Transthoracic Ultrasound for the Categorization of Pleural Effusions as Malignant: An Adjunct, but Not the Answer?

In this issue of Respiration , Bugalho et al. [6] report the findings of a prospective, observational study examining the diagnostic value of transthoracic ultrasonography for predicting malignancy in undiagnosed pleural effusions in a relatively large, unselected study population. Patients were scanned by chest physicians with expertise in thoracic ultrasound who were blinded to clinical and radiological details. Static images and video clips were reviewed by at least three chest physicians with similar expertise in order to generate a consensus opinion. A total of 133 patients with pleural effusion of unknown etiology were included; 66 were eventually diagnosed with malignant effusion and 67 had benign disease. Transthoracic ultrasound had an overall sensitivity of 80.3%, specificity of 83.6%, PPV of 82.8% and NPV of 81.2% for pleural malignancy. The study confirmed the strong association between malignant effusions and the presence of pleural and diaphragmatic nodularity (OR 29.02 and 95% CI 7.65– 110.01). However, unlike the report by Qureshi et al. [4] (where the pretest probability of malignancy was higher), these features were not 100% specific for malignant disease and were also found in 9% of benign effusions due to tuberculosis or pneumonia. The presence of chest wall invasion, a peripheral lung lesion associated with the effusion or hepatic metastasis were pathognomonic for malignancy, while air bronchograms and a septated pattern The detection, quantification and assessment of the nature of pleural effusions and the evaluation of associated pleural thickening and tumors remain the main indications for the use of transthoracic ultrasonography by respiratory physicians [1, 2] . The echogenicity of malignant pleural effusions is known to vary and is therefore considered nonspecific, but the ultrasonographic appearance of the pleura itself may often suggest a neoplastic etiology [2] . Several morphological criteria, based on those previously validated for contrast-enhanced computed tomography (CECT) [3] , were evaluated in a landmark study by Qureshi et al. [4] . The authors found that a radiologist, blinded to clinical and other radiological investigations, was able to correctively categorize 26 of 33 effusions as malignant and 19 of 19 as benign in patients referred under suspicion of malignancy. Ultrasound had an overall sensitivity of 79%, specificity of 100%, positive predictive value (PPV) of 100% and negative predictive value (NPV) of 73% [4] . The presence of parietal pleural thickening >10 mm, diaphragmatic nodularity or thickening >7 mm, visceral pleural thickening and pleural nodularity/irregularity were associated with malignancy. It was, however, suggested that these results may not be generalizable, given the high level of expertise at the author’s tertiary referral center and the low percentage of patients with benign disease [5] . Published online: February 13, 2014