Background: To better understand colorectal cancer etiology and prognosis, archived surgical tissues were collected from Cancer Prevention Study II (CPS-II) Nutrition Cohort participants who were diagnosed with colorectal cancer. Herein, the methodology for this collection is described to help inform other efforts to collect tissues.

Methods: The main components to accruing tissue were: (i) obtaining consent from participants or next-of-kin; (ii) contacting hospitals to request materials; and (iii) pathology review and laboratory processing.

Results: In CPS-II, we identified 3,643 participants diagnosed with colorectal cancer between 1992/1993 and 2009. Of these, tissue could not be sought from cases verified through state cancer registry linkage (N = 1,622), because of insufficient information on tissue location. We sought tissue from the 2,021 cases verified using medical records, and received tissue from 882. When hospitals were contacted within 10 years of diagnosis, we received 87% of tissue materials; beyond that 10-year mark, we received 32%. Compared with the 2,761 colorectal cancer cases without tissue, the 882 cases with tissue were more likely to be alive, diagnosed more recently during follow-up, and had less-advanced staged disease. Cases with and without tissues were similar with respect to age at diagnosis, smoking, body mass index, physical activity, and other epidemiologic factors.

Conclusions: Some of the most important elements in forming a tissue repository included having the cases' hospital contact and surgical accession information as well as contacting patients/next-of-kin and hospitals within 10 years of surgery.

Impact: This tissue repository will serve as an important resource for colorectal cancer studies.

See all the articles in this CEBP Focus section, “Biomarkers, Biospecimens, and New Technologies in Molecular Epidemiology.”

Cancer Epidemiol Biomarkers Prev; 23(12); 2694–702. ©2014 AACR.

Colorectal cancer is a complex disease that evolves through the acquisition of genetic and epigenetic instability. Sources of this instability include tumor microsatellite instability (MSI), chromosomal instability (CIN), and epigenetic modification (e.g., CpG island methylator phenotype, or CIMP). These sources of instability can be used to define colorectal cancers according to molecular subtypes (1, 2), some of which have implications for risk factor identification and prognosis (3).

Molecular pathologic epidemiology (MPE) combines traditional approaches in epidemiology and pathology to identify whether risk or prognostic factors for a given disease differ according to molecular phenotype of the disease (3). Recent findings from MPE studies suggest that associations of obesity and smoking with colorectal cancer risk differ by tumor MSI or CIMP (4–8). Motivated, in part, by these findings, we sought to establish a repository of colorectal tissue specimens from participants enrolled in the Cancer Prevention Study-II (CPS-II) Nutrition Cohort. Here, we describe the methods developed to accrue tissues and compare characteristics of cases with and without tissues to help inform other epidemiologic efforts to collect these materials.

### The CPS-II baseline and nutrition cohorts

Men and women in the CPS-II Nutrition Cohort (N = 184,194) were recruited from among the 1.2 million U.S. adults enrolled in the CPS-II Baseline Cohort, a study of cancer mortality that was initiated in 1982 (9). In 1992 and 1993, a detailed questionnaire was mailed to a subgroup of the Baseline Cohort. Respondents to this questionnaire were enrolled into the CPS-II Nutrition Cohort (9). Participants in CPS-II Nutrition are followed for cancer incidence and mortality; they have received additional mailed questionnaires in 1997 and every 2 years thereafter to update exposure information and to obtain self-reported cancer diagnoses. We ask each participant who self-reports a cancer diagnosis to grant us permission to obtain her/his medical records to verify the diagnosis. Fatal cases are also identified through linkage with the National Death Index. When medical records cannot be obtained, often because the participants died before being able to self-report their cancer or provide consent for access to medical records, cancer diagnoses are verified through computerized linkage with state cancer registries.

Blood samples were collected from 39,380 members of the CPS-II Nutrition Cohort from 1998 to 2001, and buccal cell samples were collected from an additional 67,000 cohort members in 2001 and 2002. All aspects of the CPS-II study are approved by the Emory University Institutional Review Board (IRB).

### Methods for establishing the Colorectal Tissue Repository in CPS-II Nutrition

With the above CPS-II resources in-hand, in 2008, we began a pilot investigation to determine the feasibility of, and optimal procedures for, collecting archived colorectal tissue blocks from hospitals. The initial pilot investigation included conferring with experts in the field of MPE who had been collecting archived surgical tissue materials (i.e., formalin-fixed, paraffin embedded tissues, FFPE) in epidemiologic studies. Literature reviews and internet searches on tumor tissue collection as well as scientific papers on MPE (e.g., refs. 10–12) supplemented the information we learned from our colleagues in the field.

Initial feasibility was determined, including projected sample size estimations and associated power calculations, and the general study procedures were established. Next, we created the standard operating procedures protocol, IRB application, and participant and hospital materials for mailings. We obtained IRB approval for tissue accrual in August 2009.

Eligible cases for tumor block acquisition in this study were participants in the CPS-II Nutrition Cohort who: (i) reported a diagnosis of colorectal cancer between baseline (1992 or 1993) and our latest incidence follow-up in mid-2009; (ii) provided us with written consent to obtain their medical records concerning their colorectal cancer diagnosis, and; (iii) had their colorectal adenocarcinoma diagnosis confirmed by a review of medical records by a Certified Tumor Registrar.

One of the first steps in tissue acquisition was to abstract medical record/pathology report information into a database. The medical record abstraction usually includes surgical pathology records from the patient's initial surgical procedure (e.g., resection) which contain the required surgical accession number and hospital contact information, both of which are required for locating the appropriate tissue blocks. The surgical accession number is crucial for identifying exactly which specimen(s) to request from hospitals, as some patients will have had several procedures at the same hospital and each procedure will have its own accession number. CPS-II case verification through state cancer registry linkage does not include acquiring information concerning hospital location or surgical accession. Therefore, in this project, we did not pursue the feasibility of collecting tissue materials from participants whose cancers were verified through state cancer registry linkage.

After identifying eligible cases, we contacted participants or their next-of-kin (if the next-of-kin was a spouse participating in the CPS-II Nutrition Cohort) to gain informed consent. A generic version of the materials we sent to participants (or their next-of-kin) is included in the Supplementary Materials S1 (online only, including a cover letter, informed consent form, HIPAA form, and revocation letter). Up to four letters were sent and, initially, these letters were followed by one phone call to nonresponders to gain informed consent. Because these phone calls had minimal impact on obtaining consent, this step was discontinued after the early pilot phase.

Initially, only participants who were still alive, or, if the participant was deceased, her/his spouse (if the spouse was also enrolled in CPS-II Nutrition), were sent a consent mailing. After the study was underway, we obtained additional IRB approval to acquire tumor specimens from hospitals directly, without seeking spousal consent, for participants who were deceased and had previously given us permission to access their medical records. For deceased patients with no separate tissue consent form, hospitals were provided with a copy of the patient's signed authorization for medical record release. A cover letter was included with the request to hospitals, explaining that the patient was deceased and previously had given the Epidemiology Research Program permission to retrieve the patient's medical records. The cover letter also explained that most institutions consider tumor blocks to be a part of a patient's medical record. If a hospital did not accept the medical record authorization release form to be valid for the release of tumor tissues, the study coordinator spoke to hospital staff and reminded them that deceased patients are not considered human subjects under the common rule and that we comply with the regulations for protected health information (PHI) for deceased patients in the same manner as we do for living patients. It was also communicated that approval was obtained from the Emory University IRB to acquire tumor specimens from hospitals directly, without spousal consent. If our Program staff determined that the hospital found the medical record consent form clearly unacceptable, then no further efforts were made.

The next major step in tissue acquisition from patients/next-of-kin who granted consent was to request the tissue materials from the hospital pathology laboratory where the participant had her/his cancer surgery. By law, in most states and in accordance with the College of American Pathologists guidelines for retention of laboratory records and materials (13), these specimens are stored for 10 years; some hospitals will keep the materials beyond 10 years, but most will discard the materials to save storage space. To simplify the process for hospitals, first all materials relating to the participants' surgery for cancer were requested. If the hospital preferred not to release the entire case materials, then one best representative cancer block and one normal block, typically selected from one of the distant surgical margins, were requested. If hospitals preferred not to release any original blocks, we requested 20 unstained tumor slides and 20 unstained normal slides from best representative cancer and normal blocks. Best representative blocks were selected by the local pathologists. We contacted hospitals with up to three letters and phoned all hospital nonresponders at least twice. Generic versions of the materials we sent to hospitals are included in Supplementary Materials S2 (online only). Block/slide materials were returned to our Program offices and logged into our tracking system. One investigator (P.T. Campbell) then entered more detailed information relating to the contents of the mailed package and the pathology report into a separate database.

The final step in block acquisition was pathology review and sample processing. Pathology review occurs at Emory University's Pathology Department. Typically, the study pathologist (A.B. Farris) reviewed batches of 20 to 50 deidentified cases at a time. Specifically, he reviewed cases to identify the best representative tumor and normal blocks; he also made notes on histologic features that may be of interest to future studies (e.g., tumor-infiltrating lymphocytes, mucinous or medullary morphology, or presence of signet ring cells). After pathology review, the best representative cancer and normal blocks from each participant were sent for processing to the Mayo Clinic Biospecimens Accessioning and Processing Core laboratory and the Pathology Research Core laboratory (collectively referred to as the “MC Core laboratory”). At the MC Core laboratory, the best representative cancer and normal blocks were sectioned onto unstained slides at thicknesses of 4 to 5 μm for future immunohistochemistry (IHC) and at 10 μm for future DNA extraction. Hematoxylin and eosin (H&E) slides at the beginning, in the middle, and at the end of block sectioning were also constructed. IHC slides were constructed for the four main mismatch repair proteins (i.e., MLH1, MSH2, MSH6 and PMS2). Tumor and normal DNA was extracted and stored separately, either from a block core or from a 10 μm slide. Initially, newly sectioned unstained slides were stored at −80°C at the MC Core lab while unstained slides that we received directly from hospitals were stored at room temperature in an air conditioned storage room. More recently, we have redipped all unstained slides in paraffin and all slides are stored at −80°C to preserve antigenicity and prevent oxidation. In the future, we may elect to construct tissue microarrays (TMA) from blocks that we were allowed to retain.

We computed means and frequencies for selected clinical and epidemiologic characteristics of CPS-II Nutrition Cohort participants who were diagnosed with a verified colorectal cancer. Detailed descriptions of these clinical and epidemiologic factors are presented elsewhere (14–20). Participants were stratified by tissue accrual status (accrued versus not accrued) and method of diagnosis confirmation (medical record versus state cancer registry linkage) and compared using Pearson χ2 tests or t tests, as appropriate.

As described in Fig. 1, after baseline (1992 or 1993) and up to mid-2009, 3,643 participants in the CPS-II Nutrition Cohort were diagnosed with a verified colon or rectal cancer. Of these 3,643 participants, 2,021 were verified through medical record abstraction while the remaining 1,622 participants were verified through linkage with state cancer registries and were not eligible for tissue accrual in this study. Of the medical record confirmed cases, 1,561 were alive (or had a living spouse who participated in CPS-II) and contacted for consent: 1,082 of these participants granted informed consent (241 declined, 194 did not respond, and 44 signed only one of the required forms and did not respond to additional requests for signature) and we subsequently obtained tissue specimens for 731 of them.

Figure 1.

Flow diagram of tissue accrual in the CPS-II Colorectal Tissue Repository.

Figure 1.

Flow diagram of tissue accrual in the CPS-II Colorectal Tissue Repository.

Close modal

The remaining 460 colorectal cancer cases whose diagnoses were confirmed by medical record review were deceased at the time of tissue accrual. To date, we have obtained tissues for 151 of these individuals, with another 59 still in progress. Of the 460 hospital contacts for deceased patients, 30 (6.5%) refused on the basis of not having direct/spousal consent and/or because the consent form for medical record acquisition was older than 1 year.

In total, we have tissue samples for 882 individuals. Of these, blocks were obtained for 649 participants, whereas unstained slides were obtained for the remaining 233 (26.4%) participants. The percentage of hospitals that sent unstained slides, compared with blocks, differed slightly by period of diagnosis: for cases diagnosed from 1992 to 2006, 184 of 752 (24%) cases for whom we have a tissue sample were sent as unstained slides, whereas for participants diagnosed from 2007 to 2009, 49 of 130 (38%) were sent as unstained slides. Of the 882 individuals with tissue specimens, 700 participants had previously donated a blood or buccal cell sample.

Hospitals requested reimbursement for 105 of the 882 cases with tissue in the repository (11.9%). The charges initially varied from $20 to$650 per case. Among hospitals that charged fees, the mean charge was $149. Of the 233 cases for which we received unstained slides, we were charged fees for 62 (26.6%). The average charge for unstained slides, when fees were requested, was$189. Of the 649 cases for which we received blocks, we were charged fees for 43 (6.6%). The average charge for blocks, when fees were requested, was $91. More recently, we have instituted a cap on fees. For cases without a previously donated blood or buccal cell sample, the cap is$200; the cap is increased to \$300 for cases that have an available blood or buccal sample. Of the first 200 cases we received in the tissue repository, we were charged fees for 25 samples (12.5%). Of the last 200 cases we have received in the tissue repository, we were charged fees for 26 (13%).

Tissue blocks are archived at the American Cancer Society Corporate Center, unless a hospital requests for the return of a specimen. When a hospital contacts us to return tissue blocks (to date, 14 cases have been returned to hospitals), if processing of the blocks is complete, all materials (less the processed slides and DNA) are returned to the hospital. If a case has not been processed, we request the MC Core laboratory to prioritize the case. However, in the rare event that a hospital requests the return of a case due to clinical care of a patient (to date this has occurred one time), the case is returned immediately, regardless of processing status. When appropriate, arrangements are made with the hospital to re-release the tissue blocks to us, once clinical care of the patient is complete.

Loan periods were not formally negotiated with hospitals. If a case was at the MC Core laboratory for processing, an estimated timeframe for return was provided to the hospital. Hospital staff was agreeable for us to return the tissue blocks over a period of months and even more than a year in some instances.

Although not shown in Fig. 1, the most common reason for not receiving tissue materials when requested from hospitals was because the tissue had been discarded since the patient's diagnosis was beyond 10 years [i.e., of the 351 participants or next-of-kin who consented to tissue accrual for whom we did not receive tissue from the hospitals, 271 (77%) had been discarded because they were beyond 10 years since surgery]. For another perspective on this, among the 1,082 participants or next-of-kin who consented to tissue acquisition, we received 570 of 650 (88%) tissues where the participant was diagnosed within 10 years of hospital contact and 161 of 432 (37%) tissues where the participant was diagnosed beyond the 10-year mark. Although tissue accrual is not yet complete for the 460 deceased persons, the pattern by time since diagnosis appears to be consistent thus far. When considering the entire series of hospital contacts thus far, excluding the 59 deceased cases that are still in progress, when hospitals were contacted within 10 years of patient diagnosis, we received 644 of 744 (87%) tissue materials, while beyond the 10-year mark, we received 238 of 743 (32%) tissue materials.

Table 1 compares clinical characteristics of cases with and without tumor tissue, both overall and when further stratified by method of diagnosis confirmation (i.e., medical record versus state cancer registry linkage). Age at diagnosis, grade, and tumor subsite in the colorectum were not meaningfully different between the two groups, although the differences for age and grade were statistically significant. More pronounced differences were observed for vital status and stage. Among colorectal cancer cases with and without tissue materials, 32.3% and 61.3% of cases were deceased, respectively. When state registry confirmed cases were excluded from this comparison, the differences were less pronounced (32.3% and 41.7%) but the difference is still statistically significant. Similarly, more advanced staged disease was observed in the group for whom tissues were not received than in the group for whom tissues were received (e.g., 17.7% vs. 5.4% diagnosed with distant metastases, respectively). Among cases confirmed via medical records, the distribution of stage is similar for those patients with tissues (localized: 46.8%; regional: 47.4%; distant metastases: 5.4%; missing: 0.3%) and without tissues (localized: 47.6%; regional: 45.0%; distant metastases: 5.4%; missing: 1.9%), although the difference is statistically significant when the missing category is included (P: 0.013): when the missing category is excluded from this comparison, the difference is no longer statistically significant (P: 0.77).

Table 1.

Clinical characteristics of colorectal cancer cases in the CPS-II Nutrition Cohort overall and stratified by tissue accrual status and method of cancer diagnosis confirmation

Medical record confirmed casesState cancer registry confirmed casesAll colorectal cancer casesMedical record confirmed casesState cancer registry confirmed casesAll colorectal cancer casesMedical record confirmed cases
CategoriesN = 2,021N = 1,622N = 3,643N = 1,139N = 1,622N = 2,761N = 882Tissue received vs. not receivedTissue received vs. not received in those with a medical record confirmation
Age at diagnosis: Mean (SD) 72.4 (6.65) 73.7 (7.13) 73.0 (6.89) 71.4 (6.65) 73.7 (7.13) 72.8 (7.02) 73.7 (6.43) P = 0.0005 P < 0.0001
Age at diagnosis, N (%)
<65 248 (12.3) 168 (10.4) 416 (11.4) 179 (15.7) 168 (10.4) 347 (12.6) 69 (7.8) P = 0.0018 P < 0.0001
65–<70 422 (20.9) 282 (17.4) 704 (19.3) 260 (22.8) 282 (17.4) 542 (19.6) 162 (18.4)
70–<75 575 (28.5) 400 (24.7) 975 (26.8) 324 (28.4) 400 (24.7) 724 (26.2) 251 (28.5)
75–<70 480 (23.8) 404 (24.9) 884 (24.3) 252 (22.1) 404 (24.9) 656 (23.8) 228 (25.9)
80+ 296 (14.6) 368 (22.7) 664 (18.2) 124 (10.9) 368 (22.7) 492 (17.8) 172 (19.5)
Diagnosis year, N (%)
1992–1997 563 (27.9) 487 (30.0) 1,050 (28.8) 444 (39.0) 487 (30.0) 931 (33.7) 119 (13.5) P < 0.0001 P < 0.0001
1998–2001 669 (33.1) 459 (28.3) 1,128 (31.0) 425 (37.3) 459 (28.3) 884 (32.0) 244 (27.7)
2002–2005 509 (25.2) 433 (26.7) 942 (25.9) 180 (15.8) 433 (26.7) 613 (22.2) 329 (37.3)
2006–2009 280 (13.9) 243 (15.0) 523 (14.4) 90 (7.9) 243 (15.0) 333 (12.1) 190 (21.5)
Vital status at last contact, N (%)
Alive 1,261 (62.4) 405 (25.0) 1,666 (45.7) 664 (58.3) 405 (25.0) 1,069 (38.7) 597 (67.7) P < 0.0001 P < 0.0001
Dead 760 (37.6) 1,217 (75.0) 1,977 (54.3) 475 (41.7) 1,217 (75.0) 1,692 (61.3) 285 (32.3)
SEER stage, N (%)
Local 955 (47.3) 473 (29.2) 1,428 (39.2) 542 (47.6) 473 (29.2) 1,015 (36.8) 413 (46.8) P < 0.0001 P = 0.013
Regional 931 (46.1) 565 (34.8) 1,496 (41.1) 513 (45.0) 565 (34.8) 1,078 (39.0) 418 (47.4)
Distant 110 (5.4) 427 (26.3) 537 (14.7) 62 (5.4) 427 (26.3) 489 (17.7) 48 (5.4)
Unknown 25 (1.2) 157 (9.7) 182 (5.0) 22 (1.9) 157 (9.7) 179 (6.5) 3 (0.3)
1 252 (12.5) 136 (8.4) 388 (10.7) 149 (13.1) 136 (8.4) 285 (10.3) 103 (11.7) P < 0.0001 P = 0.19
2 1,244 (61.6) 832 (51.3) 2,076 (57) 687 (60.3) 832 (51.3) 1,519 (55.0) 557 (63.2)
3 326 (16.1) 340 (21.0) 666 (18.3) 186 (16.3) 340 (21.0) 526 (19.1) 140 (15.9)
4 25 (1.2) 24 (1.5) 49 (1.3) 10 (0.9) 24 (1.5) 34 (1.2) 15 (1.7)
Unknown 174 (8.6) 290 (17.9) 464 (12.7) 107 (9.4) 290 (17.9) 397 (14.4) 67 (7.6)
Subsite, N (%)
Colon 1,511 (74.8) 1,217 (75.0) 2,728 (74.9) 831 (73.0) 1,217 (75.0) 2,048 (74.2) 680 (77.1) P = 0.08 P = 0.034
Rectum 510 (25.2) 405 (25.0) 915 (25.1) 308 (27.0) 405 (25.0) 713 (25.8) 202 (22.9)
Medical record confirmed casesState cancer registry confirmed casesAll colorectal cancer casesMedical record confirmed casesState cancer registry confirmed casesAll colorectal cancer casesMedical record confirmed cases
CategoriesN = 2,021N = 1,622N = 3,643N = 1,139N = 1,622N = 2,761N = 882Tissue received vs. not receivedTissue received vs. not received in those with a medical record confirmation
Age at diagnosis: Mean (SD) 72.4 (6.65) 73.7 (7.13) 73.0 (6.89) 71.4 (6.65) 73.7 (7.13) 72.8 (7.02) 73.7 (6.43) P = 0.0005 P < 0.0001
Age at diagnosis, N (%)
<65 248 (12.3) 168 (10.4) 416 (11.4) 179 (15.7) 168 (10.4) 347 (12.6) 69 (7.8) P = 0.0018 P < 0.0001
65–<70 422 (20.9) 282 (17.4) 704 (19.3) 260 (22.8) 282 (17.4) 542 (19.6) 162 (18.4)
70–<75 575 (28.5) 400 (24.7) 975 (26.8) 324 (28.4) 400 (24.7) 724 (26.2) 251 (28.5)
75–<70 480 (23.8) 404 (24.9) 884 (24.3) 252 (22.1) 404 (24.9) 656 (23.8) 228 (25.9)
80+ 296 (14.6) 368 (22.7) 664 (18.2) 124 (10.9) 368 (22.7) 492 (17.8) 172 (19.5)
Diagnosis year, N (%)
1992–1997 563 (27.9) 487 (30.0) 1,050 (28.8) 444 (39.0) 487 (30.0) 931 (33.7) 119 (13.5) P < 0.0001 P < 0.0001
1998–2001 669 (33.1) 459 (28.3) 1,128 (31.0) 425 (37.3) 459 (28.3) 884 (32.0) 244 (27.7)
2002–2005 509 (25.2) 433 (26.7) 942 (25.9) 180 (15.8) 433 (26.7) 613 (22.2) 329 (37.3)
2006–2009 280 (13.9) 243 (15.0) 523 (14.4) 90 (7.9) 243 (15.0) 333 (12.1) 190 (21.5)
Vital status at last contact, N (%)
Alive 1,261 (62.4) 405 (25.0) 1,666 (45.7) 664 (58.3) 405 (25.0) 1,069 (38.7) 597 (67.7) P < 0.0001 P < 0.0001
Dead 760 (37.6) 1,217 (75.0) 1,977 (54.3) 475 (41.7) 1,217 (75.0) 1,692 (61.3) 285 (32.3)
SEER stage, N (%)
Local 955 (47.3) 473 (29.2) 1,428 (39.2) 542 (47.6) 473 (29.2) 1,015 (36.8) 413 (46.8) P < 0.0001 P = 0.013
Regional 931 (46.1) 565 (34.8) 1,496 (41.1) 513 (45.0) 565 (34.8) 1,078 (39.0) 418 (47.4)
Distant 110 (5.4) 427 (26.3) 537 (14.7) 62 (5.4) 427 (26.3) 489 (17.7) 48 (5.4)
Unknown 25 (1.2) 157 (9.7) 182 (5.0) 22 (1.9) 157 (9.7) 179 (6.5) 3 (0.3)
1 252 (12.5) 136 (8.4) 388 (10.7) 149 (13.1) 136 (8.4) 285 (10.3) 103 (11.7) P < 0.0001 P = 0.19
2 1,244 (61.6) 832 (51.3) 2,076 (57) 687 (60.3) 832 (51.3) 1,519 (55.0) 557 (63.2)
3 326 (16.1) 340 (21.0) 666 (18.3) 186 (16.3) 340 (21.0) 526 (19.1) 140 (15.9)
4 25 (1.2) 24 (1.5) 49 (1.3) 10 (0.9) 24 (1.5) 34 (1.2) 15 (1.7)
Unknown 174 (8.6) 290 (17.9) 464 (12.7) 107 (9.4) 290 (17.9) 397 (14.4) 67 (7.6)
Subsite, N (%)
Colon 1,511 (74.8) 1,217 (75.0) 2,728 (74.9) 831 (73.0) 1,217 (75.0) 2,048 (74.2) 680 (77.1) P = 0.08 P = 0.034
Rectum 510 (25.2) 405 (25.0) 915 (25.1) 308 (27.0) 405 (25.0) 713 (25.8) 202 (22.9)

Table 2 compares participants with and without tissues on select epidemiologic risk factors for colorectal cancer. Modest differences were noted for baseline age, alcohol intake, and diabetes status. No statistically significant differences in means or frequencies were observed for sex, body mass index, physical activity, red meat intake, cigarette smoking, and use of nonsteroidal anti-inflammatory drugs.

Table 2.

Demographic characteristics at baseline (1992/1993) of colorectal cancer cases in the CPS-II Nutrition Cohort overall and stratified by tissue accrual status and method of cancer diagnosis confirmation

MRRegistryAllMRRegistryAllMR
CategoriesN = 2,021N = 1,622N = 3,643N = 1,139N = 1,622N = 2,761N = 882Tissue received vs. not receivedTissue received vs. not received in those with a MR
Baseline age, mean (SD) 64.2 (5.91) 65.6 (6.1) 64.8 (6.03) 64.6 (5.91) 65.6 (6.1) 65.2 (6.04) 63.7 (5.86) P < 0.0001 P = 0.0005
Baseline age, N (%)
<60 459 (22.7) 264 (16.3) 723 (19.8) 238 (20.9) 264 (16.3) 502 (18.2) 221 (25.1) P < 0.0001 P = 0.055
60–<65 558 (27.6) 429 (26.4) 987 (27.1) 306 (26.9) 429 (26.4) 735 (26.6) 252 (28.6)
65–<70 642 (31.8) 506 (31.2) 1,148 (31.5) 375 (32.9) 506 (31.2) 881 (31.9) 267 (30.3)
70–<75 294 (14.5) 322 (19.9) 616 (16.9) 175 (15.4) 322 (19.9) 497 (18) 119 (13.5)
75+ 68 (3.4) 101 (6.2) 169 (4.6) 45 (4.0) 101 (6.2) 146 (5.3) 23 (2.6)
Gender, N (%)
Women 927 (45.9) 749 (46.2) 1,676 (46) 506 (44.4) 749 (46.2) 1,255 (45.5) 421 (47.7) P = 0.24 P = 0.14
Men 1,094 (54.1) 873 (53.8) 1,967 (54) 633 (55.6) 873 (53.8) 1,506 (54.5) 461 (52.3)
BMI, mean (SD) 26.4 (4.31) 26.5 (4.4) 26.5 (4.35) 26.4 (4.42) 26.5 (4.4) 26.5 (4.41) 26.5 (4.16) P = 0.97 P = 0.85
BMI, N (%)
<18.5 25 (1.2) 16 (1.0) 41 (1.1) 19 (1.7) 16 (1.0) 35 (1.3) 6 (0.7) P = 0.21 P = 0.22
18.5–<25 770 (38.1) 609 (37.5) 1,379 (37.9) 436 (38.3) 609 (37.5) 1,045 (37.8) 334 (37.9)
25–<30 837 (41.4) 671 (41.4) 1,508 (41.4) 460 (40.4) 671 (41.4) 1,131 (41) 377 (42.7)
30+ 365 (18.1) 293 (18.1) 658 (18.1) 208 (18.3) 293 (18.1) 501 (18.1) 157 (17.8)
Unknown 24 (1.2) 33 (2.0) 57 (1.6) 16 (1.4) 33 (2.0) 49 (1.8) 8 (0.9)
MET hours of exercise, mean (SD) 12.3 (12.96) 12.1 (12.63) 12.2 (12.81) 12.3 (12.86) 12.1 (12.63) 12.2 (12.72) 12.2 (13.09) P = 0.94 P = 0.91
MET hours of exercise, N (%)
<3.5 219 (10.8) 221 (13.6) 440 (12.1) 130 (11.4) 221 (13.6) 351 (12.7) 89 (10.1) P = 0.22 P = 0.73
3.5–<4.5 614 (30.4) 465 (28.7) 1,079 (29.6) 331 (29.1) 465 (28.7) 796 (28.8) 283 (32.1)
4.5–<14 322 (15.9) 230 (14.2) 552 (15.2) 182 (16) 230 (14.2) 412 (14.9) 140 (15.9)
14–<24.5 468 (23.2) 366 (22.6) 834 (22.9) 271 (23.8) 366 (22.6) 637 (23.1) 197 (22.3)
24.5+ 360 (17.8) 308 (19.0) 668 (18.3) 203 (17.8) 308 (19.0) 511 (18.5) 157 (17.8)
Unknown 38 (1.9) 32 (2.0) 70 (1.9) 22 (1.9) 32 (2.0) 54 (2.0) 16 (1.8)
Servings of red meat/week, mean (SD) 5.1 (4.21) 5.2 (4.28) 5.2 (4.24) 5.2 (4.37) 5.2 (4.28) 5.2 (4.32) 5.0 (4.01) P = 0.25 P = 0.30
Servings of red meat/week, N (%)
<1.5 318 (15.7) 238 (14.7) 556 (15.3) 171 (15) 238 (14.7) 409 (14.8) 147 (16.7) P = 0.10 P = 0.60
1.5–<3 352 (17.4) 268 (16.5) 620 (17.0) 201 (17.6) 268 (16.5) 469 (17.0) 151 (17.1)
3–<5 419 (20.7) 317 (19.5) 736 (20.2) 244 (21.4) 317 (19.5) 561 (20.3) 175 (19.8)
5–<8 382 (18.9) 299 (18.4) 681 (18.7) 204 (17.9) 299 (18.4) 503 (18.2) 178 (20.2)
8+ 373 (18.5) 294 (18.1) 667 (18.3) 214 (18.8) 294 (18.1) 508 (18.4) 159 (18.0)
Unknown 177 (8.8) 206 (12.7) 383 (10.5) 105 (9.2) 206 (12.7) 311 (11.3) 72 (8.2)
Alcohol drinks/week, mean (SD) 4.7 (8.25) 4.2 (8.24) 4.5 (8.24) 4.4 (8.09) 4.2 (8.24) 4.3 (8.17) 4.9 (8.44) P = 0.056 P = 0.17
Alcohol drinks/day, N (%)
None 769 (38.1) 665 (41.0) 1,434 (39.4) 435 (38.2) 665 (41.0) 1,100 (39.8) 334 (37.9) P = 0.044 P = 0.98
<1 188 (9.3) 174 (10.7) 362 (9.9) 105 (9.2) 174 (10.7) 279 (10.1) 83 (9.4)
1+ 999 (49.4) 690 (42.5) 1,689 (46.4) 561 (49.3) 690 (42.5) 1,251 (45.3) 438 (49.7)
Unknown 65 (3.2) 93 (5.7) 158 (4.3) 38 (3.3) 93 (5.7) 131 (4.7) 27 (3.1)
Cigarette smoking, N (%)
Never 791 (39.1) 627 (38.7) 1,418 (38.9) 445 (39.1) 627 (38.7) 1,072 (38.8) 346 (39.2) P = 0.82 P = 0.84
Former 1,040 (51.5) 799 (49.3) 1,839 (50.5) 590 (51.8) 799 (49.3) 1,389 (50.3) 450 (51.0)
Current 164 (8.1) 179 (11.0) 343 (9.4) 88 (7.7) 179 (11.0) 267 (9.7) 76 (8.6)
Unknown 26 (1.3) 17 (1.0) 43 (1.2) 16 (1.4) 17 (1.0) 33 (1.2) 10 (1.1)
NSAID pills/month, N (%)
No use 931 (46.1) 728 (44.9) 1,659 (45.5) 516 (45.3) 728 (44.9) 1,244 (45.1) 415 (47.1) P = 0.21 P = 0.77
<15 302 (14.9) 232 (14.3) 534 (14.7) 178 (15.6) 232 (14.3) 410 (14.8) 124 (14.1)
15–<30 190 (9.4) 119 (7.3) 309 (8.5) 101 (8.9) 119 (7.3) 220 (8.0) 89 (10.1)
30–<60 352 (17.4) 310 (19.1) 662 (18.2) 201 (17.6) 310 (19.1) 511 (18.5) 151 (17.1)
60+ 169 (8.4) 147 (9.1) 316 (8.7) 100 (8.8) 147 (9.1) 247 (8.9) 69 (7.8)
Unknown 77 (3.8) 86 (5.3) 163 (4.5) 43 (3.8) 86 (5.3) 129 (4.7) 34 (3.9)
Diabetes
No 1,859 (92.0) 1,444 (89.0) 3,303 (90.7) 1,037 (91.0) 1,444 (89.0) 2,481 (89.9) 822 (93.2) P = 0.003 P = 0.08
Yes 162 (8.0) 178 (11.0) 340 (9.3) 102 (9.0) 178 (11.0) 280 (10.1) 60 (6.8)
MRRegistryAllMRRegistryAllMR
CategoriesN = 2,021N = 1,622N = 3,643N = 1,139N = 1,622N = 2,761N = 882Tissue received vs. not receivedTissue received vs. not received in those with a MR
Baseline age, mean (SD) 64.2 (5.91) 65.6 (6.1) 64.8 (6.03) 64.6 (5.91) 65.6 (6.1) 65.2 (6.04) 63.7 (5.86) P < 0.0001 P = 0.0005
Baseline age, N (%)
<60 459 (22.7) 264 (16.3) 723 (19.8) 238 (20.9) 264 (16.3) 502 (18.2) 221 (25.1) P < 0.0001 P = 0.055
60–<65 558 (27.6) 429 (26.4) 987 (27.1) 306 (26.9) 429 (26.4) 735 (26.6) 252 (28.6)
65–<70 642 (31.8) 506 (31.2) 1,148 (31.5) 375 (32.9) 506 (31.2) 881 (31.9) 267 (30.3)
70–<75 294 (14.5) 322 (19.9) 616 (16.9) 175 (15.4) 322 (19.9) 497 (18) 119 (13.5)
75+ 68 (3.4) 101 (6.2) 169 (4.6) 45 (4.0) 101 (6.2) 146 (5.3) 23 (2.6)
Gender, N (%)
Women 927 (45.9) 749 (46.2) 1,676 (46) 506 (44.4) 749 (46.2) 1,255 (45.5) 421 (47.7) P = 0.24 P = 0.14
Men 1,094 (54.1) 873 (53.8) 1,967 (54) 633 (55.6) 873 (53.8) 1,506 (54.5) 461 (52.3)
BMI, mean (SD) 26.4 (4.31) 26.5 (4.4) 26.5 (4.35) 26.4 (4.42) 26.5 (4.4) 26.5 (4.41) 26.5 (4.16) P = 0.97 P = 0.85
BMI, N (%)
<18.5 25 (1.2) 16 (1.0) 41 (1.1) 19 (1.7) 16 (1.0) 35 (1.3) 6 (0.7) P = 0.21 P = 0.22
18.5–<25 770 (38.1) 609 (37.5) 1,379 (37.9) 436 (38.3) 609 (37.5) 1,045 (37.8) 334 (37.9)
25–<30 837 (41.4) 671 (41.4) 1,508 (41.4) 460 (40.4) 671 (41.4) 1,131 (41) 377 (42.7)
30+ 365 (18.1) 293 (18.1) 658 (18.1) 208 (18.3) 293 (18.1) 501 (18.1) 157 (17.8)
Unknown 24 (1.2) 33 (2.0) 57 (1.6) 16 (1.4) 33 (2.0) 49 (1.8) 8 (0.9)
MET hours of exercise, mean (SD) 12.3 (12.96) 12.1 (12.63) 12.2 (12.81) 12.3 (12.86) 12.1 (12.63) 12.2 (12.72) 12.2 (13.09) P = 0.94 P = 0.91
MET hours of exercise, N (%)
<3.5 219 (10.8) 221 (13.6) 440 (12.1) 130 (11.4) 221 (13.6) 351 (12.7) 89 (10.1) P = 0.22 P = 0.73
3.5–<4.5 614 (30.4) 465 (28.7) 1,079 (29.6) 331 (29.1) 465 (28.7) 796 (28.8) 283 (32.1)
4.5–<14 322 (15.9) 230 (14.2) 552 (15.2) 182 (16) 230 (14.2) 412 (14.9) 140 (15.9)
14–<24.5 468 (23.2) 366 (22.6) 834 (22.9) 271 (23.8) 366 (22.6) 637 (23.1) 197 (22.3)
24.5+ 360 (17.8) 308 (19.0) 668 (18.3) 203 (17.8) 308 (19.0) 511 (18.5) 157 (17.8)
Unknown 38 (1.9) 32 (2.0) 70 (1.9) 22 (1.9) 32 (2.0) 54 (2.0) 16 (1.8)
Servings of red meat/week, mean (SD) 5.1 (4.21) 5.2 (4.28) 5.2 (4.24) 5.2 (4.37) 5.2 (4.28) 5.2 (4.32) 5.0 (4.01) P = 0.25 P = 0.30
Servings of red meat/week, N (%)
<1.5 318 (15.7) 238 (14.7) 556 (15.3) 171 (15) 238 (14.7) 409 (14.8) 147 (16.7) P = 0.10 P = 0.60
1.5–<3 352 (17.4) 268 (16.5) 620 (17.0) 201 (17.6) 268 (16.5) 469 (17.0) 151 (17.1)
3–<5 419 (20.7) 317 (19.5) 736 (20.2) 244 (21.4) 317 (19.5) 561 (20.3) 175 (19.8)
5–<8 382 (18.9) 299 (18.4) 681 (18.7) 204 (17.9) 299 (18.4) 503 (18.2) 178 (20.2)
8+ 373 (18.5) 294 (18.1) 667 (18.3) 214 (18.8) 294 (18.1) 508 (18.4) 159 (18.0)
Unknown 177 (8.8) 206 (12.7) 383 (10.5) 105 (9.2) 206 (12.7) 311 (11.3) 72 (8.2)
Alcohol drinks/week, mean (SD) 4.7 (8.25) 4.2 (8.24) 4.5 (8.24) 4.4 (8.09) 4.2 (8.24) 4.3 (8.17) 4.9 (8.44) P = 0.056 P = 0.17
Alcohol drinks/day, N (%)
None 769 (38.1) 665 (41.0) 1,434 (39.4) 435 (38.2) 665 (41.0) 1,100 (39.8) 334 (37.9) P = 0.044 P = 0.98
<1 188 (9.3) 174 (10.7) 362 (9.9) 105 (9.2) 174 (10.7) 279 (10.1) 83 (9.4)
1+ 999 (49.4) 690 (42.5) 1,689 (46.4) 561 (49.3) 690 (42.5) 1,251 (45.3) 438 (49.7)
Unknown 65 (3.2) 93 (5.7) 158 (4.3) 38 (3.3) 93 (5.7) 131 (4.7) 27 (3.1)
Cigarette smoking, N (%)
Never 791 (39.1) 627 (38.7) 1,418 (38.9) 445 (39.1) 627 (38.7) 1,072 (38.8) 346 (39.2) P = 0.82 P = 0.84
Former 1,040 (51.5) 799 (49.3) 1,839 (50.5) 590 (51.8) 799 (49.3) 1,389 (50.3) 450 (51.0)
Current 164 (8.1) 179 (11.0) 343 (9.4) 88 (7.7) 179 (11.0) 267 (9.7) 76 (8.6)
Unknown 26 (1.3) 17 (1.0) 43 (1.2) 16 (1.4) 17 (1.0) 33 (1.2) 10 (1.1)
NSAID pills/month, N (%)
No use 931 (46.1) 728 (44.9) 1,659 (45.5) 516 (45.3) 728 (44.9) 1,244 (45.1) 415 (47.1) P = 0.21 P = 0.77
<15 302 (14.9) 232 (14.3) 534 (14.7) 178 (15.6) 232 (14.3) 410 (14.8) 124 (14.1)
15–<30 190 (9.4) 119 (7.3) 309 (8.5) 101 (8.9) 119 (7.3) 220 (8.0) 89 (10.1)
30–<60 352 (17.4) 310 (19.1) 662 (18.2) 201 (17.6) 310 (19.1) 511 (18.5) 151 (17.1)
60+ 169 (8.4) 147 (9.1) 316 (8.7) 100 (8.8) 147 (9.1) 247 (8.9) 69 (7.8)
Unknown 77 (3.8) 86 (5.3) 163 (4.5) 43 (3.8) 86 (5.3) 129 (4.7) 34 (3.9)
Diabetes
No 1,859 (92.0) 1,444 (89.0) 3,303 (90.7) 1,037 (91.0) 1,444 (89.0) 2,481 (89.9) 822 (93.2) P = 0.003 P = 0.08
Yes 162 (8.0) 178 (11.0) 340 (9.3) 102 (9.0) 178 (11.0) 280 (10.1) 60 (6.8)

Abbreviation: MR, medical record.

The CPS-II Nutrition Cohort colorectal tissue repository has been established. The methods used to establish this repository are described in detail to motivate and inform future collections of tissue materials in other epidemiologic studies. With tissue samples accrued thus far from 882 participants diagnosed with colon or rectal cancer, the CPS-II colorectal tissue repository is on par with or larger than other tissue collections from comparable prospective cohort studies, including the Nurses' Health Study and Health Professionals Follow-up Study (21–23), European Prospective Investigation into Cancer and Nutrition (EPIC) (24), the Iowa Women's Health Study (8, 25), the Melbourne Collaborative Cohort Study (6, 26), the Netherlands Cohort Study (6), and the Northern Sweden Health and Disease Study (27).

Cases with and without tumor tissue differed according to some clinical characteristics such as stage and vital status. These differences were expected and caused primarily by two factors. First, many medical record confirmed cases lacked tissue because the required 10-year storage time had been exceeded and their tissue had been discarded. Such cases were diagnosed earlier during the follow-up period, and, as a result, were more likely to be deceased than cases diagnosed later. Second, many of the other cases who lacked tissue were verified through state cancer registries rather than through medical records. These cases often died before having the opportunity to self-report their cancer on a biennial follow-up questionnaire or giving us permission to access medical records. When we exclude participants who were confirmed through state cancer registry linkage, the clinical features of the 2,021 participants with and without tissue materials are more similar. Nonetheless, in future prognostic studies using these resources, it will be important to compare the relevant clinical features of participants that have tissues to those who do not have tissues to assess any potential for bias.

One of our primary interests in MPE is to assess whether lifestyle and environmental risk factors are differentially associated with specific molecularly defined tumor phenotypes. Therefore, it is particularly encouraging to see that the distribution of risk factors among participants with tumor tissues is similar to that among the overall population of colorectal cancer cases in this cohort.

In conclusion, we have established a large tissue repository of participants in a prospective cohort study who were diagnosed with colon or rectal cancer. Combined with previously collected blood and buccal cell biospecimens and extensive questionnaire data on lifestyle, demographic, and medical factors, this tissue repository will serve as an important resource for conducting pathoepidemiology studies. From this experience, we found that some of the most important elements in forming a tissue repository included having the cases' hospital contact information and surgical accession numbers as well as contacting patients or their next-of-kin and hospitals within 10 years of surgery. Ultimately, the goal of these resources is to better understand colorectal carcinogenesis and lessen the burden of disease associated with this cancer. While the methods described here were specific to colorectal tissue collection, we have applied similar methodology to breast, hematologic, and prostate tissue collections. Therefore, the methodology described here should be applicable across tumor sites in other studies.

No potential conflicts of interest were disclosed.

Conception and design: P.T. Campbell, P. Briggs, A.B. Farris, S.M. Gapstur

Development of methodology: P.T. Campbell, P. Briggs, A.B. Farris, S.N. Thibodeau

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): P.T. Campbell, A. Deka, P. Briggs, A.B. Farris, L. Tillmans, S.M. Gapstur

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): P.T. Campbell, A. Deka, A.B. Farris, M.M. Gaudet, E.J. Jacobs, C.C. Newton

Writing, review, and/or revision of the manuscript: P.T. Campbell, A. Deka, P. Briggs, M.S. Cicek, A.B. Farris, M.M. Gaudet, E.J. Jacobs, C.C. Newton, A.V. Patel, L.R. Teras, S.N. Thibodeau, L. Tillmans, S.M. Gapstur

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): P.T. Campbell, P. Briggs, A.B. Farris

Study supervision: P.T. Campbell

This work was supported by grants from the American Cancer Society (to P.T. Campbell and S.M. Gapstur).

1.
Jass
JR
.
Classification of colorectal cancer based on correlation of clinical, morphological and molecular features
.
Histopathology
2007
;
50
:
113
30
.
2.
Ogino
S
,
Goel
A
.
Molecular classification and correlates in colorectal cancer
.
J Mol Diagn
2008
;
10
:
13
27
.
3.
Ogino
S
,
Chan
AT
,
Fuchs
CS
,
Giovannucci
E
.
Molecular pathological epidemiology of colorectal neoplasia: an emerging transdisciplinary and interdisciplinary field
.
Gut
2011
;
60
:
397
411
.
4.
Slattery
ML
,
Curtin
K
,
Anderson
K
,
Ma
KN
,
Ballard
L
,
Edwards
S
, et al
Associations between cigarette smoking, lifestyle factors, and microsatellite instability in colon tumors
.
J Natl Cancer Inst
2000
;
92
:
1831
6
.
5.
Campbell
PT
,
Jacobs
ET
,
Ulrich
CM
,
Figueiredo
JC
,
Poynter
JN
,
McLaughlin
JR
, et al
Case-control study of overweight, obesity, and colorectal cancer risk, overall and by tumor microsatellite instability status
.
J Natl Cancer Inst
2010
;
102
:
391
400
.
6.
Hughes
LA
,
Williamson
EJ
,
van Engeland
M
,
Jenkins
MA
,
Giles
GG
,
Hopper
JL
, et al
Body size and risk for colorectal cancers showing BRAF mutations or microsatellite instability: a pooled analysis
.
Int J Epidemiol
2012
;
41
:
1060
72
.
7.
Poynter
JN
,
Haile
RW
,
Siegmund
KD
,
Campbell
PT
,
Figueiredo
JC
,
Limburg
P
, et al
Associations between smoking, alcohol consumption, and colorectal cancer, overall and by tumor microsatellite instability status
.
Cancer Epidemiol Biomarkers Prev
2009
;
18
:
2745
50
.
8.
Limsui
D
,
Vierkant
RA
,
Tillmans
LS
,
Wang
AH
,
Weisenberger
DJ
,
Laird
PW
, et al
Cigarette smoking and colorectal cancer risk by molecularly defined subtypes
.
J Natl Cancer Inst
2010
;
102
:
1012
22
.
9.
Calle
EE
,
Rodriguez
C
,
Jacobs
EJ
,
Almon
ML
,
Chao
A
,
McCullough
ML
, et al
The American Cancer Society Cancer Prevention Study II Nutrition Cohort: rationale, study design, and baseline characteristics
.
Cancer
2002
;
94
:
2490
501
.
10.
Newcomb
PA
,
Baron
J
,
Cotterchio
M
,
Gallinger
S
,
Grove
J
,
Haile
R
, et al
Colon Cancer Family Registry: an international resource for studies of the genetic epidemiology of colon cancer
.
Cancer Epidemiol Biomarkers Prev
2007
;
16
:
2331
43
.
11.
Ogino
S
,
Nosho
K
,
Kirkner
GJ
,
Kawasaki
T
,
Chan
AT
,
Schernhammer
ES
, et al
A cohort study of tumoral LINE-1 hypomethylation and prognosis in colon cancer
.
J Natl Cancer Inst
2008
;
100
:
1734
8
.
12.
Poynter
JN
,
Siegmund
KD
,
Weisenberger
DJ
,
Long
TI
,
Thibodeau
SN
,
Lindor
N
, et al
Molecular characterization of MSI-H colorectal cancer by MLHI promoter methylation, immunohistochemistry, and mismatch repair germline mutation screening
.
Cancer Epidemiol Biomarkers Prev
2008
;
17
:
3208
15
.
14.
Campbell
PT
,
Deka
A
,
Jacobs
EJ
,
Newton
CC
,
Hildebrand
JS
,
McCullough
ML
, et al
Prospective study reveals associations between colorectal cancer and type 2 diabetes mellitus or insulin use in men
.
Gastroenterology
2010
;
139
:
1138
46
.
15.
Campbell
PT
,
Newton
CC
,
Dehal
AN
,
Jacobs
EJ
,
Patel
AV
,
Gapstur
SM
.
Impact of body mass index on survival after colorectal cancer diagnosis: the Cancer Prevention Study-II Nutrition Cohort
.
J Clin Oncol
2012
;
30
:
42
52
.
16.
Campbell
PT
,
Patel
AV
,
Newton
CC
,
Jacobs
EJ
,
Gapstur
SM
.
Associations of recreational physical activity and leisure time spent sitting with colorectal cancer survival
.
J Clin Oncol
2013
;
31
:
876
85
.
17.
McCullough
ML
,
Gapstur
SM
,
Shah
R
,
Jacobs
EJ
,
Campbell
PT
.
Association between red and processed meat intake and mortality among colorectal cancer survivors
.
J Clin Oncol
2013
;
31
:
2773
82
.
18.
McCullough
ML
,
Patel
AV
,
Kushi
LH
,
Patel
R
,
Willett
WC
,
Doyle
C
, et al
Following cancer prevention guidelines reduces risk of cancer, cardiovascular disease, and all-cause mortality
.
Cancer Epidemiol Biomarkers Prev
2011
;
20
:
1089
97
.
19.
Hannan
LM
,
Jacobs
EJ
,
Thun
MJ
.
The association between cigarette smoking and risk of colorectal cancer in a large prospective cohort from the United States
.
Cancer Epidemiol Biomarkers Prev
2009
;
18
:
3362
7
.
20.
Jacobs
EJ
,
Thun
MJ
,
Bain
EB
,
Rodriguez
C
,
Henley
SJ
,
Calle
EE
.
A large cohort study of long-term daily use of adult-strength aspirin and cancer incidence
.
J Natl Cancer Inst
2007
;
99
:
608
15
.
21.
P
,
Kuchiba
A
,
Imamura
Y
,
Liao
X
,
Yamauchi
M
,
Nishihara
R
, et al
Microsatellite instability and BRAF mutation testing in colorectal cancer prognostication
.
J Natl Cancer Inst
2013
;
105
:
1151
6
.
22.
Morikawa
T
,
Kuchiba
A
,
Yamauchi
M
,
Meyerhardt
JA
,
Shima
K
,
Nosho
K
, et al
Association of CTNNB1 (beta-catenin) alterations, body mass index, and physical activity with survival in patients with colorectal cancer
.
JAMA
2011
;
305
:
1685
94
.
23.
Nishihara
R
,
P
,
Kuchiba
A
,
Jung
S
,
Yamauchi
M
,
Liao
X
, et al
Aspirin use and risk of colorectal cancer according to BRAF mutation status
.
JAMA
2013
;
309
:
2563
71
.
24.
Gay
LJ
,
Arends
MJ
,
Mitrou
PN
,
Bowman
R
,
Ibrahim
AE
,
Happerfield
L
, et al
MLH1 promoter methylation, diet, and lifestyle factors in mismatch repair deficient colorectal cancer patients from EPIC-Norfolk
.
Nutr Cancer
2011
;
63
:
1000
10
.
25.
NJ
,
Vierkant
RA
,
Tillmans
LS
,
Wang
AH
,
Weisenberger
DJ
,
Laird
PW
, et al
Associations between colorectal cancer molecular markers and pathways with clinicopathologic features in older women
.
Gastroenterology
2013
;
145
:
348
56
e1–2
.
26.
Walsh
MD
,
Clendenning
M
,
Williamson
E
,
Pearson
SA
,
Walters
RJ
,
Nagler
B
, et al
Expression of MUC2, MUC5AC, MUC5B, and MUC6 mucins in colorectal cancers and their association with the CpG island methylator phenotype
.
Mod Pathol
2013
;
26
:
1642
56
.
27.
Van Guelpen
B
,
Dahlin
AM
,
Hultdin
J
,
Eklof
V
,
Johansson
I
,
Henriksson
ML
, et al
One-carbon metabolism and CpG island methylator phenotype status in incident colorectal cancer: a nested case-referent study
.
Cancer Causes Control
2010
;
21
:
557
66
.