Dysplasia and risk of further neoplastic progression in a regional Veterans Administration Barrett's cohort.

OBJECTIVES: No published data are available on the risk of further neoplastic progression in Barrett's patients stratified by baseline dysplasia status. Our aims were to estimate and compare the risk of progression to high-grade dysplasia or cancer in groups of Barrett's patients stratified by baseline dysplasia status.
METHODS: Consecutive Barrett's cases from 1988-2002 were identified via pathology databases in a regional VA health-care system and medical record data were abstracted. The risk of progression to high-grade dysplasia or cancer was measured and compared in cases with versus without low-grade dysplasia within 1 yr of index endoscopy using survival analysis.
RESULTS: A total of 575 Barrett's cases had 2,775 patient-years of follow-up. There were 13 incident cases of high-grade dysplasia and two of cancer. The crude rate of high-grade dysplasia or cancer was 1 of 78 patient-years for those with baseline dysplasia versus 1 of 278 patient-years for those without (p= 0.001). One case of high-grade dysplasia in each group underwent successful therapy. One incident cancer case underwent successful resection and the other was unresectable. Two cases with high-grade dysplasia later developed cancer, one died postoperatively, the other was unresectable. When these two cases were included (total of four cancers), the crude rate of cancer was 1 of 274 patient-years for those with baseline dysplasia versus 1 of 1,114 patient-years for those without.
CONCLUSIONS: In a large cohort study of Barrett's, incident malignancy was uncommon. The rate of progression to high-grade dysplasia or cancer was significantly higher in those with baseline low-grade dysplasia. These data may warrant reevaluation of current Barrett's surveillance strategies.