CHEST Original Research PULMONARY PROCEDURES journal.publications.chestnet.org CHEST / 142 / 2 / AUGUST 2012 377 D epending on the geographic region, 26% to 51% of participants in multidetector CT scan lung can- cer screening trials showed at least one noncalcified pulmonary nodule on their CT scan. 1-4 The likelihood of these nodules being malignant depends on size. 1,5 The Fleischner Society guideline recommends a recall CT scan, PET scan, or biopsy for nodules . 8 mm detected on a CT scan 5 but not by bronchoscopy. The American College of Chest Physicians (ACCP) guide- line recommends only evaluation by bronchoscopy under the condition that an air bronchogram is pre- sent on CT scan or in centers with expertise in newer techniques. 6,7 Literature on the role of newer tech- niques, such as ultrathin bronchoscopy, autofluores- cence bronchoscopy, and CT scan-guided bronchoscopy in lung cancer screening settings is sparse. To our knowledge, a study by McWilliams et al 8 is the only one Background: Up to 50% of the participants in CT scan lung cancer screening trials have at least one pulmonary nodule. To date, the role of conventional bronchoscopy in the workup of suspi- cious screen-detected pulmonary nodules is unknown. If a bronchoscopic evaluation could be eliminated, the cost-effectiveness of a screening program could be enhanced and the potential harms of bronchoscopy avoided. Methods: All consecutive participants with a positive result on a CT scan lung cancer screening between April 2004 and December 2008 were enrolled. The diagnostic sensitivity and negative predictive value were calculated at the level of the suspicious nodules. In 95% of the nodules, the gold standard for the outcome of the bronchoscopy was based on surgical resection specimens. Results: A total of 318 suspicious lesions were evaluated by bronchoscopy in 308 participants. The mean SD diameter of the nodules was 14.6 8.7 mm, whereas only 2.8% of nodules were .30 mm in diameter. The sensitivity of bronchoscopy was 13.5% (95% CI, 9.0%-19.6%); the specificity, 100%; the positive predictive value, 100%; and the negative predictive value, 47.6% (95% CI, 41.8%-53.5%). Of all cancers detected, 1% were detected by bronchoscopy only and were retro- spectively invisible on both low-dose CT scan and CT scan with IV contrast. Conclusion: Conventional white-light bronchoscopy should not be routinely recommended for patients with positive test results in a lung cancer screening program. Trial registration: Nederlands Trial Register; No.: ISRCTN63545820; URL: www.trialregister.nl. CHEST 2012; 142(2):377–384 Abbreviations: ACCP 5 American College of Chest Physicians; NELSON 5 Dutch-Belgian Randomized Lung Cancer Screening Trial; NPV 5 negative predictive value; VDT 5 volume-doubling time The Role of Conventional Bronchoscopy in the Workup of Suspicious CT Scan Screen-Detected Pulmonary Nodules Susan C. van ’t Westeinde, MD; Nanda Horeweg, MD; René M. Vernhout, MD; Harry J. M. Groen, MD, PhD; Jan-Willem J. Lammers, MD, PhD; Carla Weenink, MD; Kristiaan Nackaerts, MD, PhD; Matthijs Oudkerk, MD, PhD; Willem Mali, MD, PhD; Frederik B. Thunnissen, MD, PhD; Harry J. de Koning, MD, PhD; and Rob J. van Klaveren, MD, PhD to report on the role of autofluorescence bronchos- copy in a lung cancer screening trial. The diagnostic yield of bronchoscopy to evaluate solitary pulmonary nodules outside a CT scan screening program varies For editorial comment see page 276 For related article see page 385 Downloaded From: http://journal.publications.chestnet.org/ on 10/23/2015