Tirzepatide vs Insulin Lispro Added to Basal Insulin in Type 2 Diabetes: The SURPASS-6 Randomized Clinical Trial

Abstract

Importance: Tirzepatide is a glucose-dependent insulinotropic polypeptide and glucagon-like peptide-1 receptor agonist used for the treatment of type 2 diabetes. Efficacy and safety of adding tirzepatide vs prandial insulin to treatment in patients with inadequate glycemic control with basal insulin have not been described.

Objective: To assess the efficacy and safety of tirzepatide vs insulin lispro as an adjunctive therapy to insulin glargine.

Design, setting, and participants: This open-label, phase 3b clinical trial was conducted at 135 sites in 15 countries (participants enrolled from October 19, 2020, to November 1, 2022) in 1428 adults with type 2 diabetes taking basal insulin.

Interventions: Participants were randomized (in a 1:1:1:3 ratio) to receive once-weekly subcutaneous injections of tirzepatide (5 mg [n = 243], 10 mg [n = 238], or 15 mg [n = 236]) or prandial thrice-daily insulin lispro (n = 708).

Main outcomes and measures: Outcomes included noninferiority of tirzepatide (pooled cohort) vs insulin lispro, both in addition to insulin glargine, in HbA1c change from baseline at week 52 (noninferiority margin, 0.3%). Key secondary end points included change in body weight and percentage of participants achieving hemoglobin A1c (HbA1c) target of less than 7.0%.

Results: Among 1428 randomized participants (824 [57.7%] women; mean [SD] age, 58.8 [9.7] years; mean [SD] HbA1c, 8.8% [1.0%]), 1304 (91.3%) completed the trial. At week 52, estimated mean change from baseline in HbA1c with tirzepatide (pooled cohort) was -2.1% vs -1.1% with insulin lispro, resulting in mean HbA1c levels of 6.7% vs 7.7% (estimated treatment difference, -0.98% [95% CI, -1.17% to -0.79%]; P < .001); results met noninferiority criteria and statistical superiority was achieved. Estimated mean change from baseline in body weight was -9.0 kg with tirzepatide and 3.2 kg with insulin lispro (estimated treatment difference, -12.2 kg [95% CI, -13.4 to -10.9]). The percentage of participants reaching HbA1c less than 7.0% was 68% (483 of 716) with tirzepatide and 36% (256 of 708) with insulin lispro (odds ratio, 4.2 [95% CI, 3.2-5.5]). The most common adverse events with tirzepatide were mild to moderate gastrointestinal symptoms (nausea: 14%-26%; diarrhea: 11%-15%; vomiting: 5%-13%). Hypoglycemia event rates (blood glucose level <54 mg/dL or severe hypoglycemia) were 0.4 events per patient-year with tirzepatide (pooled) and 4.4 events per patient-year with insulin lispro.

Conclusions and relevance: In people with inadequately controlled type 2 diabetes treated with basal insulin, weekly tirzepatide compared with prandial insulin as an additional treatment with insulin glargine demonstrated reductions in HbA1c and body weight with less hypoglycemia.

Trial registration: ClinicalTrials.gov Identifier: NCT04537923.

Figure 1.. Flow of Participants in the SURPASS-6 Trial

See Table 3 for the details of the adverse events that led to study treatment discontinuation. ^aIncludes 1 participant each who did not meet baseline age criteria, had chronic or acute pancreatitis or hepatitis, receiving drugs that directly affect gastrointestinal motility, or have a known clinically significant gastric emptying abnormality, have history of history of diabetic ketoacidosis or hyperosmolar state/coma, have family or personal history of medullary thyroid carcinoma or multiple endocrine neoplasia syndrome type 2, have evidence of an autoimmune abnormality, have a history of any other condition that may preclude the participant from following and completing the protocol, are pregnant or breastfeeding, have participated in a clinical study involving an investigational product, or have previously completed/discontinued from this study or any other study investigating tirzepatide. ^bStudy discontinuations before, after, or at the primary end point visit were included.

Figure 2.. Effect of Tirzepatide vs Insulin Lispro on Hemoglobin A_1c (HbA_1c) and Body Weight

A, Lines show the values for least-squares mean HbA_1c levels over time derived from a mixed-model repeated-measures analysis (efficacy estimand). To convert HbA_1c values to mmol/mol, use the equation 10.93 × HbA_1c) − 23.50. Boxplots represent observed HbA_1c actual values over time for randomized patients from the pooled tirzepatide and insulin lispro groups, irrespective of treatment discontinuation or initiation of rescue therapy, using the full analysis set. The middle lines within each box represent the median observed HbA_1c, the symbols in the boxes represent the mean observed HbA_1c, the top and bottom of the boxes represent the interquartile range, the whiskers extend to the most extreme observed values with 1.5 times the IQR of the nearer quartile, and the symbols beyond these points represent the observed values outside that range. B, Lines show least-squares mean change from baseline in body weight over time, derived from a mixed-model repeated measures analysis (efficacy estimand). Boxplots represent observed changes from baseline in body weight over time for randomized patients from the pooled tirzepatide and insulin lispro groups, irrespective of treatment discontinuation or initiation of rescue therapy, using the full analysis set. The middle lines within each box represent the median observed changes from baseline; the symbols in the boxes represent the mean observed percentage change; the box tops and bottoms represent the interquartile range; the whiskers extend to the most extreme observed values with 1.5 times the IQR of the nearer quartile; and the symbols beyond these points represent the observed values outside that range. More negative values indicate greater reductions. Due to the large y-axis scale and number of individual data points shown in panel B, some outliers are overlapping and not visualized within the boxplot.