Generating a robust prediction model for stage I lung adenocarcinoma recurrence after surgical resection

Yu Chung Wu*, Nien Chih Wei, Jung Jyh Hung, Yi Chen Yeh, Li Jen Su, Wen Hu Hsu, Teh Ying Chou


研究成果: Article同行評審

4 引文 斯高帕斯(Scopus)


Lung cancer mortality remains high even after successful resection. Adjuvant treatment benefits stage II and III patients, but not stage I patients, and most studies fail to predict recurrence in stage I patients. Our study included 211 lung adenocarcinoma patients (stages I-IIIA; 81% stage I) who received curative resections at Taipei Veterans General Hospital between January 2001 and December 2012. We generated a prediction model using 153 samples, with validation using an additional 58 clinical outcome-blinded samples. Gene expression profiles were generated using formalin-fixed, paraffin-embedded tissue samples and microarrays. Data analysis was performed using a supervised clustering method. The prediction model generated from mixed stage samples successfully separated patients at high vs. low risk for recurrence. The validation tests hazard ratio (HR = 4.38) was similar to that of the training tests (HR = 4.53), indicating a robust training process. Our prediction model successfully distinguished high- from low-risk stage IA and IB patients, with a difference in 5-year disease-free survival between high- and low-risk patients of 42% for stage IA and 45% for stage IB (p < 0.05). We present a novel and effective model for identifying lung adenocarcinoma patients at high risk for recurrence who may benefit from adjuvant therapy. Our prediction performance of the difference in disease free survival between high risk and low risk groups demonstrates more than two fold improvement over earlier published results.

頁(從 - 到)79712-79721
出版狀態Published - 2017


深入研究「Generating a robust prediction model for stage I lung adenocarcinoma recurrence after surgical resection」主題。共同形成了獨特的指紋。