Abstract
Background:
Inheritance of a germline mutation in one of the DNA mismatch repair (MMR) genes MLH1, MSH2, MSH6, and PMS2 causes a high risk of colorectal and other cancers (Lynch Syndrome). Use of aspirin has been shown to be associated with a reduced risk of colorectal cancer for the general population as well as for MMR gene mutation carriers. The aim of this study was to determine whether use of aspirin and ibuprofen in a nontrial setting is associated with the risk of colorectal cancer risk for MMR gene mutation carriers.
Methods:
We included 1858 participants in the Colon Cancer Family Registry who had been found to have a pathogenic germline mutation in a MMR gene (carriers). We used weighted Cox proportional hazards regression to estimate hazard ratios (HRs) and 95% confidence intervals (CIs). All statistical tests were two-sided.
Results:
A total of 714 carriers (38%) were diagnosed with colorectal cancer at a mean age of 42.4 (standard deviation 10.6) years. A reduced risk of colorectal cancer was associated with aspirin use (for 1 month to 4.9 years: HR = 0.49, 95% CI = 0.27 to 0.90, P = .02; for ≥5 years: HR = 0.25, 95% CI = 0.10 to 0.62, P = .003) and ibuprofen use (for 1 month to 4.9 years: HR = 0.38, 95% CI = 0.18 to 0.79, P = .009; for ≥5 years: HR = 0.26, 95% CI = 0.10 to 0.69, P = .007), compared with less than one month of use.
Conclusion:
Our results provide additional evidence that, for MMR gene mutation carriers, use of aspirin and ibuprofen might be effective in reducing their high risk of colorectal cancer.
Lynch Syndrome, previously termed Hereditary Non-Polyposis Colorectal Cancer, is an inherited susceptibility to colorectal and other cancers. Lynch Syndrome is caused by germline mutations in the DNA mismatch repair (MMR) genes MLH1, MSH2, MSH6, and PMS2, which encode proteins that identify and excise DNA mismatches (1). The prevalence of MMR gene mutation carriers in the general population has been estimated to range from approximately one in 370 to one in 3100, depending on the population and calculation assumptions (2,3). Lynch Syndrome causes 2% to 4% of all colorectal cancers and 10% to 15% of colorectal cancers diagnosed before the age of 50 years (4,5).
The average cumulative risk of colorectal cancer for MMR gene mutation carriers is substantial: about 50% by age 70 years for MLH1 and MSH2 carriers (6). However, there is evidence that there is very wide variation in risk even across carriers of mutations in the same gene (6), such that the majority of carriers are either at only modestly increased risk or at very high risk, rather than being clustered around the average risk. These observations raise the possibility that risk for mutation carriers is strongly determined by other genetic and/or environmental factors and that risk might be reduced substantially if relevant modifiable factors can be identified.
The main risk-reduction strategy for mutation carriers is regular screening with colonoscopy and polypectomy, which has been shown to reduce both colorectal cancer incidence and mortality (7). Accordingly, screening guidelines recommend that MMR mutation carriers have colonoscopy every one to two years from age 25 years, or five years earlier than the youngest colorectal cancer diagnosis in the family, whichever comes first (8–12).
Another potential colorectal cancer risk reduction approach, for which evidence has been accumulating over the last two decades, is chemoprevention using aspirin or other nonsteroidal anti-inflammatory drugs (NSAIDs). Several randomized controlled trials (RCT) and large cohort studies have consistently shown that use of aspirin and other NSAIDs reduces the risk of colorectal cancer for the general population (13–16). Five RCTs have also shown a protective effect of aspirin against colorectal adenoma for individuals with a personal history of colorectal neoplasia (17,18).
For MMR gene mutation carriers, evidence of a protective effect of aspirin use was provided by the CAPP2 RCT, the first trial of aspirin chemoprevention with cancer as the primary endpoint. At the end of two years of treatment, there was no evidence for an effect of daily high-dose (600mg) aspirin on colorectal cancer risk (19). However, an analysis after a mean follow-up of 56 months found that the mutation carriers who had been randomly assigned to aspirin had about a 40% reduction in colorectal cancer incidence compared with the control patients (20). The effect of low-dose aspirin and other NSAIDs including ibuprofen on colorectal cancer risk for carriers outside a trial setting is unknown.
We report here results from the largest observational study to date investigating the association between use of aspirin and the risk of colorectal cancer for MMR gene mutation carriers and present the first analysis of a potential association with use of ibuprofen, a widely available nonaspirin NSAID.
Methods
Study Sample
The study sample consisted of participants in the Colon Cancer Family Registry who had been genetically tested and found to be carriers of a germline pathogenic mutation in an MMR gene. A detailed description of the Colon Cancer Family Registry design and recruitment strategy has been presented elsewhere (21) and is available at http://coloncfr.org. Briefly, between 1997 and 2012, families were recruited through: 1) population-based probands with newly diagnosed colorectal cancer identified by state or regional population cancer registries in the United States (WA, CA, AZ, MN, CO, NH, NC, and HI), Australia (Victoria), and Canada (Ontario); and 2) clinic-based probands from multiple-case families referred to family cancer clinics in the United States (Mayo Clinic, Rochester, MN, and Cleveland Clinic, Cleveland, OH), Canada (Ontario), Australia (Melbourne, Adelaide, Perth, Brisbane, Sydney), and New Zealand (Auckland). We obtained permission from probands to contact their relatives to seek their enrollment in the Colon Cancer Family Registry. Informed consent was obtained from all study participants, and the study protocol was approved by the research ethics review boards of each recruitment center.
Data Collection
Epidemiologic and demographic information was collected using in-person interviews, telephone interviews, or mailed questionnaires at the time of recruitment. The questionnaires asked about participants’ characteristics; personal and family history of cancer; history of cancer screening; history of polyps, polypectomy and other surgeries; and use of aspirin, ibuprofen, and other medications. The detailed questionnaires used by each Colon Cancer Family Registry center are available at: http://coloncfr.org/questionnaires. Reported cancer diagnoses and ages at diagnosis were confirmed if possible using pathology reports, medical records, cancer registry reports, and death certificates. We attempted to obtain blood samples from all study participants and tumor tissue from colorectal cancer–affected participants.
MMR Gene Mutation Testing
Screening for germline mutations in MLH1, MSH2, MSH6, and PMS2 was performed for all population-based probands who had a colorectal tumor that showed evidence of impaired MMR function by either tumor microsatellite instability (MSI) or a lack of MMR protein expression by immunohistochemistry, or both. For clinic-based families, screening was performed for the youngest colorectal cancer case patient regardless of MSI or MMR protein expression status. For the MLH1, MSH2, and MSH6 genes, mutation testing was performed by Sanger sequencing or denaturing high performance liquid chromatography (dHPLC), followed by confirmatory DNA sequencing. Large duplication and deletion mutations were detected by Multiplex Ligation Dependent Probe Amplification (MLPA) according to the manufacturer’s instructions (MRC Holland, Amsterdam, the Netherlands) (21–23). For the PMS2 gene, mutation testing involved a modified protocol from Senter et al. (24), in which exons 1–5, 9, and 11–15 were amplified using three long-range polymerase chain reactions (PCRs) followed by nested exon-specific PCR/sequencing, while the remaining exons (6, 7, 8, and 10) were amplified and sequenced directly from genomic DNA. Large-scale deletions in PMS2 were detected using the P008-A1 MLPA kit (MRC Holland, Amsterdam, the Netherlands). Relatives of probands with a pathogenic MMR germline mutation (25) who had provided a blood sample were tested for the specific mutation identified in the proband. In this study, we included all probands and their relatives who had been found to be carriers of a germline pathogenic mutation in a MMR gene.
Exposure Definitions
Aspirin intake and ibuprofen intake were the two main exposure variables. Never use was defined as answering “no” to “Have you ever taken (aspirin/ibuprofen) at least twice a week for a month or longer?” Ever use was defined as answering “yes” to “Have you ever taken (aspirin/ibuprofen) at least twice a week for a month or longer?”
Duration of aspirin/ibuprofen use was based on the question “How long, in total, have you taken this medication for at least twice a week for a month or longer?” We calculated the age at first use of aspirin/ibuprofen (under the assumption that duration was continuous and recent) by subtracting the reported duration of use from the age at interview. Then, we calculated the duration of use as the total number of years of (aspirin/ibuprofen) use between the age at first use and the age at colorectal cancer diagnosis or censoring.
Those who answered “yes” to the question “Have you ever taken (aspirin/ibuprofen) at least twice a week for a month or longer?” and reported a duration that was shorter than the time interval between age at colorectal cancer diagnosis or censoring and age at interview, were classified as never users.
Statistical Analysis
We conducted weighted Cox proportional hazards regression analyses, with age as the time scale, to estimate the associations of use of aspirin only, ibuprofen only, and aspirin and/or ibuprofen with the risk of colorectal cancer. Time at risk started from birth and ended at age at first diagnosis of colorectal cancer (n = 714), age of diagnosis of non-colorectal cancer (n = 304), polypectomy (n = 306), or age at interview (n = 534), whichever occurred first. We chose to censor at age at any cancer diagnosis because cancer treatment could alter subsequent colorectal cancer risk. Similarly, we censored at age at first polypectomy because removal of polyps markedly reduces colorectal cancer risk (7). This prevents the potential confounding of colonoscopy screening on the associations between the exposure variables and colorectal cancer risk.
Colorectal cancer cases with a strong family history and/or early onset were preferentially tested for MMR gene mutations in the Colon Cancer Family Registry. Therefore, selection of the mutation carriers to be included in this study was not random with respect to disease status—i.e., there was an excess of colorectal cancer case patients compared with what would be expected in an unselected cohort of carriers. To adjust for this nonrandom sampling, we created a “synthetic” cohort using the weighted cohort approach described by Antoniou et al. (26), which has been previously applied in studies of disease-risk modifiers for carriers of rare genetic mutations (27–29). This method involves giving affected and unaffected carriers age-specific relative weights based on their sampling probabilities so that the age-specific incidence in the synthetic cohort is consistent with the incidence estimated for carriers in the general population (26).
In this study, as described in detail by Pande et al. (30), population-based estimates of colorectal cancer incidence for five-year intervals (31) were multiplied by the relative risks of colorectal cancer for MMR gene mutation carriers (32) to obtain age-specific incidences (see Supplementary Table 1, available online). These age-specific incidences were used to calculate sampling fractions in order to weight mutation carriers so that for each five-year interval the proportion of affected carriers in the synthetic cohort equaled the proportion of affected carriers in the general population. A simulation study of this approach using Cox regression showed that allowing for nonrandom sampling by using external age-specific incidences led to unbiased estimates of risk modifiers when the external incidences were correctly specified (26).
Aspirin-only, ibuprofen-only, and aspirin and/or ibuprofen use were fit separately as binary (ever vs never), and categorical (<1 month, 1 month, to 4.9 years, and ≥5 years) variables, and corresponding hazard ratios (HRs) and 95% confidence intervals (CIs) were estimated for associations with colorectal cancer risk. The proportional hazards (PH) assumption was tested by examining the relationship between the scaled Schoenfeld residuals and survival time (33).
Model-building strategies were based on the “purposeful selection” method described by Hoshmer and Lemeshow (34). Selection of covariates for inclusion in the multivariable models was based on statistical significance at the 25% level in the univariate models and on clinical importance for any variables not selected with this criterion. The following factors were considered as potential confounders: 1) year of birth (1914–43, 1944–54, 1955–65, 1966–90); 2) sex; 3) country of recruitment; 4) ethnicity; 5) education; 6) smoking status (never, former, current); 7) number of alcohol drinks per day; 8) body mass index two years before interview (underweight [<18.5kg/m2], normal [18.5–24.9kg/m2], overweight [25.0–29.9kg/m2], obese [≥30kg/m2]); 9) history of diabetes (yes, no); 10) multivitamin supplement use (<1 month vs ≥1 month); 11) regular physical activity (at least 30 minutes per week for at least three months); 12) acetaminophen (<1 month vs ≥1 month); and 13) laxatives(<1 month vs ≥1 month) for both sexes and 14) hormone replacement therapy use (<6 vs ≥6 months); and 15) number of live births (none, 1, 2, ≥3) for women. Variables that did not meet the PH assumption were included as time-dependent covariates in the model. Missing values of covariates were coded as an additional category to allow inclusion of as many carriers as possible in the multivariable model fits. Tests for interactions between exposures and potential effect modifiers were performed using the log-likelihood ratio test.
To identify any potential survival bias on recruitment, we further conducted an analysis restricted to carriers who were diagnosed with colorectal cancer, or censored, within five years prior to the interview. To determine whether censoring at polypectomy influenced the estimated associations, we conducted an analysis using colorectal neoplasia (i.e., either colorectal cancer or colorectal polyps) as the outcome. We also conducted an analysis restricted to carriers for whom there was complete covariate data (i.e., carriers with a missing value for any covariate were excluded).
We applied the Huber-White robust variance estimation by clustering on family membership to allow for a correlation in risk between family members (35,36). All statistical tests were two-sided, and, following convention, nominal statistical significance was determined at P values of less than .05. Analyses were performed using Stata 13 (37).
Results
From the Colon Cancer Family Registry, we identified a total of 2003 carriers of a pathogenic mutation in a MMR gene. Of these, we excluded 11 (0.5%) because they were censored before age 18 years and 134 (6.7%) because they had missing data on both main exposure variables. Excluded carriers did not differ in main characteristics from those who were entered into the analyses (details not shown).
The study therefore involved 1858 MMR gene mutation carriers from 748 families (25% population-based), of whom 1038 (55.8%) were female, who contributed a total of 77 494 person-years. Of these, 681 carried a mutation in MLH1, 873 in MSH2, 204 in MSH6, and 100 in PMS2, and 714 (38%) were diagnosed with colorectal cancer at a mean age at diagnosis of 42.4 years (SD = 10.6). A total of 653 (91%) reported cancers were confirmed by pathology, clinical, or cancer-registry records. The time interval between age at colorectal cancer diagnosis or censoring and age at interview was on average 9.2 (SD = 9.9) years for carriers with colorectal cancer and 9.9 (SD 9.4) years for carriers without colorectal cancer (P = .14).
Characteristics of participating MMR gene mutation carriers are presented in Table 1. Of all carriers, 117 (6.7%) reported having used aspirin only for at least one month, 126 (7.2%) reported having used ibuprofen only for at least one month, and 286 (14.3%) reported having used aspirin and/or ibuprofen for at least one month. The prevalence of aspirin and ibuprofen use by country, age group, and sex is shown in Table 2. The prevalence of aspirin use for carriers affected with colorectal cancer and unaffected carriers was 6.5% and 6.9%, respectively (Supplementary Table 2, available online).
Table 1.
Characteristics of mismatch repair gene mutation carriers by colorectal cancer diagnosis status
| Variables | No colorectal cancer (n = 1144, 61%) |
Colorectal cancer (n = 714, 38%) |
Total (n = 1858) |
|---|---|---|---|
| Study centers | |||
| Australia or New Zealand USA Canada |
687 (60.0) 304 (26.5) 153 (13.3) |
342 (47.9) 265 (37.1) 107 (14.9) |
1029 (55.3) 569 (30.6) 260 (13.9) |
| Ethnicity | |||
| White Other Missing |
1086 (94.9) 41 (3.5) 17 (1.4) |
650 (91.0) 57 (7.9) 7 (0.9) |
1736 (93.4) 98 (5.2) 24 (1.2) |
| Age, y* | |||
| Mean (SD) Median (range) |
41.2 (13.0) 41.0 (18–85) |
42.4 (10.6) 42.0 (19–75) |
41.7 (12.2) 42 (18–85) |
| Year of birth | |||
| 1914–1943 1944–1954 1955–1965 1966–1990 |
206 (18.0) 262 (22.9) 281 (24.5) 395 (34.5) |
189 (26.4) 218 (30.5) 201 (28.1) 106 (14.8) |
395 (21.2) 480 (25.8) 482 (25.9) 501 (26.9) |
| Sex | |||
| Men Women |
430 (37.5) 714 (62.4) |
390 (54.6) 324 (45.3) |
820 (44.1) 1038 (55.8) |
| Education | |||
| Primary or less Some high school Completed high school/some tertiary study Vocational/technical school University degree Missing |
11 (0.9) 221 (19.3) 389 (34.0) 207 (18.0) 307 (26.8) 9 (0.7) |
18 (2.5) 156 (21.8) 229 (32.0) 119 (16.6) 183 (25.6) 9 (1.2) |
29 (1.5) 377 (20.2) 618 (33.2) 326 (17.5) 490 (26.3) 18 (0.9) |
| MMR gene mutated | |||
| MLH1
MSH2 MSH6 PMS2 |
373 (32.6) 575 (50.2) 142 (12.4) 54 (4.7) |
308 (43.1) 298 (41.7) 62 (8.6) 46 (6.4) |
681 (36.6) 873 (46.9) 204 (10.9) 100 (5.3) |
| BMI† two years before CRC diagnosis or censoring | |||
| Normal Overweight Obese Underweight Missing |
334 (29.2) 230 (20.1) 109 (9.5) 23 (2.0) 448 (39.1) |
98 (13.7) 101 (14.1) 58 (8.1) 10 (1.4) 447 (62.6) |
432 (23.2) 331 (17.8) 167 (8.9) 33 (1.8) 895 (48.1) |
| Diabetes | |||
| No Yes Missing |
1110 (97.0) 30 (2.6) 4 (0.3) |
679 (95.1) 30 (4.2) 5 (0.7) |
1789 (96.2) 60 (3.2) 9 (0.4) |
| Cigarette smoking‡ | |||
| Never Former Current Missing |
623 (54.4) 260 (22.7) 256 (22.3) 5 (0.43) |
325 (45.5) 152 (21.2) 237 (33.1) 0 (0) |
948 (51.0) 412 (22.1) 493 (26.5) 5 (0.27) |
| Number of alcohol drinks per day | |||
| None 1 ≥2 Missing |
289 (25.2) 376 (32.8) 329 (28.7) 150 (13.1) |
175 (24.5) 226 (31.6) 233 (32.6) 80 (11.2) |
464 (24.9) 602 (32.4) 562 (30.2) 230 (12.3) |
| Regular physical activity§ | |||
| <3 mos ≥3 mos Missing |
35 (3.0) 999 (87.3) 110 (9.61) |
4 (0.5) 631 (88.3) 79 (11.0) |
39 (2.1) 1630 (87.7) 189 (10.1) |
| Multivitamin intake‖ | |||
| <1 mo ≥1 mo Missing |
788 (68.8) 306 (26.7) 50 (4.3) |
547 (76.6) 134 (18.7) 33 (4.6) |
1335 (71.8) 440 (23.6) 83 (4.4) |
| Acetaminophen intake‖ | |||
| <1 mo ≥1 mo Missing |
1033 (90.3) 91 (7.9) 20 (1.7) |
656 (91.8) 40 (5.6) 18 (2.5) |
1689 (90.9) 131 (7.0) 38 (2.0) |
| Laxative intake‖ | |||
| <1 mo ≥1 mo Missing |
1079 (94.3) 53 (4.6) 12 (1.0) |
663 (92.8) 41 (5.7) 10 (1.4) |
1742 (93.7) 94 (5.0) 22 (1.2) |
| Hormone replacement therapy use¶ | |||
| <6 mos ≥6 mos Missing |
604 (84.5) 93 (13.0) 17 (2.3) |
277 (85.4) 41 (12.6) 6 (1.8) |
881 (84.8) 134 (12.9) 23 (2.2) |
| Number of live births¶ | |||
| None 1 2 ≥3 Missing |
209 (29.2) 73 (10.2) 199 (27.8) 215 (30.1) 18 (2.5) |
56 (17.2) 43 (13.2) 94 (29.0) 119 (36.7) 12 (3.7) |
265 (25.5) 116 (11.1) 293 (28.2) 334 (32.1) 30 (2.8) |
* Age at first diagnosis of colorectal cancer for affected carriers; age at first polypectomy or diagnosis of non-colorectal cancer or interview for unaffected carriers (whichever came first). BMI = body mass index; CRC = colorectal cancer; MMR = mismatch repair.
† Underweight (<18.5kg/m2), normal (18.5–24.9kg/m2), overweight (25.0–29.9kg/m2), obese (≥30kg/m2).
‡ Former smokers defined as carriers who had smoked at least one cigarette per day for at least three months and had quit more than two years before age at colorectal cancer or censored age; current smokers defined as carriers who had smoked at least one cigarette per day for at least three months and continued within two years of age at colorectal cancer or censored age.
§ Regular physical activity defined as any physical activity for at least 30 minutes per week for at least three months.
‖ At least twice a week.
¶ Only for female carriers.
Table 2.
Prevalence of aspirin and ibuprofen intake in mismatch repair gene mutation carriers by country, sex, and age category*
| NSAID use | Country/age group, y | Men No. (%) |
Women No. (%) |
Overall No. (%) |
|---|---|---|---|---|
| Aspirin only | USA | |||
| ≤35 | 4 (9.3) | 1 (3.0) | 5 (6.5) | |
| 36–45 | 5 (10.8) | 4 (7.1) | 9 (8.8) | |
| 46–55 | 7 (9.2) | 5 (6.4) | 12 (7.8) | |
| >55 | 20 (20.6) | 12 (11.5) | 32 (15.9) | |
| All | 36 (13.7) | 22 (8.1) | 58 (10.9) | |
| Australia/New Zealand | ||||
| ≤35 | 3 (2.6) | 6 (4.1) | 9 (3.4) | |
| 36–45 | 5 (6.0) | 5 (5.2) | 10 (5.5) | |
| 46–55 | 4 (3.6) | 3 (2.4) | 7 (3.0) | |
| >55 | 7 (6.4) | 16 (8.8) | 23 (7.9) | |
| All | 19 (4.5) | 30 (5.5) | 49 (5.1) | |
| Canada | ||||
| ≤35 | 1 (5.2) | 1 (3.7) | 2 (4.3) | |
| 36–45 | 2 (10.0) | 0 (0) | 2 (4.8) | |
| 46–55 | 0 (0) | 1 (2.5) | 1 (1.5) | |
| >55 | 2 (5.0) | 3 (4.8) | 5 (4.9) | |
| All | 5 (4.8) | 5 (3.3) | 10 (3.9) | |
| Overall | 60 (7.6) | 57 (5.9) | 117 (6.7) | |
| Ibuprofen only | USA | |||
| ≤35 | 4 (9.0) | 6 (15.3) | 10 (12.0) | |
| 36–45 | 2 (4.7) | 5 (8.3) | 7 (6.8) | |
| 46–55 | 7 (9.5) | 7 (8.6) | 14 (9.0) | |
| >55 | 3 (3.9) | 3 (3.2) | 6 (3.5) | |
| All | 16 (6.8) | 21 (7.7) | 37 (7.3) | |
| Australia/New Zealand | ||||
| ≤35 | 6 (5.1) | 11 (7.4) | 17 (6.4) | |
| 36–45 | 7 (8.3) | 13 (12.5) | 20 (10.7) | |
| 46–55 | 8 (7.1) | 14 (10.6) | 22 (9.0) | |
| >55 | 9 (8.0) | 11 (6.3) | 20 (6.9) | |
| All | 30 (7.0) | 49 (8.8) | 79 (8.0) | |
| Canada | ||||
| ≤35 | 0 (0) | 1 (3.7) | 1 (2.2) | |
| 36–45 | 1 (5.0) | 2 (8.0) | 3 (6.6) | |
| 46–55 | 2 (7.4) | 2 (4.1) | 4 (5.3) | |
| >55 | 0 (0) | 2 (3.1) | 2 (2.0) | |
| All | 3 (3.0) | 7 (4.2) | 10 (3.8) | |
| Overall | 49 (6.4) | 77 (7.7) | 126 (7.2) | |
| Aspirin and/or ibuprofen | USA | |||
| ≤35 | 10 (19.2) | 10 (22.7) | 20 (20.8) | |
| 36–45 | 10 (19.2) | 12 (17.9) | 22 (18.4) | |
| 46–55 | 16 (18.3) | 16 (15.9) | 32 (16.9) | |
| >55 | 27 (25.2) | 22 (17.4) | 49 (21.0) | |
| All | 63 (21.1) | 60 (17.7) | 123 (19.3) | |
| Australia/New Zealand | ||||
| ≤35 | 10 (8.2) | 19 (12.0) | 29 (10.3) | |
| 36–45 | 12 (13.3) | 18 (16.3) | 30 (15.0) | |
| 46–55 | 13 (10.8) | 18 (12.9) | 31 (11.9) | |
| >55 | 18 (14.5) | 31 (15.3) | 49 (15.0) | |
| All | 53 (11.6) | 86 (14.1) | 139 (13.0) | |
| Canada | ||||
| ≤35 | 1 (5.2) | 2 (6.9) | 3 (6.2) | |
| 36–45 | 3 (13.6) | 2 (7.6) | 5 (10.4) | |
| 46–55 | 2 (6.6) | 4 (7.8) | 6 (7.4) | |
| >55 | 2 (4.7) | 8 (11.1) | 10 (8.7) | |
| All | 8 (7.0) | 16 (8.9) | 24 (8.2) | |
| Overall | 124 (14.3) | 162 (14.3) | 286 (14.3) | |
* NSAID = nonsteroidal anti-inflammatory drugs.
Table 3 presents the results of the univariate and multivariable Cox regression models and the confounders included in each model. A reduced risk of colorectal cancer was associated with ever use of: aspirin only (HR = 0.43, 95% CI = 0.25 to 0.75, P = .003), ibuprofen only (HR = 0.35, 95% CI = 0.19 to 0.63, P = .001), and aspirin and/or ibuprofen (HR = 0.41, 95% CI = 0.28 to 0.61, P < .001). A longer duration of use was associated with a lower risk of colorectal cancer; aspirin-only use (for 1 month to 4.9 years: HR = 0.49, 95% CI = 0.27 to 0.90, P = .02; for ≥5 years: HR = 0.25, 95% CI = 0.10 to 0.62, P = .003) and ibuprofen-only use (for 1 month to 4.9 years: HR = 0.38, 95% CI = 0.18 to 0.79, P = .009; for ≥5 years: HR = 0.26, 95% CI = 0.10 to 0.69, P = .007) (Table 3). These inverse associations are illustrated in Figures 1 and 2 for aspirin and ibuprofen, respectively.
Table 3.
Univariate and multivariable hazard ratios for associations between aspirin and ibuprofen use and the risk of colorectal cancer for mismatch repair gene mutation carriers
| NSAID use | Univariate model | Multivariable model | |||||
|---|---|---|---|---|---|---|---|
| n | N | Person-years | HR (95% CI) |
P | HR* (95% CI) |
P | |
| Never user† | 622 | 1572 | 25 743 | 1 (Ref) | 1 (Ref) | ||
| Aspirin-only user | 48 | 117 | 2505 | 0.48 (0.33 to 0.71) |
<.001 | 0.43 (0.25 to 0.75) |
.003 |
| 1 month to 4.9 years | 38 | 96 | 1904 | 0.49 (0.31 to 0.76) |
.002 | 0.49 (0.27 to 0.90) |
.02 |
| ≥5 years | 10 | 21 | 601 | 0.45 (0.21 to 0.92) |
.03 | 0.25 (0.10 to 0.62) |
.003 |
| Ibuprofen-only user | 30 | 126 | 1373 | 0.33 (0.20 to 0.55) |
<.001 | 0.35 (0.19 to 0.63) |
.001 |
| 1 month to 4.9 years | 22 | 92 | 1010 | 0.36 (0.20 to 0.66) |
.001 | 0.38 (0.18 to 0.79) |
.009 |
| ≥5 years | 8 | 34 | 363 | 0.25 (0.11 to 0.58) |
.001 | 0.26 (0.10 to 0.69) |
.007 |
| Aspirin and/or ibuprofen user | 92 | 286 | 4548 | 0.40 (0.30 to 0.54) |
<.001 | 0.41 (0.28 to 0.61) |
<.001 |
| 1 month to 4.9 years | 66 | 210 | 3196 | 0.42 (0.29 to 0.59) |
<.001 | 0.44 (0.27 to 0.69) |
<.001 |
| ≥5 years | 26 | 76 | 1352 | 0.36 (0.22 to 0.57) |
<.001 | 0.34 (0.19 to 0.62) |
<.001 |
* Adjusted for year of birth (1914–43, 1944–54, 1955–65, 1966–90) and average lifetime alcohol intake (0, 1, ≥2 drinks per day) and stratified by sex, country (USA, Australia/ New Zealand, Canada), cigarette smoking status (never, former, current), regular physical activity (at least 30 minutes per week for at least three months), and multivitamin intake (<1 month, ≥1 month). All statistical tests were two-sided. CI = confidence interval; HR = hazard ratio; n = number of colorectal cancer cases; N = total number of carriers; NSAID = nonsteroidal anti-inflammatory drugs.
† Never users defined as carriers who reported not having taken either aspirin or ibuprofen or both for at least one month.
Figure 1.
Cox regression analysis for association between aspirin only use and the risk of colorectal cancer over time for mismatch repair gene mutation carriers. Adjusted for year of birth (1914–43, 1944–54, 1955–65, 1966–90) and average lifetime alcohol intake (0, 1, ≥2 drinks per day) and stratified by sex, country (USA, Australia/ New Zealand, Canada), cigarette smoking status (never, former, current), regular physical activity (at least 30 minutes per week for at least three months), and multivitamin intake (<1 month, ≥1 month). Upper and lower dashed lines represent upper and lower limit of 95% confidence interval, respectively. Solid line represents the hazard ratio.
Figure 2.
Cox regression analysis for association between ibuprofen only use and the risk of colorectal cancer over time for mismatch repair gene mutation carriers. Adjusted for year of birth (1914–43, 1944–54, 1955–65, 1966–90) and average lifetime alcohol intake (0, 1, ≥2 drinks per day) and stratified by sex, country (USA, Australia/ New Zealand, Canada), cigarette smoking status (never, former, current), regular physical activity (at least 30 minutes per week for at least three months), and multivitamin intake (<1 month, ≥1 month). Upper and lower dashed lines represent upper and lower limit of 95% confidence interval, respectively. Solid line represents the hazard ratio.
Table 4 presents the results of the multivariable analyses separately for men and women. The inverse associations with aspirin only, ibuprofen only, and aspirin and/or ibuprofen were similar to those from the combined analysis, and there was no evidence that they differed by sex (P > .7 for all interactions). Table 5 presents the results of the multivariable analyses separately for mutation carriers by the specific MMR gene mutated. There was no evidence that the strengths of association differed by gene (P > .2 for all interactions).
Table 4.
Sex-specific multivariable hazard ratios for associations between aspirin and ibuprofen use and the risk of colorectal cancer for mismatch repair gene mutation carriers
| NSAID use | Men | Women | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| n | N | Person-years | HR* (95% CI) |
P | n | N | Person-years | HR† (95% CI) |
P | |
| Never user‡ | 338 | 696 | 14 137 | 1 (Ref) | 284 | 876 | 11 606 | 1 (Ref) | ||
| Aspirin-only user | 29 | 60 | 1501 | 0.43 (0.21 to 0.86) |
.02 | 19 | 57 | 1004 | 0.42 (0.17 to 1.03) |
.06 |
| 1 month to 4.9 years | 23 | 50 | 1128 | 0.45 (0.21 to 0.97) |
.04 | 15 | 46 | 776 | 0.54 (0.21 to 1.41) |
.21 |
| ≥5 years | 6 | 10 | 373 | 0.32 (0.10 to 0.96) |
.04 | 4 | 11 | 228 | 0.15 (0.01 to 1.32) |
.09 |
| Ibuprofen-only user | 15 | 49 | 693 | 30 (0.14 to 0.67) |
.004 | 15 | 77 | 680 | 0.35 (0.14 to 0.86) |
.02 |
| 1 month to 4.9 years | 11 | 36 | 517 | 0.38 (0.14 to 1.01) |
.05 | 11 | 56 | 493 | 0.36 (0.12 to 1.09) |
.07 |
| ≥5 years | 4 | 13 | 176 | 0.21 (0.05 to 0.86) |
.03 | 4 | 21 | 187 | 0.33 (0.09 to 1.15) |
.08 |
| Aspirin and/or ibuprofen user | 52 | 124 | 2580 | 0.41 (0.24 to 0.68) |
.001 | 40 | 162 | 1968 | 0.42 (0.23 to 0.78) |
.006 |
| 1 month to 4.9 years | 38 | 94 | 1808 | 0.42 (0.24 to 0.74) |
.003 | 28 | 116 | 1388 | 0.39 (0.18 to 0.84) |
.02 |
| ≥5 years | 14 | 30 | 772 | 0.20 (0.06 to 0.62) |
.005 | 12 | 46 | 580 | 0.52 (0.21 to 1.28) |
.16 |
* Adjusted for year of birth (1914–43, 1944–54, 1955–65, 1966–90) and average lifetime alcohol intake (0, 1, ≥2 drinks per day) and stratified by country (USA, Australia/ New Zealand, Canada), cigarette smoking status (never, former, current), regular physical activity (at least 30 minutes per week for at least three months), and multivitamin intake (<1 month, ≥1 month). All statistical tests were two-sided. CI = confidence interval; HR = hazard ratio; n = number of colorectal cancer cases; N = total number of carriers; NSAID = nonsteroidal anti-inflammatory drugs.
† Adjusted for year of birth (1914–43, 1944–54, 1955–65, 1966–90) and number of live births (none, 1, 2, 3, or more) and stratified by country (USA, Australia/ New Zealand, Canada), cigarette smoking status (never, former, current), regular physical activity (at least 30 minutes per week for at least three months), hormone replacement therapy use (<6 months, ≥6 months), and multivitamin intake (<1 month, ≥1 month).
‡ Never users defined as carriers who reported not having taken either aspirin or ibuprofen or both for at least one month.
Table 5.
Multivariable hazard ratios for associations between aspirin and ibuprofen use and the risk of colorectal cancer by mismatch repair gene mutated
| MMR gene/NSAID use | n | N | Person-years | HR* (95% CI) |
P |
|---|---|---|---|---|---|
| MLH1 | |||||
| Never user† | 276 | 600 | 11 043 | 1 (Ref) | |
| Aspirin-only user | 13 | 30 | 609 | 0.34 (0.13 to 0.86) |
.02 |
| Ibuprofen-only user | 13 | 40 | 603 | 0.91 (0.32 to 2.55) |
.86 |
| Aspirin and/or ibuprofen user | 32 | 81 | 1499 | 0.43 (0.21 to 0.87) |
.02 |
| MSH2 | |||||
| Never user† | 266 | 751 | 10 975 | 1 (Ref) | |
| Aspirin-only user | 18 | 51 | 887 | 0.52 (0.19 to 1.45) |
.22 |
| Ibuprofen-only user | 8 | 53 | 344 | 0.27 (0.08 to 0.85) |
.03 |
| Aspirin and/or ibuprofen user | 32 | 122 | 1505 | 0.39 (0.20 to 0.74) |
.004 |
| MSH6 | |||||
| Never user† | 50 | 148 | 2330 | 1 (Ref) | |
| Aspirin-only user | 6 | 21 | 352 | 0.23 (0.04 to 1.06) |
.06 |
| Ibuprofen-only user | 5 | 24 | 242 | 0.16 (0.03 to 0.72) |
.02 |
| Aspirin and/or ibuprofen user | 12 | 56 | 643 | 0.27 (0.04 to 1.75) |
.17 |
| PMS2 | |||||
| Never user† | 30 | 73 | 1395 | 1 (Ref) | |
| Aspirin-only user | 11 | 15 | 657 | 1.78 (0.56 to 5.67) |
.33 |
| Ibuprofen-only user | 4 | 9 | 184 | 1.60 (0.33 to 7.59) |
.55 |
| Aspirin and/or ibuprofen user | 16 | 27 | 901 | 0.70 (0.33 to 1.47) |
.35 |
* Adjusted for sex, country (USA, Australia/ New Zealand, Canada), average lifetime alcohol intake (0, 1, ≥2 drinks per day), year of birth (1914–43, 1944–54, 1955–65, 1966–90), cigarette smoking status (never, former, current), regular physical activity (at least 30 minutes per week for at least three months), and multivitamin intake (<1 month, ≥1 month), where possible. All statistical tests were two-sided. CI = confidence interval; HR = hazard ratio; n = number of colorectal cancer cases; MMR = DNA mismatch repair; N = total number of carriers; NSAID = nonsteroidal anti-inflammatory drugs.
† Never users defined as carriers who reported not having taken either aspirin or ibuprofen or both for at least one month.
In the analysis restricted to carriers diagnosed with colorectal cancer or censored within five years prior to their interview (n = 1213, 65%), the results were similar to those from the main analysis (Supplementary Table 3, available online). Similar results were also observed when we used colorectal neoplasia (either polyp or colorectal cancer) as the outcome (Supplementary Table 4, available online). Finally, in the analysis restricted to carriers with complete covariate data, the results were similar to the main analysis (data not shown). All models were tested and there was no evidence that the PH assumption was violated.
Discussion
This study includes the largest sample of confirmed MMR gene mutation carriers assembled to date to investigate the association between aspirin use and colorectal cancer risk in the context of Lynch Syndrome. We estimated that, for carriers of pathogenic germline mutations in MMR genes, aspirin use for at least one month was associated with a decrease of approximately 60% in the risk of colorectal cancer. A recent meta-analysis of 10 cohort studies investigating the association between duration of aspirin use and the risk of colorectal cancer for the general population estimated 10% (95% CI = 8% to 12%) and 18% (95% CI = 14% to 22%) reductions in colorectal cancer risk for five and 10 years’ use, respectively (38). In our analysis, the magnitude of the association for MMR gene mutation carriers was substantially stronger than that observed for the general population, particularly for aspirin use for more than five years (P = .002 for difference between estimates based on the test of interaction proposed by Altman and Bland [39]). Our findings are more consistent with the risk reduction found by Burn et al. in the CAPP2 trial, in which MMR gene mutation carriers had a 58% (95% CI = 14% to 81%) reduction in colorectal cancer risk from use of 600mg of aspirin daily for an average period of two years (this result was detected after a mean follow-up period of 4.6 years) (20). Our finding that the associations were similar when colorectal neoplasia was the outcome is consistent with use of NSAIDs delaying initiation of adenomas. This finding contrasts with the CAPP2 study, which did not find evidence for a short-term (two-year) reduction in the risk of adenoma (19). However, it is consistent with several RCTs of people with familial adenomatous polyposis (FAP), which found evidence that use of NSAIDs reduces the size and the number of colorectal polyps (40–44). It should be noted, however, that FAP and Lynch Syndrome are very different clinically and genetically.
Our analysis is the first reported investigation of a potential association between ibuprofen and the risk of colorectal cancer in the context of Lynch Syndrome. Use of ibuprofen only for at least one month was associated with a 65% decrease in the risk of colorectal cancer. This finding is consistent with previous evidence of an inverse association between nonaspirin NSAIDs, particularly those of the propionic acid derivatives class such as ibuprofen, and the risk of colorectal cancer for the general population. A case-control study by Reeves and colleagues has shown that ibuprofen might have a greater beneficial effect against colorectal cancer than aspirin (57% vs 21% risk reduction) (45).
The potential chemopreventive effect of NSAIDs on colorectal cancer has been recognized since the late 1980s (46) and is strongly supported by experimental data (47). The mechanism of action that defines NSAIDs as a class is their ability to inhibit the cyclooxygenase (COX) activity of the enzyme prostaglandin G/H-synthase, which is expressed as two distinct isoforms, designated COX-1 and COX-2 (48). However, the exact mechanism by which aspirin reduces colorectal cancer risk is yet to be understood. Most colorectal cancers progress through the action of multiple molecular mechanisms that involve cyclooxygenases, Wnt–β-catenin, MAP kinase, cytokines, and growth factors (14). The effect of NSAIDs on colorectal neoplasia could be mediated by their ability to restore apoptosis through the inhibition of the COX pathway (49,50). In addition to the COX-inhibition pathway, several other possible mechanisms have been described (51,52).
Recently, Greenspan and colleagues investigated the effect of ibuprofen on biochemical pathways in both human adenomas and colorectal cancers in vitro and showed that ibuprofen is effective in inhibiting Wnt–β-catenin nuclear translocation in both human adenomas and in colorectal cancer cells. This study confirmed that a NSAID can suppress β-catenin, a key mediator of colon tumorigenesis, in sporadic adenomas following long-term treatment in patients (53). Overall, it should be noted that it is currently unknown whether the biological processes that have been discovered so far and that potentially underlie the association between NSAIDs and sporadic colorectal cancers also apply to Lynch Syndrome–associated colorectal cancers.
One of the strengths of our study is that we were able to adjust for recognized potential confounding variables, including alcohol consumption, cigarette smoking, hormone replacement therapy, and multivitamin use. The Colon Cancer Family Registry had established standardized and uniform protocols for collection of epidemiologic data, family history, and cancer history. It had also undertaken comprehensive mutation screening (21). To minimize bias because of patients having been selected on the basis of phenotype, we used a weighted cohort approach (26).
A limitation of our study is the inability to assess the gradient in risk with specific doses of aspirin and ibuprofen, as this information was not collected by the questionnaires. Long-term follow-up of a current trial could help resolve this issue and determine optimal aspirin treatment regimen. No studies investigating the dose-response effect of ibuprofen on colorectal cancer risk have been published. There have been concerns about the potential side effects of long-term aspirin use, particularly regarding the risk of hemorrhagic strokes and gastrointestinal bleeding beyond age 70 years (54,55). NSAIDs (perhaps particularly nonaspirin NSAIDs) are also associated with side effects, including what is called diaphragm disease—a non-neoplastic overgrowth in the small intestine (56,57). More research is needed for a better understanding of the determinants of these side effects (58). Recently, a comprehensive review and synthesis of the available evidence on the benefits and harms has found a strong net benefit for the prophylactic use of aspirin to reduce the incidence of colorectal and other cancers, and mortality, in the general population (59). However, recent results from the Women’s Health Study have shown that the potential benefit of aspirin use in relation to cancer outcomes might only apply after age 64 years (60). In the context of Lynch Syndrome, where absolute cancer risk is high, use of aspirin as a chemoprevention strategy for colorectal cancer is likely to present a more favorable and consistent benefit-harm profile than in the general population.
Another limitation of this study is the possibility that the findings may have been affected by recall or response bias, given that data on the intake of aspirin and ibuprofen use and other variables were collected at the time of recruitment, and therefore, for those affected, after their diagnosis of colorectal cancer. Differential recall on exposures between participants diagnosed with colorectal cancer (as well as other cancers or polypectomy) and those who were censored at the time of interview is also a potential source of bias in our study. Similarly, our definition of duration of aspirin/ibuprofen use was based on the assumption that the use was continuous and recent. However, participants might have used aspirin/ibuprofen intermittently and we were unable to account for this variation in our analysis. Further, study participants might have used NSAIDs other than aspirin and ibuprofen that were not included in the Colon Cancer Family Registry questionnaire. It is therefore possible that some participants were misclassified as nonusers in the analysis. Finally, as the carriers with poorer survival were less likely to be included in this study (as they were less likely to be alive to be recruited and give a blood sample), there is the possibility of survival and selection bias if age at onset and survival of case patients were related to exposure variables and/or mutation status. However, our sensitivity analysis restricted to carriers who were diagnosed with colorectal cancer or censored within five years prior to the interview showed findings similar to the main analysis.
In summary, this study provides additional observational evidence that aspirin use is associated with a reduction in risk of colorectal cancer for MMR gene mutations carriers. Our results suggest that regular long-term use of aspirin or ibuprofen might be an effective way to reduce colorectal cancer risk for this group of highly susceptible people who currently rely only on frequent colonoscopies to reduce their risk.
Funding
This work was supported by grant UM1 CA167551 from the National Cancer Institute, National Institutes of Health (NIH), and through cooperative agreements with members of the Colon Cancer Family Registry and Principal Investigators. Collaborating centers include Australasian Colorectal Cancer Family Registry (U01/U24 CA097735), Mayo Clinic Cooperative Family Registry for Colon Cancer Studies (U01/U24 CA074800), Ontario Familial Colorectal Cancer Registry (U01/U24 CA074783), Seattle Colorectal Cancer Family Registry (U01/U24 CA074794), Stanford Consortium Colorectal Cancer Family Registry (U01/U24 CA074799), and University of Hawaii Colorectal Cancer Family Registry (U01/U24 CA074806). This work was also supported by Centre for Research Excellence grant APP1042021 and program grant APP1074383 from the National Health and Medical Research Council (NHMRC), Australia. AKW is a NHMRC Early Career Fellow. MAJ is an NHMRC Senior Research Fellow. JLH is a NHMRC Senior Principal Research Fellow. DDB is a University of Melbourne Research at Melbourne Accelerator Program (R@MAP) Senior Research Fellow. CR is the Jeremy Jass Pathology Fellow.
Supplementary Material
The content of this manuscript does not necessarily reflect the views or policies of the National Cancer Institute or any of the collaborating centers in the Colon Cancer Family Registry, nor does mention of trade names, commercial products, or organizations imply endorsement by the US Government or the CFR. Authors had full responsibility for the design of the study, the collection of the data, the analysis and interpretation of the data, the decision to submit the manuscript for publication, and the writing of the manuscript.
The authors have no conflict of interest to declare with respect to this manuscript.
The authors thank all study participants of the Colon Cancer Family Registry and staff for their contributions to this project.
References
- 1. Lynch HT, Lynch PM, Lanspa SJ, et al. Review of the Lynch syndrome: history, molecular genetics, screening, differential diagnosis, and medicolegal ramifications. Clin Genet. 2009;76 (1):1–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Dunlop MG, Farrington SM, Nicholl I, et al. Population carrier frequency of hMSH2 and hMLH1 mutations. Br J Cancer. 2000;83 (12):1643–1645. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Hampel H, de la Chapelle A. The search for unaffected individuals with Lynch syndrome: do the ends justify the means? Cancer Prev Res (Phila). 2011;4 (1):1–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Aaltonen LA, Sankila R, Mecklin JP, et al. A novel approach to estimate the proportion of hereditary nonpolyposis colorectal cancer of total colorectal cancer burden. Cancer Detect Prev. 1994;18 (1):57–63. [PubMed] [Google Scholar]
- 5. Hopper JL. Application of genetics to the prevention of colorectal cancer. Recent results in cancer research 2005;Fortschritte der Krebsforschung. Progres dans les recherches sur le cancer. 166:17–33. [DOI] [PubMed] [Google Scholar]
- 6. Dowty JG, Win AK, Buchanan DD, et al. Cancer risks for MLH1 and MSH2 mutation carriers. Hum Mutat. 2013;34 (3):490–497. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Jarvinen HJ, Aarnio M, Mustonen H, et al. Controlled 15-year trial on screening for colorectal cancer in families with hereditary nonpolyposis colorectal cancer. Gastroenterology. 2000;118 (5):829–834. [DOI] [PubMed] [Google Scholar]
- 8. Cairns SR, Scholefield JH, Steele RJ, et al. Guidelines for colorectal cancer screening and surveillance in moderate and high risk groups (update from 2002). Gut. 2010;59 (5):666–689. [DOI] [PubMed] [Google Scholar]
- 9. Levin B, Lieberman DA, McFarland B, et al. Screening and surveillance for the early detection of colorectal cancer and adenomatous polyps, 2008: a joint guideline from the American Cancer Society, the US Multi-Society Task Force on Colorectal Cancer, and the American College of Radiology. CA Cancer J Clin. 2008;58 (3):130–160. [DOI] [PubMed] [Google Scholar]
- 10. NHMRC. Clinical practice guidelines for the prevention, early detection and management of colorectal cancer. The Australian Cancer Network website. http://www.cancer.org.au/guidelines or http://www.nhmrc.gov.au/publications/synopses/cp106/cp106syn.htm.
- 11. Vasen HF, Blanco I, Aktan-Collan K, et al. Revised guidelines for the clinical management of Lynch syndrome (HNPCC): recommendations by a group of European experts. Gut. 2013;62 (6):812–823. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Lindor NM, Petersen GM, Hadley DW, et al. Recommendations for the care of individuals with an inherited predisposition to Lynch syndrome: A systematic review. JAMA. 2006;296 (12):1507–1517. [DOI] [PubMed] [Google Scholar]
- 13. Chan AT, Giovannucci EL, Meyerhardt JA, et al. Aspirin dose and duration of use and risk of colorectal cancer in men. Gastroenterology. 2008;134 (1):21–28. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Cuzick J, Otto F, Baron JA, et al. Aspirin and non-steroidal anti-inflammatory drugs for cancer prevention: an international consensus statement. Lancet Oncol. 2009;10 (5):501–507. [DOI] [PubMed] [Google Scholar]
- 15. Flossmann E, Rothwell PM Effect of aspirin on long-term risk of colorectal cancer: consistent evidence from randomised and observational studies. Lancet. 2007;369 (9573):1603–1613. [DOI] [PubMed] [Google Scholar]
- 16. Rothwell PM, Wilson M, Elwin CE, et al. Long-term effect of aspirin on colorectal cancer incidence and mortality: 20-year follow-up of five randomised trials. Lancet. 2010;376 (9754):1741–1750. [DOI] [PubMed] [Google Scholar]
- 17. Cole BF, Logan RF, Halabi S, et al. Aspirin for the chemoprevention of colorectal adenomas: meta-analysis of the randomized trials. J Natl Cancer Inst. 2009;101 (4):256–266. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Ishikawa H, Mutoh M, Suzuki S, et al. The preventive effects of low-dose enteric-coated aspirin tablets on the development of colorectal tumours in Asian patients: a randomised trial. Gut. 2014;63 (11):1755–1759. [DOI] [PubMed] [Google Scholar]
- 19. Burn J, Bishop DT, Mecklin JP, et al. Effect of aspirin or resistant starch on colorectal neoplasia in the Lynch syndrome. N Engl J Med. 2008;359 (24):2567–2578. [DOI] [PubMed] [Google Scholar]
- 20. Burn J, Gerdes AM, Macrae F, et al. Long-term effect of aspirin on cancer risk in carriers of hereditary colorectal cancer: an analysis from the CAPP2 randomised controlled trial. Lancet. 2011;378 (9809):2081–2087. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Newcomb PA, Baron J, Cotterchio M, et al. Colon Cancer Family Registry: an international resource for studies of the genetic epidemiology of colon cancer. Cancer Epidemiol Biomarkers Prev. 2007;16 (11):2331–2343. [DOI] [PubMed] [Google Scholar]
- 22. Rumilla K, Schowalter KV, Lindor NM, et al. Frequency of deletions of EPCAM (TACSTD1) in MSH2-associated Lynch syndrome cases. J Mol Diagn. 2011;13 (1):93–99. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Southey MC, Jenkins MA, Mead L, et al. Use of molecular tumor characteristics to prioritize mismatch repair gene testing in early-onset colorectal cancer. J Clin Oncol. 2005;23 (27):6524–6532. [DOI] [PubMed] [Google Scholar]
- 24. Senter L, Clendenning M, Sotamaa K, et al. The clinical phenotype of Lynch syndrome due to germ-line PMS2 mutations. Gastroenterology. 2008;135 (2):419–428. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Win AK, Young JP, Lindor NM, et al. Colorectal and other cancer risks for carriers and noncarriers from families with a DNA mismatch repair gene mutation: a prospective cohort study. J Clin Oncol. 2012;30 (9):958–964. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Antoniou AC, Goldgar DE, Andrieu N, et al. A weighted cohort approach for analysing factors modifying disease risks in carriers of high-risk susceptibility genes. Genet Epidemiol. 2005;29 (1):1–11. [DOI] [PubMed] [Google Scholar]
- 27. Andrieu N, Easton DF, Chang-Claude J, et al. Effect of chest X-rays on the risk of breast cancer among BRCA1/2 mutation carriers in the international BRCA1/2 carrier cohort study: a report from the EMBRACE, GENEPSO, GEO-HEBON, and IBCCS Collaborators’ Group. J Clin Oncol. 2006;24 (21):3361–3366. [DOI] [PubMed] [Google Scholar]
- 28. Antoniou AC, Spurdle AB, Sinilnikova OM, et al. Common breast cancer-predisposition alleles are associated with breast cancer risk in BRCA1 and BRCA2 mutation carriers. Am J Hum Genet. 2008;82 (4):937–948. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Win AK, Dowty JG, Antill YC, et al. Body mass index in early adulthood and endometrial cancer risk for mismatch repair gene mutation carriers. Obstet Gynecol. 2011;117 (4):899–905. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Pande M, Lynch PM, Hopper JL, et al. Smoking and Colorectal Cancer in Lynch Syndrome: Results from the Colon Cancer Family Registry and The University of Texas M.D. Anderson Cancer Center. Clin Cancer Res. 2010;16 (4):1331–1339. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. American Cancer Society. Cancer Facts & Figures 2013. In. Atlanta: American Cancer Society; 2013. [Google Scholar]
- 32. Jenkins MA, Baglietto L, Dowty JG, et al. Cancer Risks For Mismatch Repair Gene Mutation Carriers: A Population-Based Early Onset Case-Family Study. Clin Gastroenterol Hepatol. 2006;4 (4):489–498. [DOI] [PubMed] [Google Scholar]
- 33. Grambsch PM, Therneau TM. Proportional hazards tests and diagnostics based on weighted residuals. Biometrika. 1994;81 (3):515–526. [Google Scholar]
- 34. Hosmer D, Lemeshow S, May S. Applied Survival Analysis. Regression Modeling of Time to Event Data. New York, NY: John Wiley & Sons; 1999. [Google Scholar]
- 35. Rogers W. Regression standard errors in clustered samples. Stata Technical Bulletin. 1999;3:19–23. [Google Scholar]
- 36. Williams RL. A note on robust variance estimation for cluster-correlated data. Biometrics. 2000;56 (2):645–646. [DOI] [PubMed] [Google Scholar]
- 37. Rebbeck TR, Mitra N, Domchek SM, et al. Modification of BRCA1-Associated Breast and Ovarian Cancer Risk by BRCA1-Interacting Genes. Cancer Res. 2011;71 (17):5792–5805. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Ye X, Fu J, Yang Y, et al. Dose-risk and duration-risk relationships between aspirin and colorectal cancer: a meta-analysis of published cohort studies. PLoS One. 2013;8 (2):e57578. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Altman DG, Bland JM. Interaction revisited: the difference between two estimates. BMJ. 2003;326 (7382):219. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Burn J, Bishop DT, Chapman PD, et al. A randomized placebo-controlled prevention trial of aspirin and/or resistant starch in young people with familial adenomatous polyposis. Cancer Prev Res (Phila). 2011;4 (5):655–665. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Giardiello FM, Hamilton SR, Krush AJ, et al. Treatment of colonic and rectal adenomas with sulindac in familial adenomatous polyposis. N Engl J Med. 1993;328 (18):1313–1316. [DOI] [PubMed] [Google Scholar]
- 42. Hallak A, Alon-Baron L, Shamir R, et al. Rofecoxib reduces polyp recurrence in familial polyposis. Dig Dis Sci. 2003;48 (10):1998–2002. [DOI] [PubMed] [Google Scholar]
- 43. Steinbach G, Lynch PM, Phillips RK, et al. The effect of celecoxib, a cyclooxygenase-2 inhibitor, in familial adenomatous polyposis. N Engl J Med. 2000;342 (26):1946–1952. [DOI] [PubMed] [Google Scholar]
- 44. Winde G, Schmid KW, Schlegel W, et al. Complete reversion and prevention of rectal adenomas in colectomized patients with familial adenomatous polyposis by rectal low-dose sulindac maintenance treatment. Advantages of a low-dose nonsteroidal anti-inflammatory drug regimen in reversing adenomas exceeding 33 months. Dis Colon Rectum. 1995;38 (8):813–830. [DOI] [PubMed] [Google Scholar]
- 45. Reeves MJ, Newcomb PA, Trentham-Dietz A, et al. Nonsteroidal anti-inflammatory drug use and protection against colorectal cancer in women. Cancer Epidemiol Biomarkers Prev. 1996;5 (12):955–960. [PubMed] [Google Scholar]
- 46. Kune GA, Kune S, Watson LF. Colorectal cancer risk, chronic illnesses, operations, and medications: case control results from the Melbourne Colorectal Cancer Study. Cancer Res. 1988;48 (15):4399–4404. [PubMed] [Google Scholar]
- 47. Thun MJ, Henley SJ, Patrono C. Nonsteroidal anti-inflammatory drugs as anticancer agents: mechanistic, pharmacologic, and clinical issues. J Natl Cancer Inst. 2002;94 (4):252–266. [DOI] [PubMed] [Google Scholar]
- 48. Chan AT, Ogino S, Fuchs CS. Aspirin and the risk of colorectal cancer in relation to the expression of COX-2. N Engl J Med. 2007;356 (21):2131–2142. [DOI] [PubMed] [Google Scholar]
- 49. He TC, Chan TA, Vogelstein B, et al. PPARdelta is an APC-regulated target of nonsteroidal anti-inflammatory drugs. Cell. 1999;99 (3):335–345. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50. Tsujii M, DuBois RN. Alterations in cellular adhesion and apoptosis in epithelial cells overexpressing prostaglandin endoperoxide synthase 2. Cell. 1995;83 (3):493–501. [DOI] [PubMed] [Google Scholar]
- 51. Chan TA, Morin PJ, Vogelstein B, et al. Mechanisms underlying nonsteroidal antiinflammatory drug-mediated apoptosis. Proc Natl Acad Sci U S A. 1998;95 (2):681–686. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52. Marx J. Cancer research. Anti-inflammatories inhibit cancer growth--but how? Science. 2001;291 (5504):581–582. [DOI] [PubMed] [Google Scholar]
- 53. Greenspan EJ, Madigan JP, Boardman LA, et al. Ibuprofen inhibits activation of nuclear {beta}-catenin in human colon adenomas and induces the phosphorylation of GSK-3{beta}. Cancer Prev Res (Phila). 2011;4 (1):161– 171. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54. Jankowski J, Barr H, Moayyedi P. Does aspirin really reduce the risk of colon cancer? Lancet. 2012;379 (9826):1586–1587; author reply 1587. [DOI] [PubMed] [Google Scholar]
- 55. Jankowski J, Barr H, Wang K, et al. Diagnosis and management of Barrett’s oesophagus. BMJ. 2010;341:c4551. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 56. Bjarnason I, Price AB, Zanelli G, et al. Clinicopathological features of nonsteroidal antiinflammatory drug-induced small intestinal strictures. Gastroenterology. 1988;94 (4):1070–1074. [DOI] [PubMed] [Google Scholar]
- 57. Ishihara M, Ohmiya N, Nakamura M, et al. Risk factors of symptomatic NSAID-induced small intestinal injury and diaphragm disease. Aliment Pharmacol Ther. 2014;40 (5):538–547. [DOI] [PubMed] [Google Scholar]
- 58. Kaufman DW, Kelly JP, Wiholm BE, et al. The risk of acute major upper gastrointestinal bleeding among users of aspirin and ibuprofen at various levels of alcohol consumption. Am J Gastroenterol. 1999;94 (11):3189–3196. [DOI] [PubMed] [Google Scholar]
- 59. Cuzick J, Thorat MA, Bosetti C, et al. Estimates of benefits and harms of prophylactic use of aspirin in the general population. Ann Oncol. 2014;26 (1):47–57. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 60. van Kruijsdijk RC, Visseren FL, Ridker PM, et al. Individualised prediction of alternate-day aspirin treatment effects on the combined risk of cancer, cardiovascular disease and gastrointestinal bleeding in healthy women. Heart. 2014;101 (5):369–376. [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.