Tirzepatide and Cancer Risk in Individuals with and without Diabetes: A Systematic Review and Meta-Analysis
Article information
Abstract
Background
Data on the carcinogenic potential of tirzepatide from randomized controlled trials (RCTs) are limited. Furthermore, no meta-analysis has included all relevant RCTs to assess the cancer risk associated with tirzepatide.
Methods
RCTs involving patients receiving tirzepatide in the intervention arm and either a placebo or any active comparator in the control arm were searched through electronic databases. The primary outcome was the overall risk of any cancer, and secondary outcomes were the risks of specific types of cancer in the tirzepatide versus the control groups.
Results
Thirteen RCTs with 13,761 participants were analyzed. Over 26 to 72 weeks, the tirzepatide and pooled control groups had identical risks of any cancer (risk ratio, 0.78; 95% confidence interval, 0.53 to 1.16; P=0.22). The two groups had comparable cancer risks in patients with and without diabetes. In subgroup analyses, the risks were also similar in the tirzepatide versus placebo, insulin, and glucagon-like peptide-1 receptor agonist groups. The overall cancer risk was also comparable for different doses of tirzepatide compared to the control groups; only a 10-mg tirzepatide dose had a lower risk of any cancer than placebo. Furthermore, compared to the control groups (pooled or separately), tirzepatide did not increase the risk of any specific cancer types. Despite greater increments in serum calcitonin with 10- and 15-mg tirzepatide doses than with placebo, the included RCTs reported no cases of papillary thyroid carcinoma.
Conclusion
Tirzepatide use in RCTs over 26 to 72 weeks did not increase overall or specific cancer risk.
INTRODUCTION
Obesity has been recognized as a risk factor for many cancers [1]. Several malignancies have shown increased occurrence in individuals with type 2 diabetes (T2D). It is widely accepted that diabetes is not mutagenic but may be mitogenic, likely due to factors such as hyperglycemia, hyperinsulinemia, or the confounding effects of adiposity. Consequently, individuals predisposed to cancer or those with undiagnosed cancer may experience accelerated disease progression in the presence of uncontrolled hyperglycemia [2]. In this context, anti-obesity and antidiabetic drugs should, at a minimum, not adversely affect cancer risk. Metformin may decrease cancer risk, whereas insulin and sulfonylureas could elevate such risks [3]. Thiazolidinediones have been linked to significant reductions in overall cancer risks; however, the association between pioglitazone and urinary bladder cancer, despite initial findings, remains controversial [4,5]. A recent meta-analysis of 157 randomized controlled trials (RCTs) indicated that dipeptidyl peptidase-4 inhibitors do not affect overall cancer risk and are associated with a significantly reduced risk of colorectal cancer [6]. Hypothetically, antidiabetic drugs that promote weight loss or improve hyperinsulinemia are likely to lower cancer risk. Dicembrini et al. [7], in a recent meta-analysis of 27 RCTs, found no difference in cancer incidence between sodium-glucose cotransporter-2 inhibitors and comparators, including placebo. However, the evidence remains limited and inconclusive regarding the cancer risk associated with emerging anti-diabetic medications such as glucagon-like peptide-1 (GLP-1)-based therapies. The U.S. Food and Drug Administration (FDA) database from 2004 to 2009 reported increased risks of pancreatic and thyroid cancer associated with the GLP-1 receptor agonist (GLP-1RA) exenatide compared to other therapies [8]. Another study using the French National Health Care Insurance System Database also reported increased risks of all thyroid cancers and medullary thyroid cancer with the use of GLP-1RAs (exenatide, liraglutide, and dulaglutide) over 1–3 years [9]. In a recent retrospective study based on a nationwide multicenter database of electronic health records of 113 million United States patients, GLP-1RAs, compared with insulin, were associated with a significant risk reduction in 10 of 13 obesity-associated cancers, including pancreatic and colorectal cancer; however, GLP-1RA had no impact on thyroid cancer [10].
Tirzepatide is the first and only GLP-1 and glucose-dependent insulinotropic peptide (GIP) agonist approved by the FDA for the treatment of diabetes and obesity [11,12]. Its effects on glucose lowering and weight reduction are promising [13,14]. Given its high efficacy in reducing glycemia and weight, tirzepatide may also have potential anti-cancer properties. However, its GLP-1-based mechanism of action could theoretically increase the risk of certain cancers [15]. Data regarding the cancer risk associated with tirzepatide are limited in the published RCTs. Additionally, there is a lack of observational studies providing long-term data on its carcinogenic potential. Given the potentially lifelong nature of such treatment, establishing the carcinogenic safety profile of tirzepatide is crucial. A recent systematic review and meta-analysis (SRM) of nine RCTs that examined the cancer risk associated with tirzepatide has been published. This SRM has several limitations, including the exclusion of some available RCTs, restriction to studies conducted among patients with diabetes, and the absence of subgroup analyses for different tirzepatide doses. Furthermore, it did not compare the cancer risk of tirzepatide with that of GLP-1RAs [16]. Therefore, there was a clear need to conduct an updated SRM that includes all relevant RCTs of tirzepatide reporting on cancer risk.
METHODS
Ethical compliance
The SRM was registered with PROSPERO (CRD42024574086), and the protocol summary can be accessed online. It adhered to the guidelines specified in the Cochrane Handbook for Systematic Reviews of Interventions and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) (Appendix 1) [17,18].
Search strategy
A systematic search was conducted across various databases and registers, including MEDLINE (via PubMed), Scopus, Cochrane Central Register, and ClinicalTrials.gov. This search spanned from the inception of each database to June 20, 2024. We employed a Boolean search strategy using the terms ‘tirzepatide’ OR ‘LY3437943,’ applying these terms exclusively to the titles of documents. The aim was to identify both recently published and unpublished clinical trials in English. Additionally, the search involved reviewing references within the clinical trials retrieved for this study, as well as relevant journals.
Study selection
The selection of clinical trials for this meta-analysis adhered to the PICOS criteria for SRM. The patient population (P) included individuals treated with tirzepatide for any clinical indication. The intervention (I) involved the administration of tirzepatide. The control (C) comprised individuals who received either a placebo or another active comparator. The outcomes (O) measured were the proportions of study subjects diagnosed with any form of cancer. The study type (S) was restricted to RCTs. This analysis focused on RCTs that lasted at least 12 weeks and involved study subjects aged 18 years or older. Each trial included at least two treatment arms/groups: one group received tirzepatide either as monotherapy or in combination with other drugs, and the other group was given a placebo or another active comparator, either alone or in combination with other drugs. Excluded from this analysis were clinical trials involving animals or healthy humans, nonrandomized trials, RCTs shorter than 12 weeks, retrospective studies, pooled analyses of clinical trials, conference proceedings, letters to editors, case reports, and articles that lacked relevant data on the outcomes of interest.
Outcomes analyzed
The primary outcome of the study was the overall risk of any cancer in the tirzepatide group compared to the control groups. Secondary outcomes included the risks of specific cancers in the tirzepatide group versus the control groups. Analyses were stratified based on the type of control groups and the dosage of tirzepatide.
Data extraction and dealing with missing data
Four review authors independently extracted data using standardized forms, as detailed elsewhere [19]. The approach to managing missing data is also described in the same source [19].
Risk of bias assessment
Four authors independently assessed the risk of bias (RoB) using version 2 of the Cochrane RoB tool for randomized trials (RoB 2) within the Review Manager (RevMan) computer program, version 7.2.0 (Cochrane Collaboration, London, UK) [20,21]. The specific biases addressed are detailed in the same source [19]. When appropriate (i.e., with at least 10 studies in a forest plot), publication bias was evaluated using funnel plots in the same software [21,22].
Statistical analysis
The results were presented as risk ratios (RRs) for dichotomous variables and standardized mean differences (SMDs) for continuous variables, each with 95% confidence intervals (CIs). Forest plots, created using the RevMan computer program version 7.2.0, illustrated the comparisons of RRs for primary and secondary outcomes. In these plots, the left side indicated a favorable outcome for tirzepatide, while the right side favored the control group(s) [21]. To accommodate the anticipated heterogeneity due to variations in population characteristics and study durations, random effects analysis models were employed. The inverse variance statistical method was utilized consistently across the analyses. The results included forest plots that incorporated data from at least two RCTs. A significance threshold of P<0.05 was established.
Assessment of heterogeneity
The evaluation of heterogeneity began with an analysis of forest plots. Subsequently, the chi-squared test with N-1 degrees of freedom and a significance level of 0.05 was conducted to determine statistical significance. Additionally, the I2 test was employed for further analysis [23]. The interpretation of I2 values has been detailed elsewhere [19].
Grading of the results
The Grading of Recommendations Assessment, Development and Evaluation (GRADE) methodology was used to assess the quality of evidence for each outcome in the meta-analysis [24]. The method for developing the summary of findings (SoF) table and determining the quality of evidence as ‘high,’ ‘moderate,’ ‘low,’ or ‘very low’ has been described elsewhere [19].
RESULTS
Search results
Fig. 1 illustrates the study selection process. Initially, 1,092 articles were identified. After screening titles and abstracts and conducting full-text reviews, the number of studies considered for this meta-analysis was reduced to 30. A detailed evaluation was performed on 13 RCTs involving 13,761 subjects, all of which met the inclusion criteria [25-37]. Seventeen studies were excluded; nine were sub-studies or post hoc analyses of an included trial, while the remaining eight did not report the outcomes of interest (Supplemental Table S1).
Flowchart on study retrieval and inclusion in the meta-analysis.
Study characteristics
All but one [25] of the RCTs included in this meta-analysis were phase 3 trials. Ten trials involved individuals with T2D [25,27,30-37], while three focused on obese or overweight subjects without diabetes [26,28,29]. Six RCTs utilized matching placebos [26-29,31,35], four used insulin [33,34,36,37], two employed GLP-1RA [30,32], and one trial included both placebo and GLP-1RA in the control groups [25]. Most RCTs featured three tirzepatide arms with dosages of 5, 10, and 15 mg [26,30-37]; one included an additional 1 mg arm [25], two had two arms of 10 and 15 mg [27,29], and one trial administered a single tirzepatide arm at the maximum tolerated dose (either 10 or 15 mg) [28]. The duration of the trials varied: one lasted 26 weeks [25], four spanned 40 weeks [31,32,35,37], five extended to 52 weeks [29,30,33,34,36], and three covered 72 weeks [26-28]. The baseline characteristics of the study subjects were consistent across all trial arms in the included RCTs. Table 1 provides a summary of the included studies.
Baseline Characteristics of the Included Randomized Controlled Trials and Participants
Risk of bias in the included studies
Supplemental Fig. S1 illustrates the RoB across the 13 RCTs included in the meta-analysis. Seven trials (53.8%) exhibited a low overall RoB. The SURMOUNT-3 study raised ‘some concerns’ regarding attrition bias due to missing outcome data. Five studies (38.7%) demonstrated high risks for overall bias, primarily due to deviations from intended interventions. Publication bias was evaluated using funnel plots for RCTs that provided data on the primary outcome, as shown in Supplemental Fig. S2.
Grading of the results
The SoF table presents the grades for the certainty of the evidence supporting the primary outcome of the meta-analysis (Supplemental Table S2).
Effect of tirzepatide on the primary outcome: risk of any cancer
Overall, tirzepatide use was associated with a similar risk of any cancer to the pooled control (RR, 0.78; 95% CI, 0.53 to 1.16; I2=0%; P=0.22, high certainty of evidence). In the subgroup analysis, the risks were also comparable in the two groups in patients with diabetes (RR, 0.70; 95% CI, 0.44 to 1.12; I2=0%; P=0.14), and without diabetes (RR, 1.02; 95% CI, 0.50 to 2.12; I2=0%; P=0.95) (Fig. 2).
Forest plot highlighting the risk of any cancer in the tirzepatide versus pooled control groups. IV, intravenous; CI, confidence interval.
When the control groups were analyzed separately, tirzepatide had indifferent risks of any cancer versus placebo (RR, 0.66; 95% CI, 0.38 to 1.16; I2=0%; P=0.15, high certainty of evidence), insulin (RR, 0.89; 95% CI, 0.49 to 1.60; I2=0%; P=0.69, high certainty of evidence), and GLP-1RA (RR, 0.97; 95% CI, 0.22 to 4.34; I2=17%; P=0.97, high certainty of evidence) (Table 2). In subgroup analysis for different doses of tirzepatide, the 10 mg dose of the drug had a lower risk of any cancer than the placebo (RR, 0.34; 95% CI, 0.13 to 0.86; I2=0%; P=0.02). However, these risks were comparable in other instances with tirzepatide and controls (placebo, insulin, and GLP-1RA) in subgroup analyses according to tirzepatide dose (Table 2).
Risks of Any Cancer in the Tirzepatide versus Control Groups
Effect of tirzepatide on the secondary outcomes: risks of individual cancers
Compared to the pooled control groups, tirzepatide did not increase the risks of breast cancer (RR, 0.59; 95% CI, 0.21 to 1.65; I2=0%; P=0.31), cholangiocarcinoma (RR, 0.33; 95% CI, 0.05 to 2.08; I2=0%; P=0.24), colon cancer (RR, 0.73; 95% CI, 0.26 to 2.04; I2=0%; P=0.54), gastric cancer (RR, 1.24; 95% CI, 0.13 to 11.86; I2=0%; P=0.85), glioblastoma (RR, 0.48; 95% CI, 0.08 to 3.04; I2=0%; P=0.44), lung cancer (RR, 0.39; 95% CI, 0.12 to 1.20; I2=0%; P=0.10), lymphoma (any) (RR, 0.18; 95% CI, 0.03 to 1.17; I2=0%; P=0.07), meningioma (RR, 0.62; 95% CI, 0.121 to 3.20; I2=0%; P=0.57), ovarian cancer (RR, 0.68; 95% CI, 0.11 to 4.32; I2=1%; P=0.68), pancreatic cancer (RR, 0.85; 95% CI, 0.10 to 7.43; I2=30%; P=0.89), prostate cancer (RR, 0.53; 95% CI, 0.14 to 1.91; I2=0%; P=0.33), renal cancer (RR, 1.33; 95% CI, 0.37 to 4.78; I2=0%; P=0.66), skin cancer (RR, 1.52; 95% CI, 0.31 to 7.34; I2=0%; P=0.61), squamous cell carcinoma (RR, 1.45; 95% CI, 0.23 to 9.17; I2=0%; P=0.70), thyroid cancer (papillary) (RR, 1.07; 95% CI, 0.22 to 5.12; I2=0%; P=0.93), urinary bladder cancer (RR, 0.49; 95% CI, 0.07 to 3.27; I2=6%; P=0.46), and uterine cancer (RR, 1.12; 95% CI, 0.23 to 5.53; I2=0%; P=0.89) (Table 3). Moreover, in subgroup analyses of the controls, tirzepatide did not increase the risks of any of these cancers compared to placebo, insulin, or GLP-1RA (Table 3).
Risks of Individual Cancers in the Pooled Tirzepatide versus Control (Pooled and Individual) Groups
Greater percent increases in serum calcitonin were observed with tirzepatide doses of 10 mg (SMD 18.28%; 95% CI, 7.45% to 29.11%; I2=100%; P=0.0009) and 15 mg (SMD 12.67%; 95% CI, 9.44% to 15.10%; I2=99%; P<0.00001) than with placebo (Supplemental Fig. S3). However, the included RCTs reported no cases of medullary thyroid carcinoma in either the tirzepatide or the control groups.
DISCUSSION
This SRM is the first in-depth analysis of cancer risks in RCTs involving tirzepatide. It includes data from 13 RCTs, which predominantly exhibit a low overall RoB and encompass 13,761 participants. The SRM assessed the cancer risks associated with tirzepatide in comparison to control groups, which included placebo, insulin, or GLP-1RAs. Our findings indicate that the use of tirzepatide was not linked to an increased risk of any cancer when compared to the pooled controls; this was consistent across subgroup analyses of the control groups. Furthermore, the risks of individual cancers did not show an increase in the tirzepatide group compared to either the pooled or individual control groups.
Tirzepatide is a dual agonist that activates both GLP-1 and GIP receptors. GLP-1 receptors are found in non-neoplastic pancreatic islet cells, duodenal glands, stomach, breast tissue, lung and kidney vasculature, and brain tissue. They are also overexpressed in insulinomas and medullary thyroid carcinomas [38]. GIP is typically expressed in the brain, bone, pancreas, and adipose tissues, and its increased expression has been noted in neuroendocrine tumors and colorectal cancer cells [39,40]. The overexpression of GIP receptors observed in obesity may be linked to the heightened risk of colorectal cancer in obese individuals [39]. In patients receiving tirzepatide, the downstream activation of GLP-1 and GIP receptor-mediated molecular pathways, especially in those with chronic pancreatitis, could increase the risk of developing pancreatic cancer [41]. Similarly, the activation of these receptors in target tissues such as the breast, liver, and colon may encourage cellular progression and growth, potentially leading to a higher risk of malignancy [42]. GLP-1 and GIP receptors are also present in thyroid C-cells, and animal studies have indicated an increased risk of medullary thyroid carcinoma with GLP-1RA treatment [43]. However, the direct applicability of these findings to human risk remains uncertain, even though these findings raise concerns [44].
Nonetheless, GLP-1–based therapies have been identified as having several anti-cancer effects. As previously discussed, diabetes and obesity are known to contribute to both the incidence and progression of cancer. The treatment with GLP-1–based therapies, which mediates a decrease in blood glucose and body weight, may inhibit cancer growth and progression [15]. Data from the Look Action for Health in Diabetes (AHEAD) Trial showed that an intensive lifestyle intervention aimed at weight loss reduced the incidence of obesity-related cancers by 16% in adults with overweight or obesity and T2D after a median follow-up of 11 years [45]. In a large recent cohort study, GLP-1RAs were associated with lower risks of specific obesity-associated cancers compared to insulins or metformin in patients with T2D [10]. A recent meta-analysis, which included data from 37 RCTs and 19 real-world studies, reported no increased risk of any cancer with semaglutide use [46]. Tirzepatide is currently the most potent weight-lowering drug approved. It also has the highest glycemic efficacy following insulin [13]. Theoretically, tirzepatide offers the best anti-cancer effects, considering the potential link between weight loss and reduced cancer risk. The current meta-analysis confirms that tirzepatide is not associated with an increased overall cancer risk. Both tirzepatide and the pooled control group showed comparable overall cancer risks in subjects with and without diabetes. Popovic et al. [16], in their previous meta-analysis, found similar cancer risks in patients with diabetes in both groups. By including more RCTs, our meta-analysis strengthens the evidence provided by Popovic et al. [16] and further confirms the carcinogenic safety of tirzepatide in patients with obesity who do not have diabetes. In subgroup analyses based on the drugs used in the control arms, the overall cancer risks were identical in the tirzepatide group compared to the placebo, insulin, and GLP-1RA groups. The lower overall cancer risks observed in the tirzepatide 10 mg arm compared to the placebo arm may be due to chance. However, this finding offers hope that the theoretical oncogenic benefits of tirzepatide could become a reality, a possibility that future trials will need to confirm.
Although, as previously mentioned, GLP-1RAs are theoretically associated with medullary thyroid carcinoma, no cases were reported in the RCTs included in our study, either in the tirzepatide or control groups. However, we did observe higher increases in serum calcitonin levels, a tumor marker for medullary thyroid carcinoma, with increased doses of tirzepatide compared to placebo. To further ensure the safety of the drug, the clinical significance of this rise in calcitonin levels must be clarified in future tirzepatide trials. Additionally, consistent with previous meta-analyses by Popovic et al. [16] on tirzepatide and Nagendra et al. [46] on semaglutide, we found no increased risk of papillary thyroid carcinoma, the most common type of thyroid cancer, associated with tirzepatide [16,36]. Our findings regarding the risks of other specific cancers with tirzepatide are reassuring and align with the results of previous meta-analyses.
This is the first comprehensive SRM to examine the carcinogenic potential of tirzepatide based on published RCTs. The evidence was found to be reasonably robust, adequately addressing this safety concern commonly associated with GLP-1-based therapies. However, we must acknowledge the limitations due to the relatively short follow-up period and the small size of the study populations, especially given the lifelong nature and high prevalence of the conditions treated by the drug (obesity and T2D) worldwide. Additionally, the proportion of participants from ethnically diverse backgrounds was relatively small, as most of the RCTs included in this SRM were conducted in Europe and America. This poses another limitation, casting doubt on the generalizability of our findings. Furthermore, the RCTs included were not specifically designed to assess the incidence of new cancer cases among the study subjects. To address these uncertainties, longer-term studies with larger and more globally diverse participant groups are necessary.
Based on current data, this systematic review provides reassuring insights into the cancer risks associated with short-term use of tirzepatide (ranging from 26 to 72 weeks) as observed in the included RCTs. Future RCTs that are larger and longer-term, along with real-world studies that appropriately involve diverse ethnic groups, are anticipated to enhance our understanding of the oncogenic or anti-oncogenic potential (if any) of tirzepatide. This promising drug molecule, known for its excellent disease-modifying properties, holds potential for managing obesity and T2D more effectively through an evidence-based approach.
Supplementary Material
Supplemental Table S1.
The Basic Characteristics of the Excluded Randomized Controlled Trials and Participants
Supplemental Table S2.
Summary of Findings Table
Supplemental Fig. S1.
(A) Risk of bias summary: review authors’ judgments about each risk of bias item for each included study. (B) Risk of bias graph: review authors’ judgments about each risk of bias item presented as percentages across all included studies.
Supplemental Fig. S2.
Funnel plot for the studies that were included in the meta-analysis of the risk of any cancer in tirzepatide versus pooled control groups. SE, standard error; RR, risk ratio.
Supplemental Fig. S3.
Forest plot highlighting the percent changes from baseline in serum calcitonin levels in tirzepatide versus placebo groups. SMD, standardized mean difference; SE, standard error; IV, intravenous; CI, confidence interval.
Notes
CONFLICTS OF INTEREST
No potential conflict of interest relevant to this article was reported.
AUTHOR CONTRIBUTIONS
Conception or design: A.B.M.K.H., D.D. Acquisition, analysis, or interpretation of data: A.B.M.K.H., M.S.A., T.S., F.T.Z.A., L.N. Drafting the work or revising: A.B.M.K.H., M.S.A., D.D., T.S., F.T.Z.A., L.N. Final approval of the manuscript: A. B.M.K.H., M.S.A., D.D., T.S., F.T.Z.A., L.N.
