Rescue Chemotherapy (Ct) After Immune-Oncology (Io) Drugs In Patients (Pts) With Refractory Solid Tumours: A Propensity Score (Ps) Matched Cohort Study

Case reports and uncontrolled studies have shown great responses to CT in pts with solid tumors and hematological malignancies after IO drugs, suggesting a chemosensitization of the checkpoint inhibitors. We explored this effect in a prospective database of pts included in IO, tyrosine kinase inhibitor (TKI) or epigenetic agent (EPI) phase I trials. All pts with solid tumors on phase I trials at VHIO are followed post-progression and CT response was assessed based on standard clinical practice. A propensity score-matched population (CT post-IO vs. CT-post TKI) adjusted for clinico-pathological factors that could influence response to CT and prognosis was identified. Endpoints were RECIST objective response rate (ORR) and progression-free survival (PFS). From 986 pts enrolled between 2010 and 2019, 132 received CT post-progression (post-IO in 55 [42%], post-TKI in 70 [53%] and post-EPI in 7 [5%]). Median age was 61y, 66% females, most common tumors were breast (22%), head & neck (12%), colorectal (10%) and melanoma (10%). Median prior lines in CT post-IO or CT post-EPI groups was 2 and in the CT post-TKI pts was 3 (P = 0.02). Partial response (PR) of 5% and stable disease of 38% were achieved with phase I trial drug. At post progression CT, ORR was 16.4% post-IO, 8.6% post-TKI and 14.3% post-EPI (P = 0.43). An unadjusted Cox model showed no differences in CT PFS across groups (2.4 months post-IO, 2.8 months post-TKI, and 10.3 months post-EPI; P > 0.14 all comparisons). Due to small numbers, post-EPI pts were excluded for PS analysis. Remaining post-IO (n = 55) and post-TKI (n = 55) PS-matched cohort showed no differences in CT ORR (odds ratio 1.96 in favor of post-IO, CI95% 0.6-6.3, P = 0.27) or CT PFS (hazard ratio 0.85 in favor of post-IO, CI95% 0.6-1.3; P = 0.44). In a multivariable model, predictors of PR with CT post-progression on phase I trials were breast cancer (OR = 16.3; p = 0.002) and clinical benefit while on the trials (OR = 5.18; p = 0.04). This unbiased controlled cohort study, despite numerically higher ORR with CT post-IO vs. post-TKI, showed no significant PFS differences with rescue CT between the groups, questioning the chemosensitization effect of IO drugs across solid tumors.