Metabolomic studies of human gastric cancer: Review

Table 1 Overview of gastric cancer detection and treatment via traditional methods compared with metabolomics

Cancer detection state/stage	Traditional methods	Metabolomics (biomarkers)	Ref.
Diagnosis	Endoscopy, biopsy	Lactic acid, butanedioic acid, malic acid, citric acids, pyruvic acid, 3-hydroxypropionic acid, serine, proline	[91,93,100,101]
Prognosis	Radiotherapy, chemotherapy surgery	Valine, isoleucine, serine, 3-indoxyl sulfate, hippurate, citrate	[96,99,102]
Metastasis	Computed tomography (CT) scanning, endoscopic ultrasonography (EUS), positron emission tomography (PET)	Sarcosine, alanine, proline, serine, myo-inositol, glycerol	[90,91,98,103]
Chemosensitivity of drugs	MTT chemosensitivity assay	1-acyl-lysophosphatidylcholines and polyunsaturated fatty acids	[75,104]

Table 2 Comparison of different analytical techniques employed in metabolomics

Method	Sampling characteristics	Sensitivity	Advantages	Disadvantages	Ref.
Nuclear magnetic resonance (NMR) spectroscopy	Non-destructive; minimum sample required	10-6	Fully automated with a high degree of reproducibility; relatively easy to identify metabolites from simple one-dimensional spectra	Lower sensitivity than mass spectrometry; co-resonant metabolites can be difficult to quantify; drug metabolites can be co-resonant with metabolites of interest	[20,41,105]
Gas chromatography-mass spectrometry (GC-MS)	Requires extraction, sample dried and chemical derivation	10-12	A relatively cheap and reproducible method with a high degree of sensitivity	Sample preparation can be time consuming; not all compounds are suitable for gas chromatography	[20,41,106,107]
Liquid chromatography-mass spectrometry (LC-MS)	Requires extraction and concentration (vacuum drying), liquid-liquid extraction	10-15	This method is increasingly being used in place of GC-MS as sample preparation is not as time consuming; has a sensitivity similar to GC-MS	More costly than GC-MS and depends on the reproducibility of liquid chromatography (more difficult to control than GC); can also suffer from ion suppression	[20,41,108,109]
Fourier-transform infrared (FT-IR) spectrometry	Uses vibrational frequencies of metabolites to produce a fingerprint of metabolism	10-6	Cheap and good for high-throughput first screening	Very difficult to identify which metabolites are responsible for causing changes; very poor at distinguishing metabolites within a class of compounds	[20,41,110,111]
Raman spectroscopy	Non-destructive; minimum sample required, occasionally hydration is needed	10-6	Has the advantage over FT-IR in that water has only a weak Raman spectrum; therefore, many functional groups can be observed	Very poor at distinguishing classes of compounds	[20,41,110,111]

Table 3 Overview of metabolomic studies on gastric cancer

Patients/xenograft model	Sample	Sample size (cancer + control)	Analytical method	Multivariate method	Major findings	Ref.
Both Xenograft model Patients	Urinary sample	33	GC-MS	PCA	Lactic acid, serine, proline, malic acid and fatty acids as potential markers for screening and early diagnosis	[93]
Patients	Serum	60	GC-MS	OPLS-DA	Sarcosine as a potential biomarker for the progression of gastric cancer metastasis	[98]
Patients	Plasma	80	GC-TOF-MS	PLS-DA	Azelaic acid, glutamate, urate, creatinine, threonate as markers for characterizing the precancerous stages and gastric cancer	[97]
Patients	Serum	50	GC-MS	PCA	3-hydroxypropionic acid and pyruvic acids as potential diagnostic markers for gastric cancer	[100]
Patients	Tissue	18	GC-MS with chemical derivatization	PCA	Valine, isoleucine, serine and phosphoserine for diagnosis and staging of gastric cancers	[99]
Xenograft model	Plasma	80	HPLC/Q-TOF-MS	PLS and hierarchical PLS	1-acyl-lysophosphatidylcholines and polyunsaturated fatty acids as potential indicators of chemosensitivity for gastric cancer	[75]
Xenograft model	Urinary sample	24	GC/MS	PCA	Lactic acid, butanedioic acid, malic acid and citric acids as potential markers for cancer screening. Alanine, proline, myo-inositol and glycerol as key markers for identifying cancer metastasis	[91]
Xenograft model	Tissue	22	GC/MS	PCA	Serine and proline metabolism pathways were enriched in cancer metastasis and may help elucidate the complex molecular mechanisms governing metastasis	[90]

PCA: Principal component analysis; PLS: Partial least squares; OPLS-DA: Orthogonal partial least squares discriminant analysis; GC-MS: Gas chromatography-mass spectrometry; PLS-DA: Partial least squares discriminant analysis; GC-TOF-MS: Gas chromatography coupled with time-of-flight mass spectrometry; HPLC/Q-TOF-MS: Uhra-performance liquid chromatography/quadrupole time-of-flight mass spectrometry.