Abstract
Background
Hypomethylation in long interspersed nucleotide element-1 (LINE-1) and high-degree microsatellite instability (MSI-high) in colorectal cancer (CRC) have been associated with inferior and superior survival, respectively; however, it remains uncertain whether the prognostic association of LINE-1 hypomethylation differs by MSI status. We hypothesized that the adverse prognostic association of LINE-1 hypomethylation might be stronger in MSI-high CRCs than in microsatellite stable (MSS) CRCs.
Methods
Utilizing 1211 CRCs in the Nurses’ Health Study and the Health Professionals Follow-up Study, we examined patient survival according to LINE-1 hypomethylation status in strata of MSI status. A Cox proportional hazards model was used to compute multivariable CRC-specific mortality hazard ratios (HRs) for a 10% decrease in LINE-1 methylation level (range = 23.1–93.1%), adjusting for potential confounders, including CpG island methylator phenotype, and KRAS, BRAF, and PIK3CA mutations. Statistical tests (log-rank test, chi-square test, and likelihood ratio test) were two-sided.
Results
In MSI-high cancers, the association of LINE-1 hypomethylation with higher mortality (HR = 2.45, 95% confidence interval [CI] = 1.64 to 3.66, P < .001) was stronger than that in MSS cancers (HR = 1.10, 95% CI = 0.98 to 1.24, P = .11) (P interaction < .001, between LINE-1 and MSI statuses). In MSI-high cases with CRC family history, the association of LINE-1 hypomethylation with higher mortality (HR = 5.13, 95% CI = 1.99 to 13.2; P < .001) was stronger than that in MSI-high cases without CRC family history (HR = 1.62, 95% CI = 0.89 to 2.94, P = .11) (P interaction = .02, between LINE-1 and CRC family history statuses).
Conclusions
The association of LINE-1 hypomethylation with inferior survival is stronger in MSI-high CRCs than in MSS CRCs. Tumor LINE-1 methylation level may be a useful prognostic biomarker to identify aggressive carcinomas among MSI-high CRCs.
Colorectal cancer (CRC) represents a group of molecularly heterogeneous diseases characterized by differing sets of epigenetic and genetic abnormalities, including DNA hypomethylation in long interspersed nucleotide element-1 (LINE-1) and microsatellite instability (MSI) (1–4). LINE-1 constitutes a substantial portion (approximately 18%) of the human genome (5), and methylation level in LINE-1 correlates with global DNA methylation status (6). LINE-1 hypomethylated CRC has been associated with advanced stage or poor prognosis (7–12) and inversely with MSI-high independent of the CpG island methylator phenotype (CIMP) (13,14).
High degree of MSI due to deficiency of DNA mismatch repair is observed in approximately 15% of CRCs and is a well established prognostic biomarker for better survival (15–21). However, a subset of patients of MSI-high CRCs still succumbs to the disease, and hence, additional biomarkers are needed to further classify MSI-high CRCs into different prognostic groups. Recent evidence suggests that LINE-1 hypomethylation is associated with higher mortality in MSI-high CRCs (9); however, it remains uncertain whether the prognostic association of LINE-1 hypomethylation differs by MSI status. We hypothesized that the adverse prognostic association of LINE-1 hypomethylation might be stronger in MSI-high CRCs (characterized by numerous somatic mutations [22]) than in microsatellite stable (MSS) CRCs.
To test this hypothesis, we utilized a molecular pathological epidemiology database of 1211 CRCs in two prospective cohort studies and examined patient survival according to tumor LINE-1 methylation level (scale 0–100%) in strata of MSI status. Considering the potential influence of Lynch syndrome (23,24), we additionally examined the prognostic role of tumor LINE-1 methylation level within MSI-high CRC cases in strata of CRC family history status.
Methods
Study Cohorts
We utilized two US nationwide, prospective cohort studies: the Nurses’ Health Study (NHS, involving 121700 women who were enrolled in 1976) and the Health Professionals Follow-up Study (HPFS, involving 51529 men who were enrolled in 1986) (25). Every two years since these studies began, the participants have been sent follow-up questionnaires to update information on potential disease risk factors, and to identify newly diagnosed cancers and other diseases in themselves and their first-degree relatives. Lethal CRC cases were identified and confirmed by searching through the National Death Index. Study physicians reviewed all medical records related to CRC, extracted clinical information including stage and tumor location, and determined cause of death in deceased individuals. We collected paraffin-embedded tissue blocks from hospitals where participants with CRC had undergone tumor resection or diagnostic biopsy specimens.
Hematoxylin and eosin-stained tissue sections from all CRC cases were reviewed by a pathologist (S. Ogino) unaware of other data. Tumor differentiation was categorized as well to moderate vs poor (>50% vs ≤50% glandular area). We used available data on tumor LINE-1 methylation level, MSI status, and survival from 1211 CRC patients diagnosed up to 2008. Given the colorectal continuum model (26,27), we included both colon and rectal carcinomas in our primary analysis; we also examined colon cancer (excluding rectal cancer) in our secondary analysis. Patients were observed until death or January 1, 2012, whichever came first. The procedures and protocols of this study were approved by the institutional review boards for the Harvard School of Public Health and the Brigham and Women’s Hospital. All subjects provided informed consent.
LINE-1 Methylation Analysis
DNA was extracted from archival tumor tissue. We performed bisulfite DNA treatment, polymerase chain reaction (PCR), and a pyrosequencing assay to quantify LINE-1 methylation level, after assay validation (28). We primarily used LINE-1 methylation level as a continuous variable (scale 0–100%) in survival analyses. To display some of our results, we categorized the degree of LINE-1 hypomethylation status into three groups, namely “severe” (<55% methylation), “intermediate” (55–64.9% methylation), and “mild/no” (≥65% methylation), consistent with our previous studies (29,30).
MSI Analysis
MSI analysis was performed utilizing a panel of 10 microsatellite markers, as previously described (20). MSI-high was defined as instability in greater than or equal to 30% of the markers, and MSS status as instability in less than 30% of the markers (20).
Analysis of CIMP, KRAS, BRAF, and PIK3CA
Assessment of CIMP, KRAS, BRAF, and PIK3CA of CRCs in our cohorts (25,31,32) is described in the Supplementary Methods (available online).
Statistical Analysis
All statistical analyses were carried out using SAS (version 9.3, SAS Institute, Cary, NC). All P values were two-sided. Statistical significance level was set at P = .05 for testing of our primary hypothesis of an interaction between LINE-1 methylation level (continuous) and MSI status (binary) in CRC-specific survival analysis. A statistical interaction was assessed by likelihood ratio test, which compared the model with the interaction term (of LINE-1 and MSI statuses) to the model without the interaction term. For secondary and exploratory analyses, we recognized multiple comparisons inherent in subgroup analyses and interpreted our data very cautiously to avoid overinterpretation. For demographic categorical data, the chi-square test was performed. A t-test or analysis of variance (ANOVA), assuming equal variances, was used to compare mean age.
Kaplan–Meier method and log-rank test were used for survival analyses. For analyses of CRC-specific mortality, deaths as a result of other causes were censored. To control for confounding, we used multivariable Cox proportional hazards regression models. In addition to the LINE-1 hypomethylation variable (continuous; 10% decrease as a unit), the multivariable model initially included sex, age at diagnosis (continuous), year of diagnosis (continuous), family history of CRC in first-degree relative(s) (present vs absent), tumor location (proximal colon vs distal colon vs rectum), tumor differentiation (well to moderate vs poor), CIMP (high vs low/negative), and KRAS, BRAF, and PIK3CA mutations. A single analysis model could estimate the effect of LINE-1 hypomethylation in each stratum of MSI status, using a reparameterization of the interaction term (of LINE-1 and MSI statuses), as previously described (33). To avoid overfitting, disease stage (I, II, III, IV, or unknown) (34) was used as a stratifying variable using the “strata” option in the SAS “proc phreg” command. A backward stepwise elimination was carried out with P = .05 as a threshold, to select variables for the final model. For cases with missing information in any of the categorical covariates (family history of CRC [0.4%], tumor location [1.1%], tumor differentiation [0.5%], CIMP [5.0%], KRAS [0.3%], BRAF [0.6%], and PIK3CA [7.3%]), we included these cases in the majority category of a given covariate. We confirmed that excluding cases with missing information in any of the covariates did not substantially alter results (data not shown). The proportionality of hazards assumption was assessed by a time-varying covariate (an interaction term of survival time and LINE-1 hypomethylation variable, P > .15).
Results
LINE-1 Hypomethylation and MSI Status in CRC
In 1211 incident CRCs within the two prospective cohort studies, the Nurses’ Health Study and the Health Professionals Follow-up Study, we measured tumor LINE-1 methylation level (scale 0–100%), which ranged from 23.1% to 93.1% with mean of 62.7% and standard deviation of 9.4%. Normal colon mucosal tissue typically showed LINE-1 methylation level of 70% to 75% in our assay (14,28). There were 190 MSI-high CRCs and 1021 MSS CRCs. Table 1 shows characteristics of the 1211 cases of CRC, stratified by LINE-1 hypomethylation status and MSI status. Severe degree of LINE-1 hypomethylation was statistically significantly associated with higher disease stage (P = .001) and inversely associated with CIMP-high (P < .001) in MSS cases.
Table 1.
Clinical, pathologic, and molecular characteristics of 1211 colorectal cancer (CRC) cases stratified by tumor LINE-1 hypomethylation status and microsatellite instability (MSI) status
| Clinical, pathologic, or molecular feature | MSS | MSI-high | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| No. of cases | LINE-1 hypomethylation status | No. of cases | LINE-1 hypomethylation status | P* | ||||||
| Mild/no (≥65% methylation) | Intermediate (55–64.9% methylation) |
Severe (<55% (methylation) |
P* | Mild/no (≥65% methylation) | Intermediate (55–64.9% methylation) |
Severe (<55% (methylation) |
||||
| No. of cases | 1,021 | 376 | 423 | 222 | 190 | 125 | 48 | 17 | ||
| Sex | .14 | .23 | ||||||||
| Male (HPFS) | 495 (48%) | 186 (49%) | 191 (45%) | 118 (53%) | 65 (34%) | 48 (38%) | 13 (27%) | 4 (24%) | ||
| Female (NHS) | 526 (52%) | 190 (51%) | 232 (55%) | 104 (47%) | 125 (66%) | 77 (62%) | 35 (73%) | 13 (76%) | ||
| Mean age ± SD, y | 68.3±8.9 | 68.9±8.9 | 68.2±8.7 | 67.3±9.3 | .11 | 70.4±7.5 | 71.2±7.5 | 70.3±6.5 | 65.2±8.5 | .008 |
| Year of diagnosis | <.001 | <.001 | ||||||||
| Prior to 1995 | 342 (34%) | 99 (26%) | 155 (37%) | 88 (40%) | 43 (23%) | 25 (20%) | 13 (27%) | 5 (29%) | ||
| 1995 to 1999 | 317 (31%) | 105 (28%) | 126 (30%) | 86 (39%) | 66 (35%) | 31 (25%) | 24 (50%) | 11 (65%) | ||
| 2000 to 2008 | 362 (35%) | 172 (46%) | 142 (34%) | 48 (22%) | 81 (43%) | 69 (55%) | 11 (23%) | 1 (5.9%) | ||
| Family history of CRC in first-degree relative(s) | .15 | .52 | ||||||||
| (−) | 788 (78%) | 303 (81%) | 320 (76%) | 165 (75%) | 138 (73%) | 94 (75%) | 32 (67%) | 12 (71%) | ||
| (+) | 228 (22%) | 72 (19%) | 100 (24%) | 56 (25%) | 52 (27%) | 31 (25%) | 16 (33%) | 5 (29%) | ||
| Tumor location | .09 | .45 | ||||||||
| Cecum | 161 (16%) | 56 (15%) | 69 (16%) | 36 (17%) | 44 (23%) | 32 (26%) | 9 (19%) | 3 (18%) | ||
| Ascending to transverse colon |
235 (23%) | 87 (24%) | 103 (25%) | 45 (21%) | 121 (64%) | 78 (62%) | 33 (69%) | 10 (59%) | ||
| Splenic flexure to sigmoid colon |
356 (35%) | 114 (31%) | 157 (37%) | 85 (39%) | 19 (10%) | 13 (10%) | 3 (6.3%) | 3 (18%) | ||
| Rectum | 256 (25%) | 113 (31%) | 91 (22%) | 52 (24%) | 6 (3.2%) | 2 (1.6%) | 3 (6.3%) | 1 (5.9%) | ||
| Tumor differentiation | .04 | .76 | ||||||||
| Well-moderate | 958 (94%) | 352 (94%) | 404 (96%) | 202 (91%) | 131 (69%) | 84 (67%) | 35 (73%) | 12 (71%) | ||
| Poor | 57 (5.6%) | 22 (5.9%) | 16 (3.8%) | 19 (8.6%) | 59 (31%) | 41 (33%) | 13 (27%) | 5 (29%) | ||
| Disease stage | .001 | .20 | ||||||||
| I | 246 (24%) | 102 (27%) | 106 (25%) | 38 (17%) | 39 (21%) | 22 (18%) | 12 (25%) | 5 (29%) | ||
| II | 240 (24%) | 81 (22%) | 109 (26%) | 50 (23%) | 103 (54%) | 73 (58%) | 24 (50%) | 6 (35%) | ||
| III | 290 (28%) | 96 (26%) | 123 (29%) | 71 (32%) | 30 (16%) | 18 (14%) | 10 (21%) | 2 (12%) | ||
| IV | 148 (15%) | 49 (13%) | 52 (12%) | 47 (21%) | 10 (5.3%) | 6 (4.8%) | 1 (2.1%) | 3 (18%) | ||
| Unknown | 97 (10%) | 48 (13%) | 33 (7.8%) | 16 (7.2%) | 8 (4.2%) | 6 (4.8%) | 1 (2.1%) | 1 (5.9%) | ||
| CIMP status | <.001 | .10 | ||||||||
| CIMP-low/negative | 904 (93%) | 320 (89%) | 376 (96%) | 208 (97%) | 45 (25%) | 30 (25%) | 8 (17%) | 7 (44%) | ||
| CIMP-high | 64 (6.6%) | 40 (11%) | 17 (4.3%) | 7 (3.3%) | 137 (75%) | 89 (75%) | 39 (83%) | 9 (56%) | ||
|
MLH1 promoter hypermethylation |
.72 | .01 | ||||||||
| (−) | 949 (98%) | 352 (98%) | 387 (98%) | 210 (98%) | 40 (22%) | 25 (21%) | 7 (15%) | 8 (50%) | ||
| (+) | 19 (2.0%) | 8 (2.2%) | 6 (1.5%) | 5 (2.3%) | 142 (78%) | 94 (79%) | 40 (85%) | 8 (50%) | ||
| KRAS mutation | .80 | .11 | ||||||||
| (−) | 606 (59%) | 225 (60%) | 246 (58%) | 135 (61%) | 164 (87%) | 104 (85%) | 46 (96%) | 14 (82%) | ||
| (+) | 413 (41%) | 150 (40%) | 176 (42%) | 87 (39%) | 24 (13%) | 19 (15%) | 2 (4.2%) | 3 (18%) | ||
| BRAF mutation | .08 | .15 | ||||||||
| (−) | 936 (92%) | 337 (90%) | 396 (94%) | 203 (93%) | 91 (48%) | 57 (46%) | 22 (46%) | 12 (71%) | ||
| (+) | 79 (7.8%) | 38 (10%) | 25 (5.9%) | 16 (7.3%) | 98 (52%) | 67 (54%) | 26 (54%) | 5 (29%) | ||
| PIK3CA mutation | .36 | .52 | ||||||||
| (−) | 793 (84%) | 292 (82%) | 323 (83%) | 178 (87%) | 144 (83%) | 96 (83%) | 35 (80%) | 13 (93%) | ||
| (+) | 156 (16%) | 62 (18%) | 67 (17%) | 27 (13%) | 30 (17%) | 20 (17%) | 9 (20%) | 1 (7.1%) | ||
* The P value for statistical significance was adjusted for multiple hypothesis testing to P = .05/24 = .002. Thus, a P value between .05 and .002 should be regarded as of borderline statistical significance. CIMP = CpG island methylator phenotype; CRC = colorectal cancer; HPFS = Health Professionals Follow-up Study; LINE-1 = long interspersed nucleotide element-1; MSI = microsatellite instability; MSS = microsatellite stable; NHS = Nurses’ Health Study; SD = standard deviation.
(%) indicates the proportion of cases with a specific clinical, pathologic, or molecular feature among cancers with each LINE-1 methylation level in MSS- or MSI-high cases.
LINE-1 Hypomethylation, MSI Status, and CRC Mortality
Among 1211 patients, there were 648 deaths, including 356 CRC-specific deaths, during a median follow-up of 151 months (interquartile range: 110 to 204 months) for censored cases. We examined the relationship between LINE-1 hypomethylation and patient survival in all cases, and in strata of MSI status. In all CRC cases, tumor LINE-1 hypomethylation was associated with higher CRC-specific mortality in Kaplan–Meier analysis (log-rank P < .001) (Figure 1) and in univariate and multivariable Cox regression analyses (for 10% decrease in LINE-1 methylation: multivariable hazard ratio [HR] = 1.16, 95% confidence interval (CI) = 1.03 to 1.30, P = .02) (Table 2). For our main hypothesis testing, we examined statistical interaction between LINE-1 methylation level (continuous) and MSI status in CRC-specific survival analysis (Table 2), which revealed a statistically significant interaction (P interaction < .001). In MSI-high CRCs, the association of LINE-1 hypomethylation with higher CRC-specific mortality was statistically significant (for 10% decrease in LINE-1 methylation: multivariable HR = 2.45, 95% CI = 1.64 to 3.66, P < .001) (Table 2). In MSS CRCs, the association of LINE-1 hypomethylation with CRC-specific mortality was weaker and not statistically significant (for 10% decrease in LINE-1 methylation: multivariable HR = 1.10, 95% CI = 0.98 to 1.24, P = .11) (Table 2). Figure 1 shows Kaplan–Meier survival curves according to LINE-1 hypomethylation categories in MSI-high CRCs and in MSS CRCs.
Figure 1.
Kaplan–Meier curves for colorectal cancer (CRC) patients according to tumor LINE-1 methylation level. CRC-specific survival in all CRC cases (A), microsatellite stable (MSS) CRC cases (B), and microsatellite instability (MSI)-high CRC cases (C). P value was calculated using log-rank test (two-sided). The tables (bottom) show the number of patients who remained alive and at risk of death at each time point after the diagnosis of CRC.
Table 2.
Tumor LINE-1 hypomethylation and colorectal cancer (CRC) mortality, stratified by microsatellite instability (MSI) status
| CRC subtype | No. of cases | CRC-specific mortality | Overall mortality | ||||||
|---|---|---|---|---|---|---|---|---|---|
| No. of events | Univariate HR (95% CI) |
Stage- stratified HR (95% CI) | Multivariable HR* (95% CI) | No. of events | Univariate HR (95% CI) |
Stage- stratified HR (95% CI) | Multivariable HR* (95% CI) | ||
| All CRC cases | |||||||||
| LINE-1 hypomethylation (10% decrease as a unit) |
1211 | 356 | 1.24 (1.11 to 1.38) |
1.19 (1.06 to 1.33) |
1.16 (1.03 to 1.30) |
648 | 1.01 (1.01 to 1.19) |
1.07 (0.98 to 1.16) |
1.06 (0.97 to 1.16) |
| P† | <.001 | .003 | .02 | .04 | .12 | .18 | |||
| MSS cases | |||||||||
| LINE-1 hypomethylation (10% decrease as a unit) |
1021 | 333 | 1.13 (1.00 to 1.26) |
1.10 (0.98 to 1.24) |
1.10 (0.98 to 1.24) |
562 | 1.04 (0.95 to 1.14) |
1.03 (0.95 to 1.13) |
1.02 (0.93 to 1.12) |
| P† | .05 | .11 | .11 | .36 | .47 | .64 | |||
| MSI-high cases | |||||||||
| LINE-1 hypomethylation (10% decrease as a unit) |
190 | 23 | 1.90 (1.30 to 2.77) |
2.23 (1.48 to 3.36) |
2.45 (1.64 to 3.66) |
86 | 1.33 (1.05 to 1.68) |
1.38 (1.08 to 1.77) |
1.42 (1.12 to 1.81) |
| P† | <.001 | <.001 | <.001 | .02 | .01 | .004 | |||
| P interaction‡ | .02 | .002 | <.001 | .06 | .03 | .02 | |||
* The multivariable, stage-stratified Cox regression model initially included LINE-1 hypomethylation variable (continuous), sex, age at diagnosis, year of diagnosis, family history of CRC, tumor location, tumor differentiation, MSI status (only for all CRC cases), CpG island methylator phenotype, KRAS, BRAF, and PIK3CA mutations. CI = confidence interval; CRC = colorectal cancer; HR = hazard ratio; LINE-1 = long interspersed nucleotide element-1; MSI = microsatellite instability; MSS = microsatellite stable.
† P value was calculated using chi-square test (two-sided).
‡ P interaction value (between continuous LINE-1 methylation level and MSI status) was calculated using likelihood ratio test (two-sided), which compared the model with the interaction term to the model without the interaction term.
In secondary analyses of overall survival as an endpoint, there was a general trend toward differential prognostic associations of LINE-1 hypomethylation by MSI status, although the differences were not as evident as those in CRC-specific mortality analyses (Table 2; Supplementary Figure 1, available online). Results of Kaplan–Meier analyses according to combined LINE-1 hypomethylation and MSI status are provided in Supplementary Figure 2 (available online).
Table 3 shows secondary analyses limited to colon cancer. The association of LINE-1 hypomethylation with higher colon cancer-specific mortality was stronger in MSI-high colon cancer than in MSS colon cancer (P interaction < .001, between LINE-1 and MSI statuses).
Table 3.
Tumor LINE-1 hypomethylation and colon cancer mortality, stratified by microsatellite instability (MSI) status
| Colon cancer subtype | No. of cases | Colon cancer-specific mortality | Overall mortality | ||||||
|---|---|---|---|---|---|---|---|---|---|
| No. of events | Univariate HR (95% CI) |
Stage- stratified HR (95% CI) | Multivariable HR* (95% CI) | No. of events | Univariate HR (95% CI) |
Stage- stratified HR (95% CI) | Multivariable HR* (95% CI) | ||
| All colon cancer cases | |||||||||
| LINE-1 hypomethylation (10% decrease as a unit) |
936 | 263 | 1.35 (1.20 to 1.53) |
1.28 (1.12 to 1.45) |
1.25 (1.09 to 1.44) |
495 | 1.16 (1.06 to 1.28) |
1.13 (1.03 to 1.24) |
1.14 (1.03 to 1.26) |
| P† | <.001 | <.001 | .001 | .002 | .01 | .01 | |||
| MSS colon cancer cases | |||||||||
| LINE-1 hypomethylation (10% decrease as a unit) |
752 | 241 | 1.21 (1.05 to 1.38) |
1.16 (1.01 to 1.34) |
1.18 (1.03 to 1.36) |
413 | 1.10 (0.99 to 1.23) |
1.03 (0.95 to 1.13) |
1.02 (0.93 to 1.12) |
| P† | .007 | .03 | .02 | .06 | .47 | .64 | |||
| MSI-high colon cancer cases | |||||||||
| LINE-1 hypomethylation (10% decrease as a unit) |
184 | 22 | 2.00 (1.36 to 2.93) |
2.37 (1.55 to 3.62) |
2.69 (1.77 to 4.11) |
82 | 1.39 (1.09 to 1.78) |
1.38 (1.08 to 1.77) |
1.42 (1.12 to 1.81) |
| P† | <.001 | <.001 | <.001 | .008 | .01 | .004 | |||
| P interaction‡ | .02 | .003 | <.001 | .10 | .07 | .05 | |||
* The multivariable, stage-stratified Cox regression model initially included LINE-1 hypomethylation variable (continuous), sex, age at diagnosis, year of diagnosis, family history of colorectal cancer, tumor differentiation, MSI status (only for all colon cancer cases), CpG island methylator phenotype, KRAS, BRAF, and PIK3CA mutations. CI = confidence interval; HR = hazard ratio; LINE-1 = long interspersed nucleotide element-1; MSI = microsatellite instability; MSS = microsatellite stable.
† P value was calculated using chi-square test (two-sided).
‡ P interaction value (between continuous LINE-1 methylation level and MSI status) was calculated using likelihood ratio test (two-sided), which compared the model with the interaction term with the model without the interaction term.
LINE-1 Hypomethylation, MSI/BRAF Status, and CRC Mortality
Although the utility of MSI/BRAF classification for prognostication in CRC has been demonstrated (20), it is an imperfect marker; ie, some patients with favorable MSI-high/BRAF–wild-type tumors may die of cancer, while other patients with unfavorable MSS/BRAF-mutant tumors may survive. Hence, additional markers are needed to refine prognostic groups of CRC.
As a secondary analysis, we examined the relationship between LINE-1 hypomethylation and patient survival in strata of MSI/BRAF subtype (Table 4). The association of LINE-1 hypomethylation with higher CRC-specific mortality appeared to be stronger in the MSI-high/BRAF–wild-type subtype (for 10% decrease in LINE-1 methylation; multivariable HR = 2.57, 95% CI = 1.47 to 4.51, P = .001) than in the other three subtypes, namely MSS/BRAF–wild-type, MSS/BRAF-mutant, and MSI-high/BRAF-mutant (for 10% decrease in LINE-1 methylation; multivariable HR = 1.06, 95% CI = 0.93 to 1.21, P = .36; multivariable HR = 1.29, 95% CI = 0.94 to 1.77, P = .12; and multivariable HR = 1.28, 95% CI = 0.82 to 2.01, P = .28, respectively).
Table 4.
Tumor LINE-1 hypomethylation and colorectal cancer (CRC) mortality, stratified by microsatellite instability (MSI) status and mutation status of BRAF
| CRC subtype | No. of cases | CRC-specific mortality | Overall mortality | ||||||
|---|---|---|---|---|---|---|---|---|---|
| No. of events | Univariate HR (95% CI) |
Stage- stratified HR (95% CI) | Multivariable HR* (95% CI) | No. of events | Univariate HR (95% CI) |
Stage- stratified HR (95% CI) | Multivariable HR* (95% CI) | ||
| MSS/BRAF–wild-type | |||||||||
| LINE-1 hypomethylation (10% decrease as a unit) |
936 | 291 | 1.12 (0.99 to 1.26) |
1.08 (0.95 to 1.22) |
1.06 (0.93 to 1.21) |
506 | 1.02 (0.92 to 1.12) |
1.00 (0.91 to 1.10) |
1.00 (0.90 to 1.11) |
| P† | .08 | .26 | .36 | .74 | .99 | .99 | |||
| MSS/BRAF-mutant | |||||||||
| LINE-1 hypomethylation (10% decrease as a unit) |
79 | 39 | 1.45 (1.05 to 2.01) |
1.42 (1.04 to 1.94) |
1.29 (0.94 to 1.77) |
51 | 1.52 (1.13 to 2.05) |
1.47 (1.10 to 1.96) |
1.26 (0.95 to 1.69) |
| P† | .03 | .03 | .12 | .006 | .009 | .11 | |||
| MSI-high/BRAF–wild-type | |||||||||
| LINE-1 hypomethylation (10% decrease as a unit) |
91 | 9 | 2.14 (1.26 to 3.64) |
2.60 (1.44 to 4.70) |
2.57 (1.47 to 4.51) |
41 | 1.28 (0.92 to 1.79) |
1.30 (0.90 to 1.88) |
1.29 (0.90 to 1.84) |
| P† | .005 | .002 | .001 | .14 | .16 | .16 | |||
| MSI-high/BRAF-mutant | |||||||||
| LINE-1 hypomethylation (10% decrease as a unit) |
98 | 14 | 1.03 (0.66 to 1.61) |
1.25 (0.79 to 1.97) |
1.28 (0.82 to 2.01) |
45 | 1.10 (0.84 to 1.44) |
1.23 (0.93 to 1.61) |
1.30 (0.98 to 1.73) |
| P† | .90 | .35 | .28 | .48 | .14 | .07 | |||
* The multivariable, stage-stratified Cox regression model initially included LINE-1 hypomethylation variable (continuous), sex, age at diagnosis, year of diagnosis, family history of CRC, tumor location, tumor differentiation, CpG island methylator phenotype, KRAS and PIK3CA mutations. CI = confidence interval; CRC = colorectal cancer; HR = hazard ratio; LINE-1 = long interspersed nucleotide element-1; MSI = microsatellite instability; MSS = microsatellite stable.
† P value was calculated using chi-square test (two-sided).
LINE-1 Hypomethylation, CRC Family History, and Mortality in MSI-High CRC
The MSI-high/BRAF–wild-type subgroup of CRCs encompasses Lynch syndrome cases, which are familial cancers due to a germline mutation in one of mismatch repair genes. As an exploratory analysis, we focused on MSI-high cases and examined the relationship between tumor LINE-1 hypomethylation and patient survival in strata of CRC family history status (Table 5). The association of LINE-1 hypomethylation with higher CRC-specific mortality appeared to be stronger in MSI-high cases with a family history of CRC in a first-degree relative (multivariable HR = 5.13, 95% CI = 1.99 to 13.2, P < .001) than in MSI-high cases without a family history of CRC in a first-degree relative (multivariable HR = 1.62, 95% CI = 0.89 to 2.94, P = .11) (P interaction = .02, between LINE-1 methylation and CRC family history status in MSI-high cases) (Table 5). Nonetheless, these results must be interpreted cautiously, given the exploratory nature of this subgroup analysis and the low event numbers.
Table 5.
Microsatellite instability (MSI)-high colorectal cancer (CRC) mortality according to tumor LINE-1 hypomethylation status, stratified by family history of CRC in first-degree relative(s)
| MSI-high CRC subtype | No. of cases | CRC-specific mortality | Overall mortality | ||||||
|---|---|---|---|---|---|---|---|---|---|
| No. of events | Univariate HR (95% CI) |
Stage- stratified HR (95% CI) | Multivariable HR* (95% CI) |
No. of events | Univariate HR (95% CI) |
Stage- stratified HR (95% CI) | Multivariable HR* (95% CI) |
||
| Family history of CRC (−) | |||||||||
| LINE-1 hypomethylation (10% decrease as a unit) |
138 | 16 | 1.36 (0.82 to 2.25) |
1.41 (0.81 to 2.47) |
1.62 (0.89 to 2.94) |
61 | 1.17 (0.88 to 1.55) |
1.18 (0.88 to 1.59) |
1.21 (0.90 to 1.63) |
| P† | .24 | .23 | .11 | .29 | .27 | .21 | |||
| Family history of CRC (+) | |||||||||
| LINE-1 hypomethylation (10% decrease as a unit) |
52 | 7 | 6.39 (2.45 to 16.7) |
4.89 (1.93 to 12.4) |
5.13 (1.99 to 13.2) |
25 | 1.95 (1.12 to 3.40) |
1.89 (1.10 to 3.25) |
2.21 (1.27 to 3.86) |
| P† | <.001 | <.001 | <.001 | .02 | .02 | .005 | |||
| P interaction‡ | .003 | .02 | .02 | .10 | .13 | .05 | |||
* The multivariable, stage-stratified Cox regression model initially included LINE-1 hypomethylation variable (continuous), sex, age at diagnosis, year of diagnosis, tumor location, tumor differentiation, CpG island methylator phenotype, KRAS, BRAF, and PIK3CA mutations. CI = confidence interval; CRC = colorectal cancer; HR = hazard ratio; LINE-1 = long interspersed nucleotide element-1; MSI = microsatellite instability.
† P value was calculated using chi-square test (two-sided).
‡ P interaction value (between continuous LINE-1 methylation level and binary status of CRC family history) was calculated using likelihood ratio test (two-sided), which compared the model with the interaction term with the model without the interaction term.
LINE-1 Hypomethylation, MSI Status, Tumor Location, and CRC Mortality
Considering the interactive prognostic association between MSI status and tumor location reported by Sinicrope et al. (35), we examined the relationship between LINE-1 hypomethylation and patient survival in strata of MSI status and tumor location (Supplementary Table 1, available online). The association of LINE-1 hypomethylation with mortality appeared to be modified by MSI status but not by tumor location, although statistical power was limited in this subgroup analysis.
Discussion
We conducted this study to test the hypothesis that the prognostic association of LINE-1 hypomethylation in CRC might be stronger in MSI-high CRCs than in MSS CRCs. Our data were consistent with this hypothesis, and the prognostic association of tumor LINE-1 hypomethylation indeed statistically significantly differed by MSI status. Although MSI-high is a well-established prognostic biomarker for better survival, a subset of MSI-high CRCs is lethal. In addition, our study has confirmed that tumor LINE-1 hypomethylation is associated with adverse prognosis in CRCs (7,8) and in MSI-high CRCs (9). Notably, our main hypothesis was that the prognostic association of tumor LINE-1 hypomethylation might be stronger in MSI-high CRCs (which are characterized by numerous somatic mutations [22]) than in MSS CRCs. This unique hypothesis has never been tested in the previous studies. To test our main hypothesis, it was necessary to utilize a large number of CRCs with detailed molecular analyses, including statuses of both MSI and LINE-1 hypomethylation, as well as other molecular features such as KRAS, BRAF, and PIK3CA mutations to control for possible confounding.
Examining tumor molecular and host factors has become increasingly important in CRC (36–42). LINE-1 methylation level and MSI are both important molecular markers in CRC. LINE-1 methylation level is used as a surrogate marker for global DNA methylation (43), and global DNA hypomethylation, indicated by LINE-1 hypomethylation, is associated with genomic and chromosomal instability (14,44). LINE-1 hypomethylation in CRC may also be a marker for familial susceptibility to CRC (30,45). MSI is a well-established biomarker that is routinely used for assessment of familial CRC (Lynch syndrome) risk in combination with BRAF testing (23). LINE-1 hypomethylation is inversely associated with MSI-high, independent of CIMP (13,14), and LINE-1 hypomethylation (7–9), and MSI-high (15–21) are associated with inferior and superior prognosis, respectively.
To the best of our knowledge, this is among the first studies to address the interactive association between LINE-1 hypomethylation and MSI status in relation to the clinical outcome of CRC patients. Although some studies have shown the association of LINE-1 hypomethylation in CRCs with inferior survival, sample sizes of those studies would unlikely be adequate to analyze this interactive effect (sample size N = 643 [7] and N = 161 [8]). Rhee et al. (9) showed that LINE-1 hypomethylation was associated with adverse prognosis in MSI-high CRCs, but they did not examine MSS CRCs. A recent study has shown that LINE-1 hypomethylation is an independent prognostic biomarker in early-stage rectal cancer, but they did not examine colon cancer (46). Our resource of a large number of CRCs (N = 1211) has provided us with reasonable power to analyze this interactive association. Furthermore, we took into account other tumor molecular data, including CIMP, and KRAS, BRAF, and PIK3CA mutations.
Considering the utility of MSI/BRAF classification for prognostication in CRC (20), as well as its routine clinical use for familial risk assessment (23), we additionally examined the prognostic association of LINE-1 hypomethylation in strata of MSI/BRAF subtype. The association of LINE-1 hypomethylation with inferior survival appeared to be stronger in the MSI-high/BRAF–wild-type subtype, which is known to imply the most favorable subtype among the four MSI/BRAF subtypes (20). Although we should interpret the results cautiously, LINE-1 methylation level can potentially be used as an additional biomarker to refine the prognostic groups by MSI/BRAF classification.
Because the MSI-high/BRAF–wild-type CRC subgroup contains Lynch syndrome cases, we focused on MSI-high cases and examined the prognostic association of LINE-1 hypomethylation in strata of CRC family history status. The association of LINE-1 hypomethylation with inferior survival appeared to be stronger in MSI-high cases with a CRC family history than in MSI-high cases without a CRC family history. Nevertheless, we must interpret the results carefully to avoid overinterpretation, considering the exploratory nature of this analysis. Since the MSI-high/BRAF–wild-type subtype and the MSI-high subtype with a CRC family history are enriched with Lynch syndrome cases (23), these intriguing results warrant further investigation to examine whether LINE-1 hypomethylation serves as an unfavorable prognostic biomarker in Lynch syndrome cases.
It is interesting, but challenging, to speculate potential mechanisms of interaction between LINE-1 methylation level and MSI status. Compared to MSS tumor, MSI-high CRC characterized by numerous somatic mutations (22) might be more influenced by genomic DNA or LINE-1 hypomethylation, which is associated with chromosomal instability (14,44). Other possible mechanisms may involve inflammatory mediators (47–50), variation in locus-specific methylation patterns (43,51–54), and non-coding RNAs (55–57). Further studies are required to elucidate the underlying mechanisms of the interactive prognostic association between tumor LINE-1 methylation level and MSI status.
Some limitations of our study deserve discussion. Because data on cancer treatment were limited, unknown bias, including differential treatment assignment, might confound results. In survival analyses, we adjusted for disease stage, on which treatment decisions are mainly based. Another caveat relates to the study population. Health professionals may not be completely representative of the general US population. Nonetheless, the pathologic and molecular features of our CRC cases are generally compatible with data from the US cancer registry and published literature. Another limitation is that we excluded cases without available tumor data, which might cause bias. Nonetheless, a previous study has shown that there are no statistically significant demographic or clinical differences between cases with and without available tumor data (58). Finally, we need to replicate the findings before implementing tumor LINE-1 methylation measurement as a clinical test following the guidelines (59).
Strengths of this study include the use of data from the two US nationwide, prospective cohort studies. Information on cancer staging, family history of CRC, and other clinicopathologic and tumor molecular data was integrated into the molecular pathological epidemiology (60,61) database. Our cohort participants were treated at hospitals throughout the US and were more representative of CRC cases in the general US population than patients in only a few academic hospitals. Finally, by virtue of our database, we could assess the prognostic association of LINE-1 hypomethylation in strata of MSI status, while controlling for multiple potential confounders, including disease stage, CIMP, and KRAS, BRAF, and PIK3CA mutations.
In conclusion, we showed a stronger association of LINE-1 hypomethylation with inferior survival in MSI-high CRCs than in MSS CRCs, further attesting biological heterogeneity of MSI-high CRCs. LINE-1 methylation may be a useful prognostic biomarker to identify aggressive cancer cases among generally indolent MSI-high CRCs.
Funding
This work was supported by US National Institutes of Health (NIH) grants (P01 CA87969 to S.E. Hankinson, P01 CA55075 to W.C. Willett, UM1 CA167552 to W.C. Willett, P50 CA127003 to CSF, R01 CA136950 to EC, R01 CA137178 and K24 DK098311 to ATC, and R01 CA151993 to SO) and by grants from the Bennett Family Fund and the Entertainment Industry Foundation through the National Colorectal Cancer Research Alliance. KI is supported by the Japan Society for the Promotion of Science Postdoctoral Fellowship for Research Abroad and by the Takashi Tsuruo Memorial Fund. PL is a Scottish Government Clinical Academic Fellow and was supported by a Harvard University Frank Knox Memorial Fellowship. SAK is supported by an early exchange postdoctoral fellowship grant from Asan Medical Center. KM is supported by a fellowship grant from the Uehara Memorial Foundation. ATC is a Damon Runyon Clinical Investigator.
A. T. Chan previously served as a consultant for Bayer Healthcare, Millennium Pharmaceuticals, Pozen Inc, and Pfizer Inc. This study was not funded by Bayer Healthcare, Millennium Pharmaceuticals, Pozen Inc., or Pfizer Inc. No other conflict of interest exists.
The content is solely the responsibility of the authors and does not necessarily represent the official views of NIH. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
We deeply thank hospitals and pathology departments throughout the U.S. for generously providing us with tissue specimens. In addition, we would like to thank the participants and staff of the Nurses’ Health Study and the Health Professionals Follow-Up Study for their valuable contributions, as well as the following state cancer registries for their help: AL, AZ, AR, CA, CO, CT, DE, FL, GA, ID, IL, IN, IA, KY, LA, ME, MD, MA, MI, NE, NH, NJ, NY, NC, ND, OH, OK, OR, PA, RI, SC, TN, TX, VA, WA, WY.
References
- 1. Lao VV, Grady WM. Epigenetics and colorectal cancer. Nat Rev Gastroenterol Hepatol. 2011;8(12):686–700 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Colussi D, Brandi G, Bazzoli F, et al. Molecular Pathways Involved in Colorectal Cancer: Implications for Disease Behavior and Prevention. International Journal of Molecular Sciences. 2013;14(8):16365–16385 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Bardhan K, Liu K. Epigenetics and Colorectal Cancer Pathogenesis. Cancers. 2013;5(2):676–713 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Goel A, Boland CR. Epigenetics of colorectal cancer. Gastroenterology. 2012;143(6):1442–1460 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Rodic N, Burns KH. Long interspersed element-1 (LINE-1): passenger or driver in human neoplasms? PLoS Genet. 2013;9(3):e1003402. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Yang AS, Estecio MR, Doshi K, et al. A simple method for estimating global DNA methylation using bisulfite PCR of repetitive DNA elements. Nucleic Acids Res. 2004;32(3):e38. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Ogino S, Nosho K, Kirkner GJ, et al. A cohort study of tumoral LINE-1 hypomethylation and prognosis in colon cancer. J Natl Cancer Inst. 2008;100(23):1734–1738 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Ahn JB, Chung WB, Maeda O, et al. DNA methylation predicts recurrence from resected stage III proximal colon cancer. Cancer. 2011;117(9):1847–1854 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Rhee YY, Kim MJ, Bae JM, et al. Clinical outcomes of patients with microsatellite-unstable colorectal carcinomas depend on L1 methylation level. Ann Surg Oncol. 2012;19(11):3441–3448 [DOI] [PubMed] [Google Scholar]
- 10. Antelo M, Balaguer F, Shia J, et al. A high degree of LINE-1 hypomethylation is a unique feature of early-onset colorectal cancer. PLoS One. 2012;7(9):e45357. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Sunami E, de Maat M, Vu A, et al. LINE-1 hypomethylation during primary colon cancer progression. PLoS One. 2011;6(4):e18884. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Matsunoki A, Kawakami K, Kotake M, et al. LINE-1 methylation shows little intra-patient heterogeneity in primary and synchronous metastatic colorectal cancer. BMC Cancer. 2012;12:574. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Estecio MR, Gharibyan V, Shen L, et al. LINE-1 hypomethylation in cancer is highly variable and inversely correlated with microsatellite instability. PLoS One. 2007;2(5):e399. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Ogino S, Kawasaki T, Nosho K, et al. LINE-1 hypomethylation is inversely associated with microsatellite instability and CpG island methylator phenotype in colorectal cancer. Int J Cancer. 2008;122(12):2767–2773 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Vilar E, Tabernero J. Molecular dissection of microsatellite instable colorectal cancer. Cancer Discov. 2013;3(5):502–511 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Gavin PG, Colangelo LH, Fumagalli D, et al. Mutation profiling and microsatellite instability in stage II and III colon cancer: an assessment of their prognostic and oxaliplatin predictive value. Clin Cancer Res. 2012;18(23):6531–6541 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Guastadisegni C, Colafranceschi M, Ottini L, et al. Microsatellite instability as a marker of prognosis and response to therapy: a meta-analysis of colorectal cancer survival data. Eur J Cancer. 2010;46(15):2788–2798 [DOI] [PubMed] [Google Scholar]
- 18. Sinicrope FA, Foster NR, Yoon HH, et al. Association of obesity with DNA mismatch repair status and clinical outcome in patients with stage II or III colon carcinoma participating in NCCTG and NSABP adjuvant chemotherapy trials. J Clin Oncol. 2012;30(4):406–412 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Popovici V, Budinska E, Tejpar S, et al. Identification of a poor-prognosis BRAF-mutant-like population of patients with colon cancer. J Clin Oncol. 2012;30(12):1288–1295 [DOI] [PubMed] [Google Scholar]
- 20. Lochhead P, Kuchiba A, Imamura Y, et al. Microsatellite Instability and BRAF Mutation Testing in Colorectal Cancer Prognostication. J Natl Cancer Inst. 2013;105(15):1151–1156 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Wangefjord S, Sundstrom M, Zendehrokh N, et al. Sex differences in the prognostic significance of KRAS codons 12 and 13, and BRAF mutations in colorectal cancer: a cohort study. Biol Sex Differ. 2013;4(1):17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. The Cancer Genome Atlas Network. Comprehensive molecular characterization of human colon and rectal cancer. Nature. 2012;487(7407):330–337 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Funkhouser WK, Jr., Lubin IM, Monzon FA, et al. Relevance, pathogenesis, and testing algorithm for mismatch repair-defective colorectal carcinomas: a report of the association for molecular pathology. J Mol Diagn. 2012;14(2):91–103 [DOI] [PubMed] [Google Scholar]
- 24. Zoratto F, Rossi L, Verrico M, et al. Focus on genetic and epigenetic events of colorectal cancer pathogenesis: implications for molecular diagnosis. Tumor Biology. 10.1007/s13277-014-1845-9 [DOI] [PubMed] [Google Scholar]
- 25. Liao X, Lochhead P, Nishihara R, et al. Aspirin use, tumor PIK3CA mutation, and colorectal-cancer survival. N Engl J Med. 2012;367(17):1596–1606 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Yamauchi M, Morikawa T, Kuchiba A, et al. Assessment of colorectal cancer molecular features along bowel subsites challenges the conception of distinct dichotomy of proximal versus distal colorectum. Gut. 2012;61(6):847–854 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Yamauchi M, Lochhead P, Morikawa T, et al. Colorectal cancer: a tale of two sides or a continuum? Gut. 2012;61(6):794–797 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Irahara N, Nosho K, Baba Y, et al. Precision of pyrosequencing assay to measure LINE-1 methylation in colon cancer, normal colonic mucosa, and peripheral blood cells. J Mol Diagn. 2010;12(2):177–183 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Schernhammer ES, Giovannucci E, Kawasaki T, et al. Dietary folate, alcohol and B vitamins in relation to LINE-1 hypomethylation in colon cancer. Gut. 2010;59(6):794–799 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Ogino S, Nishihara R, Lochhead P, et al. Prospective study of family history and colorectal cancer risk by tumor LINE-1 methylation level. J Natl Cancer Inst. 2013;105(2):130–140 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Ogino S, Kawasaki T, Brahmandam M, et al. Sensitive sequencing method for KRAS mutation detection by Pyrosequencing. J Mol Diagn. 2005;7(3):413–421 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Ogino S, Nosho K, Kirkner GJ, et al. CpG island methylator phenotype, microsatellite instability, BRAF mutation and clinical outcome in colon cancer. Gut. 2009;58(1):90–96 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Nosho K, Irahara N, Shima K, et al. Comprehensive biostatistical analysis of CpG island methylator phenotype in colorectal cancer using a large population-based sample. PLoS One. 2008;3(11):e3698. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Edge SB, Byrd DR, Compton CC, et al. AJCC cancer staging manual. 7th ed. New York (NY): Springer; 2010 [Google Scholar]
- 35. Sinicrope FA, Mahoney MR, Smyrk TC, et al. Prognostic Impact of Deficient DNA Mismatch Repair in Patients With Stage III Colon Cancer From a Randomized Trial of FOLFOX-Based Adjuvant Chemotherapy. J Clin Oncol. 2013;31(29):3664–3672 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Reimers MS, Zeestraten ECM, Kuppen PJK, et al. Biomarkers in precision therapy in colorectal cancer. Gastroenterology Report. 2013;1(3):166–183 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Sinicrope FA, Foster NR, Thibodeau SN, et al. DNA mismatch repair status and colon cancer recurrence and survival in clinical trials of 5-fluorouracil-based adjuvant therapy. J Natl Cancer Inst. 2011;103(11):863–875 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Roth AD, Delorenzi M, Tejpar S, et al. Integrated analysis of molecular and clinical prognostic factors in stage II/III colon cancer. J Natl Cancer Inst. 2012;104(21):1635–1646 [DOI] [PubMed] [Google Scholar]
- 39. Behl AS, Goddard KA, Flottemesch TJ, et al. Cost-effectiveness analysis of screening for KRAS and BRAF mutations in metastatic colorectal cancer. J Natl Cancer Inst. 2012;104(23):1785–1795 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Theodoratou E, Montazeri Z, Hawken S, et al. Systematic meta-analyses and field synopsis of genetic association studies in colorectal cancer. J Natl Cancer Inst. 2012;104(19):1433–1457 [DOI] [PubMed] [Google Scholar]
- 41. Toiyama Y, Takahashi M, Hur K, et al. Serum miR-21 as a Diagnostic and Prognostic Biomarker in Colorectal Cancer. J Natl Cancer Inst. 2013;105(12):849–859 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Galon J, Mlecnik B, Bindea G, et al. Towards the introduction of the ‘Immunoscore’ in the classification of malignant tumours. J Pathol. 2014;232(2):199–209 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. Kitkumthorn N, Mutirangura A. Long interspersed nuclear element-1 hypomethylation in cancer: biology and clinical applications. Clin Epigenetics. 2011;2(2):315–330 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44. Gaudet F, Hodgson JG, Eden A, et al. Induction of tumors in mice by genomic hypomethylation. Science. 2003;300(5618):489–492 [DOI] [PubMed] [Google Scholar]
- 45. Goel A, Xicola RM, Nguyen TP, et al. Aberrant DNA methylation in hereditary nonpolyposis colorectal cancer without mismatch repair deficiency. Gastroenterology. 2010;138(5):1854–1862 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46. Benard A, van de Velde CJ, Lessard L, et al. Epigenetic status of LINE-1 predicts clinical outcome in early-stage rectal cancer. Br J Cancer. 2013;109(12):3073–3083 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47. Palmqvist R, Wikberg M, Ling A, et al. The Association of Immune Cell Infiltration and Prognosis in Colorectal Cancer. Current Colorectal Cancer Reports. 2013;9(4):372–379 [Google Scholar]
- 48. Dahlin AM, Henriksson ML, Van Guelpen B, et al. Colorectal cancer prognosis depends on T-cell infiltration and molecular characteristics of the tumor. Mod Pathol. 2011;24(5):671–682 [DOI] [PubMed] [Google Scholar]
- 49. Di Caro G, Marchesi F, Laghi L, et al. Immune cells: plastic players along colorectal cancer progression. J Cell Mol Med. 2013;17(9):1088–1095 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50. Nosho K, Baba Y, Tanaka N, et al. Tumour-infiltrating T-cell subsets, molecular changes in colorectal cancer, and prognosis: cohort study and literature review. J Pathol. 2010;222(4):350–366 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51. Pobsook T, Subbalekha K, Sannikorn P, et al. Improved measurement of LINE-1 sequence methylation for cancer detection. Clin Chim Acta. 2011;412(3–4):314–321 [DOI] [PubMed] [Google Scholar]
- 52. Xie H, Wang M, de Andrade A, et al. Genome-wide quantitative assessment of variation in DNA methylation patterns. Nucleic Acids Res. 2011;39(10):4099–4108 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53. Szpakowski S, Sun X, Lage JM, et al. Loss of epigenetic silencing in tumors preferentially affects primate-specific retroelements. Gene. 2009;448(2):151–167 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54. Hansen KD, Timp W, Bravo HC, et al. Increased methylation variation in epigenetic domains across cancer types. Nat Genet. 2011;43(8):768–775 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55. Bartley AN, Yao H, Barkoh BA, et al. Complex patterns of altered MicroRNA expression during the adenoma-adenocarcinoma sequence for microsatellite-stable colorectal cancer. Clin Cancer Res. 2011;17(23):7283–7293 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 56. Earle JS, Luthra R, Romans A, et al. Association of microRNA expression with microsatellite instability status in colorectal adenocarcinoma. J Mol Diagn. 2010;12(4):433–440 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 57. Schetter AJ, Leung SY, Sohn JJ, et al. MicroRNA expression profiles associated with prognosis and therapeutic outcome in colon adenocarcinoma. JAMA. 2008;299(4):425–436 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 58. Nishihara R, Lochhead P, Kuchiba A, et al. Aspirin use and risk of colorectal cancer according to BRAF mutation status. JAMA. 2013;309(24):2563–2571 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 59. Febbo PG, Ladanyi M, Aldape KD, et al. NCCN Task Force report: Evaluating the clinical utility of tumor markers in oncology. J Natl Compr Canc Netw. 2011;9 Suppl 5:S1–32 [DOI] [PubMed] [Google Scholar]
- 60. Ogino S, Stampfer M. Lifestyle factors and microsatellite instability in colorectal cancer: the evolving field of molecular pathological epidemiology. J Natl Cancer Inst. 2010;102(6):365–367 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 61. Ogino S, Chan AT, Fuchs CS, et al. Molecular pathological epidemiology of colorectal neoplasia: an emerging transdisciplinary and interdisciplinary field. Gut. 2011;60(3):397–411 [DOI] [PMC free article] [PubMed] [Google Scholar]