Predictors of long-term outcomes in pediatric focal segmental glomerulosclerosis

Background Available data on primary focal segmental glomerulosclerosis (FSGS) in children usually report on short follow-up and small samples. Furthermore, the application of the Columbia classification for FSGS in children has not yet been fully agreed. We aimed to confirm the prognosis and risk factors of FSGS in a large cohort of Chinese children. Methods Two hundred seventy-four children with primary FSGS from a single center were enrolled from 2003 to 2018. Long-term renal survival and related risk factors were evaluated by the Kaplan–Meier method and Cox multivariate regression analysis. Receiver operating characteristic (ROC) curve analysis further tested the effect of various risk factors in predicting renal outcomes. The composite end-point included ≥ 50% reduction in estimated glomerular filtration rate and/or end-stage renal disease or death. Results One hundred twenty-five children were diagnosed with not otherwise specified (NOS) (45.6%) variant; 79 with tip lesions (28.8%), 32 with collapsing (11.7%), 31 with cellular (11.3%), and 7 with perihilar lesions (2.6%). The renal survival rate was 80.73% at 5 years, 62.58% at 10 years and 34.66% at 15 years. Multivariate analysis showed that chronic tubulointerstitial damage ≥ 25% (HR 4.14, 95% CI 1.49–11.50, P < 0.01), collapsing variant [(reference: NOS) HR 2.16, 95% CI 1.10–4.27, P = 0.03], segmental sclerosis (HR 1.03, 95% CI 1.01–1.04, P < 0.01) and age at biopsy (HR 0.91, 95% CI 0.85–0.98, P = 0.01) were significantly associated with renal outcomes. ROC curve analysis showed an excellent diagnostic yield of the Columbia classification. The combination of Columbia classification, CTI ≥ 25% and segmental sclerosis had the best predictive value for renal outcomes (AUC = 0.867, sensitivity = 77.78%, specificity = 82.27%, P < 0.01). Conclusions This study reports a renal survival rate of Chinese children with FSGS of 62.58% at 10 years and 34.66% at 15 years. Prognosis is poorer in patients with collapsing variant or CTI ≥ 25% and good in patients with tip variant. The Columbia classification is confirmed as a valuable tool for predicting prognosis of Chinese children with FSGS.