CODHy 2018

February 21-22, 2018; Tel Aviv, Israel; Day #1 Highlights – Draft

Executive Highlights

  • We have six outstanding highlights from our first day at CODHy 2018 in Tel Aviv, starting off with Dr. Mikhail Kosiborod on a CV disease-based paradigm for treatment selection. Later, Dr. Enzo Bonora advanced a similar sentiment of focusing on CV and renal effects when choosing therapy – it’s so exciting to witness this collective shift in thinking away from only glucose control.
  • Dr. Ralph DeFronzo debuted the “Fab Four.” This quartet of therapies includes SGLT-2 inhibitors, GLP-1 agonists, TZD pioglitazone, and metformin, comprising the only diabetes drugs that Dr. DeFronzo believes are worth using. Read on below for his take on each of these therapies, including some interesting commentary on the misunderstanding of pioglitazone. We loved the simplicity of this approach.
  • Prevention received notable airtime, which was great to see: Esteemed prevention expert Dr. Jaakko Tuomilehto explored the unexpected controversies surrounding screening for diabetes and prediabetes, and a lively prevention-focused session delved into the “evidence-free zone” of repurposing diabetes drugs (namely SGLT-2 inhibitors, GLP-1 agonists, and DPP-4 inhibitors) for prevention and the regulatory challenges surrounding this. In our view, this should all be about risk-stratification – for those careening toward a type 2 diagnosis, treatment of prediabetes should be different from someone at low risk for many years.

Day #1 of a two-day CODHy 2018 has wrapped up here in Tel Aviv, and we’re bringing you our top six highlights from the day, with commentary from Drs. Mikhail Kosiborod, Ralph DeFronzo, Jaakko Tuomilehto, and many more!

In case you missed it, here’s our CODHy preview for a look at what we’ll be seeing tomorrow – highlights include Dr. Simon Heller on hypoglycemia, Prof. Moshe Philip on progress toward an artificial pancreas, and a serious presence of digital therapy on the agenda.

Top Six Highlights

1. Dr. Kosiborod Outlines Compelling Case for Choosing Diabetes Therapy Based on CV Status; Argues in Favor of GLP-1 and SGLT-2 CV Class Effects

Dr. Mikhail Kosiborod presented a new diabetes management paradigm, where treatment is determined based on a patient’s status of CV disease – see the image below for an overview. He arrived at this algorithm based on findings from CVOTs of GLP-1 agonists, SGLT-2 inhibitors, DPP-4 inhibitors, and TZD pioglitazone. For people with CV risk factors only, Dr. Kosiborod suggested that GLP-1s and SGLT-2s are appropriate. Individuals with recent stroke should take pioglitazone, per his guidelines, while GLP-1s and SGLT-2s are also ideal for diabetes patients with a history of stroke/MI or with established coronary disease. Why should therapy be chosen according to the CV disease spectrum? As Dr. Kosiborod put it, CV disease remains the leading cause of death for people with type 2 diabetes, and it’s well-accepted that a diabetes diagnosis leads to residual CV risk. While comorbidities, contraindications, and A1c/glycemic targets should also be factored into treatment decisions, Dr. Kosiborod pointed out happily that the agents we now know to reduce CV risk are also some of the more effective glucose-lowering drugs. This paradigm shift in diabetes care, with cardioprotection at the center, is starting to take hold: The ADA’s 2018 Standards of Care contains a formal recommendation for cardioprotective drugs (SGLT-2 empagliflozin and GLP-1 liraglutide are specifically called out) in people with established CV disease. Despite the fact that better glucose control has not been empirically linked to improved CV outcomes or lower all-cause mortality, we do see value in the two-for-one combination of glucose-lowering and cardioprotection, especially when it comes to convincing regulators, payers, and diverse HCPs of the utility of these newer agents. As Dr. Kosiborod explained, the way you lower glucose is extremely important: DPP-4 inhibitors don’t improve CV outcomes, but GLP-1 agonists and SGLT-2 inhibitors do.

  • Dr. Kosiborod also spoke in favor of cardioprotective class effects for GLP-1 agonists and SGLT-2 inhibitors. He defined a class effect as results from multiple trials going in the same direction. Across LEADER (liraglutide), SUSTAIN 6 (semaglutide), ELIXA (lixisenatide), and EXSCEL (exenatide), GLP-1 agonists have given a 10% relative risk reduction (HR=0.90, 95% CI: 0.82-0.99, p=0.033) in three-point MACE (non-fatal MI, non-fatal stroke, CV death) and a 12% relative risk reduction in all-cause mortality (HR=0.88, 95% CI: 0.81-0.95, p=0.002) – Dr. Kosiborod reviewed this meta-analysis, which we also saw at IDF. The notion of a cardioprotective class effect around GLP-1 agonists has generated some controversy in the field since AZ’s Bydureon (exenatide once-weekly) just missed the threshold for MACE superiority in EXSCEL (HR=0.91, 95% CI: 0.83-1.00, p<0.001 for non-inferiority, p=0.06 for superiority). Since then, however, we’ve heard several experts attribute EXSCEL’s neutral result to its pragmatic trial design, and thought leaders seem to be leaning toward class effect (this is certainly our view, and Dr. Ralph DeFronzo argued at WCTD 2017 that EXCEL was “probably positive”). We were pleased to hear additional commentary in support from Dr. Kosiborod. Consensus on CV benefit to GLP-1s could improve uptake of all products in this class.
  • For SGLT-2 inhibitors, Dr. Kosiborod emphasized that both CANVAS (canagliflozin) and EMPA-REG OUTCOME (empagliflozin) demonstrated significant reductions in three-point MACE with the SGLT-2 vs. placebo. That said, canagliflozin (J&J’s Invokana) was not found to significantly reduce CV or all-cause death, and empagliflozin (Lilly/BI’s Jardiance) is now indicated for the reduction of CV death. Given this mild heterogeneity, Dr. Kosiborod suggested that all eyes are on DECLARE, the CVOT for dapagliflozin (AZ’s Farxiga) expected to read out in 2H18. DECLARE will be the largest and longest CVOT when it completes in July 2018, enrolling only ~40% of participants with baseline CV disease, compared to 100% in EMPA-REG and 66% in CANVAS. We hope there is lots of sub-group analysis. Moreover, DECLARE is the first-ever CVOT with a co-primary endpoint that includes heart failure, so results could help clear up (i) what CV benefit SGLT-2 inhibitors offer in primary prevention and (ii) the impact of SGLT-2s on heart failure. We look forward to both these insights, and the latter could help fill an unmet need in diabetes care – as you can see on Dr. Kosiborod’s graphic below, he listed symptomatic heart failure, hospitalization for heart failure, and end-stage heart failure as “not adequately studied.”  To share optimism for positive DECLARE results, Dr. Kosiborod pointed out that CVD-REAL (for which he was a primary investigator) found significant mortality and heart failure benefits  with SGLT-2 inhibitors in both initial and recurrent heart failure patients in the real world. We’re eager for findings from CVD-REAL 2 (encompassing Asia Pacific, Canada, and the Middle East), reporting at ACC 2018 in just a couple of weeks!

2. Dr. DeFronzo Debuts the “Fab Four”: SGLT-2s, GLP-1s, Pioglitazone, Metformin; Emphasizes Paramount Need to Target Beta Cell Dysfunction in Type 2 Diabetes Care; Denounces Sulfonylureas

Dr. Ralph DeFronzo challenged diabetes treatment algorithms and current prescription habits by introducing his “Fab Four” of therapy – SGLT-2 inhibitors, GLP-1 agonists, TZD pioglitazone, and metformin. He particularly praised GLP-1 agonists: “Why would you not use this as first-line therapy in every single one of your patients?” While this line of thinking isn’t new from Dr. DeFronzo (read about his ringing endorsements of GLP-1s, SGLT-2s, and pioglitazone at last year’s CODHy gathering in Buenos Aires and at the WCTD conference in Berlin), this was the debut of the term “Fab Four.” We have to say: The name is catchy, highlighting the superior effectiveness of more advanced drug classes (GLP-1s, SGLT-2s), as well as the growing importance of cardioprotection in diabetes care. As Dr. DeFronzo pointed out, gut peptides GLP-1 and GIP account for 60%-70% of the insulin secreted after a meal, making them absolutely essential for normal glucose tolerance. Moreover, they counteract beta cell failure, which is an important role of anti-hyperglycemic therapy and is all but necessary for sustained A1c reduction. Dr. DeFronzo pointed to a 2003 study showing that liraglutide (Novo Nordisk’s GLP-1 agonist Victoza) returns beta cell function to the levels expected of someone without diabetes. He expressed distinct enthusiasm for Novo Nordisk’s oral semaglutide as well as Intarcia’s ITCA 650 (implantable exenatide mini pump). While we think most patients (and providers) would be resistant to first-line therapy with an injectable, we also fully appreciate and endorse the ethos of earlier, more aggressive treatment to fight diabetes progression. Oral semaglutide and an implantable mini pump that lasts for three-six months (initiation phase and maintenance phase, respectively) could very well increase the number of patients who start on GLP-1 earlier in the course of disease, if not first-line, given the reduced or eliminated injection burden. That said, before we get too far ahead of ourselves, we note that awareness of and reimbursement for GLP-1s will have to improve remarkably to increase prescription volume. According to a Diabetes Care article published last year, only 7% of second-line diabetes prescriptions in the US go to a GLP-1 agonist; the majority (46%) still go to an SU, and only 7% go to an SGLT-2 inhibitor.

  • Turning to SGLT-2 inhibitors, Dr. DeFronzo highlighted a key advantage of SGLT-1/2 dual inhibitors (like Sanofi/Lexicon’s sotagliflozin) – these dual inhibitors block nearly 100% of glucose reabsorption, while SGLT-2 inhibitors only block ~55% of reabsorption. Sotagliflozin is in phase 3 for type 2 diabetes, and we’re interested to see what glycemic and weight loss data look like in patients with type 2. Notably, the agent will also be filed with FDA and EMA shortly for a type 1 indication (FDA submission is scheduled for 1Q18, per Sanofi’s 4Q17 update).
  • Dr. DeFronzo maintained his status as one of diabetes’ most outspoken proponent of TZDs (specifically, pioglitazone). While pioglitazone gives a ~35% increase in insulin sensitivity (and is the only available drug for sensitization), he explained that he sees the TZD as most useful for increasing insulin secretion. Pioglitazone has also been shown to reduce MI/stroke in people with and without diabetes, in PROactive and IRIS. On the horizon, Dr. DeFronzo forecast a serious gap in the development of new insulin sensitizers. He acknowledged early-stage Lyn kinase activators, but said he is otherwise disappointed at the lack of therapeutic options for this aspect of diabetes pathophysiology, for this aspect of the ominous octet.
    • In Dr. DeFronzo’s view, pioglitazone and its effects continue to be misunderstood. Dr. Itamar Raz challenged Dr. DeFronzo’s argument for combination therapy, noting that pioglitazone worries some doctors due to lingering concern over heart failure. Dr. DeFronzo seemed quite disappointed at this “misperception.” He pointed out that pioglitazone is actually a potent insulin sensitizer in the myocardium, and has been shown to markedly improve both systolic and diastolic function. However, as the label stipulates, pioglitazone should not be given to those with class III or IV heart failure because it causes sodium retention, increasing fluid volume and “[throwing people into heart failure]." Dr. DeFronzo actually cast doubt over whether some participants in PROactive had heart failure, as it was never adjudicated – he attributed some events to fluid retention. As Dr. DeFronzo has previously stated, the weight gain associated with pioglitazone is also partially due to fluid retention and can be combated with distally-acting diuretics. Moreover, lower doses of pioglitazone – never above 30 mg – are preferred, and weight gain can even be countered with a GLP-1 agonist. Other thought leaders have similarly advocated for low-dose pioglitazone at 15 mg (never above 30 mg) for type 2 diabetes, including Drs. Jay Skyler, Robert Eckel, and George Bakris at CMHC 2016. In our view, it’s important to think about how patients experience and perceive weight gain, because this can lower treatment satisfaction, adherence, and overall health outcomes. We hope providers are vigilant in explaining and treating these effects when prescribing pioglitazone, and we’d love more concrete guidance for HCPs on what this communication should look like.
  • Dr. DeFronzo once again denounced sulfonylureas and over-reliance on metformin, noting that they correct none of the triumvirate leading to hyperglycemia (impaired insulin secretion, decreased glucose uptake, increased hepatic glucose production) and none of the ominous octet. A compilation of long-term studies on SUs demonstrates that they are completely ineffective in the long term, Dr. DeFronzo explained. This is a crucial point – many clinical trials, and even head-to-head studies, show strong A1c-lowering efficacy with sulfonylureas because of study duration <two years, but beta cell burnout is a known side-effect of SUs and this attenuates their efficacy over time. Interestingly, Dr. DeFronzo laid out a similar argument for metformin. He suggested that metformin is commonly misperceived as an insulin sensitizer (rather, the generic drug prevents gluconeogenesis at the liver). He pointed out that UKPDS is traditionally cited as evidence for metformin use, but the CV analysis contained only 342 participants and A1c slowly rose over time, surpassing baseline levels at ~five years. Indeed, we were somewhat surprised by Dr. DeFronzo’s inclusion of metformin in the Fab Four – he’s called its use “archaic” in the past – though he later clarified it was the least important of the four. Metformin has been incredibly well studied, is remarkably affordable (just dollars a month), and has a mild side-effect profile, keeping it at the top of treatment algorithms. In the end, even Dr. DeFronzo isn’t ready to say goodbye to metformin, though he argued that it should only be used in combination with other agents. We imagine the inclusion of metformin could actually promote faster acceptance of the Fab Four within the diabetes world (patients, providers, and payers alike), since the agent is so familiar to this community.

3. Panel Delves into the “Evidence-Free Zone” of Repurposing Diabetes Drugs for Prevention; Arguments for SGLT-2s, GLP-1s, and DPP-4s; Regulatory Challenges Persist

During an afternoon panel, Drs. John Wilding, Agostino Consoli, and Baptist Gallwitz discussed the plausibility of using SGLT-2 inhibitors, GLP-1 agonists, and DPP-4 inhibitors to treat prediabetes. Said Dr. Wilding: This is an evidence-free zone, but that doesn’t prevent any of us from having something to talk about.” We love this perspective, and we agree that productive conversation should continue around possible prediabetes drugs, despite resistance from regulators to consider a prediabetes indication (to be sure, there are challenges here, but many thought leaders have criticized FDA’s hesitation to approve metformin for prediabetes). Outlining the case for SGLT-2 inhibitors, Dr. Wilding cited compelling reductions in A1c and body weight (a key risk factor for type 2) with these agents in the diabetes population, though there are no published trials to-date on the use of SGLT-2s in prediabetes. Additionally, Dr. Wilding pointed to the much-discussed CV benefit of two SGLT-2 inhibitors – Lilly/BI’s Jardiance (empagliflozin) and J&J’s Invokana (canagliflozin) – as seen in EMPA-REG OUTCOME and CANVAS. He suggested that these benefits for people with diabetes at high CV risk could very well extend to the prediabetes population. Risk reduction for atherosclerotic CV events might be somewhat muted if prediabetes patients face lower CV risk to start; then again, prediabetes is a known risk factor for CV disease, and these individuals may already have significant atherosclerosis. Moreover, there have been some hints that SGLT-2s may exhibit a heart failure benefit entirely independent of glycemia: A post-hoc of CANVAS found significant risk reduction for heart failure in the primary prevention cohort (whereas three-point MACE was not significantly reduced among these lower-risk patients), and Dr. David Fitchett shared data from EMPA-REG OUTCOME attributing empagliflozin’s heart failure benefit to volume effects, which apply outside the context of diabetes. Dr. Subodh Verma has explained that diastolic dysfunction (a precursor to heart failure) appears earlier in the course of diabetes progression than does atherosclerosis, defending earlier use of SGLT-2 inhibitors – possibly even in prediabetes. All this said, Dr. Wilding returned to the issue of plausibility. He noted that SGLT-2s for prediabetes is a milestone “unlikely to happen in the foreseeable future,” given the time it would take to complete a large-scale trial in prediabetes and the steep cost of SGLT-2 inhibitors relative to other candidate drugs for prevention, namely metformin. There may be some opportunity to investigate the impact of SGLT-2 therapy in prediabetes within the EMPEROR and Dapa-HF trials, of course, and we very much hope this is happening. These studies are recruiting participants with chronic heart failure, with or without type 2 diabetes, so presumably, there will be a cohort of patients with prediabetes at baseline randomized to SGLT-2 inhibitor (Jardiance or Farxiga) or to placebo. We’re eager to find out how these advanced oral agents perform as heart failure treatment distinct from diabetes and glucose-lowering, and to probe for effects in the prediabetes population. We’ll have to wait until June 2020 for the EMPEROR program to wrap up, while Dapa-HF is expected to complete in December 2019. We are counting the months …

  • Dr. Consoli presented a similar argument for GLP-1 agonists, adding that beyond glycemic and weight loss benefits, this therapy class also boasts strong evidence for the potential to improve beta cell function and “comfort” the beta cell failure that drives type 2 pathogenesis. That said, he acknowledged that it is still an open question if GLP-1 agonist therapy can truly restore beta cell function and prevent beta cell failure – we would be happy, of course, for some preservation of beta cell function. As for investigating a GLP-1 molecule in prediabetes, we see opportunity in the upcoming SELECT CVOT of Novo Nordisk’s semaglutide in people with obesity (n=~17,500). This trial is slated to start sometime in 2018, and will presumably enroll a decent cohort of patients with baseline prediabetes, allowing for interesting sub-group analyses. Novo Nordisk’s SCALE program has already shown high-dose liraglutide (branded Saxenda for obesity) to effectively delay type 2 diabetes onset over three years in people with prediabetes – so, perhaps we shouldn’t call this area an entirely “evidence-free zone,” though certainly much more research is needed.
  • Given the particularly high cost of SGLT-2 and GLP-1 products, we found Dr. Gallwitz’ argument for DPP-4 inhibitors to be quite compelling. As another incretin-based therapy (alongside GLP-1s), these agents could potentially foster a similar beta cell-protective effect in a more adherence-friendly oral (rather than injectable) formulation and at lower cost. We add that generic DPP-4s are closer around the corner than are generic GLP-1s, and widespread access will be key in promoting any prediabetes therapy (given that this patient population is even more numerous than the type 2 diabetes patient population). On the other hand, DPP-4 inhibitors are less potent than GLP-1 agonists and do not come with weight loss benefit or CV/renal protection, though they do boast a solid safety profile and excellent tolerability.
  • With rising prediabetes prevalence (estimated at 84 million adults in the US alone, according to CDC) we imagine the pursuit of a prediabetes indication could be a worthy investment for any company with an SGLT-2 inhibitor, GLP-1 agonist, or DPP-4 inhibitor on the market. That said, we understand that the lack of a clear regulatory pathway for prediabetes (which is not yet classified as a disease) represents a substantial obstacle – for now. In all likelihood, it will take collaboration from multiple manufacturers to sway regulatory authorities in the US and elsewhere to consider any prediabetes indication, regardless of therapy class.
  • It’s also worth mentioning that the ACE CVOT recently showed the potential for diabetes prevention with alpha-glucosidase inhibitor acarbose (Bayer’s Glucobay, also generic). Though Glucobay did not meet any of its CV endpoints in the study, it showed a compelling 18% risk reduction vs. placebo for the secondary endpoint of new-onset type 2 diabetes (HR=0.82, 95% CI: 0.71-0.94, p=0.005) in a Chinese population with prediabetes and a previous history of CV events (n-6,526). We found it notable that acarbose was barely mentioned in this discussion (Dr. Wilding did mention it in passing), because it is low-cost and it’s commonly-used in China, where prediabetes affects an astonishing 500 million people (half the adult population). We’d expect this from a meeting like CODHy with such a global focus, especially. That said, we understand the GI side-effects of acarbose make it an unpopular therapeutic choice for many patients in Europe and North America.

4. CV/Renal Benefits Should Dictate Therapy Selection in Diabetes, According to Dr. Enzo Bonora; Uncertainty Abounds in How to Treat Primary CV Prevention Patients (Until DECLARE and REWIND Report?)

Echoing earlier sentiments from Dr. Mikhail Kosiborod, Dr. Enzo Bonora argued that the number of different therapy classes now available enables genuine personalization of diabetes care, but that CV and renal benefits should dominate therapy selection (as below). Dr. Bonora called this “treat to benefit” rather than “treat to target,” which we love and will be repeating. He drew the most significant distinction in therapy choice between primary and secondary prevention of CV disease. Evidence for secondary prevention (i.e. those with established CV disease) is quite robust given positive SGLT-2 and GLP-1 CVOTs, despite the fact that this reflects only ~25% of patients with type 2 diabetes in the real world. Nevertheless, Dr. Bonora characterized the evidence as quite strong for adding SGLT-2 inhibitors, GLP-1 liraglutide (Novo Nordisk’s Victoza), or TZD pioglitazone to metformin (unless not indicated/contraindicated) to prevent subsequent CV events in these patients with a prior history. In contrast, he suggested that there’s a relative dearth of CVOT evidence when it comes to patients without established CV disease: While there are many options to choose from, there are few certainties and few head-to-head comparisons to inform decision-making. By our count, all reported CVOTs of GLP-1 agonists and SGLT-2 inhibitors enrolled 66%-100% secondary prevention cohorts, with most enrolling >80% in secondary prevention. As such, choosing therapy for primary prevention is complicated and less evidence-based right now (studies like DECLARE for AZ’s SGLT-2 dapagliflozin and REWIND for Lilly’s GLP-1 dulaglutide could start to change this, as both CVOTs – reporting this year – have recruited a majority primary prevention population). For the primary prevention patient today, an HCP might balance various benefits (glucose-lowering, CV risk factor improvement, microvascular disease prevention) and risks (hypoglycemia, other adverse effects) – there’s room for more personalization here, but also more room for interpretation.

  • Interestingly, Dr. Bonora contextualized this current focus on CV and renal outcomes as the next step in a long series of treatment foci, starting with an emphasis on body weight in the 1980s and progressing through postprandial glucose, insulin resistance, beta and alpha cell dysfunction, and then multiple targets and hypoglycemia before present-day. This historical perspective piqued our interest. What might be the focus of diabetes treatment in 10 or 20 years, as science furthers our understanding of the etiology and mechanisms underlying the disease? For example, what role will inflammation play in the future? In the meantime, we think CV disease, CKD, and heart failure are particularly efficacious targets for reducing morbidity and mortality, and we’d still like to see greater focus on hypoglycemia risk reduction in diabetes care. Dr. Bonora has developed the flow chart below to illustrate the concept of treatment personalization with respect to these various factors. We imagine most providers use similar mental heuristics when selecting therapy, but we would love to see more detailed and formal decision support tools aimed at medication selection.

5. Dr. Jaakko Tuomilehto on Diabetes Screening: “It Sounds Simple, But It’s Not”

In a compelling session on diabetes prevention, Dr. Jaakko Tuomilehto discussed the surprising controversy that surrounds diabetes screening and its potential benefits. As the lead investigator on the first-ever diabetes prevention RCT, the Finnish Diabetes Prevention Study, Dr. Tuomilehto is an esteemed expert on prevention, a field that is still very much emerging. When it comes to screening for hyperglycemia, Dr. Tuomilehto highlighted several unresolved questions: What measure or combination of measures should be used as the basis of diabetes screening? Should we screen for prediabetes as well as overt diabetes? How often should screening take place within a population, or for an individual? Perhaps most importantly, does screening actually produce long-term improvements in health outcomes?

  • On which screening method to use, Dr. Tuomilehto expressed a preference for the one-hour, 50 g oral glucose tolerance test, which is low-cost and shows excellent sensitivity in identifying people with undiagnosed diabetes (whereas A1c and fasting plasma glucose, though easier, may occasionally miss some diabetes cases). We are wondering about a two-week blinded wear of Libre Pro …
  • On the nuances of “to screen or not to screen,” things are less conclusive, according to Dr. Tuomilehto. He noted that while prevention studies (his own Finnish DPS, the US DPP study, and the Chinese Da Qing study) show convincingly that screening for prediabetes can prevent or delay the onset of type 2, it has not been definitively shown that screening for diabetes itself improves hard outcomes. Though it seems intuitive that diabetes screening would lead to earlier detection and therefore a wider therapeutic window to slow disease progression, Dr. Tuomilehto pointed out subtleties in how we assess the outcomes of screening programs. Due to the phenomenon of “length bias,” screening can appear to correlate with better prognosis as a byproduct of the fact that screening tends to preferentially pick up slow progressors (since they have a longer window of opportunity to be screened), whereas rapid progressors with a more aggressive form of the disease are more likely to already have a clinical diagnosis and therefore not be included in the screening cohort. Also at play is “lead time bias,” the idea that early disease detection only prolongs the length of time a person is aware that they have a disease, not necessarily the length of time they live with the disease (i.e. two people who live until age 80 with diabetes could “survive” 30 years or 10 years with the disease depending on whether they were diagnosed at age 50 or age 70, without any difference in actual lifespan).
  • “Screening sounds simple, but it’s not.” This session certainly opened our eyes to the hidden complexities of diabetes screening, and we’re left with the open question of whether screening programs are actually the best avenue to diabetes prevention. Still, we note that under-diagnosis of diabetes (and obesity, for that matter) is a serious public health problem. Education and awareness should not be overlooked or de-prioritized, even though there are kinks to be worked out. For instance, we’d love to find solutions to “length bias” (could better detection methods identify even the rapid progressors?) and “lead time bias” (this seems very solvable, by setting standards for data analysis in screening/prevention studies), rather than shying away from screening altogether.

6. NASH: The Epitome of “Controversies to Consensus” at CODHy

Israeli gastroenterologist Dr. Ran Oren outlined the numerous disputes surrounding NAFLD/NASH (fitting for a conference on “controversies to consensus”). As a relatively uncharted and under-recognized disease state, NAFLD/NASH is riddled with controversy at every level – from screening to diagnosis to management to recommended health systems approaches to the growing epidemic of fatty liver disease. Dr. Oren hit on the following key issues:

  • Defining NAFLD. The current definition (>5% liver fat in people who drink little to no alcohol) is very vague and does not systematically encompass any concrete measure of alcohol intake to differentiate NAFLD from alcoholic liver diseases.
  • NAFLD vs. NASH. Less urgency surrounds NAFLD than full-fledged NASH, but Dr. Oren argued that NAFLD deserves serious attention in and of itself, given that it’s a precipitating risk factor for the more severe NASH. Emerging evidence suggests that improvement or resolution of NAFLD can have beneficial effects for comorbid diabetes, all the more reason that NAFLD should be taken seriously in its own right.
  • Understanding the prognosis of lean NAFLD/NASH. Studies counterintuitively suggest that the progression of NAFLD to NASH is much more aggressive in lean people with fatty liver disease, despite their lower prevalence of fibrosis and more favorable risk factor profile.
  • Epidemiology of NAFLD/NASH. Prevalence of NAFLD in the general population is commonly estimated at around 25%-30%, of which 20%-30% eventually progress to NASH. Due to the difficulty of diagnosis, the “real” prevalence statistics are yet to be determined.
  • How to diagnose NAFLD/NASH. Non-invasive diagnostic tools such as MRI and liver enzyme tests provide an incomplete picture of NASH, largely because its three pathophysiological features – steatosis, inflammation, and cirrhosis – are difficult to simultaneously assess. Composite measures like the NAFLD fibrosis score, the BARD score, and the Hepascore (which integrate various combinations of liver biochemistry, glycemic measures, and anthropometric measurements) are not sufficiently sensitive and are typically unable to definitively characterize >15% of people. According to Dr. Oren, even the gold standard liver biopsy diagnostic is “neither gold, nor standard” because there can be incredible variability between different biopsy samples of the same liver. What’s more, liver biopsy is quite expensive and invasive, adding another deterrent for an already under-diagnosed disease.
  • Population health. Given what appears to be epidemic rates of fatty deliver disease, should countries call for nationwide NAFLD/NASH screening? Would such a policy be cost-effective, given the lack of available treatments? (In other words, after screening, what next? There are no FDA-approved treatments for NASH on the market.) Would population-level screening be feasible given the lack of consensus over diagnosis?
  • Treatment for NAFLD/NASH (in the context of no formally-approved therapies). Currently, the only officially-recommended treatments for fatty liver disease (besides diet and exercise) are pioglitazone and vitamin E, though preliminary evidence suggests that GLP-1 agonists and metformin could also be effective. A wide array of therapies studied specifically for NAFLD/NASH are in various stages of development, but even the most advanced phase 3 candidates (Allergan/Norvartis’ cenicriviroc, Genfit’s dual PPAR alpha/gamma agonist elafibranor, Intercept’s FXR agonist obeticholic acid, and Gilead’s ASK-1 inhibitor selonsertib) are years away from reaching the market (and much further from being affordable and well-reimbursed).
  • As Dr. Oren described, NAFLD/NASH represents the epitome of controversy and lack of consensus in the field metabolic disease. We’d even add one more key issue to this list: 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. 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. To this end, the Global Liver Institute’s Dr. Brian Harvey has advocated that hepatic data be systematically collected in diabetes CVOTs (~10,000-patient datasets ripe for the picking). 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. 


-- by Ann Carracher, Abigail Dove, Payal Marathe, and Kelly Close