血浆代谢组学特征与晚期和致命性前列腺癌的风险。

Plasma Metabolomic Profiles and Risk of Advanced and Fatal Prostate Cancer.

机构信息

Behavioral and Epidemiology Research Group, American Cancer Society, Atlanta, GA, USA.

出版信息

Eur Urol Oncol. 2021 Feb;4(1):56-65. doi: 10.1016/j.euo.2019.07.005. Epub 2019 Aug 1.

DOI:10.1016/j.euo.2019.07.005

PMID:31378665

Abstract

BACKGROUND

Little is known about the underlying molecular mechanisms of prostate cancer, especially advanced and fatal prostate cancer.

OBJECTIVE

To examine associations of prediagnostic plasma metabolomic profiles with advanced and fatal prostate cancer.

DESIGN, SETTING, AND PARTICIPANTS: In a case-cohort study of the Cancer Prevention Study-II Nutrition Cohort, of 14 210 cancer-free men with a blood sample in 1998-2001, 129 were diagnosed with advanced prostate cancer (T3-T4 or N1 or M1) through June 2013 and 112 died from prostate cancer through December 2014. Plasma samples from advanced and fatal cases, and a randomly selected subcohort of 347 men were metabolically profiled using untargeted mass spectroscopy-based platforms.

OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS

Prentice-weighted Cox proportional hazards regression models were used to assess associations of 699 known metabolites with advanced and fatal prostate cancer.

RESULTS AND LIMITATIONS

Two metabolites derived from fatty acid metabolism (ethylmalonate and butyrylcarnitine), aspartate, sphingomyelin (d18:1/18:0), and two γ-glutamyl amino acids (γ-glutamylmethionine and γ-glutamylglutamine) were statistically significantly associated (false discovery rate <0.2) with fatal prostate cancer. One standard deviation (SD) increase in each γ-glutamyl amino acid was associated with 34-38% decreased risk, whereas one SD increase in each of the other metabolites was associated with 45-53% increased risk. A metabolic risk score based on four of these metabolites (excluding butyrylcarnitine and γ-glutamylglutamine, which were not independent predictors) was strongly associated with fatal prostate cancer (relative risk per SD: 2.72, 95% confidence interval: 2.05-3.60). No metabolites were statistically significantly associated with advanced prostate cancer. These results were observational and may not be causal.

CONCLUSIONS

These findings identified metabolic pathways that are altered in the development of fatal prostate cancer. Further research into these pathways may provide insights into the etiology of fatal prostate cancer.

PATIENT SUMMARY

In a large follow-up study of cancer-free men, those with a certain metabolomic profile had a higher risk of dying from prostate cancer.

摘要

背景

人们对前列腺癌的潜在分子机制知之甚少，尤其是晚期和致命性前列腺癌。

目的

研究前列腺癌患者诊断前的血浆代谢组学特征与晚期和致命性前列腺癌之间的关系。

设计、地点和参与者：在癌症预防研究 II 营养队列的病例-队列研究中，纳入了 1998 年至 2001 年期间的 14210 名无癌症男性，他们的血液样本中有 129 人在 2013 年 6 月前被诊断为晚期前列腺癌（T3-T4 或 N1 或 M1），112 人在 2014 年 12 月前死于前列腺癌。从晚期和致命性病例以及随机选择的 347 名男性的亚队列中抽取血浆样本，使用非靶向质谱技术平台进行代谢组学分析。

测量结果和统计分析

采用 Prentice 加权 Cox 比例风险回归模型评估 699 种已知代谢物与晚期和致命性前列腺癌之间的关系。

结果和局限性

两种源自脂肪酸代谢的代谢物（丙二酸和丁酰肉碱）、天冬氨酸、神经鞘磷脂（d18:1/18:0）和两种γ-谷氨酰氨基酸（γ-谷氨酰甲硫氨酸和γ-谷氨酰谷氨酰胺）与致命性前列腺癌具有统计学显著相关性（错误发现率<0.2）。每个γ-谷氨酰氨基酸增加一个标准差与风险降低 34%-38%相关，而其他代谢物中每个标准差增加与风险增加 45%-53%相关。基于其中 4 种代谢物（不包括丁酰肉碱和γ-谷氨酰谷氨酰胺，它们不是独立的预测因子）的代谢风险评分与致命性前列腺癌密切相关（每标准差的相对风险：2.72，95%置信区间：2.05-3.60）。没有代谢物与晚期前列腺癌有统计学显著相关性。这些结果是观察性的，可能不是因果关系。