INVITED EDITORIALS Editorial: metabolomic biomarkers for colorectal adenocarcinoma and in the differentiation between irritable bowel syndrome and ulcerative colitis in clinical remission— confounded by the gut microbiome? With advances in spectrometry techniques, metabolomics is becom- ing a robust strategy for the identification of biomarkers for the diagnosis of disease or disease activity. In gastrointestinal diseases, faecal and urinary samples are ideal sources of metabolic signatures of divergent pathogenic pathways; the noninvasive nature of their collection makes them fitting for novel diagnostics. Two recent pub- lications in AP&T 1,2 aimed to identify molecules that accurately diag- nose gastrointestinal diseases based on metabolic profiling. Using direct infusion (DI‐)/liquid chromatography (LC‐) tandem mass spectrometry (MS/MS) and gas‐chromatography (GC‐) MS to define urinary amino acid profiles, Keshteli et al 1 were for the first time able to discriminate between patients with irritable bowel syn- drome (IBS) and ulcertive colitis (UC) in remission. Lysine was signifi- cantly higher in IBS compared to UC (where reduced levels had been previously reported 3 ), while proline was significantly higher in UC. Differences in four metabolites were marked with area‐under‐ the‐curve on receiver operator characteristic (AUROC) of 0.99 (95% CI: 0.95‐1.0) in the validation set. The implication is that new labora- tory assays may be developed to differentiate the two conditions. Of interest, several urinary metabolites correlated with intestinal microbial species on testing of paired 16S rRNA amplicon sequenc- ing of DNA extracted from stool samples. In UC, urinary lysine cor- related moderately with the abundance of Clostridiales (corrected r = -0.50) while proline correlated positively with Erysipelotrichales (- corrected r = 0.39). These findings suggest that the metabolic profile of the urine might be influenced by differences in the composition of the gut microbiome associated with the respective diseases. This is further supported by observations that cadaverine, a product of lysine metabolism, is increased in faecal extracts from patients with UC when compared to patients with IBS and healthy controls. 4 Using Headspace GCMS to define profiles of volatile organic compounds (VOCs) in faeces, Bond et al 2 were able to accurately differentiate patients with colorectal adenocarcinoma from controls by summation of three compounds (AUROC 0.73). Propan‐2‐ol was the strongest single biomarker with an AUROC to predict adenocar- cinoma of 0.76. While this might not imminently translate into screening due to the cost and complexity of capturing VOCs in comparison to cheaper techniques, it raises the possibility that the intestinal microbiome might be the source of relevant VOCs and dri- ver of colorectal dysplasia. Fusobacterium nucleatum has previously been associated with colorectal cancer, 5,6 with a key role for its adhesin FadA in binding E‐cadherin, activating β‐catenin signalling and inducing inflammation and oncogenesis. 7 Increased levels of pro- pan‐2‐ol in adenocarcinoma could be associated with increased levels of F nucleatum since they have observed (data not shown) the bacterium to produce this compound. Before these novel metabolomic studies can translate into useful diag- nostic tools for gastrointestinal diseases, larger studies are required to account for potential confounding factors. For example, dietary and other environmental factors that drive a disease process might modify the intesti- nal microbiome and be independent risk factors of disease. Also, cause and effect cannot be implied from cross‐sectional studies. Demonstrating dose‐ response in accordance with biological plausibility and replicating results across laboratories would increase their credibility. ACKNOWLEDGEMENTS Declaration of personal interests: Rupert Leong has served as a speaker or an advisory board member for Aspen, AbbVie, Celgene, Ferring, Hospira, Janssen, MSD, Pfizer, Takeda, and has received research funding from NHMRC, Shire, Janssen, Takeda, and owns patent PCT/AU2014/000682, PCT/AU2017/050580 relating to diag- nostic methods. Yunki Yau, owns patent PCT/AU2014/000682, PCT/ AU2017/050580 relating to diagnostic methods. Nadeem Kaakoush has no declaration Valerie Wasinger, owns patent PCT/AU2014/ 000682, PCT/AU2017/050580 relating to diagnostic methods. Declaration of funding interests: None. ORCID Rupert W. L. Leong https://orcid.org/0000-0001-5944-3488 Nadeem O. Kaakoush https://orcid.org/0000-0003-4017-1077 Yunki Y. Yau https://orcid.org/0000-0001-8832-8200 Valerie C. Wasinger https://orcid.org/0000-0001-5338-0869 AP&T invited editorial columns are restricted to discussing papers that have been published in the journal. An editorial must have a maximum of 500 words, may contain one table or figure, and should have no more than 10 references. It should be submitted electronically to the Editors via http:// mc.manuscriptcentral.com/apt. DOI: 10.1111/apt.15171 1086 | © 2019 John Wiley & Sons Ltd wileyonlinelibrary.com/journal/apt Aliment Pharmacol Ther. 2019;49:1086–1102.