Clinical utility of the modified Glasgow prognostic score in lung cancer: A meta-analysis