Novo Nordisk’s oral semaglutide meets primary endpoint in PIONEER 4 (vs. Victoza) + PIONEER 7 (vs. Januvia) – June 21, 2018

Executive Highlights

  • Novo Nordisk has released topline results from PIONEER 4 and PIONEER 7, which come on the heels of topline data from PIONEER 1 and 2. This phase 3 program is evaluating oral semaglutide (the likely first-to-market oral GLP-1 agonist) for type 2 diabetes.

    • After 26 weeks in PIONEER 4, 14 mg oral semaglutide was non-inferior to Novo Nordisk’s Victoza (liraglutide) on A1c-lowering and was superior to Victoza on weight loss. These findings are key, suggesting that oral semaglutide is at least as efficacious as an injectable GLP-1. While oral delivery is a tremendous win, we don’t want patients to have to sacrifice on efficacy.
    • In PIONEER 7, oral semaglutide showed superiority to Merck’s DPP-4 inhibitor Januvia (sitagliptin) on percent of patients reaching A1c <7% after 52 weeks. Unsurprisingly, oral sema was also a more effective weight-lowering agent. PIONEER 7 allowed for dose adjustment of oral semaglutide, meaning providers could change a patient’s dose based on glycemic trajectory and side-effects.
  • Novo Nordisk has emphasized on-treatment analyses for all PIONEER studies reporting topline data so far. This secondary statistical approach doesn’t reflect patients who discontinued treatment or who required rescue medication. Our main question now (and we imagine many people in the field want to know) is what exactly is causing the discrepancy between primary and secondary results? Can this be attributed to early treatment discontinuation? How is efficacy impacted if patients don’t adhere to all fasting requirements? To our knowledge, PIONEER protocol asks participants to take oral semaglutide in a fasting state in the morning and to refrain from eating for another 30 minutes thereafter.

    • In the on-treatment analysis of PIONEER 4, oral semaglutide gave A1c reductions of 1.3% and 1.2% at weeks 26 and 52, vs. -1.1% and -0.9% with Victoza. Weight loss with oral semaglutide was ~10 lbs and ~11 lbs at weeks 26 and 52, vs. ~7 lbs at both time points with Victoza. No p-values were reported, though Novo Nordisk’s announcement states that all comparisons were statistically significant.
  • In the on-treatment analysis of PIONEER 7, A1c declined by 1.4% with oral semaglutide after 52 weeks vs. -0.7% with sitagliptin (statistically significant, no p-value reported). Weight loss was ~6 lbs vs. ~2 lbs (statistically significant, no p-value reported). At 52 weeks, 60% of participants were receiving 14 mg oral semaglutide, 31% were taking 7 mg, and 9% were taking 3 mg – this likely explains the lower magnitude of weight loss compared to other trials, though A1c reduction was on par with PIONEER 1, 2, and 4 and nausea/treatment discontinuation were just as common. We’re looking forward to seeing these results broken down by dose.
  • In both PIONEER 4 and 7, nausea was the most common side-effect and ~10% of patients discontinued oral sema treatment due to adverse events. Oral semaglutide’s side-effect profile seems to match that of existing GLP-1 agonists. PIONEER 4 offers compelling evidence of this – 20% of oral sema patients experienced nausea vs. 18% of liraglutide patients, and 11% discontinued vs. 9% of liraglutide patients. That said, as seen in PIONEER 2 and 7, premature treatment discontinuation was much less common with SGLT-2 or DPP-4 agents (4% and 3%, respectively). This could come into play if Novo Nordisk continues with its previously-outlined plan to position oral semaglutide alongside existing oral agents rather than existing injectable GLP-1s.

Yesterday, Novo Nordisk announced positive topline results from PIONEER 4 and PIONEER 7, the third and fourth phase 3 trials to report for oral semaglutide. PIONEER 4 (n=711) was a 52-week head-to-head trial of oral sema vs. injectable GLP-1 agonist Victoza (liraglutide). PIONEER 7 (n=504) was a 52-week study comparing oral sema with dose adjustment vs. DPP-4 inhibitor Januvia (Merck’s sitagliptin), and both sets of results could have major implications on marketing and commercialization of this first oral GLP-1 agonist.

Oral semaglutide met its primary endpoint in both trials – non-inferior A1c reductions vs. Victoza at 26 weeks and a significantly greater proportion of patients achieving A1c <7% vs. Januvia after one year. Across both studies, participants on oral semaglutide experienced significantly greater weight loss vs. their counterparts on a comparator drug (at 26 weeks vs. Victoza in PIONEER 4 and at 52 weeks vs. Januvia in PIONEER 7).

This announcement follows topline data from PIONEER 1 (vs. placebo) released in February and from PIONEER 2 (vs. Lilly/BI’s SGLT-2 inhibitor Jardiance) released in May. Full PIONEER 1 results will be presented as a late-breaking poster at ADA on Sunday, June 24.

Overall, we’re impressed by the phase 3 data that we’ve seen so far on oral semaglutide. This candidate seems to offer profound glucose-lowering (superior A1c reductions vs. placebo, SGLT-2 Jardiance, and DPP-4 Januvia) and meaningful weight loss – in other words, the potency of a GLP-1 agonist in an oral rather than injectable formulation. Novo Nordisk is targeting 2019 for an NDA, meaning an oral GLP-1 could be available to type 2 patients by 2020; that’s beyond exciting.

We do think that demonstrating strong safety, tolerability, and ease of use for oral semaglutide may be a challenge, albeit a conquerable one for Novo Nordisk. This company is very experienced in the GLP-1 market; Victoza leads the class in sales and we hear the Ozempic launch is going well.

Before we dive into the data, it’s important to note that Novo Nordisk conducted primary and secondary statistical analyses in assessing PIONEER 4 and 7 (and 1 and 2, for that matter). The primary approach evaluates data from all randomized participants regardless of treatment discontinuation or rescue medication use. The secondary approach describes on-treatment effects, excluding data from those who discontinued treatment or who required rescue medication. The company only provided numerical results for the secondary statistical analysis.

The key question to consider: If findings are significant under the secondary statistical approach but not the primary one, does that mean that dropout rate is high for oral semaglutide, implying low tolerability/adherence? We’ll be digging deeper as we hear full results from PIONEER trials, but for now, we’re optimistic that Novo Nordisk can create a patient-friendly product.

Next up, we look forward to results from PIONEER 3 (n=1,864), expected by June 30. PIONEER 3 randomized patients to 3 mg, 7 mg, or 14 mg oral semaglutide once-daily or to 100 mg sitagliptin once-daily for 78 weeks.

PIONEER 4 Deep Dive

  • In PIONEER 4, 14 mg oral semaglutide (the highest dose being investigated) met its primary endpoint by demonstrating non-inferiority to Victoza on A1c-lowering at 26 weeks. Novo Nordisk’s topline release doesn’t disclose baseline A1c, margin of A1c drop, or p-value under the primary statistical approach. Weight loss with oral semaglutide was also superior to weight loss with Victoza at 26 weeks, using the primary approach.

  • Turning to the on-treatment analysis: Oral semaglutide gave A1c reductions of 1.3% and 1.2% at week 26 and week 52, respectively, vs. 1.1% and 0.9% with Victoza and 0.1% and 0.2% with placebo. Novo Nordisk states that oral semaglutide was superior to both Victoza and placebo on this glycemic endpoint, but no p-values were reported. In the oral sema group, 69% of participants achieved target A1c <7% after 52 weeks compared to 63% of patients in the liraglutide group and only 18% of patients on placebo. Oral semaglutide led to a mean ~10 lbs weight loss from baseline at 26 weeks and ~11 lbs at 52 weeks vs. ~7 lbs and ~7 lbs with Victoza. No baseline body weight or p-values were included in the topline release, but both comparisons met statistical significance, implying greater weight loss efficacy with oral semaglutide compared to the current market-leading GLP-1. Weight loss in the PIONEER 4 placebo group was ~2 lbs and ~3 lbs at 26 and 52 weeks, respectively.

    • For comparison, in the PIONEER 1 on-treatment analysis, 14 mg oral semaglutide gave a 1.5% reduction in A1c at 26 weeks (vs. 0.1% with placebo; baseline A1c 8.0%). Moreover, 80% of this treatment arm achieved A1c below 7% vs. 35% of patients on placebo. The 14 mg oral sema dose also gave ~9 lbs weight loss at 26 weeks vs. only ~3 lbs with placebo. Taking these results together, oral semaglutide seems to lower A1c by 1.0%-1.5% over 26 weeks, also stimulating ~10 lbs weight loss in that time frame. We’re particularly intrigued by the continued/non-plateauing weight loss over 52 weeks (~10 lbs at 26 weeks and ~11 lbs at 52 weeks in PIONEER 4), because this was also seen in phase 2 obesity studies of injectable semaglutide, and it could make this molecule a groundbreaking weight loss therapy for people with diabetes and/or obesity.

  • Adverse events/discontinuations: 20% of patients on oral semaglutide experienced nausea vs. 18% on Victoza and 4% on placebo. Notably, 11% of those randomized to oral semaglutide discontinued treatment due to adverse events vs. 9% on Victoza and 4% on placebo; 7% of patients on 14 mg oral sema discontinued therapy prematurely in PIONEER 1, 11% discontinued from PIONEER 2 (compared to only 4% of patients taking the SGLT-2 inhibitor), and 9% discontinued from PIONEER 7 (see below). This is the main criticism we’ve picked up on when it comes to oral semaglutide – that real-world adherence (and therefore efficacy) may not be as high as anticipated for an oral pill due to adverse events and fasting protocol. Some have expressed concern that ~6.5 hours of required fasting (6 hours pre-dose + 30 minutes post-dose) may impose undue burden on patients, canceling out the benefits of lower injection burden. Lilly management has espoused this view, though oral semaglutide will compete with the company’s GLP-1 Trulicity (dulaglutide) and/or its BI-partnered SGLT-2 Jardiance (empagliflozin). In LEADER (the CVOT for Victoza), 9.5% of trial participants on liraglutide discontinued due to adverse events vs. 7.3% of placebo participants. While it’s difficult (if not impossible) to draw comparisons between studies with such different designs and time courses, it is interesting to see from LEADER and PIONEER 4 that a ~10% dropout rate is not unheard of for a GLP-1 agonist.

    • Nevertheless, we imagine Novo Nordisk will have some work to do in promoting oral semaglutide’s tolerability and ease of use. This will be especially crucial if the company positions the drug as a competitor to current oral agents like SGLT-2s and DPP-4s, since discontinuation from PIONEER was much less common among participants on Jardiance or Januvia. Our sense is that oral semaglutide’s safety profile is more or less consistent with other GLP-1 agonists (e.g. Victoza), which means more frequent GI side-effects at onset of treatment compared to SGLT-2s or DPP-4s (these symptoms subside over time), but that Novo Nordisk will have to manage expectations for how convenient an oral therapy should be, given the fasting restrictions.
  • PIONEER 4 is an especially interesting (and important) study within this entire phase 3 program because it shows oral semaglutide to be at least as efficacious as an injectable GLP-1 agonist for patients with type 2 diabetes. Oral delivery of GLP-1 is an attractive proposition, as needle phobia prevents many patients from starting on a GLP-1 agonist and high injection burden precludes optimal adherence. That said, we’re not sure oral semaglutide would be successful if patients, HCPs, and payers were asked to sacrifice on efficacy. And so, these positive PIONEER 4 results are key. We also think it’s worth repeating that nausea and discontinuation were not terribly imbalanced between the two treatment arms in this study, suggesting that tolerability with oral semaglutide is not drastically different from that of Victoza or other existing GLP-1 options.

PIONEER 7 Deep Dive

  • In PIONEER 7, oral semaglutide met its primary endpoint by showing superiority to sitagliptin at 52 weeks on an endpoint of proportion of participants achieving A1c <7%. Oral semaglutide was also superior to sitagliptin on weight loss, both according to the primary statistical approach. No p-values were reported, and we’ll be looking for this info when PIONEER 7 data is presented in full.

  • Turning to the on-treatment analysis: A1c dropped by 1.4% after 52 weeks of oral semaglutide vs. a 0.7% drop with Januvia (baseline A1c 8.3%).Novo Nordisk mentioned that this was a statistically significant benefit to oral semaglutide therapy. In addition, 63% of participants receiving oral semaglutide achieved an A1c <7% after 52 weeks, significantly greater than the 28% of people reaching this A1c goal with sitagliptin. The on-treatment analysis revealed ~6 lbs mean weight loss with oral semaglutide at week 52 vs. ~2 lbs with sitagliptin. No p-values were shared for A1c reduction, proportion of patients reaching target A1c, or weight loss.

  • Adverse events/discontinuations: 21% of people on oral semaglutide experienced nausea, compared to 2% on sitagliptin. Further, 9% receiving oral semaglutide discontinued treatment due to adverse events vs. 3% receiving sitagliptin. In PIONEER 4, 11% of the oral semaglutide group prematurely stopped therapy due to adverse events – see above for this data and for our thoughts thus far on oral semaglutide’s safety/tolerability profile. In brief, we believe the current fasting requirements would be difficult for some patients to follow, but we don’t view these as an insurmountable obstacle for all patients. If oral semaglutide is offering the glucose-lowering, weight loss, and possibly even cardioprotection that typically comes with an injectable GLP-1 agonist, we imagine this is a trade-off that many patients might be willing to make.

  • Of note, PIONEER 7 was an open label trial that allowed for dose adjustment of oral semaglutide based on glucose control and drug tolerability. This more closely mimics real-world clinical scenarios. At 52 weeks, 60% of participants were receiving 14 mg oral semaglutide, 31% were receiving the 7 mg dose, and 9% were receiving the 3 mg dose. While it’s difficult to make judgments without knowing baseline characteristics, we do immediately notice that PIONEER 7 saw lower weight loss with oral semaglutide vs. PIONEER 2and PIONEER 4, where only the 14 mg dose was administered. It’s reassuring that A1c reduction was similar despite a proportion of patients receiving lower doses, but we note that nausea and dropout rates were also about the same; presumably, dose adjustment would allow providers to help minimize GI symptoms for their patients. We’re eager for more detail on the relationship between dose, efficacy, safety, and discontinuations with oral semaglutide – this will be critical information as Novo Nordisk moves toward regulatory submission.

Close Concerns Questions

Q: Which dose of oral semaglutide will Novo Nordisk commercialize? Is the highest, 14 mg dose associated with greater efficacy alongside more frequent/intense GI side-effects?

Q: What instructions and support did HCPs receive in PIONEER 7 to support oral semaglutide dose adjustment?

Q: How will the primary vs. secondary statistical approaches play out as Novo Nordisk seeks regulatory approval for oral semaglutide? How will different stakeholders react to this (payers, HCPs, etc.)?

Q: What exactly explains the differences in results between the primary and secondary statistical approaches? Is it low adherence/high dropout rates from the oral semaglutide arm? Is the efficacy of oral semaglutide attenuated if patients don’t fulfill all fasting requirements (e.g. eating before 30 minutes have passed)?

Q: Will oral semaglutide compete more so with existing GLP-1s or with existing oral agents (SGLT-2s, DPP-4s)? Previously, Novo Nordisk has outlined a promotional strategy wherein oral semaglutide would compete with other oral agents, which appear earlier in the traditional diabetes treatment paradigm than injectable GLP-1s. PIONEER 2 results support that this could be an effective commercial strategy, as oral semaglutide showed significant benefit on A1c and body weight vs. Lilly/BI’s Jardiance (empagliflozin) at one year – though, admittedly, the A1c difference was much more impressive than the weight loss results. That said, it is our understanding that Novo Nordisk’s strategy for commercializing and marketing oral semaglutide remains “up in the air.” Price will play a huge part (and to this end, Lilly management has argued that oral semaglutide won’t be able to match the value proposition of Jardiance and other SGLT-2s in terms of efficacy/list price).

Q: How will oral semaglutide be priced on the market (assuming FDA approval comes through)? As is the case for any new diabetes drug, pricing questions loom large, but management emphasized on the company’s 1Q18 call that they will make no additional comments on pricing until the therapy is approved.

Phase 3 PIONEER Program


Estimated Enrollment






Completed December 2017; Topline results announced February 2018; Full results coming on ADA poster (2-LB)



Lilly/BI’s Jardiance (empagliflozin)

Completed March 2018; Topline results announced May 2018



Merck’s Januvia (sitagliptin)

Completed March 2018, Topline results expected by June 30, 2018



Novo Nordisk’s Victoza (liraglutide)

Completed March 2018; Topline results announced June 2018



Moderate renal impairment

Expected to complete May 2018




Expected to complete September 2018



Flexible dose escalation

Completed March 2019; Topline results announced June 2018



Insulin add-on

Expected to complete August 2018



Placebo and liraglutide in Japan

Expected to complete August 2018



Lilly’s Trulicity (dulaglutide) as an add-on to oral agents in Japan

Expected to complete July 2018


-- by Ann Carracher, Payal Marathe, and Kelly Close