Memorandum

FDA releases briefing documents for Advisory Committee for Sanofi’s lixisenatide and LixiLan (lixisenatide/insulin glargine) – May 23, 2016

Executive Highlights

  • The FDA has posted briefing documents for the May 25th Advisory Committee meeting for Sanofi’s lixisenatide and LixiLan (lixisenatide/insulin glargine).
  • Sanofi’s briefing document reveals the full phase 3 data for LixiLan for the first time, demonstrating greater A1c reductions vs. both components that were driven by robust postprandial glucose reductions. Results also showed a neutral to positive effect on body weight, a comparable hypoglycemia risk vs. Lantus (rates were low in both groups), and fewer GI side effects vs. lixisenatide.
  • The FDA briefing document focuses on LixiLan and largely echoes the same concerns around GLP-1 agonist/basal insulin combinations raised in the briefing document for Novo Nordisk’s Xultophy (insulin degludec). The document does raise two LixiLan-specific concerns related to the two different proposed pen formulations, with different lixisenatide to insulin glargine ratios, and the design of the pen.
  • The FDA’s draft questions ask the panel to discuss any efficacy/safety issues related to lixisenatide, the benefits of treatment with LixiLan in patients not on basal insulin or a GLP-1 agonist and in patients already treated with one of the components, and clinical concerns related to dosing flexibility.

The FDA has posted briefing documents for the May 25th Advisory Committee meeting for Sanofi’s Lyxumia (lixisenatide) and LixiLan (lixisenatide/insulin glargine). As a note, the briefing documents refer to “insulin glargine and lixisenatide fixed ratio combination” and “iGlarLixi” rather than “LixiLan.”

The FDA’s briefing document focuses its concerns on the LixiLan combination product rather than the standalone Lyxumia. The document raises many of the same concerns as the FDA’s briefing document for Novo Nordisk’s Xultophy (insulin degludec/liraglutide) and takes a slightly more negative view of GLP-1 agonist/basal insulin combination products than we had initially expected. In addition, the briefing documents for Lyxumia and LixiLan raise additional concerns specific to the proposed commercial pen formulations of LixiLan. Sanofi proposed that LixiLan be available in two different pens, one for patients with lower insulin requirements and one for those with higher insulin requirements. However, the ratio of lixisenatide to insulin glargine is lower in the higher dose pen than the lower dose pen, leading the FDA to cite possible confusion on this front. Furthermore, the FDA’s briefing documents take issue with the controls in place with the delivery pen to prevent under-dosing of the combination. Overall, we believe that concerns raised by the FDA are manageable and do not preclude approval of the combination.

Sanofi’s briefing document portrays LixiLan as an attractive option that can help meet the unmet need for greater intensification of type 2 diabetes treatment. It also reveals the full data from the phase 3 LixiLan-O and LixiLan-L trials for the first time, demonstrating greater A1c reductions vs. both components that were driven by robust postprandial glucose reductions. Results also showed a neutral to positive effect on body weight, a comparable hypoglycemia risk vs. Lantus (rates were low in both groups), and fewer GI side effects vs. lixisenatide. The document also reviews data from the GetGoal phase 3 trial program and the ELIXA CVOT for lixisenatide – we expect the discussion there to be very straightforward.

The FDA’s draft discussion questions ask the panel to discuss any efficacy/safety issues related to lixisenatide, the benefits of treatment with LixiLan in patients not on basal insulin or a GLP-1 agonist and in patients already treated with one of the components, and clinical concerns related to dosing flexibility. As with Xultophy, we expect that the panel will ultimately vote in favor of approval, but the discussion may be a bit more contentious than we had anticipated. Read on below for more details and thoughts on the agenda and the list of panelists, and stay tuned this week for our day-of coverage of both of this week’s meetings.

FDA Briefing Document

  • The FDA’s briefing document largely focuses on LixiLan rather than Lyxumia and raises several concerns specific to the combination product. The FDA is most concerned with potential limitations of dosing flexibility associated with the combination product. To minimize the impact of the combination on dosing flexibility, the agency prefers that the new combination product offer all possible combinations that would be available by combining the individual products. The agency also raises the question of whether the increase in hypoglycemia with the combination product vs. a GLP-1 agonist alone or the increase in nausea with the combination vs. insulin alone were justified by a meaningful reduction in A1c. That said, phase 3 trials (see below) have shown that LixiLan is associated with a lower risk of nausea than lixisenatide alone, though hypoglycemia rates were comparable to those with Lantus (insulin glargine) – we suspect this might be different in the real world. Furthermore, LixiLan has been shown to mitigate the weight gain associated with standalone insulin glargine therapy. The briefing document also questions the clinical basis justifying the initiation of a dual combination therapy when a single product may have been adequate to achieve glycemic control for a particular patient. Of course, this concern only applies to patients initiating LixiLan as a first injectable therapy and is less relevant for those who turn to LixiLan as an intensification option after standalone GLP-1 agonist or basal insulin therapy, which we believe will be the main initial target demographic for the combination. It is possible that this concern could lead to a narrower indication for LixiLan, though we see this as unlikely.
  • Most notably, the briefing document raises concerns related to the specific proposed doses of lixisenatide and the design of the LixiLan pen.
    • The FDA took issue with the reduction of the lixisenatide/insulin glargine ratio in the “high dose” formulation of LixiLan compared to the “low dose” formulation. The low dose version offers doses ranging from 10 U insulin glargine/5 μg lixisenatide to 40 U insulin glargine/20 μg lixisenatide. The high dose version offers doses ranging from 30 U insulin glargine/10 μg lixisenatide to 60 U insulin glargine/20 μg lixisenatide. Sanofi’s briefing document (see below) explains that this was done to ensure that the lixisenatide dose remains within the recommended range, as doses above 20 μg increased side effects without adding additional efficacy in clinical trials. Because of the different ratio, a patient switching from the low dose pen to the high dose pen would be reducing their dose of lixisenatide if they began titration at the same dose of insulin, For example, a patient receiving 40 units of insulin via the low dose pen would receive 20 micrograms of lixisenatide while a patient receiving 40 units of insulin via the high dose pen would receive only 13 micrograms of lixisenatide (a 35% reduction). The briefing document suggests that the use of two pens and the recommended increments for the two pens were complex and not intuitive. For comparison, Novo Nordisk’s GLP-1 agonist/basal insulin fixed-ratio combination Xultophy (insulin degludec/liraglutide) is only offered in a single pen with a fixed-ratio of insulin to GLP-1 agonist regardless of the insulin dose. Of course, the simplicity of that device also prompted concerns from the FDA about the wide range of possible liraglutide doses and the lack of flexibility to titrate insulin.
    • The briefing document also raised concerns about accidental under-dosing due to the pen design. The pen relies on a dial color scheme to warn users of under-dosing rather than a physical blocking mechanism that prevents under-dosing. The FDA expressed concern that patients and providers may accidentally use the high dose version of the pen to deliver an insulin dose that would be more appropriately delivered with the low dose pen. Since the lixisenatide to insulin glargine ratio is lower in the high dose pen, inappropriate use of the high dose pen to deliver low doses of insulin expose patients to suboptimal and potentially ineffective doses of lixisenatide.
    • We believe the possible confusion from both of these points can be mitigated through adequate patient and provider education. While the two separate pens may initially cause some confusion, we imagine the ability to offer a less aggressive dose of GLP-1 agonist as patients require higher insulin doses could allow for greater individualization by allowing more combinations of lixisenatide and insulin glargine doses. Sanofi developed a robust patient education and provider assistance program for its PCSK9 inhibitor Praluent (alirocumab) and we expect that the company will invest heavily in patient and provider education for LixiLan as well, especially given that its strategic plan through 2020 hinges on extensive promotion of its insulin glargine franchise.
  • The FDA’s briefing document also describes several specific concerns related to LixiLan that echo similar concerns raised in the FDA’s briefing document for Xultophy.
    • It combines a fixed-dose product with a titratable product dosed on a continuous scale. All other fixed-dose or fixed-ratio combinations that have been approved to date combined either two individual fixed-dose products (such as SGLT-2 inhibitor/metformin, DPP-4 inhibitor/metformin, or SGLT-2 inhibitor/DPP-4 inhibitor combinations) or two individual titratable products (such as pre-mixed insulin).
    • The proposed dosage for the lixisenatide component of the combination is different from the established effective dosage of standalone lixisenatide. The proposed dose range includes low doses that have been demonstrated to be ineffective on their own, meaning patients could be exposed to a component that provides no benefit and causes adverse events.
    • Insulin dosing was artificially constrained in the clinical trials for LixiLan, limiting the clinical applicability of the results. In trials investigating LixiLan against an insulin comparator, the starting dose, dose adjustments, and maximum allowable dose of the insulin were all fixed, whereas providers have much more dosing flexibility in real-world practice.
    • Inadequate insulin dosing may have biased efficacy results in favor of LixiLan in clinical trials. The FDA notes that the majority of participants in the clinical trial program never reached their fasting glucose targets, suggesting that insulin dosing may have been inadequate. Therefore, it believes that Sanofi’s claims of clinical superiority for LixiLan vs. insulin are difficult to interpret. 
    • Patients already on a GLP-1 agonist or basal insulin would need to reduce the dose of their current treatment when switching to the combination. This could translate to poorer glycemic control compared to adding the second component independently.
    • Patients not on either component would be exposed to the adverse effects of two products rather than one. Some of those patients might have been able to achieve good glycemic control with a single agent, running a lower risk of adverse events. 
    • The complexity of the proposed combination (such as the fact that the two components have different dosing measurements) may lead to clinical errors. The high number of clinical errors today with insulin alone is a sobering concern, of course, although it was not discussed in the briefing materials.
  • The outcome of the Advisory Committee meeting for Xultophy, which will take place the day before the Lyxumia and LixiLan meeting, will likely provide an indication of the impact of these concerns on the committee’s vote. While the FDA’s briefing documents for both meetings were more critical than we had initially expected, we believe the concerns are manageable and won’t ultimately preclude approval of either drug. It does appear that the FDA might have more concerns about LixiLan than it does for Xultophy, largely related to the dose offerings that the agency perceives as being more complicated. Xultophy also has the advantage of real-world experience in the EU, though we are not sure to what extent this can or will be part of the discussion around US approval. The patient and provider feedback for Xultophy has been positive thus far, though its penetration remains fairly limited – see our coverage of the Xultophy briefing documents for more on this front. We imagine the positive LEADER results demonstrating a cardioprotective benefit for the liraglutide component and the SWITCH 1 and SWITCH 2 results demonstrating reduced hypoglycemia for the insulin degludec component only strengthen the case for Xultophy’s approval. That said, we expect the Advisory Committee will look favorably upon lixisenatide and LixiLan’s demonstrated effects on postprandial glucose and believe that LixiLan could meet a differentiated need from Xultophy. Overall, we would be very surprised if the committee voted for approval of one of the combination products but not the other.
  • We expect the discussion surrounding Lyxumia will be fairly straightforward and the committee will vote in favor of approval. The FDA did not raise any concerns specific to lixisenatide in its briefing document. Lyxumia has already been available in Europe for around two years and will be the sixth GLP-1 agonist to reach the US market, so the efficacy and safety of the class as a whole and the specific product are already fairly well established.

Sanofi Briefing Document

  • Sanofi describes lixisenatide as offering clinically relevant A1c reductions, robust postprandial glucose reductions, beneficial effects on weight, and low hypoglycemia risk. With an eye on the combination, the document also emphasizes that lixisenatide has been shown to be an effective add-on to basal insulin. The document reviews the results from the GetGoal phase 3 trial program and the ELIXA CVOT, all of which support a favorable risk/benefit profile for the product. We expect the discussion for lixisenatide to be brief and straightforward given that the FDA’s concerns appear to be focused almost entirely on LixiLan.
  • The document portrays LixiLan as an attractive option that can help meet the unmet need for greater intensification of type 2 diabetes treatment. Making arguments similar to those Novo Nordisk used in its briefing document for Xultophy, Sanofi emphasizes the importance of sustained glycemic control for preventing complications and the challenges (weight gain, hypoglycemia, therapeutic complexity) of achieving that goal with current therapies. The document explains that phase 3 data demonstrated a significantly larger effect on A1c and postprandial glucose with LixiLan vs. Lantus, along with reduction or mitigation of weight gain, no additional hypoglycemia risk vs. insulin alone, and fewer GI side effects than a GLP-1 agonist alone. It also characterizes the simple dosing scheme as an advantage, noting that therapeutic simplicity is an important consideration for patients on multiple medications.
  • The document reveals the full phase 3 results for LixiLan for the first time. Sanofi announced topline results from both the LixiLan-O and LixiLan-L trials in the second half of 2015 and plans to present the full datasets at ADA in June. Overall, the results tell the same impressive story as the data from the DUAL program for Xultophy: greater A1c reductions with the combination compared to either of its components along with mitigation of side effects. We are somewhat disappointed that LixiLan was not able to demonstrate a clear hypoglycemia advantage over Lantus – we expect there might be more of a difference in the real world given how little hypoglycemia there was in the Lantus groups in the trials.
    • LixiLan-O (n=1,170) demonstrated significantly greater A1c reductions with LixiLan (1.6%) vs. Lantus (1.3%) and lixisenatide (0.8%) after 30 weeks in patients poorly controlled on oral agents. Baseline A1c was 8.1% in all groups. In addition, 74% of participants in the LixiLan group achieved an A1c <7% compared to 60% with Lantus and 33% with lixisenatide. LixiLan also outperformed Lantus in terms of the percentage of patients achieving composite endpoints, though the low absolute numbers in both groups were a good illustration of how much unmet need remains. 43% of patients on LixiLan achieved an A1c <7% without weight gain vs. 25% of patients on Lantus, and 32% of the LixiLan group achieved an A1c <7% without weight gain or symptomatic hypoglycemia vs. 19% of the Lantus group. Fasting plasma glucose reductions were comparable with LixiLan (62 mg/dl) and Lantus (59 mg/dl) and less impressive (as expected) with lixisenatide (27 mg/dl). Two-hour postprandial glucose reductions with LixiLan (42 mg/dl) fell between those achieved with lixisenatide (58 mg/dl) and Lantus (3 mg/dl). LixiLan was essentially weight neutral (0.3 kg loss), whereas Lantus led to a 1.1 kg weight gain and lixisenatide to a 2.3 kg weight loss. The rate of documented symptomatic hypoglycemia was low and comparable between the LixiLan (26% incidence, 1.4 events/patient-year) and Lantus (24% incidence, 1.2 events/patient-year) groups.
    • LixiLan-L (n=736) demonstrated significantly greater A1c reductions with LixiLan (1.2%) vs. Lantus (0.6%) in patients poorly controlled on basal insulin. Baseline A1c was 8.1% in both groups. 55% of participants in the LixiLan group achieved an A1c <7% vs. 30% in the Lantus group. LixiLan also beat Lantus in the percentage of patients achieving an A1c <7% without weight gain (34% vs. 13%) and the percentage achieving an A1c <7% without weight gain or hypoglycemia (20% vs. 9%). Fasting plasma glucose reductions were comparable between the groups (6 mg/dl with LixiLan vs. 8 mg/dl with Lantus), while postprandial glucose reductions were significantly greater with LixiLan (70 mg/dl vs. 8 mg/dl). LixiLan led to an average 0.7 kg weight loss while Lantus led to an average 0.7 kg weight gain. The incidence of documented symptomatic hypoglycemia was 40% with LixiLan and 43% with Lantus; the rate was 3.0 events/patient-year with LixiLan and 4.2 events/patient-year with Lantus.
    • Pooled safety data from the two trials demonstrated fewer GI side effects with LixiLan vs. lixisenatide and no new concerning signals. The total incidence of adverse events was 55% with LixiLan vs. 50% with Lantus and 67% with lixisenatide, explained mainly by higher rates of GI side effects in the lixisenatide group. The incidence of nausea was 10% with LixiLan vs. 24% with lixisenatide, and the incidence of vomiting was 3% with LixiLan and 6% with lixisenatide.
  • Sanofi is proposing two different devices for LixiLan with different fixed ratios of the two components. Both will use the same SoloStar platform as Lantus. The first pen (which will be a yellow/peach color) is intended for both insulin-naïve patients and those already on basal insulin and would offer doses ranging from 10 U insulin glargine/5 μg lixisenatide to 40 U insulin glargine/20 μg lixisenatide. The second pen (which will be a green/olive color) is intended only for patients switching from basal insulin and would offer doses ranging from 30 U insulin glargine/10 μg lixisenatide to 60 U insulin glargine/20 μg lixisenatide. The goal is to give providers flexibility to titrate insulin glargine across a wide range of doses while keeping the lixisenatide dose within the clinically effective range demonstrated in clinical trials. This could help Sanofi avoid some of the concerns the FDA expressed about Xultophy: an overly simplistic titration scheme, the likely need to reduce insulin doses when switching, and the potential for ineffective doses of liraglutide. However, this choice has also clearly prompted other concerns from the FDA, as several of the issues raised in the Agency’s briefing documents for LixiLan relate to these devices.

FDA Discussion Questions

1) Discussion: Discuss any issues related to the efficacy or safety of lixisenatide for the treatment of patients with type 2 diabetes mellitus. Please comment on whether any of these issues preclude approval of lixisenatide.

2) Discussion: Discuss the benefit(s) of starting the fixed-combination drug product containing lixisenatide and insulin glargine in patients with type 2 diabetes mellitus not treated with either a basal insulin or a GLP-1 agonist (i.e., starting two new drugs at once). In your discussion, identify the patient population in whom this use would be particularly useful, and address why you would select the fixed combination over use of an available GLP-1 agonist or basal insulin in these patients. Explain your rationale using data from the briefing materials, presentations, or your own clinical experience.

3) Discussion: Discuss the benefit(s) of using the fixed-combination drug product containing lixisenatide and insulin glargine in patients with type 2 diabetes previously treated with either a basal insulin or a GLP-1 agonist (i.e., adding a single new drug to an existing regimen). In your answer, identify the patient population in whom use of the fixed-combination drug product in this manner would be particularly useful. Explain your rationale using data from the briefing materials, presentations, or your own clinical experience.

4) Discussion: Discuss clinical concerns related to the use of the fixed-combination product which combines a drug that, when used alone, has a wide effective dose range and is titrated to effect on a continuous scale (i.e., insulin glargine) with a drug that, when used alone, has one or two recommended effective dose(s) (i.e., lixisenatide).

Specifically discuss:

A. Issues related to loss of dosing flexibility including but not limited to: Use of potentially ineffective doses of one agent in populations with low insulin requirements, inability to dose the two drugs independently with the device presentation proposed, inability to increase the insulin dose beyond 60 units.

B. Issues related specifically to product presentation/devices including but not limited to: use errors that may occur in the care setting related to a lack of clarity on the amount of each product delivered with each given dose, insufficient understanding that, unlike insulin products, the maximum dose for the combination is capped, inadequate understanding of the role of the two devices.

5) Vote: Based on data in the briefing materials and presentations at today’s meeting do you recommend approval of the lixisenatide/glargine fixed-combination drug delivered using the proposed pen devices for the treatment of adult patients with type 2 diabetes mellitus?

A. If you voted yes, explain your rationale and discuss whether use of the combination should be approved for patients not treated with a basal insulin or a GLP-1, for patients who are inadequately controlled on either a basal insulin or a GLP-1 analog or for both populations. Recommend additional post-approval studies if you think these are needed.

B. If you voted no, explain your rationale and recommend additional pre-approval studies if you think these are needed.

Panelists

  • Voting Members
    • Robert Smith, MD (Chairperson, Brown University, Providence, RI) – Also on the Advisory Committee for Xultophy
    • Diana Hallare, MPH (Consumer Representative, Visalia, CA) – Also on the Advisory Committee for Xultophy
    • Daniel Budnitz, MD, MPH (CDC, Atlanta, GA) – Also on the Advisory Committee for Xultophy
    • Brendan Everett, MD, MPH (Brigham and Women’s Hospital, Boston, MA) – Also on the Advisory Committee for Xultophy
    • James Neaton, PhD (University of Minnesota, Minneapolis, MN) – Also on the Advisory Committee for Xultophy
    • Charles Stanley, MD (University of Pennsylvania, Philadelphia, PA) – Also on the Advisory Committee for Xultophy
    • Peter Wilson, MD (Emory University, Atlanta, GA) – Also on the Advisory Committee for Xultophy
    • Susan Yanovski, MD (NIDDK, Bethesda, MD) – Also on the Advisory Committee for Xultophy
    • Barbara Berney (Patient Representative, Rockford, IL) – Also on the Advisory Committee for Xultophy
    • Timothy Lesar, PharmD (Albany Medical Center, NY) – Also on the Advisory Committee for Xultophy
    • Martha Nason, PhD (NIH, Bethesda, MD) – Also on the Advisory Committee for Xultophy
    • Ellen Seely, MD (Harvard Medical School, Boston, MA)
    • Kenneth Burman, MD (MedStar Washington Hospital Center, DC) – Also on the Advisory Committee for Xultophy
    • Steven Meisel, PharmD (Fairview Health Services, Minneapolis, Minnesota) – Also on the Advisory Committee for Xultophy
    • Michael Reed, PharmD (University Hospitals Case Medical Center, Cleveland, OH) – Also on the Advisory Committee for Xultophy
  • Non-Voting Members
    • LaToya Bonner, PharmD (Designated Federal Officer, FDA, Silver Spring, MD) – Also on the Advisory Committee for Xultophy
    • Reshma Kewalramani, MD (Industry Representative, Amgen, Thousand Oaks, CA)
    • Mary Parks, MD (FDA Participant, FDA, Silver Spring, MD)
    • Lubna Merchant, PharmD (FDA Participant, FDA, Silver Spring, MD)
    • Suchitra Balakrishnan, MD, PhD (FDA Participant, FDA, Silver Spring, MD)
    • Jean-Marc Guettier, MDCM (FDA Participant, FDA, Silver Spring, MD)
    • William Chong, MD (FDA Participant, FDA, Silver Spring, MD)

-- by Helen Gao, Emily Regier, and Kelly Close