Lights and shadows of mass spectrometry-based metabolomics studies in cancer diagnosis: is there any marker beyond the literature background noise?

Mass spectrometry (MS)-based metabolomics is considered as one of the gold-standards for the discovery of clinical biomarkers. However, metabolites proposed by MS-based metabolomics do not usually fulfil all the quality requirements to be considered as clinical biomarkers. This is especially relevant in cancer research, where there is an urgent need for robust diagnostic biomarkers. In this work, we performed an in-depth revision of recently published MS-based metabolomics studies for cancer biomarker discovery published between 2017 and 2024. We also presented a workflow for assuring robustness and reproducibility in the proposed biomarkers. The key steps of the workflow to ensure transferability of the biomarkers to clinical routine are discussed based on published literature. Our work revealed important aspects on how research has been performed recently: (i) MS-based metabolomic studies are often focused on high-incidence cancers, rather than on those lacking early diagnostic biomarkers, (ii) there is insufficient consistency in the proposed metabolites among different studies, (iii) most studies do not perform an accurate identification of the features detected by untargeted approaches, (iv) targeted analytical methods are scarcely validated, (v) most studies fail to present a clear protocol for data interpretation (e.g. concentration thresholds for proposed metabolites), and (vi) biomarkers are not usually validated using an independent cohort. This work illustrates the need for a consensus in MS-based metabolomics for the discovery of clinical biomarkers and their implementation for the early-diagnosis of cancer. The proposed workflow aims to guide researchers toward achieving robustness and reproducibility between studies, thus increasing confidence in the results.