The Journal of the Arkansas Medical Society Issue 9 Vol 114

SCIENTIFIC ARTICLE Prevalence and Characteristics of Primary Lung Cancer Among Large Lung Masses Kshitij Chatterjee, MD 1 ; Nishi Shah, MD 1 ; Amy Joiner, MD 2 ; Yogita Rochlani, MD 1 ; Nikhil Meena, MD, FCCP 1,3 Department of Internal Medicine, UAMS 1 Department of Pathology, UAMS 2 Division of Pulmonary and Critical Care Medicine, Department of Medicine, UAMS 3 ABSTRACT Background and objective: Lung mass has been traditionally defined as a focal pulmonary lesion > 3 cm and is considered malignant until proven otherwise. There is a lack of recent data on prevalence of primary lung cancer among lung masses. We attempt to establish the proportion of lung masses that are primary lung cancer and determine their CT char- acteristics. Methods: Patients with a lung mass were classified in three groups: primary lung cancer, non-lung cancer malignancy, and benign or infectious causes; and CT findings were compared among them. Results: About 63% of lung masses represented primary lung cancer, 18% were non-lung cancer malignancies, and 18% accounted for benign tumors and in- fectious pathologies. A mass that crossed anatomical boundaries on CT was more likely to represent lung cancer (OR 5.54 {1.47-20.86}, p=0.01). Conclusion: Although lung carcinoma remains the most common pathology among lung masses, a significantly in- creasing proportion of masses now repre- sent benign and infectious etiologies. Keywords/MeSH terms: Biopsy, needle; diagnostic imaging; granuloma/ pathology; lung neoplasms; lung/radiog- raphy. INTRODUCTION (approval number 202086). olitary pulmonary nodule (SPN) has been well defined in litera- ture as a spherical, well-circum- scribed, radiographic opacity less than or equal to 3 cm surrounded by aerated lung; without associated adenopathy, at- electasis or pleural effusion. 1 Focal lesions Population: Retrospective chart review was performed on all patients who underwent transthoracic needle aspiration (TTNA) or endobronchial ultrasound-guided trans-bronchial needle aspiration (EBUS-TBNA) at our institution between Jan. 1, 2009 and Dec. 31, 2012. Cases where thoracic CT scans were not available for re- view or size of lesion was not documented were excluded. Cases where a lung mass (>3 cm) was sampled were identified. These cases were cat- egorized in three groups based on the final patho- logical diagnosis: (1) primary lung cancer; (2) non- lung cancer malignancy; and (3) benign or infec- tious pathologies. S larger than 3 cm are described as lung masses and are considered malignant until proven other- wise. 2 The prevalence of lung cancer among such masses was more than 90% in older studies, but there has been no recent re-evaluation. 3-4 Clinical models and calculators described for predicting the probability of lung cancer among SPN might not be applicable for lung masses. 5 Nodule size is a signif- icant predictor for malignancy, but other predictors like margins and contours are difficult to delineate in large lung masses spanning across different lobes. Predictors based on computed tomography (CT) features of large lung lesions will encourage optimum use of non-invasive testing to comple- ment tissue sampling and decrease the need for multiple sampling procedures. Knowledge of prevalence of lung cancer and benign patholo- gies among lung masses is necessary to avoid premature prognostication of patients with lung masses undergoing diagnostic workup. Hence, we investigated lung masses to determine prev- alence of primary lung cancer and CT features predicting a higher risk of malignancy in a retro- spective cohort study. MATERIALS AND METHODS This is a single center, retrospective observa- tional study at a tertiary care hospital. The Institu- tional Review Board of UAMS approved this study 210 • THE JOURNAL OF THE ARKANSAS MEDICAL SOCIETY Data collection: Detailed chart review was performed for the selected cases with a lung mass. Patient characteristics including age at diagnosis, sex, and smoking status were noted. Smoking status was classified as current smoker, former smoker and never smoker. Former smoking was defined as a history of smoking at least 100 cigarettes in lifetime but not at the time of diag- nosis. 6 Thoracic CTs of all patients with lung mass were reviewed and presence or absence of specific features was noted. These findings were: presence of central cavitation, central necrosis, air-bron- chogram, well-defined borders, satellite lesions, lymphadenopathy, pleural effusion, and if the mass crossed anatomical boundaries. Statistical analysis: Number of patients in each group, namely group 1 (primary lung cancer), group 2 (non-lung cancer malignancy) and group 3 (benign or infectious pathologies) were noted. We conducted the analyses using proc freq and univariate methods for categorical and numeric variables respectively. The presence of above- VOLUME 114

The Journal of the Arkansas Medical Society Issue 9 Vol 114 | Page 18