Millions of people now take semaglutide for weight loss or cardiovascular risk reduction, and a growing body of post-marketing surveillance data is raising questions about pancreatitis and kidney injury that the drug’s largest clinical trial did not settle. Two independent analyses of the FDA Adverse Event Reporting System (FAERS) have flagged disproportionate safety signals for both conditions among semaglutide users, even as the SELECT trial, which enrolled more than 17,000 participants, recorded identical pancreatitis rates in drug and placebo groups and no excess kidney failure. The gap between controlled trial results and real-world reporting patterns has become a central tension for prescribers, patients, and regulators.
Why post-marketing semaglutide signals diverge from SELECT trial data
The SELECT trial tested semaglutide 2.4 mg in people with obesity but without diabetes, and its primary report in the New England Journal of Medicine recorded pancreatitis events in 29 semaglutide patients (0.3%) versus 30 placebo patients (0.3%). On the kidney side, a dedicated analysis of SELECT’s long-term renal outcomes found no excess of acute kidney failure in the semaglutide arm compared with placebo. By the standards of a large randomized controlled trial, semaglutide looked clean on both fronts.
Real-world pharmacovigilance tells a different story, or at least a more complicated one. A FAERS-based study covering the period from the fourth quarter of 2017 through the fourth quarter of 2023, according to a pharmacovigilance analysis published in Frontiers in Pharmacology, detected safety signals for pancreatitis and renal events across different semaglutide administration routes. A separate disproportionality and Bayesian analysis of FAERS data, covering 2018 through mid-2023, also identified reporting patterns consistent with kidney and pancreatic adverse events. The two analyses used overlapping but not identical time windows, which means their case counts are not directly comparable, but both pointed in the same direction.
The reason these signals did not appear in SELECT likely comes down to who was studied. Randomized trials apply strict enrollment criteria. Patients with advanced chronic kidney disease, those taking multiple medications that stress the kidneys, and those who escalate doses outside protocol guidance are typically excluded or closely monitored. In the real world, semaglutide is prescribed to a far broader population, including people with baseline renal impairment and those on drugs like NSAIDs or diuretics that can compound kidney stress. That mismatch between trial populations and pharmacy counter populations is where the hypothesis of concomitant nephrotoxic medications and rapid dose escalation gains traction. FAERS cannot prove causation, but it can detect patterns that controlled trials were not designed to capture.
Case reports and systematic reviews add clinical detail
Beyond aggregate signal detection, individual case evidence fills in clinical specifics. A published case series documented acute kidney injury temporally linked to semaglutide use, with presentations including AKI in patients who already had chronic kidney disease. Several of those patients improved after the drug was stopped, which strengthens the temporal association even if it falls short of proving a causal mechanism. The reports also describe common precipitating factors such as volume depletion from severe nausea, vomiting, or diarrhea, which can further compromise renal perfusion in vulnerable patients.
A systematic review that searched the literature through late 2025, according to a Frontiers in Medicine publication, cataloged reported presentations including acute kidney injury, acute renal failure, and acute interstitial nephritis across a range of semaglutide doses. The review mapped a clinical profile that extends beyond what FAERS disproportionality metrics alone can describe, highlighting that some patients required temporary dialysis while others recovered with conservative management. Patients with preexisting kidney disease appeared repeatedly in these reports, suggesting that baseline renal function is a meaningful risk modifier and that even modest additional insults may tip them into overt AKI.
For pancreatitis, the picture is less developed at the case-report level but still present in pharmacovigilance databases. The SELECT trial’s 0.3% rate in both arms suggests that semaglutide does not add pancreatitis risk in a carefully selected population with controlled comorbidities. Whether that holds for patients with histories of gallstone disease, heavy alcohol use, or hypertriglyceridemia-groups often excluded or underrepresented in trials-is a question FAERS signals raise but cannot answer on their own. Some clinicians suspect that rapid weight loss and gallstone formation could mediate a portion of the risk, but definitive mechanistic studies are lacking.
Unanswered questions about semaglutide kidney and pancreas safety
Several gaps in the evidence remain open. No published analysis has linked FAERS-reported kidney events to confirmed electronic health record data, so the severity and clinical course of these cases rest on voluntary reports rather than verified medical records. The exact total of unique patients across the two FAERS analyses is not specified in either study, and the frequently cited figure of tens of thousands of reports depends on secondary aggregation that neither primary paper fully details. Without patient-level linkage, duplicate entries and incomplete follow-up remain persistent limitations.
SELECT’s trial publications do not release individual-level adjudication data for pancreatitis or kidney events, which limits the ability to examine whether specific subgroups-such as those with mild baseline renal impairment or prior gallbladder disease-experienced different risks. It is also unclear how closely trial participants mirror the current wave of off-label or cosmetically driven semaglutide use, where dose escalation may be faster and comorbidities more complex than in a cardiovascular outcomes study. This disconnect matters because safety profiles established in one population may not transfer cleanly to another.
Another unresolved issue is the role of concomitant medications. Many FAERS reports list nonsteroidal anti-inflammatory drugs, diuretics, ACE inhibitors, or contrast agents alongside semaglutide, but the database structure makes it difficult to determine temporal ordering or relative contribution. Without controlled comparisons, it is impossible to know whether semaglutide amplifies kidney risk in the presence of these agents or simply co-occurs in patients who are already medically complex. Similar questions surround pancreatitis risk in patients taking other drugs known to affect the pancreas or biliary tree.
How clinicians are navigating the signal-versus-noise problem
In practice, prescribers are left to navigate between the reassuring clarity of SELECT and the noisier, more ambiguous signals emerging from FAERS and case reports. Many clinicians now screen more carefully for baseline kidney disease before initiating semaglutide, check creatinine and estimated glomerular filtration rate at baseline, and repeat testing after dose escalations or episodes of significant gastrointestinal illness. Some avoid the drug or use lower starting doses in patients with advanced chronic kidney disease, particularly when other nephrotoxic medications cannot be discontinued.
For pancreatitis, clinicians often ask about prior episodes, gallstones, and heavy alcohol use, and they may choose alternative therapies in patients with a strong history of pancreatic disease. Patients starting semaglutide are frequently counseled to report severe, persistent abdominal pain, especially if it radiates to the back or is accompanied by vomiting, so that pancreatitis can be ruled out promptly. While such precautions are not mandated by trial data, they reflect a precautionary response to the accumulating post-marketing signals.
Regulators face a parallel challenge. FAERS is designed to detect rare or unexpected harms, not to quantify risk precisely, and yet its signals are increasingly scrutinized in the context of blockbuster drugs used by millions. Balancing the substantial cardiovascular and weight-loss benefits of semaglutide against uncertain but plausible kidney and pancreatic risks will likely require more than spontaneous reports. Large, linked database studies that integrate pharmacy records, laboratory data, and clinical outcomes could help clarify whether the FAERS signals represent true excess risk, and if so, in which patients and at what doses.
Until those data arrive, the tension between SELECT and real-world pharmacovigilance will persist. For now, the evidence supports continued use of semaglutide in appropriate patients, but with a more individualized approach to kidney and pancreas monitoring, especially in those whose baseline risk is already elevated. The story of semaglutide safety is still being written, and how quickly the next chapters appear may determine whether today’s signals are remembered as early warnings or as noise that never fully resolved into risk.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
