fulltext.pdf 509 KB
In order to elucidate factors influencing the prognosis of small-cell lung cancer (SCLC), we reviewed the records of 253 patients with SCLC and evaluated 20 pretreatment prognostic factors by univariate analysis and Cox's multiple regression analysis. Recursive partitioning and amalgamation (RPA) was employed to identify subgroups with similar survival rates. Cox's multiple regression analysis identified five significant factors: extent of disease, number of metastatic sites, serum albumin, serum lactate dehydrogenase, and presence of weight loss. Among these, extent of disease was the most influential factor. RPA analysis revealed three subgroups predicting significantly different prognoses. The median survival time and 3-year survival rate were 18.4 months and 20.6%, respectively for the good-risk group (limited disease without weight loss), 13.5 months and 9.1%, respectively for the intermediate-risk group (limited disease with weight loss or extensive disease with less than two metastatic sites), and 9.2 months and 0%, respectively for the poor-risk group (extensive disease with two or more metastatic sites). These results will be useful for development of new staging system or subsequent stratification for randomized trials.
Cox's multiple regression analysis
recursive partitioning and amalgamayion method
small-sell lung canser
Acta Medica Okayama
Okayama University Medical School
|Web of Science KeyUT|