A2007 Smoking, occupational risks factors and bronchial tumor location: a possible impact for lung cancer CT-Scan screening

Thursday, March 22, 2012
Ground Floor (Cancun Center)
Maria Gonzalez, Occupational Diseases Department, University Hospitals Of Strasbourg, Strasbourg, France
Jean-michel Vignaud, Pathology Department- Inserm U954, University Hospital of Nancy, Vandoeuvre Les Nancy, France
Christelle Clement-duchene, Respiratory Medecine Department, University Hospital of Nancy, Vandoeuvre les Nancy, France
Amandine Luc, Inserm U954, Nancy University Hospital Faculty of Medicine, Vandoeuvre les Nancy, France
Pascal Wild, Statistics, PW Statistical Consulting, Laxou, France
Odile Bertrand, Occupational Medecine Department, University Hospital of Nancy, Vandoeuvre les Nancy, France
Luc Thiberville, Respiratory Medecine Department, University Hospital, Rouen, France
Yves Martinet, Respiratory Medecine Department, University Hospital of Nancy, Vandoeuvre les Nancy, France
Jacques Benichou, INSERM U657, University Hospital, Rouen, France
Christophe Paris, Occupational Diseases / U954, University Hospital / INSERM, Vandoeuvre Les Nancy, France
Introduction
The aim of this study was to describe associations between lung tumor location and smoking as well as selected occupational exposures. In the context of lung cancer screening by CT-Scan, tumor location may have an interest. CT-Scan is known to better detect more peripheral tumors.

Methods
Lung cancer cases diagnosed in two French University hospitals between 1997 and 2009 were included. Tumors visible on white-light bronchoscopy were defined as central. Occupational exposures were assessed by the same expert. Data were analyzed by case–case comparisons using unconditional logistic regressions.

Results
1701 cases were included, mainly men (86.3 %), current smokers (52.8 %) or former smokers (42.8 %). Main histological subtypes of cancer were adenocarcinomas (33.8 %) and squamous cell carcinomas (32.6 %). The tumor location was found to be central in 61 % of cases, never smokers and women had more often peripheral tumors. Exposure to asbestos was significantly associated with central location with dose-response relationship (OR for peripheral tumors = 0.45 CI 95 % [0.29-0.70]) for the highest level of exposure. Exposure to silica dust was significantly associated with peripheral tumor (OR for peripheral tumors = 3.28 CI 95 % [1.50-7.17]) for the highest level of exposure. Exposure to welding fumes was associated with central location (OR for peripheral tumors = 0.51 CI 95 % [0.26-0.96]) for the first level of exposure).

Discussion
Our findings suggest that the benefit of lung cancer CT scan screening may be less high in some specific populations, namely heavy smokers or occupationally exposed to asbestos subjects. Specific strategies of screening have to be designed. Not only smoking characteristics but also occupational exposures have to be considered to define more accurately high-risk populations suitable for lung cancer screening or early detection programs.