Plasma phosphocreatine (PC) as a predictive biomarker for immune checkpoint inhibition in patients with refractory metastatic colorectal cancer (mCRC): Analysis of the CCTG CO.26 trial.

183 Background: In response to energetic stress, colorectal cancer cells secrete creatine kinase brain-type (CKB). CKB converts creatine and ATP from the extracellular matrix to PC, which is imported intracellularly to sustain survival and metastatic spread. In addition, PC modulates immune cell functions and may play a role in mediating responses to immune checkpoint inhibition. CO.26 was a phase II trial (NCT02870920) that randomized patients (pts) with refractory mCRC to durvalumab plus tremelimumab (D+T) versus best supportive care (BSC). In an exploratory, post-hoc analysis, we investigated the role of plasma PC as a predictive biomarker for response to D+T. Methods: PC concentrations were determined from pre-treatment blood samples with HPLC-tandem mass spectrometer. A minimum p-value approach was used to select an optimum cut-off value which dichotomized patients into low (< 95.6 ng/ml) versus high (≥ 95.6 ng/ml) groups predictive for benefit. Cox proportional hazard models were used to analyze predictive impacts of PC on progression free survival (PFS) and overall survival (OS). Results: Of 180 pts enrolled, pre-treatment blood samples were available for 162 pts (N =115 for D+T; 47 BSC). Pre-treatment PC was low in 15% (N=24) and high in 85% (N=138). There were no differences in baseline characteristics between pts included in this analysis and the total study pts, or PC low and high pts. D+T improved OS significantly in PC low pts (median OS 4.7 months vs 2.3 months; Hazard Ratio (HR) 0.32, 95% confidence interval (CI): 0.11 – 0.95, p = 0.03). There was no improvement in PC high pts with D+T (median OS 6.8 vs 5.2 months; HR 0.80, 95% CI: 0.55 – 1.17, p = 0.24. Interaction p < 0.0001). Plasma PC values had no impact on PFS and rates of disease control. Conclusions: Pts with low plasma PC derived more benefit from immune checkpoint inhibition with D+T in pts with refractory mCRC. Further prospective validation studies are needed. Predictive analysis for OS with pre-treatment PC levels dichotomized by minimum p approach. [Table: see text]