Urinary proteome of steroid-sensitive and steroid-resistant idiopathic nephrotic syndrome of childhood.

Research paper by Robert P RP Woroniecki, Tatyana N TN Orlova, Natasha N Mendelev, Ibrahim F IF Shatat, Susan M SM Hailpern, Frederick J FJ Kaskel, Michael S MS Goligorsky, Edmond E O'Riordan

Indexed on: 05 May '16Published on: 10 Jun '06Published in: American journal of nephrology


The response to steroid therapy is used to characterize the idiopathic nephrotic syndrome (INS) of childhood as either steroid-sensitive (SSNS) or steroid-resistant (SRNS), a classification with a better prognostic capability than renal biopsy. The majority (approximately 80%) of INS is due to minimal change disease but the percentage of focal and segmental glomerulosclerosis is increasing. We applied a new technological platform to examine the urine proteome to determine if different urinary protein excretion profiles could differentiate patients with SSNS from those with SRNS. Twenty-five patients with INS and 17 control patients were studied. Mid-stream urines were analyzed using surface enhanced laser desorption and ionization mass spectrometry(SELDI-MS). Data were analyzed using multiple bioinformatic techniques. Patient classification was performed using Biomarker Pattern Software and a generalized form of Adaboost and predictive models were generated using a supervised algorithm with cross-validation. Urinary proteomic data distinguished INS patients from control patients, irrespective of steroid response, with a sensitivity of 92.3%, specificity of 93.7%, positive predictive value of 96% and a negative predictive value of 88.2%. Classification of patients as SSNS or SRNS was 100%. A protein of mass 4,144 daltons was identified as the single most important classifier in distinguishing SSNS from SRNS. SELDI-MS combined with bioinformatics can identify different proteomic patterns in INS. Characterization of the proteins of interest identified by this proteomic approach with prospective clinical validation may yield a valuable clinical tool for the non-invasive prediction of treatment response and prognosis.