Thoracentesis for the assessment of lung cancer with pleural effusion

OBJECTIVE: To analyze the pleural and mediastinal effect of thoracentesis tumor-positive cytology in pleural effusions (PE) detected by chest X ray of lung cancer patients. PATIENTS AND METHODS: The study was performed in patients with lung cancer for whom PE was evident in chest X ray films, who then underwent thoracentesis followed by video-assisted thoracoscopy (VAT) to evaluate direct pleural tumor infiltration, mediastinal node involvement and the existence of pleural metastasis. Patients without contraindication underwent the procedure, even if tumor positive cytology was present. When pleural metastasis was found the treatment employed was talc pleurodesis and chemotherapy. Descriptive statistics were compiled and the validity of VAT for pleural metastasis diagnosis, of thoracentesis pleural cytology to detect infiltration of the tumor-adyacent pleura, N2 disease and pleural metastasis were calculated. Survival was also analyzed. RESULTS: PE was present in 188 of 971 consecutive lung cancer patients. Seventy two PEs were visible in the chest X ray films. Volume exceeded 425 mL. Tumor positive pleural cytology was detected in 29 cases (40%). Pleural metastasis were found in 54 patients, 23 of whom had tumor positive pleural cytology. In the other 6 patients with positive cytology the primary neoplasm infiltrated the visceral pleura, completely in 5. In 4 of those 5, the mediastinal pleura was also involved. The primary tumor and diseased lymph nodes were removed from 11 patients, 3 of them with tumoral pleural cytology. Visual pleural inspection by VAT had a sensitivity of 93%, specificity of 82%, positive predicted value (PPV) of 94% and negative predicted value (NPV) of 78% for the diagnosis of pleural metastasis. Thoracentesis cytology showed a sensitivity of 43%, specificity of 67%, PPV of 79% and NPV of 28% for pleural metastasis. For the evaluation of adjacent pleura infiltration, without pleural metastasis, the sensitivity of cytology was 40%, specificity 100%, PPV 100% and NPV 25%. For mediastinal node invasion clinically evaluated, the sensitivity of cytology was 55%, specificity of 62%, PPV 18% and NPV 90%. Survival after thoracotomy was 39% after 2 years, and the median survival time was 14.5 months. In the 11 resected patients, survival was 53% at two years. The difference in survival between patients treated by thoracotomy and those treated by tale pleurodesis after VAT was significant (p < 0.01). The 3 resected patients with pleural tumor-positive cytology survived 84, 39 and 25 months. CONCLUSIONS: Nineteen percent of patients with lung cancer have PE, of which 7% can be seen in chest X ray films. In such patients the likelihood of pleural metastasis is 75%. Pleural metastasis is not necessarily present when PE cytology indicates that tumor is present. VAT can be considered the ideal technique for the assessment of direct pleural invasion by the tumor or of pleural metastasis.