Skip to main content

Influential FDA Panelist Dr. Sanjay Kaul Discusses Diabetes, Obesity, and the Benefit-Risk Assessment for Drug Development

Updated: 8/14/21 12:00 pmPublished: 4/30/11

by vincent wu, sanjay trehan, joseph shivers, and kelly close

We recently had the privilege of speaking with Dr. Sanjay Kaul, a noted cardiologist at Cedar-Sinai Medical Center's Heart Institute in Los Angeles and  cardiology expert who is often called on to advise the FDA's Endocrine and Metabolic Advisory Committee. In the past year, Dr. Kaul has played a highly influential role in the panel discussions on numerous diabetes and obesity medications, including AvandiaQnexa, Meridia, Lorqess, and Contrave. Dr. Kaul also serves as a member of the FDA's Cardiorenal Advisory Committee, a role he has held for many years.

During our discussion, Dr. Kaul emphasized that the FDA pays more attention to the discussions at advisory panel meetings than the votes themselves, helping to explain the seeming disconnect between some advisory panel votes and the FDA's subsequent decisions (e.g., Contrave). To reduce public confusion, he suggested that the FDA remove the voting process altogether from advisory panel meetings – we at diaTribe hope this does not happen since we like to see explicitly how each member of the panel is leaning. In addition, Dr. Kaul focused on the need to assess cardiovascular outcomes for diabetes and obesity drugs, given that surrogate markers (e.g., A1c and weight loss) do not necessarily translate to benefits for the heart; notably, he didn't believe long-term outcomes trials should be recommended for all obesity drugs, but where there is a "signal" (i.e., a worry), there should be such trials. Overall, he was very positive on the need for more drug development in obesity, making us believe that representatives from companies and the government should talk more frequently about how to make better treatments available.

advisory committee meetings

Kelly Close: Hello, Dr. Kaul. Thank you so much for joining us! This is really an honor for diaTribe to speak with you. To start off, can you tell us how you would like to see the advisory panel process improved?

Dr. Sanjay Kaul: One area that would benefit from a change is the voting process. Although the current simultaneous electronic voting is an improvement over the old "going around the table" verbal voting process in the sense that it tends to eliminate potential biases introduced by "group think" or "herd mentality," it nevertheless has its own limitations. Quite often, the yes/no vote introduces an artificial dichotomy with a fine line separating the two. There are times when the vote fails to capture or accurately reflect the content of the panel discussions… I am sometimes struck by how the voting process is viewed as a contest between the absolute certainties of yes or no. It is anything but. Calls for caution, and the ifs and buts that mark the deliberations appear to de drowned out. Thankfully, the FDA pays greater attention to the discussions around the vote than the simple vote count. This might help explain why nearly one-third of the time the FDA rejects approvals despite a "yes" vote by the advisory panel. Thus, I don't think it is unreasonable to question the utility of the voting process. Perhaps, the FDA should consider eliminating it entirely, especially since our vote is not binding and our job is only to offer advice.

Kelly: We hope from a patient perspective they don't eliminate it because we like seeing how the different panelists vote to get a sense of their priorities! This is harder to do without a vote, but we very much see your point. Many have criticized advisory panel compositions because so few panelists have been involved in drug development itself; panel members, due to concerns about possible bias, tend to be those who have less experience developing drugs from a commercial perspective. While that does allay fears about bias, it raises the question about whether there is enough experience on advisory committees to properly advise on what companies can practically implement and not implement. I wonder if you could share with us your thoughts on this.

Dr. Kaul: While it is true that experts most intimately involved with obesity drug development are not represented in the advisory committee, the FDA has a difficult responsibility to limit the biases and the conflicts of interest of the advisory committee. These are not just limited to financial conflicts but also include intellectual biases and conflicts. In my experience, the best advice often comes from expert "consumers" rather than expert "generators" of information and evidence. Nonetheless, in my opinion, the FDA does a very good job in assembling a panel with a broad spectrum of expertise in drug development, evidence appraisal, benefit-risk assessment, clinical decision-making, and the regulatory approval process.

obesity drug development, assessment, and the regulatory landscape

Kelly: Some of the panelists on the Contrave advisory committee expressed concern that obesity drug development would slow dramatically or come to a halt if a pre-approval cardiovascular outcomes trial was required of Contrave. Can you share your thoughts on this with us?

Dr. Kaul: In my assessment, a pre-approval cardiovascular trial is warranted, and the FDA made the right call, and not a very surprising one in my opinion.

Kelly: I think what surprised some of us was that panelists voted that a pre-approval trial was not needed and that Contrave should be approved; in effect, the FDA voted against the panel, which we're certainly aware can happen, but we don't view as common – we didn't realize as you just said that this happens up to one-third of the time! On a related note, can you discuss your view on the need for obesity drug development broadly speaking? And specifically, what is your thought on what will happen with Contrave?

Dr. Kaul: Yes, I do think that obesity drugs need to be developed, but rather than repackaging old ones, I also feel that the whole field would be better served by focusing on innovative, safe, and effective therapies. Perhaps the recent negative decisions by the FDA regarding the obesity drugs will help spur this innovation. With regards to Contrave, it all depends on whether the company has the fortitude and the fiscal health to carry out the evaluation that that FDA has asked for. Obviously it will take time, effort, and money. The critics lament that the FDA is setting a very high bar requiring the sponsors to prove their drugs are "whiter than white" in terms of safety. However, given the checkered history of weight-loss drugs, I believe extra caution is warranted. Otherwise, there is a price to pay every time the FDA pulls a weight-loss drug off the market because of anticipated or unanticipated risks – erosion of public trust. No one likes to be second-guessed. The FDA often walks a tightrope reconciling the dual goals of protecting and promoting public health.

Kelly: Another view is that since there are fewer options for people who are obese and since the epidemic is arguably worse [than the diabetes epidemic], it's reasonable for the standards to differ [from the standards for diabetes drugs].

Dr. Kaul: I am sensitive to that argument. However, if there are already signals for cardiovascular risk in early phase development, and an identified plausible mechanism, then it is incumbent upon us to make sure that we are not increasing the cardiovascular risk of those exposed. I am aware of the criticisms that the Qnexa panel was sensitized by the Avandia hearings that immediately preceded it, and that the FDA is risk-averse and not approving the drug is a sure way to limit exposure and risk. Frankly, such criticisms are off target. Each panel member does his or her best to adjudicate benefit-risk based on the specific evidence at hand. And the FDA, to the best of my knowledge, takes its job of protecting and promoting public health very seriously. Obesity drugs should be required to rule out some degree of unacceptable cardiovascular harm pre-approval, wherever appropriate, i.e., depending on the mechanism of action and the signal for increased cardiovascular risk (such as elevated heart rate, blood pressure, or valvular side effects) identified during early phase development. I also generally agree with the current guidance that requires all diabetes drugs to exclude cardiovascular harm pre-approval. However, rather than fixing the boundary of unacceptable harm, it should be flexible based on the individualized benefit-risk tradeoffs. Ideally, pre-approval trials should assess durability of weight-loss efficacy by extending the follow-up to at least two years. However, whether the weight loss translates into meaningful health benefits should be assessed post-approval.

Kelly: We will certainly be eager to see more on this, if drug development can continue. Can you give us your thoughts on conditional approval for weight-loss medications?

Dr. Kaul: Conditional or accelerated approvals are granted by the FDA under Subpart H regulations for a new drug product on the basis of adequate and well-controlled clinical trials establishing that the drug product has an effect on a surrogate endpoint that is reasonably likely, based on different sources of evidence to predict clinical benefit. Approval under this section is subject to the requirement that the applicant study the drug further, to verify and describe its clinical benefit. This could be potentially implemented for weight-loss medications. However, the FDA currently lacks the legislative authority to enforce withdrawal of conditional approval if postmarketing clinical studies fail to verify clinical benefit (two recent examples that come to mind include midodrine and Avastin). At what cost? The advisory members and the FDA reviewers (who are loath to be second guessed) are going to be hesitant to consider conditional approval in the future, thereby denying patients timely access to potentially life-saving therapies. It's a double-edged sword. Thus, it is vitally important to allow the regulatory process to take its due course of action.

Sanjay Trehan: Another broad question. Do you think a formal structure to give the FDA that legislative authority would be helpful?

Dr. Kaul: Absolutely. They already have that for safety assessment under the FDA Amendment Act of 2007, which provides a mandate to require drug sponsors to implement specific measures to reduce safety risks through REMS (Risk Evaluation and Mitigation Strategy) programs. I believe it should also be extended to assessment of clinical efficacy of drugs that are approved based on surrogate measures of efficacy. Ideally, the FDA should grant only "provisional approval" when surrogate endpoints are used. The conditions for full approval should be timely demonstration of efficacy and safety in large, well-designed clinical-outcomes trials that examine cardiovascular endpoints. Furthermore, Congress should empower the FDA to enforce post-marketing commitments, including withdrawal of approval if these post-marketing studies fail to confirm clinical benefits. Only then can we hope to avoid potential controversies.

Kelly: Dr. Kaul, we are also really interested in your opinion of warnings and restricted access programs.

Dr. Kaul: Labeling changes do not reliably mitigate risk. Elevating warnings to contraindications has a poor track record of improving inappropriate patient selection.

cardiovascular risk and cardiovascular outcomes trials

Sanjay: We would love to talk more about cardiovascular outcomes trials. To start, how do you think one can assess the risk of cardiovascular outcomes for an obesity drug, given that the benefits of modest weight loss may accrue over time, and the target population may not be at a high enough short-term risk to make what a traditional outcomes study, at least in diabetes, feasible?

Dr. Kaul: Most obesity trials have been conducted in women in their 40s, and the risk level in this demographic is lower than what would typically be seen in diabetes trials. In general, cardiovascular risk is higher for men than women. So this poses a challenge in trial design. What I've seen so far with pre-approval trials of obesity drugs in general is the opposite – a "sanitized" population that is not representative of the real-world patients, and a follow-up limited to about one to two years. One of the major shortcomings of randomized controlled trials (RCTs) is their limited external generalizability. These trials enroll highly selective and homogeneous populations who receive standardized optimal care. The real-world scenario is quite different. The population is heterogeneous (in terms of underlying risk and treatment response), and there is substantial variability in the quality of care the patients receive. I am a strong believer in large and simple, so-called "pragmatic" RCTs that minimize exclusionary criteria and faithfully capture the real-world efficacy and safety of therapeutic interventions. In such trials, the most reliable estimate of a treatment effect within a particular subgroup is the overall treatment effect in the trial.

Kelly: That's unsettling and points to a different problem, a REMS issue. Why would the FDA even approve any drugs indicated for specific populations, if the view is that the indications will be ignored? 

Dr. Kaul: Well, that's the reality. There's a regulatory world, and then there's a real world. Regulators are responsible for approving drugs. They are not in the business of mandating the practice of medicine. Physicians are free to use approved drugs in any fashion they deem safe. And, in some instances, the art-of-medicine philosophy might even encourage off-label use as proper exercise of clinical judgment. Controlling the use of drugs beyond the label indication is a challenge. The FDA always worries that diet drugs will not be used or prescribed appropriately. None of this should come as a surprise. Let us not forget the diet clinics ("pill mills") that popped up during the fen-phen years. It is very difficult to prevent widespread usage of weight-loss drugs in inappropriate populations. One of the biggest concerns is that since obesity drugs are not typically reimbursed, claims data are not typically available to conduct observational studies post-launch and to monitor outcomes, making surveillance all the more difficult. These challenges are further compounded by lack of a reliable REMS program to mitigate anticipated risks. So, while limited approval might sound intuitively appealing, the difficulty lies in implementing these programs, controlling use, and restricting access.

Sanjay: For diabetes medications, do you think the pre-approval cardiovascular outcomes requirement has been constructive thus far? It seems as if we've completely reversed where the regulators stand in terms of the information they receive.

Dr. Kaul: There is a cloud over new diabetes drugs as a result of the new guidance, which has discouraged and will continue to discourage the pharmaceutical industry from venturing into this area. Some have remarked that in our great concern about safety, we may be standing in the way of developing even better drugs in the future. I personally believe that the guidance should be refined in that it should be more anchored in both the science and in the art-of-medicine philosophy. The safety concerns should be balanced against the treatment benefit. In my opinion, ruling out the same degree of fixed cardiovascular harm for drugs that provide varying degrees of efficacy is not appropriate. The guidance should be flexible, thereby allowing tolerability of a greater degree of harm in return for a greater degree of benefit. For example, drugs that improve hemoglobin A1c by 1% or more should not be expected to rule out 30% harm post-approval. I would argue that ruling out perhaps 50% harm is quite acceptable. On the other hand, drugs that only offer less than half a percent improvement in hemoglobin A1c should require more stringent criteria for approval.

Sanjay: What are your thoughts on the existing obesity guidance?

Dr. Kaul: I think it is in the best interest of the FDA and the sponsors to come up with a more comprehensive guidance that captures the true benefit-risk of weight-loss intervention in its intended "real-world" environment. Our committee felt it had enough criteria to work with efficacy, but lacked sufficient guidance from FDA on safety of weight-loss drugs. Changing the goal posts in the middle of the game is obviously not fair to the sponsors. On the other hand, I can't necessarily fault the FDA for being risk-averse, given their goal to protect the public, and especially given the dubious past track record of weight-loss drugs. So I think a standardized guidance document that clearly outlines the metrics for a favorable benefit-risk assessment, including weight loss, cardiometabolic profile, and morbidity and mortality outcomes a priori, would help the sponsors navigate the stormy waters of drug development and approval.

final thoughts

Kelly: Cardiovascular outcomes in the overall population may look better if the 27 million people with diabetes in the US and hundreds of millions more all over the globe could start doing better earlier on in their disease progression. It's likely that diabetes rates would drop dramatically if obesity could be better addressed. Are there any final thoughts on the obesity epidemic and implications for diabetes outcomes overall? The diabetes community is concerned about macrovascular disease, and they're also worried about people going blind, losing limbs, and developing other microvascular complications. What I'm most worried about is that we're creating an increasingly "elderly unwell" population that is growing dramatically, driven in large part by obesity that is so challenging to address.

Dr. Kaul: I agree that minimizing the risks of both microvascular and macrovascular disease is a critical clinical goal in the management of patients with diabetes. And lifestyle modification, including eating healthier, getting more physically active, and losing weight (aided by pharmacologic and non-pharmacologic interventions wherever appropriate), together with adequate control of risk factors such as blood pressure, lipid, and blood glucose, should be the major focus of our efforts to stem the rising tide of the obesity epidemic.

Kelly: Thank you so much for all of your insights, Dr. Kaul. It has been an incredible privilege to spend time better understanding your thinking about obesity, cardiovascular disease, and the US regulatory environment.

Dr. Kaul: It has been a pleasure speaking with you and Sanjay. Thank you.

What do you think?