[547] Outcome Analysis of 15-Year Follow-Up in a Regional Veterans Affairs Barrett's Esophagus Cohort with Respect to the Updated 2008 Guidelines for Surveillance
S Askari, KK Dunning, BR Lindgren. Veterans Affairs Medical Center, Minneapolis, MN; University of Minnesota, Minneapolis, MN
Background: The grade of dysplasia determines surveillance interval in Barrett's esophagus (BE). American College of Gastroenterology & Practice Parameters Committee published 2008 guidelines: 3-yr surveillance with 2 biopsies (bxs) (within 1 year) negative for dysplasia (NEG); 1-yr surveillance until no dysplasia in 2 bxs when low-grade dysplasia (LGD) is the highest grade on repeat bx (within 6 months). We performed an outcome analysis of our BE cohort to investigate same for our patients (pts). Design: BE cases between 7/90 and 12/95 were identified. Records were retrospectively reviewed through 6/06 to select cases NEG on 1st bx and with at least 1 follow-up bx. Bxs were categorized as NEG, indefinite (INDF), LGD, high-grade (HGD), or carcinoma (CA). Event was defined as progression to HGD/CA. Time-to-event (TTE) analysis was performed by comparing event free rates between group 1-NEG on 2nd bx and group 2-INDF/LGD on 2nd bx. Kaplan-Meier method with log-rank test was used. Results: A total of 699 bxs occurred in 131 pts with 859 pt-yr follow-up (1.3 events/100 pt-yrs). 8 of 11 events occurred in group 1 (N=113) with 2 of 8 occurring in <3 years of index bx.
 Five of 8 events occurred without INDF/LGD with mean period of 3.4 yrs (R=0.6-7.6) from index bx. The mean time interval between first two NEG bxs was 1.6 yrs (R=0.003-10.04). In 32 of 113 cases, a subsequent diagnosis of INDF/LGD was made with mean of 4 (R=2-13) interval NEG bxs over 5.5 yrs (R=0.5-12.5). In group 2 (N=16), a single event of HGD, identified on the 7th bx, occurred 8.5 yrs from index bx following 2 consecutive NEG bxs.
Table 1: Time-to-event Analysis by 2nd Bx; p=0.942 (N=129)* | Event Free Rate | | 1-Year | 3-year | 5-Year | 10-Year | 15-Year | | NEG (N=113) | 0.99 | 0.98 | 0.94 | 0.89 | 0.89 | | INDF/LGD (N=16) | 1.00 | 1.00 | 1.00 | 0.83 | 0.83 | *Two patients with HGD on 2nd bx were excluded.
Conclusions: Time-to-event rate is unpredictable and selected BE-NEG pts may benefit from annual surveillance endoscopy. Continued surveillance may be considered for selected LGD pts despite consecutive NEG bxs. Category: Gastrointestinal
Wednesday, March 11, 2009 1:00 PM
Poster Session VI # 96, Wednesday Afternoon
|