Background The type of histological class of lupus nephritis dictates treatment approaches among lupus patients with renal involvement. Risk stratification may facilitate therapeutic decisions, when renal biopsy is contraindicated or poses high patient risk.
Objectives To develop risk scores predicting distinct classes of lupus nephritis by using clinical variables and simple laboratory measures.
Methods Demographic, clinical, serological and histopathological variables were recorded from 163 consecutive patients with biopsy-confirmed lupus nephritis. A risk score was developed to estimate the risk for developing distinct histological classes of lupus nephritis. Two-sided Fisher's exact and Mann-Whitney or ANOVA tests were used to compare qualitative and quantitative characteristics, respectively, between patient groups. Both univariate and multivariate models were considered. Measures of discrimination [Receiver operating characteristic (ROC) statistic] were also calculated
Results Variables independently associated (p<0.10) with type II class included age ≤44 years (p=0.087), no malar rash (p=0.051), negative anti-dsDNA (p=0.008) and ≤5 urine leucocytes (p=0.027). The presence of musculoskeletal (MSK) manifestations (p=0.002), new-onset hypertension (p=0.056), positive anti-dsDNA (p<0.0001), >5 urine leucocytes/hpf (p=0.002), creatinine levels>1.2mg/dl (p=0.028) and absence of nephrotic range proteinuria (NRP) (p=0.002) were independently associated with III/IV histological classes. Age>44 years (p=0.011), malar rash (p=0.002), absence of MSK complaints (p=0.002) or leucopenia (p=0.027), NRP (p=0.001), and ≤9 erythrocytes/hpf in urinalysis (p=0.001) were associated with type V lupus nephritis. A risk score for the prediction of specific histological classes was calculated for each patient, with OR [95%CI] for the presence of ≥3 of the aforementioned risk factors of 4.86 [1.99-11.85] for type II, 9.30 [4.09-21.13] for types III/IV and 7.61[2.76-21] for type V, respectively. When ROC curves for the predictive models were fitted, area under the curve (AUC) were 0.711 (p=0.001) for type II nephritis, 0.819 (p<0.0001) for type III/IV nephritis and 0.823 (p<0.0001) for type V.
Conclusions The identification of independent factors associated with specific types of lupus nephritis and a risk score for prediction can provide guidance in selecting specific therapeutic modalities, particularly in cases where renal biopsy is contraindicated or poses unacceptably high patient morbidity and eventually mortality.
Disclosure of Interest None declared