乙型肝炎病毒相关性肾炎药物治疗的Meta分析

目的 乙型肝炎病毒相关性肾炎（HBV-GN）是最常见的继发性肾小球疾病之一，目前尚无统一的治疗方案，本研究对HBV GN药物治疗的疗效进行Meta分析，以期为临床治疗提供参考。方法检索PubMed、EMBASE、Cochrane Library、中国生物医学文献数据库、中国期刊全文数据库和学术会议论文汇编、学位论文集，收集关于HBV-GN药物治疗的中、英文文献。应用国际Cochrane中心推荐的方法进行文献质量评价。使用STATA 9.0和RevMan 4.2软件进行Meta分析，计算治疗组与对照组缓解率的相对危险度（RR）及其95%CI，并对不能进行Meta分析的数据进行描述性分析，系统评价各种治疗方案的疗效。结果共检出中、英文文献978篇，依据纳入和排除标准，最终有9篇文献（英文8篇，中文1篇）纳入本研究，其中仅1篇为3分的RCT，其余均为队列研究。抗病毒治疗的疗效分析共纳入6项研究（1项为单中心RCT，5项为队列研究），其中4项研究以重组人干扰素-α（rIFN-α）进行治疗，1项以拉米呋啶进行治疗，另1项包括了上述两种治疗，纳入患者159例（抗病毒治疗组72例，对照组87例）。临床应答的Meta分析显示：RR=1.69，95%CI：1.08~2.65，抗病毒治疗组蛋白尿缓解率（91.0%）高于对照组（56.0%），两组差异有统计学意义（P=0.02）；儿童亚组分析显示抗病毒治疗组蛋白尿缓解率（86.7％）与对照组（61.1％）差异无统计学意义。病毒学应答的Meta分析显示：RR=6.44，95%CI：3.11~13.35，抗病毒治疗组的HBeAg阴转率（59.7%）高于对照组（8.33%），两组差异有极显著统计学意义（P<0.000 01）；儿童亚组分析结果一致，且一致性检验发现血中HBeAg的清除与蛋白尿缓解呈显著相关（kappa=0.285, P=0.002）。糖皮质激素治疗的疗效分析共纳入5项研究，均为队列研究，纳入患者76例（激素治疗组37例，对照组39例）。临床应答的Meta分析显示：RR=1.45，95%CI：0.68~3.11，激素治疗组与对照组蛋白尿缓解率的差异无统计学意义（P=0.34），儿童亚组分析结果一致。结论已有多种药物被用于治疗HBV-GN，但迄今为止最佳的治疗方案尚无定论。本研究显示抗病毒（包括rIFN-α和拉米夫啶）治疗HBV-GN能有效缓解蛋白尿、抑制HBV复制，且HBeAg的清除与蛋白尿的缓解密切相关；并能在一定程度上延缓肾功能恶化的发生；大多数患者耐受性好。而糖皮质激素治疗HBV-GN，其预后无明显改善。由于受限于原始研究的质量、数量以及各研究间治疗方案的差异，上述结论仅供临床医生参考，确切的结论有待今后由多中心、大规模RCT研究获取更多高质量EBM证据对本系统评价进行更新。

A meta-analysis of drug therapy for hepatitis B virus-associated nephritis

Objective Hepatitis B virus associated nephritis (HBV-GN) is one of the common secondary glomerular diseases in China. However, up to now, the optimal therapy is undefined although several approaches have been made. The aim of this study was to evaluate the efficacy of drug therapy for HBV GN by performing a systematic review of the literature with a meta analysis method.Methods Medical electronic databases including PubMed, EMBASE, Cochrane Library and CNKI were searched in two languages (English and Chinese) till December 2007, and other sources were manually screened. Two investigators independently identified literatures according to eligibility and exclusion criteria, assessed quality by the Sackett assessment plus Jadad scale, and extracted data. The primary outcome was clinical response (remission of proteinuria), the secondary outcome was virological response (clearance of HBeAg). The fixed effect model or random effect model were used to pool the data, with heterogeneity and sensitivity analyses by software STATA 9.0 and RevMan 4.2. In addition, other data were reviewed and described.ResultsNine studies (8 in English, 1 in Chinese) met the predefined criteria, one of which was randomized controlled trial (RCT), and others were cohort studies. The efficacy of antiviral therapy was assessed by synthesizing 6 out of 9 articles. There were 1 RCT and 5 cohort studies (159 unique patients). Recombinant IFN-α (rIFN-α) was given in most (5/6) studies, and lamivudine therapy in two study. It was found by the meta analysis that, compared with the control, antiviral therapy could significantly elevate the remission rate of proteinuria ［91.0% vs 56.0%, RR=1.69, 95%CI：1.08-2.65, P=0.02］, and the clearance rate of HBeAg ［59.7% vs 8.33%, RR=6.44, 95%CI：3.11-13.35, P<0.000 01］. Furthermore, Kappa analysis showed a significant link between proteinuria remission and HBeAg clearance after antiviral therapy (P=0.002). On the other hand, the efficacy of corticosteroid treatment was assessed by synthesizing 5 out of 9 articles. All of them were cohort studies (76 unique patients). Meta analysis showed there was no significant difference between corticosteroid treatment group and control group in the remission rate of proteinuria (RR=1.45, 95%CI：0.68-3.11, P=0.34).Conclusions This study showed the efficacy and safety of antiviral therapy (including IFN and lamivudine) in HBV GN. IFN-α was efficacious in remission of proteinuria, clearance of HBeAg and delay of renal function deterioration. In addition, remission of proteinuria was accompanied by clearance of HBV replication marker, which demonstrated the benefit of antiviral therapy. However, corticosteroid treatment couldn′t improve renal outcome in patients with HBV-GN. There were some shortcomings in our meta-analysis duo to limited high-quality RCTs, thus well designed RCTs are still needed before final conclusions is made.