病毒量於慢性B型肝炎帶原者肝臟疾病進展之重要性

Abstract

背景 近來的研究發現測定B型肝炎病毒量對於慢性B型肝炎病人的處置有愈發顯著之重要性，然而目前對於病毒量對於肝臟疾病各階段進展之影響仍未臻完全。 研究方法 我們在台灣七個鄉鎮於1991-1992年間開始追蹤之前瞻性世代中選取了四個代表性族群進行研究。這四個族群均為B型肝炎表面抗原陽性及C型肝炎抗體陰性，但有著下列不同代表性特徵，他們分別是：一) 不活動B型肝炎病毒帶原者，於納入追蹤初始時具有B型肝炎e抗原陰性、病毒量少於10000copies/mL、血清ALT值正常、同時無肝硬化或肝癌的對象；二) 持續血清ALT值正常者（PNALT），所有對象為1991-1999追蹤期間維持血清ALT 值正常且未發生肝硬化或肝癌者；三) 慢性肝炎族群，對象包括所有於追蹤起始前兩年內血清ALT值維持正常但於後續追蹤出現持續或間歇性異常、且未發生肝硬化或肝癌者；四) 肝硬化族群，對象包括所有於追蹤起始前兩年內血清ALT值維持正常但於後續追蹤發生肝硬化、且未發生肝癌者。 另外也於原本世代中選取了所有B型肝炎表面抗原陰性及C型肝炎抗體陰性、且具有與不活動B型肝炎病毒帶原者臨床上相似肝臟特徵者以供在一)中作為對照。對於二)~四)族群的對象，分別在符合族群選取定義時的追蹤採集血清檢體中量定B型肝炎病毒量。研究中的出象（outcome）包括了慢性肝炎、肝硬化、肝癌、肝臟相關死亡、以及末期肝臟疾病（ESLD，定義為發生肝硬化、肝癌或肝臟相關死亡事件中一件或一件以上者），分別依不同的族群而選定適當的出象進行分析。出象的發生是由追蹤複檢之血清ALT值、腹部超音波檢查結果、以及連結國家癌症登錄與死亡資料庫而確認。肝臟疾病各階段進展速率，係由各族群陳現的資料中據以推估。危險預測因子之多因子校正危險比率（multivariate-adjusted hazard ratio）則由Cox回歸模式中引導而得。 結果 一) 在平均追蹤了13.1±1.8年中，一共有51例肝癌及62例肝臟相關死亡新發個案，肝癌發生率及肝臟相關死亡率，不活動B型肝炎病毒帶原者為每百人年有0.06及0.04例，而對照組為每百人年有0.02及0.02例。相較於對照組，不活動B型肝炎病毒帶原者發生肝癌及肝臟相關死亡之危險比率（95%信賴區間），多因子校正的結果為4.6 (2.5–8.3) 及 2.1 (1.1–4.1)。此外高齡及具有飲酒習慣也是具有預測肝癌發生之危險因子。二) 在持續血清ALT值正常者之族群中，於平均追蹤的4.9±0.3年中，一共有96例進展發生慢性肝炎。慢性肝炎之發生率，依不同之病毒量（copy/mL）而論，在B型肝炎病毒e抗原陽性者中每百人年分別為－3.0 (<105)、10.0 (105-<108)、11.8 (≥108)例，而對於e抗原陰性者則為－2.0 (<300)、 1.8 (300-<104)、 4.4 (104-<106)、 以及11.3 (≥106)例等。其不同病毒量下相應之慢性肝炎發生危險比率（95%信賴區間），多因子校正之後，在e抗原陽性者而言（以<105 copies/mL者為基準時）為2.6 (0.9-7.3) / (105-<108) 及 3.3 (2.6-6.5) / (≥108)；而對於e抗原陰性者（以<300 copies/mL為基準時），則分別為0.9 (0.5-1.6) / (300-<104)、 2.1 (1.2-3.5) / (104-<106)、以及 4.8 (2.1-11.0) / (≥106)等，均有著隨病毒量增加的現象。三) 對於慢性肝炎族群而言，在平均5.1±2.1年中共有21例末期肝臟疾病發生。末期肝臟疾病之發生率，對於e抗原陰性者而言每百人年有1.5例，而e抗原陽性者則有3.6例。整體而言，當病毒量逾越105 copies/mL時（以低於103 copies/mL為基準時），其多因子校正後之危險比率方有統計上顯著之增加，其數值（95%信賴區間）為9.2 (1.2-70.8)。病毒量之外，高齡者也是預測慢性肝炎發生之重要因子。四) 在肝硬化族群裡，在平均追蹤的6.1±3.7年裡，共有25例肝癌及23例肝臟相關死亡個案。肝癌發生率及肝臟相關死亡率而言，每百人年，e抗原陽性者為3.1及2.5例，而e抗原陰性者為3.0及2.9例。肝癌之發生，高齡為唯一具顯著性之預測因子，而與病毒量並無顯著相關。然而，對於肝臟相關之死亡而言，當病毒量高於104 copies/mL以上時（相較於低於104 copies/mL）仍有獨立於肝癌發生影響之外的顯著預測能立，其危險比率（95%信賴區間），多因子校正的結果為5.6 (1.6-2.6)。 結論 不活動B型肝炎病毒帶原者，相較於非B型肝炎病毒帶原者而言，仍具有發生肝癌及肝臟相關死亡之實質危險性；只有當B型肝炎病毒表面抗原消失以後，所增加之危險性方可消滅。至於慢性B型肝炎病毒感染自然史中肝臟疾病之各個階段，包括持續肝功能正常、慢性肝炎、乃至肝硬化等階段而言，較高之病毒量都是導致肝臟疾病快速惡化之重要因素。

Background The determination of viral load is becoming of greater importance in the management of chronic hepatitis B virus infected persons. The implication of the viral load on liver disease progression for chronic hepatitis B virus infection remains incomplete. Methods Four HBsAg-positive-anti-HCV-negative subsets were selected from the prospective cohort in Taiwan enrolled in 1991-1992. 1) Inactive HBV carrier subcohort, characterized by HBeAg-negative, a viral load <10000 copies/mL and normal serum ALT level, free of liver cirrhosis (LC) and hepatocellular carcinoma (HCC), was selected at baseline. An HBsAg-negative-anti-HCV-negative control subcohort with similar clinical liver feature was selected for comparison. 2) Persons with persistently serum ALT levels, free of LC and HCC through 1991-1999, were selected as the persistently normal serum ALT (PNALT) subcohort. 3) Chronic hepatitis (CH) subcohort comprised persons who developed chronic hepatitis from the persons with PNALT, free of LC and HCC in the first two years of cohort enrollment. 4) Liver cirrhosis (LC) subcohort included persons who developed liver cirrhosis from the persons with PNALT, free of liver cirrhosis and HCC in the first two years of cohort enrollment. HBV DNA levels were determined through follow-up serum samples at each subcohort entry. The outcomes included CH, LC, HCC, liver-related death (LD), and end-stage liver disease (ESLD, defined as at least one event of LC, HCC and LD). They were determined through follow-up tests and data-linkage with the computerized National Cancer Registry and Death Certification profiles. The disease progression rates were estimated. Multivariate-adjusted hazard ratios (HRa) of risk predictors were derived by the Cox regression models. Results 1) During a mean (±std) follow-up of 13.1 (±1.8) years, there were 51 HCC and 62 LD cases－the incidence rates (/100 person-years) of HCC and LD were 0.06, 0.04 for inactive HBV carriers, and 0.02, 0.02 for controls. The HRa (95%CI) for the inactive HBV carriers, compared to controls, to develop HCC and LD were 4.6 (2.5–8.3) and 2.1 (1.1–4.1). Older age and an alcohol drinking habit were additional independent predictors of HCC. 2) In PNALT subcohort, there were 96 CH cases during a mean (±std) follow-up of 4.9 (±0.3) years, the incidence rates of CH (/100 person-years) by viral load (copy/mL) were 3.0 (<105), 10.0 (105-<108) and 11.8 (≥108) for HBeAg- positive persons and 2.0 (<300), 1.8 (300-<104), 4.4 (104-<106), and 11.3 (≥106) for HBeAg-negative persons. The HRa (95% CI) by viral load (copy/mL) of developing CH for HBeAg-positive persons were 2.6 (0.9-7.3) (105-<108) and 3.3 (2.6-6.5) (≥108) (<105 copies/mL as reference); while for HBeAg-negative persons, they were 0.9 (0.5-1.6) (300-<104), 2.1 (1.2-3.5) (104-<106) and 4.8 (2.1-11.0) (≥106) (<300 copies/mL as reference). 3) In the CH subcohort, there were 21 newly developed ESLD cases during a mean (±std) follow-up of 5.1 (±2.1) years. The overall incidence rates (/100 person-years) of ESLD for HBeAg-negative and HBeAg-positive persons were 1.5 and 3.6. A viral load above 105 copies/mL (vs. <103 copies/mL) overall predicted ESLD with a HRa of 9.2 (1.2-70.8). Older age was an additional risk predictor. 4) In the LC subcohort, there were 25 HCC and 23 LD cases during a mean (±std) follow-up of 6.1 (±3.7) years. The incidence rates (/100 person-years) of HCC (LD) for HBeAg- negative and HBeAg-positive persons were 3.1 (2.5), and 3.0 (2.9). Older age was the only important risk predictor of HCC. Independent of HCC, a viral load above 104 copies/mL (vs. <104) remained important in predicting LD with a HRa of 5.6 (1.6-2.6). Conclusions Carriers of inactive HBV, compared with controls, have a substantial risk of HCC and LD, which risk would be diminished after loss of HBsAg. High viral loads lead to more rapid liver disease progression, state by state, throughout the natural history of chronic hepatitis B virus infection.