Study objectives: Despite advances in the surgical treatment of spontaneous pneumothorax, the timing of surgical intervention continues to be a subject of controversy. We test the hypothesis that CT scanning can help to predict the probability of the occurrence of primary spontaneous pneumothorax (PSP) by detecting lung bullae.
Design: Prospective, longitudinal cohort study.
Patients and methods: Between May 1994 to March 1995, 28 consecutive patients (23 men; age range, 18 to 47 years; mean, 29 years) with unilateral PSP who were to undergo video-assisted thoracic surgery (VATS) received preoperative CT of the thorax. CT scans were interpreted by one radiologist blinded to the clinical data for the presence of bullae in both lungs. All patients were followed-up in our outpatient clinic for an average of 59.0 months (range, 54 to 64 months).
Results: Eighty-eight percent of the blebs or bullae identified intraoperatively were demonstrated on preoperative CT scans. CT scans also showed the presence of lung blebs or bullae in the contralateral lung in 15 patients (53.6%). During the follow-up period, 4 of these 15 patients (26.7%) with contralateral blebs developed PSP in the untreated lung; none of the patients who did not have contralateral blebs (n = 13) developed PSP (p = 0.04 [chi(2) analysis]).
Conclusions: The detection of lung bullae by CT scanning in the contralateral lung following unilateral PSP is associated with a higher rate of subsequent occurrence of pneumothorax in that lung. Thus, CT scanning can be used to predict the risk of occurrence of this condition, allowing preemptive surgical intervention in selected patients.