Merck Investor Briefing – Glucose-responsive insulin moving into phase 1; once-weekly DPP-4 to be submitted in Japan by FY13 – May 7, 2014

Executive Highlights

  • Merck thinks it has a glucose-responsive insulin (L-490), and is planning to move it into phase 1 testing (undisclosed timeline).
  • Merck plans to files its once weekly DPP-4 inhibitor, omarigliptin, in Japan first, and intends to do so by the end of the year.
  • Januvia has a 65% share of the global DPP-4 inhibitor market, and 74% of the US market.

Merck hosted a four-hour investor briefing in its Boston research laboratories yesterday. Merck’s presentations delved into its four areas of focus: (i) diabetes; (ii) vaccines; (iii) acute care hospital; and (iv) oncology. The big news of the diabetes section was Merck’s optimism about its glucose-responsive insulin L-490, and its intentions to move the agent into phase 1 testing (timeline for doing so not specified). They discussed it for five minutes! Below we have detailed this encouraging news, as well as four other key highlights from the call.

Top Five Highlights

1. The highlight of the meeting was hearing that Merck thinks it has a glucose-responsive insulin (L-490), and is planning to move it into phase 1 development. Management explained that L-490 is an insulin-oligosaccharide conjugate (an insulin molecule attached to a small chain of carbohydrates). Basic science alert – the following sentences detail more basic science elements of the compound. Briefly, as we understand it, when a person has a normal or low glucose level, L-490’s oligosaccharide component binds to the lectin receptor (a carbohydrate-binding protein), preventing L-490’s insulin portion from binding with the insulin receptor to lower glucose levels. When a person’s glucose level is high, the glucose in his or her body blocks the lectin receptor so that L-490 cannot bind to it and its insulin side instead binds to the insulin receptor, lowering glucose levels. In a steady-state study in dogs (i.e., the amount of glucose delivered to the dog is altered such that the dog’s blood glucose level remains in zone following insulin injection), the glucose infusion rate (GIR) required to keep the dogs in range (90-120 mg/dl) following the injection of L-490 was lower than that for Lilly’s Humulin, so long as the baseline glucose level was in range. This indicates that when the dogs’ were euglycemic L-490 did not work to lower their blood glucose level as much as Humulin. For example, when the baseline glucose level was 100 mg/dl the GIR required to maintain euglycemia was ~2.5 mg/kgŸmin for L-490 and ~6 mg/kgŸmin for Humulin (p-value not disclosed). However, when the dogs were hyperglycemic (e.g., 200 mg/dl), the GIR was slightly higher for L-490 than for Humulin (~20 mg/kgŸmin vs. ~23 mg/kgŸmin), suggesting that L-490 had an edge on Humulin in terms of glucose lowering efficiency. We think a key question will be the specificity and consistency with which L-490 accounts for people’s glucose levels – given that current insulins do not respond to one’s glucose level at all, we suspect L-490 cannot do much worse, unless is very unpredictable in its action. Given the dramatic impact this technology could have on patients’ lives if successful, we are very encouraged to see it moving into the clinic and have our fingers tightly crossed that it lives up to its potential.

  • Based on Merck’s description of L-490’s mechanism of action, it does not sound like the glucose-dependent insulin Merck acquired when it purchased SmartCells in December 2010 for $500 million in cash and milestones (most was based on milestones we believe). As a reminder, SmartCells was a private company developing a glucose-dependent insulin SmartInsulin. In the SmartInsulin formulation, insulin was bound to a biodegradable polymer that includes sugar groups. The insulin-polymer conjugate was connected to a multivalent glucose-binding molecule that can attach to the polymer’s sugar groups, leading to an aggregate that prevents insulin from entering solution when blood glucose is low. During hyperglycemia, however, glucose would bind to the glucose-binding molecule, displacing the insulin-polymer conjugate so that active insulin enters the bloodstream.

2. Merck plans to first file its once weekly DPP-4 inhibitor omarigliptin in Japan, and to do so by the end of the year. Merck plans to also register omarigliptin in the US, EU, other, so far, other undisclosed countries. Ten phase 3 studies of omarigliptin (total n=nearly 8,000) are currently proceeding, including a CVOT and trials to support its use in patients with renal insufficiency.  The CVOT has a primary estimated completion date of November 2017 (Clinical Identifier: NCT01703208).

  • Merck briefly presented 24-week data comparing omarigliptin 25 mg, Januvia 50 mg, and placebo in Japanese people. Omarigliptin and Januvia had near identical impacts on A1c, and were associated with an A1c decline of ~0.7% from a baseline of 7.9% and 8.0%, respectively. In contrast, people on placebo saw a ~0.15% rise in A1c from a baseline of 8.1%. Merck did not disclose how many people were in this trial; however, indicates that a 52-week study (with a 24-week, primary outcome of change in A1c) of omarigliptin, Januvia, and placebo in Japanese people enrolled 410 people. The primary completion date is listed as April 2014 ( Identifier: NCT01703221).

3. Seven phase 3 studies (total n=8,300 people) are ongoing for Merck/Pfizer’s SGLT-2 inhibitor ertugliflozin, and we counted another two studies on that are being planned. The seven currently ongoing trials are all recruiting. Notably, one of the planned studies (Identifier: NCT02099110) is looking at the co-administration of ertugliflozin and Januvia (sitagliptin) – a step in the development of Merck/Pfizer’s fixed-dose combination (FDC) of sitagliptin and ertugliflozin (in addition to a FDC of ertugliflozin and metformin). The study profile notes the use of a biological insulin glargine rescue medication – we are curious if this is the phase 3 insulin glargine Merck and Samsung Bioepis are developing. This study has a primary estimated completion date of May 2016. Merck/Pfizer are investigating the safety and efficacy of ertugliflozin in people with inadequate glycemic control on metformin (NCT02033889), on metformin and Januvia (NCT02036515), and on diet and exercise (NCT01958671). Additionally, Merck/Pfizer are studying the safety and efficacy of ertugliflozin in people with (i) hepatic impairment, with and without diabetes (NCT02115347); (ii) type 2 diabetes and stage 3 kidney disease (NCT01986855); and (iii) renal impairment and type 2 diabetes (NCT01948986). Additionally, Merck/Pfizer are conducting a CVOT of ertugliflozin, which is expected to end in 2020 (NCT01986881). Finally, a trial is being enrolled to compare ertugliflozin against glimepiride (NCT01999218) – we look forward to results from that one as presumably it will crystallize well the benefits of ertugliflozin on weight and hypoglycemia. Excluding the FDC and CVOT trials, the last primary estimated completion date for a phase 3 trial of ertugliflozin appears to be January 2016 (for the currently not enrolling study of ertugliflozin in people with stage 3 kidney disease).

4. Merck highlighted Januvia as a “triumph,” noting that it has a 65% share of the global DPP-4 inhibitor market and a 74% of the US market. We agree – the drug achieved just under $6 billion in sales in 2013, up from $42 million in 2006.  Januvia’s share in the US declined from 81% in January of 2012 to reach this level. Based on Merck’s graphic on the topic, it appears that since 2012, BI/Lilly’s Tradjenta (linagliptin) is the franchise that has made the most inroads on Januvia’s share going from <1% in January 2012 to around 10% in January 2014. That is the most differentiated DPP-4 inhibitor in our view, given that there is only one dose and it is not limited in renally impaired patients (meaning no testing is required, etc). Notably, AZ’s Onglyza (saxagliptin) appears to have maintained the second largest US share at roughly flat 15% over the time period; we are curious to see how it does when it is unencumbered by partners, although of course Pfizer may be in its future (especially thinking about the agreement with Pfizer on ertugliflozin, we assume an AZ partnership will be messy given all the joint CVOT investments, etc.). Similar to Merck’s comments on its 1Q14 call, the company underscored that Januvia’s position in the US market appeared to have stabilized in 1Q14. Januvia’s new-to-brand share (i.e., the proportion of people taking a new drug who are taking Januvia) improved, according to management, and its new-to-brand volume (i.e., the number of people taking Januvia for the first time) increased in March – the first time Merck has seen this measure rise in “a very long time” acknowledged management.

  • As a reminder, Merck characterized Januvia’s 3% YOY growth in 1Q14 (5% constant currencies) to $1.3 billion as “solid,” suggesting single-digit growth for the franchise could be a new normal. Januvia franchise sales increased in the US for the first time since 2Q13 in 1Q14, experiencing over 4% YOY growth to reach $687 million in 1Q14 (this was an easy comparison as 1Q13 sales fell 1%, the first decline ever for the franchise). However, internationally the franchise performed more poorly in 1Q14, posting its second weakest ex-US YOY growth ever: ex-US sales rose just 2.1% in 1Q14 to reach $647 million.

5. There is a lot of interest in what Merck is doing in insulin and in diabetes more broadly. In the meeting, Merck reaffirmed that it is developing a biosimilar insulin glargine that is already in phase 3 (MK-1293), and that it is working on GPR40 agonists. However, Merck did not detail its efforts in these areas in any detail, regrettably. For more background on this class, please see our 2011 ADA coverage of next-generation drugs.

Table 1: Merck Diabetes Drug Pipeline

Drug Name




Other Remarks

Omarigliptin (MK-3102)

DPP-4 inhibitor

Type 2 diabetes

Phase 3

Once-weekly formulation; Japan submission by FY14


SGLT-2 inhibitor

Type 2 diabetes

Phase 3

Partnered with Pfizer


Biosimilar insulin glargine

Type 1 and type 2 diabetes

Phase 3

Partnered with Samsung Bioepis


Glucose-responsive insulin

Type 1 and type 2 diabetes


To enter phase 1 development (undisclosed timeline)


GPR40 agonists

Type 2 diabetes



--by Hannah Deming and Kelly Close