Patients with autosomal dominant polycystic kidney disease were excluded from the pivotal SGLT2 inhibitor trials, and the theoretical concern is real: these drugs increase cyclic AMP signalling, which is the pathway driving cyst growth. Yet the drugs are being prescribed in this population on the strength of general chronic kidney disease benefit. This meta-analysis asked what the evidence actually shows.
It pooled six clinical studies covering 451 patients with autosomal dominant polycystic kidney disease. Compared with each patient's own pre-treatment slope, starting an SGLT2 inhibitor was associated with an attenuation of first-year estimated glomerular filtration rate decline of 0.65 mL/min/1.73 m2 per year (95% CI 0.05 to 1.26).
That single positive result does not survive a proper comparison. Against patients receiving non-SGLT2 inhibitor therapy, the first-year eGFR slope difference was 2.12 mL/min/1.73 m2 per year with a confidence interval from -8.13 to 12.38 — uninformative. Total kidney volume, the disease-specific measure, showed no difference: -31.31 mL per year (95% CI -184.00 to 121.37).
No major safety concerns emerged, but the authors are clear that the evidence is limited by small samples, predominantly observational designs and substantial heterogeneity.
The honest position for clinic is that there is neither evidence of benefit nor evidence of harm. An SGLT2 inhibitor started in a patient with polycystic kidney disease for a separate indication — diabetes, heart failure — is reasonable. Started specifically to slow cyst disease, it is not supported, and the patient should be told so rather than left with an unfounded expectation.
- Do not start an SGLT2 inhibitor specifically to slow autosomal dominant polycystic kidney disease — benefit is unproven
- Continuing or starting one for diabetes or heart failure in a patient with polycystic disease remains reasonable
- Total kidney volume, the disease-specific endpoint, showed no change at all
- Tell patients plainly that this is an evidence gap, not a proven treatment
- Note the safety signal is reassuring but rests on only 451 patients across six mostly observational studies
The statistics, in plain English
Three results, and only one of them is a fair comparison. The positive figure compares each patient's post-treatment slope with their own earlier slope, which is a before-and-after design vulnerable to regression to the mean and to the natural non-linearity of eGFR decline. The comparison that matters — against patients on other therapy — produced an interval from -8.13 to 12.38, meaning the data are compatible with substantial benefit and substantial harm alike. That is what an uninformative result looks like, and it should not be read as a null. With 451 patients across six studies, mostly observational, this is a call for a trial rather than an answer.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for nephrology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free