve in the US? What impact will FreeStyle Libre have on Dexcom’s US business – will it help build the CGM category or is it a threat? Most importantly, we’d emphasize that the US CGM market remains very underpenetrated in insulin users, leaving lots of runway for multiple companies with great products to succeed. Like any consumer product category, we expect continuous glucose data to evolve into several product form factors with various pros/cons at different price points. Indeed next year we may see four accurate standalone sensors competing in the US market, including Dexcom’s G5 and under-FDA-review G6 (launch in 2018), Abbott’s FreeStyle Libre, Medtronic’s under-FDA-review Guardian Connect (launch by April 2018), and Senseonics’ under-FDA-review Eversense (FDA advisory committee expected in early 2018). All these products have pros and cons, though we’ll be keenly listening to hear which points of differentiation drive the largest gains for which products. We include several key product features and points of differentiation in the table below. We have noticed a lot of noise around Libre vs. G5 in recent weeks, especially in consensus statements that differentiate the products (e.g., “intermittent scanning” vs. “real-time CGM”). To this point of Libre vs. G5, we’d again highlight how important it is for many, many CGM products to succeed in the US, given the category’s low penetration rate and the number of consumers who could actually benefit from more continuous glucose data.
• In the US, recent T1D Exchange data (see below) suggests that CGM penetration at top US centers is only at 24%! This has more than tripled in the past five years, but it means three out of four type 1s at the top US centers are still NOT on CGM. This also means CGM penetration across the US and globally is far lower than 24% in type 1, with negligible penetration in type 2 intensive insulin users – to say nothing of non-intensive users with type 2, who can also benefit from CGM. There is clearly plenty of runway for many companies to succeed! (Caveat: It is striking that CGM penetration is up to 45% in the <6-year-old group in the T1D Exchange– a clear sign of how much value this technology is adding for parents.)

• In a surprisingly aggressive move, Abbott has rolled out US marketing plans directly targeting Dexcom users. On the US FreeStyle Libre website, Dexcom users can sign up to get a FreeStyle Libre reader and one sensor for free. The website asks for a user’s Dexcom transmitter serial number, presumably to avoid scamming. Notably, Dexcom users will be contacted by phone or email to complete a survey regarding their experiences with the FreeStyle Libre system – we wonder what Abbott will learn and whether that will drive any product decisions on the planned continuous communication CGM (to be included as part of Bigfoot’s systems, with the Loop AID pivotal study expected in 2018). We’ve also heard about radio ads with similar marketing.
• Though it is not perfect, we found the below Google Trends analyses to be very interesting – this shows search engine queries in the US and globally for “Dexcom” (blue), “FreeStyle Libre” (red), and “CGM” (yellow) over the past 90 days and 12 months. Obviously, “FreeStyle Libre” is a product and “Dexcom” is a company, so perfect comparisons aren’t possible – “Abbott” does far more than diabetes, and “G5” is the name of many non-diabetes products. Limitations aside, we thought this was an insightful comparison of search volume, as it suggests: (i) an exponential increase in FreeStyle Libre searches in the week of December 17-23, followed by a fall in the past week; (ii) Dexcom and FreeStyle Libre’s US search volume are almost identical as the year closes out; and (iii) fairly constant searches for “CGM” over the course of 2017. It’s impossible to know if this is a reliable indicator of interest, sales, and adoption, since not everyone who searches converts to a user, and many don’t Google search at all (e.g., they might just go straight to the website or might just get a prescription). It’s also hard to know if “CGM” searches are a reliable indicator of category growth/interest. Globally, the story is different over the past 12 months (second picture) – “CGM” is more searched than either “FreeStyle Libre” or “Dexcom.” “FreeStyle Libre” has seen higher overall search volume globally since September (likely reflecting the US), but the gap has closed recently.

The numbers below represent search interest relative to the highest point on the chart for the given region and time. A value of 100 is the peak popularity for the term. A value of 50 means that the term is half as popular. A score of 0 means the term was less than 1% as popular as the peak.

United States Searches in Past 90 Days for “Dexcom” (blue), “FreeStyle Libre” (red), “CGM” (yellow) – Link

Worldwide Searches in Past 12 months for “Dexcom”, “FreeStyle Libre”, “CGM” – Link

Automated Insulin Delivery Goes Commercial with Gradual MiniMed 670G launch; Other companies converging on ~2019-2020 timing

• After years of discussion and painstaking research, we finally have a commercial, hybrid closed loop system on the market! Medtronic initiated the much-anticipated US launch of the MiniMed 670G in June following a short and small-scale “customer training phase.” The launch rolled out gradually to Priority Access Program participants (630G->670G upgrade), and as of November 21, over 20,000 people were on the 670G in the US. The rollout went far slower than we expected, given the planned “Spring 2017” launch and FDA approval in September 2016. Sensor manufacturing shortages played a big role in this slowdown, which only became clear during summer 2017. Medtronic expects to resolve these in February-April 2018 (F4Q18), at which point we’d guess the launch will go more full scale. Management has high double-digit growth expectations for the coming year, which should be achievable given the easy year-over-year comparisons. Still, Medtronic needs to execute and prove it can scale Guardian Sensor 3 manufacturing and the 670G onboarding process – the system still requires a fair amount of training. Although Medtronic is still well ahead of competition to launch hybrid closed loop, its first-to-market status has so far not driven the growth or adoption we would have expected at this point last year. The 670G international launch has also taken longer than expected, now expected by April 2018.
  
  • In last year’s Reflections, we were interested to see how well the 670G fared in the real world…in brief, it performs about as well as it did in the pivotal trial, at least for those who choose to share their data. As of November, Medtronic reported that ~13,000 patients are uploading 670G data into CareLink, with outcomes and performance “very similar to the pivotal trial.” At CDTM, 670G experts Dr. Jennifer Sherr (Yale) and Ms. Laurel Messer (Barbara Davis Center) discussed on-the-ground experiences with 670G, noting it makes a big difference in patient quality of life and glycemia, especially overnight. However, there are still plenty of people who would find it too burdensome or that it requires that they think too much about their diabetes.
  • As a first-in-class product, the 670G patient response has been more modest than we would have guessed. We look forward to seeing outcomes once the launch expands beyond early adopters, as well as what Medtronic’s plans are for 670G follow-up products – with a multi-year head start, Medtronic could quickly improve the product months before others make it to market. Key updates (in our view) would include fewer sensor calibrations, fewer alarms, a more aggressive algorithm/customizable target, automatic correction bolusing, a simpler menu structure, and smartphone connectivity.
• There are currently five companies aiming to launch first-gen AID systems in the US in ~2019-2020: Tandem, Insulet, Bigfoot, Beta Bionics, and Lilly – see our competitive landscape here and read specifics below. For a long time, it seemed that Animas could have been second or even first to market with its hypoglycemia-hyperglycemia minimizer, but given its closing (see below) and delays from most of the other manufacturers, Medtronic will now have the market to itself for at least another year (with the exception of Tandem’s PLGS system – more below). With so many options available in the next few years, we wonder how patients, providers, and payers will choose among them – even at this point, the products will have meaningful differences in form factor, burden, algorithm aggressiveness, personalization, and beyond. The four-year warranty cycle will continue to be a challenge, since it prevents a fluid consumer market – in any given year, only a fraction of current users can switch to a new pump and get it reimbursed (unless they switch insurance). We’d expect some will evaluate the automated options on the market and choose among them (potentially favoring Medtronic), while others will hold off on committing to a system until more are available (e.g., staying on a pump post-warranty expiration or going back to MDI – remaining a “free agent”). Choice is obviously good for patients, though it will still be a few years until this is a reality in AID.
  
  • Tandem’s treat-to-target AID system with Dexcom’s G6 and a TypeZero algorithm is expected to launch in 1H19, following a pivotal trial to start in 1H18. This represents a ~six-month delay from last year’s expected “2018” launch timing. It has taken much longer than expected to ramp up the iDCL trial, as FDA submission of the system was at one time expected by the end of 2017. Despite the delays, Tandem could still be second-to-market in the US with a hybrid closed loop system, although we would not be surprised to see further adjustments to the schedule. Nearer-term, the launch of Tandem’s predictive low glucose suspend (PLGS) device with Dexcom G5 integration is now expected in the summer 2018, also back ~6 months from the previous “2017” launch timing. On the plus side, the PLGS pivotal trial (PROLOG) is underway (expected to finish by the end of 2017), meaning if the submission goes in quickly early next year and review/approval happens fast, next summer is achievable. Inpatient feasibility data at ADA 2017 looked strong, and we continue to view this as a low-risk product with a proven algorithm, clear FDA path, and solid upside for Tandem to get a first-gen product to market quickly. As a huge perk, Tandem customers will also be able to upgrade their t:slim X2 with G5 to the PLGS algorithm at no cost and in the comfort of their homes. It will be interesting to see how the system competes with 670G, since it only has the bottom half (hypoglycemia mitigation) of a hybrid closed loop.
  • Insulet made meaningful AID progress in 2017: it’s currently conducting a five-day hotel study (n=48) of the Omnipod Horizon Automated Glucose Control System, and eyeing a launch by the “end of 2019” or “early 2020.” As of DTM in November, 118 subjects have been studied in the hotel setting, accounting for 7,584 hours of wear time. At that meeting, powerhouse Medical Director Dr. Trang Ly shared interim outcomes from 11 patients, showing strong 74% time-in-range (70-180 mg/dl), 1.8% of time <70 mg/dl, and 25% of time >180 mg/dl. Mean glucose was 150 mg/dl. Following completion, the system will be tested in a pre-pivotal study, and finally a pivotal trial as the last clinical development step. See our ADA coverage for a summary table of the many studies conducted as of June (through IDE2 – the hotel study is IDE3) – they are many in number and will certainly help Dr. Ly and team develop a comprehensive understanding of the system’s strengths and weaknesses. That each three-day, disposable, Bluetooth-enabled OmniPod will contain the algorithm is a big win for patients (no need to keep the PDM nearby to stay in closed loop). Though it may not be second to market, Horizon could be among the strongest and most differentiated AID products when it comes out.
  • As of July, Bigfoot’s automated insulin delivery system (Bigfoot Loop) will use a second-gen Abbott FreeStyle Libre sensor with continuous communication – the company now anticipates a US pivotal trial to commence in 2018, followed by a PMA submission at the “end of 2019” and an anticipated “2020” launch. The timing is unquestionably ambitious, as the system will progress straight to the pivotal trial without an intermittent feasibility study incorporating FreeStyle Libre. Bigfoot has done extensive simulation modeling, but just one feasibility trial (n=20) with prior partner Dexcom G5. We were very impressed at DTM 2017 when Director of Clinical Innovation Ms. Jen Block explained that Loop’s algorithm is capable of adjusting basal rates, carb ratios, insulin sensitivity factors, and glucose targets on an impressive hour-to-hour basis. The 48-hour feasibility data (n=20) was promising, and Bigfoot’s impressive vClinic ran over 100 million subject-days of simulation prior to the human trial. Notably, the vClinic can run the equivalent of a three-month trial (200 subjects, 100 days) in “less than a minute” according to Ms. Block. With recent FDA draft guidance indicating interest in leveraging computer modeling and simulation, we wonder how much Bigfoot can leverage in the regulatory process. We also remain optimistic about the plan to shake-up the business and service models of pumps and AID with a single-prescription, monthly subscription model, and reimbursement based on outcomes.
    • Bigfoot had quite the year: Along with the exciting Abbott FreeStyle Libre partnership announced in July, Bigfoot acquired Timesulin and its insulin pen dose capture technology in June. We like the vision of “Automated Insulin Delivery for Everyone,” which includes both the “Bigfoot Inject” (MDI auto-titration system) and “Bigfoot Loop” (automated insulin delivery system with the pump acquired from Asante) platforms. Both include the next-gen Libre and the Bigfoot smartphone app serving as the user interface.
  • Beta Bionics told us in an email that it aims to begin pivotal trials of both its insulin-only and bihormonal systems in the beginning of 2019, with a possible insulin-only launch in 1H20. FDA submission and launch timing for the bihormonal version depend on a stable glucagon – a Zealand dasiglucagon home-use trial in the iLet integrated device is expected in 2018. Like other companies in our landscape, Beta Bionics has seen a lot of pushback in the timelines. For instance, at this time in 2016, an insulin-only PMA submission was expected by mid-2018 and a bihormonal pivotal was expected in 1H18. Still, as of December 21, the public benefit corporation has received nice votes of confidence from three diabetes pharma players: Lilly ($5 million in April 2016), Novo Nordisk ($5 million in February 2017), and Zealand (a $1.5 million equity investment this month). In the other potential partner news, we learned at IDF that Beta Bionics is exploring Senseonics’ Eversense as a possible “additional sensor option,” in addition to Dexcom, after Dr. Steven Russell noted a strong Eversense performance vs. G5 and Libre in a six-week outpatient study.
  • Lilly’s Cambridge Innovation Center came out of the woodwork in the past 30 days, sharing plans to commercialize AID with a proprietary pump and a smart pen with insulin titration in the next two to three years. A phase 1 study of the AID system is already underway. The AID system is composed of Lilly’s own tubed pump (developed by DEKA; founder Dean Kamen is known for inventing the Segway and the first portable insulin pump in 1973), Dexcom CGM, and an acquired hybrid closed loop algorithm from Montreal’s Class AP. The pump is described as “a white disk about the size of a shoe-polish tin,” holds three days of insulin, has an infusion set, and can be worn direct on the body or in the pocket. It will be controlled wirelessly from a handheld and comes with a companion app. We had guessed there were big things happening inside the Cambridge Innovation Center since its 2015 inception, but we did not expect Lilly to jump full in and become a diabetes tech company so quickly! Of the three major insulin producers, Lilly has arguably taken the biggest public-facing plunge into technology.
• In Europe, Medtronic (670G launch by April 2018 per 3Q17) and Diabeloop (launch expected in early 2018) are vying to be the first hybrid closed loop products to market. Roche/Senseonics/TypeZero (pivotal completion by late 2018 to support CE mark) and Cellnovo (TypeZero integration currently being finalized, with pilot study for validation to follow in a limited number of patients) also hope to launch in the next couple of years, at least as the landscape currently stands. We look forward to gauging the different health systems’, providers’, and patients’ appetite for closed loop therapy in the heterogeneous European market.
• In February, NIH announced ~$41 million in grants for four “pivotal” multi-center artificial pancreas trials starting in 2017-2018: (i) the International Diabetes Closed Loop trial (including a US arm with Tandem/Dexcom and an EU arm with Roche/Senseonics); (ii) DAN05, a Cambridge pediatric trial in 6-18 year olds slated to begin earlier this year (n=130, 12 months!, using the Medtronic 640G/Enlite 3 and an Android phone); (iii) an IDC-Schneider’s Children’s collaborative comparing Medtronic’s MiniMed 670G vs. advanced hybrid closed loop with automated correction boluses (DreaMed algorithm) in a crossover study previously slated to start in late 2017 (n=100, three months,); and (iv) BU/MGH’s bihormonal pivotal trial now scheduled to begin in early 2019 (n=312, six months, iLet device with insulin+glucagon). At DTM, we got a run-down on the slow-to-launch iDCL trial, which now has four protocols and two primary ones: (i) The EU pivotal trial (n=72) with Roche/Senseonics/TypeZero (to begin in 1Q18); and (ii) the US pivotal trial for Tandem/Dexcom/TypeZero (to begin in 1H18). The NIH grants demonstrate tremendous commitment to the field – we hope this can continue, assuming the Special Diabetes Program is renewed by Congress. We look forward to an entire session devoted to updates on the four studies at ATTD in February in Vienna.

The Insulin Companies Go More Digital: Lilly, Novo Nordisk, and Sanofi Move into Diabetes Tech with Unique Approaches

• Strategic moves from the three large insulin manufacturers over the past year have made it clear that traditional pharmaceutical companies no longer envision a secure future just manufacturing drugs. We agree! Interestingly, each company is taking a slightly different approach to their technology/digital health forays: (i) Lilly opened the Cambridge Innovation Lab a couple of years ago, and recently announced that it was developing a smart pen/insulin dose titration platform and an AID system all in house (pulling in Dexcom CGM); (ii) Novo Nordisk has teamed up with Glooko to develop jointly-branded digital health solutions, including insulin titration, and is currently piloting a connected pen in Sweden – progress from an IBM Watson partnership has not yet been shared; and (iii) Sanofi formed a $496 million joint venture with Alphabet’s Verily late last year (Onduo), has initiated a payer pilot of its own with Common Sensing and One Drop, and recently announced results from an under-the-radar IBM Watson collaboration. All three have also invested in diabetes tech startups (Lilly in Companion Medical, Novo Nordisk and Lilly in Beta Bionics, and Sanofi in Common Sensing). Of course, despite all of the conference announcements and press releases, it is too early to know what will actually come to fruition. How will large pharma companies merge with the fast pace of technology development and iteration? What approach will win out – fully in-house development, or a partnering approach – or will both be successful? Which business models will be most successful? Will companies create “walled gardens” of software and hardware that are only compatible with their drugs? If so, will these digital add-ons be provided for free with a drug purchase? Might companies take on full risk and be reimbursed solely for outcomes and population management? If we are heading down that road, how will the pharmaceutical company interact with traditional payers and providers? Will innovation in future diabetes drug molecules be dwarfed by differences in the quality of supporting software, hardware, services, and human touch? One thing is certain: This is the most complicated, nuanced, and exciting time to follow the diabetes market that we can recall, and we’re hopeful for these new combinations of drugs and tech.
• Lilly announced plans in November to develop two tech-enabled insulin delivery systems and bring them to market in ~two to three years. The move represents the biggest public commitment from an insulin company to automating insulin delivery, as Lilly will submit the PMA and commercialize the systems itself. As noted above, this includes (i) an automated insulin delivery system with its own pump and Dexcom CGM (discussed above) and (ii) smart pens with a paired dose titration app, supplemented by a partnership with Rimidi to develop provider-focused tools. While still a non-exclusive, non-commercial agreement, Dexcom’s development agreement with Lilly is also a strong response to Bigfoot/Abbott’s partnership, which has similar aims. Lilly does have some clear advantages (ability to bundle with insulin, large sales rep footprint, payer relationships, etc.) but Bigfoot also comes with deep digital and algorithm expertise, a market-ready pump (acquired from Asante), and Abbott’s factory-calibrated CGM. We’ll be watching the market dynamics of MDI+dose titration vs. AID with a pump – the segmentation is hard to call for now, but there will presumably be differences in outcomes, ease of use, and cost. For companies offering both smart pens and pumps, patients can conceivably switch back and forth between systems, potentially encouraging loyalty within a given company’s ecosystem. Lilly will bring strong potential competition on the product and business model fronts, though it must also prove it can build and run a pump/AID company within a pharma company – not a given!
• At the Novo Nordisk-sponsored HITLAB Symposium in Palo Alto, Novo Nordisk shared a 90-second video of the company’s digital health vision, complete with a connected insulin pen, patient-facing app with dosing guidance, automatic carb counting, real-time CGM, and a healthcare provider dashboard with clinical decision support. The first tip off toward this vision came in the form of an IBM Watson partnership late in 2015. Early in 2017, the company struck an alliance with Glooko to develop jointly-branded digital tools, and has since delivered a first-gen app in Cornerstones4Care (the framework upon which other functions will be built…presumably including insulin titration). In that vein, Glooko is currently recruiting for a trial evaluating its mobile insulin dosing system (MIDS) with Tresiba – MIDS previously proved efficacious in a three-week feasibility study. Novo Nordisk also has connected NovoPen 5 Plus pens piloting in Sweden, and a connected Tresiba pen is being used in a three-month NIH-funded study of TypeZero’s inControl MDI Advisor at UVA, Stanford, and Mt. Sinai. Additionally, the company’s recent (and final) Triple Bottom Line quarterly magazine focused exclusively on digital health. We’d guess Novo Nordisk will remain fairly quiet until it has more to share – perhaps in 2018.
• Onduo (Sanofi + Verily) and the Blue Cross Blue Shield Association announced plans to pilot a type 2 diabetes management program in three states in 1Q18, including connected CGM, apps, and continuous lifestyle and medication support through Onduo’s healthcare experts. Sanofi is also involved in a pilot with Innovation Health, Common Sensing, and One Drop and has its own FDA-cleared My Dose Coach app. At IDF 2017, Sanofi shared results from an exciting collaboration with IBM Watson (apparently beating Novo Nordisk to the punch) – the two companies used big, real-world data and analytics to answer the questions: What are the predictors of success with insulin+GLP-1 combination therapy? What are the chances of a patient having success on basal insulin after 12 months? Deriving (and updating) these answers based on big data sets, and then providing recommendations to providers either in the form of written guidance (i.e., labels) or automated decision support is an exciting frontier.

Fragile US Insulin Pump Market for Animas, Roche, Medtronic, and Tandem; Slow Type 2 Diabetes Insulin Delivery Progress

• 2017 was a year of three tiers in the US pump market: (i) Roche discontinued US pump sales and Animas exited the market entirely; (ii) Medtronic and Tandem experienced YOY US sales declines in 1Q17-3Q17 (-3% and -2%, respectively), despite launching important new products; and (iii) Insulet, the lone bright spot, saw US Omnipod sales grow 17% YTD. The overall market fragility is hard to pin on one specific factor, but it could speak to:
  
  • Not enough gamechanging innovation on the product or business model fronts– Why is getting an insulin pump compelling for someone with diabetes? How are insulin pumps making life far better, relative to what they ask of the user and the system? What are pumps offering payers, who have fewer dollars to spend on more people with diabetes? What are pumps offering healthcare providers, who have less time to spend each day on more people with diabetes? Automated insulin delivery will clearly improve the value proposition meaningfully, but at the current rate, it will be several years until all insulin pumps have these upgrades.
  • The difficulty of running a sustainable, profitable insulin pump business – J&J and Roche could not make it work, we’re not sure if Medtronic has ever made significant margins on the MiniMed acquisition, Tandem is still far from profitability, and Insulet is only just getting close. How can companies rethink manufacturing, operations, and the cost structure of running a pump company?
  • Better substitutes? – More type 1s are likely moving to CGM before a pump, and we’d guess next-gen basal insulins like Tresiba and Toujeo are making MDI far better for these users – both trends could make pumps less attractive as a further add-on.
  • Market saturation? – How many insulin pump companies can the US market support? With no meaningful runway into type 2 diabetes (yet), only a handful of companies can presumably find enough patients to run a sustainable business.
• In short, insulin pumps still hit a minority of people with type 1 diabetes, with many of the same barriers we’ve talked about for years: upfront cost/reimbursement hassle (including the four-year reimbursement cycle); size on the body; complexity in prescribing, training and using; clinical and economic value of non-automated pump therapy over MDI; patient quality of life benefit vs. burden; infusion set hassles; and beyond. Indeed, Insulet has been the only company to grow its US pump business in the first three quarters of the year, arguably because it is addressing more of these barriers than others.
• On the plus side, the pump market should meaningfully change in the coming years as automation, novel device designs, connectivity, personalization, integration with consumer electronics, and new companies with new business models change the paradigm. Indeed, it is compelling to think of a single-prescription, outcomes-based subscription model for automated insulin delivery, which would include all supplies, a pump/sensor/algorithm that learns the user, adjusts inputs as needed, alerts the user when maintenance is required, offers remote monitoring, ships supplies automatically, and makes life easier for HCPs. We see this period in pumps as we see pretty much the rest of healthcare: in the midst of revolutionary change and full of potential for the right product combinations that blend software, hardware, and services.
• Here’s a company-by-company breakdown of the US pump market in 2017:
  
  • With the cessation of Roche’s and Animas’ sales this year, patients in the US exit 2017 with just three pump options: Medtronic, Insulet, and Tandem. The Animas exit was one of the biggest diabetes tech stories of the year – J&J’s is now officially out of the pump business, leaving its large base of ~90,000 Animas pumpers to “preferred partner” Medtronic. (Tandem and Insulet stepped up with appealing welcome programs of their own, and we’ll see in the coming quarters what impact those have.) As we said in October, the pump market’s fragility continues to be a cautionary tale for new entrants, and the future of sustainable, profitable companies may rest as much on product innovation as business model innovation. J&J had been considering sale of Animas and LifeScan since at least January, but this straight exit implied there was not a viable buyer/offer for the business. Continuing to operate Animas was presumably not profitable, and we’d guess there was no near-term path forward for improving things. The bar for product excellence continues to rise in diabetes devices (just look at CGM), and Animas did not move quickly enough on automating insulin delivery – it could have been first to market with a hybrid closed loop, but did not execute as many had hoped. On the other hand, a Roche re-entry to the US is not out of the question – it still sells pumps outside of the US, and the Accu-Chek Insight pump and Senseonics’ 180-day Eversense CGM will be used in an arm of the International Diabetes Closed Loop trial.
  • One of the more surprising performances of the year came from Medtronic, which has had a trying three quarters in 2017 despite gradually rolling out the world’s first hybrid closed loop system to high expectations. US sales still fell 5% YOY in 3Q17 ($258 million), fell 8% YOY in 2Q17 ($243 million), and grew just 3% in 1Q17 ($303 million). CGM sensor capacity issues and accounting considerations (i.e., revenue recognized at the time of enrollment in the Priority Access Program) likely posed headwinds, but 670G didn’t explode out of the gate in a way that we thought it could. Management expects double-digit diabetes growth in May 2018-April 2019, but this will rest on excellent execution – especially on scaling sensor manufacturing. 
  • Tandem struggled in the first half of 2017 (particularly against the backdrop of strong t:slim G4 sales in 2016), but picked up steam in Q3 with launch of the t:slim X2/Dexcom G5 in August. Interest from former Animas customers could drive further growth in Q4. Many wondered about the fate of the company this year, but the user-friendly touchscreen pump, strong pipeline, remote software upgrade capability, Bluetooth compatibility, and improved cash position suggest high potential. Now, it comes down to execution, a strong upcoming Q4 and 1Q18, and raising enough cash to sustain operations. In addition to the automated insulin delivery pipeline (discussed above), Tandem is also seeking regulatory approval outside of the US soon, with Canada as the number one priority. The user base was over ~60,000 in late October, and we see especially high potential for Tandem to take advantage of JDRF’s open protocol initiative – could it be the first interoperable Bluetooth-enabled open protocol pump available in the US?
  • Insulet has been the lone bright spot in the US pump market, with growth in the mid-to-high teens in each of the first three quarters of 2017. In both 2Q17 ($65 million) and 3Q17 ($70 million), US Omnipod sales hit new records, and the installed base – ~80% of new users from MDI – is expected to grow to ~130,000 by the end of 2017. In the pipeline, Insulet has the Bluetooth-enabled Omnipod Dash PDM and pod (FDA submission around the end of this year, with a limited market release in mid-2018), and a well-studied hybrid closed loop algorithm expected to launch by the end of 2019 or early 2020 (per the 3Q17 call). The company has maintained its lofty goals for 2021: $1 billion in revenue, gross margins of 70%, and above-market profitability – a lot of things have to go right by then.
• Insulin delivery devices for type 2 diabetes (full-featured pumps and simple on-body devices) have not taken off, despite a sizable potential market. J&J’s promising OneTouch Via bolus-only patch device has yet to launch, though it is technically ready (will it be sold?); Valeritas V-Go sales have hovered around $5 million since 2Q15; CeQur’s Paq has yet to be filed with the FDA, to our knowledge (though it is CE marked); launch of BD’s type 2 pump has been pushed back to October 2018-September 2019; and Insulet’s Lilly U200 and U500 Omnipods (see screenshots) are tracking to 2019 and 2020 launches, respectively. The roundup of this area is very similar to that of last year – Valeritas has clearly had the most commercial experience, but uptake has been slow. Meanwhile, Medtronic’s type 2 diabetes efforts are far more focused on CGM, and Tandem’s t:flex pump has not seen a lot of uptake. Discreet and more consistent insulin delivery seems like a no-brainer for type 2s, so we wonder (i) if there is a true market and (ii) what is holding back the field. Reimbursement/cost? Awareness? Lack of demonstrated outcomes? On-body design and user experience? Clinical inertia? We also wonder how connected insulin pens will compete with on-body insulin delivery devices in type 2 – how will the market segment? Can on-body devices get to much lower costs and big scale? Will any insulin players seek to offer a prefilled V-Go-like device, bringing commercial muscle, strong reimbursement, and more convenience?

Smart Pens/Caps and Insulin Dose Titration – Now 20+ Companies, On the Cusp of a Commercial Market?

• This year saw a notable acceleration in the pipeline of smart insulin pens/caps and titration software – combined, we count more than 20 companies in our competitive landscapes, roughly doubling year-over-year (Smart Pens/Caps Competitive Landscape, Dose Titration Landscape). There’s no question interest has exploded, but at this stage, no products are available at scale or widely used, at least from what we can tell. The field remains in “pilots,” “limited launch,” R&D, and/or awaiting commercial partners – not a bad thing, since getting this right will drive huge gains for the diabetes ecosystem. Companion Medical’s InPen was an important story this year, limited launching in December as the first commercially available smart insulin pen in the US. InPen came out with stronger-than-expected reimbursement from major payers (many patients will have copays of $0 or $50), hopefully an encouraging sign for the field. Meanwhile, at least four patient-facing insulin dose titration apps are now FDA cleared, including Voluntis’ Insulia (basal-only), Amalgam’s iSage Rx (basal-only), Lilly’s Go Dose (Humalog only), and Sanofi’s My Dose Coach (basal-only) – this quadrupled year-over-year, when only Insulia was FDA cleared. Two companies – Lilly and Bigfoot Biomedical – have publicly committed to launch combined smart insulin pens/caps + paired insulin dose titration in roughly 2-3 years. Both of these were major news items this year – Lilly announcing its connected diabetes ecosystem in late November, and Bigfoot announcing the acquisition of Timesulin in June. In the pilot or R&D phase, Novo Nordisk is testing connected pens in Sweden and in the US, while BD expects to launch its own smart pen needles in October 2018-September 2019 (FY19, delayed from the previous 2H18 expectation). Traditionally-quiet Common Sensing finally became more public about its own Bluetooth-enabled Gocap this year, sharing one of our favorite posters at ADA in June and announcing a type 2 diabetes pilot with Sanofi, Innovation Health (Aetna), and One Drop. Glooko has its own Mobile Insulin Dosing System under FDA review, and closed-loop algorithm developers DreaMed and TypeZero also became more public about their injection titration plans – we saw data at ATTD (MD Logic Advisor Pro) and DTM (inControl MDI Advisor). Glytec continues to amass positive inpatient and outpatient data for its Glucommander clinical decision support software, now with four integration partnerships in the works (Livongo, AgaMatrix, Telcare, Onduo). Hygieia made progress too, expanding access to its d-Nav Insulin Guidance Service across the UK and continuing to run a pilot with Blue Cross Blue Shield of Michigan. Last, Onduo (Sanofi/Verily) revealed plans to use both Glytec and Voluntis’ dose titration solutions in 2018 pilots, along with separate news to use connected CGM + experts to review medication doses. What a year!
• We’re glad to see the pipeline progress, though it should be noted that titration apps and smart pens are not widely available or routinely prescribed. The major insulin companies will play a big role here in raising awareness, though most are still in the development phase (see theme above). Outcomes data is also somewhat limited at this point, meaning more studies will need to show smart pens/caps and dose titration drive benefits – we have no doubt these systems will show non-inferiority to current care and likely superiority.  

Big tech companies advance further into diabetes, especially Verily (Alphabet), Apple, and Fitbit

• As we write this, two of the five “FAANG” (Facebook, Apple, Amazon, Netflix, and Google) have a diabetes presence, and two more are involved peripherally. By far, Alphabet (Google/Verily) and Apple showed the most commitment this year. Apple is laying a strong software and hardware groundwork for better, more comprehensive, on-the-go, interoperable data on health (HealthKit with glucose and insulin data, Dexcom partnership for Apple Watch, Apple Watch EKG band, etc.), while Alphabet (Google/Verily) is taking a more direct partnership approach with at least five bets in diabetes (Dexcom, Onduo, Novartis, GSK, Project Baseline). This large tech move into healthcare shows that data and low-cost, consumer-friendly devices to collect it will be key currencies pushing the field forward. It’s unclear at this point how far – or how successfully – these consumer companies will push into healthcare or diabetes. Will they stop at ancillary support or eventually replace the traditional channels of care delivery and payment? Will they take on care delivery themselves, or focus on propelling others? Since many of these partnerships have lofty ambitions – see below – it will be some years until we know for sure. However, we love seeing these successful consumer companies come to our field, since they bring deep tech-sector knowledge that should improve hardware and software and data in our field. Below, we’ve highlighted the key happenings in 2017 for Verily, Apple, Fitbit, Amazon, and Facebook.
• Verily now has at least 15 projects in healthcare, of which five have direct ties to diabetes: (i) the factory calibrated, fully disposable CGM with Dexcom; (ii) a diabetes management platform from the $496 million joint venture with Sanofi (Onduo); (iii) a smart glucose-sensing contact lens with Novartis; (iv) a mini implanted nerve modulation to treat chronic disease with GSK; and (v) retinal imaging work leveraging machine learning to detect diabetic retinopathy and diabetic macular edema with Nikon. Verily also recently announced a new partnership with NHS to focus on risk prediction models for chronic disease, including diabetes. The Dexcom, Onduo, and retinal imaging projects saw the most updates this year:
  
  • The first-gen sensor with Dexcom is expected to complete development in 1H18 as of Dexcom’s 3Q17 call, but the commercial launch now depends on the G6 no-calibration regulatory strategy. While the device has historically been positioned as a type 2/non-intensive product, we’d guess it will also be available to type 1 users – since it uses the G6 sensor, it should maintain the insulin dosing claim. This Dexcom/Verily gen one sensor will also offer a big upgrade in form factor (reportedly smaller on the body than FreeStyle Libre) and a much-needed fully disposable design (no hassle of reordering transmitters). Meanwhile, the Dexcom/Verily R&D teams are “accelerating” efforts on the smaller, second-gen, bandage-like sensor – this has “aggressive” cost targets and quite a compelling form factor. Previously, it was slated for a 2020-2021 launch, which feels reasonable – assuming Verily gen one launches in 2019 (our speculation), 1-2 years later to follow with a much smaller and lower cost version that is already in development seems quite possible. This partnership remains just as exciting as it was when the debut announcement came in 2015, though the initial plan to launch within “~2-3 years” is coming up in 2018 – will the companies bring the first-gen sensor to market next year? Will it be strong competition for FreeStyle Libre in the US? Will it be offered to both type 1 and type 2 users, or will Dexcom bifurcate its offerings? We hope to learn more at JPM in January.
  • After a quiet year, Onduo announced an intriguing pilot with Blue Cross Blue Shield (beginning in 1Q18), as well as a slew of BGM and insulin titration partnerships. We like the platform/curation vision to combine connected glucose monitoring, apps, coaching, and medication review to help type 2s. Can Onduo pull it all together and scale it? We should get more data points next year.
  • Verily is also doing interesting diabetes work on automated retinal screening and Project Baseline. We last heard Diabetes Clinical Lead Dr. Howard Zisser speak at DTM 2017, where he shared that there are ongoing clinical trials of the automated retinal screening system in India – crucially, both investigating sensitivity/specificity as well as workflow. We’re excited to see Verily strive towards launching products in nations with dire shortages (India alone has a shortage of 127,000 eye doctors). Verily also just announced a diabetes initiative of the mammoth Project Baseline study. The 12-month, n=200 study will investigate a diabetes management platform including an app, coaching, and connected devices (including Dexcom’s G5 and Telcare’s BGM). 
• Apple showed stronger commitment to entering the diabetes field in the past year. In an unexpected win, Apple’s high profile June special event shared on stage that Apple Watch will talk directly to Dexcom’s CGM without the phone nearby – this Native Core Bluetooth update will be a massive convenience win and could unlock a powerful data stream pairing glucose with activity, heart rate, and beyond. According to a December New York Times piece on Apple Watch, Dexcom’s direct G5-transmitter-to-Apple Watch communication is actually under FDA review now – the first specific timing update in six months (Dexcom has been notably quiet about this). Meanwhile, Apple’s HealthKit can now read/write insulin dosing data, expanding interoperability in the digital diabetes ecosystem as smart pens and Bluetooth-enabled pumps come to market. FDA also recently cleared an EKG Apple Watch band, continuing the move to turn the watch into a medical device. Rumors continued to swirl of Apple’s drive to make non-invasive CGM, a trend that continued in the December New York Times piece. If this is actually in development, could it actually come to market and be accurate enough?
• Fitbit bolstered its diabetes presence this year, supplementing its Medtronic partnership (professional CGM) to add Dexcom and One Drop. Both new partnerships will build apps for the Fitbit Ionic smartwatch in 2018, plus bring obvious potential to combine diabetes data with Fitbit activity, sleep, and heart rate data.
• With Apple Watch and Fitbit expanding into diabetes, users will now have seamless access to valuable data streams on the wrist – a particularly compelling, discreet hub from which to manage diabetes and motivate behavior change. Importantly, these advances also represent a push toward smartwatches as standalone devices in health – might payers eventually cover smartwatches as medical devices? Recall that, in August, rumors were swirling that Apple and Aetna might bring the Apple Watch to Aetna members. We envision the day when the DIY-only ability to control insulin delivery from a watch is also an industry staple – something users of the DIY Loop system are already doing. Apple Watch has the clear edge over Fitbit given superior form factor, more functionality, and cellular, but Fitbit’s Blaze and Ionic smartwatches have consistently occupied the top spots on Amazon’s Best Selling Smartwatches list, boasting 4+ day battery lives and sleep tracking.
• Fitbit, Apple, Verily, and Samsung were all accepted into the FDA’s digital health software Pre-Cert pilot program, marking these big tech companies as stronger players in digital healthcare – especially as they move into regulated product that speak to medical devices. (Tidepool, Roche, and J&J were also accepted.) Pilot participants will help the FDA in shaping the Agency’s Pre-Cert protocol, which aims to expedite product reviews from vetted companies, either by minimizing the burden of submissions or eliminated them in some cases. Not only does this move indicate big-tech commitment to healthcare from Fitbit, Apple, Verily, and Samsung, it also demonstrates a potential vote of confidence from the FDA.
• Facebook and Amazon have also stuck their toes into the healthcare pool, though the ties to diabetes are more peripheral. Facebook, according to CNBC, held a meeting in September on clinical trial recruiting, and the company has deployed AI to halt preventable suicides. Meanwhile, Amazon has created a frenzy of buzz in healthcare by obtaining pharmacy-wholesaler licenses in a dozen states – with its size and scale, even just in distribution, Amazon could disrupt retail drugstores and PBMs. That said, this has been a rumor for years, so we’ll have to see if it’s true. Amazon’s move likely urged CVS to launch free same-day home prescription delivery starting in select cities, with next-day home delivery to go nationwide in 2018 – we love the potential there for getting drugs and devices to people faster and with greater convenience.

Payers More Interested in Diabetes Tech/Digital Health? Positive Pilot and Coverage News in 2017

• Three payer/industry pilots have popped up in the past three months, indicating more payer interest in diabetes technology, digital health, and service models with human coaching – especially for type 2 diabetes. (i) Innovation Health (Aetna+Inova Health Systems) and Sanofi have initiated a pilot using Common Sensing’s dose capture technology and One Drop’s app, coaching, and Bluetooth-enabled BGM; (ii) Onduo (Verily+Sanofi JV) and Blue Cross Blue Shield followed with a three-state pilot with connected CGM, apps, lifestyle, medication review, and Onduo experts; and (iii) UnitedHealth and Dexcom are ramping an unprecedented pilot of CGM and coaching in 10,000 type 2s over the next six to nine months (the announcement came a few months after former UHC president Richard Collins was appointed to Dexcom’s Board of Directors). We’re excited about all three pilots, especially because they will focus on using technology and data to help people with type 2 diabetes. Of course, they may also signal major US payers are more open to keeping people with diabetes healthy and productive via connectivity, real-time data, on-the-go smartphone access, personalized guidance, human touch, and in-between-appointment support/coaching. We expect these pilots will drive learning, experimentation, and answers to important questions – e.g., Do these combination approaches work? How do payer or payer-contracted coaches interact with PCPs and the traditional channels of care delivery? What should the reimbursement contracts look like? What outcomes should serve as the basis for payment? Clinical studies of the components (CGM, coaching, etc.) alone show they are effective, so when integrated in a personalized way, we have high hopes!
  
  • One question is whether these “pilots” are translatable at scale – can these approaches reach and support millions of people with diabetes?  It’s a bit of a paradox: Iteration and agility require testing these integrated approaches in smaller settings, and it is inherently easier to deliver personalized care in these settings, particularly if human coaching/review is involved. If you scale up, you know more about how the solutions behave in a more realistic environment, but then it’s more difficult to iterate and learn. We’ve even heard some people refer to the tendency toward small-scale pilots as “pilotitis” – mentioned both at The diaTribe Foundation’s D17 event and in Q&A at DTM. We look forward to monitoring how close each of these come to striking the “right” balance. Regardless, we’re very excited by this initial payer enthusiasm, as curation, connectivity, and coaching are promising frontiers in population diabetes management. These pilots might even be informative for general population health and wellness management.
• In June, Medtronic Diabetes announced an outcomes-based deal with Aetna: if MDI users moving to pumps don’t achieve a certain A1c benefit, Aetna gets a rebate. Will we see more risk sharing/outcomes-based deals in the years ahead? As part of the deal, Medtronic will be paid for pumps like it is now, but if pre-agreed A1c outcomes are not met, Aetna will receive a “rebate payment.” There are no details on the rebate payment size or magnitude of A1c improvement. This deal represents an initial foray into outcomes-based reimbursement in diabetes technology, since Medtronic will be paid whether it achieves outcomes or not. Short term, we don’t imagine it will have a major impact on the commercial environment, but longer term, we expect the payer environment will increasingly be a key battleground on which diabetes technology is fought.
  
  • Notably, Medtronic’s MiniMed 670G is now covered by all 25 major private payers in the US. Anthem was the lone high-profile holdout for a number of months, calling hybrid closed loop “investigational and not medically necessary under all circumstances.” Possibly spurred by the JDRF #Coverage2Control petition that amassed 50,000+ signatures, Anthem revised its Policy in November, citing “expert clinical opinion.” In the same Policy, there were 32 peer-reviewed publications on automated insulin delivery, an indication of the clinical evidence in this field. It is an encouraging victory that the first-to-market automated insulin delivery system has quickly achieved widespread reimbursement – will it help other systems in the competitive landscape? Will systems’ outcomes be differentiated enough to drive different payment? Besides Bigfoot Biomedical, will any players seek to change the reimbursement model for AID?
• Payers are also beginning to de-risk product development by letting their voices be heard early in the process. At MDIC’s Annual Public Forum, BCBS Executive Director of Clinical Evaluation, Innovation, and Policy Dr. Naomi Aronson introduced the payer working group Excite International, whose mission is to accelerate pathways for novel, beneficial technologies. The international working group is also populated by reps from Kaiser Permanente, Aetna, Ontario, NICE, Norway, and CMS (as an observer). She urged manufacturers to go to the group to get a consensus opinion on a likely coverage decision. This (along with FDA/NICE’s Payer Communication Task Force) is long overdue – regulatory guidance on the device side is increasingly well fleshed out, but getting clearance/approval doesn’t guarantee access. Early payer input should ideally allow manufacturers to get to initially launch with strong reimbursement.

Big 4 BGM Companies Seek to Reinvent Themselves, Taking Different Paths

• Resembling the trend in diabetes pharma, glucose monitoring is undergoing a lot of change – and with it, the Big 4 BGM players (Roche, Abbott, J&J, and Ascensia) are taking different paths to reinvent their businesses. For Roche and Ascensia, this has meant going digital and signing data integration partnerships. For Abbott, it’s been about investing in FreeStyle Libre and capitalizing on the move to CGM. And for J&J, it’s increasingly looking like an exit from the BGM business (if Animas was any indication…). The upside to the BGM field’s upheaval is a chance for these companies to take a new approach and rethink glucose monitoring. In recent years, these companies’ diabetes talk has mostly focused on pricing and competitive pressures. Now, we hear a bit more optimism (especially from Abbott), and the name of the game is adding value and context to glucose values – either through continuous sensing, connectivity, partnerships, dose titration, or even new business models. With the promise of more factory-calibrated CGM devices coming down the pipeline, the BGM market is growing increasingly fragile. Then again, CGM won’t be for everyone, and the $4+ billion BGM market won’t disappear overnight.
  
  • Abbott has taken the boldest and most successful tack so far, doubling down on FreeStyle Libre. Abbott took the ahead-of-its-time Navigator sensor technology and built a compelling product around it – setting a new standard for what effective, less burdensome, less costly, more valuable diabetes technology looks like. Future integration into Bigfoot’s Loop and Inject systems – plus potential for a standalone continuous CGM version (our speculation) – Abbott is well positioned for the next decade in glucose monitoring. Although the company has not officially separated out BGM from Libre sales, we’ve estimated that quarterly FreeStyle Libre sales now exceed $100 million. We wonder if the coming quarters we’ll get more distinction to determine how much BGM sales have been deprioritized.
  • Roche took several strong moves in 2017 to go digital, acquiring mySugr in June and featuring the company’s open diabetes management ecosystem at EASD, complete with Senseonics’ Eversense implantable CGM, mySugr, and even a reusable smart pen available in cooperation with Pendiq in Germany in 2H17. Roche currently distributes Eversense in several countries, and invested $30 million in the company in May. An EASD rep told us that Roche hopes to make its platform device-agnostic in the near future, and that it would even be interested in taking on population-management and risk-share approaches – now that’s a new look! Roche was also chosen to be one of nine participants in the FDA’s digital health PreCert pilot program, demonstrating both commitment on Roche’s end to prioritizing technology, as well as FDA’s recognition of Roche as a serious player in the space.
  • Meanwhile, Ascensia has focused on integrating the Bluetooth-enabled Contour Next One and Plus One BGM systems with attractive digital and device offerings, including Voluntis’ Insulia basal titration app, Insulet’s upcoming Omnipod Dash PDM, Dexcom’s Medicare bundle, and Glooko’s diabetes data management platform. (A big loss in 2017 was with partner Medtronic, who choose Roche as its partner for future Bluetooth-enabled pumps.) Ascensia is also hosting a global innovation competition, crowdsourcing digital solutions that improve the lives of people with type 2 diabetes.
  • It is hard to know what will happen with J&J’s LifeScan portfolio, though a sale or exit in 2018 would not be too surprising. Animas closed in October and J&J maintained it was evaluating strategic options for LifeScan and Calibra (maker of OneTouch Via). No additional detail was provided in 3Q17. That said, J&J wrote in an October press release that “The OneTouch Reveal mobile app is the central component of Diabetes Care from the OneTouch brand.” It is possible to read this as a sign that J&J is still committed to LifeScan, or that it is publicly sharing the value of the brand before a sale. We think the OneTouch Verio Flex Bluetooth-enabled BGM’s WellDoc BlueStar integration is particularly promising.
• Our worst fears were confirmed by the results of The Diabetes Technology Society's BGM Surveillance Program, which identified only six of the 18 tested BGMs passed the accuracy bar. The cleared devices included: (i) Ascensia’s Contour Next; (ii) Roche’s Accu-Chek Avivia Plus; (iii) Arkray’s Walmart ReliOn Confirm/Micro; (iv) AgaMatrix’s CVS Advanced; (v) Abbott’s FreeStyle Lite; and (vi) Roche’s Accu-Chek Smart View. The study, a huge undertaking by DTS, was clearly needed, exposing some highly concerning trends – ~68% of the Medicare mail order BGM market in 2016 was for meters that didn’t pass the accuracy bar in DTS’ evaluation! For now, the DTS is awaiting some response from the FDA before repeating the study - the study obviously isn’t worth repeating if the Agency can’t do anything about the results. We hope to see the FDA or at least payers (especially Medicare!) take some action to punish inaccurate meters.

Are “Digital Diabetes Clinics” the Future? Remote Coaching/Care companies emphasize real-world outcomes

• While providers commonly cite a lack of evidence as reason to distrust digital health tools, Roche’s mySugr, Livongo, One Drop, Virta, Omada, and others have shared very encouraging data over the past year. Virta had some of the most impressive data of the year at ADA 2017: in an interim analysis of 111 patients with data at one year, A1c dropped a significant 1.3% from a baseline of 7.4% (p<0.0001), and 58% of patients achieved an A1c <6.5% while taking no diabetes medications or metformin only. Wow! Insulin was reduced or halted in 97% of users, and weight was reduced an impressive 14% from baseline, equating to a mean 35 lbs of weight loss (from 255 to 212 lbs; p<0.0001). The company’s “full stack” healthcare vision – combining a low-carb/high-fat diet (to induce nutritional ketosis), connected devices, and tech-enabled remote care with physicians – is one of the strongest examples of a digital diabetes clinic we’ve seen. mySugr, Livongo, Omada, and One Drop have all published observational (often retrospective) data too, with fairly new One Drop leading the charge at over a dozen publications/posters so far. Though the products are all different, they seem to be converging on a combination of connected diabetes devices, an app, in-app education/community, and remote coaches that provide guidance and support. Combined, these companies are still reaching a tiny fraction of the diabetes population – but will we see them scale meaningfully in 2018? Are “digital diabetes clinics” the most likely future of diabetes care?
• We’ve been excited to see more consumer-friendly business model innovation from many companies this year. A few months after being acquired by Roche in June, Roche mySugr launched a direct-to-consumer (DTC) Pro bundle in the US, consisting of unlimited test strips delivery, an Accu-Chek Guide BGM, the mySugr app, and 24/7 CDE access for $39.99/month. Livongo’s DTC offering is $65/month (the steeper price may be explained by higher touch, more hands-on coaching), and One Drop launched “One Drop Plus,” in May, bringing lower-priced subscription BGM strips + coaching offerings to the US and Europe, in addition to its Premium plan offering unlimited strips at $33-$39.95/month. Will we see more of these bundled subscription services in the new year, perhaps with advanced features (insulin dose titration!)? They are certainly more convenient, lower hassle, and often more affordable than traditional approaches. Meanwhile, Virta and Omada’s models rest on selling to employers and getting paid upon delivering outcomes – a very appealing business model for payers. As these companies reach more patients, we’ll be keen to hear how the traditional healthcare system views them – will doctors gladly outsource their patients to players like Virta? Where does the line between medical and lifestyle advice fall, and how will coaches interact with traditional channels of care delivery? In terms of product development:
  
  • One Drop has been busy diversifying, announcing pending CDC recognition of its “Revive” digital diabetes prevention program in October, as well as an exciting multi-part collaboration with Fitbit in November and integration with Amazon Alexa Voice Technology in June. Launch of direct Chrome BGM-to-Apple-Watch communication was expected in 4Q17, though we don’t believe this is available yet.
  • Livongo has also formed key partnerships, collaborating with Voluntis in February to bring the Insulia basal insulin titration app to Livongo members with Type 2 diabetes (no launch that we are aware of), establishing another insulin titration partnership with Glytec in April (no launch yet that we are aware of), and acquiring Diabeto in September to facilitate glucose data capturing from 40+ non-connected BGMs, dramatically expanding its potential market size.
  • Roche mySugr’s bolus calculator feature is under FDA review, as of early October. The company also received reimbursement for its bundle in Germany and launched in the US.
• Amid the potential, there is also some fear and misinformation surrounding digital health apps and companies. One EASD panel of international physicians in particular stood out, with many expressing concern regarding regulation and maintenance. None of the panelists were aware of any regulatory body overseeing digital health solutions, despite the FDA releasing plenty of clear guidance on digital health regulation. We wonder if physician trust is among the biggest barriers to the field – will regulatory approval prove sufficient, or is the idea of an algorithm titrating insulin too obscure or scary for many physicians? We expect some providers will be quite excited by the new tools, while others will remain late adopters. There are lots of questions too, as argued in an excellent recent New York Times piece: If an algorithm causes an adverse event, who would be held responsible – the company that designed the algorithm, or the provider who prescribed it? What if the algorithm is based on machine learning and is impossible to explain? Like any other new therapy or technology, we suspect that continuing to publish positive outcomes will be key in convincing providers of the merits. But clinical inertia is a huge force in the field and something that won’t go away any time soon.

Roundup of Major Diabetes Tech Partnerships/Acquisitions in 2017

• 2017 saw a handful of major partnerships in diabetes tech, plus encouraging progress on deals announced in 2016 and before. Even the companies with the most in-house resources are reaching out to others with different areas of expertise. Underlying the partnership wave is the trend away from simple device manufacturing – the bar is constantly getting higher, since devices ideally need to flow into a connected, plug-and-play, interoperable ecosystem with constantly-changing consumer electronics, new displays, and rising consumer expectations for ease-of-use. Few companies can be world-class on all these fronts, hence the flurry of partnerships. We’ll be watching closely to see how well these partnerships execute and deliver value – thus far, the highest profile deals have yet to launch products. But the ultimate vision (at least from the patient perspective) is an open ecosystem where a spectrum of devices and apps are widely interoperable – just like Bluetooth headphones can work with any device, so too should it be for pumps, smart pens, CGMs, BGMs, diabetes apps, EMRs, etc.
  
  • In the summer, Bigfoot acted quickly to acquire Timesulin (injection dose capture) and partner with Abbott to use the second-gen FreeStyle Libre with continuous communication in its automated insulin delivery and MDI auto-titration systems. The Timesulin deal was a great match for the Bigfoot mission and vision to make insulin delivery far easier, safer, and less burdensome. The Abbott deal was more unexpected (though also a great fit), as Bigfoot had previously inked a development deal with Dexcom and performed its only feasibility trial with the G5. On Abbott’s 2Q17 call, CEO Mr. Miles White had very positive things to say about the Bigfoot team, product, and business. Notably, the partnership is exclusive on the Bigfoot side (Bigfoot can only use FreeStyle Libre), but not on the Abbott side (FreeStyle Libre can be used in other AID systems). Now coming off a preliminary Series B raise of $37 million (highly regarded Janus co-led the round), Bigfoot ambitiously expects beginning a US pivotal trial of Loop (pump-based AID) in 2018, followed by a PMA submission at the “end of 2019” and a “2020” launch. Regulatory approval of the Inject system (MDI auto-titration) is expected “in the same timeframe” as that of Loop, placing launch around 2020.
  • Dexcom CGM is currently in use in four of six automated insulin delivery systems in the US pipeline – Tandem, Insulet, Beta Bionics, and Lilly (freshly announced) – all of which made strong progress this year. (For those counting, the obvious exceptions are Medtronic and Bigfoot.) Dexcom’s broad integration strategy positions it well to capitalize on AID, though it will need partners to execute – something Animas could not do.
  • The Dexcom-Verily CGM partnership is moving along nicely, as the first-gen sensor (with a slightly flatter on-body profile displayed at DTM) will complete development in 1H18. Launch timing depends on the G6 no calibration regulatory strategy – depending what happens with the G6, we could imagine gen-one Verily launching in 2019. (2018 would be quite ambitious, but technically possible if G6 no-cal goes very fast at FDA.) As a reminder, the product will move to a fully-disposable transmitter and have either 14-day or 10-day life (Dexcom’s call), along with factory calibration. Efforts on the smaller, second-gen, bandage-like, lower cost sensor are now accelerating – it has previously been slated for a 2020-2021 launch. It’s remarkable to see how far Dexcom has come in 11 years – diaTribe’s Test Drive of the STS in 2006 provides a great zoom out: “The product is clearly valuable, specifically the trending data that allow us to catch extreme highs or lows; but we don't see it replacing finger sticks for a while. Inaccurate glucose values undermine its reliability, and insurers don't yet cover it, so its high cost makes it impractical for all but the wealthiest.” It’s taken a decade, but we’re now finally seeing the paradigm of CGM truly changing.
  • In one of the more unexpected deals this year, Roche acquired mySugr to the (rumored) tune of ~$75-$100 million. We saw this as a victory for both companies – Roche gains a strong foothold into digital diabetes care (mySugr has >1 million registered users), while mySugr retains its autonomy as an open platform and no longer has to worry about raising a large financing round. mySugr also now has Roche’s payer and distribution influence, which could prove critical in achieving further reimbursement for its new bundled offering – Accu-Chek BGM, coaching, unlimited strips, app, and population management. The bundle is reimbursed by large insurers in Germany and is direct-to-consumer in the US.
  • Roche also deepened its relationship with Senseonics, showing off the implantable Eversense CGM in conference exhibit halls, investing $30 million in the company, expanding the distribution partnership, and co-developing an automated insulin delivery system with a Roche pump and TypeZero algorithm. The latter will be evaluated in an arm of the NIH-funded iDCL, which is slated to wrap in 2018 and support CE marking. Roche has invested heavily in Eversense, which could make a full-out acquisition possible. Senseonics still has much to prove in Europe – where full-year 2017 sales are expected to be ~$6-$7 million – but it could have a big 2018 if things scale as expected and FDA approval comes through. Meanwhile, Roche seems confident, as it has essentially shelved its own Accu-Chek Insight CGM in favor of Eversense.
  • Up until the announcement of the pilot with Blue Cross Blue Shield (BCBS), Onduo – the $496 million joint venture between Verily and Sanofi announced ~15 months ago – had been under the radar. The BCBS pilot seems to be just the tip of the iceberg, as soon-to-follow announcements shared that Telcare (cellular-enabled BGM, now owned by BioTelemetry), Voluntis (Insulia basal insulin titration app), and Glytec (Glucommander provider-facing outpatient insulin dose titration) will integrate into Onduo’s “virtual diabetes clinic” in type 2 diabetes pilots with “US providers and payers” beginning in early 2018. Rather than building things from scratch, Onduo is combining and curating connected devices (BGM and CGM), apps, and lifestyle and medication support through a team of healthcare experts. The strong 12-person team is not yet ready to share more information on the pilots, but the vision to combine offerings into a single, streamlined service for payers seems promising – especially if Onduo can launch it at scale with a strong business model.
  • Medtronic/IBM Watson’s high-profile Sugar.IQ app has been quite delayed, but expectations remain high. US launch is now expected with the long-awaited Guardian Connect standalone CGM by April 2018. We saw the first batch of Sugar.IQ outcomes data from the ongoing “limited learning launch” at ADA (n=81), and then an updated data set at DTM 2017 (n=256). In the latter expanded set, Sugar.IQ’s glycemic insights helped increase time in range by a modest 33 minutes per day. Users spent an additional 2% of the day in range (baseline not given, not statistically significant). Within a week after a low glucose “insight” delivered by Watson, 55% of users had fewer lows, with an overall 2.4% less time spent below target; within a week after a high glucose insight, 54% of users had fewer highs, with an overall 1.7% less time spent above target – we assume the latter were absolute reductions, meaning ~30-minute daily improvements in time spent low/high after insights. (That is very meaningful on the hypoglycemia side.) Participants in the study used the app an impressive 2.1 times per day, indicating it is engaging and early adopters perceived high ROI from the pattern recognition. We can’t wait to see this in patient hands, and wonder how big of a differentiator it will be for Guardian Connect, relative to Dexcom’s G5 and Abbott’s FreeStyle Libre, when it comes to market. We imagine Medtronic will bring out serious marketing/advertising guns for the IBM partnership as soon as it secures FDA approval. 
  • The MiniMed Pro-set Infusion Set with BD’s FlowSmart technology launched in limited fashion last September, but was quickly pulled off the market due to a higher-than-anticipated rate of complaints associated with kinked cannulas and elevated blood glucose. BD has since conducted another trial to learn more about how patients can optimally use the set, and the last update in August called for a launch by September 2018. The infusion set has several nice features relative to others on the market, and we’re looking forward to seeing revamped instructions/training (and product design?) to ensure safer use and a more successful re-launch. All eyes will be on exclusive partner Medtronic to see how it handles this re-launch – Will Medtronic go out with major advertising? How are Medtronic’s margins on the BD set vs. Unomedical? Recent Medtronic quarterly updated have cited market pressure in infusion sets, which could be a threat or opportunity for BD.
  • Glooko and Novo Nordisk went public with a partnership to develop jointly-branded digital tools in January, and launched the first free app (“Cornerstones4Care Powered by Glooko”) in July. Novo Nordisk also has a partnership with IBM Watson, which wasn’t leveraged, to our knowledge, in Cornerstones4Care. The currently-available app enables tracking and some data insights – very similar to Glooko – but will eventually serve as the framework for future jointly-developed tools with Novo Nordisk and Watson. The partnership is moving toward further insight-generation, such as insulin dose titration, which could come in the form of Glooko’s Mobile Insulin Dosing System (MIDS). In a feasibility study of MIDS, type 2s saw an 18 mg/dl drop in mean glucose. MIDS is in a larger 240-patient study with Novo Nordisk’s Tresiba.
  • Adding to its 2016 Medtronic partnership, Fitbit moved further into diabetes by establishing partnerships with UnitedHealthcare (Motion program), One Drop, and Dexcom. We are intrigued by a name-brand wellness/fitness company throwing itself into diabetes, and look forward to seeing how it influences product design, passive data acquisition, consumer-friendly data displays, engaging motivation (“Way to go!”), and social networking. 2018 will see the launch of a Dexcom CGM data app on Fitbit’s Ionic smartwatch, along with a One Drop integration on Ionic. While the smartwatch integration is convenient, we see even bigger potential for other diabetes products/programs that drive behavior change and engagement.
  • Plus, there were many additional partnerships in insulin dose titration, which you can read about here.

Diabetes Data Liberation, DIY, and the Software Ecosystem: Votes of Confidence from Larger Players

• 2017 saw continued progress on a more interoperable, user-friendly, insightful diabetes data/app ecosystem. There were several major news items that exemplified this theme, in our view:
  
  • Dexcom launched its public API for third-party apps to leverage retrospective CGM data. Partners One Drop, Tidepool, Nutrino (food), App Practice (physician monitoring), Rimidi (population management), Achievement (rewards), Ensa (supplement recommendations), and Evidation debuted with the API integrated into their apps. Dexcom users can authorize which apps access their CGM data, enabling patient choice. Dexcom will provide significant support for developers, and third-party developers will have access to quite a bit of information to develop novel retrospective CGM data apps (class I), including glucose data, statistics, device info, and calibrations. These are industry standard “OAuth2” and “RESTful APIs.” The bold move is intended to drive an ecosystem of creative CGM data displays/apps. Dexcom recognizes it cannot (and should not) build all useful CGM software internally. This could serve as a competitive advantage for Dexcom, assuming an app ecosystem takes off and patients/HCPs/payers derive meaningful value from it.
  • JDRF announced an initiative to accelerate “open protocol” automated insulin delivery. This major win for the DIY movement encourages manufacturers to update their devices for seamless connectivity. JDRF will contribute financial resources (more than $1 million) and coordinate with regulators and legal advisors to develop a predictable FDA pathway and frameworks. Insulin pumps are currently the biggest area of opportunity. The ultimate hope is an ecosystem of interoperable pump and CGM devices that allow patients to mix and match different system components, including DIY apps/algorithms they may choose to use. This news shows JDRF’s patient-centricity and desire to drive creative device innovation far more quickly – a terrific win for the field. We hope to hear of many companies applying.
  • The DIY community (e.g., OpenAPS, Loop) had a bigger presence at diabetes conferences this year – the already-impressive user experience of these systems continues to improve! We heard about the communities at ADA 2017 (poster on AutoTune, D-Data Panel, The diaTribe Foundation’s Musings After Hours), AADE 2017 (talks from Dana Lewis and Adam Brown), IDF 2017 (a debate on #WeAreNotWaiting), Verily’s Ground Rounds on Diabetes, Health 2.0, and Friends for Life. Plus, the DIY community was mentioned in many talks throughout the year from Dr. Aaron Kowalski and FDA’s Dr. Courtney Lias, among many others. It seemed like nearly every conference someone would stand up in Q&A and share how much a DIY AID system had helped them! There is no question that this community is adding value to existing diabetes devices, iterating very quickly on user experiences and algorithms, and integrating into consumer smartphones and watches. For now, these systems are reaching only a tiny fraction of the type 1 diabetes population – those willing to go DIY and take some time to set up the system. A big question is how these efforts will evolve and scale in the coming years – How can the DIY community be harnessed to drive outstanding commercial products? How can industry propel DIY efforts? What can the DIY community teach industry? 
  • Apple added insulin delivery doses to Apple Health in iOS 11, paving the way for Bluetooth-enabled pumps and pens to easily share data between apps. The DIY AID app, Loop, is already posting insulin delivery to Health, and we expect to see more products do this in the future. Companion Medical expects to add this functionality in the coming weeks, and we hope Bluetooth-enabled pumps from Tandem, Insulet, and others will follow.
  • The standardized, one-page, ambulatory glucose profile (AGP) report is now in CGM analysis software from Abbott, Dexcom, Glooko/Diasend, and Roche (to display Senseonics data)– a terrific win for standardizing data display. We hope Medtronic follows suit soon.
  • In a sign of what can be learned from a population of connected devices, Tidepool shared the first CGM data from its Big Data Donation Project. Though it was only from 354 individuals, we found the results fascinating for benchmarking average glucose, time-in-range, time-below range, and time-above range for different age groups. We hope to see far more of this data from the CGM and diabetes data players – much could be learned about the current state of diabetes care, Bright Spots (for those doing well, what do they do differently?), user behavior, and device interaction. 

FDA’s CDRH Drives Diabetes Digital Health and Medical Device Innovation With New Guidance And Pre-cert Program

• With Commissioner Dr. Scott Gottlieb, CDRH Director Dr. Jeffrey Shuren, and Digital Health lead Mr. Bakul Patel leading the charge, the FDA has been busy with efforts to better match its regulatory processes with the fast-paced world that is digital health. In July, the FDA announced a new pilot, the Digital Health Software Precertification (PreCert) Program – a “firm-based approach” for expediting product review sort of like a TSA Pre-Check for digital health companies. Currently, the pilot has enrolled nine companies (seven of which have ties to diabetes) who will help shape the program’s requirements. Given the blazing pace of digital health innovation, we’re thrilled to see a more streamlined FDA process in the works and see high potential for faster reviews, more iteration, and smarter use of FDA’s limited resources in diabetes/digital health. CDRH also released an eight-page Digital Health Innovation Plan, outlining five upcoming draft/final guidance documents. One such document, the Clinical and Patient Decision Support Software draft guidance, came out ahead of schedule in December, clarifying the types of clinical and patient decision support software no longer defined as a medical device. From what we can tell, patient-facing insulin dose titration will still be regulated, while provider-facing-only tools are more nuanced (basically, a tool that allows doctors to review the basis for their recommendations are proposed to not be regulated, while those that do not, such as those that process or analyze medical images, will still be under the purview of FDA). A second draft guidance released in December clarified the exclusion of low-risk digital health solutions from regulation, another promising win for innovation, and a third draft guidance released in the same month expanded and clarified “least burdensome” regulatory processes (focusing on the need-to-know rather than the nice-to-know). Throughout this most recent document, the Agency stresses flexibility and efficiency – two exceptional qualities that have characterized the FDA’s Device division and its approach in the past few years.
• The FDA also recently released final guidance on real-world data (RWD), supplying criteria for determining what types of RWD constitute valid scientific evidence worthy of regulatory consideration. The document specifically mentions that FDA has begun implementation of NEST (National Evaluation System for health Technology), which many hope will quickly identify safety issues and better elucidate the benefit-risk profile of devices in the real-world, possibly reducing pre-market requirements. While RCTs certainly have their place, and won’t be replaced by RWD, they take substantial time to complete and the technology in question may be outdated. This has been particularly problematic in CGM and digital health, which move on far faster timelines than robust, randomized clinical trials. Speakers at the Medical Device Innovation Consortium (MDIC) 2017 Annual Forum in October were particularly enthusiastic regarding RWE collection, especially in the post-market setting. Clearly, both RWE and RCTs will have a place ahead, and we applaud FDA’s CDRH division for lighting the way.
• While we were sad to see President Obama’s choice for FDA Commissioner, Dr. Robert Califf, depart the Agency (he’s now at Verily), Dr. Gottlieb has provided energy, efficiency, and innovation. Dr. Gottlieb seems to be pumping out blog posts and guidances at an unprecedented rate, in the interest of more wisely allocating FDA’s limited resources. Among his other initiatives discussed above, are the allowance of computer modeling and simulations; the use of historical control groups, observational/registry studies, the subject as his/her own control, and adaptive study designs; and leveraging data and decisions from other countries’ regulatory bodies. We heard him speak at the 2017 Milken Institute Future of Health Summit discussing the influence of his business background on his regulatory philosophy and innovation – you can watch the entire conversation here. FDA, and CDRH in particular, seem to be in extremely capable hands.
• We couldn’t write an FDA highlight without acknowledging the work of Drs. Courtney Lias, Stayce Beck, and the FDA Artificial Pancreas Team for taking home a 2017 Samuel J. Heyman Service to America Medal (“Sammie”) for the faster-than-expected three-month review/approval of Medtronic’s MiniMed 670G. The amount of work and listening Drs. Lias and Beck and their teams continue to do on patients’ behalf is ever-deserving of recognition. This year, Dr. Lias continued to stress that interoperability will drive innovation, delivering enthusiastic talks on the matter at both Keystone and DTM. In the latter, she hinted that FDA is working on “smart solutions” for seamless device interoperability (“low hanging fruit”), and discussions are ongoing with stakeholders – including JDRF and the Helmsley Charitable Trust. Dr. Lias said FDA will “hopefully be announcing some public opportunities for discussions on that front. “We’re trying to remove barriers that don’t actually help.” Nice!

Next-Gen Tech – Excitement for Machine Learning and Voice Technology

• Examples of machine learning (aka artificial intelligence or AI) applications in diabetes are starting to add up, and we’re looking forward to seeing them roll out to clinics, health systems, investigators, and patients. AI can be very powerful in diabetes and healthcare in general, especially for screening for retinal damage, recognizing patterns in glucose and insulin, identifying at-risk patients from the EHR, and more. This year…
  
  • We saw positive preliminary data from Medtronic/IBM Watson’s Sugar.IQ pattern recognition app, which is expected to launch by April 2018. See the section above for more detailed coverage.
  • Sanofi and IBM Watson Health presented posters at IDF demonstrating the power of real world data and advanced analytics to answer meaningful clinical questions – What are the main predictors of success with insulin+GLP-1 combination therapy? What are the chances of basal insulin success after 12 months?
  • Verily provided updates on its automated diabetic retinopathy detection, sharing that there are ongoing trials looking at performance and clinic workflow in India.
  • At DTM, a Bigfoot poster and subsequent presentation shared how the company’s automated insulin delivery system leverages machine learning to optimize and personalize insulin delivery.
  • JDRF and IBM Watson teamed up to identify factors contributing to type 1 diabetes risk and onset.
  • The Helmsley Charitable Trust funded a machine-learning predictive analytics project (software by Cyft, Inc.) for type 1 diabetes at Children’s Mercy Kansas City and Joslin Diabetes Center.
  • At the recent New York HITLAB symposium, interim ADA CEO Ms. Martha Clark shared that the Association’s work with IBM Watson is chiefly concerned with mining EHRs for real-world evidence on the efficacy of various drugs/devices. Nice!
  • Plus, we assume many companies are using machine learning under the hood to power their software and services. Apple’s new iOS 11 includes a machine learning module for developers (Core ML), which should make it easier for more apps to leverage this technology. 
• There was more buzz than ever about voice technology (e.g., Amazon Alexa, Google Home) for diabetes management. How promising is it really? One Drop announced an integration with Alexa in June; Our report on the Merck/Amazon Diabetes Challenge (voice solutions for type 2 diabetes management) was one of the highest read pieces of the year; and Novo Nordisk and HITLAB teamed up for the 2017 HITLAB World Cup of Voice-Activated Technology in Diabetes. On one hand, voice could become a powerful health interface, reducing burden, improving patient-provider relationship, enabling better in-the-moment-logging, etc., but the application in diabetes is not yet well-defined or tested. Where is voice going to add significant, differentiated, life-improving value in diabetes? Will it be able to accurately answer real questions/tasks people are going to ask, like “What should I eat for breakfast?” or “How much insulin should I take?” or “What medications am I supposed to take today?” or “Can you reorder my prescriptions?” (though this is admittedly more an issue of underlying software and machine learning, not the viability of using voice for user experience). How well will voice interface with existing medical devices? Where will the regulatory line be drawn as voice tools drift into medical data and therapeutic advice? As we understand it, voice is not HIPAA compliant at the moment, but we hope this barrier will fall.
• There are also growing concerns about the black box nature of AI and algorithms’ ability to exacerbate biases – see this NYT article and the book, Weapons of Math Destruction for more on these topics. 

Most Highly Read Reports

• Abbott’s FreeStyle Libre hits major US pharmacies with ambitious pricing; cash-pay at CVS, Walgreens: ~$43-$53 per sensor, $85-$97 per reader – Nuanced pricing analysis between Abbott and Dexcom – WAC vs. net price, Dexcom out-of-pocket with our Aetna insurance vs. Abbott cash price; our questions for US CGM category
• Abbott and Bigfoot partner to include next-gen FreeStyle Libre in Bigfoot’s automated insulin pump, MDI titration systems; US pivotal in 2018; next-gen Libre to keep factory calibration and add real-time communication – Automates insulin delivery without fingersticks –major win for Abbott and Bigfoot; Abbott is exclusive US sensor partner; next-gen Libre retains same form factor (picture enclosed); Abbott provides Bigfoot sensors, same monthly service model; interview with Jeffrey Brewer
• Lilly developing smart pen with insulin titration and AID with its own pump and Dexcom CGM; launch expected in ~2-3 years, trials to start by end of year – Lilly has own disk-sized pump (designed by Segway inventor Dean Kamen); dose capture from disposable pens, app with dose titration (bolus, basal); Dexcom “development agreement”; interview with Dr. Howard Wolpert, Marie Schiller
• UnitedHealth and Dexcom’s unprecedented pilot: CGM and coaching in 10,000 type 2s! Ramping over next 6-9 months; could save “billions” – Wow! BIG win for CGM in type 2; Coaching and clinician house calls; the biggest US Private payer win for CGM in T2D?; transcripts from UnitedHealth’ Dr. Deneen Vojta and Dexcom’s Steve Pacelli; our 30 questions
• Novo Nordisk and Glooko partner to develop jointly-branded digital tools; initial launch this year; smart pen to pilot in EU in 2017! – Insulin dose titration is a major goal; IBM Watson to work with Glooko, NN for subsequent products; Glooko has 1 million users! NN clear commitment to digital health; interview with both companies
• Dexcom launches public API for third-party apps to leverage retrospective CGM data; major win for driving diabetes app ecosystem – Data partners One Drop, Tidepool, Nutrino, App Practice, Rimidi, Achievement, Ensa, Evidation already using API; major internal investment, commitment to patient choice, app ecosystem; pictures and questions
• Medtronic issues urgent field safety notification for MiniMed Pro-set; approximately 500 customers in initial limited launch – Higher than expected complaints related to kinked cannulas, elevated BG levels; “majority” of customers pleased; unclear if full launch timing for ~Jan-Mar impacted
• Apple to add insulin delivery doses to HealthKit in iOS 11 this fall – Screenshot enclosed; watch developer video here (43:28 mark); mentioned at ADA from Companion Medical’s Sean Saint; important step for data/app interoperability as Bluetooth pens + pumps come out
• Beta Bionics partners with Novo Nordisk, receives $5 million investment (complementing Lilly!); pivotal trial timelines slip slightly; two R&D team additions – Lilly and Novo Nordisk now both have seats on the BoD – unprecedented!; Insulin-only pivotal pushed to late 2017/early 2018; two new additions to R&D, forming “Beta Bionics West”; our questions
• J&J’s Animas closes, exiting insulin pump market; Medtronic chosen as preferred partner for transition; sad news reflects fragile, competitive pump market and high innovation bar – ~90,000 Animas pumpers worldwide; free MiniMed 630G transition for Animas warranties expiring on/after September 30, 2019; Implications for Medtronic, Insulet, Tandem? Reminder of tough business model, cost structure in pumps

What We Got Right/Expected

Glucose Monitoring

• FreeStyle Libre uptake accelerating globally, obtaining a non-adjunctive label in the US, and coming to the US with strong cash pricing. Though some were surprised by the label, we always felt it was unlikely the FDA would approve a factory-calibrated sensor with an adjunctive label.
• Dexcom obtaining Medicare coverage following the FDA non-adjunctive approval last December. The rollout took some time and required a deal with Ascensia to bundle a BGM, but the FDA approval was a big gating factor to this milestone. We hope to see Dexcom get remote monitoring approved in some capacity for the Medicare population in 2018.
• Momentum for outcomes beyond A1c movement, more discussion at scientific meetings. Though 2016 definitely made progress on Beyond A1c, this year felt like a turning point – beyond A1c discussion took place at ATTD, ADA, AADE, EASD, IDF, and the July consensus conference in Washington – wow!
• CGM publication and reimbursement victories – Dexcom (DiaMonD, GOLD), FreeStyle Libre (UK NHS, Japan, France), CONCEPTT, Belgium data, etc. More accuracy, more connectivity, and better insight are starting to drive better, now-published outcomes.
• Dexcom’s continued open approach to diabetes data, with launch of public API for retrospective data and more partnerships.
• Slow sales in traditional SMBG market, but a drive to reinvent themselves in a more digital, CGM-driven era.
• Apple not launching non-invasive glucose monitoring, despite the rumors it would do so. However, a New York Times piece published this week suggests Apple still has a team working on it.
• Very gradual EU sales ramp for Senseonics’ Eversense CGM
• No movement on Intuity Medical’s POGO automatic BGM, despite raising $55 million in Series 3 funding to launch this year.

Insulin Delivery (AID, Pumps, Smart Pens, Devices)

• A lot more enthusiasm for smart pens/caps, injection dose capture, and titration, especially from the major insulin companies. Companion Medical launch InPen to close out the year, a first-in-class product in the US – and hopefully the first of many to come in this category.
• Challenging and highly competitive US pump market, with a bright spot in Insulet’s Omnipod patch pump.
• Continued fundraising momentum and bold moves from Bigfoot Biomedical – raising $37 million in Series B funding, acquiring Timesulin, signing with Abbott to use FreeStyle Libre in its two systems.
• Novo Nordisk and Sanofi continuing to take a partnership approach to digital health (e.g., Glooko/IBM Watson; Onduo/Voluntis/Glytec).
• Insulet and Ypsomed terminating their agreement for Omnipod distribution in Europe – it felt unlikely this would continue, given Ypsomed’s own pump and the profitability advantages of Insulet taking over distribution.
• Slow uptake of Cellnovo’s pump, continued delays in manufacturing and US clearance

Data, Software, and Digital Health

• Exploding partnerships for insulin titration – it’s a no-brainer that algorithms will titrate insulin far better than humans!
• Strong FDA moves to minimize burden, maximize speed to market
• Fitbit moving further into diabetes and healthcare, as consumer sales slow and it realizes it can make a meaningful difference in our field
• More focus on diabetes coaching and strip subscription models (One Drop, mySugr, Livongo), though scalability remains a question mark.

What We Got Wrong/Did Not Expect

Glucose Monitoring

• Medicare’s strong ~$8/day pricing ($250/month) for Dexcom’s G5. We also didn’t expect Medicate to cover CGM for BOTH type 1 AND type 2 CGM users on intensive insulin therapy; Medicare could have restricted it to type 1.
• Bigfoot partnering to use a next-gen Abbott FreeStyle Libre CGM, leaving initial partner Dexcom. Though Dexcom and Abbott were in negotiations for a long time, we guessed they would come to a commercial agreement.
• FreeStyle Libre’s shorter 10-day wear and longer 12-hour warmup in the US – while the accuracy rationale makes sense, we did not see these fairly dramatic product changes coming.
• Apple’s June event discussing direct Apple Watch-to-Dexcom transmitter CGM communication, no phone needed – we would not have guessed Apple would devote so much airtime to its work with Dexcom.
• Roche acquiring mySugr and moving deeper into digital health – Roche is clearly far more committed than just SMBG.
• Medtronic signing an exclusive deal with Roche for future Bluetooth-enabled BGMs+pumps, leaving long-time partner Ascensia
• Livongo’s continued fundraising success in a very tough, price-sensitive BGM market – can it continue in 2018?

Insulin Delivery (AID, Pumps, Smart Pens, Devices)

• J&J Animas completely exiting the insulin pump business – while the Animas business was moving slowly on innovation, this was a solvable problem. We did not expect a complete exit from J&J, given the user base of ~90,000 customers and potential to improve things.
• Lilly jumping headfirst into diabetes tech and digital health, planning to commercialize its own hybrid closed loop (with its own pump!) and smart pen/titration systems with Dexcom CGM.
• Medtronic’s US struggles in 2Q17-3Q17 (e.g., US sales, infusion set recall, reduced guidance), despite the first-to-market MiniMed 670G in the US.
• JDRF’s Open Protocol Automated Insulin Delivery initiative – this unexpected (and welcome!) vote of confidence could fuel DIY efforts with help from the major non-profit.
• BD already pushing back the launch timing on its patch pump and smart pens – we would have guessed the device manufacturing juggernaut had things well lined up following the announcement last fall.
• Beta Bionics securing an investment from Novo Nordisk, complementing Lilly. Like some of its competitors, Beta Bionics did push back its pivotal timelines this year, so we’ll have to see how execution goes in 2018.
• Valeritas successfully going public, though it has not been able to scale the product.

Data, Software, and Digital Health

• At least three basal-only insulin titration products cleared, but true availability seems to be limited to pilots/feasibility trials – will this become a thriving commercial market?  
• Amalgam Rx launching its own basal insulin titration app, founded by former WellDoc execs Ryan Sysko and Dr. Suzanne Clough.
• Apple adding insulin delivery tracking to Health in iOS 11 – while this was an obvious thing to do (on top of glucose), we’re not sure we would have guessed it was coming at the beginning of the year.

Other

• JDRF launching the T1D Fund – new and improved venture philanthropy approach, taking lessons from the PureTech fund that did not pan out

Key Questions for 2018

Glucose Monitoring

• What will the US CGM market look like in 2018? Will Abbott’s FreeStyle Libre help build the category? What kind of commercial success and reimbursement will Libre see in the US?
• Will Dexcom launch a no-calibration G6? Will Dexcom and Verily launch their sensor in 2018?
• Will Senseonics’ Eversense and Medtronic’s Guardian Connect/Sugar.IQ get approved in the US?
• What will CGM adoption look like at the end of 2018, in the US and Europe? Will we see any CGM reimbursement victories in type 2 diabetes?
• Can Medtronic fix its sensor manufacturing shortages? How much is this truly gating sales and 670G adoption? Will Sugar.IQ with IBM Watson be a valuable differentiator for Medtronic’s Guardian Connect CGM? Or will the hardware/calibration limitations of Guardian limit adoption relative to Dexcom and Abbott?
• What will happen to J&J’s LifeScan business? Will it sell the division to a large player? Will it shut it down, like Animas?
• Who will have more digital BGM reinvention success in 2018 – Roche or Ascensia?
• Will we hear more about some of the exciting type 2 diabetes pilots, including Dexcom/UHC’s in 10,000 type 2s and Onduo’s work with connected CGM and coaching?
• Will Dexcom find a way for Medicare beneficiaries to use the G5 app and remote monitoring? What solutions are in the works? Will Abbott make progress on Medicare coverage for FreeStyle Libre?
• Will direct smartwatch communication with CGM and BGM be compelling? How much will this benefit Apple Watch and Fitbit, and how much will it benefit Dexcom and One Drop? Will these integration motivate more data engagement and behavior change?
• Will any CGMs seek support from JDRF’s open protocol initiative? Which upcoming systems will be the most interoperable?
• Will we see any novel CGMs make progress in 2018? Will AgaMatrix get a CE Mark for its CGM in 2018, as it expects? Will we hear any updates on the Verily/Novartis glucose monitoring contact lens?
• What can big data teach us in CGM? What health systems and centers are using CGM well? What can we learn? What lessons can we learn from the top 100 prescribers of professional CGM?
• Why are CGM endpoints not accepted by FDA’s CDER division? What could change this?
• Will subscription BGM approaches with coaching see more adoption in 2018? Who will see the most success – Livongo, One Drop, mySugr, someone else?
• Will DTS’ BGM surveillance program drive FDA to take action against poorly performing meters? What about payers? Will it influence patient or HCP choice?

Insulin Delivery (AID, Pumps, Smart Pens, Devices)

• Will the MiniMed 670G launch hit full scale in 2018? Will Medtronic hit guidance for double-digit growth next year? What will the international response be to the 670G? How many could be using the 670G by the end of 2018? Will any follow-on products in 2018 improve the 670G experience?
• Will Tandem’s t:slim X2 with PLGS launch next summer? Will it successfully compete with the 670G? Can Tandem raise more cash in 2018? Will the iDCL pivotal study actually start on time in 1H18?
• How will Insulet users receive Omnipod Dash? Will it drive even stronger US growth for Insulet? Will it compete with 670G and Tandem’s PLGS?
• Will the US pump market return to growth in 2018, following a slow 2017?
• Will Medtronic/BD’s MiniMed Pro-set relaunch in 2018? Will it see strong uptake relative to existing Unomedical offerings? How will infusion set pricing pressures impact the field?
• Will Unomedical launch its competitive all-in-one hidden needle insertion device (see ATTD 2017)?
• Will Bigfoot successfully enter its pivotal trial for Loop in 2018 with Abbott’s second-gen FreeStyle Libre and its own pump? Will we see progress on the Inject platform in 2018?
• Will we see data from Lilly’s first feasibility studies of AID and injection titration?
• Will any insulin pumps apply for JDRF’s exciting open protocol initiative? We think this would prove to be a huge competitive advantage, assuming the regulatory path can be worked out.
• Will we see FDA progress on interoperability, as Dr. Courtney Lias has been promising?
• Will type 2 insulin delivery devices see any meaningful adoption or launches in 2018? What will happen with Valeritas? Will BD share more about its type 2 patch pump? Will CeQur file with FDA? What will happen with J&J’s One Touch Via (Finesse)?
• What will uptake of Companion Medical’s In Pen look like in 2018? Will any other smart pens/caps launch next year outside of pilots?
• What will smart pens teach us about insulin delivery in the real world? What can we learn for populations? How much better and safer will smart pens + titration make insulin delivery?
• Will any of the four FDA-cleared insulin titration apps see wider adoption in 2018 – Voluntis’ Insulia (basal-only), Amalgam’s iSage Rx (basal-only), Lilly’s Go Dose (Humalog only), and Sanofi’s My Dose Coach (basal-only)?
• How will automated insulin delivery with a pump compare to smart pens with dose titration? Which system will deliver the best outcomes/value? How will the market segment? How will companies like Bigfoot and Lilly juggle the two offerings in payer conversations?
• Which of the insulin companies will make the most digital progress in 2018? How will insulin companies keep up with the fast pace of technology development and iteration? How will pharma companies sell their tech-enabled offerings? Will they exclusively bundle the whole package, including insulin?
• For pharma, will technology (delivery device, dose tracking, and titration) become the key battleground for future innovation, rather than the molecule itself? How will companies split R&D between the two fields, particularly as technology becomes more important? Which of the three major insulin companies will be the most successful?
• What’s the right business model for automated insulin delivery, smart pens, and insulin titration? There is a spectrum of approaches we’ve heard about: payment for a standalone device (e.g., Companion Medical’s InPen, MiniMed 670G), monthly or yearly fees per user (Common Sensing, Voluntis), bundled hardware/software service offerings (Bigfoot Biomedical), and beyond. Risk sharing/outcomes-based pricing is also very likely in this field, and paired with it, hard decisions about which outcomes should be tracked and used as the basis of payment. There are a variety of audiences that could conceivably pay for smart pens and insulin titration, including payers (to get better outcomes for their patients), providers (clinical decision support, population monitoring), and even patients themselves. Business models are a critical question for everyone in this field, since commercializing these enhanced insulin delivery products will require additional R&D resources, education, marketing beyond what is already happening. Along with it, we expect the value will be much higher.
• What’s the ideal insulin dose capture product form factor? Reusable smart pen vs. smart cap? Most seem to be pursuing solutions that add-on to disposable pens, which has the clear advantage of fitting into the existing disposable insulin pen infrastructure and perhaps being lower cost (e.g., Bigfoot, BD, Common Sensing). The advantage to reusable pens, however, might be a more integrated product experience that one company can control (Companion Medical, Novo Nordisk from what we can tell, Lilly based on this study page); the downside is hassle on the user’s end to obtain separate insulin cartridges. Since both classes of dose capture have pros and cons, it’s possible we’ll see a variety of different form factors that appeal to different users.
• What will most differentiate different automated insulin delivery systems, either pump or injection based? Will it be the algorithms? Major differences in patient outcomes? Patient user experience? Ease of integration with HCPs’ clinical workflow? Ease of prescribing? Payment models? Something else?
• Provider- vs. patient-facing insulin injection dose titration? Most seem to pursuing a hybrid model, where an HCP sets up the titration parameters, and then the app runs on its own giving patient-facing recommendations. From a regulatory perspective, it should be easier to get fully provider-facing clinical decision support cleared/approved (Glytec). That said, HCP-facing approaches are likely to be less scalable than more patient-facing designs. Which approach will drive the greatest improvement in outcomes? Which will most extend the reach of HCPs?
• What’s the role of software vs. humans? Some companies are combining insulin titration software with human coaching as a combined intervention (e.g., Onduo, Hygieia), while others are relying more on the software alone with some HCP input upfront (e.g., Voluntis, Amalgam, Bigfoot from what we can tell). Will the addition of coaching dramatically improve outcomes and adherence to recommended doses? Will coaching prove scalable? Will companies be able to segment patients into software-only or software+coaching?
• Where are outcomes trials needed? When will we see more studies comparing insulin titration software (with or without smart pens/caps) to standard of care insulin dosing? Will anyone run a head-to-head trial comparing automated insulin delivery with a pump vs. smart pens with titration software? What will the difference in outcomes look like? How will companies with both products – e.g., Bigfoot, Lilly – price the two options? Will patients switch between these options?

Data, Software, and Digital Health

• Will we see 2018 progress on time-in-range goals and benchmarks? What data could help validate CGM endpoints for regulators?
• Will “digital diabetes clinics” and coaching/remote monitoring see more momentum in 2018, including more data and payer pilots? Between Roche’s mySugr, Livongo, One Drop, and Virta, who will make the most progress next year? Who will reach the widest scale?
• Will Fitbit return to stronger growth in 2018, buoyed by its diabetes technology partnerships and a greater move into healthcare? Will it move into diabetes behavior change programs?
• Will Dexcom’s public API see more compelling app integrations in 2018? How many patients are currently using the API, and how many more could we see in a year with stronger value-added apps?
• What else might Verily, Apple, and Fitbit do in diabetes technology? What about Amazon, Facebook, and Microsoft?
• Will voice and machine learning see a compelling diabetes application in 2018 – perhaps Sugar.IQ on the AI/machine learning front?
• Will IBM Watson be more public about its diabetes work in 2018? How will it compete with Verily’s machine learning efforts?
• Will we see more acquisitions like Roche/mySugr? Or will diabetes tech companies continue to mostly rely on partnerships? Nimble digital health companies can provide a source of innovation, experimentation, and human factors expertise to larger organizations, while providing them with resource security and a broader platform. Thus far, however, acquisitions are rare in diabetes tech – most are going the partnership route, presumably because it is faster and less expensive.
• Will the FDA PreCert program finalize criteria in 2018? What companies will it most help? Outside of the pilot participants, how hard will it be to get pre-certification?

What’s Coming in 2018

Below, we enclose the pipeline products we are aware of, though we acknowledge this list is likely not 100% complete. We have done our best to be as comprehensive as possible based on the most recent public updates we’ve heard.  

Automated Insulin Delivery, Pumps, Delivery Devices, Pens and Caps

BD

• MiniMed Pro-set with BD FlowSmart re-launch could occur by September, per the last update in August
• Type 2 basal-bolus patch pump could launch by the end of the year, though it’s technically slated for the fiscal year running from October 2018-September 2019
• Smart pen needle technology could also launch by the end of the year, though it’s technically slated for the fiscal year running from October 2018-September 2019

Beta Bionics

• Phase 2b study of Zealand dasiglucagon (liquid stable glucagon analog) in iLet integrated device expected in early 2018

Bigfoot Biomedical

• Pivotal trial of Loop automated insulin delivery system to start in 2018 with Abbott’s FreeStyle Libre. We assume this would come toward the end of the year, as PMA submission is expected at the “end of 2019,” followed by a “2020” launch.
• Inject MDI auto-titration system pivotal could start in 2018, per VP John Sjolund’s DiabetesMine remarks. However, CEO Jeffrey Brewer said the regulatory timeline is still in development and the system will undergo at least one feasibility study before going to a pivotal trial.

Biocorp

• 2018 EU launch with partners following late 2017 CE marking of its Easylog smart pen attachment.

Cellnovo

• FDA clearance of patch pump system expected in 2018 – initially submitted in November 2016, but FDA requested additional information on the submission.
• Possible pilot AID study in limited number of patients with TypeZero algorithm. A launch was previously slated for 2018, but we could see this being revised.
• Possible 2018 launch in Europe in Diabeloop’s closed loop system.

Common Sensing

• Ongoing Joslin GoCap study, which will run through August 2018
• Pilot underway in Northern Virginia with Sanofi, Innovation Health (Aetna+Inova Health), and One Drop
• Plan to add compatibility with four more insulins (only works with Lantus and Apidra at the moment)

Companion Medical

• Broader retail pharmacy launch for Bluetooth-enabled InPen in early 2018
• CE Mark expected in 2Q18
• Android app expected in 2Q18
• Will post insulin delivery data to Apple Health in the coming weeks.

CeQur

• Timing unknown for the PAQ patch insulin delivery device. The device has yet to be filed with FDA, as far as we know.

Diabeloop

• EU launch of automated insulin delivery system (with Cellnovo pump, Dexcom CGM, algorithm on a wireless handheld) expected in early 2018. The ongoing CE mark pivotal trial is expected to complete in February/March 2018 per ClinicalTrials.gov.

Insulet

• Limited market release of Bluetooth-enabled pod and touchscreen PDM, OmniPod Dash, expected in mid-2018. FDA submission slated for around the end of 2017, per 3Q17 update.
• Possible 2018 pre-pivotal study of Horizon automated glucose control system. As of early November, the system was in the midst of a five-day hotel study (IDE3) – see interim data from DTM – a pre-pivotal study will be next, followed by the pivotal study. Launch is slated for “end of 2019” or “early 2020.”
• Data from phase 3 study of Lilly U-500 OmniPod in type 2 patients to possibly report at ADA 2018
• Will assume Europe Omnipod distribution in July 2018

LifeScan OneTouch (J&J)

• Plans for OneTouch Via (formerly Calibra Finesse) bolus patch delivery device unknown. FDA clearance for a new manufacturing process was achieved in June and reps told us in May that a focused US launch was expected “in the coming months,” but it’s hard to know if, how, and when LifeScan will launch the product, J&J could sell this asset along with the LifeScan business, though there have not been recent updates following Animas’ exit from the market.

Lilly

• Feasibility study of AID system (with proprietary DEKA-designed pump, Dexcom CGM, and Class AP algorithm) underway, to wrap up in April 2018
• “Assessment of Mealtime Bolus Insulin Behavior” study with Lilly reusable smart pen and CGM (presumably Dexcom) recruiting, set to complete in June 2018
• Other studies with AID and “Integrated Insulin Management” system. Both set to launch in next ~two-to-three years.

Medtronic

• Expects to resolve sensor constraints in February-April, after which remaining 670G Priority Access Program participants can get on hybrid closed loop and broader launch can presumably commence.
• 670G international launch expected by April 2018 (countries not specified).
• 670G pediatric study to complete in April 2018, per ClinicalTrials.gov.
• 670G 1,000 patient outcomes RCT to continue; study has begun randomization.
• NIH-funded trial of 670G vs. 670G+DreaMed algorithm previously slated to start on December 1, 2017.
• MiniMed Pro-set with BD FlowSmart could re-launch by September, per the last update in August

Novo Nordisk

• No firm timing shared on smart pens or the partnerships with Glooko or IBM Watson. It’s possible we could see a broader launch of durable smart pens (NovoPen Echo 5 currently piloting in Sweden), updates to the Cornerstones4Care app with Glooko, and/or updates on the IBM Watson partnership.

Roche

• Pivotal trial of long-term AID system with Accu-Chek Insight pump, 180-day implantable Senseonics Eversense CGM, and TypeZero inControl AP algorithm to enroll in early 2018 and complete by late 2018. This is part of the NIH-funded iDCL trial and will support CE mark submission.

Tandem

• PLGS pivotal study (PROLOG) previously expected to complete by end of 2017, followed by launch in summer 2018. Customers with t:slim X2 G5 will receive a free remote upgrade to this system.
• Pivotal trial for second-gen AP product (TypeZero treat-to-target algorithm) to start in 1H18

Unomedical

• Possible 2018 launch of all-in-one, fully disposable, hidden needle insertion set device? Per ATTD 2017, launch was expected “after” summer 2017, first coming to market in an exclusive partnership with a Luer Lock pump. We’re not sure who that would be, given Animas’ exit, Tandem’s switch to t:lock, and Roche’s US exit. It’s possible it could come to market with Tandem in the US or with Roche OUS.
• Possible launch of novel catheter (Lantern) in 2018? Per ATTD 2017, launch was expected in 3Q17. To compete with BD’s FlowSmart, Lantern includes several slits along the side that allow insulin to flow out of multiple places in the case of occlusion or bending.

Valeritas

• 40% larger sales force starting 2018 vs. 1Q17
• No firm timing shared on V-Go Link and/or pre-filled V-Go, but these could theoretically progress in 2018.
• Deals with distributors overseas?

Ypsomed

• Termination of Omnipod distribution in mid-2018.
• YpsoPump expansion into 10 additional countries (possibly up to 20) by end of 2018.
• Ypsomed Orbit infusion set to be included in 2018 Bigfoot Loop pivotal trial

Glucose Monitoring

Abbott

• FreeStyle Libre reimbursement details from US payers? The device launched in pharmacies earlier this month, but is cash pay to start. Abbott CEO Miles White said there is potential for $50-$100 million in US FreeStyle Libre revenue in 2018.
• US FreeStyle Libre label updates? – pediatric use? LibreLink? 14-day wear? Shorter warmup?
• Medicare coverage progress, given therapeutic CGM designation?
• Use in Bigfoot Loop and Inject trials (see above)

AgaMatrix / WaveForm

• CE Mark expected for its own CGM, per October announcement
• Bluetooth-enabled BGM distributed with Companion Medical InPen

Ascensia

• Integration of Voluntis Insulia basal titration previously expected in 4Q17; possibly pushed back to 2018

DarioHealth

• Partners to deliver “fully-connected and real-time solution combined with coaching”?
• Scale a B2B business?

Dexcom

• G6 Launch expected in 2018, with one calibration per day and/or factory calibration – both currently under discussion with FDA.
• Direct Apple Watch-to-transmitter communication (“Native Core Bluetooth”) currently under FDA review
• Fitbit Ionic smartwatch integration in 2018
• Gen one Verily-partnered sensor to complete development in 1H18. Launch timing depends on G6 pathway, but could conceivably launch in 2018.
• Used in AID studies /products with: Tandem, Lilly, Insulet, Beta Bionics, Diabeloop.

GlySens

• Last year, we were told a CE approval trial could possibly start in 1Q17 and contributed to US pivotal evaluation. We haven’t heard an update since and wouldn’t be surprised if this timing has been pushed back.

Intuity Medical

• The FDA cleared POGO automatic BGM was expected to launch in 2017, but didn’t, to our knowledge. Is a 2018 launch in the cards?

LifeScan (J&J)

• The WellDoc BlueStar-OneTouch Verio Flex BGM integration is technically complete, but has not rolled out. Will this occur in 2018, or might LifeScan be looking for a buyer for its BGM business?

Medtronic

• Guardian Connect standalone mobile CGM and IBM Watson-partnered Sugar.IQ to launch by April 2018
• iPro3 professional CGM to launch in FY19 (May 2018-April 2019)

Roche

• Accu-Chek Guide Link BGM integration with future Bluetooth-enabled Medtronic pumps. As early as mid-2018, per a Roche webinar (never confirmed by Medtronic)

Senseonics

• FDA Advisory Committee expected in early 1Q18, followed by a controlled US launch two months after approval. Will primary approval come with the second-gen transmitter and a dosing claim, or will those be pursued after?
• 180-day pediatric study submitted for publication in January and presentation at ADA 2018
• Broad EU rollout of 180-day Eversense XL in 1Q18
• Use in iDCL pivotal AID study with Roche pump and TypeZero algorithm. Study will wrap up by the end of 2018 and support CE mark of system.
• Beta Bionics is considering adding Eversense to its system, in addition to Dexcom CGM – will we see movement on this front in 2018?

Ypsomed

• Introduce the new mylife Unio Neva BGM, which will link to mylife app via Bluetooth. Plans to also integrate CGM into mylife in 2018 (discussions ongoing with Abbott and Dexcom)

Digital Health/Insulin Titration

Amalgam

• Basal+GLP-1, basal-bolus FDA clearance previously expected by end of 2017; possibly 2018 events/launches.

DreaMed

• Two ongoing evaluations of Dreamed Advisor Pro to wrap up in 2018. We’ll hear more from DreaMed at ATTD 2018 in February.  
  
  • Advice4U Pro Study (pivotal) for automated insulin pump setting adjustments (funded by Helmsley Charitable Trust) recruiting and set to complete in December. 
  • Evaluation of automated insulin pump setting using the Advisor Pro with collaborator Novo Nordisk recruiting and to complete in April.

Glooko

• Mobile insulin dosing system under FDA review as of the last update; no launch timing ever shared.
• N=240 trial of Mobile Insulin Dosing System (MIDS) for insulin degludec titration recruiting and to complete in November 2018.

Glytec

• Launch of insulin dose clinical decision support with Livongo?
• Used in Onduo payer/provider pilots in early 2018.
• Wider integration with AgaMatrix
• Integration with CGM players? A KOL hinted that Glytec is in talks with CGM companies at AADE.

Hygieia

• Integration of iSage Rx titration app in 2018
• Present results from nine-month US reimbursement study with BCBS Michigan? These were previously expected in September; we’re not what’s caused the delay.

Livongo

• Incorporate Voluntis, Glytec insulin dose titration into subscription service?
• Expand device compatibility through Diabeto acquisition?

mySugr (Roche)

• Further reimbursement for population health management bundle?
• FDA clearance of bolus calculator? The calculator is currently under FDA review.

Omada

• A June press release previously indicated that preparations are underway to offer services to Medicare beneficiaries in 2018…we assume CMS’ decision to not reimburse virtual/digital DPP (for now, at least) will prevent this being paid for at this time.

One Drop

• One Drop to develop app for Fitbit Ionic smartwatch in 1Q18
• “Revive” digital DPP to launch in 1Q18
• A-One RCT (n=400) of One Drop vs. One Drop+Mannkind Afrezza inhaled insulin ongoing, to complete in February 2019
• Use of app, coaching/education, and BGM in pilot with Sanofi, Innovation Health (Aetna+Inova Health), and Common Sensing
• Direct Chrome BGM-to-Apple Watch transmission previously slated for 4Q17 – possibly coming in 2018?

Virta

• “Transformational” one-year clinical data “in the process of rolling up” – publication forthcoming in 2018?

Verily

• Onduo payer/provider pilots in early 2018 with Glytec, Voluntis, and Telcare.
• Onduo and Blue Cross Blue Shield to begin T2D pilot in 1Q18 in three states; will use connected CGM, apps, lifestyle, medication review, Onduo experts.
• Launch of gen-one Dexcom sensor? Timing depends on G6 regulatory path, so could be a 2019 event.
• Updates on Novartis glucose-sensing contact lens?
• Updates on other diabetes-related projects: Project Baseline sub-study, Nikon (retinal imaging), and GSK (bioelectronics)?

Voluntis

• Scale in the US through Insulia Savings Program
• Expanded use/launches with partners Livongo, Sanofi, Onduo, and Ascensia

WellDoc

• Launches/scale of J&J partnership (?), Samsung partnership, BlueStar Rx (requires prescription because it includes an insulin calculator), and BlueStar (same as Rx minus insulin calculator and prescription), and BlueStar C (consumer)?

Other

NeuroMetrix

• Third-gen Quell expected to launch in 2018

Baxter

• An ongoing clinical trial in the US for an expanded hemodialysis (HDx) therapy enabled by THERANOVA

Big Picture

Themes

Diabetes Shows No Sign of Slowing: 425 Million Globally, ~50% Undiagnosed, $1.3 Trillion in Total Spending

• The IDF’s eighth edition Diabetes Atlas, released in November, paints a most concerning picture of the state of diabetes in the world. Today, there are estimated to be 425 million people living with diabetes, of which ~50% are undiagnosed. Since the last Atlas was released in 2015, prevalence has gone up 2.4% (+10 million). The number living with diabetes is expected to reach 629 million in 2045, and mortality is also high, as someone dies from diabetes every eight seconds (four million in a year). A particularly concerning region of the world, in terms of prevalence, is India – currently, the nation boasts the second-highest prevalence (73 million people with diabetes), but this number is expected to nearly double to 134 million in the next 28 years. For comparison, China currently has the most people with diabetes globally (114 million), expected to rise 5% to 120 million by 2045. From a financial perspective, one in eight global healthcare dollars ($727 billion) goes directly toward diabetes, up from $673 billion in 2015 (+8%). The US is responsible for nearly half of these costs (!) and is driving most of the expense growth. By 2045, IDF very conservatively expects spending to increase 7% to $776 billion in direct costs. We are sure this estimate is low, when considered alongside a 48% increase in prevalence. A separate analysis published in the Lancet Diabetes & Endocrinology calculated that, in 2015, the global direct and indirect costs of diabetes summed to $1.31 trillion. This degree of human, productivity, and financial cost is unsustainable, and we need to continue to push toward innovative thinking in incentives alignment, public-private partnership, and especially prevention to stem this tide before it reaches the severity IDF expects in 2045.   
  
  • A large degree of type 2 diabetes and other metabolic diseases can be attributed to the astronomical and climbing obesity rates worldwide. The Cities Changing Diabetes initiative revealed a “bold goal” at its Summit in Houston: Cut obesity by 25% in cities so that diabetes rates go no higher than one in 10 worldwide (currently, prevalence is one in 11). By their projections, instead of 11.7% prevalence by 2045 (approximately one in nine), which is where we’re headed now, achieving this bold goal will prevent 111 million people from developing type 2 diabetes (800 million fewer people will have obesity). Instead of spending >$1 trillion on type 2 diabetes in 2045, annual expenditures will be $872 billion, saving the healthcare system $200 billion. In the US, new data from the CDC reported obesity rates in 2016 exceeded 20% in all US states and territories, ranging from a low of 22% in Colorado to a high of 38% in West Virginia. In order to bend the curve of diabetes, it is of utmost importance to address the obesity burden.

A Monumental Year for Outcomes Beyond A1c Movement – Core CGM metrics agreed upon! – and CGM Report Standardization

• At this time last year, we noted that calls for outcomes beyond A1c had hit a fever pitch and the field was asking a lot of the right questions; today, stakeholders have agreed on a lot pertaining to core CGM metrics and display. What a year of progress! At ADA, our own Ms. Kelly Close and Bruce Buckingham set the tone for the rest of 2017, chairing a star-studded (and p-a-c-k-e-d) panel that led Dr. Buckingham to proclaim “You were all in the room where it happened” (borrowing a line from the smash musical Hamilton to illustrate the significance of the moment). He was referring to the fact that many major stakeholders – ATTD, AACE/ACE, EASD/ADA Technology Working Group, International Hypoglycaemia Study Group, and the JDRF T1D Outcomes Program – had agreed on a set of core CGM metrics to be measured and used in clinical trials with CGM: <54, <70, 70-180, >180, and >250 mg/dl for time spent in different ranges. The diaTribe Foundation hosted a consensus meeting in Washington just over a month later, convening many of the same faces and more, plus FDA representatives (both CDER and CDRH). The esteemed group verified the metrics settled on at ADA, and during the day’s work, reached further consensus on glucose metrics, namely using coefficient of variation for glycemic variability, mean glucose, AGP for CGM visualization, and more. The core metrics were again corroborated in consensus statements from ATTD and JDRF’s T1D Outcomes Program in an important December Diabetes Care issue. We were thrilled to see the amazing work done by so many to push the field beyond A1c finally put into print. And in ADA’s 2018 Standards of Care, we were pleased to see improved mention of the limitations of A1c (despite an oversight definition of level 1 hypoglycemia as “≤” 70 mg/dl vs. the consensus “<” 70 mg/dl…we expect to see this changed in time).
• There is still a lot of work to do, including:
  
  • Get regulators and payers on board with CGM outcomes, which is easier said than done – and probably the next step that will have the most impact. It is our understanding that CDER doesn’t accept CGM data (unlike their device division counterparts), and we wonder what it will take to not only have them be considered in regulatory review, but also included on labels. This is solvable problem in our view – the field could collect evidence that more time-in-range (or less time-below-range or above-range) is strongly correlated to A1c, to quality of life, to healthcare costs, to hospitalization, to complications, to adherence, or to other secondary markers of success. The bigger question is how to do so in a way that payers and regulators will accept. We were heartened, however, to hear former FDA official Dr. Zan Fleming say the “day of reckoning is coming” for CDER.
  • How should the field think about benchmarks for time in each of the bins? Should there be specific guidance, and should this vary depending on – age? Diabetes duration? Access to tools? Zip code? Socioeconomic status? Etc.; See our two pieces in diaTribe on this topic: one collecting a slew of fascinating key opinion leader views, and a column offering Adam’s perspective on this topic.
  • What constitutes a clinically meaningful improvement in time-in-range, hypoglycemia, or hyperglycemia? Answering this could be addressed with the data from the first bullet linking CGM data to outcomes – what changes in CGM outcomes are linked to important outcomes?
  • Should CGM use in drug trials be blinded or unblinded? We assume blinded will be acceptable in the next few years, as CGM is still not close to used by a majority of people with type 1 diabetes in the US. Once it moves to a majority of users, we’d guess real-time CGM will need to be a standard part of control and interventions arms.
  • What is the role of other metrics, such as estimated A1c, overall glycemic variability (coefficient of variation, standard deviation) vs. variability at a given time of day?
  • The field has plenty to do, but we are certainly more energized than at the end of 2017 – the above challenges (and others) will take time and resources, but they feel addressable with the right data, commitment, and continued drive to consensus!
• On a related note, the respected Dr. Phil Cryer penned a viewpoint in the same December Diabetes Care issue that argued for the definition of “severe hypoglycemia” to be expanded from an “event requiring assistance” to include a measured glucose concentration <50 mg/dl (a level associated with sudden death). This piece created some confusion, as many thought it was in opposition to the <54 mg/dl threshold – it’s not. However, we have yet to hear opinions on Dr. Cryer’s view to expand the severe hypoglycemia threshold to include very low measured glucose values. Of course, people who find themselves <50 mg/dl are in real danger, especially if they are recurrently at that level. Plus, this expanded definition of severe hypoglycemia would result in far more “events” in trials, a win for therapies that improve hypoglycemia. On the flip side (as UAE physician Dr. Mohamed Hassanein argued at IDF), including <50 mg/dl in the definition of severe hypoglycemia may cause certain patients to overreact to serious lows or even fear starting/continuing on insulin. More research is clearly needed in this area.
• As mentioned in the tech section, the AGP (Ambulatory Glucose Profile) is gaining more momentum from professional associations and key opinion leaders, and industry has listened: In June, Dexcom joined Abbott, Roche (to display Senseonics Eversense data), and Glooko/Diasend, in adopting and licensing the display by incorporating it into Clarity. We have yet to see Medtronic officially hop on board, but they are using something close in the new CareLink reports for MiniMed 670G. We’d love to see them make it official, license the report from IDC, and thereby make the display universal to help patients and healthcare providers unpack CGM data in the same way across device manufacturers. Clinicians have told us for a long time that the AGP is a great way to display data – we love that the momentum continues to build and that more companies see the value of adding this report on top of their own innovative ways to display data – the field needs both standardization and innovation.

Payer Trends: Shift to Value-Based Care, More Integrated Care Delivery (Full Stack), and Consolidation

• In early December, CVS Health announced plans to acquire Aetna in a $69 billion blockbuster deal. This was undoubtedly the most impactful example of movement toward more integrated, end-to-end models of care delivery this year. The deal is the first partnership of its kind between a major retailer/PBM and an insurance company. The possible implications are many, from better and speedier care, to lower long-term costs for patients and employers, to lower volume for over-burdened physicians and hospitals (but also lower revenue), possible questions related to drug formularies (as CVS Health is now the sole gatekeeper to drug access for the 22 million Aetna beneficiaries), to more outcomes-based contracting. To be sure, the full impact of the merger is TBD at this stage – one of our absolute favorite parts of the deal is CVS’s newfound incentive to remove junk food from its stores. Just days after CVS’s announcement, the New York Times reported that UnitedHealth’s Optum had purchased the physician group from DaVita (a large for-profit chain of dialysis centers) for $4.9 billion. As the article notes, the acquisition brings UnitedHealth closer to the direct delivery of medical care. While there are certainly caveats related to patient choice in this sort of vertical consolidation, seeing payers get more involved in direct patient care seems like a positive – aside from patients themselves, payers are financially interested in delivering timely, high-quality care, and when possible, preventing the need for it. Some analysts believe that the move toward consolidation is a response to the threat of competitors (e.g., Amazon) entering the pharmacy/healthcare arena – and it may be in part – but there is also no way that the current system, as currently organized, can cope with the chronic disease avalanche facing the country. Will the US see further consolidation between payers and healthcare providers, closing the gap between those paying for care and those delivering it?
  
  • When people ask, “Who does chronic care well?”, the answer is often “Kaiser Permanente.” Kaiser offers more end-to-end care, as it owns the value chain from payment to hospitals to salaried doctors. Is this the future of diabetes care? Virta is taking a similar approach (minus the payer component), bringing devices, software, and healthcare provider services in house and sharing risk with payers. The idea of integrating care and keeping all parties invested in the patient’s wellbeing seems to hold water. Then again, several large health systems who began offering health plans under the ACA quickly began losing money and got out of the business.
  • When will we see payers involving in more public-private partnerships? It’s one thing to foray into the pharmacy and care delivery space, but the real holy grail for insurance companies is prevention. It would be very much in payers’ interest to come together to put forth education campaigns, make cities healthier and more walkable, or even subsidize fresh, nutritious foods for low-income families.
• Calls for value-based care remain frequent, though the year saw only a few deals announced. Will we see more in 2018?  Medtronic and Aetna forged an outcomes-based contract, a number of payer pilots in glucose monitoring and digital health (see tech theme above) are presumably exploring the option, and population management companies such as Livongo, Virta, Omada, etc. sell their products to employers and payers with some risk/outcomes component, at least some of the time. Bigfoot CEO Mr. Jeffrey Brewer recently told us that payers are “quite eager for medical device companies to approach them with value-based pricing models” and that “The idea that payers are resisting value-based pricing models is woefully incorrect.” This was heartening to hear, though this area of the field still needs time to mature. ON the therapy side, we wonder how much connected devices will help. A key piece to address is outcomes – what should be tracked and used as the basis for payment, besides A1c? To patients, value can mean more time with family or less time thinking about diabetes. To payers, it ultimately comes down to lower cost or surrogate outcomes that denote lower cost. To employers, it means higher employee productivity. All of these outcomes can line up in many cases, but how are they reconciled when they don’t?  

Major disappointment as CMS decides not to reimburse digital diabetes prevention programs. Is there hope?

• We were extremely disappointed to learn that Medicare would not be covering digital DPP, as announced this summer. We’re not alone in our sentiments either – at AADE 2017, Dr. Deborah Greenwood referred to the decision as “a huge blow.” In fact, the ADA’s recently released Standards of Care included an update underscoring the benefits of digitally delivered diabetes management programs. An additional statement highlighted the advantages of telemedicine, especially for rural populations or those with limited physical access to healthcare. We couldn’t agree more. DPPs have the potential to make a meaningful dent in diabetes prevalence, but they’re only as useful as they are accessible. For the working single mother holding two jobs, finding the time to attend weekly classes just isn’t a realistic priority. Digital DPP providers like Omada Health and Noom shift the paradigm, meeting patients where they are and reducing hassle in the process. We’re hopeful that Medicare will take note as more players enter the space and drive positive outcomes. Demonstrating cost-effectiveness and ROI will be key here – an arguably taller challenge in prediabetes vs. later stage type 2 as it’s harder to say, “this person would have cost the system boatloads of money if not for our program” if the person doesn’t yet have the first signs of complications. Then again, Dr. Jaakko Tuomilehto argued at IDF that prediabetes itself confers CV risk. But hope is not lost for digitally-administered DPP: the Center for Medicare and Medicaid Innovation is undertaking a large-scale pilot to investigate virtual DPPs, and many close to the field see reimbursement as possible in the near future. We wonder which DPP providers are to be included in the pilot, and whether the trial has begun recruiting – news has been fairly quiet on this front.

Insulin Pricing Controversy Remains at a Standstill

• Following the frenzy of controversy over insulin pricing in 2016, we were surprised and disappointed not to see more meaningful progress toward any solutions this year. All the major insulin manufacturers released position statements over the past 12 months pledging to cap list price increases for their products. Sanofi’s commitment to limiting list price increases to 5.4% (so as not to exceed the national health expenditures growth projection), is the most generous; Novo Nordisk has committed to no more than single digit annual increases in list price for its medicines, and Lilly similarly committed to an annual increase of no more than 10%. These are promising initial steps forward, and we applaud these pharma players for the attempt at transparency, but it’s unclear how these caps on list price will actually affect access for patients. For one, these statements are commitments not to make the situation worse, but there’s little hope in these pledges that insulin will become more affordable for patients in the coming years. Moreover, list price only means so much without considering patient discounts and public/private reimbursement. PBMs are still a black box in the complicated web of pharmaceutical pricing, and make no mistake, we need all stakeholders to come to the table if we have any chance at unraveling drug cost to improve access. There’s a clear discrepancy between list price and sales realized by the manufacturer. Illustrating this, a JAMA article published in July showed that 27% of total pharmaceutical sales in the US – a grand sum of $115 billion – was paid by industry to payers and PBMs in 2015. This rebate structure, embedded in the current price-setting system, leads to complexity and a lack of transparency, particularly around decisions made by the middlemen PBMs. We’re eager to know the total amount in rebates paid by diabetes companies, specifically. Even though industry has started to shed light on profits lost to payers/PBMs, each company reports this a little bit differently, which makes it hard to compare or to derive concrete insights (for example, Merck reported 41% profit loss to rebates, discounts, and returns, and we haven’t seen other companies include “returns” in their calculations; Lilly reported ~50% profit loss to rebates and discounts).
• The lack of large-scale action is not to say that the issue of insulin pricing has faded from the field’s consciousness. At ADA 2017, for the third year running, a headlining symposium was dedicated to the topic. Chaired by Dr. Irl Hirsch (one of the first to sound the alarm bells on this issue), the discussion focused largely on the complex interplay between PBMs and pharma. Former ADA Chief Scientific and Medical Officer Dr. Robert Ratner noted that rebates to PBMs rose substantially from $67 billion in 2013 to a whopping $106 billion in 2015. He argued that we should get rid of these “middlemen with no added value increasing expenditures.” We aren’t sure they add no value, but we agree that we need more transparency – this is urgent. ADA also published a statement of concern in August regarding PBM formularies and exclusive contracts that restrict patient choice and that often force people with diabetes to switch medications and interrupt continuity of care. Formulary exclusions add another layer of complexity to price-setting. Ultimately, it’s essential that all stakeholders across the system – industry, payers, and PBMs alike – stop oversimplifying the issue of drug pricing and start sharing the responsibility. With so little transparency, it’s impossible to know where to begin to address the rising cost of insulin, but we suspect it will take collaborative action from all participants. One of our fondest wishes for 2018 is to see representatives from all these stakeholder groups speaking face-to-face.

The Movement for Person-First, Non-Stigmatizing Language Picks Up Steam in Diabetes & Obesity

• When the ADA/AADE jointly published a paper detailing the importance of language in diabetes, a movement that had been a long time coming emerged fully-fledged. And when the ADA edited a guideline to promote the use of person-centered, strength-based language, that movement gained a mandate. To be clear, this work has been building for some time: The first groundbreaking came in 2012 when Diabetes Australia published a position statement on the language that should be used in communicating with and about people with diabetes. This effort was led by Professor Jane Speight, and we had the pleasure of hearing from her recently at IDF in Abu Dhabi. Another main event came this October, when work spearheaded by Dr. Jane K. Dickinson, with representatives from the ADA and AADE, was published simultaneously in Diabetes Care and The Diabetes Educator. This paper outlines a set of specific recommendations about language use in diabetes, advocating for the adoption of empowering, non-judgmental, non-stigmatizing vocabulary to motivate, educate, and foster patient/provider collaboration. For example, “adherence” and “compliance” can be replaced with “engagement,” “participation,” or “involvement.” A practical guide is available here. Most recently, in the ADA’s 2018 Standards of Care, the association has altered communication guidelines to emphasize person-centered and strength-based language. We were excited to see this addition, and we’re enthusiastic about how positive, person-first language might improve the lived experience of people with diabetes. While we think many real-world providers would benefit from a more specific definition of “person-centered” and “strength-based,” we’re glad to see the ADA taking a stand on language, especially as it relates to the psychosocial wellbeing of patients. While 2017 certainly offered much to celebrate in terms of the language movement, we should note that there’s a long way to go. At AADE 2017, Dr. Dickinson shared an anecdote about presenting on language at ADA and then immediately afterward, attending a dinner where healthcare professionals were using the term “non-compliant” casually and abundantly. And these were HCPs – getting the media to adopt non-stigmatizing vocabulary in depictions of diabetes is another, greater challenge. The joint ADA/AADE paper clarifies that the language recommendations are meant not only for people who treat diabetes every day, but also for the media and for society in general. Again, this was an important stride, but it will take persistent effort to uproot and change language habits in diabetes care.
• The language movement is also accelerating in obesity. In July, the AMA adopted a resolution to destigmatize obesity with person-first language. This resolution simply promotes the use of “person with obesity” rather than “obese person,” and asks providers to avoid terms such as “morbidly obese” and “fat.” We see this resolution as reflective of the increasing perception of obesity as a physiological disease, rather than a result of poor lifestyle decisions. Better language enables better patient-provider relationships, which is a key piece of improving treatment for obesity, and we were glad to see the AMA take a stand here like the ADA has done as well. The language movement recently received attention from the mainstream press in a New York Times article, which focused specifically on language use for people who treat childhood obesity. The author, pediatrician/writer Dr. Perri Klass, suggests we should avoid “obesity” entirely in talking to children about excess weight, adding another layer of nuance to the conversation. On this, Dr. Lee Kaplan shared his perspective, which we can’t help but share again here. Dr. Kaplan agreed that “obesity” can be off-putting but also explained that “if we are to take obesity seriously as a disease, it has to have a name, and that name needs to be a noun.” He reiterated the critical importance of person-first language, and shared practical advice for navigating these challenging patient/provider conversations: “In talking with patients about their own weight, I often ease the discussion by avoiding ‘obesity,’ but I generally come back to it when describing the implications of the disease more abstractly – such as ‘many of the medications that you are on are for diseases that are caused by obesity, so we need to address your weight as well’.” Dr. Kaplan drew a parallel to cancer, “two diseases that no one wants to have”: “In cancer as in obesity, we may use euphemisms to soften the initial diagnosis. But the greater determinant of how our efforts will be perceived comes in the words and actions that follow.”
  
  • AACE released a position statement this year promoting a switch from “obesity” to “adiposity based chronic disease” (ABCD). We do think this could have profound societal impact in reducing stigma, though to our knowledge, there hasn’t been much real-world uptake of the term yet (and it’s certainly a long acronym for patients to keep track of). That said, it’s noteworthy that another professional society has taken a stand on language and has offered a very specific alternative for a word (“obesity”) that is steeped in stigma by now, especially considering that we didn’t find ADA’s definitions of “person-centered” and “strength-based” to be quite specific enough.

Novo Nordisk/Cities Changing Diabetes Set Ambitious Goals for Prevention; Four New Cities Added for 12 Total, Now Covering Population of 100 Million+

• The Cities Changing Diabetes (CCD) Program scaled up in 2017, growing 50% from eight to 12 cities, and issuing an ambitious call-to-action for all cities to reduce their obesity rate 25% by 2045 to keep global type 2 diabetes prevalence below one in 10. Three cities from China were added to the program – Hangzhou (population ~nine million), Beijing (~21.5 million), and Xiamen (~3.5 million) – which is notable considering the sheer size of the diabetes epidemic there. Nearly 11% of the adult population in China has diabetes, and nearly half of all adults have prediabetes (that’s ~500 million people). Next, we hope to see participation from India, another country with skyrocketing obesity and diabetes rates, not to mention one of the world’s largest populations at baseline. Leicester in the UK was the fourth CCD addition in 2017, and Dr. David Napier shared at IDF that the 12 participants together now cover >100 million urban citizens. The premise of the program is that cities house two-thirds of the global population, but are also a small enough entity to implement thoughtful public health solutions that consider sociocultural norms and economic factors, making them an ideal target for largescale diabetes prevention. At the second-ever Cities Changing Diabetes Summit, this year in Houston, former Philadelphia Mayor Michael Nutter argued that local leaders should elevate health to the top of their agenda because a city’s vitality is directly affected by the health of its citizens. Also at this gathering, we saw early success stories from the eight existing CCD participants: Houston, Mexico City, Shanghai, Tianjin, Copenhagen, Johannesburg, Rome, and Vancouver. Each uses a tailored approach based on the major variables spurring diabetes prevalence and incidence in that particular city. CCD has also committed to sponsoring and conducting vulnerability assessments in at least five more cities in 2018, so we could see this program grow even more next year.
• As for the ambitious call-to-action, Novo Nordisk CMO Dr. Alan Moses called this a “bold goal,” reviewing the rationale behind this target during an opening talk at the Houston CCD Summit. The numbers – cut obesity by 25%, keep diabetes below 10% – are not at all arbitrary. Currently, one in 11 adults globally (9.1%) has type 2 diabetes. CCD is advocating for action so that worldwide prevalence does not rise above one in 10 (10%). This means a higher proportion of the urban population will have type 2 diabetes in 28 years, but we’ll be “bending the curve” so to speak, changing an upward trajectory to point down. Instead of 11.7% prevalence by 2045 (~one in nine), which is where we’re headed now, achieving this goal will prevent 111 million people from new-onset diabetes, and 800 million fewer people will have obesity. Instead of spending >$1 trillion on type 2 diabetes in 2045, annual expenditures will be $872 billion, saving the healthcare system $200 billion. As Dr. Moses put it, “this bold goal makes fiscal sense, it makes public health sense, it makes sense for the individual citizen at risk.” He went on to explain the emphasis on obesity, which isn’t the only risk factor for new-onset type 2, but it does account for 44% of diabetes burden. As both the diabetes and obesity epidemics continue to grow, in size and scope, we couldn’t be happier about Novo Nordisk’s and CCD’s leadership in prevention. This is one of the most intimidating problems in public health, considering the spread of obesogenic environments, and it will certainly take ambition and a “bold goal” to make a difference. That said, and we can’t underscore this enough, diabetes prevention is possible and it is urgent. We look forward to seeing what CCD accomplishes next, and we applaud the progress made so far. In fact, Novo Nordisk management recognized the improvements made between the first CCD Summit (2015, Copenhagen) and this year’s gathering in Houston, which was one of the best meetings we attended in 2017, both inspired and inspiring.

Soda Tax Movement Presses On; Some Wins, Some Losses

• Following a wave of landmark 2016 votes to establish per-ounce taxes on sugar-sweetened beverages, 2017 saw positive readouts on the early effectiveness of these policies to quell soda purchases (which are the number one vehicle for added sugar in the American diet). Preliminary results on the impact of the Philadelphia Beverage Tax (which went into effect in January 2017) were unveiled at Obesity Week 2017 in October. There was a 57% drop in the volume of sugar-sweetened beverages sold in Philadelphia vs. the comparator city of Baltimore, and there was a 66% drop vs. surrounding counties in Pennsylvania. Respectively, this translates to a whopping 769,000 and 838,000 fewer ounces of sugar-sweetened beverages sold. Another recent analysis found that soda sales have declined nearly 10% in Berkeley, CA since the city’s sugar-sweetened beverage tax was introduced in November 2014. Looking outside the US, Health Affairs reported that the one peso/liter excise tax on sugar-sweetened beverages in Mexico reduced purchases by 5.5% in 2014 and by an even greater 10% in 2015. As a reminder, Mexico and Berkeley pioneered the soda tax movement in 2014, followed by Philadelphia (PA), Boulder (CO), Albany (CA), Oakland (CA), San Francisco (CA), and Seattle (WA) in the US, and France, Norway, and Hungary abroad.
• Despite these early signs of success, the soda tax is still a difficult political proposition, as illustrated by Cook County’s repeal of its sugar-sweetened beverage tax and persistent roadblocks in Mexico. Cook County, IL (home to the city of Chicago) voted 15-2 to repeal the area’s penny-per-ounce tax on sugar-sweetened beverages in October. The tax, which narrowly passed in November 2016, had been contentious from the start. Opponents argued that it burdens consumers and small businesses, and an onslaught of messaging from the “Can the Tax Coalition” (funded by the American Beverage Association) petitioned that “beverage taxes just don’t work.” Only time will tell whether this is an isolated incident or rather foreshadowing of a larger backlash against soda taxes – we certainly hope other locales don’t follow suit. Duke’s Dr. Kelly Brownell expressed some optimism on this front, and we were excited to quote him in Closer Look: “The soda companies showed their muscle in Cook County and won this skirmish. But like the companies that won some early skirmishes with cigarette taxes and then could not keep their finger in the dike, soda companies are losing the tax fight all around the world. It’s just a matter of time until these taxes are very common.” In Mexico, despite the data to support the success of the tax in reducing soda consumption, Mexican Diabetes Federation President Mr. Juan José Irazabal Lujambio argued at IDF 2017 that the tax’s impact has been hampered in part by government corruption and lack of widely available clean drinking water as a substitute for soda.
• The balance of wins and losses in 2017 shows that a soda tax will not work in isolation – several factors need to be in place to support success. This includes availability of substitutes, a non-corrupt government, and elastic consumption (a small change in price should correspond to a large reduction in consumption. The tax must be passed down to consumers (not incurred internally by a soda company or distributor), and there must be sufficient enforcement against arbitrage (when soda is sold across borders due to price difference). Above all, there needs to be resilient political will. We are optimistic that soda taxes can pack a mighty punch if implemented thoughtfully. To this end, we were encouraged by a meta-analysis presented at ENDO 2017, which estimated that a national sugary drink tax would, in one year, reach 307 million lives (the US population), prevent 576,000 cases of obesity, cost only $47.6 million, and save ~$31 healthcare dollars per dollar spent – now that’s impressive! We’d also note that even if the taxation of sugary beverages doesn’t directly diminish obesity, this additional tax revenue could be used to subsidize bottled water or invest in bike lanes, parks, and community nutrition/health programs, which could have downstream health impacts at scale.

New Data Highlights Treatment Inertia as a Persistent Problem in Diabetes Care; Thought Leaders Scrutinize Treatment Guidelines

• It’s no secret that many people with diabetes do not meet an A1c target of ≤7%, and we’ve heard serious criticism of clinical inertia as a factor contributing to this problem. Most recently, the DISCOVER trial (n=15,992) put hefty data behind the blame: Results from this global real-world study indicate that mean A1c of patients starting second-line therapy is 8.4%, with nearly one-third of this worldwide patient population exhibiting A1c ≥9%, and only 15% meeting goal of A1c <7%. Major treatment guidelines recommend that second-line treatment be initiated after three months if a patient doesn’t meet glycemic targets, and yet people with diabetes are far, far above goal before starting on a second drug. What’s worse is that this treatment inertia is impacting outcomes: Worldwide, 18% of DISCOVER patients had microvascular complications (this number is likely higher in reality, as the investigators also pointed out widespread under-diagnosis), despite an average diabetes duration of only 5.6 years. We also saw evidence this year quantifying how patients resist treatment intensification with insulin (and possibly other injectables). A study conducted at Brigham and Women’s hospital in Boston found that 30% of type 2 diabetes patients decline insulin therapy after their provider recommends it. This paper in Diabetic Medicine underscored the need to overcome the stigma of “failure” or sense of punishment associated with insulin, and also advocated for technology to improve the experience of taking insulin. There are behavioral strategies that providers can employ to support patients in insulin initiation, which is understandably scary, but we also understand that providers, especially the PCPs who often provide sole diabetes care, are already overworked. This year, we’ve continually lamented the sluggish uptake of combination therapies, especially basal insulin/GLP-1 fixed-ratio combinations Xultophy (Novo Nordisk) and Soliqua (Sanofi). Still, we know PCPs struggle to prescribe and titrate insulin, making it all understandable that they would be intimidated by two injectables in one new product – educating HCPs on fixed-ratio injection is an essential step in this commercial story, and Sanofi has outlined efforts to this end (we can’t say the same for Novo Nordisk, since the company has explicitly de-prioritized Xultophy in favor of basal insulin Tresiba and GLP-1 agonist Victoza, which we still find hard to believe).
• Thought leaders were also vocal this year about the need for more explicit recommendations in professional treatment guidelines, often comparing the ADA vs. AACE algorithms. At CMHC 2017, Dr. Yehuda Handelsman praised the comprehensive approach AACE has on treating weight, dyslipidemia, hypertension, and glycemia in turn. In particular, AACE’s glycemic control algorithm has listed medications in a suggested hierarchy of usage, de-prioritizing insulin in favor of GLP-1 agonists, SGLT-2 inhibitors, DPP-4 inhibitors, and TZDs in dual therapy. Sulfonylureas are at the very bottom of every part of the treatment algorithm, and AACE has placed explicit cautions on SUs, TZDs, and insulin (due to hypoglycemia and CV risks). We’re glad to see an algorithm actively incorporate safety info – while ADA provides this information, its algorithm is more neutral in recommending specific drugs over others. As Dr. Ralph DeFronzo has said, “you can’t have drugs all on one line and say they’re all equivalent – that’s not helpful to providers, and they’re not all equivalent. Sulfonylureas should not be on that line.” In its 2018 Standards of Care, ADA has moved away from the infamous line and now has a chart with details on CV effects, body weight effects, hypoglycemia risk, A1c-lowering efficacy, cost, administration (oral vs. subcutaneous), renal effects, and additional considerations for each pharmacotherapy class. Still, there’s not a lot of clear guidance for providers on which therapies are more advanced, safer, and more efficacious overall. The issue is certainly complicated, and we understand that prescription decisions have to be individualized taking all these factors plus patient preferences into account. That said, we feel the tide slowly turning toward more aggressive treatment algorithms, as new drugs become more established through real-world and RCT data, and as providers become more familiar with newer agents. There’s also growing consensus that earlier intervention has substantial payoff, in line with the criticisms of treatment inertia that we discuss above. In an ideal world, precision medicine will enable patients to easily find the safest, most efficacious medicines that fit their lifestyle via an algorithm.

Physician burnout remains a “serious problem” – studies show the unsustainable demands placed on physicians

• A concerning JAMA Viewpoint published in February disclosed disturbing statistics regarding the HCP environment in the US: (i) the prevalence of burnout among practicing physicians in the US exceeds 50%; (ii) in one recent study, physicians spent approximately 33% of their work hours performing direct clinical work vs. 49% completing clerical tasks and interfacing with the EHR (insanity!); and (iii) 44% of US physicians work >60 hours per week vs. 8% of US workers that work >60 hours per week. The article noted (and we agree) that EHRs have caused serious burden for physicians, distracting from meaningful patient interaction. The epidemic led AMA’s Dr. Marie Brown to dub burnout an “unsustainable crisis” at Health 2.0. We hope voice recognition technology like Amazon Alexa can record what happens in encounters automatically. Moving forward, we also hope more digital care delivery models will leverage passively collected data, and therefore, require less documentation and can enable actual focus on patients. However, even with these potential advances in technology, provider uptake may prove challenging. A Medscape survey of physicians (n=1,153) published in September identified current technology to be shockingly underutilized – when asked which forms of communication they use with patients (providers could select more than one), 93% indicated the telephone, while only 34% chose email, 8% text, and 2% videoconferencing. Two-thirds of physicians don’t use dictation or voice recognition for EHR notes, which could obviously help cut down on note capture time. We believe technology must play a larger role in improving physician workflow, but it’s clear from this report that even basic functions are not sufficiently optimized. In an ideal world, the diabetes clinical care is going through a transition period with growing pains, but will come out on the other side in a more collaborative, data driven, patient-centered, value-based fashion.
• Burnout is particularly rampant among family physicians, as a study published in JAMA this summer reported that 25% of family physicians in the US experience burnout. After controlling for workplace variables, the study concluded that the EMR in itself is not an explanation for physician burnout. Rather, added tasks stemming from EMR, including a time-consuming documentation process often spilling over from the clinic to home, show statistically significant correlation to burnout. An EMR that saves time and maximizes patient-provider interactions is of utmost importance, particularly in diabetes care, which requires a complex interplay of drugs, devices, and coordination between multiple members of a care team. We wonder if Apple, Alphabet (Google, Verily), Microsoft, Amazon, or Facebook could actually launch a EMR 2.0 that is far better from everyone. Boy does the field need it!

Most Highly-Read Reports

• IDF Diabetes Atlas 8th Edition: 425 million have diabetes globally, with 629 million projected by 2045; 1 in 2 undiagnosed – One person dies every eight seconds from diabetes; Prevalence expected to climb to 629M by 2045; $727 billion spent on diabetes this year (direct costs only)
• CDC releases annual report on health and healthcare trends – 488-page document details US health trends over 40 years; Diabetes prevalence has grown from 5% to 10% from 1999 to 2014; Obesity prevalence has risen from 23% to 38% over the same time period
• Virta Health launches to reverse T2D in 100 million people by 2025; Combines nutritional ketosis and technology; 1% A1c reduction at 10 weeks; Impressive investors include Venrock, Redmile – <30 grams of carbs per day + coaching; 87% of insulin users decreased dose or eliminated insulin; currently available through some self-insured employers; out-of-pocket: ~$400/month
• Pennington’s Dr. William Cefalu named ADA Chief Scientific & Medical Officer; Replaces the recently-retired Dr. Bob Ratner – Tenure begins on February 20; 30 years experience in diabetes research + care; Emphasizes precision medicine, advocacy, development, and helping PCPs treat PWD
• JDRF Announces New T1D Venture Philanthropy Fund to Drive Commercialization; Bigfoot Biomedical is the First Investment – New “venture philanthropy” model for funding early-stage, high-impact type 1 opportunities; Goal to raise over $80 million by 2018; Big win for Bigfoot as the first investment; Quotes and questions
• Merck/Amazon Alexa Diabetes Challenge – Wellpepper’s Sugarpod (Alexa-enabled foot scanner and care plan) takes prize; Alexa not a HIPAA-enabled technology?; Four other finalists include chatbot nutritionist & mood-sensing diabetes coach
• JAMA viewpoint calls out rebates, highlights complexity of pharmaceutical distribution and financing – Points to lack of transparency around PBMs; 27% of 2015 pharmaceutical sales revenue ($115B) captured in rebates from manufacturers to payers/PBMs; Illustrates major consolidation, with CVS + Express Scripts + UnitedHealth controlling 73% of PBM market
• CVS Health to acquire Aetna in a $69 billion transaction; Major move to reduce costs and improve access to health care – Partnership represents an effort to provide greater integration of health care and more effectively meet patient needs; Savings could total $117/month for people with chronic diseases; Transaction expected 2H18
• What’s the ROI of patient engagement in drug development? New projection reports potential for 500-fold ROI – Research conducted through Duke, Tufts, J&J collaboration; Facilitated discussions between patients/investigators to revise clinical protocols could accelerate development timeline by ~2 years, increase product value by $35-$75M
• Study reports that higher A1c goals are not associated with decreased hypoglycemia in older patients – Higher A1c goals do not protect against hypoglycemia in older patients with type 2 diabetes; Authors recommend CGM for hypoglycemia reduction

What We Got Right/Expected

• Diabetes prevalence and costliness continues to climb. Today, there are 425 million adults with diabetes in the world, only half of whom are diagnosed. One in eight global healthcare dollars goes toward diabetes ($727 billion), with the US responsible for 48% of these costs.
• Globally, the average A1c for people with type 2 diabetes remains far above goal at 8.4%. The DISCOVER study collected data from nearly 16,000 type 2s from 38 countries, reporting this mean A1c for patients initiating second-line therapy.
• PBMs cast a shadow over the pharmaceutical industry, showing little-to-no transparency with regard to rebates. ADA released a statement of concern in August, criticizing how PBM formularies are designed in a way that limits patient choice, interrupts continuity of care, treats diabetes as “one-size-fits-all,” and drives up cost.
• Outcry continues surrounding rising cost of insulin, but progress toward a solution seems to have all but halted.
• New outcomes-based payer contracts emerge for diabetes drugs/devices. We see this as part of the larger movement, happening slowly but surely, toward value-based care in diabetes.
• Conversation continues on how we can reduce stigma surrounding type 2 diabetes, though these ideas have yet to spread meaningfully to the general public.

What We Got Wrong/Did Not Expect

• Few drastic changes to the US healthcare system make it through Washington. Following the surprise election of President Donald Trump, many anticipated swift changes to the Affordable Care Act (ACA). As we understand it, healthcare funding has been significantly impacted, but the ACA remains intact after multiple attempts by Congress to “repeal and replace.”
• HHS Secretary Dr. Tom Price, a strong critic of the ACA, resigns <one year into his tenure. Dr. Price was appointed by President Trump as part of a “dream team” to repeal and replace the ACA. Former Lilly executive Mr. Alex Azar is currently President Trump’s leading pick to fill the open cabinet position.
• FDA Commissioner Dr. Scott Gottlieb focuses on lowering barriers to drug development. We weren’t sure what to expect when Dr. Gottlieb took over for former Commissioner Dr. Robert Califf. To-date, we’ve been impressed by his blog posts promoting generic drug development, his work to establish less burdensome regulatory review processes, and the new digital health initiative that aims to spur innovation and clarify FDA’s role in the field.
• In an unprecedented merger, CVS/Aetna joined forced in a $69 billion deal with the potential to save patients money and improve access to care. We’ve heard a mix of optimism and skepticism surrounding this move. While we’re confident that Aetna CEO Mr. Mark Bertolini has patients in mind, the integration of a PBM and a payer is a novel concept, one that gives CVS Health serious power.
• Blink Health calls off partnership with Express Scripts to offer Lilly insulins at a 40% discount. As we understand it, Lilly and Blink Health remain partners and continue to offer discounted insulin at participating pharmacies. It seems that the middleman PBM was simply cut out of the deal, for reasons unclear.
• Cook County (home of Chicago) repeals the soda tax, which passed in November 2016. This was disappointing, though we acknowledge that the tax was contentious from the start.

Key Questions for 2018

• What’s next for the #beyondA1c movement?
• How can we get PBMs to the table to discuss the rising cost of insulin? A solution to this insulin pricing controversy will almost certainly require collaboration and transparency from all stakeholders (PBMs, payers, manufacturers). Even as industry players come forward with position statements on insulin affordability, PBMs have been largely silent and absent from the conversation.
• Will the new outcomes-based payer contracts forged in 2017 be successful? Will these deals prove that outcomes-based reimbursement can be a win-win-win for payers, manufacturers, and patients? How many new outcomes-based contracts will we see in 2018?
• Will virtual DPP platforms win reimbursement from major payers? Medicare’s decision to restrict coverage only to in-person iterations of the DPP was extremely misguided, in our opinion.
• How broadly will Medicare DPP be utilized?
• How many more cities will pass a soda tax? How will soda consumption change as a result? What new data can we collect to show the impact on a population’s obesity, diabetes, and overall health?
• Will Virta Health make meaningful progress toward reversing type 2 in 100 million people by 2025?
• Will there be a resurgence of interest in repealing and replacing the ACA? We certainly can’t take our eyes off Washington these days, and we’ll continue to monitor healthcare politics/policy closely in 2018.
• What changes are in store for FDA and NIH as President Trump’s term continues?
• Will new FDA guidances have meaningful impact on drug approvals and prices? Commissioner Dr. Scott Gottlieb made strides this year, coming out with new draft guidance on generic drug approvals, less burdensome regulatory review processes, etc. But will these changes translate in the near-term for industry? For patients?
• Will the mainstream media catch wind of the language movement around diabetes? To be sure, there’s a lot of work to be done in shifting vocabulary around obesity as well. We wonder what positive change will come through Novo Nordisk’s “Get Real About Diabetes” program, with spokesperson Anthony Anderson (star of the television show “Black-ish”).

What’s Coming in 2018?

• Medicare reimbursement of the DPP. This is a highly-anticipated move.
• The CVS Health/Aetna merger is expected to close in 2H18, pending regulatory approval.
• We’ll continue to monitor the breadth and depth of changes made to the US healthcare system under President Trump. The ACA’s fate still seems very tenuous, and there are a host of other decisions on funding, research priorities, health policy, etc. that could impact diabetes and obesity.
• A new HHS secretary will be appointed, presumably former Lilly executive Mr. Alex Azar, to replace Dr. Tom Price.
• ADA could make more frequent updates to its Standards of Care, for example, incorporating semaglutide as an available GLP-1 agonist. We learned that this will be a more flexible document in 2018, and ADA will consider making more frequent updates where appropriate/substantiated before the official 2019 Standards of Care.

Obesity

Themes

Obesity Rates Hit 40% Among Adults, 19% Among Youth in the US

• In October, the CDC released its latest obesity rate estimates, reporting 40% prevalence among US adults and 19% prevalence among US youth for the years 2015-2016. Based on NHANES data, these rates represent a 3% increase for adults and a 2% increase for youth from 2013-2014. Even though these changes were not statistically significant vs. 2013-2014, we note that obesity prevalence has trended in the wrong direction for far too long (since at least the 1980s) and shows no real signs of plateauing. The message here is stark but important: Obesity remains an urgent public health crisis. Obesity rate in 2015-2016 was higher among non-Hispanic blacks and Hispanics (47% for adults) vs. non-Hispanic whites (38%) and non-Hispanic Asians (13%). These figures are the most reliable national obesity estimates, to our understanding, as they’re based on standardized physical exams conducted in-person by professionals.
  
  • The CDC also released state-by-state estimates of obesity rate for 2016 this year – prevalence exceeded 20% in all US states and territories. Colorado was at the low end, with 22% adult obesity prevalence, while West Virginia was at the high end, with 38%. These data also show that minority populations are hit particularly hard by the obesity epidemic, with 38% and 33% prevalence among black and Hispanic populations, respectively, compared to 28% among whites. Based on these findings, the Robert Wood Johnson Foundation took a tentatively optimistic tone and suggested that obesity rates seem to be leveling off. However, we’re reluctant to take these data as a sign of progress. For one, we would expect the growth rate to flatten as prevalence grows to affect such a large segment of the population (roughly another one-third of Americans are overweight). Further, these estimates are based on BRFSS data, which is a self-report system utilizing phone interviews. By our calculations via a weighted average, the BRFSS data yields a ~30% national obesity rate – indicating serious underestimation of individual weight (or, less likely, overestimation of height) and pointing to harmful stigma and persistent under-diagnosis of obesity. Nevertheless, this report served as a reminder of the heterogeneity in obesity prevalence across the country and how certain socioeconomic and cultural factors can profoundly impact local health.
  • Despite efforts to combat childhood obesity, prevalence continues to rise, highlighting the need for better and more treatment and prevention efforts. Obesity rates increase as children get older, from 14% among those 2-5 years-old, up to 18% for 6-11 year-olds and 21% for 12-19 year-olds. Prevalence among children 2-5 years-old rose from 9% in 2011-2014 to 14% most recently. We learned this year that childhood obesity quadruples risk of later developing diabetes, and though some have argued that obesity rates are leveling off, we’re not convinced the data support that conclusion. The obesity epidemic begs for increased efforts to both treat and prevent obesity. On the former, we hope to see increased industry investment to innovate in pharmacological treatments, devices, and surgery for obesity, particularly as basic science continues to build our understanding of the physiological mechanisms underlying obesity. We expect Novo Nordisk’s semaglutide CVOT in obesity to go a long way in promoting an understanding of obesity as a medical, treatable disease, but we’re hoping to see change long before that study reads out. On prevention, we need far bigger investment and collaboration across public and private sectors to tackle the obesogenic environment that contributes so strongly to establishing obesity in humans – these efforts include everything from promoting activity in cities to making healthy food choices more accessible, a la Cities Changing Diabetes, another Novo Nordisk effort. You’ll find dedicated themes on both the first-ever obesity CVOT (below) and the Cities Changing Diabetes program (above) in this reflections piece.

GLP-1 Agonists – Present (and Future?) of Obesity Care; Novo Nordisk Leads Class with Saxenda, Continues to Innovate with Semaglutide

• GLP-1 agonists gained favor as obesity treatments this year. Novo Nordisk’s Saxenda (high-dose liraglutide) continued to lead the market for obesity pharmacotherapy, while the company’s second-gen GLP-1 agonist semaglutide showed a “new level of efficacy” in phase 2 obesity trials. Saxenda had another standout year, with revenue >doubling YOY in 1Q17 to $77 million, rising 82% YOY in 2Q17 to $105 million, and growing 53% YOY in 3Q17 to $101 million (although this marked a slight 7% sequential drop). The product held 63% of the market by value in 1Q17, 73% in 2Q17, and 73% again in 3Q17. Not only is Saxenda the indisputable frontrunner within the obesity class, but it also remains the primary growth driver. The class grew ~66% YOY in 1H17, totaling $122 million in Q1 and $146 million in Q2 – excluding Saxenda, however, YOY growth was diluted to ~21% YOY, and pooled sales totaled only $45 million in Q1 and $41 million in Q2. Saxenda achieved key regulatory milestones in 2017 as well. In April, FDA added three-year SCALE data to the product label, showing long-term weight loss maintenance and safety, and in June, the EMA approved inclusion of LEADER data on the Saxenda label to reflect cardioprotection with liraglutide. These label updates could spur further franchise growth. Weight maintenance has proven to be more challenging and more frustrating for patients/providers than is initial weight loss, and real-world HCPs are understandably wary of weight loss drugs considering the host of safety issues that arose with the first generation of obesity pharmacotherapy – three-year SCALE data provides reassurance on both fronts. Moreover, as CV risk management becomes a central component of diabetes and obesity care, we imagine any association with LEADER data is a positive sign for Saxenda uptake.
• Novo Nordisk isn’t stopping with Saxenda, but has plans to launch a robust phase 3 program (STEP) for semaglutide in obesity next year. Details on four STEP studies were announced during the company’s Capital Markets Day last month: ~4,500 patients with obesity, with or without type 2 diabetes, will be enrolled (this doesn’t include the first-ever obesity CVOT – more on this below). Semaglutide has shown profound weight loss in clinical trials to-date. In the phase 2 obesity study (n=957), once-daily semaglutide produced ~16% weight loss (mean drop of 39 lbs from a baseline 244 lbs) among participants who completed one full year of treatment. This was double the efficacy seen with Saxenda, which produced ~8% weight loss in people who completed the full year. Participants randomized to placebo experienced ~2% weight loss (all treatment was adjunct to diet/exercise). Without a doubt, these are tantalizing results, leading CSO Dr. Mads Thomsen to refer to a “new level of efficacy” in weight loss with semaglutide. Of note, the phase 3 STEP trials will use once-weekly semaglutide injections rather than once-daily, but management seems confident that this will not impact weight loss efficacy (and if anything, could improve adherence and treatment satisfaction). STEP will be a foremost clinical program to watch in 2018, and we’re thrilled to see ongoing innovation in GLP-1 for obesity.
  
  • Several early-stage obesity candidates combine GLP-1 with another peptide, which contributes to our sense that GLP-1 agonism could eventually become a mainstay of obesity care (of course, this will also require overcoming stigma and the other commercial obstacles affecting this therapy class). Sanofi plans to launch two phase 3 trials of its GLP-1/glucagon dual agonist in people with overweight/obesity. OPKO Health, Novo Nordisk, J&J/Hanmi, and Zealand/BI are also investigating GLP-1/glucagon dual agonists for obesity (in phase 1 or 2). Novo Nordisk’s obesity pipeline additionally features a phase 1 GLP-1/GIP/glucagon tri-agonist. See our obesity drug competitive landscape for the complete picture.

New Semaglutide CVOT Could Establish Medical Obesity as a Chronic Disease; Obesity, Prediabetes as Independent CV Risk Markers

• As further proof of Novo Nordisk’s leadership in obesity, the company has announced plans to conduct the first-ever obesity CVOT. This is truly wonderful news that gives us hope for the future of obesity care, and we note that five years ago, no one would have thought that an obesity CVOT would be funded. The study will enroll ~12,500 patients with obesity (a remarkable clinical effort), randomizing them to GLP-1 agonist semaglutide or placebo. Although trial design is still under discussion with FDA, management expressed distinct excitement for this CVOT during Novo Nordisk’s late November Capital Markets Day. CSO Dr. Mads Thomsen underscored that no study to-date has established medical obesity as a chronic disease, as UKPDS and DCCT did for type 2 diabetes and for type 1 diabetes: “Novo Nordisk wants to be the company that does that landmark study.” This CVOT could show once and for all that people with obesity stand to benefit enormously from pharmacotherapy, especially a drug as potent and efficacious as semaglutide. The study could lead to a CV indication applicable to patients with obesity. We’d have access to hard data supporting the long-term impact of obesity on overall health, showing how the condition correlates with meaningful, costly outcomes like MI, stroke, and death. Results could translate into a vastly improved reimbursement landscape for obesity products. Moreover, we imagine a subset of participants in this CVOT will have prediabetes. Data could reveal semaglutide’s potential in diabetes prevention (as SCALE did for Saxenda), and we’re generally intrigued by the prospect of intervening even earlier in the course of hyperglycemia to lower CV risk, given this is ultimately the leading cause of mortality for diabetes patients, and perhaps for obesity patients as well. Indeed, many thought leaders are positioning prediabetes as a CV risk marker in and of itself, and are advocating for more screening and earlier intervention with a CV risk reduction strategy – Dr. John Buse spoke on this topic at Keystone 2017, while Drs. Jaakko Tuomilehto and Rury Holman shared similar perspectives at IDF 2017.

Strategic, Anti-Stigma Marketing Campaign Provides Major Boon for Orexigen’s Contrave Franchise

• On the strength of a new marketing campaign, Orexigen’s Contrave (naltrexone/bupropion extended-release) has pulled into a solid second place in the overall obesity market (behind the definitive leader, Novo Nordisk’s Saxenda). In 3Q17, Contrave captured 14% of the $139 million obesity market by value, trailing Saxenda (73%) but comfortably ahead of Vivus’ Qsymia (7%) and Arena/Eisai’s Belviq (6%). Notably, with $19 million in quarterly sales (a near-doubling YOY), Contrave held an impressive 50% of the $38 million obesity market excluding Saxenda. The product also posted strong sales in 1Q17 and 2Q17, with revenue growing 36% YOY and then 90% YOY. In contrast, Orexigen sales fell 7% YOY between 2015 and 2016, which leads us to believe that the marketing campaign launched at the beginning of 2017 has made a real commercial impact. This initiative includes direct-to-consumer advertising that emphasizes the neural mechanisms that make weight loss so difficult. These ads tackle the stigma surrounding medical management of obesity – this is such an important message, not only to stimulate greater uptake of Contrave, but to support obesity care more generally and to boost the obesity pharmacotherapy class as a whole. Orexigen has opened a telemedicine platform (currently available in 46 states) through which patients can consult a provider online and can receive Contrave by mail, allowing for both convenience and anonymity. Additionally, this marketing plan includes an employer activation program to improve insurance coverage for Contrave, as well as a physician activation program to boost the number of HCPs prescribing the product. Contrave now has nearly 100,000 unique prescribers in the US – quite solid for an obesity drug. In fact, Orexigen management shared in the company’s 3Q17 update that Contrave is the #1 prescribed branded weight loss drug (we attribute the discrepancy between this and the product’s comparatively low value share of the obesity market to the higher list price for Saxenda). We are certainly heartened to see this success story in the challenging obesity market, plagued by deep-seated social stigma around obesity, weight-based bias from providers, under-diagnosis of obesity, patient/provider reluctance to consider non-lifestyle interventions for weight loss, and HCP concerns regarding the safety of these agents. Only time will tell whether this DTC campaign can provide lasting momentum for Contrave. Certainly, much work remains to truly overcome all obstacles to obesity care, but we applaud Orexigen’s concerted efforts in 2017 and we look forward to seeing where the company goes from here.

Anti-Stigma Movement Gains Advocates, Evidence-Based Support: Insights from ACTION, Obesity in America Survey

• Though progress in overcoming weight-based stigma continues at a slow pace, 2017 did offer new data highlighting the pervasiveness of stigma and misconceptions of obesity, which is a necessary foundation for the anti-stigma movement. The latest findings from Novo Nordisk’s ACTION study were presented at Obesity Week 2017. This study, the first to systematically explore barriers to medical management of obesity, revealed that while most stakeholders – 63% of people with obesity (n=3,008), 80% of healthcare providers (n=606), and 62% of employers (n=153) – agree that obesity is a serious disease, there are egregious gaps in care. Only 55% of Americans with obesity have been formally diagnosed. A startling 82% feel “completely responsible” for their body weight, while only 72% of HCPs and only 18% of employers report feeling any responsibility to support the weight loss efforts of their patients or employees, respectively. Notably, 71% of people with obesity have spoken with an HCP about their weight in the past five years, and yet only 24% (!) were offered follow-up care. We can’t imagine this little follow-up as an acceptable standard of care for any other chronic illness. As Dr. Lee Kaplan put it, though obesity may be increasingly identified as a disease, it is far less often treated as one.
  
  • The American Society for Metabolic and Bariatric Surgery’s Obesity in America survey, unveiled at ENDO 2017, corroborates the findings from ACTION. While a majority of Americans surveyed (81%, n=1,509) consider obesity to be the most serious health problem facing the nation – tying cancer as the top issue ahead of diabetes (72%), heart disease (72%), mental illness (65%), and HIV/AIDS (46%) – only 38% consider obesity to be a disease. Troublingly, this misperception of obesity’s disease status was accompanied by misperception in what actually constitutes obesity: 89% of survey respondents whose BMI placed them in the range of obesity did not consider themselves to have obesity. Furthermore, respondents tended to overestimate the effectiveness of diet/exercise for long-term weight loss and to underestimate the effectiveness (and safety) of medical and surgical treatments: 78% of Americans consider diet/exercise to be the most effective weight loss method, with weight loss surgery (59%) and prescription obesity medications (26%) far behind.
  • Taken together, these studies underscore widespread misconceptions about obesity among the general public and obesity stakeholders alike. These findings are certainly consistent with the continuing commercial challenges in the obesity market and the uphill battle for widespread patient, provider, and public acceptance of medical interventions for obesity.
• As another sign of weight-based stigma, we noticed a glaring discrepancy in US obesity prevalence when calculated from self-report surveys vs. physical exam records. In September of this year, the CDC published maps of the US showing state-by-state obesity rate derived from self-report. West Virginia had the highest adult obesity prevalence (38%) in 2016, while Colorado had the lowest (22%). No state or territory had prevalence <20%. We were disheartened by these results – that is, until NHANES came out with its 2015-2016 obesity statistics ~two months later, and showed the problem to be even worse than we thought before. According to NHANES, the nationwide adult obesity rate was 40% between 2015-2016, which is notably higher than the highest statewide prevalence found via self-report. We took this as a signal that people are largely unaware of what constitutes medical obesity, and that persistent stigma precludes us from talking openly about weight and getting people the care they need. The individuals who don’t self-identify as having overweight/obesity, or who don’t recognize it in their peers/loved ones, will delay care and might miss an opportunity for effective obesity treatment and possible diabetes prevention. From the CDC’s September release, our team used a weighted average to calculate a national adult obesity rate of ~30%, which is 10% off from the true prevalence, per NHANES. Admittedly, our analytical methods weren’t as rigorous as those used in ACTION or the Obesity in America survey, but we point out that a 10% difference is not trivial, and highlights how stigma so desperately needs to be addressed.
• We were encouraged to see more attentiveness to obesity-related language this past year, which is another important step in tackling stigma. In July, the American Medical Association (AMA) approved a resolution calling on all healthcare providers to use unbiased, person-first language when interacting with patients who have obesity. The resolution recommends “a person with obesity” vs. “an obese person.” The mainstream media caught wind of this language movement with a New York Times article illuminating the difficult line providers must walk in addressing childhood obesity without contributing to the distress many children already feel around body weight. In line with a joint statement from the American Academy of Pediatrics and The Obesity Society, the article advises the use of “weight” and “body mass index” in place of “obese” and “fat,” as well as person-first language (e.g. “child with excess weight” rather than “obese child”). Earlier this year, AACE proposed an entirely new (and stigma-free) name for obesity: adiposity-based chronic disease (ABDC). Dr. Tim Garvey presented on this at AACE 2017, arguing that “obesity” is severely lacking as a label because it’s not actionable, it’s not informative for overall health and wellness, nor is it sufficiently specific. “Obesity” doesn’t tell you anything about adipose tissue mass or the long-term complications an individual patient will encounter. Transitioning language in this therapeutic area to talk about “adiposity-based chronic disease” could circumvent stigma and bias by medicalizing obesity and encouraging people to think about it in the same category as any other disease, with symptoms and adverse health effects. Indeed, promoting the concept of obesity as a biological disease is crucial in overcoming stigma. Orexigen launched a new marketing campaign this year that emphasizes the biological, physiological foundations of obesity – read more details above.

A Year of High-Profile Obesity Partnerships

• Several partnerships for the development of novel obesity therapies were either forged or solidified in 2017 – further proof that this vastly underserved therapeutic area is worth investing in. ReShape Lifesciences (formerly EnteroMedics, renamed after its acquisition of ReShape Medical in October) joined forces with Galvani Bioelectronics, a high-profile joint venture between GSK/Verily that aims to pioneer next-gen bioelectronic therapies for chronic disease. Galvani’s ambitious goal is the development of miniaturized devices that can be attached to individual nerves to precisely control the nervous system’s electrical signaling to a specific organ. Through this partnership, ReShape will modify its signature vBloc neuromodulation system for use in Galvani’s preclinical research. A pacemaker-like device that targets the nerve pathway between the brain and stomach to reduce hunger signaling, vBloc technology is a logical starting point for the eventual creation of a device with the ability to precisely “read” the distorted nerve signals underlying obesity and also “write” corrected firing patterns to moderate this electrical conversation. While it’s too early to know the odds of success, this innovative, potential new therapeutic modality has certainly piqued our interest, and we look forward to future updates from Galvani on the progress of these preclinical trials. We are happy to see investment from such a heavy-hitting company into obesity – where there is need for therapeutic options that fill the wide treatment gap between mildly effective lifestyle-based therapies and maximally effective but expensive (and poorly reimbursed), invasive, and side-effect-ridden options of metabolic surgery. In theory, bioelectronic medicine could reduce side-effects (because its action would be confined to precise neural circuits), though efficacy and future cost remain open questions. Bioelectronics also boasts ease of adherence since these implantable devices would require no daily interaction – a major win in terms of patient friendliness.
• Many notable partnerships announced this year are centered around early-stage obesity drugs. In November, BI and Gubra (a Danish biotech) teamed up to develop peptide-based treatments that regulate food intake. This complements BI’s collaboration with Zealand around a once-weekly GLP-1/glucagon dual agonist and a long-acting amylin analog (both in phase 1 for obesity), and positions BI to eventually commercialize combination therapies for obesity. Zealand/BI’s candidates each entered phase 1 clinical trials early in 3Q17, and topline results from these studies are expected in early-mid 2018, which gives us plenty to look forward to in the year ahead. Though Zealand/BI’s cardiometabolic partnership was established in 2014, the amylin target was disclosed only in March of this year, after being kept under wraps for more than two years as the companies conducted preclinical investigations. Dr. Daniel Drucker has posited that amylin may be the logical next step as a therapeutic target for type 2 diabetes/obesity, given its association with robust weight loss and improved cardiometabolic health. Novo Nordisk also has an amylin analog in phase 1 for obesity, with an ongoing study expected to complete in January 2018. We will be closely following the progress of these partnerships. While it’s exciting to see new players and partners arrive on the scene, committed to next-gen obesity pharmacotherapy, we acknowledge that all of these candidates face a long road ahead through clinical and commercial development – we can’t wait for the days when we’re writing about what these partnerships have accomplished, rather than their mere existence. Click here for our obesity drug competitive landscape.

Late-Stage Obesity Pipeline Now Includes Phase 3 Candidates from Saniona/Medix and Rhythm

• At the start of 2017, there were no phase 3 treatments featured in our obesity competitive landscape, and now we have two: (i) Saniona/Medix’s triple monoamine reuptake inhibitor tesofensine and (ii) Rhythm’s MC4R agonist setmelanotide. Saniona/Medix’s 24-week phase 3 trial (n=372) will be conducted in Mexico, with completion expected within two years of study start (to our knowledge, the trial has not yet begun). Rhythm’s 52-week phase 3 trial will evaluate the safety/efficacy of setmelanotide in pro-opiomelanocortin (POMC) deficiency obesity, a rare genetic condition that affects ~100-500 people in the US. The study is expected to complete in October 2018. Rhythm plans to launch an additional phase 3 trial in LepR deficiency obesity, another rare genetic condition. Insights from Rhythm’s ongoing Genetic Obesity Project demonstrate that a substantial proportion of people with obesity may have a form of the disease that is setmelanotide-responsive, suggesting that this agent could have much broader applicability down the line. Importantly, tesofensine and setmelanotide represent the most advanced candidates in the obesity competitive landscape, and both are positioned to be first-in-class. This points to increasing variety of obesity treatment options coming to market in the mid-term future, which is definitely exciting, although considering the sheer size of the global obesity epidemic, we should be writing about more than two late-stage therapy candidates on the eve of 2018. Our fingers are crossed for positive readouts from these phase 3 studies, and we also hope to see progress from phase 2 and earlier-stage candidates in the competitive landscape.

Tackling Obesity – and Diabetes – in Adolescents

• As childhood obesity prevalence continues to rise, we’re noticing more activity around tackling both obesity and diabetes in adolescents. This was reflected in the ADA’s 2018 Standards of Care, which feature a markedly expanded section on type 2 diabetes in children. ADA called out this addition near the very start of its press release on the new Standards, sending a clear message to the field: We need to pay more attention to type 2 diabetes in youth. Screening for type 2 diabetes is now recommended in children ≥10 years-old with a BMI in the 85th percentile or above and with at least one risk factor (family history, race/ethnicity, symptoms). While not frequently discussed, ~5,000 cases of youth-onset type 2 are diagnosed each year. We find it devastating that so many are living with diabetes at such a young age, putting them at higher risk for complications in the future. The rate of youth-onset diabetes is projected to increase four-fold by 2050, and it should be noted that gestational diabetes is the strongest predictor for this, adding another layer of complexity to this growing health concern (tackling childhood obesity and diabetes starts before a baby is even born). To be clear, the obesity epidemic is driving this disturbing increase in childhood type 2 diabetes, and we’re already seeing other obesity-related comorbidities exact their toll on the healthcare system through young people. What’s more, we imagine the problem is seriously underdiagnosed, as few pediatricians are likely taught to watch for type 2 diabetes in their patients. As of now, only metformin and insulin (in addition to lifestyle) are indicated for treating diabetes in this population <18 years-old – only orlistat is indicated to treat obesity – but we are seeing more investigations of these agents in adolescents.
  
  • DPP-4 inhibitors, SGLT-2 inhibitors, and GLP-1 agonists have all been investigated or are currently being investigated in children or adolescents with type 2 diabetes. Linagliptin (Lilly/BI’s DPP-4 Tradjenta) was not found to be significantly efficacious in reducing A1c in a 12-week phase 2 trial (n=40); to our knowledge, only phase 1 trials have been conducted with other DPP-4 inhibitors in adolescents. A currently-recruiting phase 3 trial (n=172) of SGLT-2 inhibitor canagliflozin (J&J’s’ Invokana) in adolescents age 10-18 with type 2 diabetes is expected to complete in 2021, but the most work has been done with GLP-1 agonists. The phase 3 Ellipse study (n=150) of liraglutide (Novo Nordisk’s Victoza) is expected to complete in May 2020, and AWARD-PEDS is currently recruiting participants for a phase 3 study (n=150) of dulaglutide (Lilly’s Trulicity). Two recruiting phase 3 trials aim to evaluate both exenatide once-weekly (AZ’s Bydureon; n=77) and twice-daily (AZ’s Byetta; n=195) in adolescents with type 2 diabetes (the latter study was initiated in 2008). Additionally, Novo Nordisk plans to conduct a trial of semaglutide (Ozempic) in adolescents, per approval requirements. We note that most of these trials are many years long but with treatment periods of ~six months, indicating to us either low priority within their sponsoring companies or difficulty in recruiting participants.
  • Novo Nordisk’s Saxenda, Arena/Eisai’s Belviq, and Vivus’ Qsymia are all under or have recently completed investigation in adolescents with obesity. A 56-week, phase 4 study of Qsymia in 42 adolescents with BMI in at least the 95th percentile for age and gender completed late in 2016, but results have yet to be posted. Arena/Eisai began a similar 52-week, 260-participant study this year, with expected completion in September 2019. A phase 3, 56-week, 228-partcipant trial of Saxenda is expected to complete in August 2019. Off-label generics have also been investigated in adolescents – for example, this phase 3 study of topiramate is ongoing, following modest efficacy in a phase 2 trial (n=34), though these studies have different non-industry sponsors.
  • Bariatric surgery in adolescents is a particularly controversial topic; that said, our sense is that the field is gradually coming to accept the value of weight loss surgery in younger people, especially in cases of severe obesity. Dr. David Cummings characterized bariatric surgery as the “only realistic hope” for adolescents with severe obesity, and emphasized that these procedures are safe overall, including in adolescents. Moreover, the cumulative lifetime risks of obesity are much higher than the risks of surgery. Some procedures are even reversible and do not cause nutritional deficiencies, though they are often less effective (e.g. the gastric band). As Dr. Sara Lappé noted at last year’s Cleveland Clinic Obesity Summit, “For severe obesity, don’t wait. They will not grow out of it, and it will get worse.” While bariatric surgery does not have the same robust, long-term data in adolescents as adults (there is some), we think it’s important to realize that obesity in children and adolescents has to be addressed. As with hyperglycemia, we believe taking swifter action with effective care – be that lifestyle intervention, pharmacotherapy, or surgery – is crucial to preventing downstream morbidity/mortality. We also hope to see more done to promote awareness and to destigmatize these chronic health issues in children as well as adults.

Most Highly-Read Reports

• CDC releases NHANES 2015-2016 obesity data, reporting 40% prevalence among adults and 19% among youth nationwide – Follows self-report obesity data (an underestimate) released last month; Childhood obesity climbs; Race persists as a risk factor, with non-Hispanic blacks and Hispanics disproportionately affected by obesity
• CDC releases state-by-state obesity data for 2016, accompanied by RWJF’s annual State of Obesity report ­– Obesity rates exceed 20% in all US states, ranging from 22% (CO) to 38% (WV); Obesity rates are over hit 38% and 33% in black and Hispanic populations vs. 28% for whites
• FDA approves label update for Saxenda to include three-year data from the SCALE trial – Data supports long-term safety/efficacy of Saxenda (liraglutide 3.0 mg); Company announcement also calls attention to Saxenda savings card
• Novo Nordisk’s semaglutide shows superior weight loss vs. placebo in phase 2 obesity trial – GLP-1 agonist associated with 13.8% body weight loss after one year of treatment vs. 2.3% on placebo; Topline data doesn’t report precise weight loss in Saxenda arm of the trial, probably ~8% from prior studies
• GSK/Verily-partnered Galvani enters research deal with EnteroMedics to investigate bioelectronic therapies for obesity – Galvani seeks to develop mini devices that modify the signaling of individual nerves to an organ; Future preclinical work in obesity will involve a modified version of EnteroMedics’ vBloc system, a pacemaker-like device that targets the nerve pathway between the brain and stomach
• JAMA study reports decline in rates of overweight/obese individuals trying to lose weight – From 1988-2014, the prevalence of overweight/obesity increased from 53% to 66% while rates of trying to lose weight among this group fell from 56% to 49%
• US medical licensing exams insufficiently test obesity, reflecting insufficient obesity education in medical schools – Recent study shows underrepresentation of obesity questions on medical licensing exams, with <1% addressing obesity management in practice 
• Milken Institute report examines the health and economic impacts of obesity – Direct and indirect medical costs of obesity totaled $1.42 trillion in 2014; Type 2 diabetes accounts for ~$320 billion of these costs
• Treat and Reduce Obesity Act proposes increased reimbursement for obesity care – Bipartisan bill would authorize Medicare coverage of obesity pharmacotherapy and counseling
• AMA adopts resolution to destigmatize obesity with person-first language – New guidelines call on all HCPs to engage in people-first dialogue, recognizing the person before his/her disease

What We Got Right/Expected

• Novo Nordisk’s Saxenda continues to lead the obesity class in sales, though its prescription volume is lower than Contrave’s due to a much higher list price. To this end, we’re eager to learn, more precisely, how many patients are on Saxenda and how much they pay out of pocket. Saxenda contributed a 14% share of growth to Novo Nordisk’s diabetes/obesity portfolio in 3Q17, despite accounting for only 3% of total portfolio sales.
• Semaglutide shows strong efficacy in phase 2 obesity trials. This was to be expected, given the profound weight loss benefit seen throughout the type 2 diabetes program. Novo Nordisk’s second-gen GLP-1 agonist will advance into phase 3 for obesity next year.
• Sanofi enters obesity with its GLP-1/glucagon dual agonist. This follows a movement of major diabetes companies expanding their R&D into diabetes-adjacent indications, recognizing obesity in particular as a vastly under-treated condition (along with NASH).
• Obesity prevalence rises, for both children and adults. According to NHANES data for 2015-2016, childhood obesity rate was ~19% in the US while adult obesity rate was ~40%.
• Novo Nordisk’s Cities Changing Diabetes (CCD) program emphasizes obesity. CCD has called on all cities to reduce obesity rate 25% by 2045. The ultimate goal of this call-to-action is to keep global type 2 diabetes prevalence below one in 10, and obesity as a risk factor accounts for 44% of diabetes burden.
• The ACTION study highlights major misconceptions of obesity as a disease, among patients, providers, employers, and the public.
• Thought leaders continue to advocate for anti-stigma efforts. Despite this, stigma persists in society, among patients, and (unconsciously) among providers.

What We Got Wrong/Did Not Expect

• Novo Nordisk commits to an obesity CVOT with GLP-1 agonist semaglutide (newly-approved for type 2 diabetes as Ozempic). This was some of the best news we heard all year on the obesity front. This study could show CV benefit to semaglutide in a broader (non-diabetes) population, but what’s even more exciting is that it could establish obesity as a medical disease with clear health consequences.
• Orexigen’s Contrave experiences surprise commercial success after a difficult 2016. Management attributes this to new anti-stigma marketing campaign.
• Merck teams up with Orexigen to market Contrave in Latin America, which could be a boon for the global franchise and for the fight against obesity in Latin America (especially Mexico). Merck’s diabetes portfolio multiplied in size this year (three Pfizer-partnered products for type 2 were just-approved), and it’s noteworthy that the company has made this investment in obesity as well.
• Zealand/BI choose to focus their cardiometabolic health partnership on obesity. This collaboration was announced back in 2014, and this May, we finally heard that amylin is the therapeutic target and obesity is the target indication.
• AACE recommends a switch from “obesity” to “adiposity-based chronic disease” (ABCD). To our knowledge, this term hasn’t seen significant uptake in the real world, but we were pleasantly surprised that AACE took a stand on obesity language and offered such a specific suggestion. We believe ABCD could make waves in reducing stigma if it becomes widely-used.
• Mainstream media picks up on person-first language movement. We encourage everyone to read this powerful NYT piece (our coverage, including expert commentary from Dr. Lee Kaplan, is here).
• Lilly moves forward with high-dose dulaglutide for type 2 diabetes, but not obesity. The company has historically maintained a tight focus on type 1 and type 2 rather than obesity. Still, we thought this might change in 2017 as management emphasized the high-dose dulaglutide formulation, given the efficacy of high-dose GLP-1 for obesity and the movement shifting R&D toward diabetes-adjacent indications. Certainly, both Novo Nordisk and Sanofi are showing commitment to obesity as a therapeutic area.
• CVS announces a new initiative for 2018 that will match obesity drug rebates to weight-based outcomes. Details were left rather vague, but we still found it quite notable that a major US formulary is taking a stab at outcomes-based reimbursement for obesity care.

Key Questions for 2018

• Which of Novo Nordisk’s six phase 1 obesity candidates will be prioritized for further clinical development? As of the company’s 3Q17 update, three of these agents (FGF21 analog, GLP-1/GIP/glucagon tri-agonist, glucagon analog) had completed phase 1 studies. Other phase 1 candidates include a PYY analog, an amylin analog, and a GLP-1/glucagon dual agonist.
• How will phase 3 trials fare for tesofensine (Saniona/Medix) and setmelanotide (Rhythm)? Could these agents be new obesity pharmacotherapies by 2019 or 2020? (We’re not expecting particularly swift regulatory filing, and for now, we’re mostly curious to see the safety/efficacy data from phase 3.)
• Which target(s) will BI/Gubra choose as the focus of their new obesity partnership? Will BI start developing combination therapies, considering its expanding obesity pipeline (also Zealand-partnered amylin analog and GLP-1/glucagon dual agonist)?
• What will come of the GSK/Verily + ReShape collaboration? Will we see any promising data on neuromodulation techniques for obesity treatment?
• Will the field start using AACE’s suggestion of “adiposity-based chronic disease” (ABCD) instead of “obesity”? What other milestones will be achieved for the language movement? Can we promote a better, more accurate understanding of obesity as a biological, medical condition? 
• Will additional professional societies take a stand on obesity as a biological, treatable disease? Will this reduce unconscious provider bias, and/or societal stigma?
• What will CVS Health’s “Transform Obesity Value” initiative look like? CVS Health will launch this program in 2018, tying obesity drug rebates to weight-based outcomes, though concrete details have been scarce.
• Can industry players work together to move the needle on reimbursement for obesity drugs? What additional data do payers want to see? What constitutes a convincing argument of long-term cost-savings?
• Will the field make any progress on turning microbiome research into new therapeutic targets for obesity? “Microbiome” has certainly grown into a buzzword, and it’s the topic of many talks at diabetes/obesity conferences, but we’re anxious for progress that helps patients.
• How can we improve how obesity care is taught in medical schools? The same applies to diabetes, of course, though the situation seems particularly bleak around obesity.
• Will the Treat and Reduce Obesity Act experience any progress in Washington? The bipartisan bill hasn’t seen any action since it was introduced to the House in mid-2015, but the implications are huge: The Act would authorize HHS to cover obesity pharmacotherapy, counseling, and diet/exercise programs under Medicare Part D for people with obesity and at least one related comorbidity. Novo Nordisk has expressed support for the bill, which was also highlighted in a Huffington Post op-ed earlier this year.

What’s Coming in 2018?

The list below covers late-stage pipeline products with expected milestones in 2018. We acknowledge that this list may not be 100% complete, though we’ve tried to be as comprehensive as possible based on the most recent public updates. If you notice anything missing, please write us.

• Novo Nordisk’s semaglutide: The once-weekly GLP-1 agonist will advance into the phase 3 STEP program (n=~4,500 patients in total). STEP 1 will investigate weight loss in people with obesity, STEP 2 aims to show weight loss efficacy in people with type 2 diabetes not on insulin, STEP 3 will combine semaglutide with intensive lifestyle intervention in an attempt to maximize weight loss, and STEP 4 will evaluate semaglutide for weight loss maintenance. Each study is 68 weeks in duration, scheduled to begin in 1H18, with estimated completion dates in 2020. Most exciting is that Novo Nordisk also has a CVOT planned for semaglutide in people with obesity.
• Sanofi’s GLP-1/glucagon dual agonist (SAR425899): Phase 3 studies in obesity will begin in 2H18. This marks Sanofi’s first foray into obesity therapy.
• Rhythm’s setmelanotide: A phase 3 trial is expected to complete in October 2018, though this investigates the agent specifically in pro-opiomelanocortin (POMC) deficiency obesity.
• Saniona/Medix’s tesofensine: The companies announced that a 24-week phase 3 trial will be conducted in Mexico, though we await the official study start – hopefully early in 2018.
• Obesity drug competitive landscape: This page offers an overview of all the obesity drugs in development, to the best of our knowledge. We frequently update this table as timelines change and as new developments arise.

-- by Adam Brown, Ann Carracher, Abigail Dove, Brian Levine, Payal Marathe, Maeve Serino, and Kelly Close