Table 2. Multivariate survival analysis in patients with lung adenocarcinoma.
Dataset | Clinical variables* | Multivariate analysis (P<0.05) | Final model† |
CANDF (N = 73) | CCP score, stage, grade, chemotherapy, smoking history, age, gender | CCP score, chemotherapy, age | CCP score, chemotherapy, age |
MKS (N = 98) | CCP score, stage, grade, radiotherapy treatment, chemotherapy, smoking history, age, gender | Grade | Grade, chemotherapy |
Takeuchi (N = 90) | CCP score, stage, grade, EGFR status, KRAS status, p53 status, smoking history, age, gender | CCP score | CCP score, stage |
Tomida (N = 116) | CCP score, stage, EGFR status, KRAS status, p53 status, smoking history, age, gender | CCP score | CCP score, stage |
Variables include the following (see Table S8 in File S4 for complete multivariate analysis):
Stage: I vs. II (CANDF), I vs. II vs. III (MKS, Takeuchi, Tomida).
Grade: Well vs. moderately vs. poorly differentiated.
Smoking history: current/former vs. never-smoker.
EGFR, KRAS, and p53 status: mutant vs. wild-type.
Final model is constructed from forward step-wise regression of significant variables (P<0.05).