11 animal evidence 'cannot be interpreted as constituting a carcinogenic hazard to women when these preparations are used as oral contraceptives'. Acknowledgment: I am grateful to Dr Frances de Souza who gave me valuable help in the preparation of this paper. REFERENCES Barnes J M (1972) British Medical Journal iv, 359 Committee on Safety of Medicines (1972) Carcinogenicity Tests of Oral Contraceptives. HMSO, London Davis B R, Whitehead J K, Gill M E, Lee P N, Butterworth A D & Roe F J C (1975) British Journal of Cancer 31, 469 Dunn T B & Green A W (1963) Journal of the National Cancer Institute 31, 425 Gardner W U (1959) Annals of the New York Academy of Sciences 75, 543 Riley V (1975) Science 189, 465

Dr J P O'Sullivan (Department ofPathology, St George's Hospital Medical School, London SW] X 7NA)

Oral Contraceptives and Liver Tumours In October 1973 Baum and her colleagues described 7 cases of hepatic adenoma which had occurred in women who were taking oral contraceptives. Five of their patients were in their 20s and the oldest of the series was 39 years of age. On the basis of this relatively large number of cases of a previously rare lesion, Baum et al. suggested that oral contraceptives might be partly or wholly the cause of the hepatic adenomas in these patients. Over the following six months reports of 4 cases of hepatic adenomas occurring in women on the pill were published (Contostavlos 1973, Knapp & Ruebner 1974, Kelso 1974, Horvath et al. 1974). Unfortunately none of these accounts had an adequate histological description of the lesions and there were no photographs of the microscopical appearances. In the middle of 1974 two papers appeared at about the same time reporting the occurrence of a rather different liver lesion in women on oral contraceptives. Mays' group in the United States reported 4 cases of what they called 'focal nodular hyperplasia' in women who had taken the pill (Mays et al. 1974). The British paper (O'Sullivan & Wilding 1974) described 3 cases of 'liver hamartomas' in similar patients. It is likely that these conditions differ only in name, and it is convenient to consider examples of both under the name 'focal nodular hyperplasia'. Whatever the nature of the lesions, there is no doubt that it is a different entity from hepatic adenoma. So at this point two distinct conditions, both previously uncommon, had been reported in moderate numbers in patients on the pill. During the rest of 1974 a few more cases of hepatic adenomas in patients on oral contra-

Section of Oncology

351

ceptives were described. Berg et al. (1974) searched the Iowa State Cancer Registry and found only 4 cases of hepatic adenoma between 1969 and 1971. These were all in young women, 3 of whom were on the pill. Significantly, the fourth patient died in the ninth month of pregnancy. They searched the files of the University Hospital back to 1938 and could find no other cases of hepatic adenoma in young women. A few more single cases appeared, one each from Greece (Tountas et al. 1974), France (Bisson et al. 1974), England (Model et al. 1974) and the United States (Antoniades & Brooks 1975). Since May 1975, several larger series of cases have been reported. Ameriks et al. (1975) described 8 cases of what they called hepatic cell adenomas. Although it is not stated in their account it is probable that 3 of their cases appeared in Baum's original paper (1973). They first read their paper at a surgical meeting, and 2 more similar cases were recalled by members of the audience in the discussion. Two cases of focal noduiar hyperplasia of liver in women on the pill were reported from Australia by O'Reilly (1975) who pointed out that in 2 of his previous cases (O'Reilly 1974) the patients were also women on oral contraceptives. A further 3 cases were described from Canada (Fisher et al. 1975, Shojania & Hogg 1975) and another example of focal nodular hyperplasia in a woman who had been on the pill appeared from Sweden (Zurbriggen & Tylen 1975). Baum (1975) referred to 6 cases additional to her original 7, but it is not clear that these are different from the patients reported by Ameriks et al. (1975). In August, Christopherson et al. (1975) added a further 7 cases of focal nodular hyperplasia and one hepatic adenoma in women who had taken oral contraceptives. In addition this group has seen a hepatocellular carcinoma in a young woman on the pill. Thus, about 50 cases of benign liver nodules have been reported in women who have taken oral contraceptives. They have, in general, presented in one of four ways: as an incidental finding at laparotomy, often for cholecystectomy; as a lump felt by the patient; as vague gastrointestinal symptoms; and, most seriously, as haemoperitoneum. A number of patients have died because the lesion has bled into the peritoneal cavity. Macroscopically there is a well-circumscribed tumour, often yellow-tan in colour, and usually single, although there may be two or even more in the liver. In focal nodular hyperplasia there is often a central scar. This is not a feature of hepatic adenoma. However, the scar may not be obvious and it is not always possible to make the diagnosis on naked-eye examination alone. Histologically there are two entities (Phillips et al. 1973). The hepatic adenoma is a benign neoplasm which consists of well-differentiated hepato-

Proc. roy. Soc. Med. Volume 69 May 1976

12

cytes and may be a relatively vascular tumour. The other lesion comprises nodules of hepatocytes separated by fibrous trabecule. There are no portal tracts within the nodules, but there are aggregates of proliferated bile ductules within the septa. There is a central scar which may be stellate and there may be quite large veins and arteries at the periphery and in some of the septa. The rest of the liver must be shown to be non-cirrhotic. This lesion has been given many names in the past (including hepatic adenoma). However, two names have come to be used more than others, 'hepatic hamartoma' and 'focal nodular hyperplasia of the liver'. It has been suggested (Christopherson et al. 1975) that the hamartoma lacks the vascular lesions and necrosis of focal nodular hyperplasia, but it is doubtful if a useful purpose is served by distinguishing two such similar lesions. The word 'hamartoma' was suggested because it well describes the condition - a tumourlike malformation composed of normal hepatic cellular elements - and because these lesions occasionally occur in children and young adults. Against this, there are objections to using the term for an entity which seems to develop in many cases in adult life, and the name 'focal noduilar hyperplasia' is probably to be preferred. There is one important difference between hepatic adenoma and focal nodular hyperplasia. Whereas the latter has no propensity to undergo malignant change, there is an area where hepatic adenoma shades into well-differentiated hepatocellular carcinoma and it may be difficult to make the distinction with certainty. However, there is no evidence at present that hepatic adenoma is a precursor of hepatocellular carcinoma (Ishak & Rabin 1975) although a very well differentiated carcinoma may be mistaken for a benign adenoma when first examined. In assessing the evidence of an association between benign liver nodules and oral contraceptive ingestion, hepatic adenorna must be considered separately from focal nodular hyperplasia. There is no doubt that hepatic adenoma was formerly rare. Henson et al. (1956) could find only 4 cases in the records of the Mayo Clinic between 1907 and 1954. Edmondson (1958) found only 2 such tumours in the autopsies done at the Los Angeles County General Hospital from 1918 to 1954. Now a large number of cases are seen over five years, all in women on the pill. Berg's data from the Iowa State Cancer Registry (Berg et al. 1974) support the belief that there is a real increase in the frequency of hepatic adenoma. In the report of the Committee on Safety of Medicines (1972) there was noted an increased incidence of liver tumours in rats given certain combinations of steroid hormones. Long-term administration of oral contraceptives leads to hypertrophy of the smooth endoplasmic reticulum

of hepatocytes (Perez et al. 1969) and it is known that this change is also associated with induction of microsomal enzymes. It is thought that in some cases enzyme induction may potentiate the carcinogenicity of certain compounds by increasing their rate of conversion to toxic metabolites. Progestogens have enzyme-inducing properties (Adlercreutz & Tenhunen 1970) and it may be significant that the synthetic progestogens in oral contraceptives are closely related to anabolic steroids such as methyl testosterone and oxymetholone. Long-term administration of these latter steroids, for example in Fanconi's syndrome, is currently thought to be associated with the development of hepatic adenomas and hepatocellular carcinoma (Farrell et al. 1975). However, some authorities (Anthony 1975) dispute the evidence in this field. With focal nodular hyperplasia also, there does seem to be an increasing number of cases being seen. In a report of 3 cases in 1974, O'Reilly (1974) said that prior to encountering the first case he was unaware of the existence of the entity, and yet he was able to add 2 cases to his original 3 within a year. At St George's Hospital, London (O'Sullivan & Wilding 1974), only 4 cases occurred in the twelve years up to 1974 - all in the last four years of this period. Three of the 4 nodules occurred in women on the pill, and one was in a man who had a large adrenal carcinoma. In further support of an association, May's group (Christopherson et al. 1975) quote the case of a patient who had a left hepatic lobectomy for focal nodular hyperplasia in July 1968. She was not taken off oral contraceptives and in November 1974 she presented with another area of focal nodular hyperplasia in her remaining right lobe. It has been suggested (Edmondson 1958) that the hyperplastic nodule represents a response to local ischmmia due to congenital or acquired vascular abnormalities. So it is possible to postulate that the thrombogenic action of cestrogen might underlie the focus of hyperplasia. However, in all the patients described there has been no note of any clotting abnormality. Alternatively, the lesion may represent a true, formerly quiescent hamartoma which has been stimulated to grow by the oestrogenic component of the pill. It has been reported (Adlercreutz & Tenhunen 1970) that small doses of cestrogens promote liver regeneration in rats. The evidence for an association between oral contraceptive ingestion and either hepatic adenoma or focal nodular hyperplasia is still anecdotal. A recent review (Sorensen & Baden 1975) of reports of both conditions during the period 1923 to 1974 showed that 68 cases were published before 1960 and 117 after that date. In spite of this we lack evidence that there is a true increase in the incidence of either condition. There is a

352

13

Section of Oncology

tendency in present circumstances for cases to be reported where the patient has been on the pill, and not otherwise unless there are clinical or pathological features of exceptional interest. Furthermore, some cases have almost certainly been reported more than once (Baum et al. 1973, Ameriks et al. 1975) and this distorts the figures. Both hepatic adenoma and focal nodular hyperplasia are well known to be commoner in women than in men, and to be commonest in women during reproductive life. Since it is estimated that, in the United Kingdom, about 30 % of women of child-bearing age have taken oral contraceptives, there is quite a high chance that a woman with any condition whatsoever is on, or has been on, the pill. In addition, many of the patients in whom the association has been suggested have been taking oral contraceptives for only a short time. For instance, one of Baum's original patients had been on the pill for only six months, and another of her patients had only a two-year history of oral contraceptive ingestion. Most recognized human carcinogens have latent periods of 10-15 years and in some the time between exposure and tumour development is considerably longer. There are two published reports of women on the pill developing hepatocellular carcinoma (Christopherson et al. 1975, Hermann & David 1973). One woman (Thalassinos et al. 1974) developed hepatocellular carcinoma after being on an 'cestrogen-like' drug for infertility, and there has been a case of hepatoblastoma (Meyer et al. 1974) in a patient on oral contraceptives. Until many more cases are reported, it would be wrong to posit any association between oral contraceptive ingestion and malignant hepatic tumours. In summary, the case for an association between the pill and either hepatic adenomas or focal nodular hyperplasia of the liver must be regarded as 'not proven'. The question is clearly important, and there is a strong case for the setting up of a panel to which all British cases of these lesions, without selection of patients either by sex or by age, be referred for pathological assessment. Only in this way will it be seen whether there is, a true predilection of either lesion for women taking oral contraceptives. Acknowledgment: I am grateful to Professor Sir Theo Crawford for his encouragement and advice.

Bisson A, Duron J-J, Fagniez P-L, Pinaudeau Y & Germain A (1974) Nouvelle Presse Mddicale iii, 2079 Christopherson W M, Mays E T & Barrows G H (1975) Obstetrics and Gynecology 46, 221 Committee on Safety of Medicines (1972) Carcinogenicity Tests of Oral Contraceptives. HMSO, London Contostavlos D L (1973) Lancet ii, 1200 Edmondson H A (1958) Atlas of Tumor Pathology. Armed Forces Institute of Pathology, Washington, DC; Section 7, fascicle 25, pp 12, 193 Farrell G C, Joshua D E, Uren R F, Baird P J, Perkins K W & Kronenberg H (1975) Lancet i, 430 Fisher A W F, Curry B & Jacques J (1975) Canadian Medical Association Journal 112, 1196 Henson S W jr, Gray H K & Dockerty M B (1956) Surgery, Gynecology and Obstetrics 103, 23 Hermann R E & David T E (1973) Surgery 74, 715 Horvath E, Kovacs K & Ross R C (1974) Lancet i, 357 Ishak K G & Rabin L (1975) Medical Clinics of North America 59, 995 Kelso D R (1974) Lancet i, 315 Knapp W A & Ruebner B H (1974) Lancet i, 270 Mays E T, Christopherson W M & Barrows G H (1974) American Journal of Clinical Pathology 61, 735 Meyer P, Li Volsi V A & Cornog J L (1974) Lancet ii, 1387 Model D G, Fox J A & Jones R W (1974) Lancet i, 865 O'Reilly K (1974) Australian and New ZealandJournal ofSurgery 44, 142 (1975) A ustralian and New Zealand Journal of Surgery 45, 76 O'Sullivan J P & Wilding R P (1974) British MedicalJournal iii, 7 Perez V, Gorodisch S, De Martire J, Nicholson R & Di Paola G (1969) Science 165, 805 Phillips M J, Langer B, Stone R, Fisher M M & Ritchie S (1973) Cancer 32, 463 Shojania N G & Hogg G R (1975) Gastroenterology 69, 28 Sorensen T I A & Baden H (1975) Scandinavian Journal of Gastroenterology 10, 1 13 Thalassinos N C, Lymberatos C, Hadjionnou J & Gardikas C (1974) Lancet i, 270 Tountas C, Paraskevas G & Deligeorgi H (1974) Lancet i, 1351 Zurbriggen S & Tylen U (1975) Fortschritte aufdem Gebiete der Rontgenstrahlen und der Nuklearmedizin 122, 404

REFERENCES Adlercreutz H & Tenhunen R (1970) American Journal of Medicine 49. 630 Ameriks J A, Thompson N W, Frey C F, Appelman H D & Walter J F (1975) Archives of Surgery 110, 548 Anthony P P (I1975) Lancet i, 685 Antoniades K & Brooks C E jr (1975) Surgery 77, 137 Baum J K (1975) Journal of the American Medlical Association 232, 1329 Baum J K, Holtz F, Bookstein J J & Klein E W (1973) Lancet ii, 926 Berg J W, Ketelaar R J, Rose E F & Vernon R G (1974) Lancet ii, 349

353

Dr N P Bishun (Research Department, Tissue Culture and Cytogenetics Unit, Marie Curie Memorial Foundation, The Chart, Oxted, Surrey)

Chromosomes and Oral Contraceptives The difficulty of anticipating the biological effects of chemicals in the environment applies to longterm effects like hereditary changes, cancer induction and degenerative cellular effects leading to poor health as well as to teratogenic effects. Also, mutagenic effects, in the broad sense, are of central importance. The ability of certain chemicals to alter hereditary material has been known for three decades; in fact, this knowledge has constituted a corner-stone in the development of modern experimental genetics. In spite of this, many geneticists were for a long period slow to appreciate the practical consequences of chemical mutagenesis and possible risks to human populations. There is now an increasing demand for chromosomal evaluation of chemicals released in the environment and for the elimination of those products which may cause mutations in human beings.

Oral contraceptives and liver tumours.

11 animal evidence 'cannot be interpreted as constituting a carcinogenic hazard to women when these preparations are used as oral contraceptives'. Ack...
573KB Sizes 0 Downloads 0 Views