Tumor Cell-Free Dna Copy Number Instability (Cni) To Predict Therapeutic Response To Immunotherapy Prior To Cycle 2

3027 Background: Tumorcell-free DNA (cfDNA) provides minimally invasive patient specific biomarkers to monitor tumor burden. Gains and losses of chromosomal regions have been detected in plasma as copy number aberrations (CNAs). Tregs are reported to be modulated by immunotherapy (immuno). We measured CNAs changes during treatment by computing a genomic copy number instability index (CNI) of cfDNA and Treg-specific demethylation region (TSDR) as measure for Tregs% of leukocytes compared to response. Methods: In this prospective study, prior to treatment and before each cycle, extracted plasma-DNA was subjected to shallow whole genome sequencing with a post mapping (HG19) read coverage of 24,000-fold per 5.5Mbp bin. Read counts were transformed into (log2 ratio) Z-values, and a CNI-score was calculated. % of TSDR+ leukocytes (TSDR%) were quantified with digital PCR from PBMC DNA. Primary endpoint was best overall response by imaging (irRECIST and RECIST 1.1). Hypotheses were: a) response to immuno is reflected by CNI change vs. baseline during therapy an b) alters TSDR%. Outcome was unblinded for analysis. Results: Of 27 enrolled patients (pts), 23 were assessable for response: 4 advanced melanoma (MEL), 2 renal cell carcinoma (RCC), 5 gastrointestinal, 4 pulmonary, 3 breast, 1 ovarian cancers, 3 pancreatic adenocarcinomas, and 1 sarcoma. MEL and RCC received interleukin-2, while the rest received anti-PD-1 with chemotherapy on trials. Median age was 58 years, 12 were women. CNI was measured in 69 samples. Mean baseline cfDNA was 8,076 (CI90th:2894-13,258cp/mL) and CNI was 2,222 (CI90th:1,162-3,282). Pts with response or stable disease (n = 12) showed a significant decrease in CNI before cycle 2 (C2) (Mean: -1,386 vs.+271; p < 0.02), before cycle #3 (-2524 vs.+698, p < 0.005), and thereafter (p < 0.002) compared to pts with disease progression. cfDNA was not correlated to CNI nor response. TSDR% showed a decrease in all patients (p < 0.02) with immuno, independent of response. Conclusions: CNI change predicted response before C2 ~3-12 weeks prior to scan results, and therefore may serve as early predictor of therapeutic response to immuno. Immuno lowers Tregs % in blood, which might reflect T-cell activation.