You are here
Metabolomic profiling of oesophago-gastric cancer: A systematic review
European Journal of Cancer, Volume 49, Issue 17, November 2013, Pages 3625 - 3637
This review aims to identify metabolomic biomarkers of oesophago-gastric (OG) cancer in human biological samples, and to discuss the dominant metabolic pathways associated with the observed changes.
A systematic review of the literature, up to and including 9th November 2012, was conducted for experimental studies investigating the metabolomic profile of human biological samples from patients with OG cancer compared to a control group. Inclusion criteria for analytical platforms were mass spectrometry or nuclear magnetic resonance spectroscopy. The QUADAS-2 tool was used to assess the quality of the included studies.
Twenty studies met the inclusion criteria and samples utilised for metabolomic analysis included tissue (n = 11), serum (n = 8), urine (n = 1) and gastric content (n = 1). Several metabolites of glycolysis, the tricarboxylic acid cycle, anaerobic respiration and protein/lipid metabolism were found to be significantly different between cancer and control samples. Lactate and fumurate were the most commonly recognised biomarkers of OG cancer related to cellular respiration. Valine, glutamine and glutamate were the most commonly identified amino acid biomarkers. Products of lipid metabolism including saturated and un-saturated free fatty acids, ketones and aldehydes and triacylglycerides were also identified as biomarkers of OG cancer. Unclear risk of bias for patient selection was reported for the majority of studies due to the lack of clarity regarding patient recruitment.
The application of metabolomics for biomarker detection in OG cancer presents new opportunities for the purposes of screening and therapeutic monitoring. Future studies should provide clear details of patient selection and develop metabolite assays suitable for progress beyond phase 1 pre-clinical exploratory studies.
Keywords: Gastric cancer, Oesophageal cancer, Metabolomics, Metabolic profiling, Mass spectrometry, Nuclear magnetic resonance imaging, Glycolysis, TCA cycle, Amino acids, Lipids.
Oesophago-gastric (OG) cancer affects approximately 1.5 million people a year worldwide and accounts for 15% of cancer related deaths. 1 In Western countries, the incidence is increasing, in particular adenocarcinomas of the distal oesophagus and gastro-oesophageal junction.2, 3, and 4 Prognosis remains poor with 5-year survival rates of 34% for localised disease and 17% for all stages combined. 5
Treatment options and overall survival depend on both the stage of the disease and a patient’s general health. The early symptoms of OG cancer are usually insidious and the majority of patients present late to their doctor with incurable disease. Surgical resection of the tumour remains the mainstay of curative treatment. In the United Kingdom (UK), only 38% percent of newly diagnosed patients are suitable for treatment with curative intent. 6 Identification of these cancers at an earlier stage could potentially improve the survival outcomes of OG cancer. There are currently no screening tests in mainstream clinical practice, despite evidence that treatment of early OG cancer has more favourable outcomes.7, 8, and 9 Oesophago-gastroduodenoscopy (OGD) and biopsy remain the gold standard procedure for the diagnosis of OG cancer. However, this is an invasive procedure and is too costly to be employed in a screening capacity. Radiological tests such as barium oesophograms and serum biomarkers are available; however their use has been limited due to poor sensitivity and specificity.
Metabolomics is defined as a quantitative description of all endogenous low-molecular-weight components (<1 kDa) in a biological sample, such as tissue, urine or plasma. 10 It is an evolving field and it has the potential to be an effective tool for the early diagnosis of OG cancer, through identification of one or more diagnostic biomarkers. The composition of these endogenous compounds is affected by the upstream influence of the proteome and genome as well as environmental factors, lifestyle factors, medication and underlying disease. The exact number of different metabolites in humans is currently unknown. Recent estimates by the human metabolite database (HDMB) suggest up to 5000 compounds, with reference to bio-fluid or tissue concentration data existing in the current literature. 11
The majority of analytical platforms for metabolomic profiling are based on spectroscopic techniques. Mass spectrometry (often combined with chromatographic separation) and nuclear magnetic resonance (NMR) spectroscopy are the two most commonly employed analytical platforms; both techniques allow extensive and rapid analysis of small molecule metabolites, 12 resulting in multi-parameter datasets containing quantitative information on a range of metabolites. Due to the complexity and volume of the resultant spectral data, computer-based pre-processing and multivariate modelling techniques have been developed to facilitate the analysis and interpretation of the data.
Metabolomics studies have primarily focussed on the identification of cancer-specific metabolites. Examples include analysis of serum for the diagnosis of colorectal cancer, 13 exhaled breath for lung cancer 14 and urine for prostate cancer. 15 The aim of these studies has been to identify a panel of metabolites that may provide a metabolic fingerprint of malignancy. In combination, their sensitivity and specificity of predicting cancer has greater statistical power than single metabolites.
This systematic review aims to identify metabolomic biomarkers of OG cancer, found in human biological samples, and to assess the quality of the published data. The dominant metabolic pathways of malignant transformation, associated with the observed changes, will also be discussed.
2.1. Search strategy
A literature search (title and abstract) of Ovid Medline(R) (1948–2012), Embase (1974–2012), Web of Science and PubMed electronic databases was conducted up to and including 9th November 2012 for studies of metabolomic profiling of oesophago-gastric cancer. The search was conducted using the MeSH terms: mass spectrometry, nuclear magnetic resonance spectroscopy, metabolomic, metabonomic, metabolic profiling in multiple combinations (AND) with gastric cancer and oesophageal/esophageal cancer.
Two reviewers (N.A. and S.K.) independently screened titles and abstracts of studies identified through the electronic search. Full texts of potentially relevant articles were retrieved. Further potentially relevant articles were identified through the searching of reference lists of relevant studies. Experimental studies investigating the metabolomic profile of human biological samples from patients with gastric or oesophageal cancers, compared to an appropriate control group were included in our analysis. Inclusion criteria for analytical platforms were mass spectrometry (MS) or NMR. Exclusion criteria were: studies analysing the proteome rather than the metabolome, in vitro cell line studies, animal studies and studies without a control group. Studies that reported the same patient population were also excluded, except for the most recent or complete publication.
Two reviewers (N.A. and S.K.) independently extracted data from selected studies including primary author; year of publication; number and types of specimen; analytical platform and significantly different metabolites in the cancer and control groups. Compounds presented with varying chemical nomenclature in different articles are described by their common name in this review. The primary outcome measure was the identification of metabolites found to have different abundances between cancer and control samples at a statistically significant level.
2.2. Study quality assessment
The QUADAS-2 tool 16 was used to assess the quality of the included studies. It consists of four key domains covering patient selection, index test, reference standard and flow of patients through the study. The reference standard is defined as the best available method to establish the presence or absence of the target condition. 17 In this review, the reference standard is considered to be histological examination. Each domain is assessed in terms of the risk of bias and the first three are also assessed in terms of concerns regarding applicability. To help reach a judgement on the risk of bias, signalling questions are included, full details of which are provided in Supplementary file 1 .
Twenty studies met the inclusion criteria and were eligible for systematic review. A kappa score of 0.85 confirmed excellent agreement between assessors. Search results are presented in Fig. 1 .
3.1. Study characteristics
There were 12 studies of gastric cancer, six of oesophageal cancer and the remaining two had a mixed cohort ( Table 1 ). Biological samples utilised for Metabolomic analysis included tissue (n = 11), serum (n = 8), urine (n = 1) and gastric content (n = 1). The analytical platforms used for metabolite detection included Gas Chromatography–Mass Spectrometry (GC–MS) (n = 10), High Resolution-Magic Angle Spinning-NMR (n = 3), Capillary Electrophoresis–Mass Spectrometry (n = 1), NMR (n = 3), Liquid Chromatography–Mass Spectrometry (n = 2) and Selected Ion Flow Tube–Mass Spectrometry (n = 1).
|Author and year of publication||Analytical platform||Cancer site||Sample type||Cancer type||NCa||Control||Nc||Metabolites of cancer compared to control||Other findings||Refs.|
|Mun CW (2004)||1H NMR||Gastric||Tissue||GC||13||NGT||22||Choline||Lipids||18|
|Tugnioli V (2006)||HR-MAS-NMR||Gastric||Tissue||GC-AC
|Ligor T (2007)||SPME–GCMS||Gastric||Tissue||GC||3||NGT||13||Carbon disulphide ⁎||No difference in VOCs of exhaled breath of patients with cancer versus control||20|
|Buszewski B (2008)||SPME–GCMS||Gastric||Tissue||GC||5||NGT||5||Propanol ⁎||Butane & carbon disulphide present in the headspace of HP culture||21|
|Hirayama A (2009)||CE–MS||Gastric||Tissue||GC||12||NGT||Lactate, Glucose, fructose, Fumurate||Citrate||22|
|Malate, Glyceraldehyde-3P and|
|All common AA apart from Gln & Asp|
|Cai Z (2010)||GC–MS||Gastric||Tissue||GC||65||NGT||65||Pyruvic acid, Lactic acid||Fumaric acid||Confirmation of findings with western blot proteomic analysis||23|
|Fructose, Glyceraldehyde and Isocitric acid|
|Wu H (2010)||GC–MS||Gastric||Tissue||GC||18||NGT||18||Val, Ile, Ser, Gln||Phosphoserine||Higher levels of l-Cysteine, hypoxanthine, l-Tryosine and lower levels of phenanthrenol and butanoic acid in late versus early stage cancer||24|
|Heptanedioic acid, Propanoic acid,||l-Altrose|
|Butenoic acid, Galactofuranoside,||l-Mannofuranose and|
|Phenanthrenol, butanetriol, Acetamide||d-Ribofuranose|
|Oxazolethione and Naphtalene|
|Song H (2011)||GC–MS||Gastric||Tissue||GC||30||NGT||30||Fumaric acid, Valeric acid,||3-Hydroxybutanoic acid||Failed to differentiate pathological stage using multivariate analysis||25|
|α-Ketoglutaric acid,||9-Hexadecanoic acid|
|Benzenepropanoic acid,||Hexadecanoic acid|
|1-Phenanthrene, Squalene,||Cis-vaccenic acid|
|Carboxylic acid, Xylonic acid and||Arachidonic acid and|
|Calabrese C (2012)||HR-MAS-MRS||Gastric||Tissue||GC-AC||5||NGT||12||Choline||Increased presence of lipid bodies found in the cytoplasm of cancer cell with TEM/SEM imaging||26|
|AAG||5||Gly, Ala and|
|Yu L (2011)||GC–MS||Gastric||Serum||GC-AC||9||GSG||19||Urate, Ornithine,||Creatinine||Surgical removal of GC tissue restored the levels of some metabolites back to those found in CSG||27|
|Pyroglutamate, Azelaic acid||Threonate|
|Glu, Asn and y-tocopherol|
|Ikeda A||GC–MS||Gastric||Serum||GC-AC||11||HV||12||3-Hydroxypropionic acid and||Octanoic acid, Phosphoric acid
|Oesophageal||EAC||15||Lactic acid, Glycolic acid, Malonic acid, Fumaric acid, Ser, Gln and Asp||Pyruvic acid|
|Song H (2012)||GC–MS||Gastric||Serum||GC||30||HV||30||Val||Gln, Hexanedioic acid,||Failed to differentiate pathological stage using multivariate analysis||29|
|Sarcosine and||9,12 Octadecadienoic acid,|
|Cholesterol, fumaric acid, mesyl-arabinose Benzeneacetonitrile,|
|Aa J (2012)||GC–MS||Gastric||Serum||CSG||20||CSG||17||Citrate, Succinate, Fumurate, Malate Glu, β-hydroxybutyrate||Glucose||Surgical removal of GC tissue restored the levels of some metabolites back to normal||30|
|Hexadecenoic acid, Docosahexaenoic acid and hepatanoic acid|
|Wu H (2009)||GC–MS||Oesophageal||Tissue||EAC 2||20||NOT||20||Val, Ile, Tyr, Asn, andAla||l-Altrose||31|
|ESC 18||Arabinofuranoside, Tetradecanoic acid,||d-Galactofuranoside|
|Hexadecanoic acid, Naphthalene,||Arabinose and|
|Myo-inositol, Phosphoric acid and|
|Yakoub D (20100||HR-MAS-NMR||Oesophageal||Tissue||EAC||35||NOT||35||Phospho-choline, Myo-inositol, Glu,||Stepwise up-regulation of metabolites from cancer to adjacent tissue to healthy distant tissue||32|
|NOTH||52||Inosine, Adenosine and uridine containing compounds|
|Djukovic D (2010)||HPLC–MS||Oesophageal||Serum||EAC||1-Methyladenosine,||Uridine||33|
|N2-methylguanosine and Cytidine|
|Zhang J (2011)||NMR||Oesophageal||Serum||EAC||68||HV||34||Citrate, lactate, a-glucose||Up-regulation of metabolites shown in BE-HGD-EAC sequence||34|
|BE||5||Gln, Lys, B-hydroxybutyrate and|
|Zhang J (2012)||LC–MS||Oesophageal||Serum||EAC||67||HV||34||Lactic acid||Val, Leu, Ile, Met, Tyr and Trp||35|
|BE||3||Carnitine and||Myristic acid, Linolenic acid and|
|Margaric acid||Linoleic acid|
|Kumar S (2012)||SIFT–MS||Gastric and Oesophageal||GastricContent||Cancer||19||HV||11||Acetaldehyde, Acetone and Acetic acid||Formaldehyde||36|
|ID||9||Hexanoic acid, Hydrogen sulphide,|
|Hydrogen cyanide, Methanol and|
|Hasim A (2012)||1H NMR||Oesophageal||Urine||SCC||108||HV||40||Mannitol, Glucose, Pyruvate,||N-acetylcysteine, Val
Citric acid and
|Metabolites in plasma positively correlated to lymph node metastatic rate||37|
Acetate and Allantoin
|Serum||Dimethylamine||Leu, Ala, Ile, Val, Glycoprotein,
Lactate, Acetone, Acetate, Choline
Isobutyrate, Unsaturated lipids
VLDL and LDL
⁎ Results not statistically significant.
1H NMR, Proton Nuclear Magnetic Resonance; HR-MAS-NMR, High Resolution-Magic Angle Spinning-Nuclear Magnetic Resonance Spectroscopy; SPME, Solid Phase Micro-extraction; GC–MS, Gas Chromatography–Mass Spectrometry; CE–MS, Capillary Electrophoresis–Mass Spectrometry; HPLC–MS, High Performance Liquid Chromatography–Mass Spectrometry; LC–MS, Liquid Chromatography–Mass Spectrometry; NCa, number of cancer patients; Nc, number of control patients; GC, gastric cancer; EAC, oesophageal adenocarcinoma; ESC, oesophageal squamous cell carcinoma; NGT, normal gastric tissue matched to the same patient; AAG, autoimmune atrophic gastritis; CSG, chronic superficial gastritis; NOT, normal oesophageal tissue from the same patients; NOTH, normal oesophageal tissue from healthy patients; HV, healthy volunteers; BE, Barrett’s oesophagus; HGD, high grade dysplasia; ID, inflammatory disorders; HP, Helicobacter pylori; VOCs, volatile organic compounds; Val, valine; His, histidine; Phe, phenylalanine; Thr, threonine; Leu, leucine; Met, methionine; Trp, tryptophan; Ile, isoleucine; Lys, lysine; Tyr, tyrosine; Gln, glutamine; Glu, glutamate; Ser, serine; Asn, asparagine; Gly, glycine; Ala, alanine; Arg, arginine; Asn, asparagine; Asp, aspartic acid; Cys, cysteine; Pro, proline; TEM, tunnel electron microscopy; SEM, surface electron microscopy.
3.2. Quality assessment of studies
The outcomes of the QUADAS-2 study quality assessment are shown in Fig. 2 . Further details regarding individual studies are presented in Supplementary file 1 . Unclear risk of bias for patient selection was reported for the majority of studies due to the lack of clarity regarding patient recruitment. It is unclear whether consecutive patient data is included or whether highly selected samples were chosen. As a consequence, 35% of studies have an unclear/high risk of applicability with respect to the target population defined by the review.
The risk of bias for the index test and reference standard was unclear for the majority of studies due to lack of clarity regarding blinding of assessors. Thirty-five percent of studies had an unclear/high risk of applicability with regard to the reference standard, mostly due to lack of information regarding histological confirmation of control samples.
3.3. Metabolites of cellular respiration
The metabolite intermediates of glycolysis, the tricarboxylic acid (TCA) cycle and anaerobic respiration that were found to be significantly different between cancer and control samples are shown in Table 2 . The majority of metabolites that have been identified show contradictory results in different studies, in terms of relative abundance in the cancer versus control groups. Lactate and fumurate were the most commonly recognised biomarkers.
|Author||Type of sample||Analytical platform||Glycolosis||Anaerobic respiration||The TCA cycle|
|Glucose||Fructose||Glyceraldehyde||Pyruvic acid||Lactic acid/Lactate||Citrate||Isocitric acid||α-Ketoglutaric acid/α-ketoglutarate||Succinate||Fumaric acid/Fumurate||Malate|
|Hirayama A (2009)||Tissue||CE–MS||↑||↑||↑||↑||↓||↑||↑|
|Cai Z (2010)||Tissue||GC–MS||↑||↑||↑||↑||↑||↓|
|Song H (2011)||Tissue||GC–MS||↑||↑|
|Ikeda A (Part B) (2011)||Serum||GC–MS||↓|
|Song H (2012)||Serum||GC–MS||↓|
|Aa J (2012)||Serum||GC–MS||↓||↑||↑||↑||↑|
|Zhang J (2011)||Serum||NMR||↑||↑||↑|
|Ikeda A (2011)||Serum||GC–MS||↓||↑||↑|
|Zhang J (2012)||Serum||LC–MS||↑|
|Hasim A (2012)||Urine||1H NMR||↑||↑||↓|
CE–MS, Capillary Electrophoresis–Mass Spectrometry; GC–MS, Gas Chromatography–Mass Spectrometry; NMR, Nuclear Magnetic Resonance Spectroscopy; LC–MS, Liquid Chromatography–Mass Spectrometry; 1H NMR, 1H Nuclear Magnetic Resonance.
3.4. Amino acid metabolites
Essential and non-essential amino acids that were found to be significantly different between cancer and control samples are shown in Table 3 . Valine, glutamine and glutamate were the most commonly identified amino acid biomarkers in descending order of frequency. The metabolite glutamine is the most consistent biomarker showing up-regulation in the serum, urine and tumour tissues of patients with OG cancer. There is a general trend of up-regulation of amino acids across the majority of studies, however in a study by Zhang et al., 35 they observed contradictory findings in serum samples from patients with oesophageal adenocarcinoma analysed by LC–MS.
|Author||Sample type||Analytical platform||Essential amino acid||Non-essential amino acid|
|Tugnioli V (2006)||Serum||HR-MAS-NMR||↑|
|Hiarayama A (2009)||Tissue||CE–MS||↑||↑||↑||↑||↑||↑||↑||↑||↑||↑||↑||↑||↑||↑||↑||↑||↑||↑||↑|
|Wu H (2010)||Tissue||GC–MS||↑||↑||↑||↑|
|Calabrese C (2012)||Tissue||HR-MAS-MRS||↑||↑|
|Yu L (2011)||Serum||GC–MS||↑||↑|
|Song H (2012)||Serum||GC–MS||↑||↓|
|Aa J (2012)||Serum||GC–MS||↑|
|Wu H (2009)||Tissue||GC–MS||↑||↑||↑||↑||↑||↑|
|Yakoub D (2010)||Tissue||HR-MAS-NMR||↑|
|Zhang J (2011)||Serum||NMR||↑||↑|
|Ikeda A (2011)||Serum||GC–MS||↑||↑||↑||↑|
|Zhang J (2012)||Serum||LC–MS||↓||↓||↓||↓||↓||↓|
|Hasim A (2012)||Urine||1H NMR||↓||↑||↑|
CE–MS, Capillary Electrophoresis–Mass Spectrometry; GC–MS, Gas Chromatography–Mass Spectrometry; NMR, Nuclear Magnetic Resonance Spectroscopy; LC–MS, Liquid Chromatography–Mass Spectrometry; 1H NMR, 1H Nuclear Magnetic Resonance; HR-MAS-NMR, High Resolution-Magic Angle Spinning-Nuclear Magnetic Resonance Spectroscopy; Val, valine; His, histidine; Phe, phenylalanine; Thr, threonine; Leu, leucine; Met, methionine; Trp, tryptophan; Ile, isoleucine; Lys, lysine; Tyr, tyrosine; Gln, glutamine; Glu, glutamate; Ser, serine; Asn, asparagine; Gly, glycine; Ala, alanine; Arg, arginine; Asn, asparagine; Asp, aspartic acid; Cys, cysteine; Pro, proline.
Several other metabolites related to amino acid metabolism have been identified in relation to OG cancer. Sulphur containing compounds, thought to be generated from the incomplete metabolism of methionine in the transamination pathway or by bacterial metabolism has been found to be up-regulated. 20 Two studies (from the same research group) investigating the volatile organic compounds (VOCs) in the headspace of tissue found a significant increase in the presence of carbon disulphide in cancer versus normal tissue. Buszewski et al., also demonstrated the presence of carbon disulphide in the headspace of Helicobacter pylori cultures. 21
3.5. Lipid metabolites
The saturated and un-saturated free fatty acids (FFA) that were found to be significantly different between cancer and control samples are shown in Table 4 . Metabolites of FFA oxidation, such as aldehydes and ketones have also been identified as potential biomarkers. Of the three endogenous ketones, 38 acetone and β-hydroxybutyrate have been described as potential biomarkers of oesophago-gastric cancer. β-Hydroxybutyrate has been shown to be up-regulated in the serum of patients with gastric cancer using GC–MS27 and 30 and in the serum of patients with oesophageal cancer using NMR. 35 The median concentration of acetone was found to be higher in the headspace of gastric content from patients with OG cancer versus controls. 36 However, it was found in lower quantities in the serum of oesophageal cancer patients when compared to healthy controls. 37 The study by Kumar et al. also identified aldehydes as potential biomarkers for cancer in the headspace of gastric content. This study demonstrated an increase in acetaldehyde in cancer patients compared with controls. 36
|Author||Sample type||Analytical platform||Un-saturated fatty acid||Saturated fatty acid|
|Palmitoleic acid||Cervonic acid||Linoleic acid||Oleic acid||Linolenic acid||Vaccenic acid||Gondoic acid||Arachidonic acid||Myristic acid||Enanthic acid||Margaric acid||Caprylic acid|
|Mun CW (2004)||Tissue||1H NMR||Decrease in lipids|
|Tugnioli V (2006)||Tissue||HR-MAS-NMR||Increase in tryiacyglycerides|
|Song H (2011)||Tissue||GC–MS||↓||↑||↓||↓|
|Calabrese C (2012)||Tissue||HR-MAS-MRS||Increase in tryiacyglycerides|
|Ikeda A (2011)||Serum||GC–MS||↑|
|Yu L (2011)||Serum||GC–MS||↑|
|Song H (2012)||Serum||GC–MS||↓||↓|
|Aa J (2012)||Serum||GC–MS||↑||↑||↑|
|Wu H (2009)||Tissue||GC–MS||↑||↑|
|Zhang J (2012)||Serum||LC–MS||↓||↓||↓||↑|
|Hasim A (2012)||Urine||1H NMR|
|Serum||Decrease in unsaturated lipids, VLDL and LDL|
CE–MS, Capillary Electrophoresis–Mass Spectrometry; GC–MS, Gas Chromatography–Mass Spectrometry; LC–MS, Liquid Chromatography–Mass Spectrometry; 1H NMR, 1H Nuclear Magnetic Resonance; HR-MAS-NMR, High Resolution-Magic Angle Spinning-Nuclear Magnetic Resonance Spectroscopy; VLDL, very-low-density lipoprotein; LDL, low-density lipoprotein.
Lipid metabolites of the cell membrane including: myo-inositol up-regulated in three studies using HR-MAS-NMR19 and 32 and GC–MS 31 ; choline and phospho-choline up-regulated in three studies of OG cancer tissue using HR-MAS-NMR18, 26, and 32 and down-regulated in the serum of oesophageal cancer using 1H NMR, 37 and squalene up-regulated in the gastric cancer tissue using GC–MS, 25 have also been identified as potential biomarkers.
3.6. Nucleotide metabolites
Studies of oesophageal cancer tissue have shown up-regulation of pyrimidine nucleotides using GC–MS, 31 and adenine and uridine containing compounds using HR-MAS-NMR. 32 A study of the serum of patients with oesophageal cancer demonstrates up-regulation of guanosine and cytidine containing compounds in comparison to healthy controls. 37
This systematic review demonstrates variation in the relative abundance of the metabolites of glycolysis, lactic acid fermentation, de novo lipid and amino acid synthesis in biological samples of patients with OG cancer versus controls. The results of several studies are contradictory; this variation may be due to sample selection and preparation, type of biological medium investigated or analytical technique. The metabolite glutamine is the most consistent biomarker, showing up-regulation in the serum, urine and tumour tissues of patients with OG cancer.22, 27, 28, 30, 32, and 37
Weinberg and Hanahan described six essential alterations in cell physiology which collectively dictate malignant transformation. 39 These characteristics were identified as the hallmarks of cancer and since 2000 reprogramming of energy metabolism has been added to the list.40 and 41 The increased rate of cell proliferation in cancer requires metabolic pathways to be redesigned to satisfy large demands for adenosine triphosphate (ATP), nicotinamide adenine dinucleotide phosphate (NADPH), nicotinamide adenine dinucleotide (NADH) and carbon skeletons. 42 The emerging field of metabolomics has aided in the identification of metabolite intermediates involved in these pathways, which have been identified as potential biomarkers in biological samples from cancer patients ( Fig. 3 ).
The metabolic intermediates of glycolysis including glucose, fructose, glyceraldehyde and pyruvic acid are shown to be up-regulated in several studies of gastric cancer tissue.22 and 23 This may be due to a phenomenon known as the Warburg effect,43 and 44 which states that even in the presence of oxygen, cancer cells reprogram their glucose metabolism to produce energy largely by glycolysis rather than oxidative phosphorylation via the tricarboxylic acid (TCA) cycle. 45 The poor efficiency of generating ATP by glycolysis is rationalised by the advantages of diverting glycolytic intermediates into various biosynthetic pathways, which are essential for the synthesis of necessary macromolecules (i.e. amino acids, nucleosides and lipids required for assembling new cells).45 and 46 The increased production of pyruvate by the Warburg effect and subsequent lactic acid fermentation propagated by the enzyme lactate dehydrogenase, may also explain the increased levels of lactate/lactic acid found in the tissues of gastric cancer,22 and 23 as well as in the serum of patients with oesophageal cancer.28, 34, and 35
The metabolite intermediates of the TCA cycle have been identified as biomarkers in the tissues of gastric cancer,22, 23, and 25 and in the serum of gastric and oesophageal cancer.28, 29, 30, 34, 35, and 37 Fumurate was the most frequently identified biomarker, but showed inconsistencies in relative abundance, with two studies observing down-regulation in cancer. The increased abundance of Krebs cycle metabolites, despite cancer cell preference of glycolysis to oxidative phosphorylation, is dependent on the process of anaplerosis. It refers to replenishment of metabolites in the cycle through the generation of α-ketoglutarate from the amino acid glutamate after its conversion from glutamine. 47 The aneplerotic flux generates macromolecules for cellular proliferation and is more specific than a high glycolytic flux, which can be induced by other factors such as hypoxia.48 and 49 This review confirms that in the majority of studies, there is an increased presence of glutamine and glutamate, suggesting an important role of this pathway in OG cancer metabolism.
In addition to higher levels of glutamine and glutamate, our review demonstrates the increase of other amino acid metabolites in the serum and tissues of oesophago-gastric cancer. The availability of amino acids in the tumour environment is important for cell proliferation and as an energy substrate. The source of these amino acids has not been determined; however it is likely to be a combination of systemic protein catabolism, 50 tumour environment degradation of extracellular matrix and de novo biosynthesis. 22 Contradictory findings of low levels of amino acids in the serum of oesophageal cancer patients in one study were attributed to the over-utilisation of amino acids in the tumour tissue. 35
Lipids and triglycerides have been shown to be present in increased levels in gastric cancer tissue by two studies using HR-MAR-NMR.19 and 26 One study confirmed these findings by demonstrating increased abundance of cytosolic lipid droplets with scanning electron microscopy. 26 Studies using GC–MS and LC–MS have demonstrated a mixed outcome of up-regulation and down-regulation, of saturated and unsaturated FFA in both tissue and serum. Low levels of FFA or lipids have been attributed to increased consumption by tumours due to their anabolic metabolism. However, much of the scientific evidence demonstrates the increased presence of fatty acids in the tumour environment as a consequence of metabolic reprogramming in the cancer disease state.51 and 52 There may be increased lipid availability from the bloodstream as a consequence of systemic lipolysis secondary to cancer cachexia or de novo fatty acid synthesis leading to FFA accumulation. 53 The role of lipid synthesis in cancer is not fully understood; however it is likely that de novo lipogenesis contributes to the formation of structural lipids for cell membrane synthesis, generation of energy by β-oxidation and cell signalling molecules in the cancer cell.51, 54, and 55
Fatty acid β-oxidation has been confirmed as a dominant pathway for energy generation in numerous malignancies including prostate and pancreatic cancer.56 and 57 Ketones and aldehydes, which are metabolic products of β-oxidation, are shown to be increased in biological samples of patients with OG cancer.27, 30, 34, and 36 Up-regulation of acylcarnitines, which are metabolic intermediates of β-oxidation involved in the transfer of FFA across the mitochondrial membrane, 58 has also been confirmed in metabolomic studies of kidney, 59 liver 60 and colorectal cancer. 61 Converse findings have been confirmed in a recent metabolic profiling study of the plasma of patients with oesophageal squamous cell cancer using LC–MS. 62
Local inflammation and oxidative stress associated with the cancer disease state have been shown to influence the production of eicosanoids from the lipid cell membrane and FFAs. 63 The increased abundance of eicosanoids have been confirmed in cell line, 64 animal 65 and human studies66 and 67 of OG cancer. Oxidative stress also promotes lipid peroxidation leading to the production of aldehydes, 68 which has been found in the headspace of gastric content for patients with OG cancer. 36 Lipid metabolic profiles have great potential in detecting OG cancer; however a limitation of current research is that only small molecular endogenous FFAs and intermediate metabolites have been investigated. Therefore, the roles of macromolecular lipids warrant further investigation.
The changes in the metabolic pathways that have been described may explain some of the observed differences in metabolites found in cancer and control samples. However, several of the hypotheses of cancer cell metabolism do not always hold true in human samples and contradictory findings are evident. The differences in study findings may be due to the type of biological sample analysed, its preparation and metabolite derivatisation, but consideration must also be given to the analytical platform employed. Variable detection limits and sensitivities of instruments may also influence the detection of metabolites that are present in a sample.
The majority of studies were unclear as to the nature of their patient selection and therefore we do not know if positive results are due to highly selected test subjects. Therefore, there can only be limited application of these results to real patient groups. A further weakness of this systematic review is the potential for publication bias and the lack of data representing no difference between compared groups. Due to lack of reported statistical data, we were not able to account for this phenomenon in the full cohort of included studies.
The application of metabolomics in cancer biomarker detection and development has largely not advanced beyond Phase 1 pre-clinical exploratory studies. Future studies have an obligation to provide transparent details of patient selection with attention being paid to variability in patient demographics, histology and treatment modalities (e.g. chemotherapy). Control subjects should be matched to cancer case subjects for factors that may potentially influence the identified biomarkers including age, sex, race, medication and lifestyle factors. Discovery of a biomarker or group of biomarkers with adequate sensitivity and specificity should be cross-validated across several analytical platforms and with different statistical methods. The development of an appropriate clinical assay suitable for screening purposes will then allow advancement of metabolomic biomarker discovery into phase 2 clinical studies and beyond.
Conflict of interest statement
- 1 A. Jemal, F. Bray, M.M. Center, et al. Global cancer statistics. CA Cancer J Clin. 2011;61(2):69-90 Crossref
- 2 M.B. Cook, W.H. Chow, S.S. Devesa. Oesophageal cancer incidence in the United States by race, sex, and histologic type, 1977–2005. Br J Cancer. 2009;101:855-859 Crossref
- 3 C. Lepage, B. Rachet, V. Jooste, J. Faivre, M.P. Coleman. Continuing rapid increase in esophageal adenocarcinoma in England and Wales. Am J Gastroenterol. 2008;103:2694-2699
- 4 C. Bosetti, F. Levi, J. Ferlay, et al. Trends in oesophageal cancer incidence and mortality in Europe. Int J Cancer. 2008;122:1118-1129
- 5 A. Jemal, R. Siegel, E. Ward, et al. Cancer statistics 2009. CA Cancer J Clin. 2009;59:225-249 Crossref
- 6 The NHS Information Centre. National oesophago-gastric cancer audit 2010. [online] < http://www.ic.nhs.uk/webfiles/Services/NCASP/audits%20and%20reports/NHS%20IC%20OGC%20Audit%202010%20interactive.pdf >; 2010 [accessed 1 November 2012].
- 7 C. Mariette, G. Taillier, I. Van Seuningen, J.P. Triboulet. Factors affecting postoperative course and survival after en bloc resection for esophageal carcinoma. Ann Thorac Surg. 2004;78(4):1177-1183 Crossref
- 8 G.B. Hanna, P.R. Boshier, A. Knaggs, R. Goldin, M. Sasako. Improving outcomes after gastroesophageal cancer resection: can Japanese results be reproduced in Western centres?. Arch Surg. 2012;147(8):738-745 Crossref
- 9 Y. Isobe, A. Nashimoto, K. Akazawa, et al. Gastric cancer treatment in Japan: 2008 annual report of the JGCA nationwide registry. Gastric Cancer. 2011;14(4):301-316 Crossref
- 10 E. Holmes, I.D. Wilson, J.K. Nicholson. Metabolic phenotyping in health and disease. Cell. 2008;134(5):714-717 Crossref
- 11 D.S. Wishart, T. Jewison, A.C. Guo, et al. HMDB 3.0 – the human metabolome database in 2013. Nucleic Acids Res. 2013;41 [database issue]
- 12 J.K. Nicholson, J. Connelly, J.C. Lindon, E. Holmes. Metabonomics: a platform for studying drug toxicity and gene function. Nat Rev Drug Disc. 2002;1:153-161 Crossref
- 13 S. Nishiumi, T. Kobayashi, A. Ikeda, et al. A novel serum metabolomics-based diagnostic approach for colorectal cancer. PLoS One. 2012;7(7) [epub ahead of print]
- 14 M. Phillips, K. Gleeson, J.M. Hughes, et al. Volatile organic compounds in breath as markers of lung cancer: a cross-sectional study. Lancet. 1999;353(9168):1930-1933 Crossref
- 15 A. Sreekumar, L.M. Poisson, T.M. Rajendiran, et al. Metabolomic profiles delineate potential role for sarcosine in prostate cancer progression. Nature. 2009;12:910-914 Crossref
- 16 P.F. Whiting, A.W. Rutjes, M.E. Westwood, et al. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med. 2011;155(8):529-536 Crossref
- 17 P. Whiting, A.W. Rutjes, J.B. Reitsma, et al. Sources of variation and bias in studies of diagnostic accuracy: a systematic review. Ann Intern Med. 2004;140:189-202 Crossref
- 18 C.W. Mun, J.Y. Cho, W.J. Shin, et al. Ex vivo proton MR spectroscopy (1H-MRS) for evaluation of human gastric carcinoma. Magn Reson Imaging. 2004;22(6):861-870 Crossref
- 19 V. Tugnoli, A. Mucci, L. Schenetti, et al. Ex vivo HR-MAS magnetic resonance spectroscopy of human gastric adenocarcinomas: a comparison with healthy gastric mucosa. Oncol Rep. 2006;16(3):543-553
- 20 T. Ligor, J. Szeliga, M. Jackowski, B. Buszewski. Preliminary study of volatile organic compounds from breath and stomach tissue by means of solid phase microextraction and gas chromatography–mass spectrometry. J Breath Res. 2007;1(1):1-6
- 21 B. Buszewski, A. Ulanowska, T. Ligor, et al. Identification of volatile organic compounds secreted from cancer tissues and bacterial cultures. J Chromatogr B Analyt Technol Biomed Life Sci. 2008;868(1–2):88-94 Crossref
- 22 A. Hirayama, K. Kami, M. Sugimoto, et al. Quantitative metabolome profiling of colon and stomach cancer microenvironment by capillary electrophoresis time-of-flight mass spectrometry. Cancer Res. 2009;69(11):4918-4925 Crossref
- 23 Z. Cai, J.S. Zhao, J.J. Li, et al. A combined proteomics and metabolomics profiling of gastric cardia cancer reveals characteristic dysregulations in glucose metabolism. Mol Cell Proteomics. 2010;9(12):2617-2628 Crossref
- 24 H. Wu, R. Xue, Z. Tang, et al. Metabolomic investigation of gastric cancer tissue using gas chromatography mass spectrometry. Anal Bioanal Chem. 2010;396:1385-1395 Crossref
- 25 H. Song, L. Wang, H.L. Liu, et al. Tissue metabolomic fingerprinting reveals metabolic disorders associated with human gastric cancer morbidity. Oncol Rep. 2011;6:431-438
- 26 C. Calabrese, A. Pisi, G. Di Febo, et al. Biochemical alterations from normal mucosa to gastric cancer by ex vivo magnetic resonance spectroscopy. Cancer Epidemiol Biomarkers Prev. 2008;17(6):1386-1395 Crossref
- 27 L. Yu, J. Aa, J. Xu, et al. Metabolomic phenotype of gastric cancer and precancerous stages based on gas chromatography time-of-flight mass spectrometry. J Gastroenterol Hepatol. 2011;26(8):1290-1297 Crossref
- 28 A. Ikeda, S. Nishiumi, M. Shinohara, et al. Serum metabolomics as a novel diagnostic approach for gastrointestinal cancer. Biomed Chromatogr. 2012;26(5):548-558 Crossref
- 29 H. Song, J.S. Peng, D.S. Yao. Serum metabolic profiling of human gastric cancer based on gas chromatography mass spectrometry. Braz J Med Biol Res. 2012;45(1):78-85 Crossref
- 30 J. Aa, L. Yu, M. Sun, et al. Metabolic features of the tumor microenvironment of gastric. Metabolomics. 2012;8:164-173 Crossref
- 31 H. Wu, R. Xue, C. Lu, et al. Metabolomic study for diagnostic model of oesophageal cancer using gas chromatography/mass spectrometry. J Chromatogr B Analyt Technol Biomed Life Sci. 2009;877(27):3111-3117 Crossref
- 32 D. Yakoub, H.C. Keun, R. Goldin, G.B. Hanna. Metabolic profiling detects field effects in nondysplastic tissue from esophageal cancer patients. Cancer Res. 2010;70(22):9129-9136 Crossref
- 33 D. Djukovic, H.R. Baniasadi, R. Kc, Z. Hammoud, D. Raftery. Targeted serum metabolite profiling of nucleosides in esophageal adenocarcinoma. Rapid Commun Mass Spectrom. 2010;24(20):3057-3062 Crossref
- 34 J. Zhang, L. Liu, S. Wei, et al. Metabolomics study of esophageal adenocarcinoma. J Thorac Cardiovasc Surg. 2011;141(2):469-475
- 35 J. Zhang, J. Bowers, L. Liu, et al. Esophageal cancer metabolite biomarkers detected by LC–MS and NMR methods. PLoS One. 2012;7(1):1-10
- 36 S. Kumar, J. Huang, J.R. Cushnir, et al. Selected ion flow tube-MS analysis of headspace vapor from gastric content for the diagnosis of gastro-esophageal cancer. Anal Chem. 2012;84(21):9550-9557
- 37 A. Hasim, H. Ma, B. Mamtimin, et al. Revealing the metabonomic variation of EC using 1H-NMR spectroscopy and its association with the clinicopathological characteristics. Mol Biol Rep. 2012;39(9):8955-8964 Crossref
- 38 L. Laffel. Ketone bodies: a review of physiology, pathophysiology and application of monitoring to diabetes. Diabetes Metab Res Rev. 1999;15(6):412-426 Crossref
- 39 D. Hanahan, R.A. Weinberg. The hallmarks of cancer. Cell. 2000;100:57-70 Crossref
- 40 D. Hanahan, R.A. Weinberg. Hallmarks of cancer: the next generation. Cell. 2011;144:646-674 Crossref
- 41 G. Kroemer, J. Pouyssegur. Tumor cell metabolism: cancer’s Achilles’ heel. Cancer Cell. 2008;13:472-482 Crossref
- 42 S. Romero-Garcia, J.S. Lopez-Gonzalez, J.L. Báez-Viveros, D. Aguilar-Cazares, H. Prado-Garcia. Tumor cell metabolism: an integral view. Cancer Biol Ther. 2011;12(11):939-948 Crossref
- 43 O. Warburg, F. Wind, E. Negelein. The metabolism of tumours in the body. J Gen Physiol. 1927;8(6):519-530 Crossref
- 44 O. Warburg. On the origin of cancer cells. Science. 1956;123:309-314
- 45 M.G. Vander Heiden, L.C. Cantley, C.B. Thompson. Understanding the Warburg effect: the metabolic requirements of cell proliferation. Science. 2009;324(5930):1029-1033 Crossref
- 46 R.A. Gatenby, R.J. Gillies. Why do cancers have high aerobic glycolysis. Nat Rev Cancer. 2004;4:891-899 Crossref
- 47 A. Schulze, A.L. Harris. How cancer metabolism is tuned for proliferation and vulnerable to disruption. Nature. 2012;491(7424):364-373 Crossref
- 48 J.M. Matés, J.A. Segura, J.A. Campos-Sandoval, et al. Glutamine homeostasis and mitochondrial dynamics. Int J Biochem Cell Biol. 2009;41:2051-2061
- 49 R.J. DeBerardinis, et al. Beyond aerobic glycolysis: transformed cells can engage in glutamine metabolism that exceeds the requirement for protein and nucleotide synthesis. Proc Natl Acad Sci U S A. 2007;104:19345-19350 Crossref
- 50 J.M. Argilés, J. Azcón-Bieto. The metabolic environment of cancer. Mol Cell Biochem. 1988;81:3-17
- 51 C.R. Santos, A. Schulze. Lipid metabolism in cancer. FEBS J. 2012;279:2610-2623 Crossref
- 52 M.T. Accioly, P. Pacheco, C.M. Maya-Monteiro, et al. Lipid bodies are reservoirs of cyclooxygenase-2 and sites of prostaglandin-E2 synthesis in colon cancer cells. Cancer Res. 2008;68:1732-1740 Crossref
- 53 J.A. Menendez, R. Lupu. Fatty acid synthase and the lipogenic phenotype in cancer pathogenesis. Nat Rev Cancer. 2007;7:763-777 Crossref
- 54 R.J. Deberardinis, N. Sayed, D. Ditsworth, C.B. Thompson. Brick by brick: metabolism and tumor cell growth. Curr Opin Genet Dev. 2008;18:54-61 Crossref
- 55 T. Mashima, H. Seimiya, T. Tsuruo. De novo fatty-acid synthesis and related pathways as molecular targets for cancer therapy. Br J Cancer. 2009;100:1369-1372 Crossref
- 56 Y. Liu. Fatty acid oxidation is a dominant bioenergetic pathway in prostate cancer. Prostate Cancer Prostatic Dis. 2006;9:230-234 Crossref
- 57 J. Khasawneh, M.D. Schulz, A. Walch, et al. Inflammation and mitochondrial fatty acid beta-oxidation link obesity to early tumor promotion. Proc Natl Acad Sci U S A. 2009;106:3354-3359 Crossref
- 58 K. Bartlett, S. Eaton. Mitochondrial β-oxidation. Eur J Biochem. 2004;271:462-469 Crossref
- 59 S. Ganti, S.L. Taylor, K. Kim, et al. Urinary acylcarnitines are altered in human kidney cancer. Int J Cancer. 2012;130(12):2791-2800 Crossref
- 60 J. Chen, W.Z. Wang, S. Lv, et al. Metabonomics study of liver cancer based on ultra-performance liquid chromatography coupled to mass spectrometry with HILIC and RPLC separations. Anal Chim Acta. 2009;650:3-9
- 61 Y.P. Qiu, G.X. Cai, M.M. Su, et al. Serum metabolite profiling of human colorectal cancer using GC–TOFMS and UPLC–QTOFMS. J Proteome Res. 2009;8:4844-4850 Crossref
- 62 J. Xu, Y. Chen, R. Zhang, et al. Global and targeted metabolomics of esophageal squamous cell carcinoma discovers potential diagnostic and therapeutic biomarkers. Mol Cell Proteomics. 2013; [epub ahead of print]
- 63 D. Wang, R.N. Dubois. Eicosanoids and cancer. Nat Rev Cancer. 2010;10(3):181-193 Crossref
- 64 S. Shimakura, C.R. Roland. Eicosanoid production by the human gastric cancer cell line ACS and its relation to cell growth. Cancer Res. 1992;52:1744-1749
- 65 X. Chen, N. Li, S. Wang, et al. Aberrant arachidonic acid metabolism in esophageal adenocarcinogenesis, and the effects of sulindac, nordihydroguaiaretic acid and alpha-difluoromethylornithine on tumorigenesis in a rat surgical model. Carcinogenesis. 2002;23(12):2095-2102 Crossref
- 66 A. Bennett, A. Civier, C.N. Hensby, P.B. Melhuish, I.F. Stamford. Measurement of arachidonate and its metabolites extracted from human normal and malignant gastrointestinal tissues. Gut. 1987;28(3):315-318 Crossref
- 67 L.M. Dong, X.O. Shu, Y.T. Gao, et al. Urinary prostaglandin E2 metabolite and gastric cancer risk in the Shanghai women’s health study. Cancer Epidemiol Biomarkers Prev. 2009;18(11):3075-3078 Crossref
- 68 F. Guéraud, M. Atalay, N. Bresgen, et al. Chemistry and biochemistry of lipid peroxidation products. Free Radic Res. 2010;44(10):1098-1124
a Department of Surgery and Cancer, Imperial College London, 10th Floor, QEQM Wing, St Mary’s Hospital, London W2 1NY, UK
b Centre for Pathology, Imperial College London, 4th Floor, Clarence Wing, St Mary’s Hospital, London W2 1NY, UK
⁎ Corresponding author: Address: Head of Division of Surgery, Department of Surgery and Cancer, 10th Floor, QEQM, St Mary’s Hospital, Praed St., London W2 1NY, UK. Tel.: +44 2078862124; fax: +44 2078866309.
© 2013 Elsevier Ltd, All rights reserved.