Objective: To compare the results of numerical chromosomal changes in chromophobe renal cell carcinoma obtained from whole tissue sections with those of tissue microarrays.
Study design: Standard sections and tissue microarrays constructed from formalin-fixed and paraffin-embedded chromophobe renal cell carcinomas from 6 patients were subjected to interphase fluorescence in situ hybridization (FISH) using centromeric probes for chromosome 1, 2, 6, 10 and 17. Tissue microarrays were constructed with 3 cores per tumor and 2 cores of normal renal tissue controls.
Results: Whole sections of chromophobe renal cell carcinoma showed multiple chromosomal abnormalities in 3 out of 6 cases (case 1, 2, 5). In 2 cases all 5 chromosomes were lost in both whole and tissue microarray cores (cases 2 and 5), and for 1 case (case 1) losses of chromosomes 2, 10 and 17 were detected on both whole and tissue microarray cores. In another 2 cases there were 2 fluorescence signals for chromosomes 1 in the majority of malignant cells (cases 3 and 6), loss of chromosome 6 (case 3), loss of chromosome 2 (case 6) and a mosaic pattern with nuclei showing a mixture of signal losses and gains for the other chromosomes. The remaining tumor did not show abnormalities (case 4). When 2 core biopsies per tumor were examined, the level of substantial concordance of results ranged from 92% to 98%.
Conclusion: In this study we have demonstrated that tissue microarrays are a valid substitute for whole tissue sections in large cohort studies of chromophobe renal cell carcinoma when FISH analysis is undertaken. Concordance of results was improved when the number of analyzed cores was increased from 2 to 3, at which point a concordance index ranging from substantial to almost perfect was observed.