Tamoxifen is by far the most clinically tested antiestrogenic drug currently used as adjuvant therapy for breast cancer and it continues to provide considerable benefit in this setting. The balance from clinical trials indicates a strong association between the use of tamoxifen and an increase in uterine tumors (three to sixfold). In rats, tamoxifen is a mutagenic, genotoxic hepatocarcinogen.
ABSTRACT: Tamoxifen is by far the most clinically tested antiestrogenic drug currently used as adjuvant therapy for breast cancer and it continues to provide considerable benefit in this setting. The balance from clinical trials indicates a strong association between the use of tamoxifen and an increase in uterine tumors (three to sixfold). In rats, tamoxifen is a mutagenic, genotoxic hepatocarcinogen. These actions are not related to its estrogen antagonist activity but have been shown to be as a result of metabolic activation of this drug by cytochrome P450 enzymes, resulting in irreversible binding to cellular DNA. The mechanism of endometrial cancer associated with tamoxifen treatment is unclear, although there are two plausible hypotheses: (1), tamoxifen causes damage and mutation to DNA in uterine cells or (2), it promotes the development of endometrial tumors through its estrogen agonist activity. The evidence for a genotoxic effect of tamoxifen in the uterus is highly contentious and, on balance, we have concluded that it is more likely that the estrogenic effects of tamoxifen promote tumor development.[ONCOLOGY(Suppl 5):14-22, 1998]
The incidence of breast cancer shows annual increases in the United Kingdom and its frequency in the United States is much higher than, for example, Thailand or Japan. The antiestrogenic drug, tamoxifen, ((Z)-2-[4-(1,2-diphenyl-1-butenyl)-phenoxy]-N,N-dimethylethanamine) (Figure 1) is a drug of choice used in adjuvant therapy for breast cancer. Women with negative lymph nodes and estrogen receptor-positive tumors, show a survival advantage that lasts for at least 5 years during tamoxifen therapy. In the United States and United Kingdom, the usual human dose is 20 mg daily, although in the rest of Europe, dosages of 20, 30, or 40 mg have been used. This variation in the dosage of tamoxifen can make comparisons between the success of treatments and adverse effects between European countries and the United States more difficult.
A meta-analysis of trials of tamoxifen in breast cancer patients involving approximately 75,000 individuals clearly shows improved disease-free survival times and an approximate 39% reduction in the incidence of contralateral primary breast cancer. This suggested that tamoxifen could have a role in the prevention of breast cancer. As adjuvant therapy, it has few serious undesirable side effects . Epidemiological evidence indicates, however, that long-term administration of tamoxifen to breast cancer patients results in an increase in the incidence of endometrial tumors [reviewed in ref 6]. After an initial report in 1985 of an association between the treatment of breast cancer patients with tamoxifen, and an increase in endometrial cancer, there have been a number of case and cohort studies to support the view that tamoxifen treatment and endometrial cancer are causally related. In 1994, the National Surgical Adjuvant Breast and Bowel Project (NSABP B-14) in Canada and the United States reported on the rates of endometrial and other cancers in 2,843 breast cancer patients randomly assigned tamoxifen (20 mg/day) or a placebo. The average time of study was 5 to 8 years. Over this period, 23 endometrial tumors were found in the tamoxifen treatment groups and two in the placebo group. It was concluded in this study that the average hazard rate through followup was 1.6 per thousand in the tamoxifen group versus 0.2 per thousand in the placebo controls. A year later these findings were confirmed by the results of the Stockholm breast cancer study group reported by Rutqvist et al. In this trial, 2,729 breast cancer patients under the age of 71 years were given tamoxifen (40 mg daily) or a placebo. The median follow up time was 9 years and showed tamoxifen treatment resulted in nearly a six-fold increase in endometrial cancer.
The effects of tamoxifen on the endometrium will have marked consequences on the risk-benefit analysis for the putative chemopreventive treatment of healthy women. In order to substantiate these data, analysis of tumor registries in the US involving a much larger cohort of 87,323 women with breast cancer has been investigated by the Surveillance Epidemiology and End Results (SEER) program. Of these, 14,358 women received hormones (tamoxifen) for their first course of therapy. Epidemiological results indicated an increase in uterine cancers in this group. As with the earlier studies, the follow-up time for the treated patients is generally less than 10 years, limiting the evaluation of long-term effects of this drug.
In view of the long time period often needed for the development of tumors in humans, it is perhaps surprising that in the case of endometrial tumors, the onset is rapid, often within 5 years from the time from start of tamoxifen. There is a strong association between estrogen replacement therapy and endometrial cancer, which is also seen in the first 5 years of initiation of treatment, suggesting an estrogen promoting-like action of tamoxifen. In the Stockholm trial, however, there was a marked time dependent increase in the cumulative incidence of endometrial tumors in treated women over a longer 15-year period.
In the Stockholm study, there was about a three-fold increase in gastrointestinal cancers in the tamoxifen treated patients. To our knowledge, this is the only study that has shown an increase in GI tract tumors in a tamoxifen treated population. This was not detected in either the SEER or the NSABP trials and none of these studies found any significant increase in liver cancers amongst the tamoxifen treated groups.
A number of other trials, results for which were published prior to the Stockholm Breast Cancer Study Group or NSABP findings and which used smaller numbers of breast cancer patients, failed to find any increase in endometrial cancers. Examples include the Nolvadex Adjuvant Trial Organization (NATO) involving 564 patients treated with tamoxifen and the Scottish trial with 374 treated women (both 20 mg/day). These studies had follow-up times of 5 or more years. While the NATO trial failed to find endometrial cancers, three uterine sarcomas were reported in the Scottish tamoxifen treated group. The numbers of patients taking part were not sufficiently large to give the analytical power of the US and Swedish investigations. It appears that it is women of less than 50 years of age who are more susceptible to tamoxifen-related endometrial tumors than women over the age of 50, who have a relative risk only slightly above that of untreated women. From the epidemiological evidence, it was concluded by the International Agency for Research on Cancer (IARC) that there was sufficient evidence for the association of tamoxifen therapy and endometrial tumors in breast cancer patients, leading to classification of this drug as a Class 1 human carcinogen.
The chlorinated structural analogue of tamoxifen, toremifene (Figure 1) is under investigation in breast cancer patients in a number of Phase III trials using daily doses of 60 to 240 mg. Toremifene was not yet registered in many countries when these trials were initiated, and the number of individuals being given this drug is small, relative to the SEER and NSABP studies. Although toremifene is not genotoxic or carcinogenic in the rat liver assay (see below), the numbers of patients in the Phase III studies are presently too small to make realistic assessments of the ability of these compounds to result in endometrial tumors in women. With larger numbers of participants, epidemiological results should give a fascinating insight into the probable mechanisms of endometrial tumors caused by tamoxifen. (Note: Toremifene is now registered in most countries.)
At the cellular level, the actions of tamoxifen are not completely understood. In breast tissue, the antiproliferative effect of the drug is primarily mediated through inhibition of estrogen activities by binding to the estrogen receptor. Tamoxifen allows dimerization of the receptor to occur but classically blocks transcriptional activation. Activation of the estrogen receptor is dependent on the differential action of the N- and C-terminal transcription activation functions AF-1 and AF2, respectively. Tamoxifen inhibits AF2 activity and functions as an estrogen antagonist where AF2 is required. However, AF2 is not required for all promoters, allowing this drug to demonstrate partial estrogen agonist activity. Recently, a second estrogen receptor gene ER-b has been reported . ER-b binds to DNA and can dimerize with ER-a. This adds further to the degree of complexity to transcription activation/inhibition in response to tamoxifen. Apart from these actions, tamoxifen can stimulate the paracrine secretion of transforming growth factor b and insulin-like growth factor 1 from stromal cells as well as inhibiting protein kinase C and calmodulin-dependent cAMP phosphodiesterase. In human endometrial tissues tamoxifen is thought to act as a estrogen agonist, inducing cell proliferation and thus promoting any endogenous DNA lesions.
In rats, several lifetime bioassays of tamoxifen have shown that this drug results in the formation of liver carcinomas but not endometrial tumors.[18,19] With respect to liver tumors, there appears to be little difference in the susceptibility between male or female animals to this drug. Mice given similar doses of tamoxifen do not get such tumors. There was a need to establish the mechanism for the formation of the liver neoplasms in the rat and to establish whether these factors operate in human liver or reproductive tissues, in order to permit better risk-benefit analysis for the many thousands of healthy women who are taking this drug in the current chemopreventative trials.
Using the 32P-postlabeling technique, it was shown that DNA adducts were formed in the livers of two rodent species following tamoxifen administration.[21,22] The level of such DNA damage in mouse liver was about one-quarter to one-third of that in rats. Following tamoxifen treatment in rats, a DNA adduct pattern consisting of approximately 12 individual spots was consistently detected in livers of three different strains. Using the postlabeling assay, no adducts could be detected in the rat liver at doses less than 5 mg/kg/day given for 7 days. Using the newly developed and highly sensitive technique of accelerator mass spectrometry, however, the binding of [14C]tamoxifen to rat liver DNA has been shown following a single dose of 0.3 mg/kg, comparable to the human therapeutic dose. It should be noted, however, that because of the very much shorter half-life of tamoxifen in rodents, rats must be given ~20 mg/kg daily to achieve similar plasma levels (~250 ng/mL) as women taking this drug therapeutically.
Three strains of female rats, Fischer, Wistar, and Lewis, fed 420 ppm tamoxifen, corresponding to a daily dose of ~40 mg/kg, for 30, 90, and 180 days, showed a cumulative increase in numbers of adducts with time, from approximately 500 adducts/108 nucleotides at 30 days to almost 3,000 adducts/108 nucleotides at 180 days. Loss of 32P-postlabeled adducts from the liver after cessation of dosing with tamoxifen takes many weeks, suggesting either that DNA repair mechanisms are saturated or that removal of tamoxifen adducts is slow. It has been shown that the loss of such adducts from mouse liver is very much faster than from rat. (Figure 2)  During 2 year exposure of mice to dietary tamoxifen, the level of the major hepatic DNA adducts did not accumulate, so that by 1 and 2 years, DNA damage was barely above background levels. The reason for this is not known, although it was suggested that it could be due to the more efficient DNA repair systems in the liver cells of mice. Species differences in the rate of repair (or stability) of DNA adducts may have important consequences in the susceptibility of women to the side effects of this drug. At present, there are no data on the stability of potential tamoxifen-DNA adducts in human cells.
After 3 months of dietary tamoxifen exposure, the numbers of altered hepatic foci increase in rat liver, compared with the background control levels.[23,25] This trend increased after 6 months, with the numbers and total area of glutathione-S-transferase P (placental form) expressing foci increasing further, in parallel with the increase in amounts of liver DNA adducts. There were marked strain differences with respect to time to tumor development. Wistar, Lewis, and Sprague Dawley animals showed a rapid onset within 6 to 10 months, whereas neoplasms in Fischer animals take longer to develop. Tamoxifen and toremifene increase liver weight as a percentage of body weight. This is associated with liver cell proliferation in some studies but not in others. No altered hepatic foci were seen in the livers of toremifene dosed rats.
If cell proliferation plays a role in promoting tumor formation in the rat liver, then an increase in labeling index would be consistent with the presence of tumors. It may be highly relevant that preneoplastic foci also showed a progressive higher incidence of estrogen receptor depletion, as they increased in size, indicating that this effect was associated with promotion of foci to tumor. Of the tamoxifen-induced adenomas and carcinomas in rats, 90% were depleted of normal hepatocyte nuclear estrogen receptor expression.
Tamoxifen given to rats as a single dose does not act as a tumor initiator in the liver, but studies have shown that after an initiation with N-nitrosodiethylamine, tamoxifen can act as a classical tumor promoter. Toremifene can also act as a tumor promoter after nitrosamine initiation, but not as an initiator. Perhaps more importantly from a human risk point of view, it has been shown that the DNA lesions formed in the liver by a short 3 month exposure to dietary tamoxifen, can be promoted by the classical promoter phenobarbitone.
After tamoxifen exposure for 3 months (420 ppm in the diet) rat uteri were decreased in size, with an absence of glands in the endometrium, reflecting its estrogen antagonist activity. Release of the antagonistic action of tamoxifen by discontinuing treatment appeared to result in a compensatory hyperplasia particularly in the myometrium. In mice, tamoxifen acts as an agonist of estrogenic activity in the uterus, resulting in cystic hyperplasia of the endometrium. Continuous exposure of B6C3F1 mice to tamoxifen for up to 2 years resulted in hyperplasia of the epithelium of the uterine endometrium for the first 3 months. From 3 months onward, there is atrophy of both the endometrial and myometrial components. Based on these results, neither the rat nor the mouse would appear to be a useful model for the uterine responses seen in women taking tamoxifen. Administration of tamoxifen at low doses (1-50 mg/pup/day) to CD-1 mice, on days 1-5 postnatally, however, resulted in subsequent proliferative lesions of the oviduct and at 16 months, adenosarcoma was observed in 58% of the mice.
Administration of toremifene to rats resulted in levels of liver DNA damage in the order of 150-fold lower than tamoxifen following short-term dosing. After 18 months dietary exposure to toremifene, DNA lesions in the liver were still 30-fold lower than with tamoxifen dosing. Low levels of DNA damage were also seen in the kidney. Neither tamoxifen or toremifene resulted in DNA damage in the uterus as assessed by 32P-postlabeling. It was noted by Li et al that treatment with either tamoxifen or toremifene induced the level of endogenous DNA adducts in uterine tissues, but the relevance of this to possible uterine tumors in women is not clear.
When given to rats, droloxifene (3-hydroxytamoxifen), which is not a natural metabolite of tamoxifen, results in no detectable DNA damage and long-term administration results in no liver tumors. These data strongly suggest that, at least in rat liver, there is a causal relationship between tamoxifen-induced DNA adduct formation and the induction of liver tumors and this is not related to the antiestrogenic properties of these compounds.
Although the postlabeling assay showed that DNA adducts were formed in rat and mouse livers following exposure to tamoxifen, the assay does not prove unequivocally that the adducts are formed by reactive metabolite(s) of tamoxifen covalently bound to DNA. To address this problem, rats and mice were treated with [14C] tamoxifen and killed 24 hours later. DNA was isolated from the livers and digested to the nucleotide level. Following HPLC analysis of rat and mouse nucleotides it was shown that [14C] label did not elute with the normal nucleotides or tamoxifen, but in the part of the gradient associated with 32P-labeled tamoxifen DNA adducts.[20,24] Nucleotides from [14C] toremifene-treated rats gave no radiocarbon peaks above background levels following HPLC analysis. Using this technique, binding of [14C] tamoxifen to extrahepatic organs, including DNA from reproductive and gastrointestinal tracts, has been demonstrated. The relevance of such low levels of binding to the assessment of carcinogenicity has yet to be established.
In view of the results from rat liver carcinogenicity studies and the detection of 32P-postlabeled DNA adducts, it was important to assess whether tamoxifen-DNA adducts were formed in human liver from patients treated with tamoxifen. Liver samples obtained from seven women receiving this drug were compared with liver taken from seven control individuals. In all but one of the treated women, tamoxifen and its N-desmethyltamoxifen metabolite could be detected in liver extracts. The total level of 32P-postlabeled DNA adducts seen in tamoxifen treated liver ranged between 18 and 80 adducts/108 nucleotides. The pattern of adducts was not the same as those seen in the tamoxifen treated rat liver. There was no significant difference in the level of DNA damage between tamoxifen and control groups. Although only a small number of subjects have so far been examined, those women who have been investigated, show no susceptibility to liver DNA damage caused by tamoxifen, in contrast to the response of rats.
Although previous investigations had demonstrated that tamoxifen is a rat liver carcinogen with genotoxic properties, the drug has consistently given a negative response in gene mutation assays in vitro. The introduction recently of the lambda/lacI transgenic rat has provided the means of studying the induction of treatment-induced gene mutations and their mutational spectra in different organs in these species. Tamoxifen caused a dose-related increase in the mutation frequency of the lacI gene in the livers of transgenic F344 rats (Big Blue) following 6 weeks oral dosing (20 mg/kg). Toremifene, in contrast, did not induce mutations or cause the formation of significant levels of DNA adducts (Figure 3).
Analysis of the spectrum of mutations at lacI showed that 59% of tamoxifen-induced mutations were point mutations compared with 80% from controls. Other types of mutation, deletions, insertions, and complex mutations, were also present: 41% in liver cells from tamoxifen-treated rats and 20% from controls. About 60% of the point mutations found in tamoxifen-treated rats were G:C > T:A transversions compared with 21% G:C > T:A in controls. The majority of the tamoxifen-induced G:C >T:A transversions were at CpG sites. These observations indicate that the spectrum of tamoxifen-induced mutations in the lacI gene is different from that occurring spontaneously. Using a dosing regime that has been shown to induce DNA adducts, hyperplasia and hepatocarcinogenesis, tamoxifen is a gene mutagen in the rat liver.
In earlier studies, Vancutsem looked at the mutations in exons 5 to 9 the p53 tumor suppressor gene in heptocellular carcinomas induced in rat liver following 12 months tamoxifen treatment. In 50% of the tumors, p53 mutations were found. These mutations were all specifically clustered in two sites, codons 231 (exon 6 to 7) and 294 (exon 8). Of the tumors investigated, nine contained A > G transitions, while four contained a C > T transition and one contained both mutations. The spectrum of mutations in the tumors that obviously occurred as a late event, after 12 months dosing with tamoxifen, clearly differs from the early changes seen at lacI following 6 weeks dosing. The p53 mutations may be important, however, as they could act as a fingerprint for tamoxifen induced DNA lesions.
Recently, the mutational spectrum of the p53 gene in 40 endometrial tumors in women indicate an excess of G:C > A:T transitions at non-CpG sites in 91% of the tumors. This differed substantially from the 19% of these mutations found in sporadic endometrial tumors. This finding, therefore, suggest, for the first time, that the tumors arising as a result of tamoxifen treatment in women may be arising from a genotoxic mechanism rather than by promoting endogenous DNA damage. These results are, however, preliminary, and because of their potential implications for an understanding of the mechanism of carcinogenicity of tamoxifen, more data are required before any conclusion is reached.
As far as liver carcinogenesis in rats is concerned, all of the evidence suggests that it is not tamoxifen itself, which is genotoxic, but a reactive metabolite of tamoxifen formed by the action of the cytochrome P450-dependent monooxygenases. The metabolism and metabolic activation of tamoxifen has been extensively studied in rodent and human liver microsomal systems [37-39]. The pattern of metabolites produced by the rat and human liver microsomes is very similar (Figure 4). There are, however, considerable quantitative differences with rates of certain metabolic pathways, for example, N-demethylation and N-oxidation, being faster in rat than in human microsomal systems. A study using a panel of 12 human liver microsomal preparations, characterized and immunoquantified for nine cytochrome P450 isoenzymes and using irreversible binding of active metabolites to protein as a surrogate for DNA binding, suggested CYP3A4 to be the major isoenzymic form involved in activation. This supports the earlier observation using a micronucleus assay in cell lines expressing cDNA for various cytochrome P450s, which also found CYP3A4 to be associated with the clastogenic effects of tamoxifen. In contrast to all of the studies in vivo, in vitro, toremifene is able to undergo NADPH-dependent metabolic activation catalyzed by either rat or human liver microsomal preparations. Microsomal systems are useful for demonstrating the potential for metabolic activation, but do not necessarily accurately reflect the extent to which this will occur in humans. The involvement of cytochrome P450 in the metabolic activation of tamoxifen at extrahepatic sites is unclear. Covalent binding of [14C] tamoxifen to microsomal preparations from human breast tumor tissue has been detected, but at levels seven times lower than in the liver, but activation in human endometrial microsomal preparation has not been reported.
Studies have shown that cytochrome P450 is not the only tamoxifen activating enzyme system. Peroxidases, including lactoperoxidase and prostaglandin synthase can both carry out N-demethylation and N-oxidative detoxication reactions and activate tamoxifen to DNA-binding species.[42,43] The major tamoxifen 32P-postlabeled DNA adducts formed when calf-thymus DNA was included in the reaction mixture were similar to the major adducts formed in rat liver. Activation of toremifene by peroxidases can also be detected. 3-Hydroxytamoxifen is completely inactive in all of the in vitro systems tested. In contrast, 4-hydroxytamoxifen can be metabolized by cytochrome P450s and peroxidases more rapidly than tamoxifen. In the case of peroxidases, this metabolism is accompanied by the formation of free radical intermediates. In rats or humans, there is little evidence for a role of peroxidases in metabolism or metabolic activation of tamoxifen with one possible exception. It was found by Hemminki et al that primary cultures of human lymphocytes incubated with tamoxifen were able to activate this drug to give low levels of DNA damage detected by 32P-postlabeling. Toremifene, too, gave positive results in this system. Lymphocytes, particularly neutrophils, are known to contain high levels of myeloperoxidase activity and this may account for the activation of both drugs. The concentration of tamoxifen used in these studies (up to 100 mg/mL) was very much higher than that seen in the plasma of women taking this drug therapeutically (0.3 mg/mL). In the latter individuals, no 32P-postlabeled adducts could be detected in lymphocyte DNA.
a-Hydroxytamoxifen (Figure 4) has been proposed as a reactive metabolite of tamoxifen. It was reported to be a major metabolite produced by rat hepatocytes and can be detected in the plasma and in the urine of patients on tamoxifen therapy. With the chemical identification of a synthetic tamoxifen-DNA adduct in which the alpha position of tamoxifen is linked covalently to the exocyclic amino group of deoxyguanosine, there is now persuasive evidence that a-hydroxytamoxifen, bioactivated as a sulphate ester, has a major role in the formation of the DNA damage in the livers of tamoxifen treated rats. In the case of toremifene, the chloroethyl function might be expected to greatly modify chemical reactivity if activation is via a-hydroxylation. For both tamoxifen and toremifene, an arene oxide intermediate has been proposed as an alternative route of activation. To account for the greater reactivity of 4-hydroxytamoxifen, at least in vitro, oxidation to the 4-hydroxytamoxifen quinone methide has also been suggested as an activation pathway. The involvement of other putative reactive species in the genotoxic and mutagenic effects of tamoxifen in humans (or rats) have yet to be established.
Since the causal relationship between the detection of DNA damage following the administration of tamoxifen and the subsequent development of tumors in rat liver has now largely been accepted, there was great interest to establish if similar activation mechanisms could occur in the reproductive tract of women taking this drug. In human endometrium, either no DNA damage  or extremely low levels of 32P-postlabeled adducts have been detected, orders of magnitude lower than the levels of DNA damage found in rat liver that lead to tumor formation. Based on the present evidence, these results suggest that endometrial tumors in women taking tamoxifen arise as a result of the estrogen agonist promoting effects, rather than as a result of genotoxic damage. A summary of the evidence for the genotoxicity, mutagenicity, and carcinogenicity of tamoxifen, comparing data from experimental rodents and humans is shown in Table 1.
Several antiestrogenic drugs are currently being developed for various clinical indications, including the adjuvant therapy for breast cancer, the treatment of osteoporosis and benign breast disease. By far the most clinically tested of these is tamoxifen, which has been widely used in all stages of breast cancer for approximately 20 years. Findings from clinical trials using tamoxifen are far from consistent, although the balance of evidence indicates a strong association between an increase in uterine tumors (three- to sixfold increase) and tamoxifen therapy. At present, the mechanism of endometrial carcinogenesis caused by this drug is not clear. The early onset of endometrial cancer after initiation of tamoxifen therapy suggests that this tumor is promoted by the partial agonist estrogenic property of tamoxifen. It is possible that other antiestrogenic drugs with partial agonist activity, such as toremifene and droloxifene, will produce similar effects on the uterus as tamoxifen. However, the antiestrogen raloxifene, has been shown to exhibit differential estrogenic agonist and antagonist effects in different tissues. This suggests that this drug will not, or is much less likely to, promote endometrial cancer and also suggests that it will be necessary to carry out clinical trials to establish whether other antiestrogenic drugs will promote uterine cancer.
At present, there are no alternative antiestrogenic drugs marketed for use in adjuvant therapy for breast cancer, and tamoxifen continues to provide considerable benefit in this setting. It is certainly unusual, however, to administer a genotoxic carcinogen to healthy humans and it remains doubtful if the prophylactic use of tamoxifen will prove acceptable.
The principle of using antiestrogens to protect against breast cancer is sound, and in the future, it seems likely that one of the nongenotoxic antiestrogens and/or a drug that is not associated with endometrial cancer will be used prophylactically. At present, it can be concluded that the risk/benefit ratio is favorable for the continued use of tamoxifen in adjuvant therapy. This will continue until alternative drugs have shown through extensive clinical trials to have equal efficacy and less toxicity.
1. Quinn M, Allen E: Changes in incidence of and mortality from breast cancer in England and Wales since introduction of screening. Br Med J 311:1391-1395, 1995.
2. Gaskell SP, Mcguire WL, Osborne CK, et al: Breast cancer mortality and diet in the United States. Cancer Res 39:3626-3637, 1979.
3. Fisher B, Dignam J, Bryant J, et al: Five versus more than five years of tamoxifen therapy for breast cancer patients with negative lymph nodes and estrogen receptor-positive tumors. J Natl Cancer Inst 88:1529-1542, 1996.
4. Anonymous: Systemic treatment of early breast cancer by hormonal, cytotoxic, or immune therapy. 133 randomised trials involving 31,000 recurrences and 24,000 deaths among 75,000 women. Lancet 339:1-15, 1992.
5. Powles TJ, Jones AL, Ashley SE, et al: The Royal Marsden Hospital pilot tamoxifen chemoprevention trial. Breast Cancer Res Treat 31:73-82, 1994.
6. Anonymous: Tamoxifen, in Some Pharmaceutical Drugs, pp 253-365. Lyon, France, IARC, 1996.
7. Killackey MA, Hakes TB, Pierce VK: Endometrial adenocarcinoma in breast cancer patients receiving antiestrogens. Cancer Treat Rep 69:237-238, 1985.
8. Fisher B, Costantino JP, Redmond CK, et al: Endometrial cancer in tamoxifen-treated breast cancer patients: Findings from the National Surgical Adjuvant Breast and Bowel Project (NSABP) B-14. J Natl Cancer Inst 86:527-537, 1994.
9. Rutqvist LE, Johansson H, Signomklao T, et al: Adjuvant tamoxifen therapy for early stage breast cancer and second primary malignancies. J Natl Cancer Inst 87:645-651, 1995.
10. Curtis RE, Boice JD Jr, Shriner DA, et al: Second cancers after adjuvant tamoxifen therapy for breast cancer. J Natl Cancer Inst 88:832-834, 1996.
11. Henderson BE, Ross RK, Pike MC: Hormonal chemoprevention of cancer in women. Science 259:633-638, 1993.
12. Novadex Adjuvant Trial Organisation (NATO): Controlled trial of tamoxfien as a single adjuvant agent in the management of early breast cancer. Br J Cancer 57:608-611, 1988.
13. Stewart HJ, (for the Scottish Cancer Trials Breast Group): The Scottish trial of adjuvant tamoxifen in node-negative breast cancer. Monogr Natl Cancer Inst 11:117-120, 1992.
14. Anonymous: Toremifene, in Some Pharmaceutical Drugs, pp 367-387. Lyon, France, IARC, 1996.
15. McDonnell DP, Dana SL, Hoener PA, et al: Cellular mechanisms which distinguish between hormone- and antihormone-activated estrogen receptor. Ann NY Acad Sci 761:121-137, 1995.
16. Kuiper GGJM, Enmark E, Pelto-Huikko M, et al: Cloning of a novel estrogen receptor expressed in rat prostate and ovary. Proc Natl Acad Sci U S A 93:5925-5930, 1996.
17. Colletta AA, Benson JR, Baum M: Alternative mechanisms of action of anti-oestrogens. Breast Cancer Res Treat 31:5-9, 1994.
18. Greaves P, Goonetilleke R, Nunn G, et al: Two-year carcinogenicity study of tamoxifen in Alderly Park Wistar-derived rats. Cancer Res 53:3919-3924, 1993.
19. Hard GC, Iatropoulos MJ, Jordan K, et al: Major differences in the hepatocarcinogenicity and DNA adduct forming ability between toremifene and tamoxifen in female Cr1: CD(BR) rats. Cancer Res 53:4534-4541, 1993.
20. Martin EA, Carthew P, White INH, et al: Investigation of the formation and accumulation of liver DNA adducts in mice chronically exposed to tamoxifen. Carcinogenesis 18:2209-2215, 1997.
21. White INH, De Matteis F, Davies A, et al: Genotoxic potential of tamoxifen and analogues in female Fischer F344/N rats, DBA/2 and C57Bl/6 mice and in human MCL-5 cells. Carcinogenesis 13:2197-2203, 1992.
22. Han X, Liehr JG: Induction of covalent DNA adducts in rodents by tamoxifen. Cancer Res 52:1360-1363, 1992.
23. Carthew P, Rich KJ, Martin EA, et al: DNA damage as assessed by 32P-postlabeling in three rat stains exposed to dietary tamoxifen: The relationship between cell proliferation and liver tumour formation. Carcinogenesis 16:1299-1304, 1995.
24. White INH, Martin EA, Vogel JS, et al: Comparison of the binding of [14C]radiolabelled tamoxifen or toremifene to rat DNA using accelerator mass spectrometry. Chem Biol Interact 106:149-160, 1997.
25. Williams GM, Iatropoulos MJ, Karlsson S: Initiating activity of the ant-estrogen tamoxifen, but not toremifene in rat liver. Carcinogenesis 18:2247-2253, 1997.
26. Carthew P, Edwards RE, Nolan BM. Depletion of hepatocyte nuclear estrogen receptor expression is associated with promotion of tamoxifen induced GST-P foci to tumours in rat liver. Carcinogenesis 18:1109-1112, 1997.
27. Dragan YP, Vaughan J, Jordan VC, et al: Comparison of the effects of tamoxifen and toremifene on liver and kidney tumor promotion in female rats. Carcinogenesis 16:2733-2741, 1995.
28. Carthew P, Martin EA, White INH, et al: Tamoxifen induces short-term cumulative DNA damage and liver tumors in rats: Promotion by phenobarbital. Cancer Res 55:544-547, 1995.
29. Carthew P, Nolan BM, Edwards RE, et al: The role of cell death and cell proliferation in the promotion of rat liver tumours by tamoxifen. Cancer Lett 106:163-169, 1996.
30. Newbold RR, Bullock B, Jefferson WN, et al: Developmental exposure of mice to tamoxifen (TAM) induced long term effects. Proc Am Soc Cancer Res 38:573, 1997.
31. Li D, Dragan Y, Jordan VC, et al: Effects of chronic administration of tamoxifen and toremifene on DNA adducts in rat liver, kidney, and uterus. Cancer Res 57:1438-1441, 1997.
32. Dahme E, Rattel B. Unlike tamoxifen, droloxifene produces no hepatic tumors in the rat. Onkologie 17:6-16, 1994.
33. Martin EA, Rich K, White INH, et al: 32P-Postlabelled DNA adducts in liver obtained from women treated with tamoxifen. Carcinogenesis 16:1651-1654, 1995.
34. Davies R, Oreffo VIC, Martin EA, et al: Tamoxifen causes gene mutations in the livers of lambda/lacI transgenic rats. Cancer Res 57:1288-1293, 1997.
35. Vancutsem PM, Lazarus P, Williams GM. Frequent and specific mutations of the rat p53 gene in hepatocarcinomas induced by tamoxifen. Cancer Res 54:3864-3867, 1994.
36. Welsh JA, Moberger B, Fornander T, et al: Tamoxifen has a characteristic mutation spectrum in human endometrial cancers. Proc Am Soc Cancer Res 38:450, 1997.
37. Lim CK, Yuan Z, Lamb JH, et al: A comparative study of tamoxifen metabolism in female rat, mouse and human liver microsomes. Carcinogenesis 15:589-593, 1994.
38. Mani C, Gelboin HV, Park SS, et al: Metabolism of the antimammary cancer antiestrogenic agent tamoxifen. 1. Cytochrome P-450-catalyzed N-demethylation and 4-hydroxylation. Drug Metab Dispos 21:645-656, 1993.
39. Kupfer D, Dehal SS: Tamoxifen metabolism by microsomal cytochrome P450 and flavin-containing monooxygenase. Methods Enzymol 272:152-163, 1996.
40. Styles JA, Davies A, Lim CK, et al: Genotoxicity of tamoxifen, tamoxifen epoxide and toremifene in human lymphoblastoid cells containing human cytochrome P450s. Carcinogenesis 15:5-9, 1994.
41. White INH, De Matteis F, Gibbs AH, et al: Species differences in the covalent binding of [14C]tamoxifen to liver microsomes and the forms of cytochrome P450 involved. Biochem Pharmacol 49:1035-1042, 1995.
42. Davies AM, Martin EA, Jones RM, et al: Peroxidase activation of tamoxifen and toremifene to cause DNA damage and covalently bound protein adducts. Carcinogenesis 16:539-545, 1995.
43. Davies AM, Malone ME, Martin EA, et al: Peroxidase activation of 4-hydroxytamoxifen to free radicals detected by EPR spectroscopy. Free Radic Biol Med 22:423-431, 1997.
44. Hemminki K, Widlak P, Hou SM: DNA adducts caused by tamoxifen and toremifene in human microsomal system and lymphocytes in vitro. Carcinogenesis 16:1661-1664, 1995.
45. Phillips DH, Hewer A, Grover PL, et al: Tamoxifen does not form detectable DNA adducts in white blood cells of breast cancer patients. Carcinogenesis 17:1149-1152, 1996.
46. Phillips DH, Carmichael PL, Hewer A, et al: a-Hydroxytamoxifen, a metabolite of tamoxifen with exceptionally high DNA binding activity in rat hepatocytes. Cancer Res 54:5518-5522, 1995.
47. Poon GK, Walter B, Lonning PE, et al: Identification of tamoxifen metabolites in human HEP G2 cell line, human liver homogenate, and patients on long-term therapy for breast cancer. Drug Metab Dispos 23:377-382, 1995.
48. Osborne MR, Hewer A, Hardcastle IR, et al: Identification of the major tamoxifen-deoxyguanosine adduct formed in the liver DNA of rats treated with tamoxifen Identification of the major tamoxifen-deoxyguanosine adduct formed in the liver DNA of rats treated with tamoxifen. Cancer Res 56:66-71, 1996.
49. Moorthy B, Sriram P, Pathak DN, et al: Tamoxifen metabolic activation: Comparison of DNA adducts formed by microsomal and chemical activation of tamoxifen and 4-hydroxytamoxifen with DNA adducts formed in vivo. Cancer Res 56:53-57, 1996.
50. Hemminki K, Rajaniemi H, Lindahl B, et al: Tamoxifen-induced DNA adducts in endometrial samples from breast cancer patients. Cancer Res 56:4374-4377, 1996.