NVO/LLY: Ozempic (semaglutide) resets the bar for glycemic efficacy; consensus GLP-1 expectations appear too low

Richard
Print Friendly
Share on LinkedIn0Tweet about this on Twitter0Share on Facebook0

SEE END OF THIS REPORT FOR IMPORTANT DISCLOSURES

Richard Evans / Scott Hinds

203.901.1631 /.1632

revans@ / shinds@ssrllc.com

@SSRHealth

February 5, 2018

NVO/LLY: Ozempic (semaglutide) resets the bar for glycemic efficacy; consensus GLP-1 expectations appear too low

  • We see global 2022 sales for NVO’s GLP-1’s (Ozempic/semaglutide and Victoza) of $6.8B to $9.7B, v. consensus of $6.2B; our low estimate assumes LLY’s GLP-1 Trulicity is labelled ‘effective’ for reduction of CV risk, the hi estimate assumes Trulicity is simply labelled CV ‘safe’. NVO’s valuation implies that buy-side expectations already are well above sell-side consensus; as such we do not conclude that NVO is undervalued
  • Ozempic resets the bar for glycemic efficacy, and in so doing makes the trade-offs associated with GLP-1 therapy (self-injection, GI side effects) that much more worthwhile. We expect the GLP-1 category to expand from its current 9.5% share of non-metformin Rx’s to circa 20% share. DPP-IV’s are weaker oral agents which address the same (‘incretin’) metabolic pathway from a different angle; these agents currently account for 18.5% of non-metformin Rx’s
  • NVO’s oral semaglutide appears to be as effective as the injectable (Ozempic); however, it must be dose-escalated much more slowly to avoid rate-limiting GI side effects. We see oral semaglutide as a similarly convenient but much more effective means of modulating the incretin pathway than DPP-IV’s; as such success of oral semaglutide should accelerate the expansion of GLP-1’s, presumably at the expense of the DPP-IV’s. Oral semaglutide completes phase 3 late this year

Investment conclusion

Consensus anticipates $6.2B in global net GLP-1 sales for NVO in 2022, equal to 26.5% of total company sales. We expect between $6.8B and $9.7B in global net GLP-1 sales for NVO, depending on results from LLY’s REWIND cardiovascular (CV) outcomes trial for Trulicity. Our higher ($9.7B) estimate for NVO is consistent with REWIND showing Trulicity to be ‘CV safe’ (i.e. not contributing to CV events, but not reducing them either); our lower ($6.8B) estimate is consistent with REWIND showing Trulicity to be ‘CV effective’ (i.e. reducing CV events; a CV profile superior to NVO’s Ozempic, but on par with NVO’s Victoza). NVO consensus appears reasonable if REWIND shows efficacy, but low if REWIND shows safety. NVO’s 2022 P/S and P/E ratios are well above peers’ (Exhibits 1, 2), which implies that buy-side expectations may track much closer to our estimates than to sell-side consensus; as such we cannot conclude NVO is undervalued[1], despite our higher GLP-1 forecasts

Consensus anticipates $4.1B in global net Trulicity sales for LLY in 2022, equal to 14.3% of total company sales. We expect between $3.7B and $5.9B; the lower estimate assumes Trulicity is ‘CV safe’, and the higher assumes Trulicity is ‘CV effective’. LLY consensus appears reasonable to slightly high if REWIND shows Trulicity to be ‘CV safe’, and too low if REWIND shows Trulicity to be ‘CV effective’. Our expectations for Trulicity bracket consensus, and being unable to reliably handicap the results of REWIND, we cannot conclude (solely on the basis of Trulicity expectations) that LLY appears misvalued

How NVO’s Ozempic fits in

The US accounts for a bit more than three-quarters of global GLP-1 sales; competition in the US market category is mainly between Victoza (NVO) and Trulicity (LLY)[2] (Exhibit 3). Both have similar efficacy and tolerability[3]; however, Victoza reduces cardiovascular (CV) risk but is dosed once-daily, whereas Trulicity is dosed weekly but its impact on CV risk is not yet established. And, Victoza appears to be slightly more effective than Trulicity at reducing body weight[4]

NVO’s Ozempic, recently approved in the US, appears to have greater efficacy than any of the current GLP-1’s, having beaten basal insulin on efficacy (reduction in A1C[5]), where Trulicity and Bydureon (AZN) showed non-inferiority to basal insulin (Exhibit 4). More to the point, Ozempic has shown superior efficacy (A1C and body weight reduction) v. Trulicity in a head to head comparison[6]. Ozempic has proven its CV safety[7], but unlike Victoza does not have a labelled indication for outright reduction of CV risk. Trulicity is conducting a major CV risk study (REWIND[8],[9]) which is due to report in late 2018; it is possible that Trulicity could show CV efficacy, giving the brand a labelled advantage (at least with respect to CV outcomes) relative to Ozempic

Tolerability across these three brands appears to be similar. Ozempic has a greater relative incidence (active: placebo) of diarrhea than Trulicity or Victoza, but a lower relative incidence of nausea at its highest dose (as compared to Trulicity and Victoza nausea rates at these brands’ high doses), and a lower relative incidence of vomiting at either dose than Trulicity or Victoza at either dose (Exhibit 5)

In a single pivotal trial (SUSTAIN-6[10]) patients on Ozempic had a higher incidence of retinopathy complications than patients on placebo. On post hoc analysis[11] retinopathy complications in Ozempic patients were largely confined to patients who had significant retinopathy at the start of the trial, and who had very large declines in A1C after initiating Ozempic. Worsening of diabetic retinopathy is a known complication for patients experiencing sudden A1C drops when initiating insulin, and it is reasonable to believe that Ozempic’s effects on retinopathy complications have more to do with the sudden large change in A1C than with any other drug-specific effects. Ozempic’s package insert calls for patients with a history of diabetic retinopathy to “be monitored for progression of diabetic retinopathy.” American Diabetic Association (ADA) guidelines[12] call for all patients with type 2 diabetes to receive a thorough retinal examination at the time of diagnosis, and for patients with evidence of diabetic retinopathy to be examined at least annually, or more frequently if retinopathy is progressive or sight threatening. Because the Ozempic labelled recommendation is consistent with existing practice standards, it is unlikely that monitoring Ozempic patients for retinopathy creates additional clinical work

Comparing relative efficacy, tolerability and safety across the class, the implication is that Ozempic becomes the preferred GLP-1 by virtue of glycemic (i.e. blood glucose lowering) efficacy, with the potential exceptions of patients with established cardiovascular disease (CVD) or very high CVD risk, and/or patients with diabetic retinopathy (DR). The relative fates of Ozempic and Trulicity hinge largely on the results of Trulicity’s ongoing REWIND CV outcomes trial. If Trulicity’s CV outcomes are no better than Ozempic’s, patients with CVD or very high CVD risks presumably would continue to be placed on Victoza at higher CVD risk levels, then on Ozempic or Trulicity (with Ozempic being preferred because of greater glycemic efficacy) at moderate to lower CVD risk levels. Conversely if Trulicity shows CV efficacy at least on par with Victoza – as is entirely possible given the ambitions and design of the REWIND trial — patients with higher CVD / risk would tend to go to Trulicity simply because of its weekly (v. daily for Victoza) dosing advantage. In this same Trulicity CV safety scenario, as degree of CVD or CVD risk declines Ozempic should become more preferred, as glycemic control arguably outweighs incremental cardiovascular efficacy at lower CVD risk levels

Relevant epidemiology – diabetic retinopathy, and cardiovascular comorbidity

Globally, about one-third (34.6%) of persons with diabetes (type 1 or 2) will have some degree of diabetic retinopathy (DR), and about one-third of persons with DR (10.2% of persons with diabetes) will have vision-threatening DR (VTDR)[13]. In the US, an estimated 28.5% to 40.3% of patients with type 2 diabetes have DR, and 4.4% to 8.2% have VTDR[14]. In Ozempic’s Western, non-cardiovascular outcomes oriented SUSTAIN clinical trials, DR rates at baseline ranged from a low of 3.7% (SUSTAIN-3) to a high of 13.9% (SUSTAIN-5)[15]. In SUSTAIN-6 (a cardiovascular outcomes trial with a more severely affected type 2 population), the baseline DR rate was 29.4%. Diabetic retinopathy tends to be more prevalent in patients with greater levels of CVD risk, which likely accounts for the higher DR rate in SUSTAIN-6 as compared to SUSTAIN 1-5[16]

The absolute prevalence of cardiovascular disease in type 1 and type 2 diabetics in the United States is approximately 26%, a rate that lies just above the mid-point of rates (8.6% to 40.5%) found in countries surveyed by the International Diabetes Federation (IDF)[17]

The potential for class expansion

Metformin is the standard starting point of type 2 diabetes treatment, thus competition among newer technologies is largely about what’s added to baseline metformin for patients who aren’t controlled on metformin alone

Sulfonylureas[18] are the most common type 2 diabetes therapy after metformin (Exhibit 6). The sulfonylureas (and meglitinides[19]) work by increasing insulin secretion; sulfonylureas are far more commonly used, with meglitinides generally being reserved for patients intolerant to sulfonylureas. Sulfonylureas offer relatively high glycemic efficacy, but tend to cause weight gain, and can cause hypoglycemia. The sulfonylureas are generically available, and so have the advantage of very low prices

Basal insulin[20] is the second most common type 2 diabetes product after metformin; efficacy is very high, but risks of hypoglycemia also are high. Also insulin tends to cause weight gain, and treatment requires daily injections as well as routine blood glucose monitoring. Basal insulin prices are falling in the wake of formulary restrictions and the recent entry of a biosimilar (LLY’s Basaglar) to SNY’s once-dominant Lantus

Because the pathophysiology of type 2 diabetes involves declining sensitivity to the body’s own (i.e., ‘endogenous’) insulin, treating with sulfonylureas and basal insulin is somewhat akin to pressing the accelerator harder to overcome the performance limits of a failing engine – the car goes faster for a while, but the engine eventually collapses under the strain. As their conditions progress patients need ever-higher doses of sulfonylureas and/or basal insulin to maintain glycemic control, with failure of glycemic control being a common mid- to longer-term result

The DPP-IV inhibitors[21] are the next most common option; they and GLP-1’s affect the same (‘incretin’) pathway at different points. Endogenous GLP-1 lowers blood glucose by improving glucose dependent insulin secretion, slowing gastric emptying, reducing post-meal glucagon (glucagon raises blood glucose levels), and reducing food intake – like tuning the engine rather than dumping more gas into it. Endogenous DPP-IV destroys endogenous GLP-1, and thus reduces these desirable effects; conversely DPP-IV inhibitors allow endogenous GLP-1 to survive longer. Exogenous GLP-1’s obviously increase GLP-1 levels by adding to what’s already there. The DPP-IV inhibitors have the advantage of being orally administered, and of being highly tolerable; however they have no significant effect on weight, have modest efficacy (considerably less than sulfonylureas and basal insulins), and some (Onglyza (AZN), Nesina (Takeda)) have been associated with increased risk of congestive heart failure (CHF). The GLP-1’s have the disadvantage of being injected (either daily (Victoza) or weekly (Trulicity, Ozempic)), and cause moderate levels of GI disturbance; however the GLP-1’s are highly effective (Ozempic has beaten basal insulin (glargine), Victoza has beaten a sulfonylurea (glimepride)), promote weight loss, and either create no additional CV risks (Ozempic), or reduce CV risks (Victoza) (Trulicity CV outcomes are still pending)

SGLT-2 inhibitors[22] promote renal excretion of glucose, thus lowering blood glucose levels. The class has relatively limited glycemic efficacy (less than sulfonylureas and basal insulins); however, these agents also offer little risk of hypoglycemia, and can lead to weight loss. At least one SGLT-2 (Jardiance (LLY)) improves CV outcomes; however, one (Invokana (JNJ)) has been shown to increase the risk of amputation

Thiazolidinediones[23] (TZDs) increase insulin sensitivity; these agents offer efficacy approaching that of sulfonylureas and basal insulin and have limited risk of hypoglycemia; however, they do tend to cause weight gain, and have been associated with increased CV risk (specifically congestive heart failure, or CHF)

The GLP-1 category currently accounts for a small share – just less than 10% – of the non-metformin type 2 diabetes prescription market (Exhibit 4, again). With best-in-class (and better than basal insulin) efficacy, CV safety, weight loss, and weekly dosing, we believe Ozempic can help drive the GLP-1 category’s share substantially higher – to 20% or more of non-metformin Rx volume. This implies the potential for a doubling of total GLP-1 Rx demand, before accounting for overall type 2 diabetes category growth

Pricing considerations

Real, sales-weighted US list price growth in the GLP-1 category was 11.1% in 3Q17, down from 16.7% in the year prior quarter. Real, sales-weighted net price growth in the category was -2.0% in 3Q17, as compared to -8.4% in the year prior (Exhibit 7). Excluding Medicaid effects, the list to net discount rate in 3Q17 was 45.1% in 3Q17, as compared to 38.7% in the year prior quarter (Exhibit 8)

Zeroing in on the net price trend for Victoza and Trulicity, Victoza shows 3.9% real net inflation in 3Q17, v. -1.4% for Trulicity (Exhibit 9). Recall that these net price values are estimated by comparing companies’ ex-factory net sales to third party reported end-user unit demand. Because Trulicity is a relatively new product with rapid growth, it is reasonable to believe that ex-factory shipments exceeded end-user pull through in the early quarters, making net price / unit look larger than it actually was; and, as ex-factory shipments fell to a level closer to end-user pull through, this would look like a fall in net price / unit, even if ‘true’ net price / unit were stable or slightly growing. Victoza, being past its launch period and thus almost certainly having a closer match between ex-factory shipments and end-user pull through, should be more representative of the category’s true net pricing trend. As such, we expect category real net inflation is close to Victoza’s 3.9% level

Because the GLP-1 category, especially with Ozempic’s launch and Trulicity’s pending CV outcomes, is still sorting itself out with respect to clinical distinctions; and, because the category isn’t yet large enough (or its rate of net inflation fast enough) for it to be a significant sales-weighted driver of total market net inflation, it’s probably too early for formulary managers to begin bidding one product against the other. Because of this, it’s reasonable to expect that US net price inflation of +/-3% real (or circa 5% nominal) can continue, at least for a couple of years

Ozempic’s list price per month of treatment matches Trulicity’s, though it’s too early to see Ozempic’s net price. For our forecasts, we assume (list and net) price parity between these two brands

GLP-1unit share: breakout by presence/ absence of CVD risk, outright CVD, and/or DR

We forecast GLP-1 unit share under two broad scenarios, one in which Trulicity’s REWIND outcomes trial shows Trulicity to be ‘safe’ (but not ‘effective’, i.e. not reducing CV risks), and one in which REWIND shows Trulicity to be CV ‘effective’

In both scenarios we assume that unit demand for GLP-1’s roughly doubles by the end of 2022, and that US net prices inflate by 5% (nominal) annually. We also assume that the proportion of GLP-1 patients with no CVD or DR is 66%, the proportion with CVD +/- DR is 26%, and that the proportion with DR but without CVD is 8%. The difference in product-level forecasts across the two scenarios is driven entirely by differences in share of CVD patients

In both scenarios we assume that Ozempic captures 65% of patients without CVD or DR, as compared to 5% for Victoza, 25% for Trulicity, and 5% for ‘other[24]’. In patients without CVD risk Victoza’s CV efficacy advantage isn’t as meaningful as glycemic efficacy and dosing convenience; dosing convenience implies most of the demand should go to Ozempic and Trulicity (as once weekly products, v. Victoza’s daily dosing); glycemic efficacy implies most of the demand should go to Ozempic rather than Trulicity. And, the absence of DR should mitigate any DR-associated concerns about choosing Ozempic

In both scenarios we also assume that Trulicity captures 55% of patients with DR but no CVD, as compared to 35% for Ozempic, 5% for Victoza, and 5% for ‘other’. In these patients the absence of CVD means glycemic control and dosing convenience trump Victoza’s daily dosing and CV efficacy, giving the majority of demand to Trulicity and Ozempic. However, because Ozempic has been associated with worsening of DR but Trulicity has not, it’s reasonable that Trulicity should be preferred to Ozempic in these patients

Patients with CVD +/- DR is the category where product shares differ across scenarios. In the ‘Trulicity CV safe’ scenario, Victoza arguably remains the preferred option (especially for patients with more severe CVD / CVD risks) because of its CV efficacy, despite its dosing disadvantage (daily for Victoza, weekly for Ozempic and Trulicity). For patients with more moderate CVD / CVD risks, dosing convenience and glycemic efficacy may trump Victoza’s profile, leading these patients to Ozempic or Trulicity. We assume Ozempic takes slightly greater share by virtue of its greater glycemic efficacy; however, its share advantage v. Trulicity is mitigated by the high co-prevalence of DR in this CVD population. Assumed 2022 unit shares of this patient segment are 48% Victoza, 28% Ozempic, and 25% Trulicity

In the ‘Trulicity CV effective’ scenario, we assume Trulicity’s CV outcomes are on par with Victoza’s, making Trulicity the preferred option in patients with CVD +/- DR. In this scenario assumed 2022 unit shares of this patient segment are 90% Trulicity, 5% Ozempic, and 5% Victoza

Under the ‘Trulicity CV safe’ scenario we forecast 2022 US net sales of $5.6B for Ozempic, $2.1B for Victoza (making $7.7B for NVO), and $2.9B for Trulicity. Under the ‘Trulicity CV effective’ scenario we forecast $5.0B for Ozempic, $400M for Victoza (making $5.4B for NVO), and $4.7B for Trulicity (Exhibit 10). Details of the two scenarios are provided in Appendices I and II

Our estimates v. consensus

Consensus calls for global 2022 sales of Ozempic and Victoza of $3.3B and $2.9B respectively, for $6.2B combined GLP-1 sales, representing 26.5% of consensus total revenues for NVO in that year. Our high case for NVO (‘Trulicity CV safe’) calls for US sales of $5.6B and $2.1B for Ozempic and Victoza, respectively, for $7.7B in total US GLP-1 sales. We assume the US will be roughly 79%[25] of global GLP-1 sales by 2022, implying $9.7B in global GLP-1 sales for NVO under the ‘Trulicity CV safe’ scenario. Under the lower (for NVO) ‘Trulicity CV effective’ scenario we call for $5.0B and $0.4B in Ozempic and Victoza US net sales in 2022; $5.4B in total US GLP-1 sales, and roughly $6.8B in global GLP-1 sales. Thus, from our perspective, 2022 consensus global estimates ($6.2B) are more consistent with a scenario in which Trulicity is proven to be CV effective ($6.8B in global NVO GLP-1 sales per our estimate), than a scenario in which Trulicity is simply shown to be CV ‘safe’ ($9.7B in global NVO GLP-1 sales per our estimate)

Consensus calls for global 2022 Trulicity sales of $4.1B, accounting for 14.3% of total consensus revenue for LLY. In the ‘Trulicity CV safe’ scenario we estimate $2.9B in US net sales for Trulicity, or roughly $3.7B in global net sales. Under the ‘Trulicity CV effective’ scenario we estimate $4.7B in US net Trulicity sales, or roughly $5.9B in global net Trulicity sales

Across all GLP-1 products, consensus calls for $11.7B in 2022 global net sales. Our estimates are marginally ($1.7B, 15%) higher; we estimate $10.6B in US net sales by 2022, which implies roughly $13.4B in global net GLP-1 sales

Oral semaglutide (Ozempic) could offer further upside for NVO, in part at the expense of DPP-IV’s

NVO is developing an orally-administered (once-daily) form of semaglutide; phase 3 development should be complete by 4Q18. The highest doses of oral semaglutide (20mg, 40mg daily) show efficacy that is comparable to the highest dose of the injectable form (1mg weekly)[26]. However to be tolerated, the oral doses must be ramped much more gradually than the injectable. The injectable is dose-escalated from 0.25mg to 0.5mg after 4 weeks, and from 0.5mg to 1.0mg after an additional 4 weeks, taking 8 weeks to reach the full injected dose. When oral semaglutide is ramped at the same rate (starting at 5mg, moving to 10mg, then 20mg, then 40mg after 4 weeks on each dose), the adverse-event related quit rate was 26% as compared to 14% for the 1mg weekly injectable. Only when the oral is ramped with 8 weeks between dose escalations did patients reach the 40mg dose with an adverse-event quit rate (also 14%) on par with the injectable. Because of this it takes a full 24 weeks to reach the full oral dose on a tolerable dose-escalation schedule, as compared to just 8 weeks for the injectable

Because the oral requires more gradual dose escalation than the injectable, it may be less suitable for patients in need of more rapid intervention. However, because type 2 diabetic patients who have not yet begun self-injecting (either insulin or a GLP-1) tend to be more needle-phobic than patients who have experience with self-injection, it’s reasonable to expect that oral semaglutide would provide a much easier entry point to GLP-1’s for a significant number of patients. And, because oral semaglutide is a more effective (but less tolerable) means of impacting the same core (incretin) pathway than oral DPP-IV inhibitors, it’s also reasonable to expect that oral semaglutide might become a preferred alternative for DPP-IV patients who are poorly regulated

©2018, SSR, LLC, 225 High Ridge Rd, 2nd Floor, Stamford, CT 06905. All rights reserved. The information contained in this report has been obtained from sources believed to be reliable, and its accuracy and completeness is not guaranteed. No representation or warranty, express or implied, is made as to the fairness, accuracy, completeness or correctness of the information and opinions contained herein. The views and other information provided are subject to change without notice. This report is issued without regard to the specific investment objectives, financial situation or particular needs of any specific recipient and is not construed as a solicitation or an offer to buy or sell any securities or related financial instruments. Past performance is not necessarily a guide to future results. In the past 12 months, through a wholly-owned subsidiary SSR Health LLC has provided paid advisory services to BioPharmX (BPMX), Pfizer Inc (PFE), Gilead Sciences (GILD), Bristol-Myers Squibb (BMY) and Sanofi (SNY) on both securities-related and non-securities-related topics. One or more of SSR Health’s analysts owns long positions in the following stocks: ACOR, AGEN, AGIO, AKAO, ALKS, ALNY, ANAB, ARRY, BMY, DERM, DOVA, DPLO, ESALY, FOLD, GWPH, INCY, IONS, KALA, LOXO, NSTG, PFSCF, PGNX, PRTK, PTLA, RARE, RHHBY, RIGL, TBPH, THERF, TRVN, TSRO, TTPH, TXMD, VRTX

  1. NVO also appears overvalued according to our Hidden Pipeline method (see “Drug & Biotech Companies with Undervalued Pipelines”, SSR Health LLC, January 15, 2018); this may also reflect a scenario in which buy-side sales and earnings expectations are well above sell-side consensus 
  2. Adlyxin (SNY) was recently approved in the US, but appears to be non-competitive (daily injection, limited glycemic efficacy, non-inferior to placebo for CV risks (where Victoza as once daily reduces CV risk)) 
  3. AWARD-6 trial: http://www.thelancet.com/pdfs/journals/lancet/PIIS0140-6736(14)60976-4.pdf 
  4. Ibid 2 
  5. I.e., glycated hemoglobin or HbA1c, an index of the degree and duration of elevated blood glucose 
  6. SUSTAIN-7, http://www.thelancet.com/journals/landia/article/PIIS2213-8587(18)30024-X/fulltext 
  7. SUSTAIN-6; reduction in MACE, no CV mortality benefit; http://www.nejm.org/doi/pdf/10.1056/NEJMoa1607141 
  8. https://clinicaltrials.gov/ct2/show/NCT01394952 
  9. http://onlinelibrary.wiley.com/doi/10.1111/dom.13028/full 
  10. http://www.nejm.org/doi/pdf/10.1056/NEJMoa1607141 
  11. http://onlinelibrary.wiley.com/doi/10.1111/dom.13172/epdf 
  12. http://care.diabetesjournals.org/content/41/Supplement_1 
  13. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3322721/ 
  14. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4657234/ 
  15. http://onlinelibrary.wiley.com/doi/10.1111/dom.13172/full 
  16. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1770969/ 
  17. https://www.idf.org/our-activities/care-prevention/cardiovascular-disease/cvd-report.html 
  18. Glipizide, glimepride (both generic) 
  19. Repaglinide, nateglinide (both generic) 
  20. Lantus (SNY), Toujeo (SNY), Levemir (NVO), Tresiba (NVO), Basaglar (LLY) 
  21. Januvia / Janumet (MRK), Tradjenta / Jentadueto (Boehringer Ingelheim, LLY), Onglyza / Kombiglyze (AZN) 
  22. Invokana / Invokamet (JNJ), Jardiance (Boehringer Ingelheim, LLY), Farxiga / Xigduo (AZN), Steglatro / Steglujan (PFE/MRK) 
  23. Pioglitazone, rosiglitazone (both generic) 
  24. Byetta (AZN), Bydureon (AZN), Tanzeum (GSK), Adlyxin (SNY) 
  25. The current ratio is roughly 75% US, 25% non-US; holding all else (e.g. unit growth) constant, at 5% US inflation and 0% ex-US price inflation, the US becomes 79% of sales after 5 years of inflation 
  26. https://jamanetwork.com/journals/jama/article-abstract/2657376?redirect=true 
Print Friendly