n-3 and n-6 PUFA are named according to the position of the first double bond from the methyl terminus of the hydrocarbon chain of the molecule. Most of the n-6 and n-3 PUFA are metabolised from precursors, linoleic acid (18 : 2n-6) and α-linolenic acid (18 : 3n-3), respectively, by a series of elongation and desaturation reactions to yield longer, more unsaturated fatty acids (FA) (Karmali, Reference Karmali1996).
n-3 and n-6 PUFA have a number of vital functions in the human body as structural phospholipids of the cell membrane; they modulate membrane fluidity, cellular signalling and cellular interaction (Kelley, Reference Kelley2001; Vancassel et al. Reference Vancassel, Barthelemy, Lejeune, Martineau, Guilloteau, Andres and Chalon2001).
While n-6 FA have potent inflammatory effects, n-3 FA have lesser pro-inflammatory effects, and these two classes of FA compete in the production of inflammatory lipid mediators. These potent immunoregulatory metabolites are synthesised from 20-C PUFA precursors. EPA or arachidonic acid are mobilised from the cell membrane by the action of phospholipase enzymes, especially phospholipase A2 and phospholipase C, and subsequently metabolised by cycloxygenase or lipoxygenase enzymes into PG, thromboxanes and leukotrienes. EPA gives rise to 3-series PG and thromboxanes and 5-series leukotrienes, the difference being the presence of an additional double bond in the structure. Because cell membrane phospholipids normally contain much higher levels of arachidonic acid than of the other 20-C PUFA, arachidonic acid is the most common eicosanoid precursor and gives rise to 2-series PG and thromboxanes and 4-series leukotrienes.
n-3 FA (e.g. EPA and DHA) reduce production of inflammatory cytokines associated with several chronic diseases, anorexia associated with these diseases and anorexia associated with immunotherapy with these cytokines (Meydani, Reference Meydani1996).
Many factors influence tumour induction and cancer growth, including a range of cytokines and growth factors, and genotoxic and oxidative stress. During cancer progression, cell turnover, differentiation and apoptosis are impaired. n-3 PUFA have emerged as anti-carcinogenic nutrients of potential benefit in cancer, through regulation of either enzyme expression and/or activity and end-product concentrations, or by modulating the levels of available precursors for biosynthetic pathways. They work through several actions to protect against the initiation and early stages of cancer, including decreasing tumour cell proliferation, enhancing tumour cell apoptosis, promoting cell differentiation, and limiting angiogenesis (Roynette et al. Reference Roynette, Calder, Dupertuis and Pichard2004).
During the past 20 years, several dozen studies have investigated the effects of n-3 PUFA on human immune and inflammatory responses. Most of these studies involved supplementing diets with marine oils containing EPA and DHA or purified EPA and DHA. These promising results prompted us to conduct this systematic review, which was first discussed by a panel of experts and finally presented in this document.
Materials and methods
The systemised review has been designed according to the Quorum statement (Moher et al. Reference Moher, Cook, Eastwood, Olkin, Rennie and Stroup1999). We located and subsequently analysed the scientific literature available from 1996 to 2006 in several of the most widely used databases, including MEDLINE, EMBASE, The Cochrane Library databases on clinical trials and the online version of the Healthstar database. The search terms used for the review were: fish oil, fatty acid, epa, eicosapentaenoic, docosahexaenoic, omega 3, tumour, neoplasm, cancer, carcinoma, appetite, cachexia, economics, cost analysis, cost benefit and quality of life (QoL). Bibliographies were checked, and experts in the field were contacted for additional studies. The realisation of a quantitative analysis of the selected papers was not considered, due to the small sample of the studied populations and also because the heterogeneousness of the papers would not assure the reliability of the results. The inclusion criteria for studies selected for analysis were those including patients of both sexes aged >18 years with malignant neoplasms associated with cachectic syndrome, a life expectancy >2 months, and not undergoing chemotherapy or radiotherapy at the time of the study. Studies on patients treated with surgery for potentially cachectising gastrointestinal malignancies were also included.
Studies on patients with potentially cachectising concomitant diseases such as renal or heart failure and autoimmune diseases, or patients with potentially hormone-sensitive or emetogenic brain, breast, ovarian, prostrate or endometrial cancer that would prevent proper oral intake were excluded.
The selected studies were analysed by two independent reviewers who coded the results separately and resolved any discrepancies by discussion and consensus between them. When there was no consensus, a third reviewer resolved the differences found. The results were presented to a panel of experts selected from among the different medical specialties related to the subject of the study (oncology, endocrinology, general surgery and intensive medicine), all of whom were opinion leaders in the nutritional implications in their field.
Studies existing in the medical literature with a higher level of scientific evidence were selected, including meta-analyses of clinical trials, clinical trials and prospective observational studies with large samples (on those issues where clinical trials were not located), economic evaluations of health technologies and qualitative studies. The study design aspects assessed to consider a study of high scientific quality were: randomised assignment of control and experimental groups, existence of a concurrent control group, prospective design, use of blinding, and sample size sufficient to detect significant differences. Studies conducted in animals or in languages other than English, French or Spanish were not considered.
The studies collected could analyse biochemical, clinical and functional parameters following nutritional support with supplements enriched with n-3 FA EPA and DHA over a variable period up to 3 months. The following clinical outcomes were noted: nutritional status, tolerance and gastrointestinal complications, incidence of post-surgical infection, length of hospital stay and survival. The functional parameters collected were: appetite, disease-specific and overall QoL (Karnofsky scale and ECOG performance status scale). Laboratory parameters analysed included plasma FA composition, pro-inflammatory response mediators: TNF, IL-1, IL-6, PG, and C-reactive protein as a marker of inflammatory response.
Two independent reviewers selected and analysed the information collected. Studies were classified according to the level of evidence based on the table prepared by the Agència d'Avaluació de Tecnologia Mèdica (Jovell & Navarro Rubio, Reference Jovell and Navarro Rubio1995) (Table 1). Based on the analysis and evaluation of the evidence collected, recommendations were subsequently formulated on the suitability of the conditions for adoption of a health technology or intervention according to the recommendation grades established by the Canadian Task Force (http://www.ctfphc.org) (Table 2). Finally, the quality of the clinical trials collected was assessed using the scale proposed by Jadad et al. (Reference Jadad, Moore, Carroll, Jenkinson, Renolds, Gavaghan and McQuay1996) which is illustrated in Fig. 1.
Level of evidence for study classification according to the Agència d'Avaluació de Tecnología Mèdica (Jovell & Navarro Rubio, Reference Jovell and Navarro Rubio1995)
Recommendation grades for specific clinical preventive actions according to the Canadian Task Force (CTF)
Grading of clinical trial quality (after Jadad et al. Reference Jadad, Moore, Carroll, Jenkinson, Renolds, Gavaghan and McQuay1996).
Results and evidence recommendations
Fifty clinical trials and prospective studies were reviewed and analysed. Of these, only seventeen met the selection criteria. The studies that were selected for evaluation are summarised in Table 3. Relevant aspects regarding the design, results and conclusions of each study are shown in the table. In the level of evidence column, the quality grade assigned to each of the clinical trials is indicated. The final assessment and evidence grade were agreed by consensus by the panel of experts. A summary of the quality of the studies by the different parameters analysed is shown in Table 4.
Summary of the selected studies
‘Level of evidence’ refers to study characteristics according to type of design, as described in Table; CT, clinical trial; E group, experimental group; C group, control group; NA, not applicable; M, megestrol acetate; WN, well nourished; MN, malnourished; PBMC, peripheral blood mononuclear cells.
* Quality grade agreed by expert panel.
Results by parameters analysed
* As described in Table 1.
Is the provision of supplements containing n-3 fatty acids beneficial in cancer patients?
Yes, in patients with advanced cancer and weight loss.
Recommendation grade: B
In which type of tumours?
Pancreas and upper digestive tract cancer
There are currently no studies on other types of neoplasms, although a recent study (Jatoi et al. Reference Jatoi, Rowland, Loprinzi and Sloan2004) showed good results in terms of weight maintenance or gain in various solid tumours.
Recommendation grade: B
Is there an appropriate fatty acid pattern?
The higher quality studies (Kenler et al. Reference Kenler, Swails and Driscoll1996; Fearon et al. Reference Fearon, Von Meyenfeldt and Moses2003; Burns et al. Reference Burns, Halabi, Clamon and Kaplan2004; Jatoi et al. Reference Jatoi, Rowland, Loprinzi and Sloan2004) used combinations of EPA and DHA in a 2 : 1 ratio. Some studies (Wigmore et al. 1996; Gogos et al. Reference Gogos, Ginopoulos, Salsa, Apostolidou, Zoumbos and Kalfarentzos1998; Zuijdgeest-Van Leeuwen et al. Reference Zuijdgeest-Van Leeuwen, Dagnelie and Wattimena2000) with positive results used EPA only.
Recommendation grade: B
What is the recommended dose?
Available data recommend the administration of 1·5 to 2 g EPA/day. Although Fearon et al. (Reference Fearon, Von Meyenfeldt and Moses2003) and Wigmore et al. (1996) found no advantages in exceeding 2 g/day, Burns administered much higher doses with good results (4·7 g EPA/day).
Recommendation grade: B
What is their tolerance and safety?
The incidence of adverse effects at the recommended doses is low. Although there is no direct recommendation, better tolerance has been reported when EPA was administered as part of a low-fat nutritional formula (Fearon et al. Reference Fearon, Von Meyenfeldt and Moses2003) than as concentrated capsules (Burns et al. Reference Burns, Halabi, Clamon and Kaplan2004, Bruera et al. Reference Bruera, Strasser, Palmer, Willey, Calder, Amyotte and Baracos2003).
Recommendation grade: B
How long should they be given?
One higher quality study in patients with advanced pancreatic cancer (Fearon et al. Reference Fearon, Von Meyenfeldt and Moses2003) recommends at least 8 weeks. Positive clinical effects have been observed in post-surgical patients from 1 week of treatment. Some biological markers have shown improvement after 1 week of nutritional treatment.
Given the low incidence of adverse effects at the recommended doses, treatment can be maintained for as long as there are objective benefits.
Recommendation grade: C
Are there markers of efficacy and effectiveness?
It is recommended to assess efficacy using anthropometric measures (weight, BMI), bioimpedance (lean mass), functional parameters and QoL scales.
Recommendation grade: B
Are there any advantages in terms of survival?
The study by Gogos et al. (Reference Gogos, Ginopoulos, Salsa, Apostolidou, Zoumbos and Kalfarentzos1998) showed statistically significant advantages in survival (P < 0·025). These differences increased (P < 0·001) when well-nourished patients fed supplement were compared to malnourished patients fed placebo. The Fearon et al. (Reference Fearon, Von Meyenfeldt and Moses2003) study did not find differences in survival as well as the study by Jatoi et al. (Reference Jatoi, Rowland, Loprinzi and Sloan2004) when compared with megestrol acetate, anyhow it was not placebo controlled.
Recommendation grade: C
Discussion
The anti-inflammatory effects of n-3 PUFA and their beneficial effects for cardiovascular health have prompted a number of studies on the management of autoimmune and inflammatory diseases. Fish-oil supplementation has reduced the symptoms of several diseases such as Ig-A nephropathy, asthma, and Crohn's inflammatory-bowel disease. Other conditions that were improved with n-3 FA supplementation include: arthritis, ulcerative colitis, lupus, psoriasis, cystic fibrosis, re-stenosis after angioplasty, and renal functions after liver or kidney transplantation. It also decreased circulating levels of IL-1, IL-2, IL-4, IL-6, TNF, and γ-interferon in patients with advanced colorectal cancer, improved survival of patients with cancers of different organs, and reduced weight loss in patients with pancreatic cancer (Wigmore et al. 1996). The intake of fish oils in those studies ranged from 4 to 20 g/d for 6 weeks to 60 months (Kelley, Reference Kelley2001; Roynette et al. Reference Roynette, Calder, Dupertuis and Pichard2004).
According to our systematic review, in neoplastic diseases the provision of diets supplemented with n-3 FA showed measurable benefits in the different biochemical, clinical and functional parameters considered. The majority of the studies support with varying grades of recommendations supplementation with diets rich in n-3 FA in this type of patients. Elia et al. (Reference Elia, Van Bokhorst-de van der Schueren, Garvey, Goedhart, Lundholm, Nitenberg and Stratton2006) reached similar conclusions in their systematic review for enterally fed patients as well as those taking oral supplements with a decrease in the complications, especially the infectious ones, as well as a decrease in hospital stays and an improvement in the inflammatory markers.
The most appropriate proportion of n-6 and n-3 FA is a controversial subject. Because of their respective anti-inflammatory and pro-inflammatory natures, the n-6 : n-3 FA ratio should be kept as low as possible (Xia et al. Reference Xia, Wang and Kang2005) One of the possible ways to achieve this is to keep the proportion of lipids low, which additionally has the positive effect of achieving a reduced satiating effect. In the studies considered, the n-6 : n-3 FA ratio was approximately 1 : 3.
Dose of EPA is an issue of great concern. Although some of the best trials have not achieved a dose greater than 1·5 g/d other studies have found increased benefit with higher doses (between 1·5 and 4 g/d). This probably means that under 1·5 g/d there is hardly any effect and that superior doses may obtain improved results.
The negative results obtained in the study by Bruera et al. (Reference Bruera, Strasser, Palmer, Willey, Calder, Amyotte and Baracos2003) should be noted, and, in our opinion, were due to the use of a patient group with tumours of very diverse origin (we found clear efficacy in cancer of the pancreas and upper digestive tract) who were treated nutritionally for a short period of 2 weeks, although they later received unmasked treatment for a longer period (treatment time was 7 weeks in the study by Barber et al. (Reference Barber, Ross, Voss, Tisdale and Fearon1999a) and 8 weeks in the study by Fearon et al. (Reference Fearon, Von Meyenfeldt and Moses2003).
Jatoi et al. (Reference Jatoi, Rowland, Loprinzi and Sloan2004) showed increased weight and appetite in their study of patients treated with EPA. Both patients with gastrointestinal and lung tumours showed lower percentages of weight gain and appetite than those previously reported by other authors and lower than those observed in their study of megestrol, another drug with proven efficacy as an anti-anorexigenic agent. However, no differences were found versus megestrol with respect to appetite, QoL or overall survival.
Finally, it should be noted that most of the studies analysed reported better tolerance when EPA was administered as part of a low-fat nutritional formula instead of in the form of concentrated capsules. On the other hand the role of DHA separated from EPA has not been clearly ascertained in human subjects as most trials deal with both products in combination.
Conclusions
In cancer patients, supplementation with FA (EPA and/or DHA) in the diet or in the form of concentrated capsules seems to be associated with an improvement in various clinical, biochemical and QoL parameters. Regarding duration of supplementation evidence is conflicting but data suggest that good results can be obtained with prolonged nutrition (8 weeks).
It is recommended that prospective studies be carried out to relate the efficacy of nutritional support with EPA in terms of both clinical parameters (body mass, survival and QoL) and biochemical parameters (plasma levels of EPA, C-reactive protein and PG).
We can conclude that, although prognosis has traditionally been defined in terms of morbidity and mortality, we should currently incorporate multidimensional concepts that include measures of functional status, QoL, patient satisfaction and economic evaluation. This is extremely important in situations where the patients, as in this systematic review, are cancer patients with associated cachexia, since application of any support in such patients must be shown to be effective (Resolution ResAP (2003)3).
We were unable to evaluate economic outcomes or cost-effectiveness/utility in this systematic review due to the lack of this information in the studies used (Voss & Gallagher-Allred, Reference Voss and Gallagher-Allred1996).
Acknowledgements
This study was conducted with an educational grant from Abbott Laboratories, Madrid, Spain.
