Lung Parenchymal Assessment in Primary and Secondary Pneumothorax

RATIONALE: The definition of primary spontaneous pneumothorax excludes patients with known lung disease; however, the assumption that the underlying lung is normal in these patients is increasingly contentious. OBJECTIVES: The purpose of this study was to assess lung structure and compare the extent of emphysema in patients with primary versus secondary spontaneous pneumothorax and to patients with no pneumothorax in an otherwise comparable control group. METHODS: We identified patients treated for pneumothorax by screening inpatient and outpatient medical records at one medical center in the United Kingdom. From this group, 20 patients had no clinically apparent underlying lung disease and were classified as having a primary spontaneous pneumothorax, and 20 patients were classified as having a secondary spontaneous pneumothorax. We assembled a control group composed of 40 subjects matched for age and smoking history who had a unilateral pleural effusion or were suspected to have a thoracic malignancy and had a chest computed tomography scan suitable for quantitative analysis. Demographics and smoking histories were collected. Quantitative evaluation of low-attenuation areas of the lung on computed tomography imaging was performed using semiautomated software, and the extent of emphysema-like destruction was assessed visually. MEASUREMENTS AND MAIN RESULTS: The extent of emphysema and percentage of low-attenuation areas was greater for patients with primary spontaneous pneumothorax than for control subjects matched for age and smoking history (median, 0.25 vs. 0.00%; P = 0.019) and was also higher for patients with secondary pneumothorax than those with primary spontaneous pneumothorax (16.15 vs. 0.25%, P < 0.001). Patients with primary pneumothorax who smoked had significantly greater low-attenuation area than patients with primary pneumothorax who were nonsmokers (0.7 vs. 0.1%, P = 0.034). CONCLUSIONS: The majority of patients with primary spontaneous pneumothorax had quantifiable evidence of parenchymal destruction and emphysema. The exclusion of patients with underlying lung disease from the definition of primary spontaneous pneumothorax should be reappraised.