Phase II study of high-dose fish oil capsules for patients with cancer-related cachexia^†

Abstract

BACKGROUND

The authors undertook a multiinstitutional Phase II cooperative group study to examine the potential of oral fish oil fatty acid supplements administered at high doses to slow weight loss and to improve quality of life in patients with malignancy-related cachexia.

METHODS

Patients with advanced malignancy and weight loss ≥ 2% of body weight in the preceding month took concentrated, high-dose omega-3 fatty acid capsules (7.5 g eicosapentaenoic acid plus docosahexaenoic acid for a 70 kg individual) that were supplied by the National Institutes of Health.

RESULTS

Forty-three patients with moderate or severe malnutrition were enrolled. The median time receiving treatment was 1.2 months. For the 36 patients who took at least 1 capsule and did not have edema, there was a weight change ranging from −6.2 kg to +3.5 kg and an overall median weight loss of 0.8 kg. Twenty-four patients had weight stabilization (a gain of ≤ 5% or a loss of < 5%), 6 patients gained > 5% of their body weight, and 6 patients lost ≥ 5% of their body weight. There was marked variability in the tolerability of the capsules, and many patients had gastrointestinal side effects. There was a correlation between time receiving treatment and weight gain for the 22 patients who were able to tolerate the capsules for at least 1 month. Quality-of-life scores were superior for patients who gained weight.

CONCLUSIONS

A majority of patients did not gain weight, and in that sense, the results of the study were unfavorable. However, a small but definite subset of patients had weight stabilization or weight gain. This suggests that omega-3 fatty acids have potential utility at the study doses, which were more than twice the doses used in published Phase III studies. Cancer 2004. © 2004 American Cancer Society.

The weight loss and inanition experienced by patients with advanced malignant disease often are as distressing as the pain and anatomic effects of the tumor. It is suspected that cytokines or other chemicals generated by, or in response to, the tumor are the mediators of cachexia.1, 2 Studies in animal models have suggested that omega-3 fatty acids may be effective in alleviating human cancer cachexia. Animals with experimental tumors had less cachexia3, 4 and less degradation of skeletal muscle5 when they were fed omega-3 fatty acids.

There is also evidence from clinical studies that fish oil may have a favorable effect on cancer-related cachexia. Wigmore et al. reported that 26 patients with advanced pancreatic carcinoma who received a fish oil supplement experienced weight stabilization or gained weight.6 Gogos et al. performed a controlled study of 60 patients with solid tumors and found that patients who took omega-3 fatty acid capsules as supplements along with 200 mg vitamin E daily had a longer mean survival and an increased T helper cell–to–T suppressor cell ratio compared with patients who received a placebo.7 In our previous Cancer and Leukemia Group B (CALGB) Phase I dose-finding study, we found that time receiving treatment was associated with weight gain.8

To our knowledge, there have been three Phase III studies of omega-3 fatty acids in the treatment of cancer cachexia. Bruera and colleagues reported on a double-blind controlled study of 60 patients with cancer cachexia, 50% of whom received gelatin fish oil capsules.9 Those authors found no effect on weight, appetite, or fatigue compared with placebo during the 2 weeks that patients took the capsules. In that study, it was determined that the ingestion of 3 g eicosapentaenoic acid (EPA) plus docosahexaenoic acid (DHA) for a brief period (14 days) had no effect on the cachectic symptoms or markers associated with terminal malignancy.

The other two Phase III studies used omega-3 fatty acids incorporated into liquid nutritional supplements. Fearon et al. randomized patients with cachexia as a result of pancreatic carcinoma to receive either a supplement that contained EPA plus antioxidants or the supplement without these additions.10 Both products slowed weight loss to some extent, but patients who received an average dose of 1.54 g EPA per day in their supplements did not have better weight-related outcomes. It is noteworthy that there was a correlation between the number of cans of supplement ingested and improvements in both weight gain and lean body mass in the EPA-enriched product group but not in the control group. In a study conducted by Jatoi and colleagues, patients with malignancy-associated wasting were randomized to receive an EPA-rich supplement, megestrol acetate, or both the supplement and megestrol acetate.11 Patients who took the EPA-enriched supplement did not have improved weight or appetite compared with patients who took megestrol acetate alone. These supplements provide only a limited amount of omega-3 fatty acids, and this may explain why they failed to yield positive effects. In any case, these supplement studies, which used daily doses that were < 50% of the doses used in the current study, do not represent the most powerful tests of the efficacy of fish oil fatty acids.

We performed a multiinstitutional Phase II cooperative group study of concentrated fish oil capsules that were obtained from the National Institutes of Health (NIH). In this study, a 70 kg patient was to receive 4.7 g EPA and 2.8 g DHA (as part of a total of 8.5 g omega-3 fatty acids) per day for > 1 month. We found that if a large dose of omega-3 fish oil fatty acid capsules was tolerable as a dietary supplement, then patients with cancer cachexia could experienced weight stabilization or weight gain.

MATERIALS AND METHODS

Patients

Patients were eligible for the current CALGB study if they were age ≥ 18 years, had a histologic diagnosis of malignancy that was not amenable to curative treatment, and experienced a loss of > 2% of body weight within the 1-month period before enrollment. Adequate renal (creatinine level < 1.5 times the upper limit of normal), liver (serum bilirubin level < 1.5 times the upper limit of normal and aspartate aminotransferase and alkaline phosphatase levels < 3 times the upper limit of normal), and bone marrow (granulocyte count > 1000/μL, hemoglobin > 8 g/dL, and platelet count > 75,000/μL) function were required, as were normal free or total thyroxine levels. Exclusion criteria included edema, ingestion of certain medications (e.g., steroids, dronabinol, megestrol acetate, diuretics), gastrointestinal obstruction, life expectancy < 2 months, pregnancy, ascites, congestive heart failure, metabolic disorders, uncontrolled infection, major surgery within 3 weeks of the study, primary or metastatic brain tumors, an Eastern Cooperative Oncology Group performance status of 3 or 4, or ingestion of omega-3 fatty acid supplements within 4 weeks of study initiation. Concurrent chemotherapy (except for investigational drugs) and radiotherapy were allowed, although abdominal/pelvic radiation was required to have been completed at least 3 weeks before study initiation. Human investigations were approved by local human investigation committees in accordance with assurances filed with the Department of Health and Human Services.

Treatment Plan

Patients received a dose of 0.15 g/kg per day divided into 2 doses: 1 dose taken with breakfast and 1 dose taken with lunch. For a 70 kg patient, this dose would consist of eleven 1 g capsules, which would contain 4.7 g EPA and 2.8 g DHA (as part of a total of 8.5 g omega-3 fatty acids) per day. The first 13 patients accrued to the study received 0.3 g/kg per day, which was the maximum tolerated dose established in our Phase I study.8 However, many patients in the current Phase II study were unable or unwilling to take the number of capsules required to achieve this dose. Therefore, a dose reduction was instituted for the remainder of the study. The treatment plan was to continue supplements for a minimum of 2 months, even in patients who continued to lose weight, to conduct an adequate trial. The use of supportive care drugs, antidiarrhea medications, antiemetics, and analgesics was allowed only if clinically necessary, except that steroids, dronabinol, and megestrol acetate were prohibited. Diuretic use was to be avoided if possible. Compliance with drug dose was monitored via patient questioning. We cannot be absolutely certain that each patient received the reported dose.

Omega-3 Capsules

Fatty acids were provided in soft gelatin capsules of fish oil omega-3 ethyl esters obtained from the NIH/National Oceanic and Atmospheric Administration (Biomedical Test Material Program, National Marine Fisheries Service, Southeast Fisheries Science Center, Charleston Laboratory, Charleston, SC). The capsules were stored in closed bottles at refrigerator temperature. Their fatty acid composition was determined using gas chromatography, as described previously.12, 13

Quality of Life

Quality-of-life (QOL) data were collected using the 53-item version of the Functional Assessment of Anorexia/Cachexia Therapy (FAACT [Version 3]) scales, which included questions on appetite and reaction to foods.14

Statistical Methods and Analysis

The primary endpoint chosen was weight change, since it was objective and quantitative. Weight change was defined as the patient's body weight at the time of the last treatment follow-up compared with baseline or reaching the precachexia baseline body weight. An indicator variable was used to define weight change as a gain of > 5%. The null hypothesis was that the weight change (response) probability (P) was ≤ 0.10, with the alternative hypothesis being that the weight change P was > 0.30. This design had a power of 96% and a type I error rate of 0.08. Assuming a 10% dropout rate, the target sample size was 43 patients. According to the protocol, ineligible patients were included in the analyses, but patients who canceled their registration before they received any therapy were not included. Response rates and 95% confidence intervals based on the binomial distribution were calculated for measured weights.

RESULTS

Patient Demographics

Forty-three patients were entered into the trial. The baseline clinical features of the patients are shown in Table 1. The low body mass index (21 kg/m²) and median weight loss (3 kg [4% of body weight]) during the month preceding the study verify that this was a poorly nourished, cachectic group of patients with malignant disease. Three patients did not take the capsules because of disease progression and were not included in the response analysis. Four additional patients were not included in the weight analysis: Three patients had moderate-to-severe edema and required diuretics that precluded evaluation of weight response, and one patient had no weight data recorded.

Table 1. Baseline Clinical Characteristics of Patients Entered in the Current Trial

Characteristic
Total no. of patients entered (%)	43 (100)
Age (yrs)
Median	67
Range	42–84
Gender (male:female)	29:14
Yrs since diagnosis
Median	1
Range	0–10
ECOG performance status (%)
0–1	22 (51)
2	21 (49)
Solid tumorsaa Patients with solid tumors included 8 patients with colorectal carcinoma; 6 patients with nonsmall cell lung carcinoma; 3 patients with small cell lung carcinoma; 4 patients with unknown primary tumors; 2 patients each with sarcoma, breast carcinoma, gastric carcinoma, renal carcinoma, neuroendocrine tumor, and myelodysplastic syndrome; and 1 patient each with pancreatic carcinoma, anal carcinoma, chronic lymphocytic leukemia, head and neck carcinoma, hepatic carcinoma, myeloma. Solid tumor type was not recorded for 4 patients. (%)	38 (88)
Weight at entry (kg)
Median	64
Range	40–89
Median weight loss over 1 mo (kg)	2.7 (4% overall)
Body mass index (kg/m^2)
Median	21.1
Range	14.6–33.7
Prior therapy (%)
Chemotherapy	24 (64)
Chemotherapy and radiotherapybb No patient received radiotherapy alone.	14 (33)
Other	3 (7)
None	2 (5)
Serum prealbumin level (mg/dL)cc The reference range at The University of Iowa as measured by nephelometry was 18–45 mg/dL.
Median	16
Range	5–58
Serum albumin level (mg/dL)dd The reference range at The University of Iowa was 3.5–5.0 g/dL.
Median	3.5
Range	0.9–4.5
Creatinine height index (%)ee The creatinine height index is the 24-hour urine creatinine excretion divided by the expected 24-hour creatinine excretion for a normal adult of the same height. This calculated quantity expressed as a percentage (n = 21).
Median	62
Range	30–107
Urine creatinine (g per 24 hrs)ff The reference range at The University of Iowa was 1.0–2.0 g/24 hours for males and 0.8–1.8 g/24 hours for females (n = 22).
Median	0.90
Range	0.4–1.71

• ECOG: Eastern Cooperative Oncology Group.
• a Patients with solid tumors included 8 patients with colorectal carcinoma; 6 patients with nonsmall cell lung carcinoma; 3 patients with small cell lung carcinoma; 4 patients with unknown primary tumors; 2 patients each with sarcoma, breast carcinoma, gastric carcinoma, renal carcinoma, neuroendocrine tumor, and myelodysplastic syndrome; and 1 patient each with pancreatic carcinoma, anal carcinoma, chronic lymphocytic leukemia, head and neck carcinoma, hepatic carcinoma, myeloma. Solid tumor type was not recorded for 4 patients.
• b No patient received radiotherapy alone.
• c The reference range at The University of Iowa as measured by nephelometry was 18–45 mg/dL.
• d The reference range at The University of Iowa was 3.5–5.0 g/dL.
• e The creatinine height index is the 24-hour urine creatinine excretion divided by the expected 24-hour creatinine excretion for a normal adult of the same height. This calculated quantity expressed as a percentage (n = 21).
• f The reference range at The University of Iowa was 1.0–2.0 g/24 hours for males and 0.8–1.8 g/24 hours for females (n = 22).

Twenty-one patients completed 24-hour urine collection for the measurement of the creatinine height index, which is a measure of muscle mass; in patients with protein-energy malnutrition, there is a decrease in this index.15 The median creatinine height index value for patients in the current study was 62%, which corresponds to a severe degree of muscle mass depletion and malnutrition.

Duration of Treatment, Dose Administered, and Omega-3 Capsule Stability

The median treatment time was 1.2 months (range, 0.5 days to 3.1 months). Patients typically received the assigned dose. One patient who was recruited early in the study and was assigned to receive 0.3 g/kg experienced dose reduction to 0.15 g/kg for the last 4 days of a 31-day treatment period as a result of side effects. One patient who was assigned to receive 0.15 g/kg experienced a dose increase to 0.2 g/kg for the last 20 days of a 77-day treatment period, as allowed by the protocol, because the patient's weight failed to increase. Two patients who were assigned to receive 0.15 g/kg received 0.2 g/kg and 0.1 g/kg, respectively, in error.

Figure 1 lists the specific fatty acids contained in the capsules, as assayed in our laboratory. The omega-3 fatty acids EPA (20:5 omega-3) and DHA (22:6 omega-3) accounted for approximately 70% of the fatty acids in the fish oil capsules. The contents of the capsules, including antioxidants (α-tocopherol, γ-tocopherol, and tertiary butylhydroquinone), as assayed by the National Marine Fisheries Service, have been reported in full elsewhere8; the fatty acid content reported in that study was similar to the results obtained in our laboratory.

Figure 1

Fatty acid composition of the fish oil capsules was determined using gas chromatography. Unsaturated fatty acids are named according to the following system: number of carbon atoms:number of double bonds, with the number immediately following the Greek letter omega (ω) indicating the number of carbon atoms from the methyl terminal end of the molecule to the first double bond. EPA: eicosapentaenoic acid; DHA; docosahexaenoic acid.

We also determined the fatty acid composition of the capsules at four subsequent time points during the study interval (Fig. 1). In the current study, the capsules were stored at refrigerator temperature and shipped at ambient temperature. The highly polyunsaturated fish oil fatty acids, which are susceptible to oxidation because of their large number of double bonds, were stable over the 6 years from the beginning of the previously reported Phase I study to the end of the current study.

Weight Response to Fish Oil Capsules

Figure 2 shows the weight change trend for all patients who took at least one capsule and did not have edema that interfered with the weight analysis, except for one patient who was ineligible due to an increase in weight before the start of therapy. The pretreatment downward trend was reversed for some (but not all) patients. For the 36 patients who took at least 1 capsule, had available weight data, and did not have interfering edema, the observed change in weight ranged from −6.2 kg to +3.5 kg. Twelve patients gained weight (range, +0.1 kg to +3.5 kg), and 22 patients lost weight (range, −0.4 kg to −6.2 kg), for an overall median weight loss of 0.8 kg.

Figure 2

Weight-change trends. Circles indicate weight measurements for individual patients. The regression line starting from Day 1 of therapy has an r value of +0.27.

Table 2 shows changes in patient weight during the study according to the criteria established before the study started. Six patients (17%; 95% confidence interval, 0.06–0.33) gained > 5% body weight or reached their typical baseline weight, and an additional 24 patients (66%) had stabilization of their weight (≤ 5% gain or < 5% loss). Therefore, it can be stated that a subset of patients had a weight response to the supplements. Six patients continued to experience weight loss of > 5% (range, 3.4–6.2 kg). All recorded weights, even after completion of therapy, are shown. For example, the patient who lost > 10 kg after > 100 days had completed therapy on Day 49. Figure 3 shows a correlation between the number of days receiving fish oil capsules and weight gain for patients who took the capsules for at least 30 days.

Table 2. Weight Response to Fish Oil Capsules

Responseaa Responses shown were relative to on-study body weight. Includes one patient who was ineligible due to inability to swallow capsules (this patient also was unevaluable due to lack of weight data) and one patient who was ineligible due to failure to meet prestudy weight loss criteria.	No. of patients (%)
Gained weightbb Weight gain > 5% or reached premorbidity body weight.	6 (17)
Lost weightcc Weight loss ≥ 5%.	6 (17)
Stable weightdd Weight gain ≤ 5% or weight loss < 5%.	24 (66)
Unevaluableee Never started treatment (n = 3); edema precluded analysis (n = 3); weight data unavailable (n = 1).	7 (—)
Total	43 (—)

• a Responses shown were relative to on-study body weight. Includes one patient who was ineligible due to inability to swallow capsules (this patient also was unevaluable due to lack of weight data) and one patient who was ineligible due to failure to meet prestudy weight loss criteria.
• b Weight gain > 5% or reached premorbidity body weight.
• c Weight loss ≥ 5%.
• d Weight gain ≤ 5% or weight loss < 5%.
• e Never started treatment (n = 3); edema precluded analysis (n = 3); weight data unavailable (n = 1).

Figure 3

Correlation between weight change and days of therapy for patients who took fish oil capsules for at least 1 month. Regression line was drawn using SigmaPlot software (SPSS Inc., Chicago, IL). The 95% confidence interval for r is 0.093–0.758 (P = 0.0191).

Survival and QOL

The median follow-up for surviving patients was 3.7 months (interquartile range, 2.3–3.9 months). Thirty-six patients died of progressive disease. The estimated median survival of all 43 patients was 3.1 months (95% confidence interval, 1.9–4.1 months).

Thirty-eight of 43 patients filled out a QOL assessment at baseline. Twenty patients had completed at least one posttreatment QOL assessment. With the exception of the additional concerns of cachexia, there were no differences in baseline characteristics or QOL scores between patients who completed only the baseline assessment and patients who completed the baseline and follow-up FAACT assessments. Patients who gained weight had higher overall FAACT scores and had overall Functional Assessment of Cancer Therapy-General Well-Being subscale scores (physical well-being, emotional well-being, functional well-being, and additional concerns) at follow-up compared with patients who did not gain weight (Table 3).

Table 3. Median and Interquartile Range Values for Changes from Baseline in Postbaseline Scores for Patients who Completed at Least One Postbaseline Functional Assessment of Anorexia/Cachexia Treatment

Scale	Median (interquartile range)
	No weight gainaa Weight loss ≥ 5% or stabilization of weight. (n = 13)	Weight gainbb Weight gain > 5% or achievement of premorbidity body weight. (n = 6)	Total (n = 19)
FAACT	−0.2 (−27.6–19.4)	7.6 (2.3–18.2)	3.3 (−14.0–18.2)
FACT-G	−2.0 (−10.4–13.5)	3.2 (1.9–5.5)	1.9 (−7.8–13.5)
Physical well-being	−2.2 (−5.0–0.0)	1.0 (0.0–3.0)	−1.0 (−14.0–3.0)
Social/family well-being	−1.0 (−4.0–2.0)	−2.5 (−2.7 to −2.3)	−2.4 (−4.0–2.0)
Relationship with physician	0.0 (−1.0–0.0)	0.0 (0.0–0.0)	0.0 (0.0–0.0)
Emotional well-being	0.0 (−3.0–3.0)	2.0 (1.0–2.0)	1.0 (−2.0–3.0)
Functional well-being	0.0 (−3.0–1.0)	0.0 (−2.0–2.0)	0.0 (−2.0–2.0)
Additional concerns	−3.5 (−13.0–8.5)	4.5 (−1.0–19.0)	0.0 (−6.8–11.0)

• FAACT: Functional Assessment of Anorexia/Cachexia Treatment scale; FACT-G: Functional Assessment of Cancer Therapy General Well-Being scale.
• a Weight loss ≥ 5% or stabilization of weight.
• b Weight gain > 5% or achievement of premorbidity body weight.

Side Effects and Early Termination of Treatment

Table 4 shows the side effects reported by the 40 patients who took at least 1 capsule. Gastrointestinal symptoms of nausea, emesis, diarrhea, and excessive belching or flatulence were common toxicities. There was considerable variability in the ability of the patients to tolerate the fatty acids, and some patients had little or no side effects. Patients were intended to continue taking the capsules for at least 2 months (nonresponders) or 3 months (responders). Table 5 summarizes reasons for terminating treatment. Ten patients received treatment for the prescribed length of time: 6 completed 3 months of therapy, and 4 continued losing weight over a 2-month period. Eight patients had progressive disease that precluded further omega-3 capsule therapy, and two patients died during the study. Many patients found the capsules difficult to ingest because of their size or difficult to tolerate because of gastrointestinal side effects. Nineteen patients ended therapy early because of side effects; these patients felt that they were not benefiting or were discouraged about their disease (n = 16) or ended treatment for unstated reasons (n = 3).

Table 4. Side Effects of Fish Oil Ethyl Ester Soft Gelatin Capsules in Order of Frequency

Symptom	No. of patients (%)
Nauseaaa Side effects were categorized according to the Cancer and Leukemia Group B Expanded Common Toxicity Criteria as follows: Grade 1, mild (n = 8); Grade 2, moderate (n = 2); Grade 3, severe (n = 1), Grade 4, life-threatening (n = 0); and Grade 5, lethal (n = 0).	11 (28)
Abnormal taste in mouth	10 (26)
Abnormal taste of foodbb N = 39.	9 (23)
Excessive belchingbb N = 39.	9 (23)
Excessive flatulencebb N = 39.	9 (23)
Edemacc Not likely to have been related to study drug. dd Six patients had Grade 1 edema, one patient had Grade 2 edema, and one patient had Grade 3 edema.	8 (20)
Emesisee Five patients had Grade 1 emesis, two patients had Grade 2 emesis, and one patient had Grade 3 emesis.	8 (20)
Diarrheaff Three patients had Grade 1 diarrhea, two patients had Grade 2 diarrhea, one patient had Grade 3 diarrhea, and one patient had Grade 4 diarrhea.	7 (18)
Abnormal body smellbb N = 39. gg Odor was sensed by patient.	4 (10)
Dyspepsiabb N = 39.	2 (5)

• a Side effects were categorized according to the Cancer and Leukemia Group B Expanded Common Toxicity Criteria as follows: Grade 1, mild (n = 8); Grade 2, moderate (n = 2); Grade 3, severe (n = 1), Grade 4, life-threatening (n = 0); and Grade 5, lethal (n = 0).
• b N = 39.
• c Not likely to have been related to study drug.
• d Six patients had Grade 1 edema, one patient had Grade 2 edema, and one patient had Grade 3 edema.
• e Five patients had Grade 1 emesis, two patients had Grade 2 emesis, and one patient had Grade 3 emesis.
• f Three patients had Grade 1 diarrhea, two patients had Grade 2 diarrhea, one patient had Grade 3 diarrhea, and one patient had Grade 4 diarrhea.
• g Odor was sensed by patient.

Table 5. Events Influencing Duration of Therapy

Event	No. of patients (%)
Completion of protocol (2–3 mos)	10 (23)
Progressive cancer	4 (9)
Adverse medical event/new problem	5 (12)
Death	2 (5)
Withdrawal of consentaa These patients eventually were removed from the protocol before 2–3 months due to side effects or refusal of further therapy (n = 16) or for reasons that were not stated (n = 3).	19 (44)
Failure to begin treatment	3 (7)

• a These patients eventually were removed from the protocol before 2–3 months due to side effects or refusal of further therapy (n = 16) or for reasons that were not stated (n = 3).

Serum Biochemical Nutritional Measurements

Serum prealbumin levels are a marker of malnutrition. This protein has a short half-life, is not affected by vitamin deficiency or hydration state, and is affected less by liver disease than other proteins. The median baseline serum prealbumin level was 16 mg/dL (n = 33). Nineteen patients among those with baseline values available had at least 1 follow-up value determined during omega-3 supplementation; the median change observed was a decrease of 3 mg/dL (mean decrease, 2.7 mg/dL). Only 5 patients had increases (range, 1–12 mg/dL), and for the group as a whole, there did not appear to be a relation between weight change and changes in prealbumin levels.

Thirty-three patients had a serum albumin level determined at baseline and at least 1 subsequent visit: 11 patients had increased levels, whereas 22 had decreased levels. The median change was a decrease of 0.4 mg/dL.

DISCUSSION

Using criteria for response that were established during the initial design of the study, equal numbers of patients gained weight and continued to lose weight. The majority had stabilization of weight in association with taking the omega-3 capsules. Many of the patients in the current study who did not respond were unable to tolerate the fatty acid capsules due to the omega-3 fatty acids themselves or due to the size of the capsules. When the weight change for the entire group of patients was examined, there was a median loss of 0.8 kg. At first assessment, this may suggest the lack of an effect. However, this statistic includes patients who took only a few capsules and patients who had a truncated treatment course. On entry into the study, patients were losing weight at a median rate of 4% body weight per month; thus, during the > 1.2-month median duration of treatment, it was predicted that patients would have at least a further 4.8% loss (1.2 months × 4% per month) if they were untreated and if their weight loss continued at the same rate. However, for the entire group, the median rate of weight loss was 1.2% (0.8 kg median loss ÷ 67 kg median baseline weight × 100), which was less than the 4.8% loss that was predicted.

Recently, Bruera et al. performed a double-blind controlled study of omega-3 fatty acid capsules that were administered to patients with cancer cachexia and found no effect on weight.9 Their study had critical differences in design compared with the current study. First, Bruera et al. administered a daily dose of 3 g EPA plus DHA, compared with > 7 g per day (for a 70 kg patient) in the current study. Second, they administered the supplement for 14 days to all patients, compared with a median duration of 1.2 months and a maximum of 3 months (for patients who tolerated the capsules) in the current study. Any dissimilar outcomes may be explained by these major differences in omega-3 dose level and treatment duration.

An adequate daily dose of omega-3 fatty acids administered over a sustained period may be necessary to affect weight gain in patients with malignant disease. In the current study, a 70 kg patient taking 0.15 g/kg (assuming the use of 1 g capsules) actually would receive 7.5 g DPA plus DHA per day. Studies in which weight loss in tumor-bearing animals was reversed used doses of 1.5–2.5 g/kg.5, 16 Lower doses did not have the same effect on weight5 or muscle protein catabolism.16 In a trial of patients with malignant disease who had experienced > 5% body weight loss, changes in weight were correlated with increases in the plasma EPA level.17 In drug development, it is considered necessary to determine the maximum tolerated dose to be used in a subsequent Phase II study. With omega-3 fatty acids, most studies have used dose levels obtained from trials that were not based on Phase I dose-finding studies. Wigmore et al. carried out a dose-escalation study,18 and we preceded the current study with a traditional Phase I study.8 Trials in which patients receive low doses may mistakenly suggest a lack of activity on the part of omega-3 fatty acids.

We also reasoned that even if a patient's appetite was stimulated immediately, it would require weeks to months for the patient to increase caloric intake and gain weight, which was our major objective indicator of response. We also wanted to provide sufficient time for fatty acid tissue modification of tumor tissues or normal tissues, in the event that such modification is important to the mechanism.19, 20 We found support for these ideas in studies involving megestrol acetate, a progestational steroid that is useful for the improvement of appetite and nutritional status of patients with malignant disease. With the use of that agent, there was little or no change in body weight at 2–3 weeks.21 Similarly, in the megestrol acetate study conducted by Simons et al., body weight did not improve significantly at 6 weeks but did improve by 12 weeks.22 Therefore, even if the appetite improves in 1–2 weeks,23 it is likely that ≥ 1 month will be required for the full effect on weight to be observed. Therefore, we studied weight change versus duration of therapy in patients who took the capsules for at least 30 days, and we found a positive correlation between the number of days taking capsules and weight gain among the 24 patients who took the capsules for at least 1 month.

Patients who gained weight had a better QOL at 30 days posttreatment compared with patients who did not gain weight. However, these QOL results should be interpreted with caution. Only a small number of patients completed the posttreatment assessments. Furthermore, a correlation does not prove a cause-and-effect relation; thus, we cannot be certain that the weight gain was directly responsible for the improvement in QOL. Although future studies will be needed to validate the findings of the QOL data, our results regarding the relation between weight gain and QOL were noteworthy.

In interpreting the doses used in fish oil studies, it is necessary to examine the actual capsule content carefully; in the literature, doses often are reported in grams based on the use of 1 g capsules, which may contain anywhere from 200 to 1000 mg of the active ingredients (EPA and DHA). Interpretation of the actual dose of active material requires calculation of the major omega-3 fatty acids actually delivered. In the current study, we used concentrated capsules, with each 1 g capsule containing 680 mg EPA plus DHA (778 mg total omega-3 fatty acids). Wigmore et al. used a capsule containing 500 mg EPA plus DHA.6 However, many other studies have used capsules with an EPA/DHA content of only ∼300 mg per capsule.7, 9, 18 The use of capsules with high EPA/DHA content may be important for optimizing the efficacy of omega-3 fatty acid treatment of cancer cachexia. In a recent study, patients in a fish oil treatment arm were able to tolerate only 10 of the planned 18 capsules per day.9 In the current study, at least 16 patients withdrew from the study due to the gastrointestinal side effects of the capsules, and similar symptoms also have been limiting in other studies.8, 9, 18 A possible way to minimize this limitation is to put as much omega-3 fatty acid as possible into each capsule so that the number of capsules taken is reduced. There are now commercially available fish oil capsules with high levels of EPA plus DHA or predominantly EPA, and we encourage their use in future trials.

Eight patients had edema recorded at some time during the study. Each of these events was evaluated carefully with regard to the degree of the condition and its relation to weight change. In three patients, edema or its resolution affected the category of response, and these patients were excluded from the analysis for that reason.

Randomized trials have shown weight increases in patients with cancer-associated cachexia who received megestrol acetate.24, 25 However, body composition studies have shown that the weight gain often is related to gains in adipose tissue and water.26 In contrast, the weight gains observed in patients with pancreatic carcinoma who were treated with omega-3 fatty acids were associated with improvements in lean body mass.27 It is believed that the mechanism of omega-3 fatty acids in averting weight loss in tumor-bearing animals and patients is related to the inhibition of cytokines, which cause muscle catabolism. In the current trial, some patients with weight loss and marked muscle loss, as evidenced by low creatinine/height index at baseline, were able to sustain improvements with omega-3 fatty acids. However, if nutritional interventions are initiated as adjuncts to treatment for malignant disease rather than as nutritional palliation in patients with preterminal disease, and if agents with different mechanisms of action are combined, then the observed benefit may be even greater.

Although there was not an increase in median weight for the entire study group, an appreciable fraction of patients with advanced cancer cachexia who could tolerate the supplement experienced a reversal of their weight loss trend or a gain in weight as a result of taking daily concentrated capsules of omega-3 fatty acids for ≥ 1 month. For future Phase III trials, we recommend a lower total daily dose (perhaps 4–6 g EPA plus DHA), with the dose remaining higher than the 1.5–3.0 g used in previous Phase III trials. It also is important to use capsules with ≥ 700 mg active omega-3 fatty acid per 1 g capsule, and the physical size of each capsule should be made as small as possible. The planned trial length should be > 30 days. Although omega-3 fatty acids showed some benefit in patients with preterminal malignant disease in the current study, future trials involving nutritional intervention may have greater success if 1) the intervention is performed earlier in the course of illness to prevent weight loss rather than as a last-minute attempt to reverse severe, established cachexia, as has been discussed by MacDonald28; 2) combinations of effective agents are used concurrently; and 3) the selected patients have a projected survival that allows them to be followed for an adequate amount of time to determine response to treatment. Patients entering future trials should understand the risk of gastrointestinal toxicity associated with the omega-3 fatty acids, as this toxicity may decrease QOL, while treatment provides no assurance of benefit.

Acknowledgements

The authors thank Kristine Johnson, R.Ph.; Janan Geick Miller, R.N.; and Brenda Werner, R.N., for their assistance in coordinating the study. The authors also thank Norman Salem, Jr., Ph.D., of the National Institutes of Health Fish Oil Test Materials Program and Patricia A. Fair, Ph.D., of the Biomedical Test Materials Program for their help in obtaining the capsules tested in the current study.