FDA approves Tresiba label update to include DEVOTE results: A partial victory – March 26, 2018

Executive Highlights

  • Novo Nordisk announced today that FDA has approved inclusion of DEVOTE data on the Tresiba (insulin degludec) product label. We consider this a meaningful victory for several reasons, if a partial one. The CV results from DEVOTE underscore compelling CV safety with Tresiba. Insulin degludec received an FDA Complete Response Letter the first time it was submitted due to CV concerns, and this latest label update puts those concerns to bed. This also marks the first time that FDA has granted a comparative claim on any endpoint, let alone hypoglycemia, which is unchartered territory in the regulatory world.
  • On the other hand, this is not the update we expected, because the SWITCH studies were left out of the label. We discuss the implications of this below – Novo Nordisk can’t promote a hypo benefit for patients with type 1 diabetes, and payers may only be compelled to enhance reimbursement for individuals who closely resemble DEVOTE study participants (type 2s at high risk for CV events). We hope this isn’t the case, and we believe patient advocates could play a role here, if they could get organized. Notably, EMA has approved inclusion of both SWITCH and DEVOTE on Tresiba’s product label in Europe.
  • We spoke at some length today with Dr. Todd Hobbs, Novo Nordisk CMO in North America, about this label update: He emphasized the value of reassuring CV safety data on the Tresiba label (though we think this has been the case for some time), and also characterized DEVOTE as the most robust dataset demonstrating hypoglycemia benefit with insulin degludec vs. insulin glargine (we certainly agree, though note that it is not Toujeo). Notably, Tresiba enrolled 7,637 patients vs. only 1,222 across both SWITCH 1 (n=501 patients with type 1 diabetes) and SWITCH 2 (n=721 patients with type 2).

This morning, Novo Nordisk announced that DEVOTE data has been added to the US product label for Tresiba (insulin degludec). These CVOT results (n=7,637) show compelling CV safety as well as significant risk reduction for severe hypoglycemia with Tresiba vs. Lantus (Sanofi’s insulin glargine). This label revision marks a regulatory milestone, as it’s the first time FDA has granted a comparative claim on any endpoint, let alone hypoglycemia. This news has exciting implications for how Tresiba is promoted and how it’s reimbursed and this is very happy news for the sales force – we expect to continue to see high growth in next-generation insulins in 2018 as a result. However, we note that this label update is not quite what we expected – data from SWITCH 1 and 2 was not included. In fact, these separate supplemental applications have been withdrawn.

As background, Novo Nordisk first submitted a supplemental application to FDA requesting that SWITCH results be added to the Tresiba label in September 2016. An FDA decision on SWITCH was expected by September 2017, but before that date arrived, DEVOTE read out at ADA 2017. The CVOT offered even more robust evidence for a hypoglycemia benefit, enrolling 7,000+ patients vs. 501 type 1s in SWITCH 1 and 721 type 2s in SWITCH 2 (for a total n=1,222). During Novo Nordisk’s 2Q17 update, management announced that FDA would now be evaluating SWITCH 1 and 2 in conjunction with DEVOTE, pushing back the timeline for regulatory feedback until 1Q18. Which brings us to today…

The reasons underlying this decision – include DEVOTE but not SWITCH – are ambiguous. We imagine there were negotiations between Novo Nordisk and FDA, and we suspect the agency is still reluctant to approach hypoglycemia as an efficacy endpoint instead of a safety one – this is speculation on our part. In a call with our team, North America CMO Dr. Todd Hobbs implied that FDA was perhaps more inclined to sign off on including DEVOTE in the label because this was an FDA-mandated CVOT. Notably, EMA added SWITCH findings to Tresiba’s European product label in 1Q17, and then added DEVOTE in 3Q17.

Dr. Hobbs added, “DEVOTE offered the most robust data to describe for clinicians as well as patients what the hypoglycemia differences are” between Tresiba and standard of care Lantus.

There’s no denying that DEVOTE was a landmark trial, and that it’s important for HCPs/patients to be aware of treatment options that come with significantly less hypoglycemia risk. Thought leaders, including Dr. Irl Hirsch and Dr. Elizabeth Seaquist, have praised Tresiba at recent conferences, recommending it for any patient who is struggling with hypoglycemia and who is taking long-acting insulin.

Moreover, we think DEVOTE provides very convincing evidence for payers that a next-generation basal is worth its higher list price because of the cost-savings that come from avoiding severe lows – namely, fewer hospitalizations. Payers, of course, will make their own decisions on this front, and we hope they see the value of this investment.

It is unfortunate that the SWITCH data didn't make it onto the label, because we believe it had valuable findings to contribute as well. As we understand it, it was a steeper challenge to get FDA to approve the inclusion of hypoglycemia data as it was defined in SWITCH vs. DEVOTE: Both SWITCH studies used a primary endpoint of "severe or blood glucose-confirmed symptomatic hypoglycemia," while severe hypoglycemia events were adjudicated in DEVOTE and required external assistance. We return to our notion of partial victory, because this indicates to us that FDA still hasn't acknowledged the full impact of hypoglycemia, and certainly hasn't acknowledged the clinical signifiance of a milder category of hypoglycemia.

Dr. Hobbs acknowledged that without SWITCH 1 reflected in product information, Novo Nordisk won’t be able to promote Tresiba’s hypoglycemia benefit for type 1 diabetes (which seems surprising, given so many advantages we’ve heard from people with type 1, but it is technically true). The DEVOTE population was comprised entirely of type 2s, and type 2s facing high CV risk at that: Participants had to be over 50 with established CV disease or CKD, or over 60 with multiple CV risk factors. As such, Tresiba’s label update won’t necessarily translate to improved reimbursement for all, as payers might only recognize the cost-savings for those who are “most sick” (older, longer duration of diabetes, history of CV events or kidney complications, etc.).

Tresiba did secure tier 2 positioning on the Medicare Part D formulary in 2018 (on par with Lilly/BI’s biosimilar Basaglar), and since this patient population is more likely to resemble DEVOTE, Novo Nordisk may be able to leverage the label update into even lower out of pocket costs for Medicare beneficiaries and/or lower rebates for the company. As an aside, we think Tresiba is very well-suited for outcomes-based contracts since hypoglycemia is such a costly complication for payers (and of course, for patients), and we hope to see these unfold in the near-future.

Importantly, none of this discussion or speculation is meant to undermine this regulatory milestone. In many ways, it is extraordinary that FDA included DEVOTE hypoglycemia data (a secondary endpoint, no less!) on Tresiba’s label. As Dr. Hobbs put it, “it’s something we questioned would ever happen.” This revision is certainly a step in the right direction as we push for FDA to more closely consider outcomes beyond A1c.

In our opinion, the best-case scenario for this FDA decision would have been (i) inclusion of both DEVOTE and SWITCH on the label; (ii) language emphasizing that hypoglycemia is a substantial burden for patients and for the health system; and (iii) a recommendation that the next-gen be used in place of earlier standard of care for hypoglycemia risk reduction. Someday, we hope to see this best-case scenario become a reality on Tresiba’s US product label.

Reassuring HCPs/Patients on Tresiba’s CV Safety

  • Dr. Hobbs emphasized that having the CV data from DEVOTE on the product label will help reassure HCPs/patients on CV safety. Novo Nordisk’s first NDA for Tresiba was met with a Complete Response Letter (CRL) from FDA related to CV risk, and Dr. Hobbs explained how some of those concerns linger today in the diabetes community. These worries can now be put to bed with DEVOTE. Tresiba showed no elevated risk for three-point MACE (non-fatal MI, non-fatal stroke, CV death) vs. Lantus (HR=0.91, 95% CI: 0.78-1.06, p<0.001 for non-inferiority); the CV results trended in favor of insulin degludec over insulin glargine, but didn’t reach statistical significance (p=0.21 for superiority). “It’s nice to finally close the issue and reassure patients and clinicians that there is no increased risk of CV events,” Dr. Hobbs said. “That’s a big win. We’re happy to see the CVOT completed and in the label.”

Next Commercial/Clinical Steps for Tresiba

  • Dr. Hobbs highlighted Novo Nordisk’s commitment to hypoglycemia education, though he told us to “stay tuned” for specifics on how this label update will influence promotional activity around Tresiba. At diabetes conferences, the company has been hosting unbranded booths focused on hypoglycemia (see our ENDO 2018 coverage for one example), showing HCPs what signs and symptoms to look out for in their diabetes patients. This is fantastic in our view, since we see an immense need for better real-world education on hypoglycemia, directed at both patients and providers. We’re eager for Novo Nordisk to ramp up this initiative now that Tresiba’s hypoglycemia benefit is reflected in product information. We’ll also be keeping a close eye on volume/sales figures for the Tresiba franchise in 2Q17 and beyond, as the new label is rolled out, although the most appreciable impact may not come until payer negotiations for 2019 formularies.
  • When we asked about using CGM to record severe and non-severe hypoglycemia with Tresiba vs. other basal insulins, Dr. Hobbs suggested that this is also a priority for Novo Nordisk, although he noted that payers today are less swayed by non-severe hypos since they don’t overtly cost money. He elaborated, “we want to do what’s right for patients, even if payers, unfortunately, don’t see as much value in lessening non-severe hypoglycemia.” That’s a pity since they certainly have an impact on productivity, absenteeism, presenteeism, and we suspect they are also more associated with severe hypo though we do not know for sure about the data on this front. We were pleased to hear Novo Nordisk’s stance on this, as we continue to push for more CGM in diabetes clinical trial. We know with certainty that severe hypo could be reduced by using CGM as well, so we would be excited by the combination. Interestingly, hypoglycemia is the primary endpoint in an ongoing head-to-head investigation of Tresiba vs. Sanofi’s next-gen Toujeo (insulin glargine U300). This study is expected to complete in December 2018 according to, and we anticipate that this will generate even more awareness/discussion around basal insulins and hypoglycemia.


-- by Payal Marathe and Kelly Close