Combined Digital Image Cytometry with Fluorescence in Situ Hybridization of Brush Cytology Identifies Patients with Barrett's Esophagus at Increased Risk for Neoplastic Progression

Abstract

Background: We have previously shown that Digital Image Cytometry (DICM) of brush cytology in Barrett's Esophagus (BE) identifies additional high risk patients with aneuploid and intermediate DNA-pattern. Similarly, flowcytometry of biopsy specimens with aneuploidy or increased 4N combined with p53 loss of heterozygosity (LOH) may identify patients at risk of progression to cancer. Methods: This prospective cohort study aimed to assess specific genetic changes of p53 by Fluorescence in situ Hybridization (FISH) as compared to DICM in cytologic specimens of patients with BE: a) no dysplasia (No D) b) low-grade dysplasia (LGD), c) high-grade dysplasia (HGD) and d) early cancer (Ca). We examined 58 brush specimens from 43 patients (27 male; mean age 64 yrs) with BE by DICM on Feulgen restained cytology smears. DNA patterns were defined as diploid, intermediate and aneuploid. Fifty cytologies (94%) were further processed by two-color FISH using a locus-specific probe for 17p13.1 (p53 gene) with an enumeration probe. A pathological FISH probe was defined as loss of p53 (LOH) in more than 6.7% of cells (> mean + 3 SD of diploid controls). Results: 15/58 (26%) of DICM examinations showed aneuploidy with p53 LOH by FISH in 10/11 (91%). 13/58 (22%) intermediate DNA-patterns were diagnosed with p53 LOH in 12/13 (92%). 30/58 (52%) brush cytologies were diploid with p53 LOH by FISH in 5/30 (17%). Thus, pathological DICM (aneuploid or intermediate pattern) correlated with p53 LOH by FISH in 23/25 (92%) of brush cytologies, yielding good agreement beyond chance (Cohen's kappa of 0.74). The combined analysis selected a high-risk constellation with p53 LOH in 21/42 (50%) of patients negative for HGD or Ca. Conclusion: Combined DICM with p53 LOH by FISH detects patients with BE without HGD or Ca who may be at increased risk for neoplastic progression according to current literature. Provided that prospective studies support an high risk prognosis of FISH positive individuals with non-dysplastic histology, brush cytology may be a promising additional diagnostic tool guiding surveillance patterns for BE.