Background: Because of the necessity for extended period and large costs until the event occurs, surrogate endpoints are indispensable for implementation of clinical studies to improve chronic kidney disease (CKD) patients' prognosis.
Methods: Subjects with serum creatinine level for a baseline period over 1-3 years were enrolled (n = 69,238) in this community-based prospective cohort study in Okinawa, Japan, and followed up for 15 years. The endpoint was end-stage renal disease (ESRD). The percent of estimated glomerular filtration rate (%eGFR) change was calculated on the basis of the baseline period.
Results: Subjects had a mean ± SD age, 55.59 ± 14.69 years; eGFR, 80.15 ± 21.15 ml/min/1.73 m2. Among the subjects recruited, 15.81% had a low eGFR (<60 ml/min/1.73 m2) and 36.1/100,000 person years developed ESRD. Cox proportional hazards models adjusted for baseline characteristics showed that the risk of ESRD tended to be high with high rates of decrease in %eGFR changes over 2 or 3 years in the high- and low-eGFR groups. The specificities and positive predictive values for ESRD based on a cutoff value of %eGFR change of less than -30% over 2 or 3 years were high in the high- and low-eGFR groups.
Conclusions: %eGFR change tends to be associated with the risk of ESRD. %eGFR change of less than -30% over 2 or 3 years can be a candidate surrogate endpoint for ESRD in the general Japanese population.
Keywords: Chronic kidney disease; Dialysis; End-stage renal disease; Glomerular filtration rate; Surrogate endpoint.