Metastatic sites and lesion numbers cooperated to predict efficacy of PD‐1 inhibitor‐based combination therapy for patients with metastatic colorectal cancer

Abstract

Abstract Background Limited data have been used to evaluate the efficacy of immunotherapy in metastatic colorectal cancer (mCRC). Furthermore, potential markers that can be used to identify responding patients and to further improve efficacy have not been fully explored. Methods and Results In our study, we included a total of 97 patients with mCRC, who each received programmed death‐1 (PD‐1) inhibitor‐based combination therapy at our center. All 12 hypermutated patients benefited from immunotherapy, with median progression‐free survival (mPFS) reaching 28.3 months, regardless of liver metastasis. The objective response rate (ORR) of non‐hypermutated patients was 16.5% (14/85), with an mPFS of 4.0 months. For non‐hypermutated patients, multivariate analysis revealed that the combination of liver metastasis and baseline lesion number significantly stratified response and survival. The lesion‐based analysis indicated that the lymph node was the most responsive, followed by the peritoneum and lung, with liver metastasis being the least responsive. None of the patients (0/7) with negative programmed ligand‐1 (PD‐L1) expression responded, and positive PD‐L1 expression may serve as a biomarker (mPFS 5.7 vs. 2.2 months, p = 0.002) that can be used to further guide treatment in non‐hypermutated mCRC with liver metastasis (CRLMs). Conclusion Patients with hypermutated mCRC benefited significantly from immunotherapy, whereas the non‐hypermutated cohort with liver metastasis and numerous lesions showed less benefit. The lesion sites reflected varying levels of efficacy, among which PD‐L1 potentially cooperated to guide the immunotherapy of CRLMs.

Keywords

• colorectal cancer
• combination immunotherapy
• liver metastasis
• PD‐L1
• predictive markers