Characteristics of Cancer Epidemiology Studies That Employ Metabolomics: A Scoping Review

Abstract

An increasing number of cancer epidemiology studies use metabolomics assays. This scoping review characterizes trends in the literature in terms of study design, population characteristics, and metabolomics approaches and identifies opportunities for future growth and improvement. We searched PubMed/MEDLINE, Embase, Scopus, and Web of Science: Core Collection databases and included research articles that used metabolomics to primarily study cancer, contained a minimum of 100 cases in each main analysis stratum, used an epidemiologic study design, and were published in English from 1998 to June 2021. A total of 2,048 articles were screened, of which 314 full texts were further assessed resulting in 77 included articles. The most well-studied cancers were colorectal (19.5%), prostate (19.5%), and breast (19.5%). Most studies used a nested case-control design to estimate associations between individual metabolites and cancer risk and a liquid chromatography-tandem mass spectrometry untargeted or semi-targeted approach to measure metabolites in blood. Studies were geographically diverse, including countries in Asia, Europe, and North America; 27.3% of studies reported on participant race, the majority reporting White participants. Most studies (70.2%) included fewer than 300 cancer cases in their main analysis. This scoping review identified key areas for improvement, including needs for standardized race and ethnicity reporting, more diverse study populations, and larger studies.

Figure 1.

PRISMA flow diagram for selecting sources of evidence. The flow diagram shows the process used to select sources of evidence to be included in the scoping review examining population-based cancer metabolomics research.

Figure 2.

Pie chart displays the breakdown of study design types used in metabolomic epidemiology studies of cancer. *Other includes meta-analysis of multiple nested case–control studies.

Figure 3.

Bar graph depicts cancer types studied in the metabolomic epidemiology literature. *Other includes extrahepatic cholangiocarcinoma, leucoma, and other not specified.

Figure 4.

Pie chart shows the breakdown of metabolomics-specific aims of included metabolomic epidemiology studies of cancer. (a) Cancer risk estimation using incident cases; (b) cancer risk estimation using prevalent cases; (c) biomarker of exposure and cancer risk estimation using incident cases; (d) biomarker of exposure and cancer risk estimation using prevalent cases; (e) biomarkers of disease diagnosis; (f) biomarkers of survival; (g) biomarkers of disease diagnosis and survival; (h) disease differentiation: cases vs. controls; (i) disease differentiation: tumor vs. non-tumor tissue; (j) risk of recurrence or death among cancer survivors; (k) cancer progression/natural history; (l) association study: prognosis/recurrence; (m) descriptive study: progression/survival.