Hallo from Berlin, Germany, where the very first World Congress on Clinical Trials in Diabetes wrapped up earlier today. Our team was lucky to join the intimate meeting of just 250 attendees, gathered to discuss both big picture perspectives on the future of diabetes therapy development and clinical trials and the practical aspects of designing trials. See below for our top ten highlights from this stellar meeting – we’re already looking forward to next year’s conference, which will again be held in beautiful Berlin!
1. Kinexum’s Dr. Alexander Fleming, formerly of the FDA, set the stage for the meeting with his opening keynote describing the ongoing trends shaping diabetes drug development. In particular, Dr. Fleming pointed to six major factors: (i) increased value-based payment and pricing pressures; (ii) the expectation of value-based medicine; (iii) the emergence of big data; (iv) the growing use of genomic and other “-omic” data to classify patients; (v) the growing adoption of individualized regulatory pathways for new drugs; and (vi) the emergence of faster and shorter product cycles.
2. In a separate talk on the FDA perspective on clinical trials, Dr. Fleming forecast a convergence toward a single regulatory definition of severe hypoglycemia, with a numerical cut point of 54 mg/dl. Dr. Fleming further suggested that broader use of SMBG and CGM in trials will be employed to meet hypoglycemia endpoints.
3. Conference organizer Dr. Stefano Del Prato provided a comprehensive overview of the history of cardiovascular outcomes and diabetes drugs, ultimately concluding that the regulatory requirement for cardiovascular outcomes trials (CVOT) has ultimately proved valuable in raising important clinical and scientific questions.
4. Conference organizer Dr. Philip Home convincingly argued that it’s not possible to identify all potential safety risks for a particular therapy in a clinical development program.
5. Dr. Robert Heine, Distinguished Lilly Scholar, offered a rebuttal to Dr. Home’s position by cautioning that post-marketing safety signals must be interpreted extremely thoughtfully to avoid unnecessarily stroking public fear.
6. Our very own Mr. Manu Venkat (now as medical student at UCSF) and Ms. Helen Gao (a senior associate at Close Concerns) offered a look at the challenges of finding and enrolling in clinical trials for the average patient.
7. “The regulators tell you whether you can use a new technology and NICE says whether you should use it.” NICE’s Dr. Amanda Adler offered an illuminating perspective on how the UK’s NICE appraises new diabetes therapies for reimbursement decisions. Dr. Adler’s presentation also offered a tantalizing glimpse at how NICE might respond to a cardiovascular death reduction indication for Lilly/BI’s Jardiance (empagliflozin), should the expanded indication receive EMA approval.
8. BI’s Global Head of Medical Affairs Dr. Maximilian von Eynatten provided a look at how novel diabetes drugs might differentiate themselves in an environment in which simple glucose lowering may no longer be sufficient for success.
9. Conference organizer Dr. Oliver Schnell provided an overview of key takeaways from the cardiovascular outcomes trials that have reported results thus far. He also advocated for the consideration of CVOT results in future iterations of treatment guidelines.
10. Mr. James Nolan, CEO of the contract research organization InClinica (Wayne, PA), highlighted seven key challenges for enrollment in modern clinical trials from an industry perspective.
Top Ten Highlights
1. Kinexum’s Dr. Alexander Fleming, formerly of the FDA, set the stage for the meeting with his opening keynote describing the ongoing trends shaping diabetes drug development. In particular, Dr. Fleming pointed to six major factors: (i) increased value-based payment and pricing pressures; (ii) the expectation of value-based medicine; (iii) the emergence of big data; (iv) the growing use of genomic and other “-omic” data to classify patients; (v) the growing adoption of individualized regulatory pathways for new drugs; and (vi) the emergence of faster and shorter product cycles. Dr. Fleming elaborated that the diabetes industry is under considerable pressure to produce clinical outcome data from payers – both national governmental payers and pharmacy benefits managers like Express Scripts. On individualized medicine, Dr. Fleming suggested that we are not quite yet at the point at which individualized medicine is a reality, but the expectation of personalization is driving clinical trials to focus on subpopulations and explore potentially predictive biomarkers. Likewise, big data and “omic” data will be increasingly important to how clinical trials are designed and how patients are characterized and selected for trials and therapies. In order to keep up with these demands, Dr. Fleming suggested that the FDA is increasingly open to individualized regulatory pathways that allow therapies to be approved in a stepwise manner as data becomes available, initially receiving approval for a smaller, high-risk patient pool in which the benefit/risk ratio is particularly favorable. We expect this could be similar to the path that PCKS9 inhibitors have taken toward approval, in which the class was initially approved for a relatively limited indication (patients with ASCVD or FH) that could potentially be expanded pending positive cardiovascular outcomes data. Dr. Fleming pointed out that adaptive regulatory pathways can shorten the time and cost of bringing a drug to market, which would presumably encourage continued drug development in an environment where product cycles in diabetes therapy are becoming shorter and faster. Dr. Fleming drew an analogy to the medical device field, noting that new drugs may only enjoy peak success for a few years before they are superseded by a new therapy. As a result, according to Dr. Fleming, clinical trials need to achieve more in less time and glucose-lowering alone is no longer sufficient from a commercial perspective. Additionally, Dr. Fleming noted that this has driven interest away from therapies for diabetes complications because of the long time needed to demonstrate benefit and has also put enormous pressure on phase 2 trials to demonstrate favorable profiles.
2. In a separate talk on the FDA perspective on clinical trials, Dr. Fleming forecast a convergence toward a single regulatory definition of severe hypoglycemia, with a numerical cut point of 54 mg/dl. Dr. Fleming further suggested that broader use of SMBG and CGM in trials will be employed to meet hypoglycemia endpoints. This was music to our ears as hypoglycemia is tremendously important to patients and has been notoriously difficult to demonstrate in clinical trials. As Dr. Fleming underscored, standardization of hypoglycemia definitions and acceptance from regulators will allow new insulin products to better differentiate themselves not on additional A1c reduction but on minimizing hypoglycemia. This is particularly important as next-generation (and next-next-generation) basal insulin analogs seek to differentiate themselves from older products like Sanofi’s Lantus (insulin glargine). Also encouragingly, Dr. Fleming acknowledged the importance of glycemic variability as an outcome, but suggested that it may be a while before glycemic variability is accepted as a regulatory endpoint because endpoints related to glycemic variability are currently not well-validated. We imagine a combination of CGM or flash glucose monitoring and use of patient-reported outcome measures could help illustrate potential benefits on glycemic variability – there certainly is massive interest and advocacy on this front from a number of patient and professional organizations, as demonstrated by the recent FDA Workshop on Outcomes Beyond A1c.
3. Conference organizer Dr. Stefano Del Prato provided a comprehensive overview of the history of cardiovascular outcomes and diabetes drugs, ultimately concluding that the regulatory requirement for cardiovascular outcomes trials (CVOT) has ultimately proved valuable in raising important clinical and scientific questions. Dr. Del Prato emphasized that the lessons from the EMPA-REG OUTCOME trial go beyond demonstrating a cardiovascular benefit for empagliflozin (Lilly/BI’s Jardiance) or even potentially for the SGLT-2 inhibitor class as a whole. The trial also demonstrated that cardioprotection in diabetes need not necessarily be associated with slowing down the progression of atherosclerosis and that the questions surrounding the mechanism of benefit has opened up new avenues for research. Regarding GLP-1 agonists, Dr. Del Prato pointed out the heterogeneity of GLP-1 CVOT results – the LEADER and SUSTAIN 6 trials for Novo Nordisk’s Victoza (liraglutide) and semaglutide, respectively, demonstrated cardioprotection while the ELIXA and FREEDOM-CVO trials for Sanofi’s Adlyxin (lixisenatide) and Intarcia’s ITCA 650 (implantable exenatide mini-pump) were neutral. He suggested that the hypothesis that the difference may be attributable to a difference in GLP-1 agonist exposure (liraglutide and semaglutide are long-acting, while lixisenatide is short-acting) doesn’t hold much weight when considering the continuous exposure offered by ITCA 650. We were glad to see Dr. Del Prato include a discussion of FREEDOM-CVO (we rarely hear speakers at conferences mention the trial, even in presentations dedicated to surveying the CVOT landscape!), though it’s important to note that FREEDOM-CVO was designed as a very small, pre-approval CVOT to meet the regulatory requirement of ruling out a 1.8 hazard ratio for cardiovascular risk for approval. The full results from the trial are not yet available so it’s unknown what the point estimates for the primary MACE endpoint and its components were and it’s quite possible that a larger post-approval CVOT could demonstrate a cardioprotective benefit. In any case, Dr. Del Prato’s point is well-taken that further results for the GLP-1 agonist class are needed to elucidate the effect of the class on cardiovascular outcomes. Further, Dr. Del Prato suggested that these CVOTs raise the possibility of the combination approaches to both sustained glycemic control and cardioprotection. In particular, he pointed to the intriguing possibilities of combining empagliflozin with semaglutide or empagliflozin with TZD pioglitazone (which demonstrated an anti-atherosclerotic effect in the PROactive and IRIS trials).
4. Conference organizer Dr. Philip Home convincingly argued that it’s not possible to identify all potential safety risks for a particular therapy in a clinical development program. He pointed out that the diabetes field has a long history of safety concerns coming as surprises after a drug is already on the market – from troglitazone’s increase in deaths from liver failure to the pioglitazone bladder cancer story to SGLT-2 inhibitors and ketoacidosis to, most recently, the increase in retinopathy seen for GLP-1 agonist semaglutide in SUSTAIN 6. He noted that post-marketing surveillance exists precisely because it is not possible to pick up on all safety signals in the regular clinical development program. Furthermore, many of these safety concerns are rare with low event rates, necessitating enormous patient-years of exposure for them to be revealed and confirmed in a clinical trial. Even cardiovascular outcomes trials don’t even come close to the exposure needed for some of these events – by his back-of-the-envelope calculations, Dr. Home suggested that 1,000 events and 100,000 patient-years of exposure would be needed to confirm the signal for increased non-fatal stroke observed in the EMPA-REG OUTCOME study (and, notably, the signal was only picked up with the >20,000 patient-years exposure of EMPA-REG OUTCOME and not previous clinical trials). In the normal development program, trials performed before licensing may record around 500 different types of adverse events and it’s very difficult with such small event numbers to determine which have a “real” connection to the drug. That said, Dr. Home concluded by emphasizing that all medications – even penicillin and statins – have safety problems and what matters is not the existence of safety concerns but the overall benefit/risk ratio. By this token, he underscored that it largely doesn’t matter if we can’t definitely rule out increased risk of certain, rare adverse events if the benefit is large enough.
5. Dr. Robert Heine, Distinguished Lilly Scholar, offered a rebuttal to Dr. Home’s position by cautioning that post-marketing safety signals must be interpreted extremely thoughtfully to avoid unnecessarily stroking public fear. In particular, Dr. Heine pointed to the pancreatitis controversy with GLP-1 agonists, posing the question “how many people were denied use of GLP-1 agonists when in hindsight it was safe and the signals were wrongly interpreted” in the six years it took to put the controversy to rest? Similarly, Dr. Heine suggested that a “flawed” analysis from Dr. Steve Nissen regarding an association between TZD rosiglitazone and cardiovascular events essentially “killed the drug.” Dr. Heine characterized latter efforts to set the record straight as “too little, too late and the drug is hardly available anymore.” The safety concerns with TZDs also extended to pioglitazone, which, according to Dr. Heine, led to patients with diabetes being denied one of the “best drugs” that now has been demonstrated to offer significant cardiovascular, diabetes prevention, and NASH benefits. Further, Dr. Heine emphasized that meta-analyses are hardly predictive and their findings are rarely corroborated by largescale cardiovascular outcomes trials (CVOTs). As such, he concluded that CVOTs, as part of a trial’s clinical development program, are ultimately essential to determining safety risks associated with drugs.
- All in all, Dr. Heine’s arguments were not so much a counterpoint as a complement to Dr. Home’s position. Certainly, many would agree that meta-analyses represent a fairly low standard of evidence and that CVOTs in recent years have elucidated a number of surprising findings – ranging from cardiovascular neutrality for drugs thought to have a potential benefit (DPP-4 inhibitors) to entirely unexpected and unprecedented benefit (empagliflozin) to impressive nephropathy findings (liraglutide and semaglutide). That said, Dr. Home’s point that it’s near impossible to identify with certainty every safety risk in a timely and cost-effective manner in clinical trials is well-taken and post-marketing surveillance and adverse event reporting continue to play an important role to drawing attention potential adverse events – such as DKA with SGLT-2 inhibitors – so that we can develop strategies to mitigate the risk in a clinical setting. Additionally, we couldn’t agree more that thoughtfulness, care, and restraint are required in the communication of scientific information to the public – too often, the mainstream media overstates scientific findings (both positive and negative) for the sake of eye-catching headlines, often stroking public fears and hopes in the process with serious consequences for patients who could benefit from certain medications when taken appropriately.
6. Our very own Mr. Manu Venkat (now as medical student at UCSF) and Ms. Helen Gao (a senior associate at Close Concerns) offered a look at the challenges of finding and enrolling in clinical trials for the average patient. Mr. Venkat illustrated the challenges of even knowing where to look for open clinical trials – the ADA website directs potential participants to ClinicalTrials.gov, which is notoriously difficult to navigate and returns over 14,000 trials when “diabetes” is inputted as a search term. Even once enrolled in a clinical trial, participation can place enormous burdens on patients. Our very own Mr. Adam Brown, for instance, participated in a closed loop trial and was required to frequently make the two hour round-trip drive down to Stanford for trial visits, which was only possible due to his enormous personal dedication and flexible job. Additionally, Mr. Venkat pointed out that clinical trials often do not answer the questions that matter most to patients. Drawing from data from market research company dQ&A, he pointed out that patients across the board care about “time in range” and that A1c often does not tell the whole story in terms of the “quality” of glycemic control.
- Based on these shortcomings, Mr. Venkat likened current clinical trials to the traditional method of hailing a taxi cab and posed the question, “So what does the Uber app version of clinical trial look like?” Taking the podium to answer this question, Ms. Gao suggested that technology and digital platforms can be harnessed to facilitate recruitment and enrollment. In particular, she highlighted five key characteristics for an ideal, patient-friendly clinical trial database: (i) it should be easy to find; (ii) with user-friendly interface; (iii) a customizable search; (iv) simple descriptions of trial requirements and eligibility criteria; and (v) clear expectations for participation. To exemplify many of these traits, Ms. Gao pointed to the Antidote platform (formerly known as TrialReach), which now features studies for both type 1 and type 2 diabetes, as well as a number of additional disease areas. Taking technology a step further, Ms. Gao also noted the potential to move parts – or even all – of clinical trials online through platforms such as Apple’s ResearchKit. In addition to easing the clinical trial burden, Ms. Gao emphasized the importance of incorporating measures of glycemic variability (via CGM or Abbott’s Freestyle Libre) and standardized patient-reported outcomes in clinical trials. From a systems standpoint, these outcomes can help inform whether or not patients will be likely to engage with and adhere to the medication once its on the market, which could prevent diabetes complications and allow more resources to be devoted to prevention or primary care. That said, she also acknowledged that the inclusion of these endpoints must also be supported at the regulatory level, and thus called for advocacy for consideration of outcomes beyond A1c and for in the inclusion of positive CVOT results on labels and guidelines. Returning to the stage, Mr. Venkat highlighted a number of practical steps trial investigators can take to today to ensure that participants feel valued. In particular, he emphasized the importance of thoughtful language and regular updates on trial progress and results to help patients feel like a member of the trial team rather than a “subject.”
- Finally, Mr. Venkat concluded with a number of pertinent questions for clinical trial improvement from the patient and public perspective: (i) How can we help the media frame clinical trial results better, to prevent the inaccurate sensationalistic headlines that terrify patients? (ii) With formulary decisions are becoming increasingly important, can payers review data in parallel with regulators? (iii) How can we give clinicians more time to speak with patients about enrolling in trials? (iv) And in addition to using technology to bring more patients to us, how can we go out and reach patients where they are? In community centers, churches, and schools? For more, see the slides and transcript of the presentation.
7. “The regulators tell you whether you can use a new technology and NICE says whether you should use it.” NICE’s Dr. Amanda Adler offered an illuminating perspective on how the UK’s NICE appraises new diabetes therapies for reimbursement decisions. As she put it, whereas the EMA determines if a drug works and if it’s safe, NICE determines how well the drug works compared to other already available drugs and whether it represents a good value proposition. She emphasized that NICE is working with a fixed budget, so the decision to reimburse a new technology comes at the expense of paying for something else – this could refer to another diabetes drugs, diabetes education, nurse salaries, or drugs for other disease areas entirely (such as Alzheimer’s or cancer), among other things. As the result, NICE uses quality-adjusted life-years (QALYs) to generalize the value of medical therapies, technologies, and services across the entire health system. Given the lack of hard outcomes in most diabetes trials, NICE relies on models of diabetes epidemiology to determine the number of QALYs associated with a given A1c reduction. The agency also conducts a comparative cost-effectiveness analysis to already available drugs. Dr. Adler then outlined five key considerations for NICE appraisals that should be taken into account when designing clinical trials: (i) how appropriate is the choice of comparator therapy? The drug should be evaluated head-to-head against comparators that reflect clinical practice; (ii) how much uncertainty is involved in extrapolating the survival rate past the end of a trial?; (iii) how plausible is the quality of life benefit as stated by the sponsor?; (iv) how believable is the cost-effectiveness data?; and (v) Are the models used to determine value transparent?
- Dr. Adler’s presentation also offered a tantalizing glimpse at how NICE might respond to a cardiovascular death reduction indication for Lilly/BI’s Jardiance (empagliflozin), should the expanded indication receive EMA approval. In particular, she shared that NICE will need to grapple with the choice of an appropriate comparator to appraise empagliflozin against for this new indication. She suggested that, in all likelihood, the comparator would be usual care without empagliflozin. In terms of the outcome, Dr. Adler pointed out that, for the first time, EMPA-REG OUTCOME measured length of life, which should presumably make the appraisal more straightforward. On the other hand, she made a point of the fact that cardiovascular death was a secondary endpoint and thus exploratory and rife with uncertainty if you were to extrapolate beyond the end of the trial. That said, at the FDA Advisory Committee meeting on this topic, biostatiscian Dr. Stuart Pocock characterized the EMPA-REG OUTCOME findings as “overwhelming evidence of benefit” and “proof beyond a reasonable doubt” of a benefit, despite the secondary endpoint status.
8. BI’s Global Head of Medical Affairs Dr. Maximilian von Eynatten provided a look at how novel diabetes drugs might differentiate themselves in an environment in which simple glucose lowering may no longer be sufficient for success. Dr. von Eynatten proposed a “better than the Beatles” test for new diabetes drugs, in which the combination of the metformin and insulin may be analogous to the Beatles as being the “best ever” from a pure A1c reduction and cost standpoint. In the case of Lilly/BI’s Jardiance (empagliflozin), Dr. von Eynatten suggested that the drug passed the Beatles test by demonstrated meaningful reductions in cardiovascular risk on top of metformin and insulin in many patients in the EMPA-REG OUTCOME trial. He attributed the drug’s ability to pass the Beatles test to three key factors: (i) it addressed an area of high unmet need, namely residual risk of premature cardiovascular mortality associated with diabetes; (ii) the EMPA-REG OUTCOME trial aimed to establish both safety and efficacy; and (iii) EMPA-REG OUTCOME was initiated right at the start of the phase 3 development program for empagliflozin, allowing outcomes data to be available just one year after the product launched. Looking to the future, Dr. von Eynatten suggested that prospective diabetes drugs may need to demonstrate cardiovascular non-inferiority versus a drug with a proven cardiovascular benefit, such as empagliflozin – even if this does not become an explicit standard of future CVOTs, it’s clear that the bar of new diabetes drugs continues to rise. Acknowledging this, Dr. von Eynatten recognized the need for truly disruptive therapies in diabetes and pointed to several opportunities for innovative diabetes drug development moving forward. (i) He argued that outcomes data will become increasingly crucial to regulatory approval as well as clinical acceptance. (ii) Ongoing payer pressure will amplify the role of comparative effectiveness analyses and performance-based-contracting in the diabetes field. (iii) Demand for patient support programs and patient relevant outcomes will intensify. (iv) Individualized diabetes care will continue to evolve and become more precise. (v) Outcomes beyond A1c – ranging from cardioprotection, weight loss, and lipid effects to CGM and time-in-range – will become standard considerations for diabetes treatment. It’s clear that the diabetes field is rapidly changing and it appears that empagliflozin is well-positioned for the emerging trends. We appreciated Dr. von Eynatten’s nuanced overview of how the diabetes landscape and outlook as shifted – we only hope that the higher bar does not make it too challenging for smaller biotech companies to invest in diabetes, however.
9. Conference organizer Dr. Oliver Schnell provided an overview of key takeaways from the cardiovascular outcomes trials that have reported results thus far. Regarding GLP-1 agonists, Dr. Schnell noted that all three agents with CVOT results (Sanofi’s Adlyxin [lixisenatide], Novo Nordisk’s Victoza [liraglutide], and its next-generation semaglutide) can safely be applied in patients with high cardiovascular risk. Furthermore, he highlighted the additional cardiovascular risk reduction demonstrated for liraglutide and semaglutide in LEADER and SUSTAIN 6, respectively, and noted that both agents also demonstrated lower rates of nephropathy in the trials as well. That said, Dr. Schnell drew attention to the finding of increased retinopathy in the SUSTAIN 6 trial as well. Regarding the EMPA-REG OUTCOME trial that demonstrated a cardioprotective benefit for Lilly/BI’s Jardiance (empagliflozin), Dr. Schnell suggested that a stronger role for empagliflozin in the treatment guidelines is warranted. Furthermore, he attributed the early emergence of the benefit to hemodynamic changes. On DPP-4 inhibitors, Dr. Schnell put forth that the agents can be used safely given their demonstrated cardiovascular safety and attributed the differences in hospitalization for heart failure risk in the trials as potentially related to differences in the structure of molecules. Dr. Schnell also briefly highlighted the hot-off-the-press topline results from the DEVOTE trial, demonstrating cardiovascular safety for Novo Nordisk’s next-generation basal insulin Tresiba (insulin degludec). Overall, Dr. Schnell summarized that a cardioprotective and renal-protective benefit can reasonably be expected of empagliflozin, liraglutide, and semaglutide in high-risk patients on the basis of these trials. That said, he emphasized that class effects are not yet readily apparent and cannot be assumed. Regarding future clinical trials, Dr. Schnell called for greater attention to risk for hospitalization for heart failure and for the inclusion criteria of CVOTs. Finally, Dr. Schnell also advocated for the consideration of CVOT results in future iterations of treatment guidelines.
10. Mr. James Nolan, CEO of the contract research organization InClinica (Wayne, PA), highlighted seven key challenges for enrollment in modern clinical trials from an industry perspective. First, he pointed to the growing popularity of orphan indications and limited targeted population, leading to few patients who are available for enrollment in trials. Second, Mr. Nolan discussed the impact of exclusion criteria on narrowing the potential participant pool. He pointed out that academic medical centers have limited patient populations and, coupled with narrow exclusion criteria, different trials sometimes end up excluding the same people over and over – leading to the “exclusion ripple effect.” Mr. Nolan even suggested that there are “clinical trial junkies,” particularly in the developing world, for whom the $300-$400 compensation for trials is significant and they enroll in multiple trials at a time as a result. On the flip side, Mr. Nolan also pointed out that a patient that has been excluded from multiple trials is less likely to want to enroll in additional trials in future. The third factor raised by Mr. Nolan was the slowdown in clinical trials caused by IRB backlogs – he shared that most IRBs are operating on a two to three-month delay and are resistant to a centralized IRB for a multi-center trial. Fourth, from a big picture standpoint, Mr. Nolan suggested that PCORI’s patient-centered focus on examining the ethics of clinical trials could threaten the traditional clinical trial architecture. Fifth, Mr. Nolan highlighted the drop in pediatric waivers at the regulatory level, making dose-ranging and other studies for children and adolescents increasingly necessary pre-approval. Sixth, he pointed out that robust competitive landscapes have led to multiple companies competing to enroll the same patient population into their trials. And finally, seventh, Mr. Nolan suggested the internet allowing patients to learn about and enroll in trials on their own, potentially introducing selection bias. Mr. Nolan’s presentation raised many specific, oft-overlooked considerations that can significantly impact the logistics of running a clinical trial. We hope that, by drawing attention to these issues, we can begin to think of methods to address these concerns while continuing to encourage and engage patients to seek out trials, for example.
--by Helen Gao and Kelly Close