Abstract
Background:
Telomeres play an important role in colorectal cancer (CRC) prognosis. Variation in telomere maintenance genes may be associated with survival after CRC diagnosis but evidence is limited. In addition, possible interactions between telomere maintenance genes and prognostic factors such as smoking and sex also remain to be investigated.
Methods:
We conducted gene-wide analyses of CRC prognosis in 4,896 invasive CRC cases from the Genetics and Epidemiology of Colorectal Cancer Consortium (GECCO).1871 common variants within 13 telomere maintenance genes were included. Cox models were fit to estimate associations of these variants individually with overall and CRC-specific survival. Likelihood ratio tests were used to test for interaction by smoking and sex. P-values were adjusted using Bonferroni correction.
Results:
The association between minor allele of rs7200950 (ACD) with CRC-specific survival varied significantly by smoking pack-years (corrected p-value=0.049), but no significant trend was observed. By sex, minor alleles for rs2975843 (TERF1), rs75676021 (POT1), and rs74429678 (POT1) were associated with decreased overall and/or CRC-specific survival in women but not in men.
Conclusions:
Our study reported a gene-wide statistically significant interaction with sex (TERF1, POT1). Although significant interaction by smoking pack-years (ACD) was observed, there was no evidence of a dose-response. Validation of these findings in other large studies and further functional annotation on these SNPs are warranted.
Impact:
Our study found a gene-smoking and gene-sex interaction on survival after CRC diagnosis, providing new insights into the role of genetic polymorphisms in telomere maintenance on CRC prognosis.
Keywords: telomere, telomere maintenance genes, colorectal cancer, survival
Introduction
Telomeres are comprised of repetitive nucleotide sequences that cap the ends of eukaryotic chromosomes (1) and protect chromosomes from deterioration or end-to-end fusion with neighboring chromosomes (2). Telomeres thus prevent aberrant chromosomal replication and help maintain chromosomal stability and genomic integrity. Telomere replication is regulated by telomerase complex, which is made up of telomerase reverse transcriptase (encoded by TERT), an RNA component (encoded by TERC), shelterin complex (encoded by TERF1, TERF2, TINF2, TERF2IP, ACD, and POT1) (3,4) and other associated proteins (encoded by TNKS, TNKS2, TNKS1BP1,TEP1 and PINX1) (5). Over time, telomeres shorten with each cell division, partly due to incomplete replication of the 3’-end of the chromosomes (1). Personal and lifestyle factors such as age, sex and cigarette smoking may also impact telomere function (6). Dysfunction in telomere replication mechanisms may result in accelerated genetic changes and cellular senescence. Hence, telomeres are considered to be a hallmark of aging.
Telomeres and telomerases may also play an integral role in cancer progression through overexpression of the telomerase enzyme. Indeed, genetic variation in telomere maintenance genes has been associated with overall and cancer-specific survival in cancers of the lung, glioma, liver, ovaries and breast (7–10). The relationship between telomere maintenance genes and CRC prognosis, however, is less clear. Further understanding of the prognostic role of telomeres and telomerases in CRC carcinogenesis may also help provide important insights into CRC treatment.
To date, no published studies have investigated whether telomere maintenance genes are specifically associated with survival after CRC diagnosis. To evaluate this association, we utilized data from the Genetics and Epidemiology of Colorectal Cancer Consortium (GECCO) (11) to elucidate the relationship between single nucleotide polymorphism (SNP) variation in 13 telomere maintenance genes (TERT, TERC, TERF1, TERF2, TINF2, TERF2IP, ACD, POT1, TNKS, TNKS2, TNKS1BP1, TEP1 and PINX1) and both overall and disease-specific survival after CRC diagnosis. We also considered whether such associations may be modified by host characteristics, such as smoking and sex, which are both involved in telomere erosion.
Methods
Study participants:
Study participants were drawn from 12 case-control and cohort studies, including data from seven cohort studies in the United States: The Seattle site of the Colon Cancer Family Registry (SCCFR), Health Professionals Follow-up Study (HPFS), Physicians’ Health Study (PHS), VITamins And Lifestyle study (VITAL); Women’s Health Initiative (WHI); Nurses’ Health Study (NHS), Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial (PLCO) and Diet, Activity and Lifestyle Study (DALS). GECCO study population and details of the participating studies have been described in detail previously (11).
For the current analysis, study subjects were restricted to participants with self-reported European descent, primary invasive CRC, and available genotype and survival information. CRC diagnosis was confirmed by medical records and pathology reports. The primary outcomes were death from any cause as well as CRC-specific deaths. Active follow-up was used to ascertain vital status in HPFS, PHS, NHS, PLCO, WHI; dates and causes of deaths were confirmed using death certificates and/or medical records. For VITAL, DALS and SCCFR, vital status was confirmed through National Death Index, cancer registries, state death records, or population registries, with cause of death verified by information from death certificates. All participants gave written informed consent and studies were approved by their Institutional Review Board (IRB) respectively. Studies were conducted in accordance with the Declaration of Helsinki. Characteristics of included studies are described in Table 1.
Table 1.
Characteristics of included studies in Genetics and Epidemiology of Colorectal Cancer Consortium (GECCO)
| Study | Case (N) | Male/Female (N /N) | Age (years)a | Smokingc | N. deaths, all-causeb | N. deaths, CRCb | Median follow-up time (days) | Platform* | ||
|---|---|---|---|---|---|---|---|---|---|---|
| Current (%) | Former (%) | Never (%) | ||||||||
| DALS 1 | 710 | 403/307 | 65.1 (32–79) | 101 (14.2) | 326 (45.9) | 283 (39.9%) | 244 (34.4%) | 135 (19.0%) | 1917.56 | 610K, 550K |
| DALS 2 | 415 | 220/195 | 65.1 (31–79) | 47 (11.4) | 168 (40.8) | 197 (47.8%) | 115 (27.7%) | 81 (19.5%) | 1674.1 | 300K |
| HPFS | 168 | 168/0 | 71.5 (50–90) | 6 (3.6) | 89 (53.9) | 70 (42.4%) | 82 (48.8%) | 47 (28.0%) | 2007.5 | 730K |
| NHS | 296 | 0/296 | 68.1 (46–85) | 37 (12.5) | 132 (44.7) | 126 (42.7%) | 118 (39.9%) | 89 (30.1%) | 2296.45 | 730K |
| PHS | 324 | 324/0 | 70.6 (44–92) | 34 (10.5) | 160 (49.4) | 130 (40.1%) | 200 (61.7%) | 131 (40.4%) | 2062.25 | 730K |
| PLCO 1 | 531 | 301/230 | 68.9 (55–82) | 54 (10.2) | 242 (45.6) | 235 (44.3%) | 180 (33.9%) | 108 (20.3%) | 2430 | 300/240S, 610K |
| PLCO 2 | 478 | 275/203 | 70.4 (55–86) | 46 (9.6) | 221 (46.2) | 211 (44.1%) | 103 (21.6%) | 75 (15.7%) | 1237.5 | 300K |
| SCCFR | 279 | 0/279 | 64.4 (47–74) | 38 (13.6) | 116 (41.6) | 125 (44.8%) | 99 (35.5%) | 54 (19.4%) | 3374 | 300K |
| VITAL | 285 | 150/135 | 69.9 (51–83) | 28 (9.9) | 152 (53.9) | 102 (36.2%) | 117 (41.1%) | 70 (24.6%) | 1847 | 300K |
| WHI 1 | 304 | 0/304 | 71.0 (52–86) | 21 (7.0) | 147 (49.0) | 132 (44%) | 103 (33.9%) | 77 (25.3%) | 1868 | 550K, 550K duo, and 610K |
| WHI 2 | 618 | 0/618 | 71.9 (50–91) | 45 (7.4) | 265 (43.5) | 299 (49.1%) | 177 (28.6%) | 132 (21.4%) | 1163.5 | 300K |
| WHI WGS | 488 | 0/488 | 71.4 (52–89) | 37 (7.7) | 227 (47.2) | 217 (45.1%) | 143 (29.3%) | 99 (20.3%) | 1337.5 | Whole genome sequencing |
Abbreviation: N: number; CRC: colorectal cancer
mean (range)
number (percentage of cases)
number (percentage of cases except for 30 missing value on smoking status
All genotyping platform, except for WHI WGS, were illumina assay
Data collection:
Data on demographic, lifestyle and environmental characteristics were collected through self-report using questionnaires and/or in-person or telephone interviews, details of which have been described previously (11). Data elements considered in the current analyses were age at diagnosis, sex, research study, cancer site, disease stage at diagnosis, smoking status, pack-years smoked and age at quitting smoking. Survival data included data on deaths from any cause, CRC-specific deaths, and time from diagnosis to death or last follow-up. Principal components analysis (PCA) was used for population stratification, to account for ancestry. Details of genotyping, quality assurance and quality control (QA/QC) and imputation are described in supplementary materials. Genomic DNA was extracted from blood samples or buccal cells by conventional methods. Genotyping platform used for each study is summarized in Table 1. Before imputation, genotyped SNPs were excluded based on call rate (< 98%), lack of Hardy-Weinberg Equilibrium in controls (HWE, P < 1×10−4), and minor allele frequency (MAF < 5% for WHI Set 1, DALS Set 1; MAF < 5/number of samples for all other studies). All autosomal SNPs were imputed to a reference panel generated from WHI whole genome sequencing (WGS).
Based on a literature search conducted through 31st December 2018, we included in our analysis genes encoding proteins that participate in telomere length regulation. Genetic variation in some of the included genes have previously been associated with risk of cancer, including CRC, (12–16) as well as with survival after cancers of lung, liver, brain, and ovary (7,8,10,17). Details on genes selected for the current analysis can be found in Supplementary Table S1. Data was available for 6,578 SNPs within our genes of interest. After focusing on common SNPs (MAF≥5%) (details can be found in supplementary materials), a total of 1,871 SNPs were included in the final analyses. All genotyping data have been published and deposited in dbGaP with accession numbers (18).
Statistical analysis:
Data from individual studies were combined for pooled statistical analyses. Within each gene being evaluated in the current study, Pearson’s correlation coefficients were computed to determine the correlations between every pair of SNPs within the gene, and principal components analysis (PCA) was performed to obtain the effect number of independent tests (Meff_G) (19). Meff_G was used for type I error control in Bonferroni correction in single-SNP model in survival analysis (19). Multivariate Cox proportional hazard regression models were used to estimate hazard ratios (HR) and 95% confidence intervals (95% CI) for the associations between single SNPs and survival. Separate models were constructed for overall and CRC-specific survival. Schoenfeld residuals were computed to check for the proportional hazards assumption. The dosage scaling from 0 to 2 represented the estimated number of copies of the count allele. All models were adjusted for age at diagnosis, sex, research study, and the first three principal components of genetic ancestry. Estimates from single-SNP models were considered to reach statistical significance if the adjusted p-value was <0.05 (adjusted for Meff_G of each gene). Host-related factors such as cigarette smoking history (yes/no), smoking status at the time of completing the questionnaire (former/current/never smoker), and categorical smoking pack-years (<12, 12–24, 25–44, ≥45, as dummy variable) were assessed for gene-environment (GxE) interaction. An interaction with sex was also assessed for the SNPs under study. Likelihood ratio tests were used to evaluate whether the interaction terms were significant. For any associations and/or interactions that had a raw p-value <0.05, we chose the SNPs with the smallest p-value as the representative SNP in the region. For those SNPs that were statistically significant in the main and interaction analyses, further sensitivity analyses were performed to evaluate potential heterogeneity by age at diagnosis, sex and smoking status across studies (more details in supplementary material). All analyses were conducted using R Version 3.4.3.
Results
This study included a total of 4,896 invasive CRC cases. After a mean follow-up of 5.5 years, a total of 1,681 deaths occurred, of which 1,098 (65.3%) were attributed to CRC. The demographic and clinical characteristics of our study participants are summarized in Table 2. The majority of the participants were women (62.4%) aged 65 years and older at diagnosis (72.5%). Smoking was common, with 56.3% reporting ever having smoked, but only 10.1% were current smokers. Most cases were colon (87.6%) versus rectal (12.4%) cancers.
Table 2.
Patients characteristics and clinical features for eligible participants from GECCO
| Characteristics | Cases | Deaths, number (percentage of cases, %) | ||
|---|---|---|---|---|
| N | % | all-cause, N (%) | CRC, N (%) | |
| Age (years) | ||||
| <65 | 1345 | 27.5 | 414 (30.8) | 321(23.9) |
| 65–69 | 1059 | 21.6 | 335 (31.6) | 218 (20.6) |
| 70–74 | 1251 | 25.6 | 461 (36.9) | 274 (21.9) |
| ≥75 | 1241 | 25.3 | 471 (38.0) | 285 (23.0) |
| Sex | ||||
| Male | 1841 | 37.6 | 727 (39.5) | 438 (23.8) |
| Female | 3055 | 62.4 | 954 (31.2) | 660 (21.6) |
| Cancer site | ||||
| Proximal | 2428 | 49.6 | 851 (35.0) | 556 (22.9) |
| Distal | 1596 | 32.6 | 493 (30.9) | 307 (19.2) |
| Rectal | 726 | 14.8 | 255 (35.1) | 181(24.9) |
| Othera | 146 | 3.0 | 82 (56.2) | 54 (37.0) |
| Cancer stageb | ||||
| In situ | 41 | 0.9 | 5 (12.2) | 1 (2.4) |
| Local | 1563 | 34.5 | 278 (17.8) | 74 (4.7) |
| Regional | 2366 | 52.2 | 745 (31.5) | 461 (19.5) |
| Distant | 563 | 12.4 | 489 (86.0) | 460 (81.7) |
| Smoking statusc | ||||
| Never | 2127 | 43.7 | 644 (30.3) | 453 (21.3) |
| Former | 2245 | 46.1 | 823 (36.7) | 522 (23.3) |
| Current | 494 | 10.1 | 204 (41.3) | 116 (23.5) |
| Ever smokerc | ||||
| Yes | 2739 | 56.3 | 1027 (37.5) | 638 (23.3) |
| No | 2127 | 43.7 | 644 (30.3) | 453 (21.3) |
| Smoking pack-yearsd | ||||
| <12 | 612 | 24.4 | 184 (30.1) | 128 (20.9) |
| 12–<25 | 645 | 25.7 | 197 (30.5) | 128 (19.8) |
| 25–<45 | 613 | 24.5 | 232 (37.8) | 145 (23.7) |
| ≥45 | 635 | 25.4 | 287 (45.2) | 156 (24.6) |
| Age quit smokinge | ||||
| <35 | 517 | 23.5 | 154 (29.8) | 111 (21.5) |
| 35–<45 | 516 | 23.5% | 165 (32.0) | 115 (22.3) |
| 45–<55 | 596 | 27.1% | 231 (38.8) | 147 (24.7) |
| ≥55 | 569 | 25.9% | 262 (46.0) | 134 (23.6) |
Other sites include those cancer sites that cannot be classified as proximal or distal colon or a rectal site
363 participants did not have data on stage at cancer diagnosis
30 study participants did not report on their smoking status
234 study participants did not report on the frequency or duration of smoking and therefore had missing data on smoking pack-years
47 former smokers did not report on the age at which they quit smoking
Associations between selected SNPs (with P <0.05) and overall and CRC-specific survival are presented in Table 3. Although SNPs located within TERT, TERF1, TNKS, TNKS1BP1, TEP1 and TERF2 were nominally associated with survival after CRC diagnosis (P <0.05), none of these associations remained significant after gene-level multiple comparison correction.
Table 3.
Hazard ratios and 95% confidence intervals for the association between selected SNPs (with unadjusted p-value <0.05) involved in telomere maintenance and survival after colorectal cancer diagnosis
| Outcome | SNP | Gene | HRa (95% CI) | P-value | adj. P-valueb | minor/ major allele | MAF |
|---|---|---|---|---|---|---|---|
| Overall survival | rs2075785 | TERT | 0.86 (0.76–0.98) | 0.018 | 0.736 | T/C | 0.123 |
| rs2981096 | TERF1 | 0.84 (0.70–0.99) | 0.046 | 0.274 | G/A | 0.053 | |
| rs10102030 | TNKS | 1.09 (1.01–1.18) | 0.036 | >0.99 | T/A | 0.230 | |
| rs4939134 | TNKS1BP1 | 1.08 (1.02–1.17) | 0.017 | >0.99 | G/C | 0.471 | |
| rs1760894 | TEP1 | 1.15 (1.04– 1.26) | 0.004 | 0.224 | C/T | 0.222 | |
| CRC-specific survival | rs2075785 | TERT | 0.85 (0.73–0.99) | 0.043 | >0.99 | T/C | 0.123 |
| rs2981096 | TERF1 | 0.79 (0.63–0.99) | 0.040 | 0.2418 | G/A | 0.053 | |
| rs10088969 | TNKS | 1.13 (1.02–1.25) | 0.018 | >0.99 | C/A | 0.226 | |
| rs2229101 | TEP1 | 0.78 (0.64–0.95) | 0.013 | 0.6783 | C/A | 0.064 | |
| rs251796 | TERF2 | 1.13 (1.02–1.25) | 0.015 | 0.2156 | G/A | 0.301 |
Abbreviation: CRC: colorectal cancer; HR: hazard ratio; CI: confidence intervals; MAF: minor allele frequency
adjusted for age at diagnosis, sex, study center and the first three principal components (pc)
p-value is adjusted using Bonferroni method using Meff_G
Next, we evaluated if genetic associations between telomere maintenance SNPs and survival differed by smoking status (Table 4). Associations with SNPs in TERT, TERF1, TERF2, PINX1, TEP1, TNKS and ACD showed suggestive differences by smoking status, but these differences were not statistically significant after correction for multiple testing. When evaluated by pack-years of smoking (Table 5), rs7200950 (ACD) was differentially associated with CRC-specific survival (adjusted P =0.049 for interaction), however, no clear dose-response was observed with increasing pack years of smoking. Therefore, this finding needs to be interpreted with caution. Additionally, comparing lowest vs. highest exposure groups (0 vs. ≥45) for pack-years of smoking suggested reduced CRC deaths with variants in TERF2IP (rs1865493) and TNKS (rs73202875). (Supplementary Table S2).
Table 4.
Hazard ratios and 95% confidence intervals for the association between selected SNPs (with unadjusted p-value <0.05) involved in telomere maintenance and survival after colorectal cancer diagnosis—stratified by cigarette smoking
| Outcome | SNP | Gene | HR a (95% CI) | P.int | adj. P-value b | minor/major allele | ||
|---|---|---|---|---|---|---|---|---|
| Non-smoker | Former smoker | Current smoker | ||||||
| Overall survival | rs56963355 | TERT | 1.27 (0.98–1.64) | 0.98 (0.75–1.22) | 0.48 (0.12–0.84) | 0.024 | >0.99 | T/G |
| rs2975842 | TERF1 | 0.86 (0.77–0.96) | 1.03 (0.93–1.14) | 1.06 (0.86–1.27) | 0.031 | 0.0942 | T/C | |
| rs2409652 | PINX1 | 0.90 (0.79–1.03) | 1.04 (0.92–1.16) | 1.23 (0.97–1.49) | 0.036 | >0.99 | T/C | |
| rs1760899 | TEP1 | 0.85 (0.71–1.01) | 1.09 (0.93–1.25) | 0.64 (0.41–0.87) | 0.007 | >0.99 | C/T | |
| CRC-specific survival | rs6420019 | TERT | 1.23 (0.98–1.54) | 1.49 (0.45–1.63) | 0.85 (0.85–1.22) | 0.013 | 0.52 | A/C |
| rs56106543 | TERF2 | 0.74 (0.51–1.09) | 1.08 (0.40–1.67) | 1.49 (0.61–1.46) | 0.014 | 0.196 | C/T | |
| rs13259648 | PINX1 | 1.27 (1.07–1.50) | 0.78 (0.75–1.32) | 0.99 (0.87–1.19) | 0.024 | 0.696 | T/G | |
| rs35656875 | TNKS | 0.63 (0.42–0.98) | 1.28 (0.09–1.97) | 1.43 (0.51–1.56) | 0.015 | >0.99 | G/C | |
| rs1760901 | TEP1 | 0.80 (0.65–0.99) | 0.48 (0.78–1.29) | 1.12 (0.83–1.24) | 0.002 | 0.102 | G/C | |
| rs6979 | ACD | 0.96 (0.83–1.11) | 0.69 (0.84–1.23) | 1.07 (0.89–1.18) | 0.023 | 0.069 | G/A | |
Abbreviation: CRC: colorectal cancer; HR: hazard ratio; CI: confidence intervals; Pint: P-value of likelihood ratio test for testing interaction by smoking
adjusted for age at diagnosis, sex, study center and the first three principal components (pc)
p-value is adjusted using Bonferroni method using Meff_G
Table 5.
Hazard ratios and 95% confidence intervals for the association between selected SNPs (with unadjusted p-value <0.05) involved in telomere maintenance and survival after colorectal cancer diagnosis—stratified by pack-years of smoking
| Outcome | SNP | Gene | HRa (95% CI), smoking pack-years | Pint | ||||
|---|---|---|---|---|---|---|---|---|
| 0 | <12 | 12–24 | 25–44 | ≥45 | ||||
| Overall survival | rs556947195 | TERT | 1.00 (0.75–1.34) | 1.48 (0.87–2.08) | 0.82 (0.39–1.25) | 0.76 (0.33–1.19) | 1.90 (1.47–2.33) | 0.013 |
| rs2975842 | TERF1 | 0.86 (0.77–0.96) | 1.13 (0.89–1.36) | 1.05 (0.83–1.27) | 1.16 (0.94–1.38) | 0.98 (0.76–1.20) | 0.033 | |
| rs251797 | TERF2 | 0.99 (0.88–1.12) | 1.08 (0.84–1.33) | 1.06 (0.82–1.30) | 1.25 (1.01–1.49) | 0.78 (0.54–1.02) | 0.019 | |
| rs67456872 | TNKS | 0.68 (0.48–0.97) | 2.42 (1.00–3.84) | 0.58 (0.17–0.99) | 1.59 (1.18–2.00) | 0.96 (0.55–1.37) | 0.002 | |
| rs76990680 | TNKS | 1.15 (1.02–1.30) | 0.82 (0.63–1.02) | 1.24 (0.99–1.49) | 0.97 (0.72–1.22) | 0.95 (0.70–1.20) | 0.096 | |
| CRC-specific survival | rs2242652 | TERT | 0.89 (0.73–1.07) | 0.76 (0.48–1.03) | 1.40 (0.93–1.87) | 0.82 (0.35–1.29) | 1.29 (0.82–1.76) | 0.025 |
| rs2853690 | TERT | 1.12 (0.93–1.33) | 1.17 (0.81–1.53) | 0.65 (0.41–0.89) | 1.30 (1.06–1.54) | 1.05 (0.81–1.29) | 0.042 | |
| rs153045 | TERF2 | 1.10 (0.95–1.28) | 1.16 (0.83–1.48) | 1.18 (0.85–1.51) | 0.74 (0.41–1.07) | 1.41 (1.08–1.74) | 0.024 | |
| rs10091836 | PINX1 | 0.88 (0.77–1.00) | 0.78 (0.59–0.98) | 1.19 (0.90–1.48) | 1.00 (0.71–1.29) | 1.26 (0.97–1.55) | 0.011 | |
| rs67456872 | TNKS | 0.63 (0.40–0.97) | 2.67 (0.91–4.43) | 0.81 (0.16–1.46) | 2.16 (1.51–2.81) | 0.95 (0.30–1.60) | 0.002 | |
| rs7200950 | ACD | 1.11 (0.83–1.48) | 1.65 (0.91–2.39) | 0.69 (0.24–1.14) | 1.45 (1.00–1.90) | 0.52 (0.07–0.97) | 0.016* | |
Abbreviation: CRC: colorectal cancer; HR: hazard ratio; CI: confidence intervals; Pint: P-value of likelihood ratio test
adjusted for age at diagnosis, sex, study center and the first three principal components (pc)
adjusted p-value is 0.0495; other adjusted p-values are not significant (>0.05)
Then, we evaluated the role of genetic variants located in telomere maintenance genes with survival after CRC, according to sex (Table 6). Two SNPs in POT1, rs75676021 and rs74429678, were differentially associated with sex, such that women had a poorer overall and CRC-specific survival (adjusted P =0.023 and 0.019 for interaction, respectively) compared to men. rs2975843 within the TERF1 gene also showed a gene-wide significant interaction with sex for the association with overall as well as CRC-specific survival (adjusted P=0.002 and 0.004 for interaction) such that women had a significantly poorer survival for both overall and CRC-specific survival than men.
Table 6.
Hazard ratios and 95% confidence intervals for the association between selected SNPs (with unadjusted p-value <0.05) involved in telomere maintenance and survival after colorectal cancer diagnosis—stratified by sex
| Outcome | SNP | Gene | HR a (95% CIs) | Pint | adj. P-value b | |
|---|---|---|---|---|---|---|
| Female | Male | |||||
| Overall survival | rs75676021 | POT1 | 1.21 (1.01–1.45) | 0.77 (0.60–0.95) | 0.002 | 0.023* |
| rs2853685 | TERT | 1.14 (1.03–1.26) | 0.89 (0.79–1.00) | 0.002 | 0.096 | |
| rs2975843 | TERF1 | 1.08 (0.99–1.18) | 0.84 (0.75–0.92) | 3.00×10−4 | 0.002* | |
| rs73615082 | TERF2IP | 0.76 (0.62–0.93) | 1.08 (0.85–1.30) | 0.019 | 0.154 | |
| rs4840518 | PINX1 | 1.21 (1.02–1.43) | 0.90 (0.72–1.08) | 0.026 | 0.748 | |
| rs77103162 | TNKS1BP1 | 0.93 (0.80–1.08) | 1.17 (0.99–1.34) | 0.0337 | 0.506 | |
| rs35259162 | TEP1 | 0.83 (0.73–0.94) | 1.02 (0.88–1.16) | 0.0293 | >0.99 | |
| rs3950296 | TERC | 0.91 (0.82–1.02) | 1.09 (0.96–1.22) | 0.0278 | 0.139 | |
| CRC-specific survival | rs74429678 | POT1 | 1.33 (1.07–1.65) | 0.75 (0.52–0.97) | 0.0019 | 0.019* |
| rs2736115 | TERT | 1.21 (1.07–1.37) | 0.90 (0.76–1.04) | 0.0027 | 0.108 | |
| rs2975843 | TERF1 | 1.12 (1.00–1.25) | 0.83 (0.71–0.94) | 6.00×10−4 | 0.004* | |
| rs73615082 | TERF2IP | 0.77 (0.60–0.98) | 1.12 (0.83–1.42) | 0.0391 | 0.313 | |
| rs10503412 | PINX1 | 1.21 (1.00–1.46) | 0.78 (0.58–0.97) | 0.0051 | 0.148 | |
| rs4416825 | TNKS | 0.97 (0.83–1.14) | 1.24 (1.01–1.47) | 0.0482 | >0.99 | |
| rs1760895 | TEP1 | 1.19 (1.02–1.40) | 0.86 (0.67–1.06) | 0.0195 | 0.995 | |
| rs9876206 | TERC | 0.89 (0.78–1.01) | 1.12 (0.95–1.28) | 0.0252 | 0.126 | |
Abbreviation: CRC: colorectal cancer; HR: hazard ratio; CI: confidence intervals; Pint: P-value of likelihood ratio test
adjustment for age at diagnosis, study center and the first three principal components (pc)
p-value is adjusted using Bonferroni method using Meff_G
Finally, we evaluated if CRC sties differentiated the association between telomere maintenance gene and survival. Some SNPs located within TEP1, TNKS2, PINX1, TERT, TNKS1BP1 and POT1 showed some suggestive association with survival (P<0.05, Supplementary Table S3), but none of them remain statistically significant after multiple comparison adjustment.
In sensitivity analyses, we did not observe any significant heterogeneity (P>0.05, Supplementary Table S6) by study in covariates including age at diagnosis, sex, and smoking. Therefore, it is reasonable to assume that the effects of age at diagnosis, sex and smoking status are common across studies.
Discussion
In this large candidate gene study of 4,896 colorectal cancer patients and variation in 13 telomere maintenance genes, including TERT, TERC, TERF1, TERF2, TINF2, TERF2IP, ACD, POT1, TNKS, TNKS2, TNKS1BP1, TEP1, PINX1, we found differential associations of CRC-specific survival with smoking pack-years (ACD) and sex (POT1 and TERF1). Specifically, rs2975843 (TERF1), rs75676021 (POT1), and rs74429678 (POT1) showed statistically significant interaction with sex, while rs7200950 (ACD) showed a suggestive association of smoking pack-years with CRC-specific survival but there was a lack of trend for dose-response. These SNPs decreased both overall and CRC-specific survival in women but not in men. Thus, the current analyses suggest that multiple variants in telomere maintenance genes may play a simultaneous role in progression of cancer, and that these variants may interact with lifestyle factors, including smoking and sex.
The dual role of telomeres and the enzyme telomerase in carcinogenesis and cancer progression is complex (20). Briefly, telomere shortening may lead to carcinogenesis but induce cell death in cancer cell lines (20); telomerase may promote tumor growth and aid in tumor progression. Indeed, several in vitro and in vivo studies have demonstrated an association between high levels of telomerase/TERT and poorer survival (21,22). Telomerase also participates in gene expression regulation, particularly NF-κB signaling, cell growth and migration, thus suggesting that telomerase may also act as a tumor-promoting factor (23,24). Taken together, it is biologically plausible that telomere maintenance genes may impact cancer prognosis, but the current evidence from existing population studies is limited. Previous reports have shown that variants in telomere maintenance genes may be associated with survival after ovarian (9,17) and breast cancer (25); however, the evidence for CRC-specific survival is lacking. There is evidence of a statistically significant association between variants in telomere maintenance genes and risk of developing CRC. SNPs within TERT (rs2736100, rs2736098) (13,16,26), and TERC (rs10936599) (15) have been associated with increased CRC susceptibility, but we did not detect a statistically significant association between these SNPs and survival in the current study. Variants within TERT, TERF2, PINX1 and TNKS have previously shown suggestive associations with overall mortality across multiple cancers, including glioblastoma, bladder, lung, breast, and ovary (8–10,25,27), although none of those associations remained statistically significant after multiple comparison corrections. These previous studies were moderately sized and, therefore, had limited power to investigate interactions with smoking and sex (where applicable). We were able to examine associations within subgroups of smoking and sex in our study, owing to the much larger sample size of our study.
In the current study, we observed a statistically significant interaction between rs7200950 (ACD) and smoking pack-years, although with no trend for dose-response. Cigarette smoking appears to accelerate telomere length shortening by oxidative stress (28) and methylation (29,30). Circulating telomere length is inversely associated with ever smoking (31) and the number of packs smoked per day (6,32) among current smokers. Short telomeres and smoking have been previously shown to jointly affect the risk of CRC (33). None of the previous studies have looked at joint associations of these genes with CRC-specific survival and smoking status. Recently, a study among non-smoking Asian women demonstrated that variants in telomere maintenance genes associated with longer telomere lengths are also associated with progression of lung cancer (34). These polymorphisms might be interacting similarly among patients with colorectal cancer.
Our study found a statistically significant association for variants within POT1 and TERF1 and decreased overall and CRC-specific survival in women, but not in men. Another study found similar results with RAP1, another telomere maintenance gene in lung cancer patients (35). These results can be partially explained by the effect of regulation of sex hormones on telomerase activity. Estrogen activates telomerase via upregulating the telomerase catalytic subunit or activating c-Myc/Max that then binds to TERT promoter to increase its activity (36). Furthermore, telomere length in men has been shown to be shorter compared to that in similarly aged women (37), and telomere length and sex are both associated with CRC risk (33).
To the best of our knowledge, this is the first study investigating the association between genetic variants involved in telomere maintenance and survival after CRC diagnosis. Our study has a large sample size with long-term follow-up and validated survival outcomes data, which permitted a robust assessment of a gene-wide main effect and GxE interactions. We had access to detailed data on smoking status allowing us to study the effect of smoking quantity, increasing the sensitivity and specificity of our analyses. We acknowledge some limitations of our work. We only included common variants in telomere maintenance genes in our analyses and, therefore, we may have missed any associations with low-frequency and rare variants. Larger sample sizes will be required to analyze such low-frequency variants. We included a comprehensive list of telomere maintenance genes, but it is possible that we missed additional genes contributing to telomere length regulation. Further, all autosomal SNPs were imputed and we used the expected number of copies of the minor allele in our analyses. However, we restricted SNPs with high imputation accuracy and previous reports show that imputed SNPs provide unbiased inference (38).
In conclusion, our large gene-wide study observed suggestive associations between genetic variation related to telomere maintenance function and overall as well as CRC-specific survival. We also observed statistically significant interactions between genes involved in telomere maintenance, smoking pack-years (ACD) and sex (POT1, TERF1) on their association with survival after CRC diagnosis. Current results need to be verified in larger studies and further functional annotation of the identified variants in this study may be of interest.
Supplementary Material
Acknowledgements
GECCO: Genetics and Epidemiology of Colorectal Cancer Consortium (GECCO) is supported by grants from the National Cancer Institute (NCI), National Institutes of Health (NIH), U.S. Department of Health and Human Services (U01 CA164930 and R01 CA059045 to U. Peters). This research was funded in part through the NIH/NCI Cancer Center Support Grant P30 CA015704. We would like to thank all people that made the study possible. We appreciate the efforts of the GECCO Coordinating Center to ensure the data collaboration.
Harvard cohorts (HPFS, NHS, PHS): Health Professionals Follow-up Study (HPFS) is supported by the National Institutes of Health (P01 CA055075 to E. Giovanucci, UM1 CA167552 and U01 CA167552 to W. C. Willett, R01 CA151993 and R35 CA197735 to S. Ogino, K07 CA190673 to R. Nishihara, R01 CA137178 and P50 CA127003 to A. T. Chan), Nurses’ Health Study (NHS) by the National Institutes of Health (P01 CA087969 to E. Giovanucci, UM1 CA186107 to M. Stampfer, R01 CA151993 and R35 CA197735 to S. Ogino, K07 CA190673 to R. Nishihara, R01 CA137178 and P50 CA127003 to A. T. Chan) and Physician’s Health Study (PHS) by the National Institutes of Health (R01 CA042182 to J. Ma). The study protocol was approved by the institutional review boards of the Brigham and Women’s Hospital and Harvard T.H. Chan School of Public Health, and those of participating registries as required. We would like to thank the participants and staff of the HPFS, NHS, and PHS 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. The authors assume full responsibility for analyses and interpretation of these data.
PLCO: Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial (PLCO) is supported by grants from the Intramural Research Program of the Division of Cancer Epidemiology and Genetics and supported by contracts from the Division of Cancer Prevention, National Cancer Institute, NIH, DHHS. Funding was provided by National Institutes of Health (NIH), Genes, Environment and Health Initiative (GEI) Z01 CP 010200, NIH U01 HG004446, and NIH GEI U01 HG 004438. The authors thank the PLCO Cancer Screening Trial screening center investigators and the staff from Information Management Services Inc and Westat Inc. Most importantly, we thank the study participants for their contributions that made this study possible.
SCCFR: The Seattle (SCCFR) site of the Colon CFR Cohort (www.coloncfr.org), is supported in part by the National Cancer Institute (NCI) of the National Institutes of Health (NIH) Award U01 CA167551. Additional support for the SCCFR, Postmenopausal Hormones and Colon Cancer (PMH) study and the SCCFR Illumina HumanCytoSNP array were through NCI/NIH awards U01/U24 CA074794 and R01 CA076366 (to P. A. Newcomb). Support for case ascertainment was provided from the Surveillance, Epidemiology and End Results (SEER) Program of the NCI. The authors wish to acknowledge the generous contributions of the study participants and dedication of study staff of the SCCFR and the PMH (CORE Studies) and the financial support from the National Cancer Institute, without which this important research was not possible. The content of this manuscript does not necessarily reflect the views or policies of the NIH or SCCFR, nor does mention of trade names, commercial products, or organizations imply endorsement by the US Government, the SEER Program, or the SCCFR.
WHI: The Women’s Health Initiative (WHI) program is funded by the National Heart, Lung, and Blood Institute, National Institutes of Health, U.S. Department of Health and Human Services through contracts HHSN268201100046C, HHSN268201100001C, HHSN268201100002C, HHSN268201100003C, HHSN268201100004C, and HHSN271201100004C. The authors thank the WHI investigators and staff for their dedication, and the study participants for making the program possible. A full listing of WHI investigators can be found at: http://www.whi.org/researchers/Documents%20%20Write%20a%20Paper/WHI%20Investigator%20Short%20List.pdf
Diet, Activity and Lifestyle Survey (DALS) is supported by grants from the National Institutes of Health (R01 CA48998 to M. L. Slattery). VITamins And Lifestyle (VITAL) is supported by grants from National Institutes of Health (K05 CA154337 to E. White). Dr. Newcomb was supported by an established investigator award (K05 CA152715) and Dr. Hardikar was supported by a career development award (K07 CA222060) from the National Cancer Institute, National Institutes of Health.
Footnotes
Conflict of Interest: The authors declare no potential conflicts of interest
References
- 1.Wentzensen IM, Mirabello L, Pfeiffer RM, Savage SA. The Association of Telomere Length and Cancer: A Meta-Analysis. Cancer Epidemiol Biomark Prev Publ Am Assoc Cancer Res Cosponsored Am Soc Prev Oncol. 2011;20:1238–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Blackburn EH, Epel ES, Lin J. Human telomere biology: A contributory and interactive factor in aging, disease risks, and protection. Science. 2015;350:1193–8. [DOI] [PubMed] [Google Scholar]
- 3.de Lange T . Shelterin: the protein complex that shapes and safeguards human telomeres. Genes Dev. 2005;19:2100–10. [DOI] [PubMed] [Google Scholar]
- 4.Collins K, Mitchell JR. Telomerase in the human organism. Oncogene. 2002;21:564–79. [DOI] [PubMed] [Google Scholar]
- 5.Savage SA, Stewart BJ, Eckert A, Kiley M, Liao JS, Chanock SJ. Genetic variation, nucleotide diversity, and linkage disequilibrium in seven telomere stability genes suggest that these genes may be under constraint. Hum Mutat. 2005;26:343–50. [DOI] [PubMed] [Google Scholar]
- 6.Sanders JL, Newman AB. Telomere Length in Epidemiology: A Biomarker of Aging, Age-Related Disease, Both, or Neither? Epidemiol Rev. 2013;35:112–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Jung Seok Won Park Neung Hwa, Shin Jung Woo Park Bo Ryung, Kim Chang Jae Lee Jong-Eun, et al. Prognostic impact of telomere maintenance gene polymorphisms on hepatocellular carcinoma patients with chronic hepatitis B. Hepatology. 2014;59:1912–20. [DOI] [PubMed] [Google Scholar]
- 8.Mosrati MA, Malmström A, Lysiak M, Krysztofiak A, Hallbeck M, Milos P, et al. TERT promoter mutations and polymorphisms as prognostic factors in primary glioblastoma. Oncotarget. 2015;6:16663–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Harris HR, Vivo ID, Titus LJ, Vitonis AF, Wong JYY, Cramer DW, et al. Genetic variation in telomere maintenance genes in relation to ovarian cancer survival. Int J Mol Epidemiol Genet. 2012;3:252–61. [PMC free article] [PubMed] [Google Scholar]
- 10.Catarino R, Araújo A, Coelho A, Gomes M, Nogueira A, Lopes C, et al. Prognostic significance of telomerase polymorphism in non-small cell lung cancer. Clin Cancer Res Off J Am Assoc Cancer Res. 2010;16:3706–12. [DOI] [PubMed] [Google Scholar]
- 11.Peters U, Jiao S, Schumacher FR, Hutter CM, Aragaki AK, Baron JA, et al. Identification of Genetic Susceptibility Loci for Colorectal Tumors in a Genome-Wide Meta-analysis. Gastroenterology. 2013;144:799–807.e24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Mocellin S, Verdi D, Pooley KA, Landi MT, Egan KM, Baird DM, et al. Telomerase reverse transcriptase locus polymorphisms and cancer risk: a field synopsis and meta-analysis. J Natl Cancer Inst. 2012;104:840–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Peters U, Hutter CM, Hsu L, Schumacher FR, Conti DV, Carlson CS, et al. Meta-analysis of new genome-wide association studies of colorectal cancer risk. Hum Genet. 2012;131:217–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Hung RJ, Ulrich CM, Goode EL, Brhane Y, Muir K, Chan AT, et al. Cross Cancer Genomic Investigation of Inflammation Pathway for Five Common Cancers: Lung, Ovary, Prostate, Breast, and Colorectal Cancer. J Natl Cancer Inst. 2015;107. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Jones AM, Beggs AD, Carvajal-Carmona L, Farrington S, Tenesa A, Walker M, et al. TERC polymorphisms are associated both with susceptibility to colorectal cancer and with longer telomeres. Gut. 2012;61:248–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Karami S, Han Y, Pande M, Cheng I, Rudd J, Pierce BL, et al. Telomere structure and maintenance gene variants and risk of five cancer types. Int J Cancer. 2016;139:2655–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Sun Y, Tao W, Huang M, Wu X, Gu J. Genetic variants in telomere-maintenance genes are associated with ovarian cancer risk and outcome. J Cell Mol Med. 2017;21:510–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Huyghe JR, Bien SA, Harrison TA, Kang HM, Chen S, Schmit SL, et al. Discovery of common and rare genetic risk variants for colorectal cancer. Nat Genet. 2019;51:76–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Gao X, Starmer J, Martin ER. A multiple testing correction method for genetic association studies using correlated single nucleotide polymorphisms. Genet Epidemiol. 2008;32:361–9. [DOI] [PubMed] [Google Scholar]
- 20.Hackett JA, Greider CW. Balancing instability: dual roles for telomerase and telomere dysfunction in tumorigenesis. Oncogene. 2002;21:619–26. [DOI] [PubMed] [Google Scholar]
- 21.Bertorelle R, Briarava M, Rampazzo E, Biasini L, Agostini M, Maretto I, et al. Telomerase is an independent prognostic marker of overall survival in patients with colorectal cancer. Br J Cancer. 2013;108:278–84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Fernández-Marcelo T, Sánchez-Pernaute A, Pascua I, Juan CD, Head J, Torres-García A-J, et al. Clinical Relevance of Telomere Status and Telomerase Activity in Colorectal Cancer. PLOS ONE. 2016;11:e0149626. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Ghosh A, Saginc G, Leow SC, Khattar E, Shin EM, Yan TD, et al. Telomerase directly regulates NF-κB-dependent transcription. Nat Cell Biol. 2012;14:1270–81. [DOI] [PubMed] [Google Scholar]
- 24.Li S, Crothers J, Haqq CM, Blackburn EH. Cellular and gene expression responses involved in the rapid growth inhibition of human cancer cells by RNA interference-mediated depletion of telomerase RNA. J Biol Chem. 2005;280:23709–17. [DOI] [PubMed] [Google Scholar]
- 25.Shen J, Gammon MD, Terry MB, Bradshaw PT, Wang Q, Teitelbaum SL, et al. Genetic polymorphisms in telomere pathway genes, telomere length, and breast cancer survival. Breast Cancer Res Treat. 2012;134:393–400. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Jannuzzi AT, Karaman E, Oztas E, Yanar HT, Özhan G. Telomerase Reverse Transcriptase (TERT) Gene Variations and Susceptibility of Colorectal Cancer. Genet Test Mol Biomark. 2015;19:692–7. [DOI] [PubMed] [Google Scholar]
- 27.Rachakonda PS, Hosen I, de Verdier PJ, Fallah M, Heidenreich B, Ryk C, et al. TERT promoter mutations in bladder cancer affect patient survival and disease recurrence through modification by a common polymorphism. Proc Natl Acad Sci U S A. 2013;110:17426–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Kawanishi S, Oikawa S. Mechanism of telomere shortening by oxidative stress. Ann N Y Acad Sci. 2004;1019:278–84. [DOI] [PubMed] [Google Scholar]
- 29.Ambatipudi S, Cuenin C, Hernandez-Vargas H, Ghantous A, Le Calvez-Kelm F, Kaaks R, et al. Tobacco smoking-associated genome-wide DNA methylation changes in the EPIC study. Epigenomics. 2016;8:599–618. [DOI] [PubMed] [Google Scholar]
- 30.Lee KWK, Pausova Z. Cigarette smoking and DNA methylation. Front Genet [Internet]. 2013. [cited 2018 Apr 23];4. Available from: https://www.frontiersin.org/articles/10.3389/fgene.2013.00132/full [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Pellatt AJ, Wolff RK, Lundgreen A, Cawthon R, Slattery ML. Genetic and lifestyle influence on telomere length and subsequent risk of colon cancer in a case control study. Int J Mol Epidemiol Genet. 2012;3:184–94. [PMC free article] [PubMed] [Google Scholar]
- 32.Patel CJ, Manrai AK, Corona E, Kohane IS. Systematic correlation of environmental exposure and physiological and self-reported behaviour factors with leukocyte telomere length. Int J Epidemiol. 2017;46:44–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Qin Q, Sun J, Yin J, Liu L, Chen J, Zhang Y, et al. Telomere Length in Peripheral Blood Leukocytes Is Associated with Risk of Colorectal Cancer in Chinese Population. PLoS ONE [Internet]. 2014. [cited 2018 May 9];9. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3912164/ [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Machiela MJ, Hsiung CA, Shu X-O, Seow WJ, Wang Z, Matsuo K, et al. Genetic variants associated with longer telomere length are associated with increased lung cancer risk among never-smoking women in Asia: a report from the female lung cancer consortium in Asia. Int J Cancer. 2015;137:311–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Lin X, Gu J, Lu C, Spitz MR, Wu X. Expression of Telomere-Associated Genes as Prognostic Markers for Overall Survival in Patients with Non–Small Cell Lung Cancer. Clin Cancer Res. 2006;12:5720–5. [DOI] [PubMed] [Google Scholar]
- 36.Kyo S, Takakura M, Kanaya T, Zhuo W, Fujimoto K, Nishio Y, et al. Estrogen Activates Telomerase. Cancer Res. 1999;59:5917–21. [PubMed] [Google Scholar]
- 37.Gardner M, Bann D, Wiley L, Cooper R, Hardy R, Nitsch D, et al. Gender and telomere length: Systematic review and meta-analysis. Exp Gerontol. 2014;51:15–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Jiao S, Hsu L, Hutter CM, Peters U. The use of imputed values in the meta-analysis of genome-wide association studies. Genet Epidemiol. 2011;35:597–605. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.