Table 1.

Studies investigating hematologic metabolic biomarkers in endoluminal gastrointestinal cancer

Ref.StudyCancerTotal n (Cancer n)Biomarker discoveryAnalytical platform(s)Final classification methodDiagnostic indices of final methodSignificant featuresFeatures after MC
Studies investigating esophageal malignancies
 (39)Djukovic 2010EAC26 (14)TargetedUPLC-TQMS8 metabolitesNot reported85a
 (40)Zhang 2011EAC118 (68)NontargetedNMRPLS-DA MV modelAUROC 0.89, Sen 88%, Sp 92%88
 (41)Zhang 2012EAC113 (67)NontargetedLC-MS (& NMR)PLS-DA MV modelAUROC 0.950, Sen 89%, Sp 90%12 + 813a
 (42)Zhang 2013“EC”50 (25)NontargetedNMR & UPLC -diode arrayOPLS-DA MV modelNot reportedNMR > 25 UHPLC = 7NMR = 12 UHPLC = 7
 (43)Hasim 2012ESCC148 (108)Nontargeted1 h-NMROPLS-DA MV modelNot reportedP values not reportedP values not reported
 (44)Liu 2013ESCC152 (72)NontargetedUPLC-ESI-TOFMS15 metabolitesNot reported1511
 (45)Xu 2013ESCC228 (124)NontargetedRRLC/ESI-MSLRM (7 metabolites)AUROC 0.961, Sen 90%, Sp 96%183a
 (46)Jin 2014ESCC110 (80)NontargetedGC-MSOPLS-DA MV modelNot reported4339a
 (47)Ma 2014ESCC111 (51)TargetedHPLCPLS-DA MV modelNot reported1211a
Studies investigating gastric malignancies
 (48)Yu 2011GAC28 (9)NontargetedGC-MSPLS-DA MV modelNot reported123a
 (49)Aa 2012GAC37 (17)NontargetedGC-MSOPLS-DA MV modelNot reported150a
 (50)Song 2012GAC60 (30)NontargetedGC-MSOPLS-DA MV modelNot reported186a
Studies investigating colorectal malignancies
 (51)Zhao 2007CRC258 (133)NontargetedLC-MSLRM (4 metabolites)Sen 82%, Sp 93%108a
 (52)Qiu 2009CRC129 (64)NontargetedGC-MS; UPLC-MSOPLS-DA MV modelNot reported3816a
 (53)Ludwig 2009CRC57 (38)Nontargeted1 H-NMRPCA MV ModelSen “about 70%”P values not reportedP values not reported
 (29)Ritchie 2010CRC443 (223)NontargetedFTICR-MS, NMR, UPLC MS/MS3 metabolitesAUROC for each 0.975050a
 (54)Kondo 2011CRC50 (42)NontargetedGC-MSPLS-DA MV modelNot reported90a
 (55)Leichtle 2011CRC117 (59)TargetedESI-MS/MSLRM (2 metabolites and CEA)AUROC 0.881111
 (56)Ma 2012CRC38 (30)NontargetedGC-MSHierarchical clustering (6 metabolites)Sen 94%62a
 (57)Nishiumi 2012CRC242 (119)NontargetedGC-MSLRM (4 metabolites)AUROC 0.91, Sen 85%, Sp 85%7646a
 (58)Li 2013CRC104 (52)NontargetedDI±ESI FTICR-MSLRM (10 metabolites)AUROC 0.99, Sen 98%, Sp 100%P values not reportedP values not reported
 (59)Li 2013CRC360 (120)NontargetedLC-MS/MSLRM (4 metabolites)Sen 89% Sp 80%1616
 (30)Ritchie 2013CRC5,883TargetedMS/MS TQMRM1 metaboliteSen 85.7%, Sp 53%b1n/a
 (60)Tan 2013CRC204 (102)NontargetedGC-TOFMS & UPLC-QTOFMSOPLS-DA MV modelSen 100%, Sp 100%7262a
 (61)Wang 2014CRC36 (16)NontargetedSPME-GC/MSPLS-DA MV modelNot reported44
 (62)Zamani 2014CRC66 (33)Nontargeted1 H-NMROPLS-DA MV modelNot reportedP values not reportedP values not reported
 (63)Zhu 2014CRC234(66)TargetedLC-MS/MSPLS-DA MV modelAUROC 0.93, Sen 96%, Sp 80%427a
Studies investigating multiple malignancies
 (64)Miyagi 2011GAC1,383 (199)TargetedHPLC-ESI-MSLDAAUROC 0.82–0.8566
CRC1,383 (199)LDAAUROC 0.87–0.881010
 (65)Ikeda 2012ESCC50 (15)NontargetedGC-MS2 metabolitesSen 80%–81%, Sp 59%–90%91a
GAC50 (11)2 metabolitesSen 70%–84%, Sp 71%–90%;50a
CRC50 (12)3 metabolitesSen 54.5%–81.8%, Sp 66.7%–91.6%120a

NOTE: Where studies included discovery and validation cohorts, diagnostic metrics of the validation set included for analysis.

Abbreviations: EAC, esophageal adenocarcinoma; ESCC, esophageal squamo-cellular carcinoma; GAC, gastric adenocarcinoma; CRC, colorectal adenocarcinoma; UPLC-TQMS, ultra-performance liquid chromatography-triple quadrupole mass spectrometry; NMR, nuclear magnetic resonance spectroscopy; ESI-TOFMS, electrospray ionization time-of-flight mass spectrometry; RRLC, rapid relaxing liquid chromatography; GC-MS, gas chromatography mass spectrometry; HPLC, high-performance liquid chromatography; FTICR-MS, Fourier transform ion cyclotron mass spectrometry; MS/MS, tandem mass spectrometry; TQMRM, triple quadrupole multiple reaction monitoring; DI, direct ionization; SPME, solid phase microextraction; PLS-DA, partial least squares discriminant analysis; ROC, receiver operating characteristic curve; PCA, principle component analysis; OPLS-DA, orthogonal projection to latent structures discriminant analysis; LRM, logistic regression model; LDA, linear discriminant analysis; MC, multiplicity correction; MV, multivariable; AUROC, area under receiver operating characteristic curve; Sen, sensitivity; Sp, specificity.

  • aWe applied a Bonferroni correction (α/n compared features).

  • bSpecificity not stated but calculated from available data.