FLOW CYTOMETRIC DNA ANALYSIS OF ULCERATIVE-COLITIS USING PARAFFIN-EMBEDDED BIOPSY SPECIMENS - COMPARISON WITH MORPHOLOGY AND DNA ANALYSIS OF FRESH SAMPLES

Objectives: The detection of aneuploidy in colonic mucosa by now cytom etric DNA analysis has been advocated as an indicator of high risk for ulcerative colitis (UC) patients developing colon carcinoma. To date, studies have primarily utilized fresh tissue and have had two limitat ions: a significant number of possible false-positive findings (aneupl oidy in the absence of detectable dysplasia) that may be due to DNA de gradation, and the inherent inability to perform retrospective studies . The latter has compromised the adequate assessment of no cp cytometr ic DNA analysis for its clinical utility in UC patients. Our objective was to attempt to overcome the limitations that have been present in DNA analysis by flow cytometry. Methods: After having established, in our laboratory, an optimal method that allowed reliable DNA analysis o n paraffin-embedded mucosal biopsy specimens, we conducted three separ ate studies to address the above problems associated with DNA analysis with fresh colonic samples: I) comparison of DNA analysis between fre sh and paraffin-embedded colonic mucosal samples from UC patients with out dysplasia, 2) correlation between morphology and DNA ploidy on par affin-embedded tissues showing no dysplasia and various degrees of dys plasia, and 3) sequential analysis of fresh, normal colonic mucosal sa mples to further evaluate possible causes of false aneuploidy. Results : We observed I) that there is discordance in DNA ploidy between paraf fin-embedded and fresh samples showing no dysplasia in that aneuploidy was found in 33/46 (72%) of fresh samples, whereas 3/40 (7.5%) were a neuploid by biopsies from the corresponding anatomic sites. 2) There i s excellent correlation between dysplasia and DNA ploidy results with paraffin-embedded tissue, i.e., none of 16 samples negative for dyspla sia, none of seven samples indefinite for dysplasia, and seven of eigh t samples positive for dysplasia were aneuploid. 3) DNA degradation pr oduced a spurious, near-diploid aneuploid peak in a normal colonic muc osal sample when it was left in saline more than 1 hr before analysis. Conclusions: The above-described results demonstrate that performing flow cytometric DNA analysis with formalin-fixed paraffin-embedded bio psy samples is feasible and that this technique may provide more relia ble ploidy results than does the use of fresh samples, when rapid refr igeration and/or freezing of the fresh samples cannot be accomplished consistently, and will permit retrospective DNA ploidy studies assessi ng risk of cancer in UC patients.