Disease Analysis: Type 2 Diabetes

Latest key takeaways

• The type 2 diabetes market is dynamic and complex. In the non-insulin segment, while older generic drugs are widely used, especially first-line metformin in the US and EU, the more novel branded drugs are used in a sizable minority of patients, generating over $30bn in sales in 2020. The DPP-IV inhibitors have been the most commonly used of these, due to their safety and tolerability, albeit with intermediate efficacy, but in the important US market, GLP-1 receptor agonists and SGLT-2 inhibitors have overtaken them in usage. The increased usage of GLP-1 receptor agonists and SGLT-2 inhibitors has been spurred by recommendations starting in a 2018 ADA/EASD consensus report for patients with relevant co-morbidities due to benefits on major adverse cardiovascular events (MACE) and, for the SGLT-2 inhibitors in particular, heart failure (HF) and renal impairment, along with related label expansions. This is expected to continue to grow these classes, though clinical inertia still needs to be overcome among a number of physicians, and intense pricing competition, with increasing rebates in the US, is typical across the indication. However, loss of exclusivity will start to take a toll on all of these classes over the next 10 years, particularly starting with patent expirations for the leading DPP-IV inhibitor, Januvia, in 2022.
• The SGLT-2 inhibitors have started to bolster their growth with data and label additions for pivotal HF and chronic kidney disease (CKD) trials, including in non-diabetics. Jardiance has taken the lead in the segment, with a strong reduction in cardiovascular death in its diabetes cardiovascular outcomes trial (CVOT), after Invokana faltered due to concerns about an increased risk of amputations. However, in pivotal HF trials, Farxiga showed a benefit on CV death, whereas Jardiance, whose study included more severe patients, did not. This could bolster the view that outcomes differences seen between the drugs are just due to differences in the studies. Jardiance also showed fairly positive data in a pivotal trial of heart failure with preserved ejection fraction (HFpEF), in both those with and without diabetes, but again without improving CV death, so it will be interesting to see how Farxiga fares on this in its HFpEF trial. While this is an area of high unmet need, some have suggested physicians need to be selective in deciding who to treat, though in diabetics, it would be easier to add the drug, given other reasons to use it as well. Side effects, including increased risk of genitourinary infections, and safety concerns have generally held the class back somewhat. Competition with Rybelsus (see below) and patent expirations starting in 2025 will also impact the class.
• Sales of the injectable GLP-1 receptor agonists have grown substantially with the addition of a more convenient weekly formulation, Eli Lilly’s Trulicity, which expanded usage of the segment rather than displacing Novo Nordisk’s daily Victoza, which was the first in the class to show a CV benefit. However, Novo Nordisk’s weekly Ozempic, introduced in 2018, is more effective, and has helped the franchise stay ahead in sales. It should also help to defend against loss of patent protection for Victoza in 2022–23 in major markets. The battle is continuing with higher doses of both drugs (Trulicity’s have been approved in the US), new CV indications, and Eli Lilly’s GIP co-agonist, tirzepatide, which will be important for the company as Trulicity loses patent protection in 2027–29 in major markets. Phase III results for tirzepatide showed better glycemic control and weight loss than the approved dose of Ozempic, albeit with some tradeoff in side effects, and suggest the high doses may still have modest advantages over Ozempic 2mg and possibly the 2.4mg dose approved in obesity, though the comparison is difficult given different weight loss in obese diabetics versus non-obese. While the GLP-1 receptor agonists cause weight loss, their main side effect is nausea, and as injectables they tend to be used after oral drugs have failed. However, an oral formulation of Ozempic – Rybelsus – approved in 2019, appears to be the most effective branded oral drug and is poised to disrupt the segment, though so far it is mostly expanding usage earlier on in competition with other oral classes. Interestingly, a Phase III study of an even higher dose of the drug started in early 2021.
• In Japan, DPP-IV inhibitors are used more commonly at first line, and guidelines are somewhat equivocal about the implications of the CV findings of the other classes for Japanese patients. Nevertheless, SGLT-2 inhibitors have grown substantially, with a lesser increase for GLP-1 receptor agonists, due to a strong preference for oral drugs.
• In the basal insulin segment, with the advent of biosimilars/follow-on copies to segment leader Sanofi’s Lantus, Sanofi and Novo Nordisk have introduced next-generation, longer-acting products – Toujeo and Tresiba, respectively. However, due to continued pricing pressures, especially in the US, these have not prevented overall sales declines, which have been particularly harsh for Sanofi’s basal insulin franchises, with a 34% drop in combined worldwide sales over the past few years despite some stability ex-US. In relation to this, Sanofi has stopped investing in diabetes research. Prescriptions of Lantus and Toujeo have stabilized somewhat over the past few years in the US, but the introduction of another follow-on copy, Semglee, in 2020 could introduce further pricing pressures, especially as Semglee has also been filed for interchangeability. Novo Nordisk’s earlier basal analog, Levemir, was not as popular as Lantus, as it is not as long-lasting and so requires twice-daily dosing in more patients. However, its successor product, Tresiba, has an even longer-acting duration than Toujeo, allowing a broader flexible dosing window, where it can be administered at different times each day. While both Toujeo and Tresiba have shown some evidence for a hypoglycemia benefit over Lantus, the data have been somewhat stronger for Tresiba, which also showed a reduction in severe hypoglycemia in a large outcomes study in type 2 diabetes, with data appearing on its US label. Nevertheless, while Tresiba appears more differentiated than Toujeo, it has been difficult to discern whether there are substantial differences in clinical benefits between the two, because each sponsor has conducted head-to-head PK/PD and longer-term hypoglycemia studies, but those studies have either shown equivalence or an advantage for the sponsor’s drug. While Tresiba has helped to increase combined usage of the Novo Nordisk franchises, it has come at the expense of increasing rebates in the US, so in the past few years, despite combined sales growth ex-US, worldwide sales have declined by 25%.
• Novo Nordisk’s insulin icodec, the most advanced weekly insulin in development, started Phase III trials in late 2020. Phase II results suggest it could be a useful alternative to daily basal insulin, and while initial data did show higher rates of hypoglycemia, other studies indicate that this can be mitigated with a less aggressive titration schedule. However, there is still a question as to whether hypoglycemia may be an issue for some patients, especially outside of the clinical trial setting. Concerns about hypoglycemia have prevented some manufacturers from entering the space, though weekly insulins could be especially useful in fixed-ratio combinations (FRCs) with weekly GLP-1 receptor agonists.
• Fast-acting/bolus insulins for mealtime administration are not as routinely used in type 2 diabetes as they are in type 1, and dynamics are somewhat different in general in the fast-acting/bolus segment of the market due to more exclusive contracting in the US which has made it more difficult for biosimilar/follow-on copies to gain a foothold, though they still contribute to pricing pressure. It has also been more difficult to differentiate next-generation, faster-acting products, though they can be injected up to 20 minutes after the start of a meal rather than just around the start. In response to the introduction of Sanofi’s Admelog, a biosimilar/follow-on copy to Eli Lilly’s Humalog, which along with Novo Nordisk’s NovoLog leads the segment, both Eli Lilly and Novo Nordisk launched their own “authorized follow-ons/generics” with 50% lower list prices in 2019–20, in order to increase competitiveness in plans where patients can end up having to pay a portion of a drug’s gross cost. Net price decreases have driven appreciable US sales declines for NovoLog and Humalog, leading to modest declines worldwide. Admelog has not been able to penetrate the US market much outside of Medicaid, but Viatris (Mylan) and Sanofi are pursuing US approvals as interchangeable insulins for their biosimilars of NovoLog, so it will be interesting to see whether that makes a difference. The more recent ultra-rapid-acting formulations have relatively marginal benefits over their predecessors, and Novo Nordisk’s Fiasp only has moderate sales, though it is still early in its ramp. Eli Lilly’s Lyumjev launched in mid-2020 in major markets, though its data are somewhat weaker than Fiasp’s, and expectations are limited.
• For the insulins space in general, while companies are looking into glucose-sensitive insulins with a lower risk of hypoglycemia, it is still uncertain how promising these will be.
• Improving insulin sensitivity is a mechanistic unmet need, but pipeline candidates so far have not appeared very compelling, though Twymeeg has shown intermediate efficacy in Japanese patients.
• Given limited budgets for many patients, there could be indirect competition with novel dyslipidemia drugs.
• The overall likelihood of approval of a Phase I type 2 diabetes asset is 6.9%, and the average probability a drug advances from Phase III is 66.7%. Type 2 diabetes drugs, on average, take 9.1 years from Phase I to approval, compared to 9.8 years in the overall endocrine space.