ESPN 50th Annual Meeting

ESPN 2017


 
LONG-TERM RENAL OUTCOME IN PEDIATRIC GLOMERULONEPHRITIS WITH CRESCENT FORMATION.
PORNPIMOL RIANTHAVORN 1 MANUNYA CHACRANON 1

1- DEPARTMENT OF PEDIATRICS, FACULTY OF MEDICINE, CHULALONGKORN UNIVERSITY
 
Introduction:

Crescent formation, a histologic marker of severe glomerular injury, is associated with unfavorable renal outcome. However, information on prognostic factors and renal outcome of pediatric glomerulonephritis with crescents is limited.

Material and methods:

We evaluated 72 children aged ≤18 years and ≥12 months follow-up (44 girls and 28 boys), in whom renal biopsy specimens showed ≥10 glomeruli and ≥10% of crescentic glomeruli, using a histological classification for ANCA-associated glomerulonephritis. Children were classified into four histological classes as six focal (≥50% of unaffected glomeruli); 21 crescentic (≥50% of glomeruli with cellular crescents); 39 mixed (<50% of normal glomeruli, <50% of crescentic glomeruli and <50% of sclerotic glomeruli); and six sclerotic (≥50% of globally sclerotic glomeruli). Time to the endpoint of chronic kidney disease (CKD), defined as ≥3 months of eGFR <60 mL/min/1.73 m2 or ESRD, was determined. Potential predictive factors for CKD including oliguria, gross hematuria, nephrotic proteinuria, hypertension, baseline eGFR <60 mL/min/1.73 m2, dialysis at onset, duration before treatment, type of immune deposits, and histological class were evaluated using Cox regression. 

Results:

At onset, gender, age, oliguria, gross hematuria and duration before treatment were similar among classes. Nephrotic proteinuria, hypertension, eGFR <60 mL/min/1.73 m2 and dialysis requirement were less common in focal than other classes (p<0.05). Type of immune deposits was different among classes (p=0.001). Focal and mixed classes were observed in patients with immune-complex deposits but not in those with pauci-immune or anti-GBM deposits. Over 629 patient-years of follow-up, CKD occurred in 24 patients (33%). The overall probability of CKD at 1, 5 and 10 years were 25%, 32% and 34%, respectively and differed among classes (Figure). The only factor independently predicting CKD was sclerotic class (hazard ratio 16.70 vs.mixed class; 95% CI 3.33–83.70, p=0.001). There was a trend towards higher probability of CKD in crescentic vs. mixed class (hazard ratio 3.24; 95% CI 1.97–10.87, p=0.06). 
 

Conclusions:

In pediatric glomerulonephritis with crescents, the histological classification applied in this study was a major determinant of CKD. Mixed and crescentic patients had similar outcome which was better than sclerotic patients but worse than focal class. Probability of CKD was substantial and inversely correlated with the number of normal glomeruli.