Tresiba History and Development: How Insulin Degludec Changed Basal Insulin Therapy

The Problem That Drove Degludec's Development

Basal insulin therapy before 2010 carried a persistent clinical limitation: nocturnal hypoglycemia. Even insulin glargine (Lantus), approved in 2000, showed measurable glucose-lowering peaks despite its "peakless" marketing claim. A 2007 meta-analysis of 11 randomized trials (N=6,264) found that glargine reduced nocturnal hypoglycemia compared to NPH but did not eliminate it, with event rates of 4.0 to 6.9 episodes per patient-year depending on diabetes type 1.

Novo Nordisk's insulin research division had already produced insulin detemir (Levemir), approved in 2004, which used albumin binding to extend duration. Detemir required twice-daily dosing in many patients and had a duration ceiling near 20 hours. The company recognized that a truly ultra-long-acting insulin with a flat pharmacokinetic profile and duration well beyond 24 hours could solve two problems at once: reduce hypoglycemia risk and allow flexible injection timing for patients who struggle with rigid daily schedules. That dual mandate shaped the molecular engineering program that became degludec.

Molecular Design: Engineering the Multi-Hexamer Chain

Degludec's mechanism begins with a single structural modification. Novo Nordisk scientists removed the threonine at position B30 of human insulin and attached a 16-carbon fatty diacid (hexadecanedioic acid) to lysine at B29 via a glutamic acid spacer 2. This acylation strategy was distinct from detemir's C14 fatty acid attachment; the longer chain and diacid linker gave degludec fundamentally different self-association behavior.

In the pharmaceutical formulation (containing phenol and zinc), degludec exists as dihexamers. That part is unremarkable. What happens after subcutaneous injection is not. As phenol diffuses away from the injection depot, the dihexamers do not simply dissociate into monomers the way conventional insulins do. Instead, degludec's fatty diacid side chains drive the formation of long multi-hexamer chains, sometimes described as "strings of pearls" in early Novo Nordisk publications 2. These soluble multi-hexamer assemblies create a slow-release depot in the subcutaneous tissue. Zinc gradually diffuses out, hexamers peel off the ends of the chains, dissociate into monomers, and absorb into the bloodstream.

The result is a terminal half-life of approximately 25 hours, roughly twice that of glargine U100 (approximately 12 hours) 3. At steady state, the duration of glucose-lowering action exceeds 42 hours. A glucose clamp study published in The Journal of Clinical Pharmacology in 2013 showed that degludec's within-day variability in glucose-lowering effect was four times lower than glargine U100 (coefficient of variation: 20% vs. 82%) 3. That flatness is the pharmacokinetic signature that distinguishes degludec from every other basal insulin on the market.

Preclinical and Phase I Development (2004 to 2009)

Novo Nordisk filed its first patent applications covering the acylated insulin analog in 2004. Preclinical work in porcine models confirmed prolonged absorption kinetics from the subcutaneous depot, and the multi-hexamer chain formation was verified using small-angle X-ray scattering 2.

Phase I trials began in 2006. Early dose-finding work in healthy volunteers and subjects with type 1 diabetes demonstrated the predicted ultra-long action profile. A key proof-of-concept finding: when subjects were given degludec once daily for five days, the glucose infusion rate profile during euglycemic clamp studies was remarkably flat across the 24-hour dosing interval, with no discernible peak 4.

The phase I program also tested whether flexible dosing was feasible. In one study, degludec was injected with intervals alternating between 8 and 40 hours (a "forced flex" design), and it maintained glycemic control comparable to a fixed-time regimen 4. This flexible dosing concept became a core element of degludec's clinical and commercial identity.

The BEGIN Trial Program: Phase III (2009 to 2012)

Novo Nordisk launched its phase III program under the umbrella name BEGIN (Basal-bolus Type 1, Basal-bolus Type 2, Once Long, Flex, and others). The program enrolled over 11,000 patients across more than 40 countries.

BEGIN Basal-Bolus Type 1 (N=629) compared degludec to glargine U100, both combined with insulin aspart at meals in type 1 diabetes. At 52 weeks, A1C reduction was comparable (degludec: -0.40% vs. glargine: -0.39%), but confirmed nocturnal hypoglycemia rates were 25% lower with degludec (rate ratio 0.75 to 95% CI 0.59 to 0.96) 5.

BEGIN Once Long (N=1,030) tested degludec versus glargine in insulin-naive type 2 diabetes patients on metformin. A1C reductions were equivalent at 52 weeks (degludec: -1.06% vs. glargine: -1.19%), with 36% fewer confirmed nocturnal hypoglycemia episodes in the degludec group 6.

BEGIN Flex explored the forced-flexible dosing schedule in type 2 diabetes, showing non-inferiority for A1C reduction when patients injected degludec at varying times each day versus glargine at the same time daily 7. The clinical implication was direct: patients who could not inject at a consistent time, whether due to shift work, travel, or other unpredictable schedules, could use degludec without glycemic penalty.

Across the BEGIN program, the signal was consistent. Degludec matched glargine on A1C and either matched or beat it on hypoglycemia endpoints. The hypoglycemia advantage was most pronounced at night, when the flat pharmacokinetic profile had its greatest practical benefit.

Regulatory Path: Rejection, Revision, and Approval

Novo Nordisk submitted its first U.S. New Drug Application (NDA) for degludec in 2012. The FDA issued a Complete Response Letter (CRL) in February 2013, declining to approve the drug. The agency did not cite efficacy or safety concerns from the BEGIN trials. Instead, the FDA requested a dedicated cardiovascular outcomes trial (CVOT) before approval, consistent with the agency's 2008 guidance requiring cardiovascular safety data for new diabetes drugs issued after the rosiglitazone controversy 8.

This was a significant setback. Degludec had already been approved in the EU (January 2013) and Japan (early 2013) based on the BEGIN data, and Novo Nordisk was generating commercial revenue in those markets. The U.S. rejection forced the company to invest several hundred million dollars and three to four years in a large CVOT.

Novo Nordisk responded by launching the DEVOTE trial. Simultaneously, the company resubmitted a revised NDA incorporating additional safety analyses. The FDA approved Tresiba on September 25, 2015, before DEVOTE's completion, based on an updated assessment of the existing data plus interim DEVOTE analyses 9. The approval covered both U100 and U200 concentrations for type 1 and type 2 diabetes in adults.

DEVOTE: The Cardiovascular Outcomes Trial

DEVOTE (Degludec Versus Glargine: Cardiovascular Safety in T2DM) randomized 7,637 patients with type 2 diabetes at high cardiovascular risk to degludec or glargine U100, treated for a median of 1.99 years. The primary endpoint was time to first occurrence of major adverse cardiovascular events (MACE: cardiovascular death, nonfatal myocardial infarction, or nonfatal stroke) 10.

Results, published in the New England Journal of Medicine in June 2017: MACE occurred in 8.5% of degludec patients and 9.3% of glargine patients (hazard ratio 0.91 to 95% CI 0.78 to 1.06, P<0.001 for non-inferiority). Degludec met the non-inferiority threshold convincingly.

The hypoglycemia findings were the headline. Severe hypoglycemia was 40% lower with degludec versus glargine (HR 0.60 to 95% CI 0.48 to 0.76, P<0.001) 10. Severe nocturnal hypoglycemia was reduced by 53% (HR 0.47 to 95% CI 0.31 to 0.73). Dr. Steven Marso, the trial's lead investigator, stated at the 2017 American Diabetes Association Scientific Sessions: "The magnitude of the hypoglycemia reduction was larger than we anticipated and was consistent across all prespecified subgroups."

DEVOTE's severe hypoglycemia data had clinical weight beyond the primary MACE endpoint. A prespecified analysis showed that episodes of severe hypoglycemia were associated with increased subsequent cardiovascular death (HR 2.05) and all-cause mortality (HR 1.80) 10. This raised the question of whether reducing severe hypoglycemia could itself reduce cardiovascular mortality, a hypothesis that DEVOTE was not powered to prove but that its data supported mechanistically.

The U200 Formulation and Delivery Device Innovation

Tresiba U200 (200 units/mL) was approved alongside U100, making it the first basal insulin available in a 200 units/mL concentration in a prefilled pen. The U200 FlexTouch pen delivers up to 160 units in a single injection (versus 80 units for U100). For patients with type 2 diabetes requiring large daily doses, often those with significant insulin resistance, U200 reduced injection volume by half and allowed single-injection dosing where two injections might otherwise have been needed.

The dose counter on both the U100 and U200 pens displays units, not volume. This was a deliberate safety decision by Novo Nordisk to prevent dosing confusion between concentrations. The 2016 American Diabetes Association Standards of Medical Care in Diabetes specifically recommended that concentrated insulins use devices that cannot lead to dose conversion errors 11.

Post-Approval Evidence and Guideline Positioning

After launch, several real-world studies expanded degludec's evidence base. The ReFLecT study, a prospective observational trial across 99 sites in seven European countries (N=2,550), found that switching from other basal insulins to degludec reduced A1C by 0.3% and halved self-reported hypoglycemia rates at 12 months 12.

A head-to-head trial against glargine U300 (Toujeo), the CONCLUDE study (N=1,609), was published in 2020. CONCLUDE had a complex statistical history: the primary maintenance period comparison did not reach significance for overall symptomatic hypoglycemia, but a 2021 post-hoc analysis of the full trial period showed lower rates of severe and nocturnal hypoglycemia with degludec versus glargine U300 13.

The 2024 ADA Standards of Care in Diabetes lists insulin degludec alongside glargine U100, glargine U300, and detemir as basal insulin options. The ADA notes the longer duration of action and flexible dosing as distinguishing features. The 2023 AACE Consensus Statement on Type 2 Diabetes Management recommends degludec as a preferred basal insulin when hypoglycemia risk is a primary concern 14.

Patent Expiry and Biosimilar Horizon

Novo Nordisk's core patents on degludec began expiring in the U.S. in 2024, with additional formulation patents extending into 2027 in some jurisdictions. As of early 2026, the FDA has not yet approved a biosimilar or interchangeable insulin degludec product, though several manufacturers have filed abbreviated pathway applications. The FDA's biologics pathway for follow-on insulins (after the March 2020 transition of insulins from drugs to biologics) applies to degludec biosimilars.

Tresiba generated approximately $3.2 billion in global sales in 2024, according to Novo Nordisk's annual report. Biosimilar entry will likely reduce costs substantially, following the pattern seen with glargine biosimilars (Semglee, Rezvoglar) which reduced per-unit prices by 40 to 65% within two years of launch.

Timeline of Key Milestones

Degludec's development spanned more than a decade from initial patent filing to mature post-marketing evidence:

• 2004: First patent applications filed by Novo Nordisk
• 2006: Phase I trials begin in healthy volunteers
• 2009: BEGIN phase III program launches globally
• 2012: NDA submitted to FDA; EU approval granted (January 2013)
• 2013: FDA issues Complete Response Letter requesting CVOT
• 2013: DEVOTE trial begins enrollment
• 2015: FDA approves Tresiba (September 25)
• 2017: DEVOTE results published in NEJM
• 2020: CONCLUDE trial published (degludec vs. glargine U300)
• 2024: Core U.S. patents begin expiring