Biomarkers to predict steroid resistance in idiopathic nephrotic syndrome: a systematic review

Abstract

In this systematic review we have sought to summarise the current knowledge concerning biomarkers that can distinguish between steroid-resistant nephrotic syndrome and steroid-sensitive nephrotic syndrome. Additionally, we aim to select biomarkers that have the best evidence-base and should be prioritised for further research. Pub med and web of science databases were searched using [steroid resistant nephrotic syndrome AND biomarker]. Papers published between 01/01/2012 and 10/05/2022 were included. Papers that did not compare steroid resistant and steroid sensitive nephrotic syndrome, did not report sensitivity/specificity or area under curve and reviews/letters were excluded. The selected papers were then assessed for bias using the QUADAS-2 tool. The source of the biomarker, cut off, sensitivity/specificity, area under curve and sample size were all extracted. Quality assessment was performed using the BIOCROSS tool. 17 studies were included, comprising 15 case-control studies and 2 cross-sectional studies. Given the rarity of nephrotic syndrome and difficulty in recruiting large cohorts, case-control studies were accepted despite their limitations. Haptoglobin and suPAR were identified as the most promising biomarkers based on their ability to predict rather than assess steroid resistance in nephrotic syndrome, their respective sample sizes and specificity and sensitivity. None of the selected papers stated whether the authors were blinded to the patients disease when assessing the index test in the cohort. These candidate biomarkers must now be tested with much larger sample sizes. Using new biobanks such as the one built by the NURTuRE-INS team will be very helpful in this regard.