FDA Adverse Event Data Reveal Distinct Safety Profiles for Feline Diabetes Drugs Bexagliflozin and Velagliflozin

Two oral medications have transformed the treatment of diabetes in cats, offering owners an alternative to twice-daily insulin injections and giving some felines a chance at a better quality of life. But until now, veterinarians have had limited real-world evidence about the safety of these drugs outside the controlled setting of clinical trials. A new pharmacovigilance study published in the Journal of Veterinary Internal Medicine has mined the United States Food and Drug Administration’s Animal Drug Adverse Events (ADAE) database to build the most comprehensive safety picture to date of bexagliflozin and velagliflozin, the only two sodium-glucose cotransporter 2 (SGLT2) inhibitors approved by the FDA for treating feline diabetes. The findings reveal that, although the two drugs share many risks, each carries a distinct signature of adverse events with different timing, duration, and severity patterns that could reshape how veterinarians monitor diabetic cats on therapy.

Diabetes affects roughly one in 200 cats, making it one of the most common endocrine diseases in the species. Traditional management relies on insulin, but success depends heavily on an owner’s ability to give injections reliably, recognize signs of poor glycemic control, and perform home blood glucose monitoring. For many owners, the financial cost, the stress of handling an unwilling cat, and the disruption of daily routines prove overwhelming; previous research has found that up to 10 percent of owners elect euthanasia rather than pursue treatment. Against that backdrop, once-daily oral SGLT2 inhibitors represented a genuine breakthrough. In clinical trials, bexagliflozin achieved a treatment success rate of 84.0 percent by day 56, while velagliflozin reached 88.4 percent by day 30, and a randomized non-inferiority trial showed that once-daily velagliflozin matched twice-daily insulin in glycemic control without episodes of clinical hypoglycemia.

Yet both drugs carry an FDA-mandated Boxed Warning for diabetic ketoacidosis (DKA) and euglycemic DKA, a life-threatening metabolic complication that has appeared in a small but clinically meaningful proportion of treated cats. Pre-market trials also documented gastrointestinal effects such as vomiting, diarrhea, and anorexia, along with other drug-specific events listed in the prescribing information. Because spontaneous reporting systems can capture safety signals that emerge only after a drug reaches a much broader population, the research team, led by Xiaoheng Lai and Maohua Chen of China, turned to the ADAE database maintained by the FDA Center for Veterinary Medicine. The database serves as an early warning system, collecting reports from veterinarians and owners about suspected adverse drug events in animals across the United States.

The researchers extracted every adverse event report submitted for cats receiving bexagliflozin between the fourth quarter of 2022 and the second quarter of 2025, and for velagliflozin between the third quarter of 2023 and the second quarter of 2025. After deduplication and merging of follow-up reports, the dataset contained 34,187 adverse event reports for cats in total, of which 2,876 involved bexagliflozin and 2,776 involved velagliflozin. Each report was coded using the Veterinary Dictionary for Drug Regulatory Activities terminology, a four-tiered hierarchy that organizes events from broad system organ classes down to specific low-level terms. The study followed the READUS-PV reporting guidelines to ensure transparent and standardized conduct of the disproportionality analysis.

At the heart of the study lies a statistical technique called disproportionality analysis, which asks a deceptively simple question: does a particular adverse event appear more often in reports for a given drug than would be expected if that drug were no more likely than any other to cause it? The team employed a case/non-case design in which the reporting proportion of each adverse event for the target drug was compared with the reporting proportion for all other drugs in the database. Four separate algorithms were applied in parallel: the reporting odds ratio (ROR), the proportional reporting ratio (PRR), the Bayesian confidence propagation neural network (BCPNN), and the multi-item gamma Poisson shrinker (MGPS). Crucially, a signal was only accepted when all four algorithms met their threshold criteria simultaneously, a conservative standard designed to suppress false positives arising from sparse data or multiple testing.

The results revealed striking differences between the two drugs. Bexagliflozin-related adverse events spanned 22 system organ classes, with statistically significant signals at the organ-class level for investigations, endocrine system disorders, and renal and urinary disorders. Among the 25 positive preferred terms, the strongest signals included weight fluctuation, glucosuria, and abnormal serum protein. Velagliflozin-related events covered 19 system organ classes, with significant signals for endocrine system disorders and investigations, and its 19 positive preferred terms were led by ketonuria, hypochloremia, and acid-base disorders. In a head-to-head comparison restricted to the overlapping reporting period, bexagliflozin showed stronger signals for diabetic ketoacidosis (ROR 7.40 versus 5.53) and ketosis (9.54 versus 7.16), while velagliflozin displayed markedly stronger signals for ketonuria (30.89 versus 7.27) and hypochloremia (8.18 versus 4.75). Ketosis was the most frequently reported event for both drugs, appearing in 1,290 bexagliflozin reports and 927 velagliflozin reports.

Timing proved to be another point of divergence. The median time-to-onset of adverse events was 9 days for bexagliflozin compared with 5 days for velagliflozin, and the median duration was 14 days versus 8 days. More than half of the events for both drugs emerged within the first 10 days of treatment, but velagliflozin showed a higher proportion of very early events, with 63.16 percent occurring within the first week compared with 16.00 percent for bexagliflozin. Stratified analysis showed that for both drugs, diabetic ketoacidosis, diarrhea, and ketosis were rapid-onset events, while bexagliflozin was additionally associated with several prolonged-duration events, including polydipsia with a median duration of 420 days and polyuria lasting a median of 90 days. The authors caution, however, that increased thirst and urination are also classic signs of diabetes itself, so those extended durations may reflect underlying disease progression rather than direct drug effects.

The mechanistic explanation for the shared ketosis signal lies in the physiology of SGLT2 inhibition. These drugs work by blocking glucose reabsorption in the kidney, causing excess glucose to be excreted in urine, but monotherapy relies on sufficient residual endogenous insulin production. When insulin is inadequate, metabolism can shift toward ketone production and ultimately ketoacidosis. SGLT2 inhibitors also stimulate glucagon secretion, further promoting ketogenesis. Notably, euglycemic DKA can occur without marked hyperglycemia, which risks diagnostic delay if owners and clinicians are not alert to the possibility. The study reinforces existing recommendations that blood beta-hydroxybutyrate concentrations be assessed at baseline and monitored closely during the first 14 days of therapy, with values above 2.4 mmol/L typically necessitating insulin and values between 1.0 and 2.4 mmol/L warranting reassessment within two to three days.

Beyond the metabolic risks, the analysis identified significant signals for urinary tract infection with both drugs, consistent with clinical trials in which positive urine cultures were among the most frequently reported events for velagliflozin, and for pancreatitis or elevated pancreatic enzymes, a finding complicated by the fact that diabetes itself is a common risk factor for feline pancreatitis. Electrolyte disturbances, particularly hypochloremia, hypokalemia, hyponatremia, and hypophosphatemia with velagliflozin, also emerged prominently, possibly arising secondarily from gastrointestinal fluid losses. An unexpected signal for hypoglycemia appeared with bexagliflozin despite its absence from the product label, prompting the authors to suggest that prescribing information and post-marketing studies be updated. The descriptive data showed predominantly male, older, heavier cats among reported cases, with euthanasia reported in roughly 7 percent of reports for both drugs and death in 1.77 percent of each.

The authors are careful to acknowledge the limitations inherent in spontaneous reporting systems. Under-reporting, reporting bias, variable data quality, and the lack of detailed clinical narratives mean that disproportionality signals represent statistical associations and hypotheses, not established causation or true incidence rates. The absence of a disease-matched comparator, such as insulin reports, may have inflated signals tied to diabetes severity itself. Nevertheless, a sensitivity analysis confirmed that the core signals were robust, and the study provides veterinarians with actionable, drug-specific guidance: vigilant ketone monitoring in the first two weeks regardless of the drug chosen, electrolyte surveillance especially with velagliflozin, and particular attention to weight changes and persistent urinary signs with bexagliflozin. As oral SGLT2 inhibitor use expands in feline practice, this real-world evidence offers a data-driven foundation for individualized treatment decisions and tailored safety monitoring in the management of feline diabetes.

Subject of Research: Post-marketing pharmacovigilance of SGLT2 inhibitors for feline diabetes using the FDA Animal Drug Adverse Events database

Article Title: Post-marketing safety monitoring of sodium-glucose cotransporter 2 inhibitors for diabetes in cats: a pharmacovigilance study based on the FDA ADAE database

Article References: Post-marketing safety monitoring of sodium-glucose cotransporter 2 inhibitors for diabetes in cats: a pharmacovigilance study based on the FDA ADAE database. (n.d.). https://doi.org/10.1093/jvimsj/aalag193

Image Credits: AI Generated

DOI: 10.1093/jvimsj/aalag193

Keywords: bexagliflozin, velagliflozin, SGLT2 inhibitors, feline diabetes, pharmacovigilance, FDA ADAE database, disproportionality analysis, diabetic ketoacidosis, ketosis, adverse events, veterinary internal medicine, drug safety