Study Questions

We convened a technical expert panel composed of distinguished basic scientists and clinicians with established expertise in omega‐3 fatty acids, human nutrition, dietary assessment methods, cancer biology, and oncology. The technical expert panel advised us on refining the preliminary questions posed to us by AHRQ, determining the proper inclusion/exclusion criteria for the study and the populations of interest, establishing the proper outcomes measures, and conducting the appropriate analyses.

Based on the original questions received from AHRQ and input from our technical expert panel, we addressed the following questions in this study:

Search Strategy

Jessie McGowan, Senior Information Scientist, and Nancy Santesso, Knowledge Translation Specialist, at the University of Ottawa were responsible for developing a common search strategy for omega‐3 fatty acids for the three participating EPCs. Nancy Santesso developed a core omega‐3 search strategy in collaboration with project librarians, biochemists, nutritionists, and clinicians, who also provided biochemical names, abbreviations, food sources, and commercial product names for omega‐3 fatty acids. The literature search was not restricted by language of publication or by study design, in order to increase sensitivity. When possible, the searches were limited to studies involving human subjects. For the SCEPC, this core search strategy was incorporated into a specific search for cancer.

In consultation with our technical expert panel and the task order officer, it was decided that, for the questions pertaining to tumor behavior, i.e., apoptosis, tumor growth, and differentiation, we would conduct a separate search focusing on review articles and meta‐analyses of animal studies and cell culture studies pertaining to both humans and animals.

The following databases were searched: MEDLINE® (1966‐October week 5, 2003), PreMEDLINE® (Nov 7, 2003), EMBASE (1980‐Week 44, 2003), Cochrane Central Register of Controlled Trials (3rd Quarter, 2003), CAB HEALTH® (1973‐October 2003). All of these databases were searched using the OVID interface, except CAB HEALTH, which was searched through SilverPlatter. Any duplicate records were identified and removed within each search question using Reference Manager® software. The citations obtained from these literature searches were sent to the SCEPC via E‐mail. In addition, we sent letters to industry experts recommended by the Office of Dietary Supplements to obtain any unpublished data.

Selection Criteria

Two reviewers independently reviewed each article considered for inclusion in the study. Any disagreements between the reviewers were resolved through consensus. For the questions pertaining to tumor incidence and response to treatment, we included any articles that pertained to the effects of omega‐3 fatty acids on cancer, presented research on human subjects, and reported the results of randomized clinical trials, controlled clinical trials, or cohort/case control studies. We were unable to identify human studies that assessed the effects of omega‐3 fatty acids on tumor behavior, i.e., cell growth, differentiation, and apoptosis. Hence, to evaluate the effects of omega‐3 fatty acids on tumor behavior, we turned to the animal and cell culture literature. The initial intent was to summarize only meta‐analyses and systematic reviews; however, because a total of only one meta‐analysis and four systematic reviews were identified, the decision was made to summarize all relevant reviews. Language was not a barrier to inclusion.

Data Extraction and Analysis

For each article on tumor incidence and response to treatment included in the study, two reviewers independently extracted data about:

• Trial design.
• Outcomes of interest.
• The quality of the trial.
• Number and characteristics of the patients.
• Details on the intervention, such as the dose, frequency, and duration.
• Types of outcome measures.
• Adverse events.
• The elapsed time between the intervention and outcome measurements.

Any disagreements between the reviewers were resolved through consensus. For each article, we then evaluated the quality of the design and execution of trials using a system developed by Jadad;³ determined a combined applicability grade based on applicability to the U.S. population and health state; performed a meta‐analysis of those studies that sufficiently assessed interventions, populations, and outcomes to justify pooling; and performed a qualitative analysis of the remaining studies. The reviews and meta‐analyses on tumor behavior were reviewed and summarized by the medical editor, a nutritional biochemist.

Tumor Incidence and Outcomes after Cancer Treatment

We screened 4,834 article titles. From these article titles, we chose to review 1,210 full‐text articles. Of these full‐text articles, 356 met our selection criteria and were chosen for data extraction. After data extraction, 52 articles met our inclusion criteria: 33 reported on cancer incidence, and 19 reported on cancer treatment (all 19 reported on surgery). The 19 cohorts that participated in the studies of tumor incidence varied widely with respect to demographics and intake of omega‐3 fatty acids.

Omega‐3 Fatty Acids and Tumor Incidence

Among 43 risk ratios calculated across the 19 cohorts for 11 different types of cancer and 5 different ways to assess omega‐3 fatty acid consumption (fish consumption, total omega‐3 consumption, alpha‐linolenic acid [ALA] consumption, docosahexaenoic acid [DHA] consumption, and eicosapentaenoic acid [EPA] consumption), only four are statistically significant.

Significant associations between omega‐3 consumption and cancer risk were reported for lung cancer in two studies; for breast cancer in one; for prostate cancer in one; and for skin cancer in one. However, for lung cancer, one of the significant associations was for increased cancer risk and the other was for decreased risk (four other risk ratios were not significant for lung cancer). For breast cancer, five other estimates did not show a significant association. Only one study assessed skin cancer risk. No effects were reported for cancers of the aerodigestive tract, bladder cancer, colorectal cancer, lymphoma, ovarian cancer, pancreatic cancer, or stomach cancer. Thus, omega‐3 fatty acids do not appear to decrease overall cancer risk.

Temporal and/or Dose‐Response Relationship Between Tumor Incidence and Omega‐3 Fatty Acid Intake

Data were insufficient to permit assessment of a temporal or dose‐response relationship.

Evidence for Involvement of Genes for Omega‐3 Fatty Acid Transport or Metabolism

No studies were identified that investigated the role of omega‐3 fatty acid transport or metabolism genes in any putative effect of omega‐3 fatty acids on tumor incidence.

Evidence for Dependence on Intake of Antioxidants or Other Bioactive Food Components

No studies were identified that allowed this question to be answered.

Evidence for Modification of Response to Omega‐3 Fatty Acids by Immune Status

No studies were identified that examined the possible modification of the effect of omega‐3 fatty acids by immune status.

Effect of Omega‐3 Fatty Acids on Clinical Outcomes

We identified 19 studies from which the effect of omega‐3 fatty acids on clinical outcomes after cancer therapy could be ascertained, all of which pertained to patients who had undergone cancer surgery for upper gastrointestinal malignancies. We did not identify any studies that assessed the effects of omega‐3 fatty acids on clinical outcomes after chemotherapy or radiation surgery. Among the identified studies, 14 described the effect on post‐operative complications, 13 on hospital length of stay, 10 on mortality, 11 on nutrition and three on weight. In pooled analyses, omega‐3 fatty acids had no effect compared to placebo on post‐operative complications, hospital length of stay, or mortality. With the exception of one study that demonstrated higher mean nitrogen intake for subjects treated with omega‐3 fatty acids relative to placebo, no significant effect on nutrition or weight loss was observed.

Evidence for Dependence of Effects on Clinical Outcomes on Intake of Antioxidants or Other Bioactive Food Components

No studies were identified that allowed this question to be answered.

Evidence for Modification of Effects on Clinical Outcomes by Immune Status

No studies were identified that examined the possible modification of the effect of omega‐3 fatty acids on clinical outcomes by immune status.

Tumor Behavior

To assess the effects of omega‐3 fatty acids on tumor growth, differentiation, and apoptosis in animal and in vitro models, we screened a total of 369 citations, of which 82 were considered relevant. Of those 82, 60 could be retrieved. Of the 60, 27 were accepted for further review because they reviewed the effects of omega‐3 fatty acids (added to the diet or to cell cultures) on cancer development, apoptosis, or cell differentiation in laboratory animals or cell culture systems.

Although much of the evidence favored a role for dietary omega‐3 fatty acid enrichment in the inhibition or prevention of colon, mammary, pancreatic, and prostate tumor growth, at least in some animal models, the quality of the reviews is not sufficient to permit strong conclusions to be drawn.

Evidence was presented in a small number of reviews that omega‐3 fatty acids can stimulate cellular differentiation and apoptosis, two proposed mechanisms for the inhibition of tumor development and proliferation; however, the evidence is insufficient to assess the relevance of these findings.

Evidence for an Inverse or Temporal Relationship with Intake

Insufficient evidence was presented to assess dose‐response effects or to ascertain the stage of tumor development that might be affected by omega‐3 fatty acids.

Evidence that Genes Involved in Omega‐3 Fatty Acid Transport or Metabolism Influence the Magnitude or Direction of the Influence on Tumor Behavior

Several reviews provided evidence that omega‐3 fatty acids may affect tumor behavior by competing with omega‐6 fatty acids for the enzymes that metabolize them to their bioactive products or by influencing the genes for these enzymes; however, other evidence suggests an effect on intracellular redox state and the integrity of membrane lipids.