Further Research Warranted on Low-fat Diets with Fish Oil Supplementation for Prostate Cancer Patients
Abstract & Commentary
By Dónal P. O'Mathùna, PhD, Dr. O'Mathùna is Senior Lecturer in Ethics, Decision-Making & Evidence, School of Nursing and Human Sciences, Dublin City University, Ireland; he reports no financial relationship to this field of study.
Synopsis: A prospective, randomized controlled trial measured the impact of a low-fat diet that included high levels of omega-3 fatty acids from fish oil on biomarkers for prostate cancer. No significant differences were found for the primary outcome during an interim analysis and the trial was stopped early. Analysis of secondary endpoints showed some significant differences between the groups, although other biomarkers did not differ.
Source: Aronson WJ, et al. Phase II prospective randomized trial of a low-fat diet with fish oil supplementation in men undergoing radical prostatectomy. Cancer Prev Res 25 Oct 2011; Epub ahead of print; DOI:10.1158/1940-6207.CAPR-11-0298.
This phase 2 randomized controlled trial examined whether lowering dietary fat and decreasing the omega-6 to omega-3 fatty acid ratio would delay the progression of prostate cancer. The hypothesis arose from previous research showing such an effect in cell lines and mouse models. In addition, some, but not all, epidemiological studies have shown a correlation between intake of dietary fat in general or omega-3 fatty acids and risk of prostate cancer.1
The proposed mechanism of action for these effects is through reduced insulin-like growth factor (IGF) signaling and changes to omega-6 to omega-3 fatty acids ratios in cell membranes. For this reason, the primary outcome for the trial reported here was the change in fasting serum IGF-1 levels between the two groups. Several secondary outcomes (to be specified below) were examined by measuring biochemical markers.
Participants were recruited from patients diagnosed with localized prostate adenocarcinoma and scheduled for radical prostatectomy at least 4 weeks after entering the trial. Participants were excluded if they had taken 5-alpha reductase inhibitors or several other drugs that might interfere with the study's biomarkers, and had to agree to stop all herbal remedies, nutritional supplements, and COX-2 inhibitors at least 1 week before the study.
The experimental intervention was designed to reduce overall fat consumption and shift the ratio of omega-6 to omega-3 fatty acids to 2:1. This was done via a low-fat diet (15% of calories from fat, 15% protein, 70% carbohydrate that included 39 g/d fiber) and supplementation with fish oil high in omega-3 fatty acids (five 1.1 gram fish oil capsules daily; each capsule contained 200 mg eicosapentaenoic acid and 367 mg docosahexaenoic). The control intervention was a Western diet, consisting of 40% of calories from fat, 15% protein, and 45% carbohydrate that included 15 g/d fiber. The omega-6 to omega-3 ratio was 15:1. All meals for both groups were prepared by the research team and delivered to the participants' homes. Unconsumed food and capsules were collected to measure compliance, which was excellent. Body weight and exercise levels were to remain unchanged. Participation lasted 4 to 6 weeks depending on when the prostatectomy was scheduled.
A power analysis was conducted which estimated that 70 participants were needed to detect a difference of 20% in serum IGF-1 levels between the groups with 80% power. After 48 subjects had completed the trial, an interim analysis showed that there was little chance that significant differences would be found for the primary outcome. Therefore, the study was ended early and secondary endpoint analyses carried out.
No significant differences were found between the groups in serum IGFBP-1, IGFBP-3, insulin, PSA levels, urine prostaglandin levels, COX-2 levels, or angiogenesis (CD-31). Comparing the Western diet and low-fat groups, respectively, triglyceride levels decreased by 13.6 and 56.5 mg/dL (P = 0.03) and cholesterol levels by 10.4 and 31.7 mg/dL (P = 0.02). Significant differences were found in the membrane levels of omega-6 and omega-3 fatty acids in benign and malignant prostate tissues and red blood cells. For example, the omega-6:omega-3 ratio in malignant tissues was 5.02 in the Western diet group and 2.68 in the low-fat group (P < 0.001). Proliferation of malignant cells was measured using Ki-67 and TUNEL methods. The former test showed significantly less proliferation in the low-fat group (P < 0.05), while the latter test showed no significant differences. Another proliferation test using subjects' serum and prostate cancer cell lines found significantly reduced proliferation in the low-fat group (P = 0.039).
The researchers concluded that the significant changes they found warranted further clinical trials with membrane omega-6:omega-3 ratios and malignant cell proliferation as the primary endpoints. Because these were not primary outcomes, however, they considered these results to be "hypothesis generating." If their current findings are validated, long-term trials of this low-fat, high omega-6 fatty acid intervention would be warranted.
As a Phase 2 trial, this study was relatively small and short, designed to show whether large-scale prospective trials were warranted. The study was rigorously designed and reported in great detail. The authors gave the rationale for their study design, based on preclinical studies conducted by their team and others. The lack of significant impact on IGF, COX-2, and prostaglandin outcomes raised questions about their postulated mechanism of action. The mechanism for the significant differences found in some of the cancer cell proliferation tests is unknown.
The authors discussed the study's strengths and limitations. Foremost among these was the study's short duration and their failure to recruit sufficient subjects as determined by their power analysis. The combined nature of the intervention meant that the influence of either the low-fat diet or the fish oil supplements could not be separated. In addition, the difference in carbohydrate content of the two diets may have had an impact, especially as this led to significant differences in the groups' fiber intake. The low-fat diet led to significantly greater reductions in cholesterol levels, and this could affect prostate cancer growth.
The most serious limitation with this study is the complex interplay of multiple factors and what is called multiple hypothesis testing.2 When statistical significance is set at P = 0.05, there is a 1 in 20 chance of getting a false-positive. The chance of a false-positive increases with each additional outcome tested. For example, with five outcomes, the calculated chance of a false-positive is almost 1 in 4.2 This study had numerous secondary outcomes, so it is not surprising that some were significantly different. Statistical tests (e.g., the Bonferonni Correction) are available to take account of these factors to minimize the risk of false-positives, but these were not reported.
Although the original study authors were cautious about the study's implications and limitations, other reports have been much more enthusiastic. The study was conducted by researchers at the UCLA Jonsson Comprehensive Cancer Center. The Center's website posted a news item about the study and this publication.3 The headline read that the diet intervention "Slows Growth of Prostate Cancer Cells." Only the significant results found in the study were discussed, and there was no mention that the study was stopped early because its primary outcome was unlikely to show significant benefit. This highlights the importance of studying original publications and not relying on news summaries of reports. As the primary investigator has stated, dietary changes for prostate cancer patients cannot be recommended on the basis of this study, although it does provide support for further research in this area.
1. Szymanski KM, et al. Fish consumption and prostate cancer risk: A review and meta-analysis. Am J Clin Nutr 2010;92:1223-1233.
2. O'Mathùna DP, et al. Critically appraising quantitative evidence. In: Evidence-Based Practice in Nursing and Healthcare: A Guide to Best Practice. 2nd ed. Ed. Melnyk B, Fineout-Overholt E. Philadelphia: Lippincott Williams & Wilkins, 2010; 81-134.
3. Anonymous. You are what you eat: Low-fat, fish oil diet slows growth of prostate cancer cells. UCLA Jonsson Comprehensive Cancer Center. Oct. 25, 2011. Available at http://www.cancer.ucla.edu//Index.aspx?page=644&recordid=532.