Radiotherapy and Oncology 96 (2010) 89–93
Contents lists available at ScienceDirect
Radiotherapy and Oncology journal homepage: www.thegreenjournal.com
Smoking and RT
Influence of smoking status on treatment outcomes after post-operative radiation therapy for non-small-cell lung cancer Sonia K.A. Nguyen *, Laurence Masson-Côté, André Fortin, Anne Dagnault Department of Radiation Oncology, L’Hôtel-Dieu de Québec, Canada
a r t i c l e
i n f o
Article history: Received 17 November 2009 Received in revised form 7 May 2010 Accepted 15 May 2010 Available online 10 June 2010 Keywords: Post-operative radiotherapy Tobacco Lung cancer Treatment outcomes
a b s t r a c t Background and purpose: The role of post-operative radiotherapy in patients with resected non-small-cell lung cancer (NSCLC) is unclear. Modifiable factors, like smoking, may help guide therapy. We retrospectively evaluated the impact of smoking on control in patients undergoing post-operative radiation therapy (PORT) for NSCLC. Materials and methods: Between 1995 and 2007, 152 patients who underwent surgery for NSCLC were analyzed (median follow-up 26 months). Non-smokers were defined as patients who never smoked or who had stopped smoking at the time of initial consultation. Sixty seven percent were non-smokers; 5% never smoked, 40% of the non-smokers had ceased smoking for a year or less, while 55% had stopped for more than a year. Results: On univariate analysis, smokers had worse 5-year local control than non-smokers (70% versus 90%, p = 0.001) and locoregional control (52% versus 77%, p = 0.002). The 5 -year survival rate was 21% for smokers and 31% for non-smokers (p = 0.2). On multivariate analysis, smokers maintained a detrimental effect on locoregional control (HR 3.6, p = 0.0006). Conclusions: Smokers at initial consultation have poorer local and locoregional control after PORT than non-smokers. In patients being considered for PORT for NSCLC, quitting smoking before treatment confers additional treatment advantage. Crown Copyright Ó 2010 Published by Elsevier Ireland Ltd. All rights reserved. Radiotherapy and Oncology 96 (2010) 89–93
Post-operative radiotherapy (PORT) has been explored in an effort to improve treatment outcomes after surgery for non-smallcell lung cancer (NSCLC) patients. A meta-analysis of nine randomized trials showed a detrimental effect of PORT on overall survival at 2 years (55 versus 48%) for the early stages of lung cancer [1]. Nevertheless, the meta-analysis has been criticized for including patient selection, suboptimal staging, old techniques and outdated dose fractionation schedules. Since the 2005 meta-analysis, retrospective data have suggested a survival benefit for PORT, mainly in N2 patients [2,3]. Trodella et al. in a randomized study using optimized radiotherapy techniques, also found good results with PORT in terms of local control in 104 pathological stage I NSCLC patients, decreasing local recurrences from 23% to 2.2%, which translated into improved 5-year survival from 58% to 67% (p = 0.048) [4]. These data are consistent with, but do not prove, a potential benefit for properly administered PORT. The Lung Adjuvant Radiotherapy Trial (LungART) is an ongoing multicenter European prospective phase III trial whose goal is to establish PORT’s safety and efficacy using modern techniques in completely resected NSCLC with N2 * Corresponding author. Address: Département de Radio-Oncologie, CHUQ-HDQ, 11, Côte du Palais, Québec, Canada G1R 2J6. E-mail address:
[email protected] (S.K.A. Nguyen).
nodal stage. LungART is currently sponsored by the Fédération Nationale des Centres de Lutte contre le Cancer. Until the results of LungART become available, however, the potential benefit of PORT remains debatable. A discussion of the risks and benefits of PORT with patients is a challenging task. In light of the renewed interest in PORT, a radiation oncologist should become familiar with the determining prognostic factors. Apart from N2 status, several factors are also known to influence outcome. Older age, male sex, large primary tumour, wedge resection and other factors are prognostic after lung surgery and PORT [5]. However, potentially modifiable factors are the most useful when counseling patients and their families. Among these, tobacco smoking is by far the most common and important factor. The impact of cigarette use has been studied in lung cancer with contradictory results [6–8], but few have studied its influence specifically on patients with PORT. In patients undergoing surgery for NSCLC, current smokers have been reported to have worse survival [9]. However, the evidence for PORT is stronger for disease control than survival. Therefore, the objective of this study was to determine the prognostic impact of smoking status on local and locoregional control in patients undergoing PORT for NSCLC.
0167-8140/$ - see front matter Crown Copyright Ó 2010 Published by Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.radonc.2010.05.008
90
Influence of smoking status on NSCLC
Materials and methods From our departmental databases, we identified 197 patients with resected lung cancer who were treated between 1995 and 2007 with adjuvant PORT. Patients were excluded if they developed metastatic disease before completing adjuvant RT (13 patients), failed to complete a significant part of planned radiotherapy for reasons other than treatment toxicity (4 patients), had missing post-irradiation follow-up information (7 patients), or had tumour histology other than NSCLC (21 patients). Of the remaining patients, 152 patients were included in the study. Every patient had an oncologic tumour resection including node dissection in the vast majority of cases. PORT was delivered for positive margins (35 patients) or for pathological positive nodes, primarily N2 (118 patients). Radiation was performed using a linear accelerator and 6–23 MV photons. Depending on the time of treatment, planning was either carried out using only fluoroscopic images via simulator (43% of patients) or using simulator and/or computed tomography-based three-dimensional planning (57% of patients). For standard two-dimensional treatment, anteroposterior opposed portals were used, which encompassed the mediastinum and tumour bed, if indicated. A boost dose to high risk areas was subsequently given with a similar technique, with oblique fields or spinal cord block. The target volume for patients who were planned with a three-dimensional conformal technique varied among radiation oncologists, but usually included the involved margin, or the pathologically involved nodal stations, the bronchial stump and ipsilateral hilum. Radiation dose varied according to the treating physician (median 50 Gy, range 30–60 Gy). A lung correction factor was introduced into our dosimetric standards in the three-dimensional conformal technique era. The dose per fraction was primarily 2 Gy/day (range, 1.8–3 Gy/day) and patients were treated five times weekly. For patients treated by simulator, the dose was prescribed at mid-plane for anterior–posterior and posterior–anterior fields. When three-dimensional planning was used, the dose was prescribed to the isodose line that best covered the planning target volume with a dose variation between +7% and 5%. Chemotherapy was given in 32 patients. The specific regimens, number of cycles and timing varied considerably. When given, chemotherapy was platinum-based in the majority of patients. Preoperative (14 patients), post-operative sequential (7 patients) or concomitant (9 patients) chemotherapy was given, while a few received both preoperative and post-operative sequential treatment (2 patients). Most received 2 cycles (range, 1–6 cycles). All clinical and RT data were collected through retrospective chart review. Non-smokers were defined as patients who never smoked or who had stopped cigarette use at the moment of initial consultation. Of those who were non-smokers, 5% had never smoked, 40% had stopped smoking for a year or less and 55% had stopped for over a year. They were included together for analysis, since there were too few true non-smokers for a meaningful analysis as a separate category. The remainder were defined as smokers. Number of pack-years and smoking status during radiation treatment were not available. The primary outcome of the study was local and locoregional control. The secondary outcome was overall survival. Survival and disease control from the beginning of treatment to the date of last follow-up or event (local or locoregional recurrence, death of any cause) were measured with a Kaplan–Meier curve. Results at 5 years were presented. Recurrences were classified as local (primary tumour location) or locoregional (primary tumour location and nodes). Statistical differences between curves were calculated by using the log-rank test for the putative prognostic factors. Comparisons between groups were made using the Pearson or maximum-likelihood test for categorical data and the Student t test
for comparison of means. Univariate and multivariate analysis were performed using the stepwise Cox proportional hazards regression model. A p value of less than 0.05 was considered statistically significant. Results are reported as mean ± SD. Statistica version 7.1 was used for data analysis (Statsoft, Tulsa, OK). Results Baseline patient characteristics are summarized in Table 1. The patients were comparable except for age, with smokers developing their cancer and presenting for PORT at a younger age compared to non-smokers (a median age of 59 and 65 years, respectively). There was a trend towards more stage III cancers in non-smokers (p = 0.07). Out of the 103 non-smokers, only 5% never smoked. Fifty five percent had stopped smoking for more than a year, while 40% of the non-smokers had ceased tobacco use for a year or less. The median duration of follow-up for all patients was 26 months (range, 5–121 months). The median followup for non-smokers and smokers was 31 and 19 months, respectively (p = 0.3). Smokers had worse local (70% versus 90%, p = 0.001) (Fig. 1) and locoregional control than non-smokers (52% versus 77%, p = 0.002) Table 1 Baseline characteristics (n = 152). Non-smoker (n = 103)
Smoker (n = 49)
p
Age (y) Median Range
64 39–80
59 39–76
0.03
Gender Female Male
37 (36) 66 (64)
17 (35) 32 (65)
0.8
Tumour histology Adenocarcinoma Bronchioloalveolar Squamous Large cell
43 (42) 5 (5) 37 (36) 18 (17)
23 (47) 0 (0) 17 (35) 9 (18)
0.4
N stage N0 N1 N2
18 (18) 28 (27) 57 (55)
10 (20) 15 (31) 24 (49)
0.7
T stage T1 T2 T3 T4 Tx
26 (25) 41 (40) 22 (22) 13 (13) 1 (0)
9 (18) 25 (51) 11 (22) 4 (8) 0 (0)
0.6
Stage I II III
7 (7) 26 (25) 70 (68)
1 (2) 19 (39) 29 (59)
0.1
Surgery Lobectomy Bilobectomy Pneumonectomy Segmentectomy
43 (42) 10 (10) 45 (44) 5 (4)
20 (41) 3 (6) 26 (53) 0 (0)
0.1
50
50
0.4
Chemotherapy Yes No
23 (22) 80 (78)
9 (18) 40 (82)
0.53
Positive margin Yes No
23 (22) 80 (78)
11 (24) 38 (76)
0.8
Radiotherapy dose Median dose (Gy) Range
Results are reported as mean ± SD. p < 0.05 is considered statistically significant. Data in parentheses are percentages.
91
S.K.A. Nguyen et al. / Radiotherapy and Oncology 96 (2010) 89–93
Table 2 Multivariate analysis of clinical and treatment parameters on locoregional control.
Gender Age N stage Chemotherapy RT dose (Gy) Tobacco use Surgery Margin status Histology
F/M >70/670 N0–1/N2 No/yes P45/
