FRI0258 Comparison of elisa and multiplex techniques for quantifying a urine biomarkers panel for lupus nephritis in children

Background A urine ‘biomarker panel’ comprising alpha-1-acid-glycoprotein (AGP), ceruloplasmin (CP), transferrin (TF) and lipocalin-like-prostaglandin-D synthase (LPGDS) has been shown to cross-sectionally perform to an ‘excellent’ level for Lupus Nephritis (LN) identification in children.1 Quantification of all four biomarkers by enzyme linked immunoabsorbant assay (ELISA) techniques is time consuming and costly. Therefore, novel methods of biomarker panel quantification are required to facilitate future urine biomarker led monitoring studies. Objectives The main objective was to compare the ability of ELISA and multiplex biomarker quantification techniques to differentiate active versus inactive LN when the biomarkers are considered individually and in combination. Methods The urinary biomarkers were quantified by both ELISA and a newly developed, custom multiplex platform in participants of the UK Juvenile Systemic Lupus Erythematosus (JSLE) Cohort Study. Multiplex assay development involved identification of appropriate antibody pairs, assessment of JSLE urine sample matrix effects and range finding in JSLE patient samples. Patients were categorised as having active LN (renal domain BILAG score of A, B and previous histological confirmation of LN) or inactive LN (renal BILAG score D or E). Firth’s penalised binary regression with AUC ROC analysis was used to compare the ability of multiplex and ELISA assays to detect active LN disease state univariately and in combination. Results Biomarker analysis was undertaken on 54 JSLE patients (13 active, 41 in-active). Assessment of each biomarker univariately demonstrated similar AUC values regardless of the biomarker quantification technique; LPGDS (ELISA AUC 0.826, multiplex AUC 0.829), TF (ELISA AUC 0.829, multiplex AUC 0.996), CP (ELISA AUC 0.901, multiplex AUC 0.983), AGP (ELISA AUC 0.934, multiplex AUC 0.979) (see table 1). Combining the multiplex biomarker data in the same order as the original ELISA based study1 led to a similar progressive increase in AUC as biomarkers were added to the model (optimal model including AGP +CP + LPGDS+TF ELISA AUC=0.951, multiplex=0.995). For all biomarker combinations, the multiplex-derived AUC was higher than the ELISA AUC. Conclusions This new LN urine biomarker panel multiplex assay has been shown to display a comparable ability for active LN disease state identification as compared to existing ELISA techniques. The major advantage to this approach is that it reduces cost, processing time and the volume of sample required, as compared to ELISA techniques, representing a key enabler for future clinical studies. Reference [1] Smith EM, Jorgensen AL, Midgley A, Oni L, Goilav B, Putterman C, Wahezi D, Rubinstein T, Ekdawy D, Corkhill R, et al. International validation of a urinary biomarker panel for identification of active lupus nephritis in children. Pediatr Nephrol2017;32:283–295. Disclosure of Interest None declared