Nutrition & Metabolism (Feb 2023)

Association of plasma lipid metabolism profiles with overall survival for patients with gastric cancer undergoing gastrectomy based on 1H-NMR spectroscopy

  • Yaopeng Qiu,
  • Zhou Xu,
  • Qingfeng Xie,
  • Renyi Zhang,
  • Luyao Wang,
  • Liying Zhao,
  • Hao Liu

DOI
https://doi.org/10.1186/s12986-023-00728-1
Journal volume & issue
Vol. 20, no. 1
pp. 1 – 11

Abstract

Read online

Abstract Background Lipid metabolism dysregulation is a prominent metabolic alteration in various cancers. The study aimed to explore the association of plasma lipid metabolism profiles with overall survival (OS) for gastric cancer (GC) patients who received gastrectomy. Methods GC patients who were treated with gastrectomy and measured with plasma lipid metabolism profiles using proton nuclear magnetic resonance (1H-NMR) spectroscopy in Nanfang Hospital between January 1, 2017, and October 31, 2018, were recruited. The Least Absolute Shrinkage and Selection Operator (LASSO) regression model was used to analyze variables selected by univariate analysis for OS. An index of plasma lipid metabolism profiles, named plasma lipid metabolism index (PLMI), was constructed by variables' coefficients in LASSO regression to explore its association with OS and its role in the prediction model. Results A total of 158 GC patients were included in this study. Four of the 110 lipid profiles, including LDL-5 Apo-B, LDL-4 Cholesterol, HDL-4 Apo-A2, and HDL-4 Free Cholesterol, were selected to construct the PLMI. The optimal cut-off value of PLMI for OS was used to classify the population into two subgroups, the high PLMI group (≥ − 0.163) and the low PLMI group (< − 0.163). The high PLMI group had a shorter OS (p = 0.0034) and was the independent risk factor for OS (Hazard Ratio = 2.13, 95% Confidence Interval (CI): 1.07–4.22, p = 0.031) after adjusting for perineural invasion and tumor stage. In subsets of the I–III stage and treating postoperative chemotherapy, high PLMI also had an unfavorable correlation with OS (p = 0.016 and p = 0.0086, respectively). The nomogram prediction models of both the training cohort and validation cohort showed good calibration and discrimination with the concordance indexes of 0.806 (95% CI, 0.732–0.880) in the training cohort and 0.794 (95% CI, 0.725–0.862) in the validation cohort. Conclusions This study found that the index derived from the LDL-5 Apo-B, LDL-4 Cholesterol, HDL-4 Apo-A2, and HDL-4 Free Cholesterol, was significantly associated with overall survival, suggesting that regulating lipid metabolisms might improve the prognosis for GC patients.

Keywords