Hepatocellular Carcinoma and Sex Hormones

The liver is morphologically and functionally modulated by sex hormones. Long-term use of oral contraceptives and androgenic steroids can induce benign and malignant hepatocellular tumors. Hepatocellular carcinoma (HCC) is more prevalent in men than in women. The role of sex hormones and their receptors in the development of HCC was reviewed. Some HCCs may be androgen dependent but others may be estrogen or even both dependent. Further studies are mandatory in order to utilize such characteristics of HCC for an effective prophylaxis and therapy of this tumor.

It is well known that cirrhotic male patients, particularly th6se with alcoholic cirrhosis, have impaired function of the hypothalamic-pituitary-gonadal axis1. All over the world, HCC is more prevalent in men than in women. This is especially true in patients with cirrhosis and in the areas where HCC is common. Such male predominance of HCC may be related to sex hormone imbalance2. On the other hand, it is undeniable that long-term use of oral contraceptives induces not only benign but malignant tumors of the liver. Also, androgenic anabolic steroids, especially 17-alkyl-substituted steroids, seem to be related to the development of liver neoplasms including HCC. These facts raise the question; what is the genuine role of sex hormones in hepatocarcinogenesis in experimental animals and particularly in humans?

Liver Tumors and Sex Steroid Administration
The effect of sex hormones on liver function and morphologic features has been pointed out over the past two decades. It is generally accepted that the liver is one of the target organs for estrogens. Since the first report by Baum et al. 3, a possible correlation of estrogens with the development of hepatic neoplasms has been suggested by numerous reports of oral contraceptive-associated focal nodular hyperplasia (FNH) or liver cell adenoma. Malignant transformation has also been reported in such patients. In addition, withdrawal of oral contraceptives occasionally induces regression of the drug induced hepatic tumors4-6. According to the recent articles by Neuberger et al. and Forman et al., the use of contraceptive pills for longer than 8 years is associated with a 20-fold increased risk of HCC.
On the other hand, the administration of exogenous androgens, particularly 17a-alkylated androgens, causes various hepatic lesions such as dilation of biliary canaliculi, loss of canalicular villi, peliosis hepatis, adenoma, and HCC9. Such changes occur less frequently with non-17-a-alkylated androgens. It is noteworthy that synthetic estrogen and progestogen used for pills have a similar chemical structure to androgenic steroids.
Although most carcinogens can be shown to be mutagenic in vitro, this has not been demonstrated for estrogens1. The Committee on Safety of Medicine (London) reported a tumorigenic action of synthetic estrogens in male rats but not in female rats1. Reznik-Schuller 2 found the development of liver tumors (hepatocellular adenomas, carcinomas, and cholangiocarcinoma) in 29% of male European hamsters treated with diethylstilbestrol (DES). This seems to be the first report showing that estrogens alone can be carcinogenic. Other authors have suggested that estrogens promote the effect of carcinogens3-5. There are several groups in Japan who have extensively studied this problem. Sumi et al. 6 initially found a synergistic effect of DES and carcinogens on the development of liver tumors in male WF rats. Later, they have found that DES itself has a direct carcinogenic effect on the liver and that this effect is not mediated by prolactin7. Higashi et al. 8 in 1980 reported that synthetic estrogen and progestogen could induce HCC in female Wistar rats.
Despite a considerable number of reports on androgenic steroid-induced hepatic tumors, there is limited evidence-for a direct etiologic relationship between administration of the drugs and HCC in humans19. Many years ago, Morris et al. 2 found that in a hepatic carcinogenesis model in rats with 2-diacetylaminofluorene HCC occurs more frequently in intact males and castrated females treated with testosterone than in intact females, castrated males, castrated females, and castrated males treated with DES. Whether or not androgenic steroids are carcinogenic remains to be elucidated.

Sex Hormone Receptors in HCC
It is theoretically assumed that the tumorigenic effect of estrogens may be through an estrogen receptor (ER It has long been believed that human liver is not a target organ for androgen. In 1985, we first observed that normal human liver possesses androgen receptors (AR) 27. Immediately after our report, Bannister et al. 28 presented a similar observation. Iqbal et al. 29 were the first to find ARs in human HCC. We have also shown that ARs are present in 74% of male and 37% of female HCCs24'27. In addition, the titers were significantly higher in HCC than in the surrounding liver parenchyma. A very similar result was presented by Ohnishi et al. 25. We thereafter investigated, with positive results, if extrincically given testosterones were taken up by AR in HCC using autoradiographic techniques. All over the world, HCC is more prevalent in males than it is in females31. This is especially true in patients with liver cirrhosis32. Male predominance is usually explained by the fact that alcoholism, chronic liver disease, and in particular chronic hepatitis B infection are more prevalent among males. In the light of our serial receptor studies, however, we have suggested that some, if not all, human HCC may be androgen-dependent. Such a speculation is supported by our clinical results which will be described in the next section.
Several investigative groups have shown that high titers of progesterone receptor (PgR) are present in FNH and liver cell adenoma whether or not they are induced by oral contraceptives33-35. Demenes et al. 36  Operatioe Results of HCC In our series of 107 radical hepatic resections, the 1-, 3-, and 5-year survival rates in male and female patients were 78% and 70%, 45% and 52%, and 19% and 52%, respectively38. The difference was significant 47 months after operation. The clinicopathologic background was not different between men and women. Thereafter, we have analyzed the recurrence and survival rates between the patients with AR-negative and AR-positive HCCs39. The recurrence rates were 33.8% in the former and 67.9% in the latter group. Also, the 5-year survival rate was significantly better in the patients with AR-negative HCCs (62.2%) than in the patients with the positive tumors (17.3%). On the other hand, the presence or absence of ER in HCC did not influence the long term results4.
Hormone Therapy of HCC Friedman et al. 22 first tried hormone therapy for HCC. Since in hormoneresponsive tissue such as breast or endometrial carcinoma that contains ERs, administration of progestins results in decreased ER titers and often causes tumor regression, they performed progestin therapy in 5 patients with HCC. Tumor regression was observed in two of them. We have treated 9 patients with nonresectable HCC with tamoxifen, but no apparent tumor response was found (unpublished data). As tamoxifen inhibits the development of hyperplastic nodules in the liver of experimental animals4, anti-estrogen therapy may be effective in selected patients with HCC. However, we probably cannot expect so much solely by antiestrogen therapy in the treatment of far advanced HCCs. Based 5 have shown that the growth of AR positive cells was enhanced by dihydrotestosterone, and was inhibited by adding cyproterone acetate.

Future Perspective
Except for oral contraceptive-associated HCCs which may be highly estrogendependent, human HCCs seem to be more androgen-dependent. Male predominance, the receptor profiles, and the clinical evidence may support this speculation. Routine receptor assays of surgically removed specimens will be useful for further elucidation of this unsolved problem. Based upon the receptor result in each individual, appropriate hormone therapy which may be anti-androgenic, antiestrogenic, or both will be possible. The poor survival in men after liver resection could theoretically be improved by adjuvant anti-androgen therapy. It may be possible to improve the survival rate of those with nonresectable HCCs by combining hormone therapy with conventional chemoor immuno-chemotherapy. If hormonal manipulation can prevent the development of HCC in high-risk patients will be a matter for future work.