Eiji Higashihara a , b , * , Shigeo Horie c , Satoru Muto c , Haruna Kawano c , Mitsuhiro Tambo b , Tsuyoshi Yamaguchi b , Satoru Taguchi b , Shinya Kaname d , Kenich Yokoyama e , Tatsuya Yoshioka e , Toshihito Furukawa f , Hiroshi Fukuhara c
23 November 2020
Background: Tolvaptan was approved for the treatment of autosomal dominant polycystic kidney disease (ADPKD). However, the official indication of “rapidly progressive disease” is described differently in the clinical guidelines. We aim to define “rapidly progressive disease” by risk of ESRD, which is evaluated using estimated height-adjusted total kidney volume (HtTKV) growth rate. Methods: The risk of ESRD was retrospectively analyzed in 617 initially non-ESRD adults with ADPKD and observed with standard of care between 2007 and 2018. The estimated annual growth rate of the HtTKV, termed as eHTKV-α (%/year), is derived from the following equation: [HtTKV at age t] = K(1 + eHTKV-α/100)<sup> t</sup>, where K = 150 mL/m is used in Mayo Imaging Classification and K = 130 mL/m is proposed for individually stable eHTKV-α value from baseline. The accuracy of eHTKV-α to predict ESRD for censored ages was analyzed using time-dependent receiver-operating characteristic curves (ROC). The cutoff point of initially measured eHTKV-α to predict ESRD was assessed using Kaplan-Meier and Cox’s proportional hazards models. Performance characteristics of the cutoff point for censored ages were calculated using time-dependent ROC and validated by the bootstrap method. Results: The area under the time-dependent ROC of eHTKV-α to predict ESRD at age 65 was 0.89 ± 0.04 ( K = 130). The mean renal survival was less than 70 years at eHTKV-α ≥4.0%/year ( K = 130). Mean renal survival was approximately 12 years shorter, and hazard ratio of ESRD was more than 5-time higher at this cutoff point than at lower point. Time-dependent sensitivity for age 65 and cutoff point of 4.0%/year ( K = 130) was 93.4 ± 0.3%. Between cutoff points ≥4.0%/year ( K = 130) and ≥3.5%/year ( K = 150), there was no significant difference in performance characteristics and accuracy to predict ESRD. Conclusion: eHTKV-α well predicts ESRD. Initially, measured eHTKV-α ≥4.0%/year ( K = 130) defines high-risk ESRD. Without additional conditions, a single eHTKV-α cutoff point identifies subjects that are most likely to benefit from tolvaptan.