文献速递消化系统肿瘤

基于身体成分和全身炎症的宿主表型预测可切除胰腺及壶腹周围癌患者生存

Prognostic host phenotypes based on body composition and systemic inflammation predict survival in patients with resected pancreatic and periampullary cancer.
PubMed 文献 · Eur J Cancer · 2026-07-09

该研究对PORSCH试验中549例接受胰十二指肠切除术的胰腺导管腺癌患者进行事后分析,基于低肌肉量、肌肉脂肪浸润和全身炎症(CRP>6mg/L)组合构建8种宿主表型。高风险表型(三项不良因素均存在)中位OS为13.0个月,显著低于中危表型的21.9个月和低危表型的35.2个月(p<0.001);多变量分析显示宿主表型与OS独立相关(中危aHR 1.39,高危aHR 2.10),且独立于肿瘤分期。

要点

  • 基于身体成分和全身炎症的宿主表型可预测可切除PDAC患者OS。
  • 高风险表型(低肌肉量+肌肉脂肪浸润+CRP>6mg/L)中位OS仅13.0个月。
  • 宿主表型的预后价值独立于肿瘤分期。

核心结论

  • 基于身体成分和全身炎症的宿主表型可独立于肿瘤分期预测可切除PDAC患者的OS。

研究设计与样本

  • 对全国性PORSCH试验进行事后分析,纳入所有因PDAC接受胰十二指肠切除术的患者,共549例。
  • 术前腹部CT采用自动分割(Mosamatic™)进行身体成分分析,低肌肉量和肌肉脂肪浸润分别通过性别标准化Z值的log-rank分层定义。
  • 基于低肌肉量、肌肉脂肪浸润和全身炎症(CRP>6mg/L)三种不良宿主因素的组合将患者分为8种宿主表型,并根据log风险比的k-means聚类分为低、中、高危表型。

主要发现

  • 高危表型(三种不良宿主因素均存在)中位OS为13.0个月(95%CI 11.4-19.3),低于中危表型([1]、[2]、[1+2]、[1+3]、[2+3])的21.9个月(95%CI 19.3-24.7)和低危表型([3]或无)的35.2个月(95%CI 29.2-45.3),p<0.001。
  • 多变量分析中,中危表型aHR为1.39(95%CI 1.08-1.80,p=0.01),高危表型aHR为2.10(95%CI 1.53-2.88,p<0.01),均独立于肿瘤分期。

关键数值与推荐

  • 全身炎症定义为C-reactive protein >6mg/L。
  • 低肌肉量和肌肉脂肪浸润分别基于性别标准化Z值的log-rank分层定义。

临床怎么用

  • 可利用术前CT身体成分分析和CRP水平构建宿主表型,辅助可切除PDAC患者的预后分层。
  • 宿主表型独立于肿瘤分期提供预后信息,提示临床生存预测需整合肿瘤-宿主交互因素。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

INTRODUCTION: Most present-day survival-models of patients with cancer do not account for tumor-host interactions. We hypothesized that host phenotypes based on systemic inflammation and body composition are prognostic of overall survival (OS) in patients with pancreatic ductal adenocarcinoma (PDAC). METHODS: We performed a post-hoc analysis of the nationwide PORSCH-trial, including all patients undergoing pancreatoduodenectomy for PDAC. Primary outcome was OS. Body composition analysis was performed using automatic segmentation (Mosamatic™) of preoperative abdominal computed tomography scans. Low muscle mass and myosteatosis were defined using log-rank stratification of sex-standardized Z-values of skeletal muscle index and radiation attenuation, respectively. Patients were clustered in eight host phenotypes based on combinations of adverse host factors: [1] low muscle mass, [2] myosteatosis, [3] systemic inflammation (C-reactive protein >6mg/L). Their association with OS was tested using multivariable-adjusted Cox-proportional hazard-analysis. Distinct combinations of adverse host factors were stratified into low-, intermediate-, and high-risk phenotypes according to k-means clustering based on log hazard-ratio's. RESULTS: 549 patients were included. The high-risk phenotype, characterized by the presence of all adverse host factors, showed lower median OS than intermediate ([1], [2], [1 +2], [1 +3], [2 +3]) and low-risk ([3]; none) phenotypes (13.0 months [95%CI 11.4-19.3] vs. 21.9 months [95%CI 19.3-24.7] vs. 35.2 months [95%CI 29.2-45.3], respectively; p < 0.001). In multivariable analysis, host phenotypes were associated with OS (adjusted [a]HR 1.39 [95%CI 1.08-1.80, p = 0.01; aHR 2.10 [95%CI 1.53-2.88], p < 0.01, for intermediate- and high-risk respectively), independent of tumor stage. DISCUSSION: Host phenotypes based on body composition and systemic inflammation predict OS independent of tumor stage in patients with resected PDAC, underscoring the importance of tumor-host interactions for clinical survival prediction.

来源:Eur J Cancer,查看原文。版权归原作者所有。

引用

  1. Hildebrand ND, van Dijk DPJ, Puik JR, et al. Prognostic host phenotypes based on body composition and systemic inflammation predict survival in patients with resected pancreatic and periampullary cancer. Eur J Cancer. 2026. PMID: 42447512. doi:10.1016/j.ejca.2026.116933

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。