studies

HEPATOLOGY, VOLUME 62, NUMBER 1 (SUPPL) AASLD ABSTRACTS 987A
mary outcome was HBV reactivation defined as an increase
in HBV DNA levels from baseline or re-emergence of HBsAg+
when previously negative. The odds ratio (OR) of reactivation
with versus without HBsAb+ was estimated with random-effects
model meta-analyses. Results: Fourteen observational studies
with 836 patients (range 8-354 with 7 patients receiving antiviral
prophylaxis) met inclusion criteria. There were 9 prospective
and 5 retrospective studies; 11 were from Asia and 3 from
Italy; 10 included only lymphoma and 4 had mixed hematologic
malignancies including leukemia and multiple myeloma.
When reported, the median patient age across studies was 62
years (range 18-90) with 334 males and 261 females. Chemotherapy
regimens varied across studies with rituximab in 8
studies. The overall pooled OR was 0.20 (95% CI 0.11–0.35;
with no heterogeneity, I 2 =0%) indicating a protective effect
of HBsAb+ on the risk of reactivation versus HBcAb+ alone.
The absolute risk of reactivation with HBcAb+ alone was 17%
(95% CI 10-25%). Similar results were found when the analysis
was limited to rituximab-based chemotherapy (OR 0.17, 95%
CI 0.08–0.34) and lymphoma (OR 0.17, 95% CI 0.09–0.32).
Sensitivity analysis found a significant protective effect of
HBsAb+ in 9 prospective studies but not in 5 retrospective studies.
The ORs continued to remain significant when analyzed by
geographic region (Asia OR 0.22, 95% CI 0.12–0.40; Italy
OR 0.07, 95% CI 0.01–0.54). Conclusions: In patients with
resolved HBV receiving chemotherapy for hematologic tumors,
HBsAb+ decreases the risk of reactivation. Our results support
the need for future studies examining the effect of HBsAb titers
and booster vaccinations prior to chemotherapy in this patient
population.
Disclosures:
The following authors have nothing to disclose: Sonali Paul, Aaron Dickstein,
Akriti P. Saxena, Norma Terrin, Kathleen Viveiros, Ethan M. Balk, John B. Wong
and the death certification system was performed. Cox proportional
hazards models were used to estimate the multivariate-adjusted
hazard ratio of developing LC and HCC. Results:
Incidence rates of LC per 100,000 person-years were 341.2,
565.5, 654.1, 920.0, 1477.9 for serum anti-HBc levels of