Tamoxifen Treatment for Breast Cancer: Concept to Gold Standard

ABSTRACT: Tamoxifen is currently the endocrine treatment of choice for all stages of breast cancer and is the gold standard for antiestrogen treatment. Over the last 25 years, the drug has revolutionized breast cancer therapy. The extension of the use of this agent has occurred because of open dialogue between the laboratory and the clinic, in which laboratory findings led to extension of clinical use. Tamoxifen was originally discovered as part of a contraceptive research program at ICI Pharmaceuticals (now Zeneca). On the basis of the estrogen dependence of many breast cancers, tamoxifen, a potent antiestrogen, was predicted to have anticancer activity. Laboratory and animal studies demonstrated efficacy in breast cancer and an ability to block binding of estradiol to the estrogen receptor of human breast cancer. Preclinical studies showed the benefit of long-term vs short-term tamoxifen treatment, a finding duplicated in the clinic. [ONCOLOGY 11(Suppl 1):7-13, 1997]

Introduction

Tamoxifen (Nolvadex) was originally used as a palliative treatment ofadvanced breast cancer. Laboratory studies demonstrating that tamoxifencould prevent progression of cancer led to studies as adjuvant treatment.The results of these studies demonstrated that tamoxifen could producesurvival advantages in appropriate patients. Tamoxifen was subsequentlyshown to prevent contralateral breast cancer with a low incidence of sideeffects. The efficacy of tamoxifen in breast cancer has prompted investigationsof it as a preventive agent in women who are at high risk of developingbreast cancer.

As with any drug, tamoxifen has benefits and risks associated with itsuse. In addition to its benefits as an effective cancer treatment, tamoxifenhas been shown to have additional benefits in relation to bone and lipids.The risk of development of endometrial cancer in tamoxifen-treated patientshas been a concern. However, clinical studies have shown that the absoluterisk is low, and may be due to methodologic biases in the studies and tothe increased risk of endometrial cancer documented in untreated breastcancer patients.

The broad laboratory and clinical database on tamoxifen compiled overthe last 25 years allows for a reasoned evaluation of the benefits andrisks of tamoxifen treatment and indicates that the benefits of tamoxifenfor breast cancer patients far outweigh the risks.

Twenty-five years ago, tamoxifen was only an experimental antiestrogenwith the code ICI 46,474. Since the approval of tamoxifen for the treatmentof advanced breast cancer in postmenopausal women, first in the UnitedKingdom in 1973 and then in the United States in 1978, the drug has revolutionizedthe approach to breast cancer therapy.

Today, tamoxifen is the endocrine treatment of choice for all stagesof breast cancer and, remarkably, it is the only single agent proven toprovide survival advantages in patients with both node-positive and node-negativedisease.[1,2]

The success and efficacy of the treatment is demonstrated by the factthat the World Health Organization has declared that tamoxifen is an essentialdrug for breast cancer therapy. Over the last quarter century, the shiftof antiestrogens from the bench to the bedside has dramatically alteredclinical practice as well as demonstrated the value of these agents.

Tamoxifen was originally discovered in the laboratories of ICI Pharmaceuticalsin Cheshire, England. Ironically, the scientists (Drs. Harper, Richardson,and Walpole, Figure 1) were primarilyinterested in new drugs to regulate fertility and they set up their screeningtests accordingly. Compound ICI 46,474 was found to be a potent antiestrogenin the rat and had high potency as an antifertility agent.[3,4]

The head of the project, the late Dr. Arthur Walpole, was also passionatelyinterested in developing drugs for cancer treatment, and earlier in hiscareer he had focused on therapies that could be useful to treat breastcancer.[5] He was also aware that oophorectomy and adrenalectomy (antiestrogenictreatment modalities) were beneficial treatments for advanced breast cancer,and concluded that it would not be unreasonable to develop an antiestrogenicdrug for breast cancer treatment.

Tamoxifen was evaluated in preliminary clinical studies for a varietyof different indications. A small phase II study against historical controlsdemonstrated activity and virtually no side effects for the treatment ofadvanced breast cancer in postmenopausal women.[6] However, it was onething to demonstrate clinical activity and entirely another thing to gaingeneral acceptance by the medical community as a safe treatment and asthe endocrine treatment of choice for all stages of breast cancer.

This process has taken 20 years and has moved forward because of opencommunication between the laboratory and the clinic. Ideas from the laboratoryhave pointed the way to clinical investigations that in turn have extendedthe utility of tamoxifen (Figure 2). Todaythere are eight million woman-years of experience with tamoxifen, and concernsthat have been raised have been addressed and placed into a reasonableperspective of risks and benefits.

The aim of this historical overview is to provide a framework to understandthe development of tamoxifen. A number of important laboratory observationshave guided the clinical evaluation of antiestrogens and have providedthe basis for the strategies used in the treatment of breast cancer andnow the clinical testing of tamoxifen as a chemopreventive agent.

ICI 46,474 to Tamoxifen

The first systematic laboratory study of tamoxifen as an antitumor agentwas conducted at the Worcester Foundation for Experimental Biology (nowthe Worcester Foundation for Biomedical Research) in Shrewsbury, Massachusetts(1972-1974). At a time when tamoxifen was being evaluated as a treatmentfor advanced breast cancer, the laboratory studies demonstrated that tamoxifeninhibits the growth of carcinogen-induced rat mammary carcinomas, preferentiallycontrols growth of estrogen receptor (ER)-positive mammary tumors, andblocks the binding of [3H] estradiol to the ER of human breast tumors.Overall, the results of the laboratory studies were used to encourage humanstudies by the Eastern Cooperative Oncology Group (ECOG), and an overview[7]of laboratory data introduced tamoxifen to the National Surgical AdjuvantBreast and Bowel Project (NSABP) in 1976.

Perhaps most important, studies were published that demonstrated tamoxifencould prevent the chemical induction of rat mammary carcinomas.[8,9] Thesedata—with the subsequent findings that tamoxifen has a low incidence ofside effects, high patient acceptability, and the unique ability to preventthe development of contralateral breast cancer—laid the foundation forthe current prevention trials.[10]

Adjuvant Use

While tamoxifen was initially approved for the endocrine treatment ofadvanced disease, the primary conceptual breakthrough was the idea of usingadjuvant therapy to destroy micrometastases throughout a patient's body.Chemotherapy was naturally the first choice because of its cytotoxicity.

The goal was to cure patients, so there was an initial reluctance touse an endocrine approach because these agents were considered only palliativetherapy for advanced disease; complete remission did not seem possiblewith endocrine therapy alone. Nevertheless, several adjuvant trials wereinitiated with one year duration of tamoxifen administration.[11-14] Short-termtherapy was chosen because of the expectation of drug resistance.

We first addressed the problem of the duration of adjuvant tamoxifentherapy in the laboratory by using a carcinogen-induced rat mammary tumormodel, the only model available in the 1970s.[15-18] We reasoned that ifthe carcinogen was administered to susceptible strains of rats and thenallowed to be promoted by ovarian hormones, the mammary glands would havemicrofoci of mammary tumor cells that would be similar to micrometastases.We asked the question of whether short-term tamoxifen could inhibit theappearance of tumors. The experiment (Figure3) demonstrated that although short-term therapy reduced the numberof tumors that developed, most animals developed at least one tumor. Incontrast, continuous therapy with a clinically relevant dose of tamoxifenmaintained the majority of animals in a tumor-free state.

These data, and the concept that long-term adjuvant tamoxifen therapyshould be evaluated in clinical trials, were first presented at a symposiumat Kings College, Cambridge in September 1977. [15] In 1977, a preliminaryclinical evaluation of long-term adjuvant tamoxifen was conducted at theUniversity of Wisconsin.

The results,[22,23] which showed that tamoxifen was safe and potentiallyeffective, were used to establish ECOG protocols of long-term vs short-termtamoxifen with chemotherapy. However, with the routine use of long-termantiestrogen therapy in node-negative patients, of whom the majority couldbe cured by surgery alone, the concern in the mid-1980s became the potentialphysiological harm that could result from tamoxifen in postmenopausal women.

Professor Michael Baum and Dr. Helen Stewart, present at the CambridgeSymposium, later compared the efficacy of two years and five years of adjuvanttamoxifen therapy, respectively (Figure 4).In the 1980s, both clinical trial groups demonstrated for the first timethat long-term adjuvant tamoxifen could produce a survival advantage inpostmenopausal women with node-positive breast cancer.[19-20] It has recentlybeen shown that five years is superior to two years of therapy.[21]

Secondary Benefits and Risks

Estrogen is essential to maintain bone mineral density and to protectwomen from coronary heart disease. Clearly, a treatment strategy of long-termantiestrogen therapy in women with node-negative disease could lead topremature osteoporosis and an increased risk of myocardial infarction.However, tamoxifen is a partial estrogen agonist and has mixed estrogenicand antiestrogenic properties.[24]

Studies in animals first demonstrated that tamoxifen could maintainbone mineral density.[25] This observation was subsequently demonstratedin postmenopausal women in the Wisconsin Tamoxifen Study[26] and laterconfirmed by other groups throughout the world.[27]

Similarly, tamoxifen was noted to lower low-density lipoprotein (LDL)cholesterol.[28] This estrogen-like effect may be the reason why tamoxifen-treatedpatients have fewer hospital- izations for any cardiac condition and havea reduced incidence of fatal myocardial infarctions if they are treatedwith at least five years of tamoxifen.[29-31]

Indeed, the risk of a myocardial infarction increases in women who havestopped tamoxifen when compared with current users.[31] Thus, the overallactions of tamoxifen appear to support a woman's physiology, like a hormonereplacement therapy, rather than place her at further risk for postmenopausalcomplications.

The target site specificity of tamoxifen did not focus on the humanendometrium until the mid-1980s, when questions regarding the potentialassociation between tamoxifen and endometrial cancer arose. Perhaps noother topic during the last eight years has been more controversial andmore misinterpreted.

In 1984, Dr. Satyaswaroop and colleagues[32] demonstrated that a humanendometrial carcinoma transplanted into athymic mice would grow in responseto estrogen and have a partial growth response to the estrogen-like propertiesof tamoxifen. Since women with a breast cancer diagnosis are at increasedrisk for endometrial cancer just because they had breast cancer, we hadconsidered that long-term tamoxifen might control only the breast cancerbut not any preexisting endometrial cancer. We illustrated this point ina 1988 paper[33] that showed the target site-specific effect of tamoxifenin blocking estrogen-stimulated breast tumor growth while stimulating growthof a cotransplanted endometrial tumor.

At that time, we recommended that patients on long-term tamoxifen therapyshould be evaluated for preexisting endometrial carcinoma; however, wefailed to point out that women not on tamoxifen should also be evaluated.This regrettably has led to sampling biases in studies on tamoxifen andendometrial cancer because tamoxifen-treated patients are more likely toexperience symptoms and are more likely to be repeatedly evaluated. Wealready know that in the uterus, occult disease is four times as commonas overt disease,[34] so increased sampling would obviously lead to anincrease in detection.

The Stockholm study[35] was the first to demonstrate the target site-specificeffect of tamoxifen in patients: tamoxifen reduces contralateral breastcancer but increases the detection of endometrial cancer. Their findingsinitiated an intense clinical evaluation of tamoxifen-treated patients,but initial estimates of a sixfold greater risk are now known to be toohigh.

Few Endometrial Cancers

During the last five years there has been intense interest in this topic;however, remarkably few endometrial carcinomas have been reported.[36-38]To date, there are fewer than 400 cases in the world literature, and themajority are observed, as one would expect, in postmenopausal women.

Additionally, an initial concern was that endometrial cancer in tamoxifen-treatedpatients would present as a high-grade tumor with a poor prognosis, butsurveys of the clinical literature and analysis of clinical trials do notsupport this assertion.[36-38] Tamoxifen does not form DNA adducts in humanuterine tissue[39] and there is, therefore, no evidence for causation.Carcinogenesis requires about 10 years, and most reports of endometrialcarcinoma in patients are noted within five years of exposure to tamoxifen.

It has taken many years to sort out the misstatements about the risksassociated with tamoxifen exposure, but the American College of Obstetriciansand Gynecologists has not recommended intense monitoring or routine biopsies.Patients should be investigated only if symptoms are present. Indeed, recentscrutiny by an agency of the World Health Organization has concluded thatalthough there is an increased risk of developing endometrial cancer duringexposure to tamoxifen, no woman should be discouraged from taking the drugbecause of the slight risk. The benefits of tamoxifen treatment far outweighany of the risks.

The discovery that endometrial carcinoma cell clones could be stimulatedto growth by tamoxifen opened the door to explore the possibility thatbreast tumors might also become dependent upon tamoxifen, a form of drugresistance. In the laboratory, MCF-7 breast cancer cells are estrogen-stimulatedto produce solid tumors in athymic mice. Tamoxifen, however, inhibits estrogen-stimulatedgrowth as long as the drug is given.[40] Nevertheless, eventually clonesof cells emerge that are ER-positive but which grow in response to eitherestrogen or tamoxifen.[41]

These data suggest that tamoxifen-dependent breast tumor growth mightultimately occur if adjuvant tamoxifen treatment is given indefinitely.

The duration of tamoxifen treatment is the subject of some controversy,but in the case of patients with risk factors for osteoporosis and coronaryheart disease, an individualized approach might be appropriate.[25] Currenttrials of node-negative, ER-positive disease demonstrate that 10 yearsof adjuvant therapy is no more beneficial in prolonging overall disease-freesurvival than five years of treatment. Nevertheless, the small number ofevents in the trial presents a degree of uncertainty that mandates furtherstudy.

In the UK, 20,000 node-positive and node-negative breast cancer patientsare being registered in either the ATTOM (Adjuvant Tamoxifen TreatmentOffer More) trial in Birmingham or the worldwide ATLAS (Adjuvant TamoxifenLonger Against Shorter) trial headquartered at Oxford. The basic designof the trials is either to stop tamoxifen at whatever duration a patienthas reached or continue for another five years. It is believed that smallbenefits will be detected by the large sample size.

What to Do After Tamoxifen

There is now clinical evidence that tamoxifen-dependent breast cancergrowth is possible. A number of anecdotal reports demonstrate a withdrawalresponse that occurs many weeks after tamoxifen is stopped.[42-44] Thistime course is almost certainly dependent on the slow (six week) clearanceof tamoxifen. Based on the laboratory studies,[39] we would predict thatdisease that was once inhibited by tamoxifen has been subjected to clonalselection, and the tumor is now weakly stimulated by tamoxifen. We wouldalso predict that the tumor will utilize a woman's own estrogen for continuedgrowth once tamoxifen is stopped. A strategy of either blocking estrogensynthesis or blocking the estrogen receptors with a novel agent would thereforebe beneficial to obtain a second endocrine response.

The specific aromatase inhibitor anastrazole (Arimidex) is availableto treat postmenopausal patients who have recently failed tamoxifen therapy.[45]Additionally, a range of new antiestrogens is being developed for use asalternatives to or following tamoxifen treatment.[46] Toremifene (or chloro-tamoxifen)has pharmacology similar to that of tamoxifen, but toremifene is threefoldless potent than tamoxifen in both preclinical[47] and clinical[48] studies.Toremifene is completely cross-resistant with tamoxifen; thus, it wouldnot be a useful agent to try after exposure to tamoxifen. Additionally,there is every indication in work that we are doing that toremifene willencourage the growth of endometrial tumors in the laboratory, so an increasein the detection of endometrial cancers during toremifene therapy is tobe expected. In contrast, new pure antiestrogens[49] will inhibit the growthof tamoxifen-stimulated tumors[50] and have shown promise in recently publishedclinical studies.[51]

Summary

Tamoxifen has revolutionized the treatment of breast cancer and hasbecome the gold standard for endocrine therapy. In the last 25 years, thevalue of tamoxifen has been proven worldwide through randomized clinicaltrials for all stages of breast cancer. The drug has been shown to be safeand effective through a close cooperation between laboratory and clinicalresearch. No other endocrine treatment for breast cancer has such extensivedocumentation. The concepts of long-term adjuvant therapy, benefits tobone, potential drug resistance, and second-line endocrine therapies haveall translated accurately from the bench to the bedside. Concerns aboutthe detection of pre-existing endometrial cancer, first illustrated inthe laboratory, have also been appropriately evaluated in patients. Theincrease in the incidence of endometrial cancer is modest (two per 1,000women per year) compared with the nontreated breast cancer patients (oneper 1,000 women per year), so no special monitoring procedures are neededoutside routine gynecological check-ups and rapid follow-up for evaluationof symptoms, such as nonmenstrual bleeding.

Finally, the concept of breast cancer prevention with tamoxifen is alaboratory fact that is being tested in North America and Europe to assessthe risks and benefits of such treatment to healthy women in double-blindrandomized clinical trials.