Table 5.
Accuracy of circulating tumor biomarkers for colorectal cancer recurrence and survival (excluding duplicate studies and those with unclear number or fewer than 20 recurrences or deaths).
| RECURRENCE | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Author, year | Biomarkers | Stage | Pre-op, post-op or time of recurrence | No. recurrence/ total | Sensitivity for recurrence | Specificity for recurrence | NPV for recurrence | PPV for recurrence | HR for recurrence (95% CI) | Adjusted HR (95% CI) | Mean/ Median recurrence-free survival |
| Ng; 2017 (117) | Mutations: based on tumor tissue findings | All stages (with mutations identified in patient tumor tissue prior to plasma testing) | Post-op; at recurrence | 26/44 | 73%; 96% |
83% | N/A | N/A | N/A | N/A | N/A |
| Ryan; 2003 (94) | Mutation: KRAS2 | All stages | Pre-op; Post-op | Pre: 20/123 Post: 20/94 |
52.6% | 92% | N/A | 62.5% | Pre: 2.07 (0.3-14.8) Post (in stage I-III): 6.37 (2.30-18.0) |
N/A | N/A |
| Tie; 2016 (99) | Mutations: based on tumor tissue findings | Stage II, no chemo (with mutations identified in patient tumor tissue prior to plasma testing) |
Post-op | 27/178 | 48% | 100% | 90.2% | 78.6% | 18 (7.9-40) | 28 (11-68) | N/A |
| Wang; 2004 (93) | Mutations: APC, p53, KRAS | All stages | Pre-op | 31/104 | 87% | 81% | 91% | 75% | N/A | N/A | N/A |
| Liu; 2016 (31) | Methylation: SST, MAL, TAC1, SEPT9, EYA4, CRABp1, NELL1 | Stage I-III | Pre-op | 43/150 | N/A | N/A | N/A | N/A |
SST: 2.40 (1.27-4.55) MAL: 1.15 (0.63-2.09) TAC1: 0.96 (0.53-1.74) SEPT9: 0.77 (0.42-1.40) EYA4:1.18 (0.65-2.15) CRABp1: 1.15 (0.63-2.10) NELL1:0.73 (0.37-1.44) |
SST: 2.60 (1.37–4.94) All others: not significant |
SST: 18.7% (vs. 38.7%), p = 0.005 |
| Young; 2016 (37) | Methylation: BCAT1, IKZF1 | All stages | At recurrence | 28/122 | 68% | 87% | N/A | N/A | N/A | N/A | N/A |
| Xue; 2017 (111) | Hypomethylation: CBS |
All stages | Pre-op | 43/95 | 74.4% | 59.6% | 74% | 62% | 1.62 (1.29-3.68) | 1.54 (1.18-3.02) | N/A |
| SURVIVAL | |||||||||||
| Author, year | Biomarkers | Stage | Pre-op or post-op | No. deaths/total | Sensitivity for death | Specificity for death | NPV for death | PPV for death | HR for poor survival (95% CI) | Adjusted HR (95% CI) | Mean/ Median survival |
| Schwarzenbach; 2008 (83) | DNA level (spectrophotometric quantification at 260 and 280nm) | Stage IV | Pre-op | 33/55 | N/A | N/A | N/A | N/A | N/A | N/A | Association with shorter survival (p = 0.02) |
| Li; 2017 (118) | cfDNA copy number variation | Stage III-IV | Post-op | 23/35 | N/A | N/A | N/A | N/A | 5.33 (6.76-94.44) | N/A | 15.87 months (vs. 68.53 months) |
| Bedin; 2017 (84) | DNA fragments (Alu83 and Alu244); Methylation: OSMR, SFRP1 |
All stages | Pre-op | 28/114 | N/A | N/A | N/A | N/A | Alu83: 3.49 (1.58-7.71) Alu244: 2.70 (1.25-5.84) |
Alu83: 2.71 (1.22-6.02) Alu244: 2.40 (1.11-5.19) |
Methylation: No association with survival |
| Lin; 2014 (90) | DNA copy number (cyclophilin); Mutations: 74 genes (including KRAS, APC, TP53, PIK3CA, BRAF) |
All stages | Pre-op | 62/191 | N/A | N/A | N/A | N/A | DNA copy number: 3.25 (1.66-6.45) Mutations: p>0.05 |
DNA copy number: 2.61 (1.31-5.19) | DNA copy number: 43% (vs. 78%), p = 0.001 Mutations: 48.8% (vs. 77%), p = 0.008 |
| Tie; 2015 (108) | Mutations: based on tumor tissue findings (including KRAS) | Stage IV (with mutations identified in patient tumor tissue prior to plasma testing) | Post-op | 20/53 | N/A | N/A | N/A | N/A | N/A | N/A | No association between change in ctDNA and survival |
| Herbst; 2017 112) |
Methylation: HPP1 | Stage IV (treated with a combination therapy containing a fluoropyrimidine, oxaliplatin, and bevacizumab) | Pre-op | 246/467 | N/A | N/A | N/A | N/A | 1.86 (1.37-2.35) | N/A | 21.9 months (vs. 35.2 months) |
| Liu; 2016 (31) | Methylation: SST, MAL, TAC1, SEPT9, EYA4, CRABp1, NELL1 | All stages | Pre-op | 58/165 | N/A | N/A | N/A | N/A |
SST: 2.40 (1.35-4.28) MAL: 2.26 (1.29-3.96) TAC1: 1.15 (0.67-1.97) SEPT9: 1.02 (0.59-1.74) EYA4:1.24 (0.73-2.12) CRABp1: 1.10 (0.64-1.87) NELL1:1.19 (0.68-2.07) |
SST: 1.96 (1.06-3.62) All others: not significant |
SST in stage II and III 16.1% (vs. 41.9%), p = 0.003 |
| Philipp; 2012 116) |
Methylation: HLTF, HPP1 | All stages | Pre-op | 190/311 | N/A | N/A | N/A | N/A | N/A | N/A |
HLTF: 36.3 months (vs. 80.2 months), p = 0.0001 HPP1: 12.6 months (104.7 months), p < 0.0001 |
| Rasmussen; 2018 (114) |
Methylation: ALX4, BNC1, HIC1, RARB, RASSF1A, SDC2, SEPT9, SRFP1, SRFP2, SPG20, TFP12, THBD, WIF1, APC, BMP3, BRCA1, CDKN2A, HLTF, MGMT, MLH1, NDRG4, NPTX2, NEUROG1, OSMR, PHACTR3, PPENK, SST, TAC1, VIM, WNT5A | All stages | Pre-op | 74/193 | N/A | N/A | N/A | N/A |
ALX4: 2.18 (1.41-3.38) BNC1: (2.93 (1.69-5.07) HIC1: 3.04 (1.46-6.32) RARB: 2.06 (1.31-3.23) RASSF1A: 2.75 (1.57-4.81) SDC2: 1.94 (1.23-3.07) SEPT9: 1.91 (1.21-3.02) SRFP1: 1.91 (1.21-3.02) SRFP2: 2.42 (1.53-3.84) SPG20: 2.66 (1.67-4.24) TFP12: 2.67 (1.38-5.19) THBD: 2.95 (1.68-5.18) WIF1: 3.26 (1.83-5.83) HLTF: 1.87 (1.04-3.38) MHMT: 2.27 (1.04-4.93) PPENK: 1.94 (1.07-3.50) TAC1: 1.56 (1.01-2.42) VIM: 1.82 (0.84-3.96) All others: not significant |
RARB: 1.99 (1.07-3.72) RASSF1A: 3.35 (1.76-6.38) All others: not significant |
N/A |
| Xue; 2017 (111) | Hypomethylation: CBS | All stages | Pre-op | 37/95 | 75.7% | 56.9% | 80% | 53% | 1.49 (1.15-2.64) | 1.35 (1.09-2.41) | N/A |
CI, Confidence interval; HR, hazard ratio; N/A, not applicable; NPV, negative predictive value; PPV, positive predictive value; post-op, post-operative; pre-op, pre-operative.