BACKGROUND: Flow-cytometric analysis of proliferation index (PI) has potential use in predicting prognosis in malignancy. Its relevance to heterogeneous tumors has not been conclusively studied. In nephroblastoma, where the epithelial components are considered more differentiated than others, potentially different PIs may exist within a single lesion based on the inverse relation between differentiation and proliferation. Proliferating cell nuclear antigen (PCNA) and MIB-1 (Ki-67 equivalent antigen) demonstration in histologic sections by immunoperoxidase methods may allow for determination of PI in relation to tissue type. EXPERIMENTAL DESIGN: A consecutive unselected series of 8 pediatric nephroblastoma patients was used to study the relation between PI and histologic differentiation as established by flow-cytometric analysis of nuclear suspensions prepared from formalin-fixed and paraffin-embedded tissue and by PCNA/MIB1 staining of parallel histologic sections. PI by PCNA/MIB1 was established using 5-microns paraffin sections, immunoperoxidase, and quantification procedures detailed in the literature. The mitotic index (MI) of tissue components was separately assessed using 5-microns hematoxylin and eosin-stained sections and counting procedures detailed in the literature. RESULTS: The 8 lesions showed a PI of 4 to 20% as determined by flow cytometry. Using PCNA staining, the epithelium showed a mean PI of 55.5% (range 40 to 80%), that was significantly higher (p < 0.001, Wilcoxon's two-tailed rank sum test) than blastema (mean PI: 34.1%, range 17.5 to 76.5%) and stroma (mean PI of 14.9%, range 5 to 24%, p < 0.001, Wilcoxon's two-tailed rank sum test). Although, probably due to tissue antigen preservation, acceptable MIB-1 staining was not achieved in all lesions (5 of 8), the results, although generally with lower labeling indices, confirmed the PCNA findings. The relative MI of epithelial components was higher than that of stroma and blastema in keeping with the immunocytochemical findings. In 6 of 8 cases, the PI by flow-cytometric analysis was lower than the lowest value for the PI (labeling index) of an individual tissue type found by PCNA or MIB staining. CONCLUSIONS: The differences found between PI of the different tissue components in nephroblastoma are difficult to understand if the epithelial components (with the highest PI values) are considered as differentiation products from the other components of the lesion. The relation between PIs as determined by PCNA/MIB-1 analysis/mitotic index, for the three components and the PI as established by flow cytometry is not simply explained by the relative volume of the tissue components.
BACKGROUND: Flow-cytometric analysis of proliferation index (PI) has potential use in predicting prognosis in malignancy. Its relevance to heterogeneous tumors has not been conclusively studied. In nephroblastoma, where the epithelial components are considered more differentiated than others, potentially different PIs may exist within a single lesion based on the inverse relation between differentiation and proliferation. Proliferating cell nuclear antigen (PCNA) and MIB-1 (Ki-67 equivalent antigen) demonstration in histologic sections by immunoperoxidase methods may allow for determination of PI in relation to tissue type. EXPERIMENTAL DESIGN: A consecutive unselected series of 8 pediatric nephroblastomapatients was used to study the relation between PI and histologic differentiation as established by flow-cytometric analysis of nuclear suspensions prepared from formalin-fixed and paraffin-embedded tissue and by PCNA/MIB1 staining of parallel histologic sections. PI by PCNA/MIB1 was established using 5-microns paraffin sections, immunoperoxidase, and quantification procedures detailed in the literature. The mitotic index (MI) of tissue components was separately assessed using 5-microns hematoxylin and eosin-stained sections and counting procedures detailed in the literature. RESULTS: The 8 lesions showed a PI of 4 to 20% as determined by flow cytometry. Using PCNA staining, the epithelium showed a mean PI of 55.5% (range 40 to 80%), that was significantly higher (p < 0.001, Wilcoxon's two-tailed rank sum test) than blastema (mean PI: 34.1%, range 17.5 to 76.5%) and stroma (mean PI of 14.9%, range 5 to 24%, p < 0.001, Wilcoxon's two-tailed rank sum test). Although, probably due to tissue antigen preservation, acceptable MIB-1 staining was not achieved in all lesions (5 of 8), the results, although generally with lower labeling indices, confirmed the PCNA findings. The relative MI of epithelial components was higher than that of stroma and blastema in keeping with the immunocytochemical findings. In 6 of 8 cases, the PI by flow-cytometric analysis was lower than the lowest value for the PI (labeling index) of an individual tissue type found by PCNA or MIB staining. CONCLUSIONS: The differences found between PI of the different tissue components in nephroblastoma are difficult to understand if the epithelial components (with the highest PI values) are considered as differentiation products from the other components of the lesion. The relation between PIs as determined by PCNA/MIB-1 analysis/mitotic index, for the three components and the PI as established by flow cytometry is not simply explained by the relative volume of the tissue components.