Jelle M. Beernink1, Niels Jongs1, Hiddo J.L. Heerspink1
1. Department of Clinical Pharmacy and Pharmacology, University Medical Center Groningen, Groningen, Netherlands.
Background: Patients with type 2 diabetes and chronic kidney disease (CKD) show large between-individual variation in albuminuria responses to the SGLT2 inhibitor dapagliflozin (DAPA). N-of-1 trials, in which individuals receive treatment and placebo repeatedly, allow for precise estimation of individual placebo-corrected effects and individual response reproducibility upon re-exposure. We developed a Bayesian model to assess if a participant’s initial placebo-corrected response predicts their response upon re-exposure.
Methods: We conducted a randomized, placebo-controlled n-of-1 trial in participants with type 2 diabetes and CKD. Each participant received DAPA and placebo twice for one week in random order, enabling within-person estimation of placebo-corrected responses during the first and second exposure. Generalized pairwise comparisons yielded 3,160 within- and between-subject ΔUACR comparisons from baseline. Bayesian hierarchical models estimated the posterior UACR response using priors from each participant’s first exposure.
Results: Twenty participants (mean age 64.9 years, mean eGFR 70.2 mL/min/1.73m², median UACR 94.7 mg/g) were included. Observed placebo-corrected UACR responses to DAPA ranged from –73.4% to 89.3% during the first exposure and –155.0% to 91.9% during the second. Model-predicted placebo-corrected UACR responses during the second DAPA exposure, based on DAPA responses during the first exposure ranged from –70.3% to 62.2% and correlated strongly with observed placebo-corrected DAPA responses upon re-exposure (r=0.70, P<0.001; Figure). Participants with initial UACR reductions <–30% or <–20% had high probabilities of repeating the response during re-exposure (75.8% and 86.6%, respectively), compared to 3.0% and 4.8% for those without such reductions.
Conclusion: Placebo-corrected UACR responses to DAPA were reproducible and predictable, paving the way for individualized DAPA response monitoring.
