Predictive Factors of Ultrasonographic Involution of Prenatally Detected Multicystic Dysplastic Kidney

OBJECTIVE To evaluate possible predictive factors of involution on ultrasonography (US) or disappearance of a prenatally detected multicystic dysplastic kidney (MCDK). PATIENTS AND METHODS Forty-five children with unilateral MCDK detected by prenatal ultrasonography between 1989 and 2002 were analysed. All patients except one had 99mTc isotopic scintigraphy to confirm the absence of renal function in the MCDK. All children were managed conservatively with follow-up visits every 6 months, with US 6-monthly during the first 2 years of life and yearly thereafter. Survival was analysed using the Kaplan-Meier method to evaluate the involution of the MCDK, with differences between subgroups assessed using the two-sided log-rank test. Cox's regression model was applied for the multivariate analysis. RESULTS The mean (range) follow-up was 50 (12–167)  months; in all, 279 ultrasonograms were taken, the mean (range) number per patient being 6 (3–10). US showed partial involution of the MCDK in 30 (67%) cases and complete involution in nine (20%). The absolute MCDK length remained almost unchanged in six children (13%). Univariate analysis showed that four variables were possibly associated with complete involution of the MCDK (gender, impalpable kidney, renal length at admission using two thresholds, 62 and 78 mm). After adjusting by Cox's model only a renal length at diagnosis of <62 mm remained associated with complete involution (relative risk 8, 95% confidence interval 0.98–68; P = 0.05). CONCLUSION These results suggest that only a renal length of <62 mm on initial US was predictive of complete involution of the MCDK during the follow-up.