Background: Sarcopenia has been associated with poor outcomes in pancreatic cancer (PC). However, published results are heterogeneous in terms of study design, oncological outcomes, and sarcopenia measurements. This meta-analysis aims to evaluate the impact of computed tomography (CT)-based sarcopenia on overall survival (OS) and progression-free survival (PFS) in patients with PC, considering potential confounders such as the CT-based method and thresholds used to define sarcopenia, as well as treatment intention.Methods: We systematically searched databases for observational studies reporting hazard ratios (HRs) for OS and PFS in PC patients stratified by CT-based sarcopenia status. Random-effects models were used to calculate pooled crude and adjusted HRs (cHRs and aHRs, respectively), with subgroup analyses based on sarcopenia measurement methods, cutoff values, sarcopenia prevalence, and treatment intention. Heterogeneity was assessed using the I2and τ2statistics, and publication bias was evaluated using funnel plots and Egger’s test.Results: Data from 48 studies were included. Sarcopenia was significantly associated with worse OS (pooled cHR = 1.58, 95% CI: 1.38–1.82; pooled aHR = 1.39, 95% CI: 1.16–1.66) and worse PFS (pooled cHR = 1.55, 95% CI: 1.29–1.86; pooled aHR = 1.31, 95% CI: 1.11–1.55). Subgroup analyses revealed significantly different, stronger associations in studies using stricter sarcopenia cutoffs (<50 cm2/m2for males) and in patients undergoing curative treatments. Heterogeneity was substantial across analyses (I2> 67%), but with generally low τ2values (0.01–0.25). Egger’s test indicated potential publication bias for OS (p< 0.001), but no significant bias was observed for PFS (p= 0.576).Conclusions: Sarcopenia determined by CT is an independent predictor of poor OS and PFS in PC, but this association varies depending on the cutoff used for its definition as well as on the treatment intention. Therefore, its routine assessment in clinical practice could provide valuable prognostic information, but future research should focus on standardizing sarcopenia assessment methods.
背景：肌肉减少症与胰腺癌患者的不良预后相关。然而，已发表的研究结果在研究设计、肿瘤学结局及肌肉减少症测量方法方面存在异质性。本荟萃分析旨在评估基于计算机断层扫描（CT）定义的肌肉减少症对胰腺癌患者总生存期和无进展生存期的影响，同时考虑潜在的混杂因素，如用于定义肌肉减少症的CT测量方法和阈值，以及治疗目的。 方法：我们系统检索了数据库，纳入报告按CT定义的肌肉减少症状态分层的胰腺癌患者总生存期和无进展生存期风险比的观察性研究。采用随机效应模型计算汇总的粗风险比和调整后风险比，并根据肌肉减少症测量方法、截断值、肌肉减少症患病率和治疗目的进行亚组分析。使用I²和τ²统计量评估异质性，通过漏斗图和Egger检验评估发表偏倚。 结果：共纳入48项研究的数据。肌肉减少症与较差的总生存期（汇总粗风险比 = 1.58，95% CI：1.38–1.82；汇总调整后风险比 = 1.39，95% CI：1.16–1.66）和较差的无进展生存期（汇总粗风险比 = 1.55，95% CI：1.29–1.86；汇总调整后风险比 = 1.31，95% CI：1.11–1.55）显著相关。亚组分析显示，在使用更严格肌肉减少症截断值（男性<50 cm²/m²）的研究中以及接受根治性治疗的患者中，相关性显著更强且更明显。各项分析间存在显著异质性（I² > 67%），但τ²值普遍较低（0.01–0.25）。Egger检验提示总生存期存在潜在发表偏倚（p < 0.001），但无进展生存期未观察到显著偏倚（p = 0.576）。 结论：基于CT确定的肌肉减少症是胰腺癌患者总生存期和无进展生存期不良的独立预测因素，但这种关联性因定义肌肉减少症所使用的截断值及治疗目的的不同而存在差异。因此，在临床实践中常规评估肌肉减少症可提供有价值的预后信息，但未来研究应侧重于标准化肌肉减少症的评估方法。
肿瘤(癌症)患者之家