Efficacy and Safety of Tirzepatide in Adults With Type 2 Diabetes: A Perspective for Primary Care Providers

The aim of the SURPASS clinical trial program was to assess the efficacy and safety of three doses of tirzepatide (5, 10, and 15 mg) once weekly in participants with type 2 diabetes. Based on the findings of the phase 2 studies, the dose escalation scheme in the SURPASS trials was as follows: all doses started at 2.5 mg once weekly for 4 weeks; then each dose was escalated every 4 weeks in increments of 2.5 mg until the corresponding maintenance dose of tirzepatide (5, 10, or 15 mg) was attained and maintained for the remainder of the study (Figure 1).

Tirzepatide was compared with placebo as monotherapy in SURPASS-1 (NCT03954834) during 40 weeks of treatment. Tirzepatide was also assessed versus placebo or active comparators as add-on to different background antihyperglycemic medications: versus once-weekly GLP-1 receptor agonist semaglutide 1 mg as add-on to metformin for 40 weeks in SURPASS-2 (NCT03987919); versus once-daily insulin degludec as add-on to metformin with or without a sodium–glucose cotransporter 2 (SGLT2) inhibitor for 52 weeks in SURPASS-3 (NCT03882970); versus once-daily insulin glargine as add-on to any combination of metformin, sulfonylurea, and SGLT2 inhibitor for 52 weeks plus a variable treatment period of up to 52 weeks in SURPASS-4 (NCT03730662); and versus placebo as add-on to insulin glargine with or without metformin for 40 weeks in SURPASS-5 (NCT04039503). All five studies were randomized trials with a 1:1:1:1 ratio for tirzepatide 5, 10, or 15 mg, or placebo/active comparator, except SURPASS-4, which had a 1:1:1:3 ratio. SURPASS-1 and SURPASS-5 were double-blind trials, whereas SURPASS-2, SURPASS-3, and SURPASS-4 were open-label trials.

The primary efficacy end point for these studies was the change from baseline in A1C. Secondary efficacy end points included the proportion of participants with an A1C <7.0% (53 mmol/mol), ≤6.5% (48 mmol/mol), and <5.7% (39 mmol/mol); mean change from baseline in body weight; and proportion of participants reaching ≥5%, ≥10%, and ≥15% weight loss.

The numbers of participants randomized and included in the modified intention-to-treat population (i.e., all participants who received at least one dose of study drug) in the placebo-controlled trials (SURPASS-1 and SURPASS-5) were 478 and 475, respectively, and 1,878, 1,437, and 1,995 in the active-controlled trials (SURPASS-2, SURPASS-3, and SURPASS-4, respectively). The mean age of participants ranged from 54 years in SURPASS-1 to 64 years in SURPASS-4 (Table 1). The number of males/females was balanced across the five trials. Because of the global nature of the SURPASS program and the countries participating in it, the majority of participants in each of the five studies were White, and the proportions of Black or African American participants were 1−5%. In the U.S. trial sites, the proportion of Black or African American participants was 13−22%. In SURPASS-1, the proportion of White participants was relatively lower (36%), whereas the proportions of Asian (35%) and American Indian or Alaska Native (25%) participants were relatively higher compared with the other SURPASS studies. The reported Hispanic or Latino population ranged from 5% in SURPASS-5 to 70% in SURPASS-2.

The mean A1C concentration at baseline ranged from 7.9% in SURPASS-1, with a relatively shorter mean duration of type 2 diabetes (4.7 years) in this younger population, to 8.5% in SURPASS-4, with a relatively longer mean duration of the disease (11.8 years) in participants with increased cardiovascular (CV) risk (87% with a history of CV disease). The overall population of SURPASS-1 had the lowest BMI (31.9 kg/m²) and body weight (85.9 kg) at baseline, whereas these parameters were slightly higher in the rest of the SURPASS studies (33−34 kg/m² and 90−95 kg).

Efficacy data presented herein are based on the efficacy estimand, which represents on-treatment efficacy, including all randomized participants who continued to receive the study drug without rescue medication. A summary of the efficacy results across the completed SURPASS trials is presented in Table 2 and Figure 2A and B.

The A1C reduction from baseline to the primary end point ranged from −1.87% with tirzepatide 5 mg in SURPASS-1 to −2.59% with tirzepatide 10 and 15 mg in SURPASS-5, considering the different participant populations, background antihyperglycemic medications, and durations of the studies. The A1C reductions were significantly greater in each SURPASS trial for all tirzepatide doses compared with placebo (SURPASS-1 and SURPASS-5) or active comparators semaglutide 1 mg (SURPASS-2) or basal insulin (SURPASS-3 and SURPASS-4) at the primary end point (Table 2 and Figure 2A). This effect was sustained in SURPASS-4 through the end of the variable treatment period (up to week 104) (25).

The American Diabetes Association (ADA) recommends an A1C target of ≤7% (53 mmol/mol) for most nonpregnant adults with type 2 diabetes and enough life expectancy to observe beneficial effects on microvascular outcomes (∼10 years) (1,27). Glycemic values must be personalized based on patient preference, risk of hypoglycemia and other AEs of therapy, and patient comorbidities.

The proportion of participants achieving an A1C value <7.0%, ≤6.5%, or <5.7% with tirzepatide across the completed SURPASS trials ranged from 81 to 97%, from 66 to 95%, and from 23 to 62%, respectively (Table 2). Significantly more participants achieved the three A1C values with all tirzepatide doses in each of the five SURPASS studies compared with placebo or active comparators at the primary end point.

Body weight loss in the treatment of type 2 diabetes is paramount, as excess body weight and obesity are linked to CV health and comorbidities. In addition, weight gain can discourage patients from initiating therapy. Therefore, broadening the treatment continuum with therapies that achieve glycemic control with weight loss effects is important. In the SURPASS trials, body weight reductions from baseline at the primary end point ranged from −6.2 kg (−6.6%) with tirzepatide 5 mg when added to insulin glargine (SURPASS-5) to −12.9 kg (−13.9%) with tirzepatide 15 mg when added to metformin with or without an SGLT2 inhibitor (SURPASS-3) (Table 2 and Figure 2B). In each of the SURPASS clinical trials, body weight reductions were significantly greater for all tirzepatide doses compared with placebo or the active comparators, including semaglutide 1 mg (SURPASS-2). In fact, weight gain was observed with basal insulin (SURPASS-3 and SURPASS-4) and with placebo in combination with insulin glargine (SURPASS-5). Tirzepatide-induced body weight loss effect was gradual up to the primary end point (40 or 52 weeks) across all SURPASS trials and was sustained through the end of the variable treatment period (up to week 104) in SURPASS-4 (25). Because SGLT2 inhibitors have favorable effects on weight, the concomitant use of an SGLT2 inhibitor with tirzepatide in SURPASS-3 may raise the question of whether tirzepatide was solely responsible for the greatest extent of weight loss observed in the SURPASS program. Subgroup analyses in SURPASS-3 showed that the reductions in body weight were similar regardless of the concomitant use of an SGLT2 inhibitor (24).

According to ADA guidelines, in patients with type 2 diabetes, a 5% weight loss is needed to achieve beneficial effects on glucose, lipids, and blood pressure levels, and these benefits can be maximized with more intensive weight loss of 15% (28). The proportion of participants achieving a weight loss of ≥5%, ≥10%, or ≥15% with tirzepatide across the completed SURPASS trials ranged from 54 to 88%, from 23 to 69%, and from 7 to 43%, respectively (Table 2). Significantly more participants achieved the weight loss of ≥5%, ≥10%, or ≥15% with all tirzepatide doses in each of the five SURPASS studies compared with placebo or active comparators at the primary end point.

There was an overall improvement in the fasting lipid profile with tirzepatide across the SURPASS trials. Namely, there was a decrease from baseline in triglycerides (by 15.2−26.7%), LDL cholesterol (by 5.2−15.5%), and VLDL cholesterol (by 14.2−25.2%) and an increase from baseline in HDL cholesterol (by 0.9−10.8%) with all tirzepatide doses at the primary end point. Decreases in systolic (of 2.8−12.6 mmHg) and diastolic (of 0.8−4.5 mmHg) blood pressure were also observed with tirzepatide in all SURPASS trials. Moreover, decreases in ALT and AST were observed with tirzepatide across the trials in which these measures were reported (SURPASS-1, SURPASS-2, SURPASS-3, and SURPASS-5). Furthermore, a specific MRI substudy was performed in the SURPASS-3 trial, and the results showed significant improvements in liver fat content and abdominal adipose tissue with all doses of tirzepatide in a subpopulation with type 2 diabetes and higher risk of elevated liver fat content (29). All of this evidence supports the potential role of tirzepatide in fatty liver disease and chronic weight management, which will be further explored in studies specifically designed for that purpose (e.g., the SYNERGY-NASH trial [NCT04166773] and the SURMOUNT program, respectively).

In SURPASS-5, insulin glargine 100 units/mL was administered as background antihyperglycemic medication once daily after an initial 4-week insulin stabilization period and a 36-week insulin titration period (26). During the insulin stabilization period, the dose of insulin glargine was unchanged except for safety reasons and additionally reduced by 20% in participants with an A1C ≤8.0% at randomization to minimize the risk of hypoglycemia. During the insulin titration period, the dose of insulin glargine was self-adjusted to a target fasting blood glucose of <100 mg/dL following a treat-to-target algorithm. Further details on the insulin titration algorithm used in SURPASS-5 can be found in Supplementary Table S1. At the primary end point (week 40), the mean change from baseline in insulin glargine dose was 4.4 units for tirzepatide 5 mg, 2.7 units for tirzepatide 10 mg, −3.8 units for tirzepatide 15 mg, and 25.1 units for placebo. This markedly lower use of insulin glargine in the tirzepatide arms versus placebo (P <0.001 for all tirzepatide doses in percentage change) was associated with the superior glycemic outcomes observed with tirzepatide.

In general, the proportion of participants reporting AEs was similar across all treatment groups in the five SURPASS trials (Table 3). The incidence of serious AEs was <20% across all treatment arms in all of the trials. There were no deaths in SURPASS-5, one (<1%) in SURPASS-1, five (<1%) in SURPASS-3, 13 (1%) in SURPASS-2, and 60 (3%) in SURPASS-4. None of the deaths were considered by the investigators to be related to the study treatment.

Overall, treatment discontinuation because of AEs was low in all SURPASS trials and more common in the tirzepatide groups (3−11%) versus placebo (3%; SURPASS-1 and SURPASS-5), versus semaglutide 1 mg (4%; SURPASS-2), and versus basal insulin (1% and 5%; SURPASS-3 and SURPASS-4, respectively). GI AEs were the most common reason for treatment discontinuation in tirzepatide-treated participants.

Nausea, diarrhea, decreased appetite, and vomiting were the most frequent treatment-emergent AEs in the tirzepatide groups across all SURPASS trials and in the semaglutide 1 mg group in SURPASS-2 (Table 3). The incidence of these GI AEs throughout each study period was higher in the tirzepatide groups (12−24% for nausea, 12−22% for diarrhea, 4−14% for decreased appetite, and 2−13% for vomiting) compared with placebo (SURPASS-1 and SURPASS-5) and basal insulin (SURPASS-3 and SURPASS-4), and similar in tirzepatide and semaglutide 1 mg (SURPASS-2). Most cases of nausea, diarrhea, and vomiting were mild to moderate in severity, with the highest incidence observed during the dose escalation period and then decreasing over time. As an example, the incidence of nausea over 40 weeks in the SURPASS-2 trial showed a comparable profile for tirzepatide and semaglutide 1 mg (Figure 3). Further details on the incidence of nausea, diarrhea, and vomiting over time for tirzepatide are published in the literature (22–26). The optimized slow dose escalation scheme used in the phase 3 SURPASS program improved the GI tolerability of tirzepatide compared with the phase 2 trials. The GI tolerability of tirzepatide was similar to that of selective GLP-1 receptor agonists.

There were no cases of severe or clinically significant hypoglycemia (blood glucose <54 mg/dL) in the tirzepatide groups in SURPASS-1 (monotherapy) and 0 or 1% with placebo, respectively (Table 3). A very low proportion of tirzepatide-treated participants reported severe (0−1%) or clinically significant hypoglycemia (<1−9%) in SURPASS-2, SURPASS-3, and SURPASS-4 (add-on to oral antihyperglycemic medications). The hypoglycemia incidence was numerically higher in SURPASS-5, as expected, as the study treatment was administered in combination with basal insulin. However, no differences were noted in severe or clinically significant hypoglycemia between the tirzepatide and placebo groups. Therefore, no increased risk of hypoglycemia was associated with tirzepatide versus comparators in any of the SURPASS trials. These findings are consistent with those for drugs in the GLP-1 receptor agonist class, which have an inherently low propensity to induce hypoglycemia because of their glucose-dependent mechanism of action. An ongoing phase 1 trial (NCT04050553) is investigating the effect of tirzepatide on the counterregulatory response to hypoglycemia.

No cases of treatment-emergent diabetic retinopathy were reported in SURPASS-1 or SURPASS-5, and very few cases were reported in SURPASS-2 (<1% with tirzepatide 10 mg), SURPASS-3 (1% with tirzepatide 5 mg and <1% with tirzepatide 15 mg), and SURPASS-4 (<1% in all treatment groups) (Supplementary Table S2). Of note, all SURPASS trials excluded participants with a history of proliferative diabetic retinopathy, diabetic maculopathy, or nonproliferative diabetic retinopathy requiring acute treatment. A substudy of the ongoing SURPASS-CVOT trial (NCT04255433) will investigate diabetic retinopathy progression on tirzepatide in a subpopulation with or at increased risk for diabetic retinopathy.

Very few cases of malignant neoplasms were reported in SURPASS-3 (up to 1%) and SURPASS-5 (up to 2%) across all treatment groups (24,26). No cases of medullary thyroid cancer were reported in any of the five SURPASS trials (22–26). There were no clinically relevant changes from baseline in mean calcitonin levels during any of the SURPASS trials.

One case (1%) of pancreatic cancer was reported in the tirzepatide 5 mg group in SURPASS-1. No cases of adjudicated pancreatitis were reported in SURPASS-1, SURPASS-3, or SURPASS-5, whereas it was reported in up to 1% of participants in SURPASS-2 and SURPASS-4 across all treatment groups (Supplementary Table S2). None of these cases were serious in SURPASS-2. It should be considered that participants with a history of pancreatitis were excluded from the five SURPASS trials.

Heart rate increases of up to 5.6 bpm were observed with tirzepatide at the primary end point, which is consistent with those seen with GLP-1 receptor agonists. To assess the CV safety of tirzepatide, a meta-analysis was performed following the FDA and the EMA guidance (30–32). Pooled data from the five SURPASS trials reviewed herein were included in the CV safety meta-analysis, in addition to data from one phase 2 trial and one regional phase 3 trial in Japan. Results from the meta-analysis indicate that tirzepatide is not associated with an increased risk of major CV events in people with type 2 diabetes (33). The hazard ratio comparing tirzepatide versus comparators was 0.80 (95% CI 0.57–1.11) for confirmed four-component major adverse CV events, including CV death, myocardial infarction, stroke, and hospitalization for unstable angina. The ongoing long-term SURPASS-CVOT trial (NCT04255433) will further assess CV safety and efficacy of tirzepatide compared with the selective GLP-1 receptor agonist dulaglutide 1.5 mg, providing a more comprehensive evaluation of the potential cardioprotective effects. This is the first CV outcomes trial comparing the CV efficacy of a novel antihyperglycemic medication for type 2 diabetes with an active comparator with proven, product-labeled CV efficacy.

Injection site reactions and hypersensitivity reactions occurred in <1–7% and 1–7%, respectively, of tirzepatide-treated participants across all SURPASS trials (Supplementary Table S2). No severe cases of hypersensitivity and injection site reactions were reported (22–24,26). Overall, the samples of tirzepatide-treated participants with treatment-emergent antidrug antibodies detected had similar pharmacokinetics and efficacy to the samples of participants with treatment-emergent antidrug antibodies not detected (22,24,25).

An important treatment element when considering injectable treatment therapy is the device that delivers the medication. Research has demonstrated that improvements in injection delivery systems have a positive outcome in terms of patients’ treatment acceptability, satisfaction, and convenience and may help decrease social stigma associated with injection therapy (34–38). Tirzepatide is administered with a single-dose pen and uses the same injection device that is used to administer dulaglutide. In a study that evaluated the safe and effective use of the dulaglutide single-dose pen, most patients (99%) found the device “easy” or “very easy” to use (39). Similarly, although no study has been conducted to assess patient preference for the tirzepatide pen, there has been research on the same device published in the literature. In the crossover PREFER study, injection-naive participants with type 2 diabetes preferred the single-dose pen to the semaglutide pen (84.2 vs. 12.3%) (40). More participants found the single-dose pen easier to use (86.8 vs. 6.8%) and were willing to use it compared with the semaglutide pen (93.5 vs. 45.8%). Shorter training time was required for participants to learn to use the single-dose pen device versus the semaglutide device (3.38 vs. 8.14 minutes). Because no active drug was administered directly into patients in this study and the device procedures are not affected by the drug dosage, these findings could be extrapolated to anticipate the patient preference between the tirzepatide pen and the semaglutide pen. Therefore, given the consistency of the single-use pen device across dulaglutide and tirzepatide and across doses, patients’ perceptions regarding device ease of use and preferences should remain the same (39,40).

Patient preferences are an important element of patient-centered care, the provision of which is essential to effective diabetes management and a cornerstone of diabetes treatment guidelines around the world. Past ADA/European Association for the Study of Diabetes guidelines have noted that, “Even in cases where clinical characteristics suggest the use of a particular medication based on the available evidence from clinical trials, patient preferences regarding route of administration, injection devices, side effects, or cost may prevent their use by some individuals” (1). Thus, ease of use and patient preference of medical devices should be considered when physicians are evaluating different therapeutic options for their patients.

Based on the clinically and statistically significant A1C reduction and weight loss with low risk of hypoglycemia versus comparators, tirzepatide can be regarded as a potential treatment choice for a broad spectrum of patients with type 2 diabetes. With current treatment, normoglycemia (A1C <5.7%) without an increased risk of hypoglycemia has not been considered attainable. The SURPASS program included patient populations that are representative of the type 2 diabetes treatment continuum observed in clinical practice, although because of the global nature of the SURPASS program, the data might not be representative of local demographics.

The range of background antihyperglycemic medications included in the SURPASS program covers relevant oral (metformin, SGLT2 inhibitors, and sulfonylureas) and injectable (basal insulins) antidiabetic drugs commonly used in primary care. The addition of tirzepatide to a basal insulin (SURPASS-5) resulted in improved glycemic control and weight loss without an increased risk of hypoglycemia and reduction in insulin dosage. It is anticipated that a more stringent insulin down-titration may result in substantial insulin sparing while maintaining glycemic control and potentially reducing hypoglycemia risk.

The choice of active comparators and dosing in the SURPASS trials has been driven by the importance these agents have and will have in clinical care. For example, semaglutide 1 mg (used in SURPASS-2) was considered a potent GLP-1 receptor agonist available on the market at the time the study was conducted (41,42). On the other hand, basal insulins degludec and glargine are widely prescribed in patient populations inadequately controlled on oral antihyperglycemic medications and followed an adequate dose titration as demonstrated by the glycemic efficacy results achieved in SURPASS-3 and SURPASS-4, respectively (24,25).

The AE profile of tirzepatide is similar to that of the GLP-1 receptor agonists. Although GI AEs were the most commonly observed AEs with tirzepatide, they affected a minority of patients with mild to moderate severity and tended to occur during treatment initiation. GI AEs can be managed with mitigating measures such as eating smaller and more frequent meals throughout the day, stopping eating when feeling full, or taking prescribed symptomatic medications per physician advice. Patients should be informed that GI AEs generally dissipate over time.

It must be noted that the durations of the completed SURPASS trials were 40 to 52 weeks, although the efficacy and safety of tirzepatide has been shown to be sustained for up to 2 years in ∼10% of participants in SURPASS-4. In any case, longer-term and real-world data will be necessary for an enhanced understanding of the clinical profile of tirzepatide in primary care.

Based on pharmacokinetic studies of tirzepatide in people with varying degrees of renal or hepatic impairment, no dose adjustment is anticipated in this patient population (43,44).

Finally, a ready-to-use single-dose pen can help to make treatment with tirzepatide an attractive option for patients with type 2 diabetes who are naive to self-injection or injecting others (39). All these factors need to be considered by clinicians in combination with the drug’s efficacy, safety, and other important treatment attributes.

Tirzepatide offers primary care providers and patients a first-in-class option to treat a wide spectrum of type 2 diabetes by achieving significant reductions in A1C (ranging from −1.87 to −2.59%) while also achieving substantial weight loss (ranging from −6.6 to −13.9%) and maintaining a low risk of hypoglycemia. The safety profile of tirzepatide is similar to that of GLP-1 receptor agonists, with GI side effects being the most common AEs. These GI AEs are transient and mild to moderate in severity, and their frequency decreases over time.

This article was supported by Eli Lilly and Company.

P.K. has received speaker’s bureau and advisory board honoraria from Abbott Diabetes Care, AstraZeneca, Bayer U.S., Boehringer Ingelheim, Eli Lilly and Company, Janssen, and Novo Nordisk. J.E.A. reports consulting, advisory board, and speaker fees from Abbott Diabetes Care, Alfasigma, AstraZeneca, Bayer, Boehringer Ingelheim, Eli Lilly and Company, Intuity, Janssen, Merck, Novo Nordisk, and Sanofi. J.S. has received advisory board honoraria and speaker fees from Amgen, AstraZeneca, Bayer Vital, Berlin-Chemie, Boehringer Ingelheim, Bristol-Myers Squibb, Dexcom, Eli Lilly and Company, Janssen-Cilag, LifeScan, Novartis, Novo Nordisk, Pfizer, Roche Diagnostics, and Sanofi. K.S.B., K.R., A.T.-G., and J.A.L. are employees of and stock shareholders in Eli Lilly and Company.

P.K., J.E.A., J.S., K.S.B., K.R., and J.A.L. interpreted the data and critically revised the manuscript for important intellectual content. K.S.B., K.R., A.T.-G., and J.A.L. contributed to the conception of the article. K.S.B., A.T.-G., and J.A.L. researched data and wrote the manuscript. All of the authors approved of this manuscript to be submitted for publication. J.A.L. is the guarantor of this work and, as such, had full access to all the data reported and takes responsibility for the integrity of the data and the accuracy of the data analysis.