Steroid related neoplasia in human liver.

Prior to the early 1970's, benign liver neoplasms were among the rarest of tumors. The seemingly rapid increase, especially in young females ingesting oral contraceptives, as well as the catastrophic presentation of many of the tumors resulting from liver rupture and hemoperitoneum, stimulated studies by several investigators. In the Liver Tumor Registry at the University of Louisville, we have examined the histologic material, and finalized the data on 227 tumors, the majority in young women. With few exceptions, they had used oral contraceptives or were either pregnant or immediately post-partum and presumably in a hyperestrogenic state. There have been 82 hepatocellular adenomas (HCA), 105 cases of focal nodular hyperplasia (FNH), and 31 hepatocellular carcinomas. The hepatocellular carcinomas occurred in non-cirrhotic livers, and 14 of the 31 cases were of a distinct, but rare type, polygonal cell carcinoma with lamellar fibrosis. While it seems reasonable to believe steroids play a role in adenomas and in FNH it is less certain that they produce hepatocellular carcinomas since malignant liver tumors are not uncommon in this age group without the use of oral contraceptives. With an estimated 50 million women either currently using or who have used oral contraceptives the risk must be very slight.

Determination of protein loss during aqueous and phase partition fixation using formalin and glutaraldehyde.

In phase partition fixation tissue is immersed in an organic solvent at equilibrium with an aqueous phase containing a fixing agent. By using radioisotope labeling techniques the effects of phase partition fixation on protein retention during fixation of tissue with formalin and glutaraldehyde have been determined and compared with those of standard aqueous fixation using these fixatives. It has been shown that retention of protein in tissue during phase partition fixation was as good or better than during aqueous fixation. Improved retention provides further evidence that phase partition fixation may be a useful alternative to aqueous fixation.

Hepatic tumors induced by sex steroids.

Our study of more than 250 women with hepatic tumors, accessioned in our tumor registry at the University of Louisville, disclosed three types of tumor: FNH , HCA, and HCC. The ingestion of sundry kinds of sex steroids by the majority of these women, chiefly for purposes of preventing conception, warrants the suspicion that such hormones induced these different types of hepatic tumors. Publications by others reporting similar hepatic tumors in men using male sex steroids lends support to this hypothesis. Rupture of the hepatic tumor and consequent hemorrhage, producing hemoperitoneum, is a major risk factor. Other presenting symptoms are pain and palpable mass. Symptomatic women using OCs should be subjected to a CT or technetium hepatic scan as an initial screening assessment. Because of the imminent possibility of rupture, large turgid vascular tumors should be resected without biopsy. Biopsy-proved HCC should also be removed surgically. All other tumors, including small multiple tumors, will usually regress when exogenous sex steroids are withdrawn and pregnancy avoided. Other significant hepatic changes observed in this study are peliosis hepatis, periportal sinusoidal dilation, and vascular lesions. The branches of the hepatic artery and the tributaries of the portal vein show combinations of intimal and smooth muscle proliferation, vascular thickening, occlusive intimal thickening, and, at times, obstructing thrombosis. Similar smooth muscle proliferation in the afferent vessels of the livers of animals treated with sex steroids suggests that there is a cause and effect relationship in women using OCs.

PIP: Previous reports describing the oncogenic capacity of estrogens and endrogens and reports suggesting a relationship between primary liver tumors and oral contraceptives (OCs) prompted a meticulous evaluation of the lifestyle and medication usage of each patient accessioned by the registry at the University of Louisville School of Medicine. More than 250 cases of hepatocyte tumors in young women, not all steroid related, have been collected since 1973. The data are sufficient for analysis in 201 patients. The intention was to assess etiological factors and critically classify the histologic features. 3 separate tumors were distinguishable: hepatocelluar adenoma (HCA), focal nodular hyperplasia (FNH), and hepatocellular carcinoma (HCC). In 9 tumors a histologic classification could not be made, generally because of massive hemorrhage and infarction of the tumor. 1 tumor was removed several weeks after hepatic artery ligation was done to stop hemorrhage from the ruptured tumor. Only dense scar tissue remained. Currently, none of the patients in the unclassified group has developed a recurrence nor have any developed metastases. None have died. HCA is relatively soft and usually solitary. The usual description of HCA stresses encapsulation. This is true of 5 tumors in the registry; in these cases the HCA occurred in women not exposed to exogenous reproductive steroids. In contrast, in women ingesting sex steroids the HCA were frequently not encapsulated. Nonencapsulation was characteristic of multiple small tumors in the group taking OCs. The typical FNH has a central scar with radiating septa and a coarsely nodular appearance. In contrast to HCA, necrosis and hemorrhage are less frequently encountered. The presence of bile duct epithelium is the single most distinguishing characteristic differentiating FNH from CHA. In women taking exogenous sex steroids, HCA and FNH may be related lesions or may be different histologic manifestations of a stimulus affecting mesenchymal and entodermal elements of the liver in variable degrees. HCC is grossly indistinguishable from FNH, having a well formed central scar, fibrous septa, and coarse nodulation. Despite the fact that all HCCs were fairly well differentiated tumors, there were no longterm survivors. There is no conclusive evidence that a cause and effect relationship exists between primary tumors of the liver and OCs. A history of prolonged sex steroid administration, usually OCs, was present in the majority of the cases studied. Similar smooth muscle proliferation in the afferent vessels of the livers of animals, treated with sex steroids suggests that there is a cause and effect relationship in women using OCs.

Are hepatic arteries end-arteries?

Studies of intrahepatic arteries in vivo and after death give dramatically different results concerning the human hepatic circulation. Corrosion casts of human liver vasculature support the existing belief that hepatic arteries are end-arteries. However, in vivo studies using angiographic techniques after ligation of either the left or the right ramus of the hepatic artery demonstrate intrahepatic translobar collateral arteries capable of perfusing the entire occluded arterial system. There is no intermediate filling of sinusoids nor of portal venules. Traditional statements that hepatic arteries are end-arteries cannot be supported by the present findings.

Options in treating trauma to the liver.

Many wounds of the liver are minor and require no corrective operations. If these patients are not killed by associated injuries, they usually survive. Hemorrhage from hepatic wounds is the most common cause of death. Selective ligation of appropriate blood vessels swiftly controls bleeding and does not cause septic hepatic necrosis because the liver in humans is sterile and collateral blood flow prevents complete infarction of hepatic tissue. Hepatic resection is dangerous in acutely hypovolemic patients, but hepatic resection is beneficial in treating the late complications of hepatic trauma. Liver-related sepsis after trauma to the liver can be obviated by elimination of gauze packs, absorbable hemostats, peritoneal drains, sutures and choledochal tubes.

Hepatic artery ligation.

The use of hepatic artery ligation (HAL) in various clinical situations is illustrated by presenting a series of eight patients. The indications for HAL included ruptured hepatic tumors, spontaneous liver rupture, delayed hemorrhage after liver trauma, hematobilia, hepatic artery aneurysm, and hemorrhage after liver biopsy. Conventional methods of hemostasis had been used in some of these patients, but failed to control hemorrhage. The reasons for the relatively late adoption of hepatic artery ligation in treating liver hemorrhage are discussed and placed in proper prospective. Hepatic artery ligation has been shown to be such an effective method of controlling hemorrhage from the liver that other methods, such as packing and mass suture, which are unsafe and ineffective, should be abandoned. Major resection should be done only when an entire lobe of the liver is reduced to pulp or when exposure and repair of the retrohepatic vena cava are necessary.

Relation of steroids to liver oncogenesis.

Experience with pathological material from 150 women with liver tumors is reviewed. The features of liver cell adenoma and focal nodular hyperplasia are sufficiently different that the vast majority of the benign tumors can be easily subclassified. Although most occurred in women ingesting steroids, the wide usage of oral contraceptives makes it difficult to prove a causative role. Nineteen of the tumors were malignant and, to date, 12 of those patients have died of their disease. Since hepatomas are much more common than benign liver tumors, one must be even more circumspect in indicting steroids in their causation. In this group of women none had cirrhosis, whereas in the general population cirrhosis is a very common precedent lesion. Further investigation of estrogens and primary liver carcinoma would be timely.

PIP: Histological specimens from 150 women with liver tumors are discussed. Of the 150 patients under consideration, 85% had ingested contraceptive steroids, most for more than 3 years. Of these 64% had taken pills containing mestranol, and 18% had used ethinyl estradiol; 18% had taken both. Average age was about 30 years, and pain was the most common presenting symptom. 19 tumors were malignant (hepatoma), 57 were adenoma, 68 were focal nodular hyperplasia, and 6 were unclassified. To date, 12 of the 19 hepatoma patients have died. In addition to presenting numerous figures depicting the pathology material and a discussion of tumor differentiation difficulties, speculation between steroid ingestation and tumor appearance is considered. Since hepatomas are much more common than benign liver tumors, circumspection is in order before indicting steroids as causative. In this group of women studied, none had cirrhosis, for example, whereas cirrhosis is very common in the general population. The authors call for further investigation of estrogens and primary liver tumores.

Selective management of post-traumatic obstructing intramural hematoma of the duodenum.

Fourteen patients with post-traumatic obstructing intramural duodenal hematoma were reviewed. Seven patients underwent exploration for associated injuries. Evacuation of the hematoma was performed in two patients. Of 12 patients treated by fasting, nasogastric decompression and parenteral fluid therapy, late obstruction of the duodenum developed in one patient, and operation was required. Criteria are discussed regarding evacuation of intramural hematoma of the duodenum found at celiotomy for associated injuries. In the typical patient with an isolated intramural hematoma in whom no other indications for operation are identified, aggressive nonoperative therapy is the preferred treatment.

Liver oncogenesis and steroids.

PIP: Data on 148 cases of liver tumor in women have been registered. Analysis of these data shows that 1) the average age is 30.3 years; 2) 85% of the patients had a history of oral contraceptive use; 3) pain was the most usual symptom followed by incidental discovery during an operation; 4) 19 were hepatomas, 56 were adenomas, 67 were focal nodular hyperplasia, and 6 were unclassified; and 5) 67% of the benign tumors were in the right lobe, there were 15 cases of multiple focal nodulat hyperplasia and 11 cases of multiple adenomas, and several of the adenomas were only partially encapsulated. The histopathologic differentiation of focal nodular hyperplasia from adenomas can be obtained by detection of the presence of bile duct epithelium in focal nodular hyperplasia; this is always absent in adenoma. Of the 19 patients with hepatomas, 12 have died (7 had metastasis, 3 deaths were related to the operative status), 2 are near death, and 5 are alive following resection. Treatment in most cases was resection or lobectomy, but biopsy only was performed in 22 cases of benign tumor. Follow-up of these cases should add to the knowledge about the necessity extent of surgery. The possible relationship of oral contraceptive use to liver oncogenesis is as yet undefined, but the incidence of tumors is very low considering the numbers of women who are current users of steroid contraceptives. Benign tumors have been reported to involute after discontinuation of steroidal medication. This therapeutic dilemma may be resolved when the patients in this series who underwent biopsy only have been followed for a longer interval.^ieng

Selectivity in the management of hepatic trauma.

Retrospective review of 178 consecutive patients who sustained hepatic injuries confirms that selective application of hepatic artery ligation is an efficient means of definitive hemostasis. Only two of 20 immediate deaths resulted from hepatic parenchymal hemorrhage. Over-all mortality (20%) was not significantly increased where hepatic artery ligation was used. Sepsis was the most frequent complication after severe hepatic trauma.

A clinicopathologic study of steroid-related liver tumors.

A registry of liver tumors was started in late 1973 in an attempt to assess the relationship of these tumors to oral contraceptives or to other environmental factors. This report is concerned with the pathological aspects and the possible pathogenesis of the first 101 tumors accessioned. There were 44 instances of focal nodular hyperplasia, 40 adenomas, four unclassified but probably benign tumors, and 13 hepatocellular carcinomas. Eighty-one patients took oral contraceptives; six were associated with pregnancy; three had taken estrogens for long periods of time; one had a thecoma; four never took sex steroids; and in five the history was unknown. Tumor rupture and intrahepatic hemorrhage were frequent complications. It is possible that the vascular lesions associated with focal nodular hyperplasia could play a part in their pathogenesis as well as with rupture. Foci of adenomatous hyperplasia may be related to the development of adenomas. The association of these tumors with sex steroids could be coincidental. The fact that none of the patients had cirrhosis of liver fibrosis, and that androgenic anabolic steroid therapy has been associated with hepatocellular carcinomas in males, suggests that further study of the problem is necessary.