B型肝炎病毒e抗原血清廓清之發生率與決定因子及其與肝硬化及肝癌發生風險的相關性

Abstract

背景 對於慢性B型肝炎的病患來說，HBeAg血清廓清對於病程會有較好的進展，也是進行抗病毒治療時的重要目標之一。HBeAg陽性是肝硬化與肝癌的重要危險因子之一。本研究利用長期追蹤的REVEAL-HBV cohort來估計HBeAg血清廓清的發生率和決定因子，並且利用長期追蹤的HBeAg重複測量資料，探索HBeAg變化與肝硬化和肝癌之間的相關性。 材料與方法 本研究利用在1991-1992年間收集之482位HBeAg陽性，沒有C型肝炎病毒感染且進入研究時沒有肝硬化的慢性B肝患者，進行HBeAg血清廓清發生率及決定因子的分析。關於HBeAg與肝硬化和肝癌間關係的探討，我們則利用3,587名有進入研究時B肝病毒量測量值、沒有C型肝炎病毒感染且進入研究時沒有肝硬化的慢性B肝患者進行研究。我們測量了參與者進入研究時的HBeAg、ALT、病毒量和基因型；我們也檢測了他們歷次追蹤時（每6-12個月）所收集血清檢體的病毒量（針對進入研究時B肝病毒量≥104 copies/mL的病患）、HBeAg、ALT和。肝硬化乃利用腹部超音波來進行診斷，肝癌的發生則是利用連結台灣癌症登記檔與死亡檔並至確診醫院調閱病歷來做確診。本研究利用Cox proportional hazards model來估計發生HBeAg血清廓清危險因子和產生肝硬化與肝癌的粗比率對比值 (RR), 多變項調整比率對比值 (RRadj) 和95% 信賴區間(confidence interval, CI)。重複測量的分析研究則是利用time-dependent Cox model來進行。 結果 482名在進入研究時HBeAg陽性的個案，共被追蹤了3321.31個人年，共有209名個案發生HBeAg血清廓清，每一百人年的發生率為6.29。多元迴歸分析顯示進入時的年齡、基線ALT、基線病毒量、病毒基因型和發生HBeAg血清廓清有相關：40-49歲的人相對於30-39歲的人發生HBeAg血清廓清的校正比率對比值 (95% CI)是0.70 (0.50-0.99)，而大於50歲的人則是0.57 (0.38-0.85)；基線ALT 15-< 45的人相對於ALT小於15者是1.37 (0.99-1.90)，基線ALT≧ 45以上的人相對於ALT小於15者為1.96 (1.29-2.97)；HBV DNA level以最高的≧108為基準，107-<108、106-<107、105-<106、104-<105、300-<104、<300發生HBeAg 血清廓清的校正比率對比值(95% CI)分別是2.06 (1.32-3.21)、 2.13 (1.37-3.29) 、3.10 (1.92-5.00) 、9.69 (5.18-18.11)、 6.78 (3.45-13.30)、 10.55 (4.76-23.40)；基因型B相對於基因型C有2.46倍發生HBeAg血清廓清的可能性。在重複測量追蹤的研究中發現，HBV DNA level有相同的結果，病毒量重複測量值越低發生HBeAg血清廓清的可能性也越高。相對於進入研究時HBeAg陰性者，追蹤過程中HBeAg血清廓清與沒有HBeAg廓清發生肝硬化的校正比率對比值(95% CI)分別是1.22 (0.78-1.89) 和3.43 (2.62-4.48)；發生肝癌的校正比率對比值(95% CI)分別是1.37 (0.68-2.74) 和3.56 (2.29-5.55)。在HBeAg重複測量的分析中發現HBeAg重複測量陽性發生肝癌的風險是陰性者的2.86倍 (1.23-6.63)；與肝硬化間的關係，HBeAg陽性有較高的風險但未達顯著差異。 結論 慢性B肝患者若年紀較輕、基線ALT不正常、感染基因型B的病毒、進入研究時病毒量較低者會有較高的可能性發生HBeAg血清廓清。追蹤時的病毒量跟HBeAg血清廓清的發生有顯著的相關，病毒量若降低就較有機會發生HBeAg血清廓清。本研究發現在發生HBeAg血清廓清的前六年病毒量就會顯著開始下降並且伴隨者ALT的急遽上升。發生HBeAg血清廓清對於慢性B型肝炎患者來說會有較好的病程進展；根據追蹤HBeAg重複測量的狀態，也發現HBeAg陽性會較HBeAg陰性較易發生肝癌。本研究也發現在發生肝硬化、肝癌時，病患也是以HBeAg陽性佔較多數，因此更加確定HBeAg陽性是產生肝硬化、肝癌的重要危險因子之一。

Background The seroclearance of hepatitis B e antigen (HBeAg) is a favorable prognostic factor and a target of antiviral treatments for patients with chronic hepatitis B. HBeAg positivity is a risk factor for liver cirrhosis and HCC. This study aimed to estimate the incidence of HBeAg seroclearance and to identify determinants associated with HBeAg seroclearance using longitudinal follow-up REVEAL-HBV cohort; and to explore the relationship between changes of HBeAg status and the development of liver cirrhosis and HCC, using data from repeated measurements of HBeAg during long-term follow-up. Materials and methods A total of 482 chronic HBV carriers who were HBeAg-positive, anti-HCV-negative and free of cirrhosis at study entry in 1991-1992 was included in the analysis for the incidence and determinants of HBeAg seroclearance. For the analysis on the association between HBeAg status and the development of cirrhosis and HCC, a total of 3,587 participants who had adequate serum sample for HBV DNA test at enrollment, anti-HCV-negative and free of cirrhosis at study entry was used. Serum samples were tested for HBeAg, alanine aminotransferase (ALT) level, HBV DNA level and HBV genotype at entry. HBV DNA level (only for participants with baseline HBV DNA level ≥104 copies/mL), serum HBeAg status, and ALT level were checked every 6 to 12 months during follow-up. The diagnosis of cirrhosis was based on abdominal ultrasonography; HCC was ascertained through data linkage with the computerized National Cancer Registry in Taiwan and with death certificates and confirmed by medical charts review. Cox proportional hazards models were used to estimate crude rate ratio (RR), multivariate-adjusted rate ratio (RRadj) with 95% confidence interval (CI) for factors associated with HBeAg seroclearance and development of cirrhosis and HCC. Analyses on repeated measurements were done using time-dependent Cox models. Results Among 482 subjects who were HBeAg seropositive at study entery, there were 209 incident cases of HBeAg seroclearance during a total of 3321.31 person-years of follow-up. The incidence rate of HBeAg seroclearance was 6.29 per 100 person-years. Multivariate analysis identified age at entry, baseline serum ALT level, baseline serum HBV DNA levels and HBV genotype associated with the seroclearance of HBeAg. The multivariate-adjusted rate ratio (95% CI) of HBeAg seroclearance was 0.70 (0.50-0.99) and 0.57 (0.38-0.85) for age at 40-49 years and ≧50 years, respectively, compared to age at 30-39 years; 1.37 (0.99-1.90) and 1.96 (1.29-2.97) for baseline serum ALT level 15-45 U/L and ≧45 U/L, respectively, compared to ALT level <15 U/L; 2.06 (1.32-3.21), 2.13 (1.37-3.29), 3.10 (1.92-5.00), 9.69 (5.18-18.11), 6.78 (3.45-13.30), 10.55 (4.76-23.40), respectively, for baseline serum HBV DNA levels 107-<108, 106-<107, 105-<106, 104-<105, 300-<104, <300, as compared with serum HBV DNA level ≧108 copies/mL as the reference group; and 2.46 (1.77-3.42) for genotype B versus genotype C. The lower the viral load in repeated measurements of HBV DNA during follow-up, the higher the probability of HBeAg seroclearance. Compared to those who were HBeAg-seronegative at entry, the multivariate-adjusted hazard ratio (95% CI) of liver cirrhosis was 1.22 (0.78-1.89) and 3.43 (2.62-4.48), respectively, for those who cleared HBeAg and not cleared HBeAg during follow-up. The multivariate-adjusted hazard ratio (95% CI) of HCC was 1.37 (0.68-2.74) and 3.56 (2.29-5.55) for patients cleared HBeAg and not cleared HBeAg during follow-up, respectively. HBeAg positivity in repeated measurements had higher risk of HCC (HRadj: 2.86, 1.23-6.63), compared with HBeAg negativity in repeated measurements. Conclusion CHB patients with younger age, abnormal ALT levels (≧45 U/L ), genotype B HBV infection, or low serum HBV DNA level at study entry had higher incidence as well as higher chance of HBeAg seroclearance. Follow-up serum HBV DNA levels were significantly associated with HBeAg seroclearance. Decreasing serum HBV DNA levels during follow-up is more likely to clear HBeAg. Serum HBV DNA level began to decrease 6 years before HBeAg seroclearance, accompanying with a flare of ALT level. Patients who cleared HBeAg had lower risk of developing liver cirrhosis or HCC compared with those does not.