Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
[CANCER RESEARCH 53, 4534--4541, October l, 1993] Major Difference in the Hepatocarcinogenicity and DNA Adduct Forming Ability between Toremifene and Tamoxifen in Female CrhCD(BR) Rats 1 G o r d o n C. H a r d , 2 M i c h a e l J. I a t r o p o u l o s , K e v i n J o r d a n , L e i l a R a d i , O l g i e r d P. K a l t e n b e r g , Anthony R. Imondi, and Gary M. Williams American Health Foundation, Valhalla, New York 10595 [G. (7. H., M. J. L, K. J., L. R., 0 P. K., G. M. W.], and Adria Laboratories, Columbus, Ohio 43216 [A. R. L] ABSTRACT The hepatoproliferative effects of 2 antiestrogens, tamoxifen and toremifene, were compared in a sequential 15-month study in which 2 doses of each compound were administered by daily gavage to female SpragueDawley rats for up to 12 months. The doses were 11.3 and 22.6 mg/kg for tamoxifen and 12 and 24 mg/kg for toremifene. There were scheduled sacrifices at 3, 6, 12, and 15 months, the latter including a 3-month recovery period from the 12th through the 14th month. In the chronic toxicity study, tamoxifen at 22.6 mg/kg produced 100% incidence of hepatoceUular carcinoma at the 12- and 15-month sacrifice intervals and 67% and 71% incidences at the ll.3.mg/kg dose. Sequential observations showed an increased incidence of glutathione S-transferase-positive foci of hepatocellular alteration by 3 months with tamoxifen in the absence of hepatotoxicity, with the first liver carcinoma appearing by 6 months of treatment. Unscheduled deaths occurring beyond 7.5 months in the tamoxifen treated groups were due in almost all cases to liver cancer. In striking contrast, toremifene did not produce any hepatoproliferative effects at 12- and 24-mg/kg dose levels, nor in a pilot study at 48 mg/kg. The 24.mg/kg dose of toremifene exerted an inhibiting effect on loci of hepatocellular alteration in rat liver detectable by glutathione S-transferase immunohistochemistry at 3 months and by conventional histology at 12 months. An antiproliferative effect was also evident in mammary gland and anterior pituitary where both toremifene and tamoxifen suppressed tumor incidence in comparison to the control group. The ability of these drugs to modify rat liver DNA after p.o. administration was investigated using the a2p-postlabeling assay. Administration of tamoxifen at 45 mg/kg for 7 days produced liver DNA nucleoside modifications represented by 7 spots on the autoradiogram. Unlike tamoxifen, toremifene did not produce any modified bases in rat liver DNA detectable by the 32p-postlabeling technique. The dose levels of tamoxifen that are strongly hepatocarcinogenic in the rat are compared with doses used in humans in various applications. Taking internal drug exposure into account, we conclude that the margin of safety for use of tamoxifen as an endocrine prophylactic agent for healthy, but breast cancer prone, women is questionable. INTRODUCTION Tamoxifen and toremifene are nonsteroidal triphenylethylene derivatives with chemical structures closely related to diethylstilbestrol (Fig. 1). They both have antiestrogenic and estrogenic properties (1, 2), depending on species and tissue type. Tamoxifen has been used extensively in the treatment of breast cancer, both in advanced cases and more recently as adjuvant treatment in early-stage disease (3, 4). Because women treated with tamoxifen were found to have a lower incidence of second and contralateral breast cancer, the possibility that tamoxifen could be an effective endocrine cancer preventive agent in women at increased risk of breast cancer has been advocated (5-9). Consequently, cancer prophylaxis trials in the United States, United Kingdom, and Italy have been initiated on groups of women at high Received4/8/93; accepted7/26/93. The costs of publicationof this article were defrayedin part by the paymentof page charges. This article must thereforebe hereby markedadvertisement in accordancewith 18 U.S.C. Section 1734 solely to indicatethis fact. 1This researchwas supportedfinanciallyby Adria Laboratories, Columbus, OH, and Farmos Group, Ltd., Turku, Finland. z To whomrequestsfor reprints shouldbe addressed,at AmericanHealthFoundation, 1 Dana Road, Valhalla, NY 10595. risk for developing breast cancer (8, 10, 11). This trial application as a chemopreventive in healthy women has occurred despite the knowledge that tamoxifen produced liver tumors in rats in the 2-year safety evaluation study conducted by the manufacturer (12). Toremifene has also been developed as a breast cancer treatment (13). It differs from tamoxifen only by the substitution of a chlorine atom for a hydrogen atom in an ethyl chain attached to one of the carbon atoms of the ethylene bond (Fig. 1). Toremifene has antiestrogenic and estrogenic properties similar to those of tamoxifen (1, 2, 14, 15), although it is cytostatic for human breast cancer cells at a dose that is 3 times higher than for tamoxifen (15). Currently toremifene is undergoing clinical evaluation in Europe and the United States (15-18). Recently tamoxifen was shown to be strongly hepatocarcinogenic to female rats (19, 20), confirming an observation noted in the manufacturer's safety evaluation report (12). Also, tamoxifen has been reported to be DNA-damaging (21, 22). The present nonclinical study was undertaken to compare the chronic effects of tamoxifen and toremifene under conditions of GLP, 3 specifically to determine whether toremifene shared the potent hepatocarcinogenic properties of its analogue. The chronic toxicity study design was a 1-year daily gavage trial with interim sacrifices, and including a 3-month recovery period. In addition, DNA adduct formation was evaluated in rat liver exposed in vivo to both compounds for 7 consecutive days using the 32p-postlabeling technique to measure modified bases (23). M A T E R I A L S AND M E T H O D S Chemicals. Tamoxifen citrate and toremifene citrate, each of 99% purity, were obtained from Adria Laboratories, Columbus, OH, and protected from light during storage. Fresh suspensions of the test compounds in 0.5% carboxymethylcellulose(Sigma, St. Louis, MO) were prepared each week and the concentration checked by UV-spectrophotometry on a monthly basis. Each compound was administered at 2 dose levels: 11.3 mg/kg (concentration 2.3 mg/ml) and 22.6 mg/kg (4.5 mg/ml) for tamoxifen, and 12 mg/kg (2.4 mg/ml) and 24 mg/kg (4.8 mg/ml) for toremifene. The dosing volume for each drug suspension was 5 ml/kg body weight. Test Animals and Animal Care. Female CrI:CD(BR) rats (Sprague-Dawley strain) aged approximately 4 weeks were obtained from Charles River Laboratories, Kingston, NY, for the carcinogenicity study. After acclimation and observation for 2 weeks, the animals were individually identified and assigned to groups using a randomization procedure based on body weights. The animals were identified by ear notch and toe clip and were housed 3 per cage on standard hardwood commercial rodent bedding (Beta Chip; Northeastern Products Corp., Warrensburg, NY) in temperature- (71 _ 5~ and humidity- (55 +-- 15%) controlled rooms with 12-h dark/low-level fluorescent light cycles changing at 7 a.m. and 7 p.m. Cage rack rotation was performed every 2 weeks. The animals were provided with NIH-07 pelleted diet (Purina Mills, Inc., St. Louis, MO) and water ad libitum supplied by Westchester County through an automated watering system. Certificates of analysis for diet, drinking water, and bedding, covering a standard range of microbial contaminants, inorganic elements, chlorinated hydrocarbons, organophosphates, 3 The abbreviations used are: GLP, good laboratory practice; H&E, hematoxylin and eosin; GST-P, glutathione S-transferase, placental form; DMBA, 7,12-dimethylbenzanthracene. 4534 Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1993 American Association for Cancer Research. LIVER EFFECTS OF TOREMIFENE AND TAMOXIFEN caac~a2 /~rOH 0H~ C=C "~-'~ \CH2CH3 Diethylstilbestrol CH3"NCH CHoO-Z~ CH3 / 2 ~ /=X ~',C:Ct~k ~, \cHzCH3 Tamoxifen CH3 ~,NCH2CH20 CH3/ "~C=CO CH2CH2CI Toremifene Fig. 1. Chemical structures of diethylstilbestrol, tamoxifen, and toremifene. nitrosamines, and aflatoxins, indicated compliance within recommended contaminant limits (24, 25). All animal maintenance and handling procedures were conducted in accordance with United States Department of Agriculture and NIH guidelines for humane care. The animal protocol was approved by the Institutional Animal Care and Use Committee prior to commencement. The American Health Foundation research animal facility is fully accredited by the American Association for the Accreditation of Laboratory Animal Care. Carcinogenicity Study Design. A total of 375 animals were assigned to one control group and 2 dose-level groups for each test compound as summarized in Table 1. Three animals from each treatment group served as sentinels for health monitoring. There were 4 scheduled sacrifice intervals: 3 interim at 3, 6, and 12 months, and a terminal sacrifice at 15 months. The animals were dosed p.o. by gavage once daily, 7 days a week for up to 12 months. Animals assigned to the 15-month sacrifice groups received no further treatment beyond 12 months, the final 3 months representing a recovery period. The results from a pilot study on toremifene are also included in this report because the data extend the upper dose-range of toremifene tested. In the pilot study, a 48-mg/kg dose of toremifene citrate was administered by daily gavage to female Sprague-Dawley rats for 12 months before termination. There was an interim sacrifice at 6-7 months. The conditions of this pilot study were identical to the bioassay described above except that it was not audited according to GLP. The numbers of animals are shown in Table l. Clinical Observations. Clinical observations for toxic or pharmacological effects were conducted daily and body weights recorded weekly for 8 weeks, biweekly for the next 16 weeks, and monthly thereafter for the study duration. Before commencement of dosing, and approximately every 2 months thereafter for the duration of the study, an ophthalmoscopic examination of all animals was conducted. Necropsy and Histopathological Evaluation. All animals were subject to a complete necropsy including those found dead or moribund. Euthanasia was performed with a CO2 atmosphere. At the scheduled sacrifices, organ weights were obtained from a range of tissues including liver, ovaries, uterus, and spleen. For histopathology, tissue specimens were taken from liver (1 section from each of 7 lobes), ovaries, oviducts, uterus, mammary gland, adrenals, pituitary, brain, tail bone, bone marrow (sternum), skin, and all grossly observable lesions, and fixed in 10% neutral buffered formalin. The eyes were also taken but fixed in Davidson's fixative (26). The carcasses and all other organs were stored in neutral buffered formalin. Fixed tissues were embedded in paraffin, sectioned, and stained with Mayer's H&E. Proliferative liver lesions were diagnosed according to the nomenclature of Maronpot et al. (27) and semiquantified. Thus, livers at each time interval were scored for foci of hepatocellular alteration, areas of hepatocellular alteration, hepatocellular adenomas, and hepatocellular carcinomas. Included in the conventional subgrouping of foci according to tinctorial variation (e.g., clear cell, basophilic, etc.) was a variant of the basophilic focus where the cytoplasmic basophilia was arranged in dense linear aggregates, known as a tigroid pattern (28). In addition, liver sections from the 3-month sacrifice were immunohistochemically stained for the demonstration of the placental form of GST-P as a marker for preneoplastic foci (29) using the avidin-biotin-peroxidase complex technique (30) and diaminobenzidine for detection of peroxidase binding. GST-P positive foci were quantitated per cm 2 of liver section by measuring the surface area of liver tissue on histological slides with a Zeiss Videoplan 2 digital image analysis system. DNA Adduct Detection by 32p Postlabeling. CrI:CD(BR) rats (3 per group) were administered toremifene and tamoxifen by gavage in 0.5% carboxymethylcellulose daily for 7 days at dose rates of 48 mg/kg and 45 mg/kg, respectively. Dose rates 2-fold higher than those used in the chronic toxicity assay were chosen because of the limited period of exposure in the 32p_ postlabeling study. On day 8, the animals were euthanatized by decapitation under CO2 sedation and the livers excised for DNA preparation by a modified chloroform-isoamyl alcohol extraction procedure (31). The isolated DNA was enzymatically digested into 3'-deoxynucleoside monophosphates. Subsequent incubation with nuclease P1 cleared the normal nucleosides leaving modified nucleosides free to be labeled with [32p]ATP. The modified nucleosides were purified and resolved by thin-layer chromatography and detected with autoradiography (32-34). Hepatic DNA from CrI:CD(BR) rats treated with 60 mg/kg of DMBA in olive oil (Bertolli USA, Inc., Secaucus, N J) was used as a positive control for DNA adduct detection. DMBA was purchased from Eastman Kodak Co. (Rochester, NY) and administered once by gavage to 3 rats, and the livers harvested 24 h later. Serum Drug Level Monitoring. During each of the scheduled sacrifices of the carcinogenicity study, and at the single sacrifice of the postlabeling assay, blood samples were obtained by cardiac puncture under CO2 anesthesia and analyzed for achieved serum drug levels. This was carried out by high-performance liquid chromatography after fluorescent activation (35). Statistical Evaluation. Statistical differences between groups in body and organ weights and lesion incidences were assessed by either Student's t test or the Xa test. Table 1 Group assignment Scheduled groups GLP study 1. Control 2. Toremifene, 12 mg/kg/day 3. Toremifene, 24 mg/kg/day 4. Tamoxifen, 11.3 mg/kg/day 5. Tamoxifen, 22.6 mg/kg/day Total animals Pilot study Toremifene, 48 mg/kg/day Total starting no. of rats 3 6 12 57 84 75 84 75 9 9 9 9 9 9 9 9 9 9 18 36 10 36 24 15 Total no. of unscheduled deaths Total no. of scheduled sacrifices 13 20 13 21 9 8 10 34 9 24 49 74 41 75 51 Sacrifice intervals (too.) and no. of rats sacrificed 290 375 36 10 5 4535 Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1993 American Association for Cancer Research. LIVER EFFECTS OF TOREMIFENE AND TAMOXIFEN Table 2 Summary of drug and metabolite levels in blood (ng/ml +- SD) Treatment (mg/kg) 3 too. Toremifene 12 24 Tamoxifen 11.3 22.6 6 too. Toremifene 12 24 Tamoxifen 11.3 22.6 12 mo. Toremifene 12 24 Tamoxifen 11.3 22.6 Parent compound N-Desmethyl metabolite 4-Hydroxy metabolite Total drug plus metabolite Metabolite to drug ratio 59.81 • 81.63 124.93 • 157.10 187.15 • 102.42 446.23 • 254.69 8.18 +-- 12.07 37.58 • 40.72 255.14 608.74 3.27 3.87 146.14 --- 92.71 244.55 • 265.51 88.96 • 39.49 129.94 • 68.19 35.97 • 19.88 47.99 • 25.08 271.07 422.48 0.86 0.73 30.25 • 4.29 53.32 • 58.92 189.46 • 13.75 487.47 • 554.22 17.29 • 0.59 21.79 • 13.16 237.00 562.58 6.84 9.55 71.97 • 9.47 119.66 • 8.66 80.31 • 4.23 132.56 • 14.19 29.87 • 3.55 39.29 • 3.46 182.15 291.51 1.53 1.44 50.32 • 27.35 144.10 • 90.98 206.67 • 87.59 579.92 • 211.48 27.31 • 17.88 56.03 -+ 35.67 284.30 780.05 4.65 4.41 137.84 + 40.65 172.01 • 102.97 136.28 --- 40.78 166.81 • 125.01 56.23 • 6.83 38.44 • 12.75 330.35 377.26 1.40 1.19 RESULTS Analytical Chemistry and Serum Drug Levels Concentrations of tamoxifen and toremifene in the dosing suspensions were within +_ 10% of the target levels. These drug preparations were homogeneous and remained stable for 7 days. Serum drug level data at 3, 6, and 12 months, shown in Table 2, indicated that for both tamoxifen and toremifene there was absorption and dose proportionality of the parent compounds and their metabolites at each time period. The serum levels of parent compound were significantly higher for tamoxifen than for toremifene, while the levels of metabolites were higher for toremifene. Thus, serum metabolite:drug ratios were much higher in the toremifene-treated groups, indicating greater or more rapid metabolic conversion of the parent compound than with tamoxifen. However, the total serum levels of parent drug plus metabolites were comparable between toremifene and tamoxifen, indicating comparable bioavailability, with a tendency towards a higher overall exposure to the former drug. With each compound, the major circulating metabolite was the N-desmethyl metabolite, the serum concentration of this product being significantly higher for toremifene than for tamoxifen at each time interval. The serum levels of the 4-hydroxy metabolite were fairly equivalent for each drug. Mortality Unscheduled deaths occurred in all treatment groups but were increased in those exposed to the high dose of each drug (Table 1). In the high-dose toremifene group the deaths were mainly gavage-related due to increased aggressiveness, which made the animals difficult to handle. With tamoxifen, mortality was attributed to gavage accidents due to increased aggressiveness as well as, in the later stages, the presence of liver tumors. logical effects known to induce aggressive behavior in rodents (37). A later observation was the presence of palpable abdominal masses in tamoxifen-treated groups by 12 months. A low incidence of drug-related, but not dose-related, alopecia was more pronounced in tamoxifen- than in toremifene-exposed animals. High-dose tamoxifen rats in particular exhibited an unthrifty condition (general wasting and unhealthy coat) related to the presence of liver tumors. Cataracts were observed by ophthalmoscopy in 4 rats given tamoxifen, 2 at each dose level. The cataracts were bilateral in 3 of these and unilateral in 1 low-dose animal. One rat in the high-dose toremifene group had a unilateral cataract at the 15-month examination i.e., after the 3-month recovery period. Postmortem Observations The most pertinent changes observed at necropsy were the presence of solitary or multiple liver masses in animals given tamoxifen at the 12- and 15-month sacrifice periods. In many cases, particularly at 15 months, almost all normal liver was replaced with tumor masses (Fig. 2). Compared with controls, other treatment-related changes with both drugs at necropsy were an increased incidence of cystic ovaries and decreased size of the uteri. Among the organ weights taken, a decrease in absolute (but not relative) liver weights was evident in all treated groups at 3 and 6 months, persisting at 12 months for both toremifene groups. In contrast, the absolute and relative liver weights for both tamoxifen groups increased significantly at 12 months, becoming more pronounced by the end of the recovery phase (Table 4). This real increase in liver weight in rats treated with tamoxifen correlated directly with the presence of liver masses. Absolute and relative uterine weights were consistently decreased in all treated groups at each sacrifice point In-Life Observations Table 3 Mean body weights of rats at scheduled sacrifices Results are expressed as means in g • SD. When compared with controls, all exposed groups in the carcinogenicity study showed significantly lower body weights, commencing from 5 weeks, continuing throughout the drug treatment period, and persisting until the end of the recovery phase (Table 3). Reduced body weight gain is a well recognized effect in female rats treated with antiestrogens (36). The most significant early clinical observation in exposed groups was increased aggressiveness, particularly in the highdose toremifene group, probably due to the drug-induced endocrino- Group no. 1 2 3 4 5 3 mo. 355.7 242.3 246.8 245.6 249.7 + 71.7 • 10.2 a "+ 20.5 a --- 24.3 a • 10.4 a 6 mo. 430.3 296.3 302.3 284.1 269.9 +• • • • 116.2 27.1 a 35.8 a 40.4 a 27.4 a 12 mo. 501.9 342.8 340.3 349.0 295.8 +• • • • 95.1 42.3 a 32.2 a 54.6 a 38.1 a a Statistically different from group 1, P < 0.01. b Statistically different from group 1, P < 0.05. 4536 Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1993 American Association for Cancer Research. 15 mo. 432.3 386.2 376.7 359.0 322.2 --- 105.8 • 48.7 • 64.6 • 43.2 b • 38.3 a LIVER EFFECTS OF TOREMIFENE AND TAMOXIFEN Table 4 Absolute (g) and relative (%) organ weights at 15 months Results are expressed as means in g or percentage _+ SD. Group no. 1 2 3 4 5 Liver Uterus Spleen 17.67 _+ 4.10g 4.21 _+ 1.01% 14.36 _+ 2.28 ga 3.72 -+ 0.41% 14.86 - 4.28 g 3.70 _+ 0.39% 30.03 -+ 30.46 g 8.32 _+ 9.34% 54.60 _+ 30.18 g" 18.00 _+ 11.23%`" 0.92 _+ 0.11 g 0.22 _+ 0.06% 0.48 -+ 0.08 a 0.13 -+ 0.03" 0.55 -+ 0.20 g`" 0.15 -+ 0.06%~ 0.55 -+ 0.32 ga 0.16 +_ 0.10%" 0.55 _+ 0.20 g`" 0.17 -+ 0.06% b 0.86 - 0.21 g 0.19 +- 0.09% 0.63 -+ 0.13 g`" 0.17 - 0.04%b 0.68 -+ 0.17g b 0.19 -+ 0.07% 0.77 +- 0.30 g 0.22 -+ 0.10% 1.02 -+ 0.24 g 0.33 + 0.10%" '~ Statistically different from group 1, P < 0.01. b Statistically different from group 1, P < 0.05. one animal had a solitary hepatocellular carcinoma (11%). The most Fig. 2. Liver from the high-dose tamoxifen group showing extensive involvement with multiple hepatocellular carcinomas. except 12 months. The only other organ weight change w a s in the relative spleen weights, w h i c h w e r e increased in the high-dose tamoxifen group at each sacrifice point. Histopathological Evaluation The pertinent pathological changes are described for each involved organ. Liver. The incidences o f hepatoproliferative lesions o b s e r v e d at each o f the 4 sacrifice points are s h o w n in Table 5. At 3 months, the only lesions present in H & E sections w e r e a low incidence o f solitary foci o f hepatocellular alteration in the controls (11%) and high-dose tamoxifen (33%). B y 6 months, the same b a c k g r o u n d o f solitary foci w a s seen in controls and toremifene groups, but the incidence of solitary foci in the high-dose tamoxifen group had risen to 89%, and c o m m o n type of focus in the tamoxifen groups at these early stages w a s the clear cell or mixed clear/eosinophilic cell type, but basophilic foci w e r e also present in the high-dose group at 6 months. It is important to note that no o b v i o u s hepatocyte necrosis, either focal or single cell death, w a s o b s e r v e d in nonneoplastic liver tissue o f tamoxifen-treated rats at these early time points. At 12 months, the b a c k g r o u n d incidence o f solitary foci in controls w a s 94% but statistically lower in the 2 toremifene groups at 5 3 % for the l o w dose and 2 0 % for the high dose (see Table 5). The l o w e r count o f foci in the latter groups w a s due to a marked reduction o f tigroid loci specifically. In the toremifene groups, there were no rats with multiple foci or areas o f hepatocellular alteration, or a d e n o m a s or carcinomas, even at the 4 8 - m g / k g dose. In contrast, both tamoxifen groups had very high incidences o f multiple foci, multiple areas, adenomas, and carcinomas, each o f these being 100% at the high dose. Thus, the incidence o f total liver n e o p l a s m s ( a d e n o m a s and carcinomas c o m b i n e d ) w a s 6 7 % for l o w - d o s e tamoxifen and 100% for the high dose. At the histological level, the t a m o x i f e n - i n d u c e d carcinomas Table 5 Incidence of hepatocellular proliferative lesions Foci Solitary Areas Multiple Solitary Multiple Adenomas Carcinomas Adenomas and carcinomas combined 3 mo. Group 1 Group 2 Group 3 Group 4 Group 5 1/9 (11)`" 0/9 0/9 0/9 0/9 0/9 0/9 3/9 (33) 0/9 0/9 0/9 0/9 0/9 0/9 0/9 6 too. Group 1 Group 2 Group 3 Group 4 Group 5 Pilot 1/9 (11) 2/9 (22) 1/9 (11) 3/9 (33) 8/9 b (89) 0/9 0/9 0/9 0/9 0/9 0/9 2/9 (22) 1/9(11) 12 too. Group 1 Group 2 Group 3 Group 4 Group 5 Pilot 17/18 (94) 19/36b (53) 2/10 ~ (20) 4/36 (11) 0/24 0/18 0/36 0/10 32/36c (89) 24/24 ~ (100) 3/18 (17) 3/36 (8) 1/10 (10) 6/36 (17) 0/24 0/18 0/36 0/10 22/36 c (61) 24/24 ~ (100) 10/13 (77) 13/20 (65) 8/13 (62) 5/21 (24) 0/9 2/13 (15) 0/20 4/13 (31) 1/20a (5) 1/13 (8) 0/20 0/13 14/21 b (67) 9/9 ~ (100) 15 mo. Group 1 Group 2 Group 3 Group 4 Group 5 0/13 16/21b (76) 9/9 c (100) 0/9 0/9 0/9 0/9 0/9 0/9 0/9 0/9 0/9 0/9 0/9 0/9 0/9 0/9 0/9 0/9 0/9 0/9 0/9 0/9 1/9 (11) 0/9 0/9 0/9 0/9 0/9 0/9 0/9 0/9 0/9 O/lO 0/9 0/9 0/9 0/9 0/9 0/9 0/9 1/9 (11) O/lO 0/9 0/9 0/13 a 3/21 (14) 0/9 0/18 0/36 0/10 21/36 c (58) 24/24 ~ (100) 0/5 0/13 0/20 0/13 13/21c (62) 9/9 c (100) 0/9 1/9 (11) O/lO 0/18 0/36 0/10 16/36c (44) 24/24 c (100) 0/5 0/18 0/36 0/10 24/36 ~ (67) 24/24 c (100) 0/5 0/13 0/20 0/13 13/21c (62) 8/9 ~ (89) 0/13 0/20 0/13 15/21 c (71) 9/9 c (100) `"Numbers in parentheses, percentages. b Statistically different from group 1 controls, P < 0.01. c Statistically different from group 1 controls, P < 0.001. d Statistically different from group 1 controls, P < 0.05. 4537 Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1993 American Association for Cancer Research. LIVER EFFECTS OF TOREMIFENE AND TAMOXIFEN were typified by trabecular areas of neoplastic hepatocytes, and the extensive involvement of liver tissue in all lobes was emphasized in the majority of cases. In one high-dose tamoxifen rat, there was metastasis of hepatocarcinoma to the lung. The pattern of lesion development differed little at 15 months after a 3-month recovery period. Toremifene groups had statistically fewer solitary areas of hepatocellular alteration than controls, and again, no liver neoplasms. The incidence of hepatocellular neoplasms in the low- and high-dose tamoxifen groups was 71% and 100%, respectively, indicating no regression during the treatment-free phase of the study. GST-P positive foci were enumerated in livers from the 3-month sacrifice. The results, summarized in Table 6, demonstrated a statistically significant increase in preneoplastic lesions in the high-dose tamoxifen group and a statistically significant reduction of foci in the high-dose toremifene group. Reproductive Tract. Compound-related effects were observed in ovaries, oviducts, and uteri, particularly at the 12-month sacrifice, with the 15-month groups exhibiting a degree of recovery in these various changes. The ovarian changes differed qualitatively between toremifene and tamoxifen treatments. Toremifene was associated primarily with an increase in the number of dilated secondary follicles and some ovarian atrophy. With tamoxifen, the changes were ovarian atrophy and increased numbers of atretic follicles. Other changes observed with both drug treatments were increased formation of cysts derived from either secondary follicles or bursa, and low incidences of stromal hypercellularity and granulosa cell hyperplasia/tumors (Table 7). Lesions diagnosed as granulosa cell tumor were observed in 1 rat of each of the treated groups at the 12-month sacrifice. There were none, however, at 15 months. Oviduct and uterine changes were evident from 3 months in all treated groups. The oviducts were typified by vacuole-like holes in the mucosal lining, possibly representing single cell loss. Uterine changes involved glandular atrophy. Mammary Glands. Tamoxifen and toremifene both produced an acinar atrophy in mammary tissue evident throughout the study. Control rats displayed a low incidence of focal glandular hyperplasia and mammary adenoma or carcinoma, which was markedly different from all drug-treated groups in which there were no proliferative lesions of mammary tissue (Table 7). Pituitary. At 12 and 15 months, there was a marked difference in the incidence of pars distalis hyperplasia and pitutary adenoma or carcinoma between control animals and those treated with either of the 2 drugs (Table 7). No neoplasms were observed in the pituitaries of drug-treated rats compared with up to a 33% incidence in controls. Pars distalis hyperplasia was observed in a few toremifene-treated rats at 15 months, but this was substantially lower than the 50% incidence in control animals. Other Tissues. Both tamoxifen and toremifene were associated with an increase in osteoporotic thinning of trabeculae at the epiphyseal growth plates of tailbone vertebrae at all sacrifice points, although a recovery trend was detectable at 15 months. Such bone loss is a Table 6 Incidence of GST-P positive liver loci at 3 months Group no. No. of rats with loci (%) No. of loci in liver lobe sections No. of foci/cm2 of liver tissue 1 2 3 4 5/9 (56) 3/9 (33) 1/9 a (11) 7/9 (78) 8 7 2 14 0.30 0.34 0.09 0.76 a Statistically different from group 1 controls, P < 0.05. b Statistically different from group 1 controls, P < 0.01. reflection of the known antiestrogenic properties of these 2 compounds in the rat (38). Avery low compound-related incidence of fiber degeneration in the optic lens was seen in tamoxifen-treated rats at 12 and 15 months. No differences were observed in the adrenal glands between control and treated animals. In the spleens of high-dose tamoxifen rats, reactive follicular hyperplasia secondary to extensive liver neoplasia was a common observation, accounting for the statistically significant spleen weight increase in this group. Unscheduled Deaths No liver neoplasms were observed in the toremifene-treated animals representing unscheduled deaths, but almost all had lung changes indicative of gavage accident. In the tamoxifen-treated groups, 1 of 9 (11%) unscheduled deaths had liver neoplasia at the low dose, and 13 of 24 (54%) at the high dose. In the latter group, all but one of the rats dying at 7.5 months or later had liver tumors. DNA Base Modification Typical autoradiograms resulting from the 32p-postlabeling assay of rat liver after 7 days of drug treatment are presented in Fig. 3. Liver samples from each animal were examined individually but the results within groups were uniform. No spots, representing DNA adducts, were found in any of the vehicle-treated control or toremifene-exposed rat livers. In contrast, 7 spots were identifiable in the autoradiograms of liver cell DNA from the 3 tamoxifen-treated rats, while 5 spots were present in liver DNA from the 3 positive control livers (DMBA-treated). DISCUSSION The most significant findings of the chronic toxicity study were the confirmation of the potent hepatocarcinogenicity of tamoxifen, and the lack of any carcinogenic activity of toremifene under the conditions of 12-month daily exposure at high dose in female rats. Tamoxifen-induced liver tumors were multiple and aggressive, leaving little normal tissue remaining by 15 months. Based additionally on clinical and necropsy observations, there was progression of the tamoxifeninduced liver tumors during the treatment-free period. In addition, the main cause of unscheduled death in tamoxifen-treated groups was liver neoplasia. The present report provides evidence, in compliance with GLP standards and in test groups of adequate size, for a major difference in the toxicology of these 2 drugs. The tamoxifen results concur with those from 2 other studies, which demonstrated liver tumors using a similar schedule of administration in Sprague-Dawley rats, but including a higher dose level (19, 20). Consistent with its hepatocarcinogenicity, tamoxifen induced an increased incidence of foci of hepatocellular alteration as early as 3 months in our sequential study. Foci have been established as the precursors of liver neoplasms, and their induction is predictive of carcinogenic potential (39), as in this case. In striking contrast to tamoxifen, toremifene was not only lacking in hepatocarcinogenic activity at daily doses ranging from 11 to 48 mg/kg but, in addition, this drug reduced the incidence of altered liver cell loci at the various interim sacrifice stages. This inhibitory effect on the development of liver foci was evident at 3 months using GST-P immunoreactivity and at 12 months by conventional H&E staining. 5 The main type of focus suppressed by toremifene was the type designated as having a tigroid pattern of staining (28). 9/9 a (100) The only proliferative change associated with drug treatment other 86 than hepatocellular involved the ovary, where a very low incidence of granulosa cell hyperplasia and tumor was observed at 12 months in 3.83 b each of the tamoxifen and toremifene groups. In the original safety evaluation of tamoxifen by the manufacturer, granulosa cell tumors 4538 Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1993 American Association for Cancer Research. LIVER EFFECTS OF TOREMIFENEAND TAMOXIFEN Table 7 Incidence of proliferative lesions in other tissues at 12 and 15 months 12 mo., group: 1 15 mo., group: 2 3 4 5 0/17 0/17 1/34 (3)a 1/34 (3) 0/10 1/10 (10) 3/33 (9) 1/33 (3) 0/24 1/24 (4) 0/13 0/13 4/18 (22) 1/18 (6) 0/36 0/36 0/10 0/10 0/36 0/36 0/24 0/24 Pituitary Pars distalis hyperplasia 6/18 (33) Adenoma 4/18 (22) Carcinoma 0/18 a Numbers in parentheses, percentage. 0/36 0/36 0/36 0/10 0/10 0/10 0/36 0/36 0/36 0/24 0/24 0/24 Ovary Granulosa cell hyperplasia Granulosa cell tumor Mammary Focal glandular hyperplasia Adenoma or carcinoma were recorded in mice (12). The absence of lesions diagnosed as granulosa cell tumors at 15 months after a 3-month recovery phase in our study suggests the possibility that this type of ovarian cell proliferation may have represented a reversible hyperplasia. The 32p-postlabeling study revealed that tamoxifen, but not toremifene, caused accumulation of modified nucleosides in rat liver, confirming evidence in similar studies from other laboratories that tamoxifen produces covalent D N A adduct formation in the target organ of carcinogenicity. Han and Liehr (21), who administered tamoxifen by i.p. injection to Sprague-Dawley rats at a rate of 20 mg/kg/day, found 2 D N A adducts after a single injection, and 4 additional adducts of increasing intensities at 3 and 6 consecutive days of treatment. Our study is in agreement in finding 7 adduct spots after 7 days of p.o. administration of a higher dose of tamoxifen. White et al. (22) also detected 7 adduct spots in Fischer 344 rat D N A following 1 2 3 4 5 0/20 0/20 1/13 (8) 0/13 0/21 0/21 0/9 0/9 2/13 (15) 2/13 (15) 0/20 0/20 0/13 0/13 0/21 0/21 0/9 0/9 6/12 (50) 4/12 (33) 3/12 (25) 1/19 (5) 0/19 0/19 2/12 (17) 0/12 0/12 0/21 0/21 0/21 0/9 0/9 0/9 p.o. gavage with 67.6 mg/kg/day for 7 days. The pattern of spots found in the latter and in our study appear to be very similar, inlcuding the relative location of the largest adduct. Tamoxifen has also been shown to cause D N A adduct formation in C57BL/6 and DBA/2 mouse strains, but at approximately one-third of the level detected in rat liver D N A (22). Furthermore, tamoxifen induced unscheduled D N A synthesis in hepatocytes from rats pretreated with the drug, although not in hepatocytes derived from untreated rats (22). In explanation, White et al. (22) suggest that the drug may induce its own activating enzyme system in vivo. T h e same authors have also demonstrated that tamoxifen is positive in the micronucleus test using a metabolically competent human cell line (22), indicating a potential for producing chromosomal anomalies in humans. On the evidence now assembled from the positive data published on carcinogenicity and short-term assays, tamoxifen should be regarded as a DNA-reactive agent. On the other Fig. 3. Autoradiogramsfrom 32p-postlabelingassay of rat liver. DNA from rats treated with tamoxifen (a), toremifene (b), dimethylbenzanthracene (positive control) (c), and carboxymethylcellulose vehicle (negative control) (d). 4539 Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1993 American Association for Cancer Research. LIVER EFFECIS OF TOREMIFENE AND TAMOXIFEN hand, toremifene appears to have no genotoxic properties to date. between rats and humans (44), then the dose level given prophylacUnlike tamoxifen, it lacks hepatocarcinogenic activity in rats, and has tically in women is higher than a hepatocarcinogenic dose in rats. been shown in our study with the Sprague-Dawley strain, and by Consequently, the margin of safety would seem insufficient to justify White et al. (22) with the Fischer 344, not to produce DNA adducts in the long-term use of this drug in relatively young cancer-free women rat liver. unless it can be proven that the rat data have no human relevance. In conclusion, our study has shown that tamoxifen is strongly Considering the minor difference in molecular structure between the 2 drugs, the strong hepatocarcinogenicity of tamoxifen, contrast- hepatocarcinogenic in female rats, with DNA adduct forming ability ing with a lack of hepatocarcinogenicity with toremifene, is striking. indicative of a DNA-reactive carcinogen. The potency of its carcinoThe basis for this difference in biological effect is not known, but we genic action coupled with DNA damaging propensity indicate that suggest that the carcinogenicity of tamoxifen may be due to activation tamoxifen has tumor initiating activity for the liver. In contrast, a close at the ethylene bond region of the molecule, as has been proposed for sructural analogue with very similar antiestrogenic properties, torediethylstilbestrol (40). This is in accordance with the finding that mifene, was without hepatocarcinogenicity or DNA modifying effects. diethylstilbestrol is also hepatocarcinogenic in female rats and pro- Like tamoxifen, toremifene suppressed spontaneous tumor formation duces DNA alterations in rat liver (41). Such a hypothesis raises the in endocrine-regulated organs, specifically mammary and pituitary possibility that toremifene has a lower capacity for generation of glands, a likely result of antihormonal effects. In addition, the inhigenotoxic metabolites because of steric hindrance by the presence of bition of liver foci by toremifene suggests that this drug may have a bulky chlorine atom adjacent to the ethylene double bond. Certainly, anticarcinogenic activity for liver as well. the carcinogenic action of tamoxifen cannot be ascribed solely to its hormonal activity because toremifene exerts qualitatively and quan- ACKNOWLEDGMENTS titatively similar hormone-related effects on female rat liver as The authors gratefully acknowledge Dr. Mirjana Djordjevic of the American tamoxifen (1, 2). In the absence of any conspicuous focal or single cell Health Foundation for preparation of the test articles, and Dr. Marjorie necrosis in nonneoplastic liver tissue, tamoxifen-induced hepatocarNeaderland, Diplomate, American College of Veterinary Ophthalmologists, cinogenicity cannot be linked to cytotoxicity. Conversely, the nonex- Animal Eye Clinic, Norwalk, CT, for performing the ophthalmological examiistent hepatocarcinogenicity of toremifene is not due to any lack of nations. biological activity in the liver. The difference in hepatocarcinogenicity is not explained by any differences in serum drug levels, as total REFERENCES drug/metabolite concentrations of the 2 test compounds were compa1. Kangas, L. Review of the pharmacological properties of toremifene. J. Steroid Biorable and the bioavailability of toremifene appeared to be higher than chem., 36: 191-195, 1990. 2. Kendall, M. E., and Rose, D. P. The effects of diethylstilbestrol, tamoxifen, and for tamoxifen. As observed here, toremifene is extensively metabotoremifene on estrogen-inducible hepatic proteins and estrogen receptor proteins in lized to N-desmethyltoremifene in the rat, but also in humans (13, 42). female rats. Toxicol. Appl. Pharmacol., 114: 127-131, 1992. Consequently, higher breast cancer therapeutic levels are required for 3. Jordan, V. C., and Murphy, C. S. Endocrine pharmacology of antiestrogens as antitumor agents. Endrocrine Rev., 11: 578-610, 1990. toremifene than for tamoxifen because the antitumor properties of 4. Riley, D., Baum, M., MacIntyre, J., Berstock, D., McKinna, A., Jackson, I., Sainsbury, N-desmethyltoremifene are slightly weaker than for the parent comJ. R. C., Wilson, A., Wheeler, T., Dobbs, J., Rees, G., Powles, T., Rubens, R., pound (13). This variation between the 2 drugs provides no basis for Haybittle, J., McPherson, K., and Houghton, J. The effect of adjuvant tamoxifen: the latest results from the Cancer Research Campaign adjuvant breast trial. Eur. J. Cancer, a difference in hepatocarcinogenicity, because toremifene proved 28A: 904-907, 1992. negative for liver pathology at a dose twice as high as that which 5. Cuzick, J., Wang, D. Y., and Bulbrook, R. D. The prevention of breast cancer. Lancet, 1: 83-86, 1986. induced 100% incidence of liver cancer by 12 months in tamoxifen6. Nayfield, S. G., Karp, J. E., Ford, L. G., Dorr, E A., and Kramer, B. S. Potential role treated rats. of tamoxifen in prevention of breast cancer. J. Natl. Cancer Inst., 83: 1450-1459, The hepatocarcinogenicity of tamoxifen must be placed in com1991. 7. Jordan, V. C. The strategic use of antiestrogens to control the development and growth parative pharmacokinetic perspective because of its medical indicaof breast cancer. Cancer (Phila.), 70: 977-982, 1992. tions in humans. Our concern regarding the cancer prophylactic ap8. Powles, T. J. The case for clinical trials of tamoxifen for prevention of breast cancer. Lancet, 340: 1145-1147, 1992. plication has been expressed recently (43). A no-observable-effect 9. Henderson, B. E., Ross, R. K., and Pike, M. C. Hormonal chemoprevention of cancer level for hepatocarcinogenicity was not established in the present in women. Science (Washington DC), 259: 633-638, 1993. study because both exposure levels were associated with liver cancer. 10. Fugh-Berman, A., and Epstein, S. Tamoxifen: disease prevention or disease substitution? Lancet, 340: 1143-1145, 1992. Notwithstanding, it is relevant to compare the hepatocarcinogenic 11. Powles, T. J., Tillyer, C. R., Jones, A. L., Ashley, S. E., Treleaven, J., Davey, J. B., and dose of tamoxifen in rats with the doses administered to humans. For McKinna, J. A. Prevention of breast cancer with tamoxifen--an update on the Royal adjuvant therapy of early-stage breast cancer, the standard dose is 20 Marsden Hospital pilot programme. Eur. J. Cancer, 26: 680-684, 1990. rag/day (6), which is the same as the daily total used long-term in 12. E. R. Barnhart (ed.). ICI Pharma Nolvadex. In: Physicians Desk Reference, Ed. 45, pp. 1070-1072. Oradell, NJ: Medical Economics, Inc., 1991. national trials of tamoxifen as a chemoprophylactic agent in women at 13. Kangas, L. Development and Biochemical Pharmacology of Toremifene, an Antiestrogenic Anticancer Drug, pp. 1-88. Turku, Finland: Farmos Group, Ltd., 1990. high risk of developing breast cancer (11). Using a scaling factor for body surface of 37 for a 60-kg woman, the therapeutic and prophy- 14. Kallio, S., Kangas, L., Blanco, G., Johansson, R., Karjalainen, R., Perila, M., Pippo, I., Sundquist, H., SodervaU, M., and Toivola, R. A new triphenylethylene compound, lactic doses of the drug convert to 12.3 mg/m2/day. Extrapolating from Fc-1157a. I. Hormonal effects. Cancer Chemother. Pharmacol., 17: 103-108, 1986. 15. Valavaara, R., Pyrh6nen, S., Heikkinen, M., Rissanen, P., Blanco, G., Th61ix, E., the rat study, the lowest hepatocarcinogenic dose of tamoxifen (11.3 Nordman, E., Taskinen, P., Holsti, L., and Hajba, A. Toremifene, a new antiestrogenic mg/kg/day), converted to 73.5 mg/m2/day using a scaling factor of 6.5 compound, for treatment of advanced breast cancer. Phase II study. Eur. J. Cancer for the rat, is 6 times higher than the standard therapeutic and proClin. Oncol., 24: 785-790, 1988. phylactic doses administered to women. Taking this a step further to 16. Gundersen, S. Toremifene, a new antiestrogenic compound in the treatment of metastatic mammary cancer. A phase II study. J. Steroid Biochem. 36: 233-234, 1990. account for internal drug exposure, our data show that the mean serum 17. Kaufman, M., Possinger, K., Illiger, H. J., Schmid, H., Hietanen, T., Johansson, R., Pyrhonen, S., Valavaara, R., and Sindermann, H. Toremifene: clinical phase II/III levels of tamoxifen in the ll.3-mg/kg group of rats ranged from about trials of a new antiestrogen in patients with advanced breast cancer. Contrib. Oncol., 72 to 146 ng/ml between 3 and 12 months of treatment. This range 37: 50-57, 1989. appears to be comparable to the 125 ng/ml of tamoxifen found by 18. Pyrhonen, S., Phase III studies of toremifene in metastatic breast cancer. Breast Cancer Res. Treat., 16: 41-46, 1990. Robinson et al. (42) in the serum of women receiving 10 mg twice 19. Williams, G. M., Iatropoulos, M. J., Djordjevic, M. V., and Kaltenberg, O. P. The daily, or 12.3 mg/m2/day. If a factor of 6 is accepted to account for the triphenylethylene drug tamoxifen is a strong liver carcinogen in the rat. Carcinogencombined differences in liver tissue volume and liver plasma flow rate esis (Lond.), 14: 315-317, 1993. 4540 Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1993 American Association for Cancer Research. LIVER EFFECTS OF TOREMIFENE AND TAMOXIFEN 20. Hirsimaki, E, Hirsimaki, Y., Nieminen, L., and Payne, B. J. Tamoxifen induces hepatocellular carcinoma in rat liver: a 1-year study with two antiestrogens. Arch. Toxicol., 67: 49-54, 1993. 21. Han, Y., and Liehr, J. G. Induction of covalent DNA adducts in rodents by tamoxifen. Cancer Res., 52" 1360--1363, 1992. 22. White, I. N. H., de Matteis, E, Davies, A., Smith, L. L., Crofton-Sleigh, C., Venitt, S., Hewer, A., and Phillips, D. H. Genotoxic potential of tamoxifen and analogues in female Fischer F344/n rats, DBA/2 and C57BL16 mice and in human MCL-5 cells. Carcinogenesis (Lond.), 13: 2197-2203, 1992. 23. Reddy, M. V., and Randerath, K. Nuclease Pl-mediated enhancement 32P-postlabeling test for structurally diverse DNA adducts. Carcinogenesis (Lond.), 7: 1543-1551, 1986. 24. National Toxicology Program. General Statement of Work for the Conduct of Toxicity and Carcinogenicity Studies in Laboratory Animals, Appendix H, p. 5. April 1987. 25. Rao, G. N., and Knapka, J. J. Contaminant and nutrient concentrations of natural ingredient rat and mouse diet used in chemical toxicology studies. Fund. Appl. Toxicol., 9: 329-338, 1987. 26. Toyoda, K., Furukawa, E, Hasegawa, R., Sato, H., Jang, J. J., Takahashi, M., and Hayashi, Y. The diagnostic value of Davidson fixation method on toxicological evaluation of eye lesion--histopathological findings of cataract formation following galactose diet intake in rats (in Japanese). Bull. Natl. Inst. Hygienic Sciences (Tokyo), 104: 82-86, 1986. 27. Maronpot, R. R., Montgomery, C. A., Boorman, G. A., and McConnell, E. E. National Toxicology Program nomenclature for hepatoproliferative lesions of rats. Toxicol Pathol., 14: 263-273, 1986. 28. Bannasch, P., Benner, U., Enzman, H., and Hacker, H. J. Tigroid cell foci and neoplastic nodules in the liver of rats treated with a single dose of aflatoxin B1. Carcinogenesis (Lond.), 6: 1641-1648, 1985. 29. Sato, K., Kitahara, A., Satoh, K., Ishikawa, T., Tatematsu, M., and Ito, N. The placental form of glutathione S-transferase as a new marker protein for preneoplasia in rat chemical hepatocarcinogenesis. Jpn. J. Cancer Res., 75: 199-202, 1984. 30. Hsu, S-M., Raine, L., and Fanger, H. Use of avidin-biotin-peroxidase complex (ABC) in immunoperoxidase techniques: a comparison between ABC and unlabeled antibody (PAP) procedures. J. Histochem. Cytochem., 29: 577-580, 1981. 31. Marmur, J. A. Procedure for the isolation of deoxyribonucleic acid from microorganisms. J. Mol. Biol., 3: 208-218, 1961. 32. Gupta, R. C., Reddy, M. V., and Randerath, K. 32p-postlabeling analysis of nonradioactive aromatic carcinogen-DNA adducts. Carcinogenesis (Lond.), 3: 10811092, 1982. 33. Gupta, R. C. Enhanced sensitivity of 32p-postlabeling analysis of aromatic carcinogen:DNA adducts. Cancer Res., 45: 5656--5662, 1985. 34. Gupta, R. C. 32P-postlabeling assay to measure carcinogen-DNA adducts. Prog. Exp. Tumor Res., 31: 21-32, 1987. 35. Holleran, W. M., Gharbo, S. A., and DeGregorio, M. W., Quantitation of toremifene and its major metabolites in human plasma by high-performance liquid chromatography following fluorescent activation. Anal. Lett., 20: 871-879, 1987. 36. Perry, B. N., McCracken, A., Furr, B. J., and MacFie, H. J. Separate roles of androgen and oestrogen in the manipulation of growth and efficiency of food utilization in female rats. J. Endrocrinol., 81: 35-48, 1979. 37. Sigg, E. B., Day, C., and Colombo, C. Endocrine factors in isolation-induced aggressiveness in rodents. Endocrinology, 78: 679-684, 1966. 38. Feldmann, S., Minne, H. W., Parvizi, S., Pfeifer, M., Lempert, U. G., Bauss, E, and Ziegler, R. Antiestrogen and antiandrogen administration reduce bone mass in the rat. Bone Miner., 7." 245-254, 1989. 39. Williams, G. M. The significance of chemically induced hepatocellular altered foci in rat liver and application to carcinogen detection. Toxicol. Pathol., 17: 663-674, 1989. 40. Metzler, M., and Degen, G. H. Sex hormones and neoplasia: liver tumors in rodents. Arch. Toxicol. Suppl., 10: 251-263, 1987. 41. Williams, G. M., Iatropoulos, M., Cheung, R., Radi, L., and Wang, C. X. Diethylstilbestrol liver carcinogenicity and modification of DNA in rats. Cancer Lett., 68: 193-198, 1993. 42. Robinson, S. P., Langan-Fahey, L., Johnson, D. A., and Jordan, V. C. Metabolites, pharmacodynamics and pharmacokinetics of tamoxifen in rats and mice compared to the breast cancer patient. Drug Metab. Dispos., 19: 36-43, 1991. 43. Williams, G. M., latropoulos, M. J., and Hard, G. C. Long-term prophylactic use of tamoxifen: is it safe? Eur. J. Cancer Prevention, 1: 386--387, 1992. 44. Iatropoulos, M. J. Accelerated rodent bioassay predictive of chemical carcinogenesis. Exp. Toxicol. Pathol., 44: 481-487, 1992. 4541 Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1993 American Association for Cancer Research. Major Difference in the Hepatocarcinogenicity and DNA Adduct Forming Ability between Toremifene and Tamoxifen in Female Crl:CD(BR) Rats Gordon C. Hard, Michael J. Iatropoulos, Kevin Jordan, et al. Cancer Res 1993;53:4534-4541. Updated version E-mail alerts Reprints and Subscriptions Permissions Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/53/19/4534 Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at [email protected]. To request permission to re-use all or part of this article, contact the AACR Publications Department at [email protected]. Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1993 American Association for Cancer Research.