Introduction
Materials and methods
Inclusion criteria
-
Primary human studies and meta-analyses in the last 5 years, which assessed serum or tissue markers’ diagnostic ability in CRC.
Exclusion criteria
-
Studies looking at familial or inherited CRC
-
Studies with less than 100 patients overall
-
Animal or in vitro studies
-
Studies that did not specify sensitivity or specificity of the markers.
Search strategy
Data extraction
Study outcomes
Results
Author (year) | Type of study | Marker | No. of patients | Blood | CRC versus control | Cutoff value | Compared to another marker | Early CRC versus control | Tissue | ||
---|---|---|---|---|---|---|---|---|---|---|---|
AUC | Sensitivity | Specificity | |||||||||
Li et al. (2016) [34] | Prospective | − 29b | 200 CRC 400 control | Plasma | 0.743 | 61.4% | 72.5% | No | No | Yes (sens 81.6%, spec 84.9%) | |
Basati et al. (2016) [35] | Retrospective | − 29b | 55 CRC 55 controls | Serum | 0.87 | 77% | 75% | 0.66 | No | No | No |
− 194 | 0.85 | 72% | 80% | 1.08 | |||||||
Imaok a et al. (2016) [11] | Retrospective | − 1290 | 211 CRC 56 adenomas, 57 controls | Serum | 0.830 | 70.1% | 91.2% | ND | No | Yes (level higher in IV compared to I–III + compared adenoma versus control) | Yes |
Zhi et al. (2015) [36] | Meta-analysis (5 studies) | − 29a | 281 CRC 299 healthy | 0.9128 | Pooled sens 59% | Pooled spec 89% | 1.33 0.545 7.2, 2 studies not specified | No | ND | No | |
Xu et al. (2015) [37] | Meta-analysis (7 studies) | − 21 | 676 CRC 417 controls | ND | 0.86 | Pooled sens 75% | Pooled spec 84% | 3.59 0.0043 0.0019 3.703 1.08 1.49, 1 study not specified | 2 studies: miR-16 U6RNA, cel-MIR-39, miR-16, miR-451, RNU6B | ND | No |
Lv et al. (2015) [38] | Retrospective | − 155 | 146 CRC 60 controls | Serum | 0.776 | 58.2% | 95% | 1.102 | No | High levels were correlated to diff/TNM but no sens/spec | No |
Chen et al. (2015) [39] | Prospective | − 106a | 100 CRC 79 cancer-free | Blood | 0.605 | 32% | 83.54% | 3.52 | No | ND | No |
− 20a | 0.590 | 46% | 73.42% | 2.44 | |||||||
Zhang et al. (2014) [40] | Meta-analysis (6 studies) | − 21 | 1071 patients | 3-Serum, 3-Plasma | 0.76 | 81% | 81% | ND | No | ND | Yes |
Yang et al. (2014) [41] | Meta-analysis (6 studies) | − 92a | 521 CRC 379 healthy | ND | 0.772 | 76% | 64% | 0.000,1 7 02972 1.231 2.87 240 | No | ND | No |
Xu et al. (2014) [42] | Prospective | − 375 | 94 CRC 46 healthy | Plasma | 0.7489 | 76.92% | 64.63% | 0.4852 | No | No | Yes |
Nonaka et al. (2014) [43] | Prospective | − 199a-3p | 84 CRC 32 non-cancer | Serum | 0.644 | 47.6% | 75% | 0.0010 | Yes (miR-21 AUC 0.675, sens 54.7%, spec 84.4%) | No | Yes |
Author (year) | Type of study | Panel of markers | No. of patients | Blood | CRC versus control | Compared to CEA/CA19-9 (sens/spec) | Early CRC versus control | Tissue | ||
---|---|---|---|---|---|---|---|---|---|---|
AUC | Sensitivity | Specificity | ||||||||
Zhu et al. (2017) [44] | Prospective | -19a-3p, miR-21-5p, -425-5p | 196 CRC 138 controls | Serum | 0.830 | ND | ND | No | No | Yes |
Vychytilova-Faltejskova et al. (2016) [45] | Prospective | -23a-3p -27a-3p -142-5p -376c-3p | 103 CRC 100 controls | Serum | 0.922 | 89 | 81 | Yes (CEA 47% sens, CA19-9 sens 27%) | Yes (I/II: AUC 0.877 sens 81 spec 81) | Yes |
Fang et al. (2015) [46] | Retrospective | -24, -320a, -423-5p | 111 CRC 59 adenomas, 24 polyps 29 IBD 130 healthy | Plasma | 0.899 | 92.79% | 70.77% | Yes (CEA sens 20.37%, spec 95%, CA19-9 sens 20.37%, spec 93.08%) | I/II: AUC 0.99 sens 90.79%, spec 70.77% | No |
Wang et al. (2014) [47] | Retrospective | -21, Let-7 g, -31, -92a, -181b, -203 | 83 CRC, 59 controls | Serum | 0.923 | ND | ND | Y (CA19-9 AUC 0.598, CEA, AUC 0.649) | No | No |
Zhang et al. (2013) [48] | Retrospective | -200c, -18a | 78 CRC, 86 normal | Plasma | 0.839 | 84.6% | 75.6% | Y (-18a: sens 73.1%, spec 79.1%, -200c: sens 64.1%, spec 73.3%) | No | Yes |
Luo et al. (2013) [49] | Prospective | -18a, -20a, -21-29a, -92a, -106b, -133a, -143, -145, -181b, -342-3p, 532-3p | 80 CRC 144 neoplasm-free | Plasma | 0.745 | ND | ND | ND | Yes (no major diff. between stages was found) | No |
Author (year) | Type of study | Marker | Number of patients | Blood | CRC versus control | Cutoff value | Compared to CEA/CA19-9 | Early CRC versus control | Tissue | ||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
AUC | Sensitivity (%) | Specificity (%) | |||||||||||
Nagai et al. (2017) [50] | Retrospective | LINE-1 hypomethylation index in cfDNA | 114 CRC 53 control | Plasma | 0.81 | 65.8 | 90 | 0.360 | cfDNA concentration (less accurate than cfDNA LHI) | Sens 63.2% spec 90% for early CRC | No | ||
Nian et al. (2017) [14] | Meta-analysis (25 studies) | Methylated SEPT 9 | 2975 CRC, 6952 adenoma | ND | 0.88 | 71 | 92 | Range of values (9 studies not specified) | N (* combined to FIT: 94% sens, 68% spec) | Sens 45% stage I, 70% stage II | No | ||
Wu et al. (2015) [15] | Prospective | Long non-coding RNA NEAT1_v1 | 100 CRC 100 controls | Whole blood | 0.787 | 69 | 79 | ND | No | Yes (no difference between mean values in early stage (I/II) compared to late stage III/IV) | Yes (no correl ation betwe en CRC tissue and whole blood) | ||
NEAT1_V2 | 0.871 | 70 | 96 | No | |||||||||
Pedersen et al. (2015) [51] | Prospective | methylatedBCAT1 + IKZF1 | 129 CRC, 685 adenoma, 1291 no neoplasia | Plasma | ND | 66 | 94 | ND | No | Yes (I/II sens 56%, III/IV 79%) | No | ||
Jin et al. (2015) [26] | Retrospective | Septin 9 | 135 CRC 169 polyps, 91 healthy | Plasma | ND | 74.8 | 87.4 | ND | No (but compared to FIT: sens 58%, spec 82.4%) | No | No | ||
Hao et al. (2014) [16] | Prospective | ALU115 of circulating free DNA | 104 CRC 63 polyps, 110 normal | Serum | 0.85 | 69.23% | 99.09% | 694 ng/ml | CEA (AUC 0.78, sens 42.31%, spec 100%). CEA + ALU115 + ALU247/1 15 = SENS 85.57% SPEC 97.27% | Yes (levels significantly different between primary CRC and polyps/controls) | No | ||
ALU247/115 of serum DNA | 0.89 | 73.08 | 97.27 | 0.52 ng/ml | |||||||||
Qi et al. (2013) [52] | Prospective | Alu-based cell-free DNA | 31 CRC 30 polyp 92 healthy | Serum | 0.904 | 64.5 | 98.9 | 634.9 ng/mL | Yes (CEA AUC 0.681, CA19-9 AUC 0.651-polyp versus CRC) | No | No | ||
Oh et al. (2013) [17] | Prospective | SDC2 methylation in DNA | 131 CRC 125 healthy | Serum | ND | 87 | 95.2 | 0.936 | No | Sens 92% stage I | Yes | ||
Rodia et al. (2016) [53] | Meta-analysis | TSPAN 8 | 67 CRC, 67 control | Whole blood | 0.751 | 83.6 | 58.2 | N/D | No | No | No | ||
LGALS4 | 0.746 | 82.1 | 61.2 | N/D | |||||||||
COL1A2 | 0.718 | 73.1 | 59.7 | N/D | |||||||||
CEACAM 6 | 0.632 | 65.7 | 61.2 | N/D | |||||||||
TSPAN8 + LGALS 4 | 0.862 | 92.54 | 67.16 | N/D | |||||||||
Li et al. (2016) [34] | Retrospective | mRNAM | 200 CRC, 200 benign disease | Tissue | 0.823 | 89.6 | 84.5 | 0.3562 | No | No | Yes | ||
0.802 | 80.5 | 76.5 | 0.3023 | ||||||||||
0.814 | 86.5 | 82.9 | 0.3243 | ||||||||||
Khales et al. (2015) [19] | Prospective | mRNA | 51 CRC, 60 healthy | Serum | 0.981 | 96.1 | 95 | 14,000 copies/1 ml of blood | Yes (CEA-correlated to SALL4) | Yes (no diff between early CRC) | Yes | ||
Wang et al. (2014) [47] | Prospective | mRNA | 92 CRC, 60 healthy | Serum | 0.855 | 84.8 | 80 | 0.128 | CEA (AUC 0.691) | No I, 28.8% Stage II | No | ||
Qi et al. (2013) [52] | Prospective | Alu-based cell-free DNA | 31 CRC 30 polyp 92 healthy | Serum | 0.904 | 64.5 | 98.9 | 634.9 ng/mL | Yes (CEA AUC 0.681, CA19-9 AUC 0.651-polyp versus CRC) | No | No | ||
Oh et al. (2013) [17] | Prospective | SDC2 methylation in DNA | 131 CRC 125 healthy | Serum | ND | 87 | 95.2 | 0.936 | No | Sens 92% stage I | Yes | ||
Rodia et al. (2016) [53] | Meta-analysis | TSPAN 8 | 67 CRC, 67 control | Whole blood | 0.751 | 83.6 | 58.2 | N/D | No | No | No | ||
LGALS4 | 0.746 | 82.1 | 61.2 | N/D | |||||||||
COL1A2 | 0.718 | 73.1 | 59.7 | N/D | |||||||||
CEACAM 6 | 0.632 | 65.7 | 61.2 | N/D | |||||||||
TSPAN8 + LGALS 4 | 0.862 | 92.54 | 67.16 | N/D | |||||||||
Li et al. (2016) [34] | Retrospective | mRNAM | 200 CRC, 200 benign disease | Tissue | 0.823 | 89.6 | 84.5 | 0.3562 | No | No | Yes | ||
0.802 | 80.5 | 76.5 | 0.3023 | ||||||||||
0.814 | 86.5 | 82.9 | 0.3243 | ||||||||||
Khales et al. (2015) [19] | Prospective | mRNA | 51 CRC, 60 healthy | Serum | 0.981 | 96.1 | 95 | 14,000 copies/1 ml of blood | Yes (CEA-correlated to SALL4) | Yes (no diff between early CRC) | Yes | ||
Wang et al. (2014) [47] | Prospective | mRNA | 92 CRC, 60 healthy | Serum | 0.855 | 84.8 | 80 | 0.128 | CEA (AUC 0.691) | No | No |
Author (year) | Type of study | Marker | Number of patient s | Blood | CRC versus control | Cutoff value | Compared to CEA/CA19-9 | Early CRC versus control | Tissue | ||
---|---|---|---|---|---|---|---|---|---|---|---|
AUC | Sensitivity (%) | Specificity (%) | |||||||||
Wang et al. (2017) [54] | Retrospective | Macrophag e Inhibitory Cytokine (MIC-1/GDF15) | 473 CRC, 25 polyps, 489 controls | Serum | 0.866 | 43.8 | 96.7 | 1000 pg/mL | CEA (AUC 0.728, sens 36.6%) CEA + MIC-1 = AUC 0.886, sens 72.7% 89% specificity | AUC 0.843 sens 38.5% | No |
Xu et al. (2016) [55] | Meta-analysis (7 for diagnostic) | IL-6 | 687 CRC, 392 controls | ND | 0.79 | 72 | 74 | 2.14, 3.06 4.24, 6.70 and 3 studies ND | n/a | n/a | No |
Zheng et al. (2015) [56] | Prospective | Growth-related gene product beta 1 | 123 CRC, 125 non-tumor, 88 healthy | Serum | 0.834 | 56.1 | 95.31 | 105 pg/mL | CEA AUC 0.739 CA19-9 AUC 0.676 | If combine d to CEA, then detects early CRC 22.2% from 5.6% for stage I, 66.7% from 41% for stage II | No |
Xia et al. (2015) [20] | Meta-analysis (5 diagnostic) | IL-8 | 725 total | ND | 0.92 | 70 | 91 | 17.71 pg/mL, 39.5 pg/mL, 44.26 pg/mL, 8.83 pg/mL | ND | No | No |
Author (year) | Type of study | Marker | Number of patients | Blood | CRC versus control | Cutoff value | Compared to CEA/CA19-9 | Early CRC versus control | Tissue | ||
---|---|---|---|---|---|---|---|---|---|---|---|
AUC | Sensitivity (%) | Specificity | |||||||||
Kunizaki et al. (2016) [57] | Prospective | A Anti-p53 | 170 | serum | ND | 30.6 | ND | ND | CEA and anti-p53: detection | 31.9% | No |
Chen et al. (2016) [58] | Prospective | Anti-p53 | 49 CRC 99 AA, 29 non-advanced adenomas, 224 controls | Serum | ND | 18 | 90% | ND | No | 26% sensitivity for early CRC | No |
Anti-IMPDH2 | |||||||||||
Anti-MDM2 | |||||||||||
Anti-MAGEA4 | |||||||||||
Wang et al. (2016) [59] | Prospective | FnIgA | 258 CRC 150 benign, 200 healthy | Serum | 0.704 | 36.43 | 92.7% | 0.45 | Yes CEA and CA19-9. If CEA + CA19-9 + anti FnIgA: AUC 0.858, sen 54.65% spec 96.6% | Sens 27.7% spec 96.21% | No |
FnIgG | 0.645 | 77.52 | 46.94% | 0.42 | ND |
Author | Type of study | Marker | Number of patients | Blood | CRC versus control | Cutoff value | CEA/CA19-9 | Early CRC versus control | Tissue | ||
---|---|---|---|---|---|---|---|---|---|---|---|
AUC | Sensitivity | Specificity | |||||||||
Fei et al. (2017) [60] | Retrospective | RBP4 | 402 CRC, 218 Normal | Serum | 0.852 | 74.9% | 81.7% | 26.7 µg/mL | Y (CEA: AUC 0.817, CA19-9: AUC 0.634) | ND | No |
THBS2 | 0.794 | 64.9% | 87.1% | 14.85 ng/mL | |||||||
RBP4 + CEA | 0.927 | 80.8% | 91.2% | ND | |||||||
Li et al. (2016) [21] | Prospective | TFF 3 | 127 CRC, 77 controls (35 polyps, 42 controls) | Serum | 0.889 | 74.2% | 94.8% | 5.591 ng/m L | Yes (CEA: AUC 0.715 sens 62.2%, spec 72.7%) | Yes (TFF3 significantly higher in stage I CRC comp to controls but not diff. if polyps) | No |
Werner et al. (2016) [9] | Retrospective | CEA Ferritin Seprase Osteoponti n Anti-p53 | 36 CRC 420 advanc ed adenom a, 1200 controls | Serum | 0.78 | 44% | 90% | ND | Yes (CEA sens 50%, spec 90%) | Yes (early cancers were detected at least as well as late-stage) | No |
CEA + anti-p53 | 0.85 | 58% | 90% | ||||||||
Wang et al. (2016) [61] | Prospective | COL3A1 | 86 CRC 21 enteritis, 3 polyps, 68 normal | Plasma Epithelial tissue | 0.92 0.975 | 98.8% 95.2% | 69.1% 91.1% | 54.23 ng/m L | Yes (CEA: AUC 0.791, sens 70.2%, spec 73%) | Yes | Yes (mRNA) |
Rho et al. (2016) [22] | Retrospective | BAG4 IL6ST** VWF EGFR BAG4 IL6ST VWF CD44 | 60 CRC 60 adenom as, 30 control | Plasma/Serum | 0.81 0.79 | 40.9% 42.44% | 90% | ND | No | No | Yes |
Overholt et al. (2016) [22] | Prospective | CA11-19 | 131 CRC, 65 polyps, 182 benign disease, 103 controls | Serum | ND | 98% | 84% | 6.4 units/mL | No | No | No |
Gezer et al. (2015) [62] | Prospective | Trimethylations of lysine 9 on histone 3 (H3K9me3) Trimethylations of lysine 20 on histone 4 (H4K20me3) | 63 CRC 40 cancer-free | Plasma | No significant difference between CRC and controls 0.715 | ND 14.3% | Yes 95% | No | No | ||
H3K27me3 | 0.620 | 17.5% | 95% | ||||||||
H3K27me3 + H4K20me3 | 0.769 | 28.6% | 95% | ||||||||
Xue et al. (2015) [63] | Prospective | Zinc-alpha-2-glycoprote in (AZGP1) | 120 CRC, 40 healthy | Serum | 0.742 | 55.8% | 85% | 2297.71 ng/ mL | Yes | No | Yes |
AZGP1 + CEA + CA1 9-9 | 0.805 | 67.5% | 82.5% | ||||||||
Wang et al. (2015) [64] | Prospective | Angiopoeti n-2 | 98CRC 90 healthy | Serum | 0.859 | 79.3% | 82.4% | 2710 pg/mL | No | No | No |
Storm et al. (2015) [65] | Retrospective | CL-L1 M-ficolin MAp44 | 99CRC 196 adenom as, 696 no cancer | Serum | 0.68 | 36% | 83% | ND | No | No | No |
Fung et al. (2015) [66] | Prospective | IGFBP2 DKK3 PKM2 | 98 CRC 99 controls | Serum | 0.91 | 73% | 95% | ND | No (but compared to FOBT and FIT and equivalent) | Yes (stage I: sens 59%, II 84%, III 71%, IV 78% for spec 95%) | No |
Sole et al. (2014) [67] | Retrospective | COL10A1 | 80CRC 23 adenom a, 77 controls | Serum | 0.76 | 63% | 85% | 208 ng/mL | ND | No | Yes |
Shin et al. (2014) [68] | Retrospective | Melanotra nsferrin | 228 CRC, 20 polyps, 77 healthy | Plasma | 0.723 | 48.2% | 92.5% | ND | Yes (CEA, PAI-1) AUC of regressed TRFM-PAI1-CEA 0.821, sens 67.5%, spec 90% | Yes | Yes |
Shirahata et al. (2014) [69] | Prospective | Vimentin methylation | 242 CRC 25 healthy | Serum | ND | 32.6% | ND | 0.0485 ng/mL | Vimentin + CEA + CA19-9 = sens 55.6% | Sens 57.1% for stage 0, 30.6% | No |
Jiang et al. (2014) [24] | Prospective | sDC-SIGN & sDC-SIGNR | 182 CRC, 101 healthy | Serum | 0.9885 | 98.7% | 94.8% | sDC-SIGN 2.226 μg/mL/sDC-SIGNR 222.7 ng/m L | Yes (CEA sens 29.22%, CA19-9 14.67%) | Yes(better at detecting early CRC compared to CEA/CA19-9; DC-SIGN sens 81.33, spec 55.56%; DC-SIGNR sens 48.65%, spec 92.5% | Yes |
Wang et al. (2013) [70] | Prospective | Kininogen | 140 CRC, 80 adenom as, 85 healthy | Serum | 0.706 | 63.64% | 65.88% | 162.99 μg/ml | Yes (CEA: AUC: 0.695 sens 38.46%, spec 85.88%)kininogen-1 or CEA: sens 79.92%, spec 58.82% | Yes | Yes |