Prognostic Effect of Performance Status on Outcomes of Patients with Colorectal Cancer Receiving First-Line Chemotherapy: A Meta-analysis.

BACKGROUND: Performance status (PS) is a variable derived from the assessment of a patient's functional status, originally proposed to predict drug toxicity. However, despite its characteristic of being subjective and unidimensional, it has become one of the most important prognostic variables for patients with metastatic colorectal cancer (mCRC). In light of the considerable progressive prolongation of median overall survival (OS) of patients with mCRC, it is unclear whether PS continues to be a valid prognostic factor. This article aims to perform a meta-analysis to verify the current prognostic role of PS. METHODS: A search on two databases of prospective trials of first-line chemotherapy in mCRC patients, published in English from 1991 to 2020, was done by predefined criteria. After the selection of phase III trials evaluating the prognostic role of PS, a meta-analysis has been performed. RESULTS: Thirteen trials were included in the meta-analysis. They reported a reduction in the risk of death with a PS 0 compared to a PS 1 or more (HR 0.63, CI 0.54-0.72; 13 studies), which was confirmed for the comparison between PS 0 and PS 1. However, the study found significant heterogeneity (Q = 68.10; p-value < 0.001) and high-grade inconsistency (I2 = 82.38%). Therefore, to explore the reasons for the heterogeneity, a univariate meta-regression was performed, which suggested a possible moderating activity for liver metastases and timing of metastasis. CONCLUSIONS: PS is a reliable prognostic factor for patients with mCRC receiving first-line chemotherapy but is poorly evaluated in phase III trials.