Customized autoantibodies (autoAbs) profiling to predict and monitor immune-related adverse events (irAEs) in patients receiving immune checkpoint inhibitors (ICI).

2528 Background: Using a customized microarray, we previously reported that patients (pts) who develop irAEs grade (G)≥2 and those who do not, have different median fluorescent intensity (MFI) levels of specific autoAbs at baseline (pre-ICI). Leveraging a larger dataset, we evaluated whether overall baseline autoAbs elevation and early increases in autoAbs after ICI can predict irAEs as well as if steroid treatment can reduce autoAbs. Methods: Plasma was obtained from pts receiving ICI in two clinical trials (MET4-IO, NCT03686202 and INSPIRE, NCT02644369) and from healthy controls (hc). Collection time points in MET4-IO and INSPIRE studies included: baseline, 3-4 weeks (w), 6-8 w, 24 w and at the end of treatment and baseline and 6 w respectively. Arrays with 162 autoAg customized for frequent irAEs were incubated with plasma and probed with Abs to detect IgG and IgM reactivity. AutoAbs with MFI >500 per individual were compared between hc and pts with and without irAEs G≥2 by the Student-t test. Results: Samples from 114 pts and 14 hc were analyzed (pts characteristics are summarized in the Table). G≥2 irAEs included: hypothyroidism (13), pneumonitis (10), colitis (7), hepatitis (4), skin toxicity (7), infusion reaction (2), pancreatitis (2), meningitis (1), hypophysitis (1), corneal ulcer (1), high creatinine (1), myocarditis (1), myositis (1), diabetes (1), mucositis (1), myasthenia (1) and adrenal failure (1). Hc had less autoAbs with MFI >500 as compared to pts at baseline (median 32 vs 62 p<0.001). IgG with MFI >500 were higher at baseline in pts who developed G≥2 irAEs vs those without G≥2 irAEs (median 39 vs 33, p=0.03). In 23 pts with plasma collected at the time of irAEs, we observed a significant increase of autoAbs with MFI>500 from baseline (median 84 vs 77 p= 0.009). Paired samples at the time of irAEs and after steroids were available for 9/23 pts, showing lower autoAbs after steroid treatment (54 vs 79 p=0.006). No differences in autoAbs with MFI>500 pre and post ICI were seen in pts without G≥2 irAEs (baseline vs first post ICI collection median 58 vs 60, p=0.13). Conclusions: We observed a higher number of IgG with MFI >500 at baseline and a greater increase after ICI administration in individuals with irAEs compared to those without irAEs. Steroid treatment resulted in a decrease in autoAbs. A prospective study is ongoing to validate the potential role of autoAbs for risk stratification and monitoring of irAEs. [Table: see text]