SUMMARY Acute interstitial nephritis (AIN) results from an immune-mediated kidney injury, which is triggered by use of medications (such as proton pump inhibitors, antibiotics, and immune checkpoint inhibitors) or by autoimmune diseases (such as Sjogren’s syndrome and sarcoidosis). An episode of AIN results in permanent kidney damage in 40-60% of patients and AIN is the cause of about 2% of all cases of chronic kidney disease, which is equivalent to 0.5-1 million U.S. adults. There are two major challenges in the clinical care of patients with AIN. First, due to lack of a reliable non-invasive biomarker, AIN diagnosis requires a kidney biopsy. Biopsy may not be feasible in some patients due to bleeding risk or is delayed in order to optimize this risk. This delay in diagnosis contributes to incomplete recovery of kidney function. Second, corticosteroid therapy, the current standard of care for AIN treatment, does not appear to benefit all patients while exposing them to the risk of short- and long-term adverse events. However, characteristics predicting a favorable response to corticosteroid therapy are currently unknown. The overall goal of this proposal is to validate urine interleukin (IL)-9 and tumor necrosis factor (TNF)-α as biomarkers to diagnose AIN (aim 1) and to identify the subgroup of patients in whom corticosteroid therapy is beneficial (aim 2). In a study of 265 patients conducted at two Yale-affiliate hospitals, we showed that the multiplication product of IL-9 and TNF-α (IL-9*TNF-α) was independently associated with AIN and showed area under receiver operating characteristic curve [AUC] of 0.79 (95% CI, 0.71, 0.88). Addition of IL-9*TNF-α to the clinician’s pre-biopsy diagnosis increased AUC from 0.62 (0.53, 0.71) to 0.85 (0.78, 0.91). We also found that corticosteroid use was associated with higher glomerular filtration rate 6-months after diagnosis only in patients with higher urine IL-9*TNF-α (by 19.7 ml/min/1.73m2). Here we propose a larger, multicenter, observational study to validate these findings. We will enroll a cohort of 580 patients at two large university medical centers (Yale and Johns Hopkins), collect urine samples for biomarker measurement, and establish presence or absence of AIN through an independent, blinded adjudication process led by three renal pathologists. In aim 1, we will validate IL-9*TNF-α as a diagnostic biomarker for AIN by demonstrating improvement in AUC for AIN with IL-9*TNF-α over a clinical model and clinician’s prebiopsy impression. We will also describe test characteristics at two pre-specified cut-offs and test accuracy in differentiating key sub-groups (drug vs. other and by drug classes). In aim 2, we will validate IL-9*TNF-α as a biomarker to identify subgroup of patients most likely to experience kidney function recovery with clinically-prescribed corticosteroids. We will follow patients with AIN for up to six months to determine kidney function after AIN. Results from this study will al...