Pos0775 combined model of renal histopathology and clinical parameters better predict one year renal outcomes in lupus nephritis: analysis of 334 kidney biopsies

BackgroundDiagnosis of Lupus Nephritis (LN) is currently based on laboratory tests and renal histopathology. Role of histopathological features in determining long term outcomes is unclear.Objectives1. To assess if clinical and biochemical parameters at baseline can identify renal histopathological class.2. To assess the clinico-histopathological predictors of renal response.MethodsThis is a single centre retrospective study comprising 334 LN renal biopsies. Clinical and biochemical parameters at the time of biopsy were noted and their association with histopathological class, activity and chronicity scores (AS/CS) (ISN/RPS classification) were evaluated. Complete, partial or no response (CR, PR, NR) for renal outcome (EULAR/EDTA) at 1 year were assessed for 293 patients. Binary logistic regression was done to look for the predictors of NR.ResultsClass III/IV LN was seen in 240(71.8%). Hypertension was seen in (52.1%) of class III/IV and <25% each with class II, V and combined class(p<0.001). Class III/IV had lower eGFR [87.6(62.75-118.8)] (p<0.001) than the other classes. Nephrotic range proteinuria was seen in 32% of class V and 21% in class III/IV (p=0.004). Among class-III/IV, AS had weak correlation with baseline UPCR (r=0.31) and eGFR (r=-0.172) (p<0.01). CS had weak negative correlation with eGFR (r=-0.212, p<0.01). NR at 1 year was higher in males (OR-4.6,95%CI-1.9-10.8, p<0.001), those with abnormal serum creatinine (OR-3.3,95%:CI1.6-7.02, p-0.001), higher renal SLEDAI (p<0.05), higher AS, CS (p<0.001), interstitial inflammation and tubular atrophy(p<0.005) (Table 1). On binary logistic regression a combined clinico-histopathological model comprising of serum creatinine, UPCR, male sex and CS performed best in predicting NR (Figure 1).Table 1.Comparison of baseline characteristics among those who attained any response (CR/PR) versus others at 1 yearParameterAny response Complete Response CR/PR (n=233)Others (No response/rescue) (n=60)OR (95% CI)P valueFemale/male, n (%)221(94.8)/12(5.2)48(80)/12(20)4.6(1.9-10.8)0.001Median age at nephritis onset28(11-65)25(13-67)0.079Median SLE duration12(0-232)18(0-144)0.770Hypertension, n (%)100(42.9)34(56.7)0.061Creatinine>1.3mg/dL (median, IQR)21(9.0)15(25)3.3(1.6-7.02)0.001eGFR categories, n (%)137(58.8)27(45)1.7(0.96-3.03)0.003>9057(24.5)15(25)61-9034(14.6)9(15)30-604(1.7)8(13.3)<30Active urinary sediments, n (%)132(56.7)44(73.3)0.019uPCR g/day (median with IQR)1.38(0.8-2.67)1.95(1.18-4.19)0.098Class III/IV, n (%)167(71.7)49(81.7)0.117Class V, n (%)17(7.3)5(8.3)0.788Combined class, n (%)7(3.0)3(5.0)0.469Activity score, median with IQR3(1-6)6(3-9)0.001Chronicity score, median with IQR0(0-1)1(0-2)5.06(1.49-17.21)0.001Presence of Crescents, no (%)43(18.5)17(28.3)0.104Fibrinoid necrosis, n (%)28(12.0)7(11.7)0.791Interstitial inflammation, n (%)86(36.9)33(55)2.08(1.17-3.70)0.003Interstitial fibrosis, n (%)23(10.7)9(15)0.273Tubular atrophy, n (%)64(27.5)27(45)0.003Blood vessel changes, n (%)2(0.9)1(1.7)0.606Fibrinoid necrosis206(88.4)50(83.3)0.339Other changes*Figure 1.ROC curve and AUC for the three different modelsModel 1: Baseline serum creatinine, urine PCR, male sex; AUC – 0.694(0.609-0.779), p <0.001Model 2: Baseline serum creatinine, urine PCR, male sex, chronicity score; AUC – 0.740(0.660-0.820), p<0.001Model 3: Baseline serum creatinine, urine PCR, male sex, chronicity score, crescents, interstitial inflammation; AUC – 0.744(0.664-0.824), p<0.001ConclusionClinical and biochemical parameters can predict the renal histological class to a fair extent but has limited value in predicting the activity and chronicity parameters. Since a combination of clinical and histopathology parameters are better in predicting renal outcomes, performing renal biopsies should be encouraged in LN.AcknowledgementsI have no acknowledgements to declare.Disclosure of InterestsNone declared