Tolvaptan for Refractory Ascites in Cirrhosis: Systematic Review and Meta-Analysis

ABSTRACT Background & Aims: Refractory ascites marks a decisive transition to decompensated cirrhosis, with limited effective therapeutic options and a persistently poor prognosis. Tolvaptan, which is an oral vasopressin V₂-receptor antagonist, offers an aquaretic mechanism distinct from conventional natriuretic diuretics. Despite broadly consistent short-term physiological effects, its clinical adoption has diverged sharply between Japan and Western recommendations. This systematic review and meta-analysis sought to quantify efficacy and safety, interrogate the evidentiary basis for the Japan–West translational gap, and define the pharmacovigilance infrastructure required for safe use. Methods: A PRISMA 2020-compliant systematic review and meta-analysis (PROSPERO: CRD420251276165). PubMed, Embase, CENTRAL, and Japanese databases (Japic-CTI, UMIN) were searched. Risk of bias was assessed using Cochrane RoB 2 and ROBINS-I. Random-effects meta-analyses were performed for physiological outcomes; certainty was graded using GRADE. Safety data were synthesised from randomised trials and large pharmacovigilance cohorts. Results: Sixteen studies (9 RCTs, 7 prospective cohorts; n=2,154) were included. Meta-analysis of RCTs provided moderate-certainty evidence for short-term improvements in urine output (mean difference [MD] +1240 mL/day, 95% CI +1050 to +1430), body weight (MD -1.48 kg, 95% CI -1.92 to -1.04), and serum sodium (MD +2.61 mmol/L, 95% CI +1.87 to +3.35). Evidence for effects on paracentesis frequency or mortality was of low or very low certainty. Hepatic function abnormalities represent a confirmed, frequent adverse drug reaction (9.6% in real-world surveillance), with clinically significant ALT elevations occurring 3–14 months after treatment initiation, mandating protocolised monthly liver function monitoring as a non-negotiable component of therapy. Japanese practice integrates surrogate endpoint evidence for symptom control within structured specialist pathways, whereas Western guidelines, citing drug-induced liver injury (DILI) risk and the absence of high-certainty outcome data, contraindicate or do not recommend its use. Conclusions: The evidence base is best understood as a gradient: moderate certainty for short-term physiological benefit alongside low certainty for patient-centred clinical outcomes. Divergent risk–benefit appraisals and differing healthcare infrastructures plausibly account for the Japan–West divide. The DILI signal is not an absolute barrier but a systems-dependent constraint, tightly linking feasibility to robust monitoring infrastructure. A definitive outcomes trial, coupled with validated, context-adapted monitoring protocols, is a prerequisite for global therapeutic reconsideration.摘要 背景与研究目的：顽固性腹水是进展至失代偿期肝硬化的关键转折点，目前有效治疗手段有限，预后持续不佳。托伐普坦（Tolvaptan）是一种口服血管加压素V₂受体拮抗剂，其利水机制与常规排钠利尿剂截然不同。尽管短期生理学效应大体一致，但其临床应用在日本与西方指南间却存在显著分歧。本系统评价与Meta分析（systematic review and meta-analysis）旨在量化其疗效与安全性，探究造成日西两地应用差异的证据基础，并明确安全用药所需的药物警戒体系。 研究方法：本研究为遵循PRISMA 2020规范的系统评价与Meta分析，已在PROSPERO平台注册（注册号：CRD420251276165）。检索了PubMed、Embase、CENTRAL以及日本本土数据库Japic-CTI、UMIN。采用Cochrane偏倚风险评估工具2.0版（Cochrane RoB 2）和非随机研究偏倚风险评估工具（ROBINS-I）评估偏倚风险。针对生理学结局指标采用随机效应模型Meta分析，证据质量采用GRADE分级体系进行评级。安全性数据则从随机对照试验（Randomized Controlled Trial, RCT）与大型药物警戒队列中综合提取。 研究结果：共纳入16项研究，其中9项随机对照试验（Randomized Controlled Trial, RCT）、7项前瞻性队列研究，总样本量为2154例。针对RCT的Meta分析显示，中等质量证据表明托伐普坦可在短期内改善多项生理学指标：尿量均数差（MD）+1240 mL/日，95%置信区间（CI）为+1050至+1430；体质量均数差为-1.48 kg，95%CI为-1.92至-1.04；血清钠水平均数差为+2.61 mmol/L，95%CI为+1.87至+3.35。而其对腹腔穿刺频次或病死率的影响证据质量则为低或极低。肝功能异常已被证实为常见的药物不良反应（adverse drug reaction, ADR），真实世界监测数据显示其发生率为9.6%，临床显著的丙氨酸氨基转移酶（ALT）升高多在治疗启动后3~14个月出现，因此标准化月度肝功能监测已成为治疗不可或缺的核心环节。日本临床实践将症状控制的替代终点（surrogate endpoint）证据整合至结构化专科诊疗路径中，而西方指南则以药物性肝损伤（drug-induced liver injury, DILI）风险及缺乏高质量结局数据为由，将托伐普坦列为禁忌或不推荐使用。 研究结论：现有证据基础可视为一个梯度：短期生理学获益具有中等质量证据，而以患者为中心的临床结局证据质量则较低。日西两地的风险获益评估差异以及医疗保健体系的不同，可合理解释两地应用的分歧。药物性肝损伤信号并非绝对禁忌，而是依赖于医疗体系的约束条件，其临床可行性与完善的监测体系紧密相关。开展确定性结局试验，并结合经过验证的适配场景监测方案，是全球范围内重新考量其治疗应用的前提条件。