OBJECTIVE: To assess the usefulness of repeat cytological examination of pleural fluid (PF) for diagnosing malignancy as well as the influence of time length between analyses, effusion's size and pleural fluid biochemistries on the diagnostic yield of cytology. METHODS: Retrospective analysis of 1,427 patients with pleural effusion (PE), including 466 patients with malignant PE. In this latter group, the time length between cytological analysis, the size of the PE, and the biochemical characteristics of PF were recorded. RESULTS: The first cytological analysis had a sensitivity of 48.5%. If this was negative, a second PF specimen was diagnostic in 28.6% of cases, whereas submission of a third PF specimen allowed 10.3% of additional diagnosis. The incidence of positive results depended on the primary tumor (e.g. 66.5% in adenocarcinomas, 30.8% in mesotheliomas), but neither on the time length between cytological analyses nor on the effusion's size. A multivariate analysis showed that a PF to serum glucose ratio </= 0.75 was associated with a higher diagnostic yield of cytology (74 vs. 47%, p < 0.001). CONCLUSION: At least a second PF specimen should be submitted immediately for cytologic analyis in all PE of unknown cause, when the first analysis is not contributory. To delay this second analysis does not increase diagnostic yield. The percentage of cases in which cytologic study of the PF established the diagnosis of malignant PE depends on the tumor type and on certain PF biochemical characteristics such as the PF to serum glucose ratio.
OBJECTIVE: To assess the usefulness of repeat cytological examination of pleural fluid (PF) for diagnosing malignancy as well as the influence of time length between analyses, effusion's size and pleural fluid biochemistries on the diagnostic yield of cytology. METHODS: Retrospective analysis of 1,427 patients with pleural effusion (PE), including 466 patients with malignant PE. In this latter group, the time length between cytological analysis, the size of the PE, and the biochemical characteristics of PF were recorded. RESULTS: The first cytological analysis had a sensitivity of 48.5%. If this was negative, a second PF specimen was diagnostic in 28.6% of cases, whereas submission of a third PF specimen allowed 10.3% of additional diagnosis. The incidence of positive results depended on the primary tumor (e.g. 66.5% in adenocarcinomas, 30.8% in mesotheliomas), but neither on the time length between cytological analyses nor on the effusion's size. A multivariate analysis showed that a PF to serum glucose ratio </= 0.75 was associated with a higher diagnostic yield of cytology (74 vs. 47%, p < 0.001). CONCLUSION: At least a second PF specimen should be submitted immediately for cytologic analyis in all PE of unknown cause, when the first analysis is not contributory. To delay this second analysis does not increase diagnostic yield. The percentage of cases in which cytologic study of the PF established the diagnosis of malignant PE depends on the tumor type and on certain PF biochemical characteristics such as the PF to serum glucose ratio.