Japanese Journal of Clinical Oncology

INTRODUCTION

Environmental, particularly dietary factors, are regarded as having strong influences on the etiology of colon cancer (1). A positive correlation between the intake of fat and the incidence of colorectal cancer has been demonstrated not only from the results of epidemiological studies (2-4), but also from animal studies (5-7). Recently, in Japan, with the spreading of a westernized lifestyle, fat intake has been increasing. Comparing Japanese in 1960 to 1985, dietary patterns have changed and the death rates due to specific diseases have become similar to the pattern of the United States in 1985; for example, death rates due to cerebral bleeding and cancer of the stomach have decreased, while deaths from cancer of the lung and colorectum have increased (8). The incidence of colon cancer has been increasing markedly and this tendency is likely to continue in Japan (9).

Although the Inuit in Greenland consume as much fat as Danes, whose incidence of colon cancer is high, the incidence rate of breast or colon cancer among Inuits was relatively low (10). This study draws attentions to the importance of n-3 polyunsaturated fatty acids (PUFA), such as docosahexanoic acid (DHA) or eicosapentanoic acid (EPA), in the development of colon cancer. Marine animals formerly eaten by Inuits were rich in n-3 PUFA. In an animal study, a high-fat diet promoted 1,2-dimethylhydrazine, methlazoxymethanol acetate or methyl nitrosourea induced colon cancer in rats (5), but this effect differed according to the type of fat used in the experiment (11). Especially the composition of fatty acids, such as corn oil containing a high amount of n-6 PUFA, promoted colon cancer development, but fish oil rich in n-3 PUFA inhibited colon carcinogenesis (12). It is not clear whether DHA or EPA has the major inhibitory effect on colon cancer. There are reports claiming that DHA is responsible for inhibition of colon carcinogenesis (13-15). In a human study, rectal cell proliferation of patients with sporadic colon adenoma was reduced by a supplement of fish oil capsules (16).

Table 1. The patients participating in the study

Case age, sex	Procedure for colectomy (age)	Compliance (follow-up period)	Number of polyps					Symptom and complication associated with capsules
			At entry	6 months	12 months	18 months	24 months
1. FAP     31, F	Subtotal colectomy with restolative ileocecal region (30)	90.2% (2 years)	Many	[rArr]	[uArr]		[uArr]	Hypermenorrhea Early colon cancer (at 24 months)
2. FAPa     58, F	Extended left hemicolectomy	98.8% (2 years)	Few	[rArr]	[rArr]	[dArr]	[uArr]	Diarrhea Early endometrial cancer (at 12 months)
3. MA     46, M		91.0% (2 years)	0 (41)b	9	12	7	1	Itching
4. MA     35, M		61.3% (1 year)	0 (31)b	0	1
5. FAP     59, M	Total colectomy with ileorectal anastomosis (53)	99.5% (1 year)	Few	[rArr]	[uArr]			Lung cancer (at 12 months)

[rArr], no change; [dArr], decrease; [uArr] increase.
FAP, familial adenomatous polyposis; MA, multiple adenoma ([ge] 30) in colorectum. ^aBecause of the operation for endometrial cancer, the patient stopped taking the capsules for 2 months. ^bThe number in parentheses represents the number of polyps at endoscopic resection before entry

Table 2. Composition of fatty acids in DHA-concentrated fish oil capsule

Fatty acids			%
14:0		(myristic acid)	2.7
15:0			0.6
16:0		(palmitic acid)	11.0
16:1		(palmitoleic acid)	6.4
17:0		(margaric acid)	1.2
18:0		(stearic acid)	1.9
18:1		(oleic acid)	17.7
18:2	n6	(linoleic acid)	1.3
18:3	n3	([alpha]-linolenic acid)	0.5
18:4	n3		1.1
20:1		(gadoleic acid)	1.9
20:4	n3		0.7
20:4	n6	(arachidonic acid)	2.4
20:5	n6	(eicosapentaenoic acid)	9.8
22:5	n3	(docosapentaenoic acid)	1.8
22:5	n6		1.5
22:6	n3	(docosahexaenoic acid)	31.5
Others			6.0

We conducted a long-term trial of DHA-concentrated fish oil capsules for patients with familial adenomatous polyposis (FAP) and with multiple colorectal polyps. We experienced three patients with FAP diagnosed as having malignant lesions during the trial.

CASE REPORTS

DISCUSSION

We previously reported the adverse effects of sulindac for patients with a high risk of colorectal cancer. Sulindac is one of the non-steroidal anti-inflammatory drugs and was expected to be a chemopreventive agent for colon cancer. We observed perforation of the stomach, multiple ulcers of the small intestine, and so on. We concluded that sulindac had untoward effects for it to be used as a chemopreventive agent, especially for Japanese (17). We sought another chemopreventive agent for patients with FAP, and expected that DHA-concentrated fish oil capsules would have a similar effect and be less harmful, because in animal studies inhibitory effects of DHA against colon neoplasm were observed (14,15). So, we chose DHA for this study. From the present study, we could obtain sufficient compliance with DHA-concentrated fish oil capsules, but found that early colon, lung and endometrial cancer developed in the patients with FAP.

The patients with FAP were originally in the high-risk population for the occurrence of malignancies, so we investigated the frequency of extracolorectal malignancies developing in the patients with FAP. In the data from the Polyposis Center, Tokyo Medical and Dental University (18), 14% of the patients with FAP also had some extracolorectal malignancies, such as gastric cancers, duodenal or small intestinal cancers and urological cancers among male patients, and also thyroid cancers and utero-ovarian cancers were observed among female patients. The rate of uterine cancer was 4.7% of the extracolorectal cancer among female patients. No lung cancer was observed among either sex.

From the data of an interventional study using wheat fiber, and vitamins C and E for 58 patients with FAP, one rectal cancer, one duodenal cancer and two thyroid cancers occurred over a two year interventional period (19). Another interventional study using ascorbic acids reported a case of cancer of the duodenal papilla and a case of breast cancer in 49 patients with FAP during a four-year trial (20). The cancer incidence of our study was higher than these reports.

DHA-concentrated fish oil capsules used in our study were made of tuna fish oil, and the composition of PUFA is richer in DHA than EPA. The form of PUFA in fish oil was given not as ethyl esters, but as triglycerides. Both forms were shown to be equally well absorbed in a human volunteer study (21). Many articles concerning the administration of fish oil to humans have been published, such as to healthy volunteers (22,23), patients with hypertension (24), psoriasis (25), IgA nephropathy (26), inflammatory bowel disease (27,28), and to patients with sporadic colon polyps (16). The trial periods in most of them were short, less than six weeks. The fish oil used in most of those studies was richer in EPA than DHA, and the forms of fish oil varied, such as triglyceride, ethyl ester or free fatty acid. From those studies, only a few documents reported on the development to cancer or adverse effect. In the trial of patients with IgA nephropathy, fish oil capsules given as the triglyceride form were administered for two years and one patient developed metastatic lung cancer and died, while in the placebo group one patient also died of metastatic lung cancer (26).

The reason why the cancer incidence in our study was so high and the reason why the inhibitory effects against colorectal polyps were not observed, in contrast to the animal studies (13-15), are unknown. The form of fish oil, the purity of DHA, other contents in the fish oil capsule used in our study, the duration of the trial or something else might be the reason. DHA was given as the ethyl ester in the animal studies and the purity of DHA was 74% in one study (13) and 97% in another two studies (14,15). The form of DHA-concentrated fish oil capsules used in our study was triglycerides and the purity of DHA was about 30%. The form of DHA or the purity of DHA might be the cause of our result, but another study with Menhaden oil (DHA composition was 11%) in rats stated that the high-fat diets with increasing level of Menhaden oil (from 30% to 60% Menhaden oil) had a lower incidence of adenocarcinoma compared to those fed the high-fat diet with corn oil (12). The inhibitory effect did not necessarily depend on the purity of DHA. So our results might have occurred by chance, but there might be some specific causes. Although we do not know whether or not DHA is the substance responsible for our result, because many DHA-adding food products were widely supplied, we believed that we should report the facts and investigate the possible mechanism. To help resolve this, we have started the basic studies, such as in multiorgan carcinogenesis models and in Apc knockout mouse with our DHA-concentrated fish oil capsules in a long-term trial.

Acknowledgments

We gratefully acknowledge Dr Yoshio Iwama of the Polyposis Center, Tokyo Medical and Dental University, for providing the data on extracolorectal malignancies in the patients with FAP. This study was supported by a Grant-in-Aid for the Second Term Comprehensive 10-Year-Strategy for Cancer Control from the Ministry of Health and Welfare, Japan.