U.S. flag

An official website of the United States government

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Kufe DW, Pollock RE, Weichselbaum RR, et al., editors. Holland-Frei Cancer Medicine. 6th edition. Hamilton (ON): BC Decker; 2003.

  • By agreement with the publisher, this book is accessible by the search feature, but cannot be browsed.
Cover of Holland-Frei Cancer Medicine

Holland-Frei Cancer Medicine. 6th edition.

Show details

Hepatotoxicity by Anticancer Therapy

, MD, PhD.

Diagnosis of Hepatotoxicity

Diagnosis of drug-induced liver toxicity is often difficult for drug toxicity may mimic any other form of liver disease. There are no histologic or biochemical features that are specific to drug-induced hepatotoxicity, and there are no simple, safe methods for diagnosing the cause in most cases. In cancer patients, this is compounded by the frequency of other causes of liver disease that may complicate the cancer itself or cancer therapy (Table 154-3). Mild or moderate transient elevations of liver tests without clinical toxicity may often be ignored. However, when clinical toxicity occurs, early diagnosis and withdrawal of the offending drug is imperative. The most important factors in early diagnosis are a high index of suspicion for drugs as a cause of liver injury and knowledge of the type of injury a particular drug can cause. Other causes of liver disease must be ruled out with viral serology, cardiovascular evaluation, and imaging of the liver. Factors that can assist in the diagnosis are the time-course of the disease, the response to rechallenge, and histologic appearance.

Table 154-3. Causes of Liver Test Abnormalities in Cancer Patients.

Table 154-3

Causes of Liver Test Abnormalities in Cancer Patients.

The chronologic criteria that have been developed to diagnose drug-induced liver disease are for drugs given continuously and do not apply to intermittent use (see below). These criteria may not apply to anticancer drugs given for brief periods of time at periodic intervals. The temporal criteria are latency, response to dechallenge and rechallenge, and the response to continuing exposure to the drug.101 Latency is the duration of time between starting the drug and onset of toxicity. Latency between 5 and 90 days is most suggestive of a drug as the etiology, but during rechallenge onset is usually accelerated to 1 to 15 days. A 50% decrease in serum aminotransferases within 8 days of stopping the drug (dechallenge) or normalization within 30 days is very suggestive, but a 50% decrease within 30 days is still suggestive of drug-related hepatitis. If a drug is continued without dose reduction, stable or increased aminotransferases are suggestive of a drug etiology, whereas a partial decrease or normalization are inconclusive.

A recurrence of liver injury after restarting a medication is the best evidence of drug-induced toxicity and a history of an unplanned reexposure can be very helpful. If a dose-dependent toxin causes liver injury typical for that drug, restarting the drug at reduced dosage may be acceptable. However for dose-independent drugs that cause severe liver toxicity during the initial exposure, deliberate rechallenge for diagnostic purposes can be extremely hazardous and should not be undertaken lightly. Clearly the need for that particular drug will need to outweigh the risk of reexposure. Although recurrence of toxicity is the strongest evidence for the diagnosis of a drug etiology, the lack of a recurrence after a short rechallenge may be misleading. For dose-dependent drugs restarted at the same dose, a rechallenge may not lead to the same blood levels that induced the original injury. For dose-independent toxins, environmental variables (concomitant drugs, dietary factors) that contributed to the formation of a minor metabolite may be absent during rechallenge or the duration of rechallenge may not be sufficient to reaccumulate the offending metabolite.

The steps outlined above usually make the diagnosis of drug-induced hepatotoxicity and there are limited indications for the use of liver biopsy to establish the diagnosis. As described earlier in “Sinusoidal Obstruction Syndrome (Hepatic Venoocclusive Syndrome),” transvenous biopsy may occasionally be needed to differentiate SOS from hyperacute graft-versus-host disease after stem cell transplantation. Liver biopsy may be helpful if there is a low likelihood that toxicity is drug-induced and drug toxicity needs to be ruled out prior to work-up of another cause. Another indication for liver biopsy would be to obtain prognostic rather than etiologic information.

Hepatotoxicity Caused by Specific Anticancer Drugs

Many case reports linking anticancer drugs to hepatotoxicity derive from a period of time when diagnostic tools were more limited. Sensitive and specific immunoassays were developed for hepatitis A and hepatitis B in the 1970s, and for hepatitis C in the early 1990s. Thus much of the early chemotherapy literature could not rule out viral hepatitis as a contributor to abnormal liver tests. Imaging of the liver has made great strides. Prior to availability of current liver imaging modalities, occult involvement of the liver by tumor may have caused liver test abnormalities falsely attributed to certain drugs. Even current imaging modalities may not always detect tumor infiltration.102 The subject is further complicated by case reports of toxicity attributed to a single drug in a multidrug regimen. In this setting, it may difficult to determine which drug is responsible or if the toxicity is a result of an interaction of the drugs. The most clear-cut evidence of hepatotoxicity comes from studies using an anticancer drug as a single agent, usually determined during the initial clinical trials. For many chemotherapeutic agents, the initial clinical trials date back to the late 1950s or 1960s. Thus much of the information originated prior to availability of diagnostic studies for viral hepatitis and of the current imaging modalities. A final complication in cancer patients is the multitude of factors that may affect the liver (see Table 154-3).

Although a high frequency of case reports may reliably link toxicity to a given drug, sporadic reports in the older literature must be examined with great care. This is particularly problematic for host-dependent, dose-independent toxicity (idiosyncratic reactions)—these reactions usually occur with low frequency and there may be few or no reported cases in the recent literature.

Chemotherapeutic agents commonly cause transient increases in liver test abnormalities without clinical evidence of liver impairment. Table 154-4 lists drugs that may cause transient liver test abnormalities at standard doses of chemotherapy. At standard doses of chemotherapy, clinically apparent liver injury is relatively uncommon and our knowledge is based on case reports for many drugs. The right-hand column of Table 154-4 provides descriptions of the type of clinically apparent liver injury, which is mostly derived from case reports.

Table 154-4. Effect of Standard-Dose Chemotherapy on Liver Tests and Liver Injury.

Table 154-4

Effect of Standard-Dose Chemotherapy on Liver Tests and Liver Injury.

The relative infrequency of significant liver injury by chemotherapeutic agents is somewhat unexpected. One factor may be that many of these compounds target rapidly proliferating cells, whereas liver cells have a slow turnover. Another factor may be the relative strength of the detoxification pathways within the hepatocyte, which protect it from electrophilic metabolites and drug-induced oxidative stress.

Selected Anticancer Drugs and Modalities

Alkylating Agents

Cyclophosphamide

Cyclo- phosphamide was discussed earlier in “Sinusoidal Obstruction Syndrome (Hepatic Venoocclusive Syndrome).” Metabolism by CYP2C9 and 3A4 in hepatocytes yields 4-hydroxy-cyclophosphamide, which appears in the circulation. 4-Hydroxycyclophosphamide equilibrates with aldophosphamide, which follows two pathways: spontaneous decomposition to phosphoramide mustard and acrolein and metabolism by aldehyde dehydrogenase 1 (ALD1) to carboxyethyl phosphoramide mustard. Phosphoramide mustard is the putative antineoplastic moiety and acrolein is the proximate toxic metabolite. There is wide interindividual variation in metabolism of intravenously administered cyclophosphamide, which may contribute to the risk of SOS.51 The presumptive mechanism of toxicity in SOS is through hepatocyte metabolism of cyclophosphamide and formation of acrolein, the proximate toxic metabolite.41 Acrolein is toxic to endothelial cells in general,103, 104 but toxicity is greatest in the sinusoidal endothelial cells because of their proximity to hepatocytes, which metabolize cyclophosphamide.

Cyclophosphamide at standard doses is an uncommon cause of liver toxicity, with few reported cases of liver test abnormalities and rare case reports of clinically significant liver disease with hepatocellular necrosis. SOS is seen almost exclusively at high doses of cyclophosphamide and in conjunction with synergistic agents, such as busulfan, total-body irradiation, or BCNU. The incidence of SOS in high-dose regimens that contain cyclophosphamide is often in the range of 20% to 40%.

Busulfan

Busulfan is a weak alkylating agent. Experimental studies with busulfan have been difficult to interpret as the doses needed in vitro to induce either interstrand cross-linking or toxicity have been significantly above therapeutic plasma concentrations. Busulfan toxicity requires glutathione S-transferase-mediated conjugation to glutathione, which leads to oxidative stress.84 In an environment high in glutathione (eg, within the hepatocyte) oxidative stress seems to be a result of the busulfan- glutathione conjugate itself, whereas in a low glutathione environment, the oxidative stress is a result of profound glutathione depletion.

Busulfan significantly depletes whole-liver glutathione levels in vivo at doses comparable to those used in hematopoietic stem cell transplantation.84 Glutathione is diminished to a similar degree in hepatocytes and sinusoidal endothelial cell glutathione in vitro. The synergistic effect of busulfan on liver toxicity by conditioning regimens may be caused by glutathione depletion, oxidative stress, or both. Much of the toxicity associated with busulfan is a result of synergism with drugs such as cyclophosphamide or melphalan, which are both glutathione detoxified.

As a single agent, liver toxicity caused by high-dose busulfan is described as cholestatic,78 but only two case reports have described cholestatic liver disease at standard chemotherapeutic doses. The literature on high-dose busulfan as a single agent is too limited to comment on whether it can induce SOS at all, but the risk seems to be low,78 and it is certainly less than with dimethylbusulfan alone.105, 106 This may reflect differences in cellular toxicity: busulfan is equally toxic to hepatocytes and sinusoidal endothelial cells, whereas dimethylbusulfan toxicity is selective for sinusoidal endothelial cells (L.D. DeLeve, unpublished observation).

Dacarbazine

Dacarbazine is an inactive prodrug that can only be activated in the liver by microsomal metabolism. The initial microsomal metabolite transforms spontaneously through several steps to the proximate methyl-donating metabolite, either a methylcarbonium or a methyldiazonium ion. This drug is selectively toxic to sinusoidal endothelial cells, which can metabolically activate it, and it is detoxified by glutathione.63 Dacarbazine-induced SOS has been reported in at least 15 case reports. The disease varies from the usual presentation of SOS in that it is associated with peripheral eosinophilia and thrombosis of the central venules and veins. The peripheral eosinophilia has been seen after the first exposure to dacarbazine, which suggests that this is related to the toxicity rather than implicating a hypersensitivity reaction.

Melphalan

At standard chemotherapeutic doses, melphalan is not hepatotoxic. As a single agent, high-dose melphalan (140 mg/m2) is associated with either mild, transient elevations of serum aminotransferase and bilirubin79 or no abnormalities at all.80, 81 In high-dose conditioning regimens, it is associated with SOS.

Antimetabolites

Methotrexate

Concern about methotrexate-induced fibrosis was raised by a much-cited 1960 study of 273 children treated for acute leukemia.107 The study described a 31% incidence of fibrosis in autopsy cases prior to the use of chemotherapy between 1940 and 1947, and an increase to 78% incidence of fibrosis between 1948 and 1951 when the folic acid antagonists aminopterin and methotrexate were used. This frequency of liver injury has not been reproduced in recent decades. In fact, more recent studies have not found a significant incidence of liver injury (see below).

In high-dose methotrexate therapy for osteosarcoma, childhood acute lymphocytic leukemia, and adult non-Hodgkin's lymphomas, methotrexate may be infused in doses of 3 to 15 g/m2. During maintenance therapy, weekly oral doses of low-dose methotrexate may be interspersed with additional infusions of high-dose methotrexate. Thus very large cumulative doses can be achieved. Although high-dose methotrexate has a high incidence of transient elevation of aminotransferase, the current literature suggests that this does not result in chronic liver disease.23, 108–110 However, the incidence of fibrosis in this population may be underestimated. Liver tests correlate poorly with histologic abnormalities and clinical studies using high-dose methotrexate have not routinely performed liver biopsies after treatment with significant cumulative doses of methotrexate. Thus, clinically asymptomatic fibrosis likely goes undiagnosed. Given the significant numbers of long-term disease-free survivors, long-term toxicity is a concern. Two case reports of hepatocellular carcinoma following methotrexate-induced fibrosis were reported in 1977 and 1987111, 112; it should be noted that these reports preceded hepatitis C testing in leukemia patients, a population with a high incidence of hepatitis C. In a large meta-analysis, patients with psoriasis and rheumatoid arthritis had a 7% chance of progression of histologic abnormalities on liver biopsy for every gram of methotrexate,113 although some of the cases of methotrexate toxicity might actually be nonalcoholic steatohepatitis.114 Even if one assumes that asymptomatic fibrosis goes undiagnosed if systematic biopsies are not done, given the very high cumulative doses of methotrexate administered in cancer therapy, the incidence of significant liver injury would seem to be much lower than that suggested by the literature for rheumatoid arthritis and psoriasis.

Thiopurines

Liver injury from 6-mercaptopurine usually presents with jaundice that may be accompanied by pruritus. The injury is most commonly cholestatic, but hepatocellular necrosis may also be present. Although liver injury caused by 6-mercaptopurine has been frequently reported, the actual incidence of toxicity is difficult to estimate, but may be lower than suggested by early studies. In current regimens for acute lymphoblastic leukemia, 6-mercaptopurine is used within multidrug regimens, usually with other potential hepatotoxins such at l-asparaginase, cytarabine, and methotrexate. Liver injury is not singled out as a major toxicity in most of the larger studies, and when it does occur, it cannot be attributed to any particular drug in the regimen. Only one clear-cut case of drug-induced cholestatic hepatitis was found in an 18-year follow-up study of 396 patients with inflammatory bowel disease treated with 6-mercaptopurine (50 mg/d or 1.5 mg/kg).115

An injury described as SOS has been linked to 6-thioguanine, often in combination with cytarabine.116–119 These cases have most often resembled radiation-induced liver disease (RILD) in some of their clinical and histologic features. Patients present with hepatomegaly that is sometimes described as painful, and ascites, but without jaundice. Bilirubin is often normal or marginally elevated, as is seen in RILD. There are venoocclusive lesions and centrilobular congestion, but centrilobular hepatocyte atrophy is described rather than necrosis. Several of the cases also had underlying cirrhosis. There is also a case report of peliosis hepatis associated with 6-thioguanine plus cytarabine.120

It is not surprising that 6-thioguanine would be linked to this group of liver injuries. SOS, nodular regenerative hyperplasia, peliosis hepatis, and sinusoidal dilatation often share similar causes, and in some cases, up to all four have been described within the same liver. Overlap of these forms of liver injury is seen with azathioprine (all four lesions), urethane (peliosis and SOS), heroin (sinusoidal dilatation, sinusoidal and perivenular fibrosis), thorotrast (peliosis, venoocclusive lesions with hepatocyte atrophy), and oral contraceptives (sinusoidal dilatation and peliosis hepatis). In patients with acquired immunodeficiency syndrome infected with Bartonella spp, the initial change in peliosis hepatitis is sinusoidal dilatation caused by damage to sinusoidal endothelial cells. Damage to sinusoidal endothelial cells, and sometimes to hepatic venular endothelial cells, also seems to be the common link in the drug-induced cases of these four types of liver injury.64, 100, 121–124

When 6-mercaptopurine or azathioprine is given orally, it is subject to extensive first-pass metabolism in the intestine and liver by xanthine oxidase, whereas in target tissues the activity is determined by the balance between detoxification by thiopurine S-methyltransferase (TPMT) activity and toxification by hypoxanthine phosphoribosyltransferase. Patients treated with thiopurines who have low TPMT activity are at increased risk for hematologic toxicity (see review in ref. 125). However, a different mechanism may apply for hepatitis caused by oral administration of these thiopurines. Studies in four patients with 6-mercaptopurine hepatitis and in two patients with azathioprine hepatitis,125, 126 demonstrated TPMT levels in erythrocytes that were normal (erythrocyte TPMT normally correlates well with activity in the liver). In four cases of hepatitis from oral 6-mercaptopurine in children with acute lymphoblastic leukemia, average peak levels and AUC of 6-mercaptopurine were significantly lower than in controls. This suggests that increased first-pass clearance of orally administered thiopurines by the liver may predispose to accumulation of toxic concentrations in the liver.126 First-pass metabolism of oral 6-mercaptopurine is subject to large interindividual variation but the mechanism has not been determined.127

Fluorodeoxyuridine

Fluorodeoxyuridine or floxuridine (FUDR) infused into the hepatic artery for chemotherapy of hepatic metastases can cause sclerosing cholangitis. The incidence varies widely in the literature. In a recent study of 32 patients who received an average of 7.3 cycles of chemotherapy, the incidence of liver test abnormalities was 15.6%, and of severe biliary sclerosis was 9.3%.128 In another recent study in which 38 patients received two cycles of FUDR plus dexamethasone, 22% had elevations of aspartate aminotransferase and/or alkaline phosphatase, and 7% had bilirubin elevations greater than three times the upper limit of normal.129 Concomitant treatment with dexamethasone has been suggested to reduce liver injury from intraarterial FUDR, but this is not well established.129–131 Cholangiogram often shows a high-grade obstruction of the common hepatic duct, which may extend into the left and right hepatic ducts. Most commonly, there are also multiple strictures of the intrahepatic ducts. The strictures are thought to be a result of ischemia secondary to drug-induced damage to the peribiliary vascular plexus. The histologic changes in the hilar vessels are suggestive of organization of occlusive thrombi. The toxicity may be reversible upon discontinuation of therapy. Patients with rising bilirubin levels, pruritus, or sepsis may be successfully palliated with percutaneous transhepatic biliary drainage.

Miscellaneous Anticancer Drugs

l-Asparaginase

l-Asparaginase is used in the treatment of acute lymphoblastic leukemia. It is a microbial product derived from Escherichia coli or from Erwinia chrysanthemi. Liver toxicity is thought to be caused by inhibition of protein synthesis because of the asparaginase and glutaminase activity. Interference with hepatic protein synthesis is manifested by decreased serum albumin, coagulation factors, and lipoproteins. In the past a particularly high incidence of liver toxicity was described with elevated alkaline phosphatase and serum aminotransferase and depression of liver synthetic function; 40% to 87% of patients were found to have hepatic steatosis on autopsy, and this could be detected up to 9 months after the last dose of l-asparaginase.132, 133 The incidence of frank liver disease has decreased with the use of lower doses, although severe and fatal cases are still reported. In adults, the incidence of transient World Health Organization (WHO) grade I or II liver test (Table 154-5) abnormalities is approximately 50%,134 with few cases of significant liver test abnormalities.134, 135

Table 154-5. WHO Grades of Liver Test Abnormalities .

Table 154-5

WHO Grades of Liver Test Abnormalities .

A recent study of 245 patients with acute lymphoblastic leukemia randomized to conventional-dose or high-dose l-asparaginase as part of a multidrug regimen found antithrombin III levels less than 50% in 0% and 2.5%, antithrombin III levels of 50% to 70% in 1.7% and 10.3%, and fibrinogen less than 100 mg/dL in 8.4% and 10.3% of patients, respectively, in the conventional- and high-dose l-asparaginase treatment groups.136 Liver test abnormalities and liver injury were not listed among the major toxicity reactions. Another recent study of 377 patients with acute lymphoblastic leukemia treated with a regimen that included l- asparaginase, liver test abnormalities, and liver injury were not listed among the more frequent toxicities; among a subgroup of 43 patients in this study who did not complete the full 30-week treatment regimen, 2% developed hepatitis.136

Actinomycin D

Actinomycin D (or dactinomycin) is associated with SOS. There seems to be some degree of synergy with abdominal irradiation and the risk may correlate with the dose of radiation and perhaps also the dose of actinomycin D.137 The disease also occurs following chemotherapy with actinomycin D and vincristine in the absence of radiation. The incidence of SOS complicating chemotherapy for left-side Wilms tumors and for rhabdomyosarcoma is low.138, 139

Most of the reported cases of actinomycin D have been in patients with right-sided Wilms tumors.139 One series noted that liver toxicity occurred in particularly large right-sided tumors.140 One must consider the possibility that mass effect from these large right-sided nephroblastomata is impeding hepatic venous outflow. Another contributing factor may be intravascular extension of the nephroblastoma, causing Budd-Chiari syndrome, which may be underdiagnosed in the older literature.141 In one study, intravascular extension of tumor was diagnosed in only 40% of cases by ultrasonography, which the authors attributed to the lesser imaging quality of older ultrasonography equipment in their early cases.141 Budd-Chiari syndrome and SOS share many of the same clinical and histologic features. The unanswered question is whether hepatic venous outflow obstruction acts in concert with actinomycin D-induced SOS or whether the increased incidence of SOS with right-sided Wilms tumors is partially a result of undiagnosed Budd-Chiari syndrome.

Monoclonal Antibodies

Gemtuzumab ozogamicin is a new drug for acute myeloid leukemia that seems to have a significant incidence of SOS.142–144 In one case series, the overall mortality rate because of SOS in patients treated with gemtuzumab ozogamicin who did not undergo stem cell transplantation was 7%.143 In patients who underwent stem cell transplantation after gemtuzumab ozogamicin, the reported overall mortality from SOS was 11%.142 In patients who first underwent stem cell transplantation and later received gemtuzumab ozogamicin, the overall mortality rate from SOS was 30%.145 Gemtuzumab ozogamicin has not been on the market long and future studies need to more clearly define the incidence of SOS and risk factors.

Gemtuzumab ozogamicin is a conjugate of calicheamicin linked to the “humanized” monoclonal anti-CD33 antibody. CD33 is a myeloid surface antigen that is expressed on more than 90% of blast cells in acute myeloid leukemia, but not on hematopoietic stem cells or lymphoid cells. The presumptive mechanism of action of gemtuzumab ozogamicin is preferential binding to cells with the CD33 antigen, internalization of the conjugate, and release of the calicheamicin moiety by acid hydrolysis within lysosomes.146, 147 Calicheamicin has a methyltrisulfide group that is reduced by glutathione. The resulting diradical species binds to the minor groove of double-stranded DNA and causes sequence-selective oxidation of deoxyribose leading to DNA strand breaks.148, 149 The mechanism leading to gemtuzumab ozogamicin-induced SOS is undefined. Because both sinusoidal endothelial cells and Kupffer cells are of bone marrow origin.67, 150 it is possible that one or both may have CD33 surface antigen that binds gemtuzumab ozogamicin.

Radiation

Radiation-induced liver disease or radiation hepatitis is characterized by anicteric ascites after hepatic irradiation. With conventional fractionation of irradiation, RILD occurs at doses in excess of 30 to 35 Gy in adults. Children or adults who have recently undergone partial hepatectomy may develop RILD at lower doses.

Long-term radiation damage in various tissues may be a consequence of damage to microvascular endothelial cells, with apoptosis of the microvascular endothelial cells. Irradiation significantly depletes mitochondrial glutathione and causes oxidative damage to mitochondrial and nuclear DNA.82 Glutathione depletion enhances toxicity of radiation in vitro and in vivo.43, 82

The signs and symptoms resemble Budd-Chiari syndrome or SOS with hepatomegaly, weight gain, varying amount of ascites, and jaundice. Common histologic features are sinusoidal congestion, sinusoidal fibrosis, and subendothelial and adventitial fibrosis of the central veins.

RILD differs from SOS because of conditioning therapy in several ways. (1) The diagnostic criteria for SOS include elevations of bilirubin > 2 mg/dL and tenderness of the liver, whereas in RILD bilirubin elevations are usually minimal and right upper quadrant pain is much less pronounced.151, 152 (2) A characteristic histologic feature of SOS is centrilobular necrosis; in contrast, by the time RILD becomes clinically apparent there is loss of centrilobular hepatocytes and coagulative necrosis is uncommon.151–153 (3) In RILD, fibrin has been identified within the central vein by electron microscopy, but fibrin has not been definitively demonstrated by electron microscopy in SOS. (4) In SOS, onset of the first clinical signs usually occurs within 1 week of administration of the conditioning regimen, although the diagnosis can be made as late as 30 days after exposure to the drugs. Onset of RILD usually occurs 1 to 2 months after irradiation, although it may occur as early as 2 weeks or as late as 7 months afterwards.152 (5) Signs of SOS resolve within 30 to 60 days of onset in patients who survive the disease, whereas in RILD evidence of liver injury can persist for months after the insult has been discontinued.151, 154 Thus clinical presentation, histology, and time-course of disease distinguish RILD from SOS.

Hepatic irradiation can also act synergistically with chemotherapy to cause liver toxicity. Total-body irradiation contributes to the risk of SOS in hematopoietic stem cell transplantation, although the doses of irradiation used (10 to 16 Gy) are well below those that are hepatotoxic. Doses of irradiation > 20 Gy may also contribute to the incidence of SOS when used in conjunction with chemotherapy for Wilms tumor. No features have been identified that distinguish SOS because of high-dose combination chemotherapy from that resulting from hepatic irradiation plus high-dose chemotherapy.

Historically, RILD has limited use of hepatic irradiation in the treatment of intrahepatic cancers. However the development of threedimensional radiation therapy treatment planning allows much higher doses of radiation to be delivered to the liver with a low incidence of RILD.155

Hormones

Tamoxifen

Tamoxifen is a nonsteroidal drug with both antiestrogenic and estrogenic properties, which is widely used as a chemopreventive for breast cancer. It is metabolized by CYP3A4. Liver injury associated with tamoxifen includes nonalcoholic fatty liver disease,156–161 peliosis hepatis,162 hepatic insufficiency,163 and, perhaps, hepatocellular cancer.164, 165

Nonalcoholic fatty liver disease is the most common form of liver injury caused by tamoxifen. A Japanese study that screened 105 women on tamoxifen by annual abdominal computed tomography examination found a 38% incidence of fatty liver, which developed during the first 2 years of therapy in 35 of 40 cases (85%).161 Of the patients with fatty liver, 40% (16 of 40) had sustained elevations of aminotransferases. In addition there are three reported cases of cirrhosis in the presence of steatohepatitis by liver biopsy.157, 159 Future studies need to determine whether there is significant risk of steatohepatitis in tamoxifen users to warrant routine screening for nonalcoholic fatty liver disease.

Amiodarone, perhexiline maleate, and diethylaminoethoxyhexestrone (coralgil) are three drugs that cause steatohepatitis because of their physicochemical characteristics. Tamoxifen shares these same structural features and likely causes steatohepatitis through the same mechanism.166 These compounds are cationic amphiphilic compounds that are lipophilic and that have an amine that can become protonated. The unprotonated, lipophilic form easily crosses the mitochondrial outer membrane. In the acidic intermembranous space of the mitochondrion, the amine is protonated and the positively charged form moves across the mitochondrial inner membrane. The compounds are trapped inside the mitochondria and accumulate there. The accumulated drug inhibits mitochondrial β-oxidation (causing steatosis) and blocks electron transfer along the respiratory chain.166

Rats treated with tamoxifen develop nodular regenerative hyperplasia, hepatic adenomas, and hepatocellular carcinomas.167 There are substantial differences between rats and humans in the rate of tamoxifen metabolism and in the metabolites formed. To achieve clinically relevant serum concentrations, rats must be given high doses, which result in very high liver concentrations of tamoxifen and its metabolites.168, 169 Rats are also more susceptible to liver DNA damage than are women, albeit at liver concentrations that are much higher than those seen in women.169, 170 The propensity for hepatocellular cancer in rats may prove to be very species specific, because liver tumors are not found in mice or hamsters. There is currently no evidence in humans of a significant increase in the incidence of liver cancer (see review in ref. 167). Given the expanded indications for use of tamoxifen, future clinical studies will undoubtedly carefully monitor the risk of hepatocellular carcinoma.

Cyproterone Acetate

Cyproterone acetate is a synthetic progesterone derivative with antiandrogenic and progesterone-like activity. It is used in the treatment of advanced prostatic cancer. In a large surveillance study of 1,685 patients receiving cyproterone acetate for indications other than prostatic cancer, elevated liver tests were noted in 10% of patients treated with 50 mg/d and 20% of those receiving > 100 mg/d.171 A retrospective analysis of 78 patients receiving 50 mg/d of cyproterone acetate for advanced prostatic cancer reported elevation of alkaline phosphatase in 14% of patients without known liver involvement and elevated aminotransferases in 2.5%.172 This study did not report any cases of hepatitis. There are 18 case reports of cyproterone-associated hepatitis with 6 fatalities. In a review article,171 a report was cited of 96 hepatotoxic events with 33 fatalities attributed to cyproterone acetate173; however, I have been unable to obtain this report. There has also been one report of cirrhosis in a pediatric patient treated for precocious puberty.174

Cyproterone acetate is mitogenic, tumorigenic and induces DNA-adducts and DNA-repair synthesis in rat liver (see reviews in refs. 175 and 176). High levels of DNA-adducts are also formed in human hepatocytes from both genders. Formation of the reactive metabolite does not appear to be mediated by P450, but it has been suggested that metabolic activation might be through hydroxysteroid-sulfotransferases. Although long-term exposure to cyproterone acetate at high doses has the potential for inducing hepatocellular carcinomas, it would seem unlikely to be clinically relevant given the current life expectancy of patients with advanced prostate cancer.

Flutamide

In a multicenter study of 905 patients with total androgen blockade, flutamide was discontinued in 0.8% of patients because of abnormal liver tests (elevation greater than four times the upper limit of normal).177 Based on postmarketing surveillance, severe liver disease caused by flutamide has occurred in 46 patients with 20 fatalities.178 Based on estimated prescriptions, the rate of serious liver injury is 3 per 10,000 flutamide users. Liver disease presents with marked elevations of bilirubin and a wide range in elevation of serum aminotransferases. The predominant histologic feature on autopsy is the marked to massive hepatic necrosis.

Flutamide is a synthetic nonsteroidal that is a competitive antagonist of the androgen receptor. After oral administration, flutamide undergoes extensive first-pass metabolism with formation of several oxidized metabolites. Metabolism through CYP3A and CYP1A leads to formation of electrophilic metabolites.179 Experimental studies in rat hepatocytes suggest that the initial toxic event is through formation of electrophilic metabolites, which deplete hepatocyte glutathione.180 The depletion of glutathione is accompanied by oxidative stress. Flutamide was also shown to be toxic at the mitochondrial level, with depression of mitochondrial respiration and ATP formation, but the study did not report whether mitochondrial toxicity was a result of depletion of mitochondrial glutathione.

By agreement with the publisher, this book is accessible by the search feature, but cannot be browsed.

Copyright © 2003, BC Decker Inc.
Bookshelf ID: NBK12879

Views

  • Cite this Page

Related Items in Bookshelf

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...