Hye Ryun Kim, Dae Joon Kim, Dae Ryong Kang, Jin Gu Lee, Sun Min Lim, Chang Young Lee, Sun Young Rha, Mi Kyung Bae, Young Joo Lee, Se Hoon Kim, Sang-Jun Ha, Ross Andrew Soo, Kyung Young Chung, Joo Hang Kim, Ji Hyun Lee, Hyo Sup Shim and Byoung Chul Cho^*

Hye Ryun Kim, Sun Min Lim, Sun Young Rha, Joo Hang Kim, Byoung Chul Cho, Yonsei Cancer Center; Hye Ryun Kim, Dae Joon Kim, Dae Ryong Kang, Jin Gu Lee, Sun Min Lim, Chang Young Lee, Sun Young Rha, Mi Kyung Bae, Se Hoon Kim, Kyung Young Chung, Joo Hang Kim, Ji Hyun Lee, Hyo Sup Shim, and Byoung Chul Cho, Yonsei University College of Medicine; Sang-Jun Ha, Yonsei University, Seoul; Young Joo Lee, National Cancer Center, Goyang, Korea; and Ross Andrew Soo, National University of Singapore, Singapore.

^*Corresponding author: Byoung Chul Cho, MD, PhD, Yonsei Cancer Center, 50 Yonsei-ro, Seodaemun-gu, CPO Box 8044, Seoul, Republic of Korea 120-752

Abstract
Purpose To investigate the frequency and the prognostic role of fibroblast growth factor receptor 1 (FGFR1) amplification in patients with surgically resected squamous cell carcinoma of the lung (SCCL) and the association between smoking and FGFR1 amplification.

Patients and Methods Gene copy number of FGFR1 was investigated in microarrayed tumors from 262 patients with SCCL who had tumor tissue as well as smoking and survival data available. Gene copy number was evaluated by fluorescent in situ hybridization, and an FGFR1-amplified tumor (FGFR1 amp⁺) was prespecified as a tumor with nine or more copies of FGFR1.

Results Among 262 patients, the frequency of FGFR1 amp⁺ was 13.0%. Patients with FGFR1 amp⁺ had significantly shorter disease-free survival (DFS; 26.9 v 94.6 months; P < .001) as well as shorter overall survival (OS; 51.2 v 115.0 months; P = .002) than those without FGFR1 amp⁺. Multivariate modeling confirmed that patients with FGFR1 amp⁺ had a significantly greater risk of recurrence and death than those without FGFR1 amp⁺ after adjusting for sex, smoking status, pathologic stage, and adjuvant chemotherapy (DFS: adjusted hazard ratio [AHR], 2.24; 95% CI, 1.45 to 3.45; P < .001; OS: AHR, 1.83; 95% CI, 1.15 to 2.89; P = .01). The frequency of FGFR1 amp⁺ was significantly higher in current smokers than in former smokers and never-smokers (28.9% v 2.5% v 0%; P_trend < .001). As the smoking dosage increased, so did the incidence of FGFR1 amp⁺ (P_trend = .002).

Conclusion FGFR1 amplification is an independent negative prognostic factor in surgically resected SCCL and is associated with cigarette smoking in a dose-dependent manner. FGFR1 amplification is a relevant therapeutic target in Asian patients with SCCL.