Construction of prognostic prediction model of resectable lung cancer and survival analysis / 肿瘤研究与临床

ABSTRACT

Objective:To explore the value of the constructed prognostic prediction model of resectable lung cancer in predicting the survival and prognosis of patients.Methods:A total of 2 267 patients with primary lung cancer in Shanxi Provincial Cancer Hospital from January 2007 to September 2018 were selected. All patients underwent primary lung cancer surgery without a second primary tumor. Gender, age, occupation, tumor site, pathological type, surgical path, surgical method, tumor stage and treatment were selected as the prognostic factors. A Cox proportional hazard model was used to construct a prognostic index (PI) equation to calculate the PI value of each patient. According to the different ranges of PI values, the low-, intermediate- and high-risk prognosis groups were divided, and the survival status of three groups were evaluated.Results:Gender ( RR= 0.684, P= 0.001), age ( RR= 0.591, P < 0.01), occupation ( RR= 1.439, P= 0.001), pathological type ( RR= 3.694, P < 0.01), surgical path ( RR= 0.734, P= 0.001), tumor stage ( RR= 0.352, P= 0.007) were independent factors affecting the prognosis of patients with resectable lung cancer. Female, ≤65 years old, thoracoscopic surgery, and tumor stage Ⅰ were prognostic protective factors, and their risks of poor prognosis were reduced by 31.6%, 40.9%, 26.6%, and 64.8%, respectively. Farmer and adenosquamous carcinoma were prognostic risk factors, and their risks of poor prognosis were increased by 43.9% and 269.4%, respectively. The PI equation was ∑β ix i=-0.380 X 1-0.526 X 2+0.364 X 31+1.307 X 55-0.309 X 6-1.045 X 81 (X 1 was the gender, X 2 was the age, X 31 was the occupation as a farmer, X 55 was the pathological type of adenosquamous carcinoma, X 6 was the surgical path, X 81 was the tumor stage Ⅰ). PI <-1 was the low-risk group, PI ≥-1 and ≤-0.5 was the intermediate-risk group, PI >-0.5 was the high-risk group, and the differences of their survival rates were statistically significant ( P < 0.05). The 1-, 3-, and 5-year survival rates for the low-, risk groups were 96.8%, 87.0% and 77.9%; the intermediate-risk group were 91.8%, 82.2% and 61.7%; the high-risk group were 86.5%, 61.7% and 50.3%. respectively. Conclusion:The prognostic prediction model of resectable lung cancer can predict the prognosis risk and the corresponding survival rate of patients with resectable lung cancer, and it can help clinicians to evaluate the prognosis and formulate subsequent treatment plans.