The phase 3 trial for semaglutide in chronic kidney disease was halted early in October 2023 because an interim analysis showed efficacy crossing the pre-specified threshold for stopping. The same drug, in the same dose range, has now produced positive results in trials for cardiovascular events, heart failure with preserved ejection fraction, sleep apnea, and Alzheimer’s disease in adolescents with obesity. Either the GLP-1 receptor turns out to be involved in the pathophysiology of nearly every chronic disease of metabolic origin, or the regulatory pathway for indication expansion has become unusually permissive at a moment when manufacturers have unusual incentive to pursue it. Both interpretations have evidence behind them.
The SELECT trial, published in 2023, established that semaglutide reduced major adverse cardiovascular events by 20 percent in patients with established cardiovascular disease and obesity but without diabetes. The result was clinically substantial and appeared in the New England Journal of Medicine in November 2023. The trial’s design, eligibility criteria, and primary endpoint had been pre-specified before the GLP-1 cardiovascular hypothesis was as commercially valuable as it has since become. SELECT, in this sense, is among the cleanest of the recent indication expansion trials, and its result is widely regarded as definitive within its specified population.
What followed SELECT, however, has been a sequence of trials whose interpretation is more contested. FLOW, the chronic kidney disease trial, showed reduction in a composite renal endpoint that combined progression of kidney disease, kidney failure, and cardiovascular death. Critics noted that most of the FLOW signal came from cardiovascular death and stroke rather than from progression of kidney disease itself, raising questions about whether the trial demonstrated a kidney-specific effect or merely an extension of the established cardiovascular benefit. Defenders pointed to the clinical importance of the composite endpoint regardless of which component drove it. The FDA approved a kidney indication for semaglutide in early 2025 on the strength of FLOW. The clinical guidelines have begun updating accordingly. Whether nephrologists prescribe semaglutide differently from cardiologists remains to be seen.
The tirzepatide indication expansion has followed a similar trajectory at slightly higher altitude. The SUMMIT trial, published in late 2024, showed tirzepatide reduced the composite of cardiovascular death and worsening heart failure events in patients with heart failure with preserved ejection fraction—a condition for which effective therapies have been historically limited. The trial result, again published in NEJM, was clinically transformative if it holds up in subsequent trials and clinical use. The patient population that may now reasonably be considered for tirzepatide includes a substantial fraction of all patients with cardiovascular disease, kidney disease, and heart failure, regardless of whether obesity is the dominant clinical concern.
What this trajectory has produced is a peculiar epistemic situation in clinical medicine. The GLP-1 drugs are accumulating evidence of benefit in indications that were not, twenty years ago, considered metabolic at all. Sleep apnea trials have shown improvement in apnea-hypopnea indices that may, in some patients, reduce the need for CPAP therapy. Cognitive trials in early Alzheimer’s disease are ongoing. Liver disease trials in MASH have produced positive interim signals. The mechanism of action that links these conditions—chronic low-grade inflammation, insulin resistance as a systemic phenomenon, weight reduction’s downstream effects on multiple organ systems—is biologically plausible. The question is whether the plausibility is doing more work than the trial data does.
There is a concept in pharmacology called pharmacological imperialism, occasionally applied to drug classes that accumulate indications over time as their commercial sponsors find new disease populations to enroll in trials. The BMJ has used the term in connection with statins, which now have evidence of benefit in conditions ranging from cancer prevention to cognitive decline. The pattern is not, in itself, evidence of fraud or even of bias—a drug that affects multiple biological pathways may genuinely produce benefit in multiple conditions, and the willingness to design trials in those conditions reflects appropriate scientific curiosity. The pattern is, however, also consistent with a less salutary dynamic: a drug class accumulates so much commercial value that the marginal cost of running a trial in a new indication is low relative to the upside of a positive result, and the negative trials are simply less likely to be published with the same prominence as the positive ones.
What distinguishes the GLP-1 indication expansion from the statin pattern is the sheer pace and the breadth of the trials. The statins took thirty years to accumulate the indication evidence that the GLP-1 class is accumulating in roughly five. Whether this is because the underlying biology is broader, or because the commercial incentive to generate the trials has been more concentrated, is genuinely difficult to disentangle. The clinicians who have followed the GLP-1 trial program closely tend to credit the underlying mechanism. The clinicians who have spent careers watching pharmaceutical indication expansion tend to be more skeptical. Both views are defensible from the available evidence.
There are practical clinical questions that the indication expansion has not yet answered, and which payers in particular are watching carefully. Whether GLP-1 therapy started for cardiovascular protection should be continued indefinitely, or whether benefit accrues only during active treatment, is unresolved. The early data on weight regain after GLP-1 discontinuation is discouraging from a chronic disease management perspective—patients who stop GLP-1 therapy regain most of their weight within a year. Whether the cardiovascular and renal benefits regress on similar timelines is being actively investigated. If they do, the implication is that GLP-1 therapy must be a chronic, lifelong intervention for the conditions in which benefit is being demonstrated, with all the implications for cumulative spending and patient burden that entails.
The clinical guideline conversation has begun to grapple with these questions. The American Diabetes Association’s 2025 standards include GLP-1 agonists as preferred therapy in a wider range of patients than previous guidelines did. The American College of Cardiology has incorporated semaglutide into its cardiovascular risk reduction recommendations. The American Society of Nephrology is considering a similar update for chronic kidney disease. The guideline updates are based on the trial evidence, but they are also being influenced by the practical reality that a substantial fraction of the relevant patient population is already taking GLP-1s for other indications, and the guideline writers are reasoning forward from observed practice as much as from de novo trial interpretation.
There is a question, easier to ask than to answer, about how clinical specialty boundaries will hold under the pressure of a single drug class affecting so many disease areas. An NEJM perspective in 2024 noted that the historical division of cardiology, endocrinology, and nephrology may become harder to defend if a single therapeutic agent is the appropriate first-line intervention across all three specialties’ principal disease populations. The administrative consequences—billing, prior authorization, prescriber qualification—are not trivial. The intellectual consequence, that the disease taxonomy may need to be reorganized around mechanism rather than affected organ system, is more interesting and more disruptive.
What the cardiometabolic sprawl will probably produce, over the next decade, is a redrawing of how primary care and specialty care divide responsibility for the most common chronic diseases. If GLP-1 therapy becomes standard of care for patients with obesity, type 2 diabetes, established cardiovascular disease, chronic kidney disease, and heart failure with preserved ejection fraction—a patient population that includes most middle-aged Americans with chronic disease—then the historical management pathways through specialist clinics will be insufficient by orders of magnitude. The clinical workforce capable of initiating, titrating, and monitoring GLP-1 therapy at population scale does not currently exist. The practical question is whether primary care can absorb this responsibility, whether new clinical roles will emerge to support it, or whether the access constraints will simply mean that most eligible patients never receive the therapy whose benefit the trials documented. The first interpretation requires substantial investment in primary care training and reimbursement. The second requires regulatory permission for non-physician prescribers in scaled telehealth settings. The third is the path of least resistance, and probably the path the system is currently on.
