慢性B型肝炎病毒感染族群非酒精性脂肪肝與慢性腎臟病之長期追蹤研究

Abstract

背景 非酒精性脂肪肝是目前全球導致肝功能異常的常見原因，並且它被認為是一種多系統的疾病，不僅會影響肝臟相關器官，甚至會影響肝外器官，包括腎臟。此外，慢性B型肝炎病毒是主要導致肝臟疾病的原因，而某些形態的腎臟功能不全也被認為是與B型肝炎病毒有關的；但慢性B型肝炎病毒與非酒精性脂肪肝之間呈負相關。 目的 本研究以回溯性縱貫性研究探討慢性B型肝炎病毒感染之男性其非酒精性脂肪肝與慢性腎臟病的長期關係；此外，由於腎絲球過濾率在長期追蹤中隨時間之進展趨勢與非酒精性脂肪肝之關聯仍不明確，因此我們進一步比較不同腎絲球過濾率trajectory patterns其非酒精性脂肪肝與其他危險因子之情形。 材料與方法 本研究以過去收集之公務人員門診中心（government employee central clinic, GECC）世代為研究對象，其參與者共有2904名慢性B型肝炎感染之男性。本研究以1989-1992年間進入研究之40歲以上，並在1993-1994年間進行第一次脂肪肝測量，且於1997–2004年間有至少兩次追蹤資料者為研究世代，最終共納入1071名慢性B型肝炎病毒感染之男性進入研究。有無脂肪肝之情形是以腹部超音波檢查之肝臟超音波來診斷；分別以CKD-EPI、台灣CKD-EPI、MDRD以及台灣MDRD四種公式估算腎絲球過濾率(eGFR)，作為判定腎臟功能之指標。以eGFR＜60 mL/min/1.73 m²定義為CKD；CKD分期則是以KDOQI提出之腎臟病指引共分為6期。以Generalized estimating equation及Generalized linear mixed model對長期追蹤資料進行分析。在追蹤期間，腎絲球過濾率的trajectory patterns是利用latent-class trajectory model進行其趨勢的模擬。並以Kaplan-Meier方法來檢驗CKD惡化的累積發生率。 結果 在基線時，有55.3%的參與者有脂肪肝的情形。分別以CKD-EPI、台灣CKD-EPI、MDRD以及台灣MDRD公式計算eGFR，eGFR＜60 mL/min/1.73 m²之比率分別為4.9%、13.2%、8.1%及19.9%。年齡、空腹血糖以及基線時的腎絲球過濾率會與慢性腎臟病的發生有關，然而，非酒精性脂肪肝與慢性腎臟病之間則不存在相關。不論以任一種公式估算之腎絲球過濾率，我們發現了三種不同的軌跡：有30.5%–52.5%的參與者持續有較高的腎絲球過濾率，其腎絲球過濾率為>70 mL/min/1.73 m²；43.9%–65.0%的參與者之腎絲球過濾率較穩定的介於60–70 mL/min/1.73 m²；3.6%–4.5%的參與者則有腎絲球過濾率下降較快的情形，而其在最初的腎絲球過濾率為＜60 mL/min/1.73 m²。在腎絲球過濾率下降較快的組別中，年齡最大、肥胖的比率較高，而有糖尿病、高血壓及心血管疾病病史的比率也都最高；然而，三組中非酒精性脂肪肝的情形則沒有顯著差異。此外，以慢性腎臟病期別定義腎臟功能惡化的情形，有糖尿病的參與者發生腎臟功能惡化顯著高於非糖尿病者，其六年累積發生率分別為51.9%及36.0%；然而，有無脂肪肝與腎臟功能惡化間依然沒有發現顯著相關。 結論 本研究之結果發現，對於慢性B肝患者，NAFLD可能不是發生CKD或CKD惡化的獨立危險因子。建議未來可以納入非慢性B肝患者進一步進行探討。

Background Non-alcoholic fatty liver disease (NAFLD) is a common cause of chronic liver disease worldwide. Furthermore, NAFLD is a multisystem disease, which does not only affect liver, but also affects extra-hepatic organs, including kidney. Besides, chronic hepatitis B virus (HBV) infection is a main cause of liver disease and has been associated with various forms of renal impairment. Paradoxically, there is an inverse association between chronic HBV infection and NAFLD. Specific Aims We conducted a retrospective longitudinal study of NAFLD and chronic kidney disease (CKD) based on a long-term cohort study of chronic HBV carriers. We aimed to investigate the longitudinal relationship with CKD for the presence of NAFLD. In addition, because the association between progressive nature of eGFR across long-term follow-up and NAFLD remains unclear, we also compared trajectory patterns of eGFR between individuals with and without NAFLD, with the incorporation of other risk factors. Materials and Methods Prospective data were from the Government Employees’ Central Clinics cohort study of 2904 male HBV carriers recruited between 1989 and 1992. A total of 1071 participants who were aged above 40, with initial ultrasonography measurement of fatty liver during 1993-1994, underwent at least 2 times of follow-up examinations between 1997-2004 were included. NAFLD was defined as the presence of ultrasonographic fatty liver. CKD was defined as eGFR＜60 mL/min/1.73 m², estimated by four equation: CKD-EPI, Taiwan’s CKD-EPI, MDRD, and Taiwan’s MDRD. CKD stages were defined according to KDOQI guidelines. Generalized estimating equation and generalized linear mixed model were used to examine longitudinal data. Trajectories of eGFR over follow-up period were determined using latent-class trajectory modeling. The Kaplan-Meier method was used to investigate the cumulative incidence rate. Results At baseline, 55.3% of the participants had NAFLD. The proportion of eGFR＜60 mL/min/1.73 m² were 4.9%, 13.2%, 8.1% ,and 19.9%, respectively, according to the formula by CKD-EPI, Taiwan’s CKD-EPI, MDRD, and Taiwan’s MDRD. Age, fasting glucose, and baseline eGFR were associated with the risk of CKD, while no association was detected for NAFLD. We identified three distinct trajectory patterns of eGFR-tracked over time, irrespective of different formulas: 30.5%-52.5% of subjects had persistently high level of eGFR of >70 mL/min/1.73 m², 43.9%-65.0% of subjects had stable levels of eGFR of 60-70 mL/min/1.73 m², and 3.6%-4.5% of subjects had accelerated decline of eGFR from level of eGFR<60 mL/min/1.73m². Subjects with accelerated decline of eGFR trajectory were more likely to be older, more obese, and have a history of diabetes, hypertension, or cardiovascular disease. However, no association was observed for fatty liver. Besides, the 6-year cumulative incidence of renal function deterioration, which was defined as a drop in CKD stages, was significantly higher in diabetes subjects (51.9%) compared to non-diabetes subjects (36.0%). Again, no association with renal function deterioration for NAFLD was observed. Conclusion We found that NAFLD might not be an independent risk factor of CKD and progressive CKD in chronic HBV carriers. Further studies are warranted to examine people without HBV.