Sanofi/Lexicon submit sotagliflozin for type 1 diabetes to FDA and EMA – March 26, 2018

Lexicon announced today that Sanofi has submitted SGLT-1/2 dual inhibitor sotagliflozin to FDA and EMA. These regulatory filings come on time with the expected 1Q18. Assuming a standard 10-12 month review process, decisions for the US and EU markets are anticipated in 1Q19 (between January and March).

If all proceeds according to plan, sotagliflozin could be the first-to-market oral adjunct therapy for type 1 diabetes in the US, bringing patients a remarkable option for additional glucose-lowering on top of insulin, not to mention weight loss, lower total daily insulin dose, and increased time-in-range (much more on the glycemic outcomes beyond A1c below). This would be a big win for patients: the only adjunct to insulin available for type 1s today is AZ’s injectable Symlin (pramlintide), which has not gained much commercial traction since launch due to sentiments about it being challenging to use. Notably, thought leaders have positioned oral adjuncts as filling an unmet need in type 1 diabetes care and our understanding from many type 1 patients taking SGLT-2s off-label is that it’s far easier to do this than to take another injectable that has itself been associated with severe hypoglcyemia (a problem for many with Symlin is getting the insulin dosing changes correct).

Given all this, it’s great to see Lexicon/Sanofi’s continued commitment to sotagliflozin. In fact, the companies previously announced a timeline of 1H18 for regulatory submissions, but were always targeting “the earliest part of 1H18” for an NDA and MAA (Marketing Authorization Application).

Earlier this month, AZ submitted SGLT-2 inhibitor Forxiga (dapagliflozin) to EMA for a type 1 indication. This means that dapa might be first-to-market in Europe, with a final EMA decision also expected in 1Q19, although sotagliflozin could still hit EU pharmacies first depending on launch timing. We thus view the two candidates as essentially neck-and-neck in Europe. AZ plans to file Farxiga for type 1 with FDA in 2H18.

Considerations for FDA: DKA

• We expect FDA will convene an Advisory Committee for sotagliflozin prior to making a decision. This is only our speculation for now (neither Lexicon nor Sanofi have indicated that an Advisory Committee meeting is forthcoming), but we have reason to suspect this will be the case. An oral drug for type 1 diabetes is unchartered territory for regulators, as sotagliflozin is the first of this class to reach FDA’s desk. That said, EMA’s new draft guidance for diabetes drugs includes – for the first time – a section on oral adjunct treatments, suggesting to us that regulators in Europe are open to this emerging therapy class and have been thinking about how to approach these applications for sotagliflozin and dapagliflozin. EMA advises that all phase 3 trials for these candidates use an insulin optimization period to isolate added benefit, which was true of inTandem1 and inTandem2, but not inTandem3 because investigators sought to create trial protocol that more closely resembled real-world clinical situations – and, the combination of inTandem1-3 data is all the more compelling for it. EMA also recommends using hypoglycemia as a co-primary endpoint in these clinical trials, which we’d love to see, particularly given reluctance from regulators to add secondary endpoint data to product labels (we hope this changes over time).
• We’ve heard significantly less from FDA on the prospect of oral adjuncts for type 1 diabetes (their drug guidance remains unchanged from over a decade ago), so it remains to be seen how the agency will respond to sotagliflozin (and eventually dapagliflozin), though we’re hoping for a swift approval. We anticipate strong HCP and patient support for this, given the type 1 success seen to-date. We are curious about AZ’s thought process in filing dapagliflozin for type 1 in Europe ahead of the US: Did the company anticipate a smoother regulatory pathway for an oral adjunct medicine in the EU, given EMA’s draft guidance? If an Advisory Committee panel does gather to discuss sotagliflozin in the US, we would expect a strong showing of patient advocates at the meeting; this will be important to ensure that the patient voice is heard on these novel additions to the type 1 diabetes treatment toolkit, and so that FDA clearly understands the need for alternatives in type 1 diabetes care.
• We strongly suspect there will be questions from FDA and EMA surrounding DKA, which was more common with sotagliflozin vs. placebo in the inTandem program (e.g. 21 events in the sotagliflozin arm of inTandem3 vs. 4 events in the placebo arm; 11 events across two doses of sotagliflozin in inTandem1 vs. zero events in the placebo group). In 2015, FDA added warnings on DKA to the labels of currently-approved SGLT-2 inhibitors. In a recent NEJM response, Lexicon’s Dr. Paul Strumph and BDC’s Dr. Satish Garg emphasize that DKA would have been more frequent in DEPICT 1 for dapa had investigators counted “possible” events (as it were, only “definite” events were positively adjudicated). While we don’t doubt that there’s elevated risk for DKA with either of these drug candidates, we firmly believe that this risk should be manageable in the real world with careful patient selection, better HCP/patient education, and diligent monitoring. Both drugs deserve approval and we are a bit surprised that at this early-stage there would be a back-and-forth over risk.
• Clearly, there are strategies that minimize DKA risk; read insights from Dr. Strumph and UNC’s Dr. John Buse here. As one example, capping reduction in insulin dose at a certain upper limit may prevent excess DKA. In DEPICT 1, total daily insulin dose was not to be reduced by more than 20%, and this likely contributed to fewer “definite” DKA episodes. That’s not to say these tactics are easily implemented in real-world clinical practice, and we see a major role for the manufacturers to play in spreading this education and possibly supplying a ketone meter alongside every prescription. At a hypothetical Advisory Committee meeting, it would bode well for Lexicon/Sanofi to have a clear and comprehensive protocol prepared for HCPs:
  
  • Who is an ideal candidate for sotagliflozin as add-on to insulin?
  • How should providers advise their patients on exercise, carbohydrate intake, and alcohol?
  • How should insulin be titrated after initiating sotagliflozin 200 mg or 400 mg once-daily?
  • How can HCPs make sure that their patients at highest risk of DKA either do not take the drug (some may try to take off-label) or put in place safety strategies to check ketones every day?
  • We see great value in an online course that patients have to take prior to starting sotagliflozin or another adjunct therapy for type 1 diabetes. We hope to see Lexicon/Sanofi collaborate with AZ and other manufacturers on this front if allowed, since it is so important that this new class for type 1 patients is well understood.
• In light of the FDA’s update to the Warnings and Precautions of all approved SGLT-2 inhibitors to include the risk of DKA, we were curious about patients’ experience with DKA and ketone testing. The risk of DKA is considered to be manageable, but have patients talked to their doctors about DKA and ketone testing? How confident are patients that they could recognize DKA? How important is the label update to patients and what do they plan on doing in response to it? We had a conversation with the team at dQ&A, the diabetes research company, to get an answer to these questions in 2016 and show these results again here. At that time, the team was particularly struck by the lack of preparedness for DKA among patients with type 2 diabetes, even those taking an SGLT-2: most patients had not talked to their doctor about it, and few type 2 patients on SGLT-2s owned a ketone testing kit or were confident that they could recognize the symptoms of DKA. It was clear at that time that there is much work that needs to be done to increase patient education about DKA and ketone testing and to increase access to ketone testing kits, and we are optimistic that some of this work has been done and there is likely more to do – we remain optimistic, as noted, that this is addressable. There’s more detail on dQ&A’s questions and answers about DKA and ketone testing here, or you can reach out directly to richard.wood@d-qa.com.
  
  • Currently, we do not believe patients are well-educated about DKA, particularly DKA associated with SGLT-2s. In particular, although numerous KOLs have said the risk is “manageable,” we emphasize the importance of active management.
  • Notably, recurrent episodes of DKA are associated with increased risk of death in type 1 patients. A small study in Diabetologia published in 2016 showed this, suggesting that risk was highest in young adults and individuals who are socially-disadvantaged or who struggle with mental health issues. These results came from this retrospective cohort study led by Dr. Fraser Gibb (University of Edinburgh, Scotland) examining 628 DKA hospitalization events for 298 patients with diabetes at the Royal Infirmary of Edinburgh. Although the results are far from definitive and cannot necessarily be applied to type 1 patients broadly speaking, this is another reminder of the importance of patient selection and management.
  • Some professional groups like AACE/ACE advised no change in recommendations for SGLT-2 inhibitors following a meeting on DKA risk assembled in late 2015; this analysis related to type 2 patients and we assume the groups will assess risk for type 1s. This is the detailed report from this meeting. We would expect that since they realize that type 1 patients have few alternatives and need more options that they will be supportive of SGLT approval for people with type 1, though we have no definitive news on this at this stage.
  • At the AACE/ACE meeting, Ms. Emily Regier emphasized the positive impact on people with type 1 taking SGLT-2s off-label and the need for further education. She cited dQ&A data on several dozen patients taking SGLT-2s off-label (write richard.wood@d-qa.com for more information) showing increased positive changes in diabetes management (lower A1c, less insulin taken, lower fasting and postprandial blood glucose, presumably weight loss though this was not measured, and higher confidence in blood glucose control). She also pointed to a 2010 study by Mackay and Knight in Journal of Diabetes Nursing (n=120 type 1s) that found that 17% of patients regularly monitor their urine for ketones, 16% regularly monitor their blood, and 58% do neither. When asked if they would monitor their ketones under specific circumstances like nausea, vomiting, or feeling unwell, only 9%-33% said yes – presumably, they either did not have ketone meters or did not know when or how to use them.
• Commercially, there is more than enough room for both sotagliflozin and dapagliflozin to be successful on the market for type 1. In fact, it goes without saying that since many type 1s are already taking SGLT-2 inhibitors off-label, approvals from FDA and EMA will make this practice safer across the board, while also allowing many more patients to benefit from these oral tablets for extra glucose-lowering, weight loss, and time-in-range.

Considerations for FDA: CGM Data and Time-in-Range Benefit

• Lexicon CEO Mr. Lonnel Coats has suggested that FDA may consider time-in-range results when evaluating sotagliflozin. Management hasn’t shared any specifics, but has alluded to productive, positive conversations with FDA on this issue. Historically, FDA’s CDER has not accepted CGM data, and we would very much like to see this re-considered given all the strides made on accuracy and usability; overall, we think this is overdue as the value of the data seems very high clinically and scientifically and practically speaking (we understand there may be issues we’re not aware of on this front, of course). In our view, improved time-in-range with sotagliflozin therapy may be the most meaningful benefit for patients: Pooled CGM data from inTandem1 and inTandem2 showed 1.3 additional hours in range with 200 mg sotagliflozin vs. placebo (p=0.026) and 2.8 additional hours in range with 400 mg sotagliflozin vs. placebo (p<0.001), which is 1-3 extra hours each day that a patient spends feeling well, not trying to correct hyperglycemia or hypoglycemia. This is enormously valuable to patient quality of life, and we also note accumulating data that links glucose variability and severe hypoglycemia to CV death and all-cause death. Mr. Coats confirmed on Lexicon’s 4Q17 earnings call that this pooled CGM data is included in the sotagliflozin NDA – we hope these numbers don’t fall on deaf ears at FDA. Again, we imagine patient advocates at a hypothetical Advisory Committee meeting could speak to the value of increased time-in-range, sending a powerful message to FDA about the need for these oral adjunct therapies in diabetes care.

Thought Leader Perspectives

• Over the past couple years, diabetes thought leaders seem to have reached a consensus on SGLT-2s and SGLT-1/2s for type 1 diabetes – that these drugs should be approved, but that the field is going to have to step up its game in educating patients, HCPs, and hospital staff on DKA management. We spoke to Dr. John Buse and Professor Thomas Danne about the sotagliflozin NDA and next steps for the field, and both experts identified the ideal patient candidate for adjunct therapy as someone who is highly engaged in diabetes self-management. At Prof. Danne’s clinic, type 1s who want to take an SGLT-2 or SGLT-1/2 inhibitor off-label have to first go through a rigorous guided program to show they can regularly check blood ketones; he also shared his view that DKA is a class effect here, and that the risk is not significantly different with sotagliflozin vs. dapagliflozin. Of course, even the most engaged patients won’t know how to properly identify and treat DKA if they aren’t well-educated. Prof. Danne pointed out that most people only receive information on DKA when they’re diagnosed with diabetes, but at that point in time, the patient likely has some residual beta cell function and is thus at low risk for DKA. By the time DKA becomes a more pressing issue, several years into diabetes development, the patient has then forgotten that early (likely cursory) education. As one possible solution to this problem, Dr. Buse suggested that reminders about DKA be incorporated alongside relevant, intermittent events, like getting a flu shot (e.g. if you get the flu and feel bad, remember to check your ketones no matter what your blood sugar is). He proposed wallet cards for patients with type 1 diabetes on an SGLT inhibitor, so that a DKA episode is more quickly and easily identified by HCPs or emergency room staff. He also envisions one-pager patient education materials, YouTube videos in different languages, and medic-alert links.

• This education will look slightly different depending on the target audience, whether it’s patients, PCPs, endos, hospitalists, etc., but Prof. Danne underscored the key message that to treat DKA you need both insulin and glucose. This is why patients on a low-carb diet may not be ideal candidates for adjunct SGLT therapy (although “low-carb” has yet to be defined precisely in this context, and it likely differs from patient to patient – when does fat oxidation begin as a means of generating energy?). He noted that DKA is often misdiagnosed in a hospital setting, and advocated that we need to spread education on best practice DKA management among providers as well as patients.

• Should blood ketone meters be supplied to all type 1s on an SGLT inhibitor? Prof. Danne argued in favor, and he suggested that this will be cost-saving in the long run for payers, considering the extraordinary expense of even a single hospitalization for DKA. “Compare ketone monitors to blood glucose monitors for a moment,” he explained. “The A1c-lowering effect of frequent blood glucose monitoring hasn’t actually been shown in any good clinical study, and nevertheless everyone believes it is the right thing to do.” We loved this sentiment and we hope payers see the value in covering blood ketone meters to support an easy-to-take oral tablet that lowers A1c and body weight while improving time-in-range. Dr. Buse also endorsed the superiority of blood ketone meters, but suggested that urine strips for counting ketones may actually suffice in most circumstances (and be more widely affordable). “Providers and patients need to understand that what works in fact works, and that perfection is the enemy of the good,” he said, adding “blood ketone meters are probably more firepower than you need to solve this issue in the vast majority of cases.” While we want all patients to have access to the best, we do have to think on the population level now that sotagliflozin and other adjuncts could be approved. Dr. Buse pointed out additional nuance in that blood ketone meters are bulkier to carry around, but that testing blood is much less cumbersome than testing urine for many patients who are accustomed to frequent fingersticks; perhaps this treatment decision needs to be individualized. Above all, Dr. Buse remarked that “the biggest issue” is if patients feel unwell, they have to test something, be it blood or urine. This discussion on blood ketone meters vs. urine strips is just one of many that will unfold over the next several months as oral adjunct treatments for type 1 diabetes become a reality.

• Dr. Buse and Prof. Danne expressed enthusiasm for the time-in-range benefit associated with sotagliflozin. “If you tell anyone – here’s a simple pill to take that gives you three hours more time-in-range – that’s very convincing,” Prof. Danne affirmed. It’s hard to predict how FDA will respond to CGM data from the inTandem program or to time-in-range findings, but Prof. Danne is pushing for more consideration of the ratio between time-in-range and hypoglycemia (which he likened to the way we measure blood pressure, looking at two values). That sotagliflozin increases time-in-range without increasing time in hypoglycemia is a major win for patients. As we wrapped up our call, Prof. Danne left us with this memorable comment: “My only concern is that this wonderful drug might somehow be shelved or delayed, and we have to really work hard to make people understand that diabetes is a lot about education. This is simply a new approach. It’d be a lot like not giving insulin because it may cause hypoglycemia.”

Competitive Landscape

• In addition to Sanofi/Lexicon, AZ and Lilly/BI also have programs ongoing for their SGLT inhibitors in type 1 diabetes. AZ was the first to file an SGLT-2 for type 1 in Europe, submitting an MAA for Forxiga (dapagliflozin) in early March. An EMA decision on Forxiga’s type 1 indication is expected in 1Q19, putting it essentially neck-and-neck with sotagliflozin in the EU. During AZ’s 4Q17 update, management also outlined plans to file Farxiga for type 1 with FDA in 2H18.
• Meanwhile, Lilly/BI’s phase 3 EASE trials for Jardiance (empagliflozin) in type 1 diabetes are slated to read out sometime in 2018. According to ClinicalTrials.gov, EASE-2 randomized ~730 type 1s to low-dose empagliflozin, high-dose empagliflozin, or placebo, while EASE-3 enrolled ~977 type 1s and featured three different doses of empagliflozin (low, medium, or high) alongside placebo. The primary endpoint in both studies was change in A1c after 26 weeks, though weight loss, hypoglycemia, blood pressure, and total daily insulin dose were all measured as secondary outcomes. Very notably, EASE-2 used CGM to measure time-in-range, and we’re particularly eager to see these findings.
• J&J reported phase 2 results on Invokana (canagliflozin) in type 1 at ADA 2016, but we haven’t heard any relevant updates since. It’s unclear whether J&J is still actively pursuing a type 1 diabetes indication for the SGLT-2 inhibitor, or whether this has been put on hold; although a boxed warning for lower limb amputations has been added, we still hope very much that type 1 and prediabetes programs will continue, despite the weak financial performance in 2017. We hope to see more momentum from Invokana’s clinical development projects whether or not franchise sales stabilize near-term – over the long term, we do expect a return to growth. In addition to studies in type 1, the company previously announced results on canagliflozin/phentermine co-administration for obesity and shared plans for a CVOT of canagliflozin in prediabetes.

-- by Payal Marathe and Kelly Close