Summary Carcinoembryonic antigen (CEA), carbohydrate antigens , cytokeratin 19 fragments (CYFRA 21-1), neuron-specific enolase (NSE) and squamous cell carcinoma antigen (SCC) were evaluated in pleural fluid for the diagnosis of malignant effusions. With a specificity of 99%, determined in a series of 121 benign effusions, the best individual diagnostic sensitivities in the whole series of 215 malignant effusions or in the subgroup of adenocarcinomas were observed with CEA, CA 15-3 and CA 72-4. As expected, a high sensitivity was obtained with SCC in squamous cell carcinomas and with NSE in small-cell lung carcinomas. CYFRA and/or CA 15-3 were frequently increased in mesotheliomas. Discriminant analysis showed that the optimal combination for diagnosis of nonlymphomatous malignant effusions was CEA + CA 15-3 + CYFRA + NSE: sensitivity of 94.4% with an overall specificity of 95%. In malignant effusions with a negative cytology, 83.9% were diagnosed using this association. The association CYFRA + NSE + SCC was able to discriminate adenocarcinomas from small-cell lung cancers. Regarding their sensitivity and their complementarity, CEA, CA 15-3, CYFRA 21-1, NSE and SCC appear to be very useful to improve the diagnosis of malignant pleural effusions.

Summary CYFRA 21-1 assay, measuring cytokeratin 19 fragments, was compared with carcinoembryonic antigen (CEA) assay, as an addition to cytological analysis for the diagnosis of malignant effusions. Both markers were determined with commercial enzyme immunoassays in pleural fluid from 196 patients. Cytological analysis and/or pleural biopsy confirmed the malignant origin of the effusion in 99 patients (76 carcinomas, nine pleural mesotheliomas and 14 non-epithelial malignancies). Effusions were confirmed as benign in 97 patients (33 cardiac failures, 39 infectious diseases -including 12 tuberculosis -and 25 miscellaneous effusions). Both markers were significantly higher in malignant than in benign effusions. All the patients with non-epithelial malignancies presented CYFRA and CEA values lower than the 95% diagnostic specificity thresholds (100 and 6 ng ml-' respectively). The diagnostic sensitivity in the group of carcinomas and mesotheliomas was similar for CYFRA (58.8%) and CEA (64.7%). However, CEA had a significantly higher sensitivity in carcinomas (72.4% vs 55.3%), while CYFRA had a clearly higher sensitivity in mesotheliomas (89.9% vs 0%). Interestingly, 12 out of the 16 malignant effusions with a negative cytology were CEA and/or CYFRA positive. Regarding their high diagnostic sensitivity and their complementarity, CEA and CYFRA appear to be very useful for the diagnosis of malignant pleural effusions when cytology is negative.

Individual adjustment of oral etoposide based on unbound pharmacokinetics after the first administration appears relevant and feasible.

Clinical symptoms and sensitization to Penicillium nalgiovense are frequent among workers exposed to mould during brushing in dry sausage plants.

The CYFRA 21-1 assay detects circulating fragments of cytokeratin 19, which is a sensitive marker for the diagnosis of lung cancers, particularly squamous cell carcinomas and adenocarcinomas. Epidermis-type proteins, such as cytokeratins 1, 2, 10/11 and 14 or filaggrin, are also expressed in squamous cell carcinomas. These could also be pertinent tumor markers, ideally as sensitive as CYFRA 21-1 and more specific for squamous cell lung cancer. To verify this hypothesis, using monoclonal antibodies produced in our laboratory, we developed immunoassays specific for these proteins. After optimization, the immunoassays were evaluated in sera from 91 controls and 138 patients with squamous cell lung cancer and compared to conventional tumor markers (CEA, SCC Ag and CYFRA 21-1). Less than 14% of the sera were above the lower limit of detection of the cytokeratin- and filaggrin-specific immunoassays. Moreover, part of these positive sera were induced by the presence of interfering heterophilic antibodies in sera. Thus, in patients with squamous cell lung cancer, we confirmed the high diagnostic sensitivity of CYFRA 21-1 (55.6%) but were unable to detect significant levels of epidermis-type cytokeratins or filaggrin.