Risk factors for renal outcomes in children with antineutrophil cytoplasmic antibody-associated vasculitis: a nationwide retrospective study in China

Abstract

Background: Pediatric antineutrophil cytoplasmic antibody-associated vasculitis (AAV) is a life-threatening systemic vasculitis featured by liability to renal involvement. However, there are few studies on the risk factors and predictive models for renal outcomes of AAV in children.

Methods: Data from 179 AAV children in multiple centers between January 2012 and March 2020 were collected retrospectively. The risk factors and predictive model of end-stage renal disease (ESRD) in AAV were explored.

Results: Renal involvement was the most typical manifestation (95.5%), and the crescent was the predominant pathological lesion (84.9%). The estimated glomerular filtration rate (eGFR) was evaluated in 114 patients, of whom 59.6% developed ESRD, and the median time to ESRD was 3.20 months. The eGFR [P = 0.006, odds ratio (OR) = 0.955, 95% confidence interval (CI) = 0.924-0.987] and the percentages of global glomerulosclerosis (pGGS; P = 0.018, OR = 1.060, 95% CI = 1.010-1.112) were independent risk factors for ESRD of renal biopsy. Based on the pGGS and eGFR at renal biopsy, we developed three risk grades of ESRD and one predictive model. The Kaplan‒Meier curve indicated that renal outcomes were significantly different in different risk grades (P < 0.001). Compared with serum creatinine at baseline, the predictive model had higher accuracy (0.86 versus 0.58, P < 0.001) and a lower coefficient of variation (0.07 versus 0.92) in external validation.

Conclusions: Renal involvement is the most common manifestation of pediatric AAV in China, of which more than half deteriorates into ESRD. The predictive model based on eGFR at renal biopsy and the pGGS may be stable and accurate in speculating the risk of ESRD in AAV children. Supplementary file 2 (MP4 18937 KB).

Keywords: Antineutrophil cytoplasmic antibody; End-stage renal disease; Glomerulonephritis; Pediatric nephrology; Vasculitis.

Fig. 1

Flow chart of the study design. CHCC Chapel Hill Consensus Conference, AAV antineutrophil cytoplasmic antibody-associated vasculitis, eGFR estimated glomerular filtration rate, ESRD end-stage renal disease, AAGN antineutrophil cytoplasmic antibody-associated glomerulonephritis

Fig. 2

Therapeutic strategies and prognosis in AAV patients. a Immunosuppressive therapies for remission induction in 174 patients; b immunosuppressive therapies for remission maintenance in 127 patients; c percentage of ESRD and non-ESRD patients at different follow-up times; d numbers of patients in different CKD stages at the end of study; e Kaplan‒Meier curves of the time from diagnosis to death in 179 patients of the whole study; f Kaplan‒Meier curves of the time from diagnosis to ESRD in 114 patients. AAV antineutrophil cytoplasmic antibody-associated vasculitis, GC glucocorticoids, CYC cyclophosphamide, TAC tacrolimus, MMF mycophenolate mofetil, RTX rituximab, PE plasma exchange, IM immunosuppressor, AZA azathioprine, ESRD end-stage renal disease, CKD chronic kidney disease

Fig. 3

Receiver operating characteristic (ROC) curves and Kaplan‒Meier (KM) curves of survival analysis of risk factors in AAV children. a ROC of Scr at baseline for predicting ESRD in 114 AAV patients; b KM curve of the time to ESRD in patients with different Scr levels at baseline; c ROC of eGFR at renal biopsy for predicting ESRD 61 AAV in patients; d KM curve of the time to ESRD in patients with different eGFR levels at renal biopsy; e ROC of percentage of global glomerulosclerosis at renal biopsy for predicting ESRD 61 AAV in patients; f KM curve of the time to ESRD in patients with different pGGS at renal biopsy; g KM curve of the time to ESRD in patients with different risk groups; h ROC of P value of ESRD in the predictive model. Scr serum creatinine, ESRD end-stage renal disease, eGFR estimated glomerular filtration rate, AAV antineutrophil cytoplasmic antibody associated vasculitis, AUC area under the receiver operating characteristic curve, pGGS percentages of global glomerulosclerosis

Fig. 4

Comparison results of serum creatinine (Scr) at baseline and the predictive model. a The accuracy and stability of Scr at baseline in internal validation and external validation; b the accuracy and stability of the predictive model in internal validation and external validation; c the accuracy and stability between Scr at baseline and the predictive model in internal validation and external validation