Fecal calprotectin as a predictor of gastrointestinal immune-related adverse events (CF-19): A prospective study.

Background: Colitis is a frequent immune-related toxicity, without any biomarker that may predict its onset. It is endoscopically similar to intestinal inflammatory diseases, where fecal calprotectin (FC) is used as a biomarker to early-detect a relapse. We found contradictory evidence about FC and immunotherapy and no prospective study was already published. Methods : We present an analytical, observational and prospective study of one year’s duration. We analyzed FC basal, and then prior to each cycle until the sixth, ending with quarterly follow-up. For evaluating the predictive value of FC we estimated the area under the ROC curve for basal absolute values and for each cycle, and calculated its relative percentage change with respect to basal. We also planned to estimate sensitivity, specificity and predictive values indexes for different cut-off points. Because of lack of recruitment we did a preliminary analysis at the end of the initially estimated period before suggesting its prolongation. Results: 24 patients (19 male) were included in the study. This included n=15 diagnosed with lung cancer, head and neck, renal, bladder and colorectal cancer (n=2, each), and melanoma (n=1). They were treated with Anti PD-1/PDL-1 mono therapy (n=18), combo with chemo (n=2), or combo with anti-CTLA4 (n=2). Three patients had G1 colitis and two, >=G2, all treated with anti-PD1 and before 6th cycle, as described on literature. ROC curve presents AUC 0,559 (CI95%:0,32-0,798) and RR for colitis taking FC value is 1,001 for each 10 units (p=0,493). Conclusion: Even though we must take into account the limitations of the study we cannot conclude that FC could be used as a predictor for detecting immune-mediated colitis.