Improved identification of malignant cells in serous effusions using a small, robust panel of antibodies on paraffin-embedded cell suspensions.

OBJECTIVE: To evaluate the expression of these markers individually and to find out which markers would be the most effective in a diagnostic panel to reliably discriminate these lesions. STUDY DESIGN: Sections from cell blocks of these fluids were stained with antibodies against calretinin, EMA, HMFG-2, BerEp4, B72.3 and CEA. A preliminary diagnosis was formulated based on cytomorphologic criteria. Subsequently, results of all 6 immunocytochemical stainings were evaluated, the most effective diagnostic pane of antibodies was proposed and staining results using this panel were compared to results obtained by solely cytomorphologic evaluation and to the ultimate diagnosis. RESULTS: Additional immunocytochemical staining with the proposed panel of calretinin, EMA, HMFG-2 and CEA improved sensitivity for malignancy in general from 78% to 96% and specificity from 73% to 91%. Sensitivity for malignant mesothelioma increased from 45% to 91%, with an increase in specificity from 87% to 96%. Sensitivity for adenocarcinoma decreased slightly from 100% to 92%, but specificity increased from 86% to 100%. Overall, diagnostic accuracy increased from 76% to 94%. CONCLUSION: Immunocytochemical staining of standardized cell block reparations of serous fluid cells with a small panel of 4 antibodies significantly improves diagnostic results compared to cytomorphologic evaluation alone.

Quantitation of Ki-67 expression in the differential diagnosis of reserve cell hyperplasia vs. small cell lung carcinoma.

OBJECTIVE: To investigate whether the detection of proliferation-associated Ki-67 antigen may be of value in differentiating between reserve cell hyperplasia (RCH) and small cell lung cancer (SCLC). STUDY DESIGN: Retrospectively, 20 Papanicolaou-stained bronchial brushes or washings from 20 patients were selected. Ten were diagnosed as RCH (and had no SCLC in follow-up) and the other 10 as SCLC (histologically confirmed). All 20 Papanicolaou-stained slides were restained with the monoclonal antibody MIB1, directed against Ki-67 antigen; that simple and reliable procedure was described recently. In each specimen 5 coherent cell groups were identified, corresponding to RCH or SCLC, respectively; photographed; and studied for Ki-67 antigen expression after MIB1 staining of the slides. At least 3 cell groups remained in each specimen. The Ki-67 labeling index (LI) of the specimens was determined as the number of MIB1-positive cells divided by the total number of cells in the remaining cell groups. RESULTS: All cases of SCLC showed a mean Ki-67 LI of at least .415 (mean .684, SD .151), whereas in the cases with RCH the mean Ki-67 LI never was more than .158 (mean .048, SD .049). The difference was highly significant (P<.001, Student's t test). Linear discriminant analysis resulted in a classifier with which we were able to discriminate correctly between SCLC and RCH in 100% of the 20 bronchial brushings and washings. CONCLUSION: The results clearly demonstrate that measuring proliferative activity in Papanicolaou-stained bronchial brushings and washings by MIB1 restaining of the slides may be of great practical value in accurately discriminating RCH from SCLC. The method is simple and can be performed in any laboratory that is able to carry out immunocytochemical staining. However, an additional (prospective) study with a series of difficult cases is necessary to confirm these findings.