A Heart Failure Warning Built on One Drug in the Class
The warning on the bottle covers the whole class. The trial that actually found the risk was run on a different drug — and the one built specifically to test his own drug came back reassuring.
Bernard K., an 81-year-old man, has kept a small vegetable garden going every summer since he retired from thirty-five years as a postal carrier, work he says "wore out my knees but never my patience." He has had type 2 diabetes for eighteen years, currently on metformin alone with an A1c of 8.2%, and a chart that shows two prior attempts at additional therapy: an SGLT2 inhibitor stopped after two recurrent urinary tract infections in three months, and a GLP-1 receptor agonist stopped after persistent nausea that he described as making him "not want to eat at all," a real concern given his BMI has already drifted down to 22 over the past year. He also has mild diastolic dysfunction on a recent echo, ordered after an incidental soft murmur, without any diagnosed heart failure.
A DPP-4 inhibitor is the class the team is now considering — well tolerated, weight- neutral, minimal hypoglycemia risk, simple once-daily oral dosing, exactly the profile that fits an 81-year-old who has already struggled with two other options. But the class carries its own complicated cardiac safety history: the SAVOR-TIMI 53 trial found a real increase in heart failure hospitalization with saxagliptin specifically, prompting regulatory warnings that were applied, in various forms, across the drug class. TECOS, the equivalent trial for sitagliptin, did not find the same signal — a genuine, documented difference between individual drugs in a class that doesn't always get read that specifically once a warning has already been applied broadly.
A class warning built on one drug's trial
I'd avoid this class given his cardiac profile. SAVOR-TIMI 53 found a real heart failure hospitalization increase, and regulatory agencies responded with warnings applied across the class. He already has mild diastolic dysfunction on echo — I'm not comfortable accepting an increased-risk drug class in that setting when other options exist, even acknowledging he hasn't tolerated them well.
I'd separate the drug from the class. TECOS, the dedicated cardiovascular outcome trial for sitagliptin specifically, did not find the same heart failure signal SAVOR-TIMI 53 found for saxagliptin. Sitagliptin has the most reassuring cardiac data in this class, and treating a warning built on one member's trial as if it applies uniformly denies him a genuinely well- tolerated option that fits everything else about his situation.
I take his diastolic dysfunction seriously, but I don't think SAVOR-TIMI 53's finding for saxagliptin should be read onto sitagliptin specifically, when the trial built to answer that exact question for sitagliptin came back reassuring.
Before either of you finalizes this, I want to look more closely at how genuinely his prior two attempts were trialed. His SGLT2 inhibitor and GLP-1 receptor agonist both have far stronger cardiovascular outcome data than any DPP-4 inhibitor — if either was stopped prematurely rather than after a genuine, dose-optimized attempt, that's worth revisiting before defaulting to the class with the weakest outcome evidence, even a well-tolerated member of it.
Reading his chart, though, both do look like real attempts — two documented UTIs on the SGLT2 inhibitor and persistent nausea on a reduced GLP-1 dose in an already underweight man. That's not a premature abandonment; that's two real intolerances. Sitagliptin is a reasonable next step for him specifically, given that history.
Sitagliptin started at a renally adjusted dose, with an explicit plan to screen for volume- overload symptoms at follow-up given his echo findings. Bernard was told directly why sitagliptin specifically was chosen over the drug-class warning he might have read about, and appreciated the distinction being explained rather than assumed.
Not agreed: whether his cardiac monitoring should include a repeat echo at a defined interval or only symptom-based reassessment. The cardiologist wanted a repeat echo at six months given his existing diastolic dysfunction; the endocrinologist felt symptom-based monitoring was sufficient given sitagliptin's own reassuring trial data. Left for the cardiology follow-up to decide.