Safety of glucagon-like peptide-1 agonists: data from the Russian National pharmacovigilance system

Introduction

According to WHO data, obesity is a chronic, relapsing disease affecting over 1 billion people worldwide and leading to significant morbidity, mortality, and substantial economic burden [1]. A key aspect of obesity treatment is the creation of a caloric deficit. Over the past decade, pharmacotherapy has trended toward expanding the use of glucagon-like peptide-1 (GLP-1) agonists. Thus, dulaglutide, liraglutide, lixisenatide, and semaglutide are included in Russian clinical guidelines for the treatment of type 2 diabetes mellitus in adults; liraglutide and semaglutide are included in clinical guidelines for the treatment of obesity in adults; liraglutide is indicated for the treatment of type 2 diabetes mellitus and obesity in children.

Fig. 1. GLP-1 agonists registered in the Russian Federation

The first GLP-1 agonist registered in the Russian Federation in 2010 was liraglutide, indicated for the treatment of type 2 diabetes mellitus to achieve adequate glycemic control in addition to diet therapy and physical exercise, including for the purpose of reducing the risk of cardiovascular events (Fig. 1). Subsequently, between 2010 and 2025, five additional GLP-1 agonists were registered, including in combination with insulin glargine (lixisenatide). In 2025, a new medicinal product — tirzepatide — was registered for the treatment of type 2 diabetes mellitus and obesity as an adjunct to a low-calorie diet and physical activity. Tirzepatide has a dual mechanism of action, acting not only as a GLP-1 agonist but also as a glucose-dependent insulinotropic polypeptide (GIP) receptor agonist. The overall trend in the development of GLP-1 agonists and their introduction into clinical practice is aimed at expanding the indications for use, primarily to include patients with obesity without type 2 diabetes mellitus, the development of dual-action medicinal products / combination drugs, and the development of tablet formulations.

The effects of GLP-1 agonists are mediated by their influence on key pathogenetic mechanisms of diabetes and obesity: enhancement of insulin secretion and survival of pancreatic β-cells; modulation of glucagon release by pancreatic α-cells in response to changes in glycemic levels; action on the satiety center in the brain; slowing of gastric emptying; regulation of lipid and glucose metabolism through effects on adipose tissue and the liver [2]. All of these lead to a statistically significant reduction in body weight and the risk of cardiovascular events, as well as normalization of glycated hemoglobin (HbA1c) and cholesterol levels [3].

At the same time, the use of this class of medicinal products is associated with a wide range of adverse reactions (ARs). The likelihood of discontinuing the medicinal product due to ARs compared to placebo was 2.86 (95% CI 1.48; 5.51) for lixisenatide; 2.15 (95% CI 1.26; 3.69) for liraglutide; 2.61 (1.56; 4.37) for semaglutide; 2.30 (95% CI 1.30; 4.09) for tirzepatide; and 2.39 (95% CI 1.14; 4.98) for exenatide [4]. Gastrointestinal disorders, primarily diarrhea, are the most commonly reported. The odds ratio for developing diarrhea compared to placebo was 2.08 (95% CI 1.62; 2.66) for dulaglutide; 2.37 (95% CI 1.84; 3.06) for semaglutide; 1.99 (95% CI 1.56; 2.56) for liraglutide; 2.88 (95% CI 2.09; 3.96) for tirzepatide; and 1.43 (95% CI 1.03; 1.99) for exenatide.

An analysis of the FDA Adverse Event Reporting System (FAERS) national pharmacovigilance database, which included 71,515 spontaneous reports regarding the use of GLP-1 agonists received between April 1, 2005, and December 31, 2021, revealed disproportionality in reporting for the following System Organ Classes (SOCs): "Gastrointestinal disorders" (n=13,104; lower bound of the 95% confidence interval for the information component (IC) [IC025]=1.34); "Investigations" (n=6,889; IC025=0.64); "Metabolism and nutrition disorders" (n=2,943; IC025=0.44); "Benign/malignant neoplasms" (n=1,989; IC025=0.01); "Hepatobiliary disorders" (n=1,497; IC025=0.38) [5]. Another analysis of FAERS data, which combined information from the introduction of GLP-1 agonists into clinical practice through the second quarter of 2023, identified statistically significant disproportionality for reports related to the SOC "Metabolism and nutrition disorders" for semaglutide (relative risk (RR) 3.34; 95% CI 3.22), liraglutide (RR 2.78; 95% CI 2.69), and exenatide (RR 2.15; 95% CI 2.11) [6].

The WHO guideline on the use of GLP-1 agonists for the treatment of obesity in adults states that GLP-1 agonists, as well as dual GIP/GLP-1 agonists, can be used long-term [1]. However, it is also emphasized that efficacy and safety in clinical trials have been evaluated over an average of 52 weeks, which complicates the assessment of the efficacy and safety of GLP-1 agonists over periods longer than two years.

Objective — to characterize the adverse reactions and assess the risks of reporting disproportionality associated with the use of GLP-1 agonists registered in the Russian Federation, based on data from the national pharmacovigilance database.

Materials and Methods

Reports submitted to the "Pharmacovigilance" database of the Automated Information System of Roszdravnadzor (hereinafter referred to as the AIS of Roszdravnadzor) during the period from January 1, 2020, to December 31, 2025, for medicinal products with the following international nonproprietary names (INNs) were studied: dulaglutide, lixisenatide, liraglutide, semaglutide, tirzepatide, exenatide. The extracted data contained the following information: unique report identification number; medicinal product (suspected, concomitant); INN; trade name; manufacturer; batch number; patient sex and age; date of report; outcome; AR; sender; sender region.

Each AR was assigned a corresponding System Organ Class according to the Medical Dictionary for Regulatory Activities (MedDRA), version 25.0 (https://www.meddra.org/). Cases of serious ARs, therapy ineffectiveness, ARs belonging to the SOC "Gastrointestinal disorders", cases of acute pancreatitis, as well as tumor manifestation, were recorded separately.

For each INN, the reporting odds ratio (ROR) and proportional reporting ratio (PRR) were calculated for the most frequent SOCs of ARs [7, 8, 9].

Formulas used for calculations:

ROR = (a/c) / (b/d), (1)

SE(lnROR) = √(1/a + 1/b + 1/c + 1/d) (2)

95% CI = e^(ln(ROR) ± 1.96√(1/a + 1/b + 1/c + 1/d)) (3)

PRR = [a/(a+b)] / [c/(c+d)] (4)

χ² = ((ad-bc)² × (a+b+c+d)) / ((a+b)×(c+d)×(a+c)×(b+d)) (5)

where:

• ROR — reporting odds ratio;

• a — number of reports of the AR of interest for the evaluated medicinal product;

• b — number of reports of the AR of interest for other medicinal products (excluding the evaluated one);

• c — number of reports of other ARs for the evaluated medicinal product;

• d — number of reports of other ARs for other medicinal products;

• SE — standard error;

• χ² — Pearson's chi-squared test;

• PRR — proportional reporting ratio.

Applicability criteria: a>5, lower limit of 95% CI >1, PRR>2, χ²>4.

The following classification of ARs was used:

• Type A (reactions due to the pharmacological properties of the medicinal product; predictable; occur frequently; dose-dependent; typically low mortality);

• Type B (reactions of immunoallergic nature; this type also includes some genetically determined reactions; unpredictable; occur rarely; not dose-related; often serious);

• Type C (reactions occurring after long-term therapy, which may lead to new additional diseases or increase the frequency of "spontaneous" diseases);

• Type D (delayed reactions (carcinogenic, mutagenic, teratogenic effects, reproductive system defects, etc., which may occur months or years after treatment));

• Type E (withdrawal reactions);

• Type F (lack of clinical effect).

Calculations were performed using Statistica software, version 6.0. The statistical significance level p was set at 0.01.

Results

A total of 181 reports on GLP-1 agonists were submitted to the AIS of Roszdravnadzor database, of which 154 were primary. The dynamics of spontaneous reports by year are shown in Fig. 2.

Fig. 2. Dynamics of the number of spontaneous reports by year (2020 – 2025)

The maximum number of reports (n=101) was associated with semaglutide. During the entire observation period, no reports were received regarding the use of exenatide; one report was received regarding the use of insulin glargine and lixisenatide; 10 reports regarding dulaglutide; and 29 reports regarding liraglutide. In 2025, 13 reports were received regarding the use of tirzepatide. Patient characteristics and the median time from the start of medicinal product use to the occurrence of ARs are presented in Table 1.

Table 1. Patient characteristics and median time to AR onset

Notes: INN — international nonproprietary name; IQR — interquartile range; AR — adverse reaction.

As can be seen from the data presented in Table 1, more than half of the patients took liraglutide and tirzepatide for the treatment of obesity/overweight. Moreover, these patients were statistically significantly younger than patients taking other GLP-1 agonists (p <0.01). For dulaglutide, semaglutide, and tirzepatide, ARs predominantly occurred within the first days of medicinal product use. ARs were mainly of type B — allergic reactions, including at the injection site, and type A — gastrointestinal disorders (nausea, vomiting, abdominal pain). Among the seriousness criteria for all GLP-1 agonists, the clinical significance of the event dominated.

A total of 3 cases of neoplasms (type D) were reported in patients taking liraglutide (one case each of ductal adenocarcinoma, breast cancer, and worsening of a brain neoplasm) and 2 cases in patients taking semaglutide (one case each of pancreatic adenocarcinoma and gastrointestinal tract carcinoma). The time to manifestation of ductal adenocarcinoma while taking liraglutide was approximately 1 year, and for breast cancer, approximately 2 years. At the same time, the time interval between the start of semaglutide therapy and the development of gastrointestinal tract carcinoma was 12 days; information on the time interval between the start of semaglutide therapy and the development of pancreatic adenocarcinoma is missing.

Data on the relative risks of developing ARs meeting seriousness criteria, cases of therapy ineffectiveness, ARs belonging to the SOCs "Gastrointestinal disorders" and "Immune system disorders", and cases of acute pancreatitis are presented in Table 2.

Table 2. Relative risks of developing adverse reactions when taking GLP-1 agonists*

*Notes: * — medicinal products with the number of events under consideration less than 5 were not taken into account; INN — international nonproprietary name; ROR — reporting odds ratio; PRR — proportional reporting ratio; AR — adverse reaction.*

As can be seen from the data presented in Table 2, there was no statistically significant disproportionality in reporting regarding the development of ARs meeting seriousness criteria; cases of therapy ineffectiveness; ARs belonging to the SOCs "Gastrointestinal disorders" and "Immune system disorders"; or cases of acute pancreatitis.

No statistically significant differences in the frequency of therapy discontinuation due to ARs were found (Table 3).

Table 3. Risk of GLP-1 agonist withdrawal due to ARs*

*Notes: * — medicinal products with less than 5 events were not taken into account; INN — international nonproprietary name; ROR — reporting odds ratio; PRR — proportional reporting rate.*

Discussion

Our study demonstrated that the frequency of AR reporting, despite increasing sales of medicinal products for obesity treatment in the Russian Federation, remains extremely low: over 5 years, no reports were received regarding ARs during exenatide use, one report during lixisenatide use, and 10 reports during dulaglutide use, making it difficult to assess the safety profile of these medicinal products based on pharmacovigilance system data [10]. According to Roszdravnadzor Order No. 3518 of June 17, 2024, "Procedure for Pharmacovigilance of Medicinal Products for Medical Use," pharmacovigilance is carried out based on reports received from medicinal product circulation entities (physicians, patients); periodic updated safety reports submitted to Roszdravnadzor by marketing authorization holders or other legal entities authorized by them; information obtained during federal state control (supervision) in the field of medicinal product circulation; risk management plans; and special notifications, including emergency safety issue notifications. In our study, information was predominantly submitted by marketing authorization holders.

Despite the absence of disproportionality in the risks of discontinuing GLP-1 agonists, our study demonstrated that more than half of the AR cases in patients receiving dulaglutide, semaglutide, and liraglutide that were reported to the pharmacovigilance system resulted in therapy discontinuation (Table 3). This trend is confirmed by numerous pharmacoepidemiological studies, in which therapy adherence after one year was only 32–50% [11, 12, 13].

Among the reasons for low adherence, a high risk of therapy discontinuation may be considered due to the fact that for patients with obesity without type 2 diabetes mellitus, the cost of therapy is not covered by healthcare systems [10, 13]. Although our study found no statistically significant disproportionality in reports of ineffectiveness, and the overall frequency of ineffectiveness reports remained low, clinical studies also demonstrate that therapy with GLP-1 agonists is ineffective in some patients: for example, the randomized clinical trial STEP 4 evaluating the efficacy of semaglutide in obesity demonstrated that after 20 weeks of therapy, 10% of patients lost less than 5% of their initial body weight, which can be assessed as therapy ineffectiveness [14]. Possible reasons for ineffectiveness currently being considered include genetic features, in particular polymorphisms of the GLP-1 receptor gene, among others [15].

Comparative safety studies demonstrate both advantages and disadvantages of using GLP-1 agonists regarding the development of specific ARs, which may also affect patient adherence to therapy. For example, a comparative study by Xie Y. et al. (2025) assessed the safety of GLP-1 agonists in patients with type 2 diabetes mellitus (n=215,970) compared to sulfonylurea drugs (n=159,465), dipeptidyl peptidase-4 (DPP-4) inhibitors (n=117,989), sodium-glucose cotransporter 2 (SGLT2) inhibitors (n=258,614), and a control group of 1,203,097 patients who continued to use antihyperglycemic drugs not containing GLP-1 agonists [16]. It was demonstrated that GLP-1 agonist therapy was associated with a reduced risk of addiction, psychotic disorders, seizures, neurocognitive disorders (including Alzheimer's disease and dementia), coagulation disorders, cardiometabolic disorders, infectious diseases, and a number of respiratory diseases. At the same time, GLP-1 agonist use was associated with an increased risk of developing gastrointestinal diseases, hypotension, syncope, arthritis, nephrolithiasis, interstitial nephritis, and drug-induced pancreatitis.

Over 5 years of observation, only 5 reports of cancer cases (type D) during GLP-1 agonist use were submitted to the Russian pharmacovigilance system. Based on the timing of administration, only two of them may have a causal relationship with GLP-1 agonist use (both during liraglutide use). At the same time, data on the risks of developing cancer during GLP-1 agonist use are conflicting. For example, the results of a case-control observational study including a total of 2,562 patients with thyroid cancer and 45,184 control group patients demonstrated that the use of GLP-1 agonists for 1-3 years was associated with an increased risk of developing all types of thyroid cancer (RR 1.58, 95% CI 1.27; 1.95) and medullary thyroid cancer (RR 1.78, 95% CI 1.04; 3.05) [17].

These data are supported by the results of a study by Mali G. et al. (2021), which found that as of January 1, 2020, a total of 11,243 cases of thyroid cancer had been reported to the European pharmacovigilance system EudraVigilance, 236 of which were associated with the use of GLP-1 agonists [18]. The strongest associations were demonstrated for liraglutide (RR 27.5 [95% CI, 22.7; 33.3]) and exenatide (RR 22.5 [95% CI, 17.9; 28.3]).

In contrast, a large population-based study including a total of 86,632 individuals (mean age — 52.4±14.5 years; 68.2% female), including 43,317 GLP-1 agonist users and 43,315 control individuals, demonstrated that the incidence rates of 14 types of cancer were 13.6 vs. 16.4 per 1,000 person-years, respectively, indicating a significantly lower overall risk of cancer among individuals taking GLP-1 agonists (RR 0.83 [95% CI 0.76; 0.91, p=0.002]) compared to those who did not take them. In particular, GLP-1 agonist use was associated with a reduced risk of endometrial cancer (RR 0.75 [95% CI 0.57; 0.99], p=0.05), ovarian cancer (RR 0.53 [95% CI 0.29; 0.96, p=0.04]), and meningioma (RR 0.69 [95% CI 0.48; 0.97, p=0.05]) [19]. However, GLP-1 agonist use was associated with an increased risk of kidney cancer (RR 1.38 [95% CI 0.99; 1.93, p=0.04]).

Thus, despite clinical efficacy, the assessment of risks of ARs associated with the use of GLP-1 agonists, especially in the long term, requires careful collection of safety information. At the same time, despite the expanding practice of using this class of medicinal products, due to the extremely low frequency of reporting, the assessment of risks of adverse reactions associated with GLP-1 agonist therapy in the Russian Federation likely requires alternative methodological approaches, such as population-based studies.

Conclusion

1. The assessment of risks of developing ARs during GLP-1 agonist use based on the analysis of the Russian pharmacovigilance database is limited due to the extremely low frequency of AR reporting.

2. For none of the GLP-1 agonists registered in the Russian Federation were statistically significant disproportionality measures obtained for ARs meeting seriousness criteria (including those with fatal outcomes), cases of therapy ineffectiveness, ARs belonging to the SOCs "Gastrointestinal disorders" and "Immune system disorders", or cases of acute pancreatitis.

3. The assessment of risks of developing serious ARs, such as malignancies, during GLP-1 agonist use likely requires studies with alternative designs (e.g., population-based studies).

Recommendations

The pharmacovigilance system is a process designed for the timely detection of potential ARs, assessment of causal relationships, and transformation of this information into risk mitigation strategies. The post-marketing phase of pharmacovigilance is currently crucial for most drug classes. Our study has shown that the passive data collection method of collecting spontaneous reports has proven ineffective for GLP-1 agonists. For this class of drugs, active monitoring in the form of post-marketing safety studies is necessary. A potential tool would be the establishment of a Russian registry of patients receiving medicinal products for obesity.