[A thirty-one year prospective follow-up program on the HBsAg carrier state and primary liver cancer in Qidong, China].

OBJECTIVE

To study the relationship between hepatitis B virus (HBV) and primary liver cancer (PLC), and to assess the outcome of PLC in the carriers of HBsAg.

METHODS

General population over age of 15 from a community in Qidong was screened for HBsAg, anti-HBs, and alanine transaminase (ALT) in 1976, and followed-up thenceforth. From January 1, 1977 through December 31, 2007, 12 351 people were enrolled in the cohort, and their occurrence, outcome of PLC and other cancers, together with all the withdrawals due to death were linked to and checked with database from Qidong Cancer Registry and Qidong Vital Registry programs.

RESULTS

The total observed person-years (PYs) were 355 305.0. One hundred and seventy-three PLC cases were identified among the HBsAg carriers, with an incidence of 361.55 per 100 000 PYs, while PLC cases were only 95 for the non-carriers, with an incidence of 30.90 per 100 000 PYs. The overall relative risk (RR) was 11.70 (95%CI: 9.06 - 15.19), with RR 12.30 for men and 10.46 for women. HBsAg carriers had high incidence at each age group, compared with the non-carriers for both men and women. Data from cross-over analysis showed that the incidence rates of PLC for the sub-cohorts of female non-carriers, male non-carriers, female carriers, and male carriers were 1.00, 3.07, 10.46, and 37.76, respectively. The cumulative rates of PLC in the 4 groups were 0.86%, 2.73%, 10.22%, and 34.19%, respectively. Results from non-conditional logistic regression model showed that the gender (male), age, HBsAg(+), and ALT(+) were risk factors for the development of PLC while anti-HBs(+) demonstrated a protective effect. No relationship was found among carriers and non-carriers for cancer sites such as lung, stomach, esophagus, intestine, pancreas, breast, cervix, bladder, and lymphoma, brain tumor, or leukemia.

CONCLUSION

Causation and its strength, together with specificity and persistency of the association were confirmed from this HBsAg-related cohort study in the general population in Qidong. Intervention measures on HBV should be highlighted for the control of PLC among the HBV infected individuals.