FDA 2014 Advisory Committee Meeting for Novo Nordisk’s Saxenda (liraglutide 3.0 mg for obesity)

September 11, 2014; Hyattsville, MD - Draft

Executive Highlights

The FDA Advisory Committee meeting for Novo Nordisk’s Saxenda (liraglutide 3.0 mg for obesity) ended with a very positive 14-to-1 vote in favor of approval. While it became much more clear by the afternoon that the vote appeared comfortably positive, we believe that the final voting tally reflected that the panel really grasped the “benefit” element of the “benefit/risk” discussion, perhaps even better than expected. As an additional win, most panelists appeared willing enough to accept outcomes data from Victoza (liraglutide at a lower dose for diabetes) and its LEADER CVOT (scheduled to end in 2016) to provide reassurance on Saxenda’s safety. The fact that the panel did not propose a new CVOT specifically for Saxenda gave many in the room reason to breathe a sigh of relief.

We were also glad to see broad recognition of Saxenda’s benefits beyond weight loss, particularly on glucose control and the prevention of diabetes – panelists Dr. Barbara Hansen (USF, Tampa, FL) and Dr. Charles Stanley (University of Pennsylvania, Philadelphia, PA) both went so far as to say that the diabetes prevention benefit is perhaps the most important element of Saxenda’s efficacy profile. We note that the importance of diabetes prevention is definitely making it into more mainstream discussions, and appeared to receive even more attention than it has in previous years. Another point that we were glad to see come up in discussion was the importance of recognizing the heterogeneity in the obese patient population, and that a drug has value even if it leads to meaningful weight loss in only a subset of super-responder patients. Patients in the Open Public Hearing (OPH) made it clear that they wanted more latitude to decide benefit/risk on an individual basis, and HCPs called for more tools to do so. In the efficacy discussion, the FDA presented a modified analysis of the phase 3 study dropouts that reduced the calculated weight loss to very slightly below the Agency’s benchmarks. While this was initially worrying, the FDA clarified that even with the revision, Saxenda’s efficacy data still meets the “spirit” of the FDA guidance and does not preclude approvability – a statistical debate, but a largely moot point in the larger scheme.

There were a number of safety questions on the agenda entering the meeting, including an increase in heart rate and imbalances in pancreatitis, gallbladder disorders, and thyroid and breast cancers – after a morning in which the discussions went positively for Novo Nordisk, the afternoon continued well (this is often when things blow up for sponsors) and none ever appeared to become deal-breakers. Partly, this is because the morning presentations addressed each of these issues and others so well: Novo Nordisk global Chief Medical Officer Dr. Alan Moses was masterful, in our view, and provided reassurance that the safety data on Saxenda does not show any definitive risk of pancreatitis or tumors. The FDA, for its part, drew upon post-marketing data on Victoza (liraglutide up to 1.8 mg for type 2 diabetes) while underscoring the limitations of non-RCT data – they didn’t ultimately have so many concerns about generalizing data from the lower diabetes dose. During the committee discussion, it was actually breast cancer that received the lion’s share of the time, which we hadn’t predicted, followed by cardiovascular concerns. Notably, panel cardiologists Dr. William Hiatt (University of Colorado, Aurora, CO) and Dr. Brendan Everett (Harvard Medical School, Boston, MA) advocated for a shift away from a rigid MACE-centric viewpoint on cardiovascular safety towards a more flexible approach focused on issues that are of greatest interest given an individual drug’s mechanism. This seems incredibly rational to us. Individual panelists emphasized the importance of getting more data on gallbladder disorders, pancreatitis, and breast cancer, but it was generally agreed, with the exception of the panel oncologist Dr. David Kelsen (who wasn’t able to sway anyone with his opinion,) that this could be done post-marketing.

Even with the day’s overwhelmingly positive vote, the FDA will still have quite a bit to consider before passing along an approval – still, we believe that is very likely and would be shocked by a non-approval or a CRL following such an overwhelmingly positive vote. Novo Nordisk’s post-marketing plan involved a set of new and ongoing studies, but the FDA may be looking for even more post-marketing data on cancer risk, among other issues – we assume that much of this could be achieved via monitoring or through the long-term studies that are already ongoing.

There is also the question of how to approve Saxenda as a first-line agent for obesity when Victoza is not approved as a first-line agent for type 2 diabetes – the labels would overlap, meaning patients with type 2 diabetes could receive Saxenda as a first-line agent to treat their obesity/diabetes. Most panelists appeared willing to accept a degree of inconsistency between labels in order to preserve a first-line indication for Saxenda in obesity. This was terrific to see from a patient perspective, since there is no universally agreed upon first-line pharmacotherapy for obesity.

There is also much to consider in terms of how to address to the multiple unanswered safety related questions on the product label and REMS program – strict warnings regarding breast cancer or a risk of spontaneous abortions that could deter female patients, though we believe there is a large audience of potential users. Despite the work yet to be done, today’s vote was certainly a major win for patients with obesity, especially those with other comorbid conditions such as diabetes, a major win for researchers, and of course, a major win for Novo Nordisk. As well, we see it as a positive for other small companies with obesity compounds since we know the area will receive far more awareness focus now that Novo Nordisk is firmly in the game.

Efficacy

  • Broad agreement emerged during the panel discussion that Saxenda was approvable at least on the basis of efficacy, as the drug met at least one of the two FDA efficacy benchmarks for obesity weight loss drugs in every phase 3 trial (>5% weight placebo-adjusted weight loss or >35% on drug achieve weight loss of >5%, double that of placebo). In fact, during voting block remarks, the majority of panelists cited this data (along with the level of unmet need in obesity) as their primary reason for approving Saxenda. In SCALE-Obesity and Prediabetes (n=3,731), using a modified last observation carried forward (LOCF) method to account for study dropouts, mean body weight loss was 8.0% in the Saxenda arm vs. 2.6% in the placebo arm. Citing the ~23-36% dropout rate in phase 3, the FDA argued that the sponsor’s LOCF approach slightly exaggerated Saxenda’s efficacy and underreported weight loss in the placebo arm. In the agency’s re-analysis, which tried to model what happened to study dropouts based on “retrieved dropouts” that had an end-of-study visit, Saxenda very narrowly missed some of the FDA weight loss benchmarks.
    • Ultimately, panelists were not deterred by the FDA analysis and concurred with the FDA’s assertion that the efficacy data still met the “spirit” of the efficacy benchmarks. During afternoon dialogue, many agreed that the true efficacy of the drug likely is somewhere between the sponsor’s analysis and that of the FDA. Dr. Brendan Everett (Harvard Medical School, Boston, MA) was one of the first to advocate this point, noting that the confidence intervals for both analyses overlap a great deal; in this sense, he suggested that “reality” was somewhere in between these findings.
  • We were pleased to hear panelists recognize the importance of positive secondary endpoints related to diabetes. Dr. Barbara Hansen (University of South Florida, Tampa, FL) went so far as to emphatically state that Saxenda’s ability to postpone or prevent the onset of type 2 diabetes is one of the drug’s most important secondary effects. Data shared in the morning demonstrated that Saxenda led to significant placebo-adjusted reductions in A1c in type 2 diabetes patients (-0.9% from a baseline of 8.0%). In SCALE-Obesity and Prediabetes, among people with prediabetes at screening, those on Saxenda were more likely to revert to normoglycemia at week 56 (70%) compared to those on placebo (32%). Similarly, of those with normoglycemia at screening, fewer of those on liraglutide 3.0 mg (7%) developed prediabetes compared to those on placebo (20%). Multiple panelists cited these data as key in assessing the comprehensive efficacy of Saxenda, as it is a more direct measure of clinical benefit as opposed to weight loss alone.
  • We thought panelists made a strong case that the more robust results in certain highly responsive subgroups should be considered, rather than solely relying on an average that includes many non-responders. According to data presented by Novo Nordisk, roughly 22% of patients experienced a >10% reduction in weight in SCALE-Obesity and Prediabetes, suggesting that there was indeed a population of super-responders – although we don’t know what percent lost 15% of weight or even more, even if it is 1%, that could translate into a most meaningful drug for millions of obese patients. Panelists noted that the obese patient population is quite heterogeneous, which calls for a heterogeneous armamentarium of treatment options – we are glad for the sake of patients to see this message get across and do think much more will be needed to educate providers on this front. Given that Saxenda non-responders are likely to – and “should,” as was noted by Dr. Hanssen – stop drug treatment when it is not working, panelists concurred that the robust weight loss in a subset of the obese population studied speaks strongly enough to the clinical efficacy of the data.
    • The issue of a potential stopping rule (which would mandate cessation of therapy if a certain amount of weight loss is not achieved) was raised during the panel discussion – this will emerge as a significant part of the labeling discussion. Some panelists, including Dr. Martha Nason (NIH, Bethesda, MD) and Dr. Susan Yanovski (NIH, Bethesda, MD) are looking to the FDA to put a stopping rule in place for Saxenda, similar to the three other recently approved anti-obesity pharmacotherapies. However, sponsor remarks were somewhat evasive on that front and suggested the labeling discussions in the coming months will determine what happens on this front. We could certainly see the rationale behind avoiding a strict stopping rule, as Saxenda has a number of benefits beyond weight reductions – it is very possible that a patient might not achieve a high percentage of body weight loss, but could see improvements in their prediabetes or other cardiovascular risk factors, and would best be served by staying on the drug.
    • This is all to say, of course, that individualizing obesity and diabetes management is sorely needed. The proponents of the “stopping” rule clearly want to avoid patients being on Saxenda for months or years if it is not working and if the patients are not receiving other benefits – we don’t imagine any clinician, any payer, or any patient would argue with that. Of course, an alternative to stopping rules is to expand education among HCPs so that management is customized for each patient. This requires more education and resources, of course, and perhaps is wishful thinking ultimately; that said, we think patients would not all benefit from the stopping rule and think other resources can be shifted to avoid patients taking a medicine that isn’t working. As a sidenote, if in fact Saxenda isn’t working, we do not think most patients would continue to take it given it’s an injectable drug.  

Applicability of Victoza Data

  • One important question before the Advisory Committee was how (or whether) to use the pivotal and post-marketing data on Victoza (liraglutide 0.6-1.8 mg for type 2 diabetes) to assess Saxenda’s safety. The FDA presentations listed the primary differences between the use of liraglutide for obesity vs. type 2 diabetes:
    • The dose being investigated for obesity is 3.0 mg, while the dose range approved for type 2 diabetes is 0.6-1.2 mg.
    • Victoza is not recommended as a first-line therapy for type 2 diabetes, whereas there is no proposed limitation on using Saxenda as a first-line therapy for obesity. Many type 2 diabetes patients would fall into the label of Saxenda, meaning they could conceivably use liraglutide 3.0 mg before any other anti-hyperglycemic medications.  
  • Although committee chair Dr. Kenneth Burman (Washington Cancer Institute, Washington, DC) expressed some uncertainty about relying too heavily on findings from a different liraglutide dose, the majority of the panel seemed to view the data on Victoza as applicable to Saxenda. A chart displayed on the FDA slides (in fact!) showed that the body weight distribution in the liraglutide trials for type 2 diabetes and obesity overlapped significantly. Meanwhile, increases in heart rate were similar between the two doses, along with many other safety findings pointed out in Dr. Alan Moses’ sponsor presentation. During discussion, Dr. William Hiatt (University of Colorado, Aurora, CO) emphatically stated that drawing data from obese patients with diabetes makes a lot of sense, as patients with both conditions might be more likely to take these drugs. He did acknowledge the question of whether findings from a higher-risk population (diabetes and obesity) can always be applied to lower-risk populations (obese patients without diabetes).
  • Much of the dialogue on the question of Victoza data centered on the ongoing LEADER cardiovascular outcomes trial (CVOT). Panelists agreed that signs of harm (cardiovascular or otherwise) from the trial would have highly worrying implications for the higher liraglutide dose in Saxenda. However, the trickier hypothetical situation at hand was what a clean bill of health for Victoza in LEADER would mean for Saxenda. Statistician Dr. Martha Nason (NIH, Bethesda, MD) suggested that she would not find a neutral LEADER topline result necessarily conclusive, but that additional sub-analyses of those results might be able to assuage any lingering concerns. See more on the LEADER discussion in the “Cardiovascular Safety” section below.
  • To conclude the discussion on the applicability of Victoza data, the FDA’s Dr. James Smith asked the panel if they thought there was any safety issue associated with Victoza that could be magnified at higher doses. Dr. Christy Boling Turer (University of Texas Southwestern Medical Center, Dallas, TX) stated that she would be worried about increased hypoglycemia with a 3.0 mg dose of liraglutide, especially in combination with a sulfonylurea. (In this case, we would assume the SFU could be easily taken out of the plan but this wasn’t discussed.)

Cardiovascular Safety

  • Discussion on the topic of cardiovascular safety was quite nuanced, with recognition of the need for responsible use of research resources when investigating macrovascular safety. Perhaps unsurprisingly, panel cardiologists Drs. William Hiatt (University of Colorado, Aurora, CO) and Dr. Brendan Everett (Harvard Medical School, Boston, MA) provided most of the comments in this topic area. Notably, Dr. Hiatt emphasized that based on the available data, he did not have any major concerns about Saxenda’s cardiovascular safety. The core consensus was that more CV outcomes data is needed on Saxenda and particularly on the increase in heart rate seen with both Victoza and Saxenda, but that the need is not as urgent as it has been for other recent diabetes drug candidates. The hazard ratio for MACE of 0.40 (95% CI: 0.15-1.05) seen in Saxenda’s phase 3 studies was based on only 17 total events, but it nonetheless did not give panelists a reason to suspect that the increased dose might lead to an increase in cardiovascular risk.
  • Although there was a slim chance coming into this meeting that panelists or the FDA would raise the possibility of a new CVOT specifically for Saxenda, there was little to no interest in recommending that approach during the panel discussion. We were struck to hear Drs. Hiatt and Everett offer very resource-conscious perspectives on the issue of running a new CVOT. Dr. Hiatt noted that it would be easy for an Advisory Committee to suggest studying a larger dose, but that doing so would likely not be the best use of research dollars. He emphasized the importance of being more thoughtful in terms of how research dollars are allocated, and of avoiding a MACE-centric cardiovascular safety evaluation paradigm if a given agent does not pose a plausible risk of ischemic heart disease but might instead (in the case of liraglutide) have a more plausible impact on heart failure due to changes in heart rate. Dr. Everett expressed wariness about requiring every obesity drug to have a CVOT, partially because it could become difficult to find enough patients to enroll in those studies. He put forth an open question about how cardiovascular safety could be better evaluated post-marketing, perhaps using study formats other than randomized control trials. Companies’ choice of endpoints for CVOTs is mandated by the FDA, and we believe that these panelists’ words will give the agency food for thought in terms of the restrictions placed on drug development – particularly when considered along with comments from the recent “FDA Public Hearing on the Confidentiality of Interim Results in Cardiovascular Outcome Safety Trials” in Washington last month.
  • A small increase in heart rate emerged as the major cardiovascular safety issue in the discussion on Saxenda. Although the increase in resting heart rate seen with Saxenda in its phase 3 trials was on the order of 2-3 bpm, a smaller continuous heart rate monitoring study found much larger increases (6-7 bpm) at certain times of the day, particularly at night. [In hindsight, we should have remembered from the EMDAC for Vivus’ Qsymia that this would be a big deal.] Data from dose-finding studies showed that the increase in heart rate was not dose-dependent, and as a result there was some cautious confidence among panelists that the LEADER trial would be able to provide information on both Victoza’s and Saxenda’s safety with regard to heart rate; cardiologist Dr. William Hiatt was particularly adamant on this front. The fact that liraglutide is associated with beneficial effects on other CV risk factors such as blood pressure and inflammation likely defused some of the potential worry on the heart rate issue – we expect that LEADER results will, of course, be very closely watched. These are expected in mid-2016.
    • Drs. Hiatt and Everett noted that increased heart rate might not have much of an adverse impact on MACE events, but could have a greater impact on heart failure or arrhythmias. As a result, they emphasized the need to adjudicate a wide range of endpoints (including heart failure and arrhythmias) in outcomes trials for obesity medications; as Novo Nordisk’s Dr. Alan Moses noted both heart failure and arrhythmias are being adjudicated in the LEADER trial.

Breast and Thyroid Cancer

  • The imbalance in breast cancer events seen in the SCALE trials was a major focus of the panel discussion, even beyond the attention given to pancreatitis and gallbladder-related adverse events. We assume this was because the breast cancer signal was relatively unexpected. Although pooled phase 3 data indicated that the incidence of breast cancer was higher in the Saxenda arm than the placebo arm, the number of events was very low (14 with Saxenda and three with placebo). Several speakers noted that there is no known plausible biological mechanism connecting liraglutide with breast cancer, and that the effect could have been due to chance or confounding factors like detection bias. Nonetheless, many panelists expressed serious concern about the potential increased risk and urged the FDA to take proactive steps to mitigate it.
    • A pooled analysis of phase 3 trial results for Saxenda found an imbalance in breast cancer incidence between the treatment and placebo groups. A total of 14 patients (0.6%) treated with Saxenda and three patients (0.3%) treated with placebo developed breast cancer; of those cases, 11 in the Saxenda group and two in the placebo group were malignant. According to the FDA’s analysis, the imbalance translated to a relative risk ratio of 2.98; neither that ratio nor any individual point estimates were statistically significant (due to the small number of events). FDA epidemiologist Dr. Christian Hampp ultimately concluded that the data does not “confirm or exclude” an increased risk of breast cancer associated with Saxenda, and we assume this will be closely watched through monitoring.
    • Several speakers questioned the biological plausibility of a link between Saxenda and breast cancer and suggested that the imbalance was likely due to chance or detection bias. Novo Nordisk Chief Medical Officer Dr. Alan Moses explained that there is no known association between GLP-1 agonists and breast cancer, and that there have been no concerning signals in any preclinical carcinogenicity studies with liraglutide. Breast cancer specialist Dr. Pamela Goodwin (University of Toronto, Toronto, Canada), who was part of the sponsor contingent, gave her view that there is no plausible mechanistic basis for a link between Saxenda and breast cancer and that the pattern of incidence (a large increase in events soon after trial initiation rather than a gradual, longer-term elevation) was not suggestive of a causal effect from the drug. Indeed, she noted that hormone replacement therapy is a known cause of breast cancer, and the progression takes several years. She hypothesized that more frequent screening among women who lost more weight was the most likely explanation for the imbalance, as past studies have shown that obesity is correlated with less frequent breast cancer screenings.
    • Despite the ambiguity surrounding the breast cancer signal, several committee members expressed serious concerns about the potential risk. Dr. David Kelsen (Memorial Sloan Kettering Cancer Center, New York, NY) cited cancer risk as one of the primary reasons for his vote against Saxenda’s approval (the only “no” vote of the day), saying that even a small increase in risk is clinically meaningful with such a large target population (more than 79 million obese patients plus those with a BMI of 27 kg/m2 and one weight-related co-morbidity). He suggested that investigators should conduct follow-up screenings of all trial participants (both drug and placebo) to determine whether breast cancer rates eventually equalized between the two groups – he would consider it a significant red flag if they did not. That proposal, along with a suggested contraindication for patients at particularly high risk for breast cancer (perhaps due to family history or a known genetic marker), received several endorsements from other panelists, though none of them ultimately considered the risks to be serious enough to preclude approval. We were surprised Dr. Kelsen would vote “no” based on this, given his comment that the sponsor likely already has this data. Breast cancer is very uncommon in men so we would at least have expected a vote for a conditional approval from him. Of course, this reflects the nuance that often lies beyond yes/no Advisory Committee votes.
  • Concerns about thyroid C-cell tumors appear to have subsided somewhat since Victoza’s approval, as worrisome preclinical findings have thus far not translated to increased risk in humans in clinical trials. As a reminder, the label for Victoza includes a boxed warning stating that the drug has been found to cause thyroid C-cell tumors in rodents and that it is unknown whether this could translate to an increased risk of medullary thyroid carcinoma (MTC), a rare form of thyroid cancer, in humans. The clinical data package for Saxenda did show a slightly higher incidence of thyroid cancer in the treatment arm, but event rates were very low (seven in the Saxenda group and one in the placebo group), and the only case of MTC occurred in the placebo group – the rest of the cases were the more prevalent papillary or follicular cancer subtypes. In marked contrast to the extensive discussion of thyroid cancer during the Advisory Committee meeting for Victoza in 2009, most panelists seemed satisfied with Novo Nordisk’s proposed risk management plan for Saxenda (the proposed label includes a similar boxed warning as Victoza’s and the drug will be added to the ongoing national MTC registry for other GLP-1 agonists).
    • The only clinical data on MTC risk comes from spontaneous post-marketing reports, which have several limitations and have not suggested a causal relationship with liraglutide. Dr. Marina Zemskova (FDA, Silver Spring, MD) explained that MTC is extremely rare (the incidence rate is ~0.22 cases per 100,000 patient-years) and that there was no evidence that the few cases reported through the national MTC registry have been associated with exposure to Victoza.
    • Panelists generally agreed that it would not be necessary to include a recommendation in Saxenda’s label against its use as a first-line therapy. Victoza’s label recommends that the drug only be used as a second-line agent for the treatment of type 2 diabetes due to the concerns about thyroid cancer, and one of the discussion questions before the Advisory Committee asked whether such a stipulation should be added to the proposed label for Saxenda. Although there was some debate over how problematic it would be for the two drugs to have inconsistent labels, the panelists largely concluded that it would be difficult in practice to advise against using Saxenda as a first-line drug therapy for obesity, given the lack of a consensus on a first-line drug in obesity comparable to metformin for type 2 diabetes. Dr. Daniel Budnitz (CDC, Atlanta, GA) was a lone voice on this issue, suggesting that the FDA should strive for consistency in its labeling rather than accepting divergent labels on the thyroid cancer risk with liraglutide.

Other Safety Concerns

  • Overall, we were relatively surprised to see that gallbladder-related adverse events were not a bit more heavily discussed as a safety concern during the discussion session, especially since Dr. Mads Thomsen (Chief Medical Officer, Novo Nordisk) suggested that the imbalances in gallbladder-related adverse events would be a main focus of the Advisory Committee during Novo Nordisk’s 4Q13 financial update. Drs. Susan Heckbert (University of Washington, Seattle, WA), Brendan Everett (Brigham and Women’s Hospital, Boston, MA), and Kenneth Burman (Washington Hospital Center, Washington, DC) mentioned that they had some lingering concerns about gallbladder disorders, but did not discuss these in detail. The most in-depth commentary came from Dr. Daniel Budnitz (CDC, Atlanta, GA), who highlighted the high rate of cholecystectomies that were needed in patients that experienced gallstones (~70% of patients with gallstones or gallbladder inflammation). He also noted that we have no long-term data on gallbladder-related events, suggesting that the risk could potentially be much higher with more chronic usage of Saxenda. In her voting rationale, Dr. Christy Boling Turer (University of Texas Southwestern Medical Center, Dallas, TX) pointed out that gallstones may only be related to weight loss, and that better tools are needed to inform patients and providers on what adverse effects are possible with significant weight loss. However, otherwise, the gallbladder imbalance was far from dominating the conversations surrounding safety concerns.
    • The FDA presentations suggested that Saxenda may increase gallstone formation dependently or independently of weight loss – the most likely answer is a combination of the two, as gallstone formation was increased with greater weight loss, but in patients with a similar level of weight loss, there was a greater incidence of gallstones with Saxenda.
    • During his morning presentation on Saxenda’s safety profile, Novo Nordisk Chief Medical Officer Dr. Alan Moses noted that the company expects to address the issue through patient education. The company’s post-marketing plan also involves a new mechanistic study on gallbladder-related adverse events.
  • Similarly, pancreatitis concerns also received little attention, although it was mentioned in a few panelists’ voting rationale statements (see below). Dr. Susan Yanovski (NIDDK, NIH, Bethesda, MD) addressed pancreatitis as an adverse event she would like addressed in a post-marketing setting. Drs. Charles Stanley (University of Pennsylvania, Philadelphia, PA) and Brendan Everett (Brigham and Women’s Hospital, Boston, MA) also listed pancreatitis as an important safety concern.
    • During his safety presentation, Dr. Moses noted that most instances of pancreatitis were mild. Novo Nordisk is conducting a post-marketing pharmacovigilance program with Victoza, and the data so far has not shown an imbalance in acute pancreatitis with Victoza. 
    • The FDA agreed that there is not enough information at the moment to determine an association between liraglutide and pancreatitis. This sentiment is in line with the message of a joint FDA/EMA article in NEJM earlier this year, which argued that current data on the incretin-pancreatitis controversy is insufficient to precipitate a large-scale change in prescribing behavior.
  • As expected, the panelists did not voice any concerns about the risk of psychiatric events. Drs. William Hiatt (University of Colorado, Aurora, CO) and Barbara Hansen (Morsani College of Medicine, Tampa, FL) both mentioned that the data surrounding psychiatric events and suicidality warranted little concern, perhaps because multiple questionnaires did not show an increase in depression or suicidal ideation with Saxenda. Following the rimonabant experience in 2007 (the drug was found to be associated with multiple neuropsychiatric adverse events), psychiatric safety has been closely evaluated for all obesity drugs (see related talk from Dr. David Heal from 2009 at the “5th Annual Effective and Advanced Treatments for Diabetes Conference”). However, given the localization of liraglutide’s action to the hypothalamus, there is less of a plausible concern about psychiatric adverse events with Saxenda as with other more neuromodulatory obesity drugs.

FDA Summary of Saxenda’s Risk/Benefit Profile

Benefits

Risks

  • Weight loss efficacy demonstrated across trials and patient population, although magnitude of treatment effect impacted by missing data
  • Expected improvements seen in glycemic parameters
  • Blood pressure and lipid changes consistent with weight loss
  • Heart rate consistently increased; overall CV risk uncertain
  • Imbalances in gallstones and pancreatitis observed; contribution of liraglutide and/or weight loss possible
  • Excess cases of suicidality observed in liraglutide arm; most patients continued on treatment
  • Medullary thyroid carcinoma observed post-marketing; contribution of liraglutide uncertain
  • Imbalance in breast cancer observed in clinical trials; cases do not support or deny the potential role of liraglutide in cancer promotion or progression
  • Hypoglycemia, renal impairment, immunogenicity, GI symptoms

Open Public Hearing

We were wowed by the turnout at the Open Public Hearing for Saxenda – over 20 speakers shared their thoughts, the vast majority of which were enthusiastically in favor of approval. There is great potential for patient advocacy before the FDA to heighten further, and we are excited to see similar or even greater turnout at future EMDAC meetings and other opportunities.

  • Ms. Kelly Close (The diaTribe Foundation, San Francisco, CA) discussed one of the biggest unmet needs in medicine today: prediabetes – see her slides here. In strongly advocating for the approval of Saxenda, Ms. Close contrasted the profound prevalence of prediabetes (86 million patients) with the silent nature of the disease – she shared data that nine out of ten individuals with prediabetes do not know that they have it. These are patients, she noted, that are likely to become type 2 patients in a decade or less, presenting a significant public health concern. In light of this burden, she highlighted the positive findings of SCALE-Prediabetes and Obesity that showed that Saxenda effectively prevented the progression from prediabetes to type 2 diabetes, and also helped prediabetes patients revert to normoglycemia. Ms. Close urged the panel to consider that there are presently zero drug classes approved to treat prediabetes, while there exist many more classes to treat fewer patients with HIV/AIDS (seven classes, ~1 million patients), cancer (11+ classes, ~13 million patients), and diabetes (11 classes, ~29 million patients). Ms. Close proposed the approval of liraglutide as the first step toward addressing this disparity.
  • Our very own Mr. Adam Brown (Close Concerns, San Francisco, CA) emphasized the dire state of obesity in the US and the need for more treatment options – see his slides here. He opened by presenting on the tremendously high medical care costs of obesity ($190 billion per year, and an astounding $1.1 million spent during the three minutes of his presentation) and referred to the recent “State of Obesity” data to show the high rates of obesity across the country (a new obesity prevalence category was needed this time around to account for two states with an obesity rate of >35%). In addition, he pointed out increasing trends in severe obesity and the higher risk of death for patients with severe obesity (a factor of 2.5 compared to those of normal weight, along with 7-14 years of life lost). With these alarming statistics behind him, Mr. Brown called for more options for patients and HCPs to combat the biological and environmental challenges of weight loss.
  • Ms. Nancy Liu (dQ&A, San Francisco, CA) presented results from a survey of over 3,000 patients with type 2 diabetes that illustrated the challenges of living with obesity – see her slides here. Several statistics demonstrated the risks associated with obesity (obese respondents were three times as likely to have an A1c >9% compared to non-obese patients) and the enormous unmet need for more treatment options. Around 77% of the patients surveyed said they were interested in losing weight, but only 2% were currently taking a weight loss medication, and a combined 82% said they would not be willing to undergo bariatric surgery. Several compelling patient testimonials (“I swear, my weight keeps going up no matter what I do”… “If I got my weight under control, my diabetes would be in a better place”) provided firsthand accounts of the challenges of managing obesity with diabetes and the need for a wider range of tools.
  • In an emotionally charged and very vivid presentation, patient Mr. Dan Wright told his personal story with obesity, asking the panel to “give doctors all the tools they need to treat us.” See the footage here. He expressed frustration with the limitations of lifestyle intervention, the misconceptions that “it’s just willpower”, and with the widespread belief that losing weight is the same as getting healthier. Mr. Wright then explained how his move to medications gave him the opportunity to not only lose weight, but to also “get healthier” by improving his A1c levels, lipid profile, and resting heart rate. Concluding, he passionately spoke of the stigma and blame that those with obesity suffer from, pleading the panel to think differently and to not “deny doctors the tools they need because the people outside this room say we [obese people] don’t deserve it.”
  • Mr. Manny Hernandez (Diabetes Hands Foundation, Berkeley, CA) presented moving testimonials from the Diabetes Hands Foundation forum, calling Saxenda a “very valuable option for diabetes patients.” In a TuDiabetes forum on Victoza that received over 1,400 views, nearly 75% of the responses found the drug’s effects helpful in losing and keeping off weight. This included a woman who had undergone bariatric surgery and regained the weight, along with an amazed mother whose daughter with diabetes lost 12 pounds after taking the drug. While acknowledging that Saxenda does come with side effects (“there’s no such thing as a free lunch”), Mr. Hernandez presented testimonial evidence of the significant benefits, which supported his call of a risk/benefit profile favoring approval.
  • Dr. Lee Kaplan’s (Massachusetts General Hospital, Boston, MA) remarks emphasized the patient-to-patient variability that exists within the obese population. He stressed that efficacy of therapies varies in different obese patient populations, and that hyper-focusing on the mean efficacy data alone is counterproductive. Dr. Kaplan noted that the limited effectiveness of therapies in the overall population has diminished enthusiasm for other therapies, despite their efficacy in certain populations. The availability of additional therapies with novel mechanisms of action, he noted, will provide options for the majority of obese individuals who do not respond to current therapeutic options. He acknowledged that it is important to specifically establish which therapies are right for which patients, but noted that we first need to make those options available; we were captivated by Dr. Kaplan’s words and feel that this appeal for greater availability of drugs resonated with the panel as well.
  • On behalf of people with diabetes, Ms. Rebecca Killion (speaking on behalf of the ADA, instead of in her usual spot at the Advisory Committee patient representative) eloquently called for “more weapons” when battling diabetes and obesity. She emphasized the “inseparable” nature of diabetes and obesity and the vast amount of harmful health consequences obesity brings. Ms. Killion described her personal battle between diabetes and weight management as a “demoralizing game of whack-a-mole” – improvements in A1c levels came with weight gain and vice versa (“insulin can really pack on the pounds”). Concluding, she emphasized that weight gain as someone with diabetes is not “frivolous” and that more tools to manage obesity and diabetes are needed.
  • AACE president-elect Dr. George Grunberger (Grunberger Diabetes Institute, Bloomfield Hills, MI) stressed the need for a new paradigm to confront the obesity epidemic, as lifestyle intervention has proven insufficient to avert the increasing medical and economic consequences of the disease. He cited the new framework that emerged from the recent AACE consensus conference, which classifies obesity as a chronic disease and calls for new interventions, including drugs, to decrease the “catastrophic costs” of this epidemic. He urged the FDA to keep the immense health benefits of weight reduction in mind as it considers Saxenda, and to give patients and providers the freedom to conduct their own assessments of the drug’s risk/benefit profile.
  • Dr. Ted Kyle (Obesity Action Coalition, Tampa, FL) implored FDA members to consider the tremendous impact their decision will have on research and innovation in the obesity arena, as well as on the lives of individual patients. He claimed that rejection of obesity drugs in the past had led companies to avoid investing in new options like Saxenda that could help fill the enormous treatment gap for obesity (between lifestyle intervention and surgery), and he argued that liraglutide’s well-understood profile provided sufficient support for approval.
  • “Don’t let fatness get in the way of your assessment of obesity,” cautioned Dr. Scott Kahan (George Washington University, Washington, DC), explaining that the view of obesity as solely a cosmetic issue or a lifestyle choice can easily interfere with objective risk-benefit calculation. He urged panelists to approach obesity as they would any other chronic health condition and to ensure that the appearance of “fatness” does not stand in the way of an “informative and humane decision process” regarding the medical issue of obesity.
  • Mr. Joe Nadglowski (Obesity Action Coalition, Tampa, FL) advocated for more obesity treatment options and called for changing the name of “weight loss” or “weight management” drugs to “obesity” drugs. He introduced the Obesity Action Coalition’s (OAC) mission as ensuring individuals with obesity have the ability to treat their condition, if they wish to do so. Noting that the sponsor has provided enough information to approve Saxenda, he pushed for obesity to be treated as any other disease and for its risk/benefit profile to be evaluated on the same standards.
  • Dr. Sean Wharton (Toronto East General Hospital, Toronto, Canada) reiterated the notion that clinicians are currently short of tools for addressing obesity. Acknowledging that bariatric surgery – among other therapies – has been a “game-changer” for patients, Dr. Wharton stated it’s not a solution for everyone. He emphasized that individual patients need individualized solutions.
  • SCALE participant Ms. Laurie Weinstein explained that after years of futile efforts to lose weight and avoid the fate of her mother and grandmother, both of whom lost their lives to complications of diabetes, taking Saxenda enabled her to improve her relationship with food and served as a catalyst for effective lifestyle changes. She said that the drug is now a “valuable piece of my puzzle” that has dramatically improved her health and quality of life, and she stressed that patients and physicians are capable of conducting their own analyses of Saxenda’s risk/benefit profile.
  • Patient Ms. Jennifer Mackenzie spoke from personal experience with Saxenda as a participant in the SCALE-Prediabetes and Obesity trial. Ms. Mackenzie was nearing the prediabetes threshold when she entered the study, but notably was nowhere near that threshold following 52 weeks of treatment. However, she characterized the psychological transformation that resulted from pharmacotherapy as the greater benefit of therapy. The treatment enabled her to escape from the vicious cycle of negative physical and emotional consequences associated with lifestyle intervention and failure. She reported that Saxenda reduced her appetite and eliminated all cravings, notably more so than therapy with Qsymia did. This was the only direct testimony we heard, all afternoon, of Saxenda’s ability to successfully counteract what we know to be the addictive neurological properties of high-sugar foods. She described Saxenda as “the crutch she needed” to overcome that addiction and develop confidence that only comes with success. We think that analogy resonated with the majority of the panel, especially due to the comparison with currently approved agents.
  • Dr. Sammy Almashat (Washington, DC) opposed the approval of Saxenda, referencing various safety concerns. He commented that the efficacy is over-estimated by LOCF analyses and that when analyzed under a “reasonable set of assumptions about missing data,” the two largest trials failed to meet FDA’s approved criteria. Dr. Almashat emphasized that Saxenda’s cardiovascular safety has not been established and also pointed out the thyroid cancer issue, stating that liraglutide is the only FDA-approved drug that has caused tumors in mice. In addition, he touched on the risks of pancreatitis and teratogenicity.

Commentary From Voting Committee Members

Following the final panel vote, each panelist was allowed a few minutes to explain his or her decision. Included below are summaries of each statement, arranged in alphabetical order by speakers’ last name. Every panel member voted in favor of approval with the exception of Dr. David Kelsen (Sloan Kettering Cancer Center, New York, NY), who voted against approval.

  • Dr. Daniel Budnitz (CDC, Atlanta, GA) voted yes primarily because of the need for new weight loss treatment options, although he emphasized the need to be vigilant about potential harms. He recommended including the FDA’s alternative efficacy analysis on the drug label in addition to the sponsor analysis. One of his main points was the need for some level of consistency between the Victoza and Saxenda labels, especially with regards to the risk of thyroid C-cell tumors. He also suggested that the REMS program for Saxenda should include monitoring of gallbladder disorders and information on the increased risk of spontaneous abortion seen with Saxenda.
  • Committee chair Dr. Kenneth Burman (Washington Hospital Center, Washington, DC) stated that the risk/benefit balance for Saxenda, in his view, is quite close. He used most of his time to explain the many issues he still feels uncertain about, including the impact of the increase in heart rate on long-term cardiovascular outcomes and long-term development of tumors. He seemed uncomfortable with the idea of patients using Saxenda chronically, especially when they are no longer experiencing weight loss. We found this view somewhat surprising to hear from an endocrinologist, as the general consensus we hear among clinicians is that weight loss medications must be used continuously to prevent weight regain.
  • Dr. David Cooke (Johns Hopkins University, Baltimore, MD) said he found the efficacy data on Saxenda very convincing and did not think the safety data raised any new concerns beyond those already known with Victoza. Although he believes there is not yet enough evidence on the long-term risks or benefits of the drug, he feels it would be “reasonable” to allow patients and physicians to consider using it on an individualized basis.
  • Dr. Brendan Everett (Harvard Medical School, Boston, MA) said that there was a great deal of evidence to support Saxenda’s weight-lowering efficacy and agreed with many OPH speakers regarding the enormous unmet need for more obesity pharmacotherapies. He qualified his support by acknowledging the risks of gallbladder disease and pancreatitis and the “lingering concerns” about breast and thyroid cancer, and he expressed hope that data from the LEADER trial will provide valuable additional information about the drug’s safety profile.
  • Consumer Representative Diana Hallare (Visalia, CA) was reassured by Saxenda’s robust efficacy data despite harboring a host of concerns regarding the drug’s safety profile. She was particularly uncomfortable regarding the consequences of Saxenda’s interactions with other medications and the effect of beginning drug therapy following surgery (e.g., bariatric surgery, heart surgery, etc.). Ms. Hallare also called for follow-up studies examining the affect of Saxenda in minority populations.
  • Dr. Barbara Hansen (Morsani College of Medicine, Tampa, FL) noted that the Saxenda’s biggest impact may be on the prevention of diabetes. She commented that Saxenda would make a “good first agent for diabetes,” even though Victoza is not currently recommended as a first-line therapy for type 2 diabetes. Dr. Hansen believes that the efficacy data met the FDA benchmarks and that the long-term effects stand to be positive. While she emphasized the need for vigilance on the breast cancer issue, she did not personally view any of the safety questions as reasons to withhold approval.
  • Dr. Susan Heckbert (University of Washington, Seattle, WA) simply stated that the overall risk/benefit assessment for Saxenda was favorable and supported approval. She believes that the ongoing LEADER trial should provide valuable and much-needed information on long-term safety with liraglutide. This is expected to report in mid-2016.
  • Dr. William Hiatt (University of Colorado School of Medicine, Aurora, CO) believes that Saxenda’s efficacy profile is strong, but he expressed a number of recommendations to the FDA about what should be included in the label. Dr. Hiatt credited that secondary endpoints suggested the potential for meaningful clinical benefit and that the risks were not disproportionate to the benefits. He provided four recommendations to the FDA: (i) to include the range of weight loss parameters on the label; (ii) to include patient-reported outcomes on the label; (iii) to not include sleep apnea recommendations in the label, pending further study; and (iv) to include recommendations of discontinuing the drug in the event that a patient does not experience weight loss.
  • Oncologist Dr. David Kelsen (Sloan Kettering Cancer Center, New York, NY) was the only “no” vote on the panel. He explained that he was looking for more information on the possibility of malignancy with Saxenda, and that his questions could possibly be answered with data that Novo Nordisk could easily provide. Specifically, in order to prove the hypothesis that the imbalance in breast cancers was due to increased diagnosis associated with greater weight loss, he suggested examining the follow-up data from the placebo group to see if the lag time in diagnoses evened out over time. He also suggested that patient populations at high risk for cancers should be addressed in the product label.
  • Dr. Corby Martin (Louisiana State University, Baton Rouge, LA) commented that efficacy of Saxenda is well established and that its true efficacy profile likely lies somewhere between the sponsor’s analysis and that of the FDA. Dr. Martin also felt that the safety profile of the drug was “decently characterized” when signals were present, and that the causal effects of the drug were unclear and not yet established. Though he harbors some concerns about neoplasms and other conditions that take time to manifest, he suggested that those concerns can be addressed with appropriate post-market surveillance.
  • Patient representative Ms. Lynn McAfee (Council on Size & Weight Discrimination, North Port, FL) ultimately voted for approval despite several lingering safety concerns. She emphasized that patients should be advised to undergo regular mammograms and other cancer screenings due to the signals from clinical trials, and said that she would like to see a separate cardiovascular outcomes trial for Saxenda, as data only from patients with diabetes may not provide sufficient information on safety.
  • Dr. Martha Nason (NIH, Bethesda, MD) voted yes, but “only barely,” as she felt that many questions remained regarding Saxenda’s long-term safety. She stressed that providers must inform patients of the need for regular mammograms and other cancer screenings, and supported Dr. Susan Yanovski’s suggestion of stopping treatment after 12 weeks if the benefits were not deemed sufficient. She also advocated for further studies that explore Saxenda’s efficacy in various subgroups; given the heterogeneity of obesity, she believes it should be possible to better predict which patients will have the greatest response to the drug.
  • Dr. Charles Stanley (University of Pennsylvania, Philadelphia, PA) was not dissuaded by the FDA’s more conservative analysis of imputed data and emphasized that one of the secondary benefits of Saxenda is the prevention of diabetes. Though quiet during the afternoon’s discussion, he urged the audience to recognize the importance of even delaying the onset of diabetes, a factor that seemed to play a significant role in his decision. Dr. Stanley cautioned that Saxenda’s REMS plan should be designed to mitigate concerns about pancreatitis and cancer risk.
  • Dr. Christy Boling Turer (University of Texas Southwestern Medical Center, Dallas, TX) called for guidance and education for providers and patients on what to expect with drugs like Saxenda. Additionally, Dr. Turer cautioned against using Saxenda with sulfonylureas and called for further research in minorities. She expressed some caution with the breast cancer concerns, recommending that providers be made aware of patients’ use of other drugs that could promote breast cancer, such as contraceptives or hormone therapy. In addition, Dr. Turer supported more management of patients undergoing weight loss by further studying how to prevent gallstones and pancreatitis.
  • Dr. Susan Yanovski (NIDDK, NIH, Bethesda, MD) voted yes, highlighting that Saxenda has a favorable risk/benefit ratio but still expressed concerns about pancreatitis. She felt that the drug will be favorable for patients with obesity and dyslipidemia, but hopes that pancreatitis will be addressed in a post-marketing setting. In addition, she recommended that the label advise providers to discontinue therapy in patients who do not attain clinically meaningful weight loss in 12 weeks at the full dose (similar to Qsymia, Belviq, and Contrave).

 

--by Melissa An, Adam Brown, Varun Iyengar, Emily Regier, Manu Venkat, and Kelly Close