Table 1.
Ref. | Authors (year) | No. of hospitals | Sample size | No. of events | Disease stage | Chemo-therapy | Loss of CDKN2A (p16) (vs. expression) or CDKN2A methylation (vs. no methylation) | |||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
OS | CS | Method | p16 lost (methylated) cases | 5-year OS or CS* | Log-rank p value | OS univariate HR (95% CI) | p value | OS multivariate HR (95% CI) | p value | Covariates examined and other notes | ||||||
14 | Liang et al. (1999) | 1 | 84 | - | - | T3 (N0M0) | 5-FU, leucovorin, levamisole | MSP | 29% | 25% (vs. 72%) | 0.0001 | - | - | - | - | age, sex, tumor location, tumor size, differentiation, pathological type, vascular invasion, lymphocytes, CEA, KRAS, TP53, LOH, MSI |
15 | Esteller et al. (2001) | 1 | 86 | - | - | Duke's A-C | - | MSP | 35% | 70% (vs. 81%) | NS | - | - | - | - | age, sex, tumor location, KRAS. CDKN2A methylation was associated with worse outcome in KRAS-wild-type cases (HR 3.3; 95% CI, 1.0-10.5). |
16 | Norrie et al. (2003) | 1 | 404 | - | - | NA | - | MSP | 25% | - | NS | - | - | - | - | age, sex, tumor location, tumor grade, histopathological type, lymphatic reaction, MSI |
17 | Maeda et al. (2003) | 1 | 90 | - | - | Duke's B-C | - | q-MSP | 13% | - | 0.022 | - | - | 1.013 (1.003-1.023) | 0.014 | age, sex, tumor location, tumor differentiation. HR for a unit increase of PMR |
18 | Kohonen-Corish et al. (2005) | 1 | 149 | - | - | Duke's C | adjuvant therapy | MSP | 35% | - | 0.966 | - | - | - | - | Age, sex, tumor size, tumor grade, venous invasion, MSI, MGMT methylation |
19 | Sanz-Casla et al. (2005) | 1 | 104 | 54 | 46 | Duke's A-D | adjuvant therapy | MSP | 18% | 75% (vs. 61%) | 0.09 | - | - | - | NS | age, sex, tumor location, tumor grade, histopathological type |
20 | Ishiguro et al. (2006) | 1 | 88 | - | - | I-IV | - | MSP | 23% | 35% (vs. 29%) | 0.57 | - | - | - | - | age, sex, tumor location, tumor differentiation, CEA, CDKN2B (p15) methylation |
21 | Shen et al. (2007) | many | 182 | 153 | - | IV | Five-arm trial | COBRA | 17% | 90% (vs. 83%) | NS | 1.51 | 0.05 | 1.42 | 0.13 | age, sex, tumor location, performance status, treatment arm, methylation (MINT-1, MINT-31, MLH1, p14) |
22 | Krtolica et al. (2007) | 1 | 47 | - | - | Duke's A-D | - | MSP | 51% | - | NS | - | NS | - | NS | sex, tumor differentiation, histopathological type, tumor location, MGMT methylation, KRAS, CD44, PCNA, Laminin, CDH1. |
23 | Mitomi et al. (2009) | 1 | 151 | 64 | - | II-IV | 5-FU, mitomycin, leucovorin | q-MSP | 30% | *43% (vs. 76%) | - | *2.95 (1.49-5.86) | 0.002 | *3.38 (1.67-6.84) | <0.001 | age, sex, tumor location, tumor differentiation, tumor size, CDKN2A (p16) expression |
24 | Kim et al. (2010) | 1 | †131 | - | - | I-IV | adjuvant therapy | Pyrosequencing | 12% | - | <0.0001 | - | - | 0.31 (0.11-0.88) | 0.028 | Rectal cancer only. Age, stage, preoperative serum CEA, lymphovascular invasion |
Current study | many | 902 | 409 | 235 | I-IV | - | q-MSP | 30% | 46% (vs. 55%) | 0.0035 | 1.36 (1.10-1.66) | 0.0035 | 1.03 (0.82-1.31) | 0.78 | age, sex, tumor location, grade, year of diagnosis, BMI, family history, CIMP, MSI, KRAS, BRAF, LINE-1, TP53, CDKN1A, CDKN1B, CCND1, CTNNB1, FASN. | |
25 | Zhao et al. (2003) | 2 | 32 | 12 | - | Duke's A-C | - | IHC | 41% | 62% (vs. 79%) | <0.001 | - | - | - | - | tumor grade, CDK4 expression, CDKN2A methylation |
16 | Norrie et al. (2003) | 1 | 161 | - | - | NA | - | IHC | 8.1% | - | NS | - | - | - | - | age, sex, tumor location, tumor grade, histopathological type, lymphatic reaction, MSI |
26 | Cui et al. (2004) | 1 | 117 | - | - | I-IV | - | IHC | 62% | 65% (vs.84%) | 0.025 | - | - | *5.99 (2.97-12.1) | <0.0001 | age, sex, tumor location, tumor grade, tumor size, RB |
27 | González-Quevedo et al. (2004) | NA | 79 | - | - | NA | - | Western blot | 33% | - | NS | - | - | - | - | CDKN2A methylation |
28 | Lyall et al. (2006) | 1 | 90 | 55 | - | III | - | IHC | 91% | MS 48 months (vs. 10 months) | 0.001 | - | - | - | - | age, sex, tumor location, tumor differentiation, expression of 22 other markers |
29 | Lam et al. (2008) | 1 | 194 | 57 | - | I-IV | - | IHC | 20% | - | NS | - | - | - | NS | age, sex, tumor location, tumor grade, histopathological type, tumor size |
30 | Wassermann et al. (2009) | NA | 82 | - | - | II-III | adjuvant therapy | IHC | - | - | - | - | - | - | - | In colorectal cancers that displayed budding at infiltrative front with CTNNB1 nuclear expression, CDKN2A (p16) expression was associated with high mortality (p<0.001). |
Current study | many | 804 | 367 | 223 | I-IV | - | IHC | 25% | 46% (vs. 54%) | 0.026 | 1.30 (1.03-1.63) | 0.026 | 1.08 (0.83-1.42) | 0.55 | age, sex, tumor location, grade, year of diagnosis, BMI, family history, CIMP, MSI, KRAS, BRAF, LINE-1, TP53, CDKN1A, CDKN1B, CCND1, CTNNB1, FASN. |
Colorectal cancer-specific survival.
Rectal cancer cases only. The HUGO Gene Nomenclature Committee (HGNC)-approved official gene/protein symbols are used (www.genenames.org).
BMI, body mass index; COBRA, combined bisulfite and restriction analysis; CEA, carcinoembryonic antigen; CI, confidence interval; CIMP, CpG island methylator phenotype; CS, colorectal cancer-specific survival; 5-FU, 5-fluorouracil; IHC, immunohistochemistry; KM, Kaplan-Meier; MS, median survival; MSI, microsatellite instability; MSP, methylation-specific PCR; NA, not available; NS, not significant; OS, overall survival; PMR, percentage of methylated reference; q-MSP, quantitative MSP.