Fish Oil Supplementation, Nutrigenomics and Colorectal Cancer Prevention

1.Rationale and Specific Aims

Colorectal cancer (CRC) is the third most common cancer and the second most frequent cause of
cancer related mortality in the United States. Animal and human studies have suggested that
the marine-derived n-3 polyunsaturated fatty acids (PUFAs), eicosapentanoic acid (EPA) and
docosahexanoic acid (DHA), have cancer inhibitory properties while conversely, n-6 PUFAs such
as arachidonic acid (ARA) may promote tumorigenesis. The mechanism behind these opposing
effects is likely due to differences in the biological activity of their eicosanoids end
products and their effects on chronic inflammation. Prostaglandin E2 (PGE2) is a
pro-inflammatory eicosanoid that is aberrantly produced in both colorectal adenomas and
cancer and is derived from ARA via the cyclo-oxygenase pathway. EPA is converted through the
same pathway into prostaglandin E3, which has 4 to 7- fold less prostaglandin E receptor
affinity, is less inflammatory, and may even be pro-apoptotic compared to PGE2. As such, it
may be the ratio of ARA to EPA and DHA rather than the absolute levels of marine-derived n-3
PUFAs that contribute most towards their antiproliferative and pro-apoptotic effects.

The ratio of ARA to EPA + DHA can be manipulated through fish oil supplementation, however;
genetic factors may play a key role on determining this ratio. Recent genome-wide association
and haplotype studies have demonstrated that up to 28% of the additive variance in tissue
levels of ARA is explained by variants in a single gene, fatty acid desaturase 1 (FADS1).
FADS1 is the rate-limiting enzyme in the conversion of linoleic acid (LA), the most commonly
consumed PUFA, to ARA, and homozygotes for the T allele (population frequency of 13%, HapMap
-CEU) in rs174537 have lower fatty acid desaturase activity and subsequently lower tissue
levels of ARA. While EPA can be produced in vivo from α-linolenic acid, in humans, this
process is extremely inefficient and most tissue level EPA derives directly from dietary
consumption of fatty fish. Thus, high activity of FAD1 and subsequently increased tissue
levels of ARA may offset some of the potential benefits of dietary supplementation with fish
oil. To date, no previously published studies have investigated how genetic variants that
influence fatty acid desaturase activity might modify the beneficial effects of fish oil
supplementation.

The investigators hypothesis is that the individuals with genetically determined lower
activity of FADS1 will derive greater benefit from fish oil supplementation than individuals
with higher FADS1 activity because of lower tissue levels of ARA and subsequently a more
favorable ARA to EPA + DHA ratio. To test this hypothesis the investigators will recruit 150
participants with recently identified adenomatous polyps and conduct a 6-month double blind 3
X 2 factorial randomized controlled trial. The first factor will be the rs174537 genotype
(GG, GT, and TT) in the FADS1 gene and the second factor will be fish oil supplementation
(fish oil versus placebo). The primary study outcome will be the change from baseline in
rectal epithelial cell proliferation as measured by Ki-67 labeling and rectal crypt apoptosis
as measured by TUNEL. Secondary endpoints will include rectal epithelial cell expression of
COX-2 and 15-PGDH, rectal cell production of PGE2 and PGE3, rectal cell fatty acid
concentrations, as well as, changes from baseline in biomarkers of inflammation (C-reactive
protein), adipokines (leptin, adiponectin), and markers of insulin sensitivity (HOMA-IR).

The Specific Aims for this research proposal are:

1. to determine the efficacy of fish oil supplements on rectal epithelial cell
proliferation indexes and markers of rectal crypt apoptosis; and,

2. to determine the effect of genetically determined fatty acid desaturase activity on fish
oil supplementation for markers of colorectal cancer risk.

The investigators long-term objectives are to determine genetic factors that might influence
the efficacy of fish oil supplementation in order to conduct a more definitive adenoma
recurrence trial using marine-derived n-3 PUFAs. The investigators anticipate that fish oil
will have anti-neoplastic effect and individuals with low FADS1 activity will have a greater
response compared to individuals with high FADS1 activity. This study will be the first to
investigate the nutrigenomics of fish oil supplementation in colorectal cancer
chemoprevention and may have implications beyond cancer prevention as fish oil is being
actively investigated for its anti-inflammatory effects in cardiovascular and psychiatric
diseases as well as diabetes mellitus and the metabolic syndrome.

2. Recruitment and Retention

The investigators will identify eligible participants based on the inclusion criteria by
reviewing study data and medical record data collected in the Tennessee Colorectal Polyp
Study(TCPS). Participants still eligible after record review will be mailed an introductory
letter inviting them to participate. One week after the letter is mailed; a trained
interviewer from the Vanderbilt Survey Research Shared Resource (SRSR) will call the
potential participants to provide more detailed information about the study, answer questions
about the study, and to see if they may be interested in participating. At that time, an
appointment will be made with the participant for the baseline in-person visit. After the
initial visit and informed consent is obtained an interviewer from the SRSR will conduct the
baseline interview survey and 24-hour dietary recalls. At the baseline visit, the
investigators will re-genotype rs174535 to confirm the accuracy of the imputation process.
This strategy will allow the investigators to efficiently and accurately identify appropriate
candidates for our study.

Eligible subjects will present to the Vanderbilt General Clinical Research Center (GCRC) for
the initial visit and baseline study procedures. Participants who are eligible for the study
and provide written consent for enrollment will have blood obtained; an adipose tissue biopsy
performed, and undergoes the baseline rectal mucosal biopsy procedure. Treatment assignment
will be obtained from the Vanderbilt Investigational Pharmacy by a coordinator. The first
dose of the study medication will be given to patients at the initial visit and the date and
time recorded. This date and time will be considered the time of randomization.

3. Randomization

Randomization will be performed according to a permuted block randomization scheme stratified
on the three genotypes. Randomization will proceed within these three strata with a block
size of balancing interval, varying randomly according to the outcome of a computer generated
random number. This ensures that the cumulative proportion of assignments to each treatment
will be balanced after each block of assignments has been made.

4. Study Procedures

1. Data Collection

Because outside dietary exposure to both n-6 and n-3 PUFAs could possibly confound the
effect of fish oil supplementation the investigators will perform a total of four
24-hour dietary recall studies for each participant over the course of the study. At
enrollment, the investigators will conduct two 24-hour dietary assessments, one on the
weekday and one on the weekend, as participants' diet may differ based on the day of the
week. In addition the investigators will conduct one 24-hour dietary recall at week 8
and at week 16. The investigators will use data collected from these 24-hour dietary
assessments along with standard food composition tables to calculate dietary exposure to
PUFAs.

The investigators will determine adherence to study drug at each in-person visit during
the study. Medication and medication changes will be recorded at these visits. Patients
who initiate a new NSAID during the study will be withdrawn and an exit visit performed.
Adherence to fish oil will also be determined through RBC phospholipid membrane fatty
acid analysis performed at month 3 and month 6. In addition, to determine whether fish
oil supplementation also influences fatty acid membrane concentration at the target
tissue, the investigators will determine the change in rectal epithelial cell
phospholipid membrane fatty acid concentration.

2. Fish oil Capsules

Participants allocated to fish oil supplementation will be instructed to take three
Lovaza® capsules each containing 465 mg of EPA and 375 mg DHA daily; this will provide a
total daily dose of 1395 mg EPA plus 1125 mg DHA for a total daily dose of fish oil of
2.5 grams. Patients will take one tablet three times a day with meals. Lovaza® capsules
are the only FDA approved preparation of fish oil, and as such, the quality of the drug
is monitored and assured. Pharmacological grade fish oil capsules have the advantage of
providing high concentrations of PUFAs, low levels of contaminants, such as mercury, and
almost no fish odor.

3. Placebo Capsules

The investigators will use oleic acid as a placebo. The reason for the use of oleic acid
is several-fold. First, oleic acid (olive oil) capsules have a similar texture, size,
color, and consistency to fish oil capsules. More importantly, oleic acid does not
undergo conversion to an eicosanoid or any other metabolically active product. This is
opposed to corn oil, which has also been used as a placebo in fish oil studies but is
primarily linoleic acid and could subsequently increase tissue levels of ARA and
confound the results of our study. Oleic acid has been used as a placebo in several
prior studies of fish oil supplementation and is well tolerated.

4. Assessment Visits

Patients will attend the GCRC clinic at the baseline (initial visit), after 3 months of study
treatment (mid-point visit), and after 6 months of study treatment (end visit). The study
coordinator will contact participants every 4 weeks over the course of the study to encourage
adherence to the study protocol. At the 3-month and 6-month visits, compliance with treatment
will be monitored by capsule count and measurement of RBC phospholipid fatty acid
concentrations. Adverse events will be recorded at these visits.

5. Data management and quality control

The Vanderbilt GCRC Informatics Core will be used as a central location for data processing
and analysis. Vanderbilt University has developed software tools and workflow methodology for
electronic collection and management of research study data. (132) REDCap (Research
Electronic Data Capture) is a secure, web-based application that provides an intuitive
interface for users to enter validated data remotely (with automated data type and range
checks), data manipulation audit trails and reporting, and an export mechanism for
end-of-study export of data to common statistical packages.

Inclusion Criteria:

- ≥ 40 and < 80 years of age

- History of 1 or more adenomatous polyps

- Consent to be contacted for future studies

- Participants with known genotype for rs174535 in FADS1

- Prior participation in the Tennessee Colorectal Polyp Study or the Personalized
Prevention of Colorectal Cancer Trial

Exclusion Criteria:

- Previously resected colorectal cancer

- Coronary artery disease or congestive heart failure

- Current metabolic or life-threatening disease

- Currently taking fish oil supplements

- Inability or unwillingness to stop NSAIDs or ASA during the study

- Allergic to fish products

- Diagnosis of inflammatory bowel disease

- Diagnosis of any cancer (except non-melanoma skin cancer)

- Diagnosis of liver or kidney disease

- Pregnant or breast feeding