Breast cancer prevention with tamoxifen - continuing education program

Val R. Adams, Pharm.D., Assistant Professor, University of Kentucky Program #401-000-99-005-H0l

Goal

Breast cancer is the most common cancer in women, with an estimated 178,700 new cases diagnosed in 1998. During her lifetime a woman has a 1-in-8 chance of developing breast cancer, which is associated with a high mortality rate. The American Cancer Society estimates that 43,500 women died of breast cancer in 1998, making it the second leading cause of cancer-related mortality, slightly behind lung cancer. [1] The risk of developing breast cancer significantly increases with age, positive family history, early onset of menarche, nuliparity and number of breast biopsies, especially if atypical hyperplasia is found. [1-3] A recent study shows that tamoxifen reduces the incidence of breast cancer by half in a population of high-risk women. [4] The goal is to identify women who may benefit from tamoxifen preventive therapy. Additionally, knowing how to monitor and manage tamoxifen-induced toxicity is crucial to optimize patient outcomes.

Objectives

After completing this CE, the participant will be able to:

1. Identify factors used in the Gail model to estimate the 5-year and lifetime risk an individual woman has of developing breast cancer.

2. List the rare, but potentially life threatening risks associated with tamoxifen therapy and how to appropriately monitor for these risks.

3. Identify non-life threatening adverse effects associated with tamoxifen and be able to recommend a plan to manage them

Introduction

Breast cancer incidence and mortality

Approximately 178,700 women were diagnosed with breast cancer in 1998, which makes it the most common cancer in females. It is estimated that 1-out-of-every-8 women will develop breast cancer in their lifetimes, with the majority of new cases diagnosed in patients over the age of 60. It is the second-leading cause of cancer-related mortality, slightly behind lung cancer, with an estimated 43,500 deaths in 1998 [1].

Risk factors associated with breast cancer

Numerous studies have evaluated factors associated with breast cancer risk (see Table 1 on page CP29), [1-3] the single largest being female gender. In 1999, the American Cancer Society estimates that 175,000 women will be diagnosed, compared to 1,300 men. [1] Due to this disparity, nearly all breast cancer research is conducted in women. Increasing age is another important risk factor for developing breast cancer. There is less than one diagnosis per 100,000 women younger than 25 years of age, while approximately 60 women per 100,000 will be diagnosed between the ages of 40 and 44, and 225 women per 100,000 will be diagnosed between the ages of 65 and 69. [2] A personal history of invasive breast cancer or lobular carcinoma in situ (non-invasive breast cancer) is also a major risk factor for developing a second primary breast cancer. Additional risk factors include a family history of breast cancer in a mother and/or sister, proliferative benign breast disease (particularly atypical hyperplasia), early age at menarche and older age at time of first live birth. [3]

Predicting breast cancer risk

Identifying risk factors associated with breast cancer is only useful if it can be used to accurately counsel patients about individual risk and potential interventions. Predicting an individual woman's risk of developing breast cancer has been difficult, with a plethora of conflicting data that come from trials varying in strength and study design. Additional difficulties in predicting individual risk include the lack of studies considering the interaction between risk factors.

Dr. Gail and colleagues [5,6] at the National Cancer Institute developed what is considered the best model for predicting breast cancer risk. They developed the model by analyzing data from 284,780 women in the Breast Cancer Detection Demonstration Project. Using logistic regression analysis, they found they could accurately predict an individual's 5-year and lifetime risk of breast cancer by considering subject age, age of menarche, number of breast biopsies, age of first live birth, number of first-degree relatives with breast cancer and pathology findings of breast biopsies. [5,6] They made the predictive model available to clinicians in 1996 in the Journal of Clinical Oncology. [7]

The model required a summation of risks, which was applied to one of a series of graphs. In 1999, the NCI, in cooperation with Zeneca Pharmaceuticals, made software based on this model available to healthcare professionals. Although the Gail model for predicting risk is the gold standard, it still has limitations. The database used to develop the model consisted of almost all Caucasian women. How well it predicts risks of minorities is unknown. The model also fails to account for the age at which a primary relative was diagnosed with breast cancer, which has been shown to be important. [5] These limitations are important to keep in mind when counseling subjects.

In recent years, significant advancements have been made with regard to interventions available for women at high risk of developing breast cancer. For several years, intense screening has been recommended for high-risk women. This screening, which consists of an annual mammogram beginning as early as age 35, allows a tumor to be detected at an early stage when highly curable. [3] In addition to screening, the National Surgical Adjuvant Breast and Bowel Project P-1 study [4] published in 1998 showed that tamoxifen could reduce the risk of developing breast cancer. The trial showed that women at high risk of developing breast cancer (greater than 1.66 percent over five years) given tamoxifen 20 mg daily for five years had a 49 percent reduction of breast cancer incidence. The Food and Drug Administration subsequently approved tamoxifen to reduce the risk of breast cancer.

Breast cancer prevention with tamoxifen

Background

In 1936, Lacassagne [8] predicted that blocking estrogen in the breast could prevent breast cancer. Unfortunately, no anti-estrogens drugs were available at that time to test his theory. In the early 1970s, ICI Pharmaceuticals began clinically testing tamoxifen, an anti-estrogen, to treat breast cancer. Based on the results of those tests, it was approved for use in Britain in 1973 and the United States in 1977. Initially, it was used to treat women with advanced disease, but only produced modest results. [9] Discovery of the estrogen receptor and methods to test for its presence on tumor cells led to better selection of patients likely to respond to tamoxifen, so response rates improved. During the late 1970s and early 1980s, investigators began testing tamoxifen as an adjuvant agent (therapy after curative surgery to reduce the number of women who relapse).

Researchers also began to recognize that tamoxifen had estrogen-like effects in other tissues. [9] They found that it increased the growth of endometrial cancer cells, decreased LDL cholesterol, increased bone density and increased the risk of thrombosis. These findings led to concerns about the benefits and risks associated with a longer duration of tamoxifen therapy. [9] To answer these concerns, a meta-analysis of adjuvant studies was conducted in 1998. [10] The analysis clearly demonstrated that tamoxifen is more effective when given for five years compared to one or two years. The studies also showed that five years of therapy is just as effective as 10 years. Adverse effects appeared minimal, but continued to accrue while patients received tamoxifen. Therefore, five years of therapy has been deemed the appropriate duration of therapy. This analysis also showed that it is effective in both premenopausal and postmenopausal women with estrogen receptor-positive tumors. [10] These overall analyses have als o revealed that adjuvant tamoxifen decreases the incidence of a new primary breast cancer in the contralateral breast by 47 percent. [10] The use of tamoxifen to prevent breast cancer has also been validated in a number of murine models. [9] The above facts combined with tamoxifen's great safety record led to several chemoprevention trials.

Clinical trials

Three large prospective randomized placebo-controlled trials evaluating the efficacy of tamoxifen to prevent breast cancer have now been reported. Dr. Prowles and colleagues [11] at the Royal Marsden Hospital in England established a study in 1986 to test the ability of tamoxifen to prevent breast cancer. In 1992, the NSABP group, with funding from the NCI, began enrolling patients in the United States and Canada to test the efficacy of tamoxifen to prevent breast cancer (the NSABP P-1 trial). [4] At the same time, Italian researchers undertook a very similar study to evaluate the ability of tamoxifen to prevent breast cancer in hysterectomized women. [12] All three studies were prospective randomized placebo-controlled trials that utilized tamoxifen at a dose of 20 mg daily.

Efficacy

The NSABP P-1 trial [4] was the largest tamoxifen breast cancer prevention study. Enrollment began in 1992 and was stopped after an interim analysis in 1997 due to a significant difference in breast cancer incidence between groups. At the time of trial termination, 13,388 high-risk women had been randomized to receive tamoxifen 20 mg daily or placebo. Subjects were defined as high-risk ([greater than or equal to]1.66 percent probability) of developing breast cancer during the next five years using the Gail model. This included women over the age of 60 with or without known risk factors and women aged 35 to 59 with risk factors placing them at risk greater or equal to a 60 year old (see Table 2 on page CP30). The patient characteristics between the groups were very similar (see Table 3 on page CP30). The study was designed to administer tamoxifen for five years; the mean and median follow-up time of subjects was 47.7 and 54.6 months. Approximately 37 percent of patients completed five years of therapy.

The primary outcome measure of this study was reduction in breast cancer occurrence. The results of the trial indicated that the risk of developing invasive breast cancer was decreased by 49 percent, 175 cases in the placebo group vs. 89 cases in the tamoxifen group (decrease of 6.76 to 3.43 cases per 1,000 women). The decrease was seen throughout the study and in all predicted risk groups (Figure 1). There was also a reduction in the cases of ductal carcinoma in situ (non-invasive breast cancer), 69 cases in the placebo group vs. 35 cases in the tamoxifen group. The findings were consistent regardless of individual risk factors such as age, breast disease and family history. Subset analysis revealed that women with a personal history of atypical hyperplasia found with breast biopsy were the most likely to benefit from tamoxifen (a reduction from 10.11 to 1.43 cases per 1,000 women per year). When breast cancer did occur, the tumors did not differ in size or nodal status. However, when analyzing estrogen rec eptor status between groups, tamoxifen only reduced the number of ER+ tumors.

The study's secondary outcome measures included analysis of coronary disease and osteoporosis-related fractures. Since tamoxifen is known to lower LDL cholesterol and increase bone density in postmenopausal women, a positive outcome in the tamoxifen group was anticipated. Myocardial infarction occurred in 28 patients receiving placebo and 31 patients receiving tamoxifen (not significant). No difference between severe angina or acute ischemia syndrome was seen between groups either. In regard to fracture rates, there was a 45 percent reduction in hip fractures (22 vs. 12), 39 percent reduction in Colles fractures (23 vs. 14) and a 26 percent reduction in spine fractures (31 vs. 23) between placebo and tamoxifen groups, respectively. These differences trended toward, but did not reach, statistical significance.

Prowles and colleagues [11] studied tamoxifen in 2,484 women, who were enrolled from the Royal Marsden Hospital. The women all had a family history of breast cancer and ranged in age from 30 to 70 years. They were randomized to receive tamoxifen 20 mg daily or placebo for a total of eight years. The trial began as a pilot project to evaluate problems in subject accrual, toxicity and safety for future large multicenter studies. [13] When it became evident that recruiting subjects was not a problem, the study continued to enroll women with the primary endpoint being prevention of breast cancer with tamoxifen. [11]

The women were followed for a mean of 70 months with approximately 70 percent remaining on the medication for five years. The incidence of breast cancer (32 cases in the placebo group vs. 30 in the tamoxifen group) did not differ between groups. Other study outcomes included a significant decrease in serum cholesterol, particularly LDL (overall 10 percent reduction). It reached a nadir at three months and remained at that level throughout the remaining five years. They also found an increased bone density in postmenopausal women. These outcomes are confounded by the fact that 42 percent of patients received hormone replacement therapy during the trial.

The Italian study [12] randomized 5,408 women aged 35 to 70 to receive tamoxifen or placebo for five years. Women were required to have had a hysterectomy to meet the inclusion criteria. These criteria were designed to remove concerns about tamoxifen-included endometrial cancer. The median length of follow-up was 46 months. Breast cancer occurrence was not different between groups (22 cases in the placebo group vs. 19 in the tamoxifen group). These results were based on initial randomization regardless of compliance or withdrawal from the study. Of the 5,408 subjects randomized, 1,422 withdrew from the study, and only 149 completed five years of treatment.

Adverse effects

Over the last 22 years, more than a million women worldwide have been treated with tamoxifen, which provides an extensive database regarding adverse effects. [14] Tamoxifen is typically well tolerated, and side effects are usually mild. Hot flashes and weight gain are most commonly reported and occur in 10 percent to 20 percent of patients. Less frequently, tamoxifen has been reported to cause headache, mild nausea with or without vomiting, skin rash and dryness, a change in menstrual period and vaginal dryness and itching. Rare hut serious adverse effects reported with tamoxifen include confusion, hepatotoxicity, ocular toxicity, pulmonary embolism, venous thrombosis, endometrial hyperplasia and endometrial cancer. Tamoxifen is also a known teratogen, which can be problematic since it increases fertility. [15]

The NSABP P-1 trial [4] confirmed the association of tamoxifen and endometrial cancer. There were 36 cases of endometrial cancer in the tamoxifen group vs. 15 in the placebo group. The increase was not seen in women under the age of 50, but it did increase the relative risk by fourfold in women aged 50 or greater (absolute annual risk increased from 0.08 percent to 0.31 percent). All of the endometrial cancers were diagnosed at a localized stage, except for one woman in the placebo group who presented with an advanced tumor. There were no deaths in the tamoxifen group from endometrial cancer. The Royal Marsden study [11] also reported a higher incidence of endometrial cancer in the tamoxifen group (four cases) than the placebo group (one case). This study did not find a statistically significant difference between these rates. In the Italian study, [12] all subjects had undergone a hysterectomy, therefore they reported no endometrial cancers.

Tamoxifen has also been associated with an increased risk of vascular events, such as stroke, transient ischemic attack, pulmonary embolism and deep vein thrombosis. In all three studies, [4,11,12] women less than 50 years of age had an equal number of vascular events. In the NSABP trial, [4] women 50 or older were more likely to experience a stroke (20 vs. 35) or deep vein thrombosis (14 vs. 24) if they received tamoxifen; these differences were not statistically significant. This study found that subjects receiving placebo had a higher rate of transient ischemic attacks (21 vs. 16), which was not statistically different. The older age group in the NSABP trial did have a significantly higher number of pulmonary embolisms, five in the placebo group vs. 16 in the tamoxifen group. The relative risk was increased by three and the absolute annual risk increased from 0.03 percent to 0.1 percent. There were three deaths in the tamoxifen group from pulmonary embolism, but all three had underlying comorbid condition s. The Royal Marsden and Italian studies [11,12] also reported slightly higher numbers of deep vein thrombosis and pulmonary embolism with tamoxifen use, but they did not reach statistical significance.

In the NSABP trial, [4] cataracts occurred more often in the tamoxifen-treated group. They occurred more frequently in older patients, with an annual incidence in the placebo group of 2.17 percent and 2.48 percent in the tamoxifen group (relative risk 1.14). Women in the tamoxifen group were also more likely to undergo cataract surgery. The Royal Marsden and Italian studies did not report cataract formation. [11,12]

A quality-of-life survey was also administered to all NSABP participants. [4] The survey quantified the incidence and severity of hot flashes, vaginal discharge, irregular menses, fluid retention, nausea, skin changes, diarrhea, weight gain or loss and depression. The only differences between groups were hot flashes and vaginal discharge. Hot flashes rated as "quite a bit or extremely bothersome" were reported by 45.7 percent of the tamoxifen group as compared to 28.7 percent of the placebo group. The proportion of patients reporting vaginal discharge rated as "moderately bothersome or worse" was 29 percent in the tamoxifen group compared to 13 percent in the placebo group. The Royal Marsden study'1 also found a significant increase in hot flashes and gynecologic abnormalities in the tamoxifen group. Event rates of hot flashes or gynecologic abnormalities were not reported in the Italian study. [12]

Summary of clinical trials

Since only one of three trials found a reduction in breast cancer incidence with tamoxifen, one could question the utility of using tamoxifen to prevent breast cancer. However, there are critical differences between the trials which likely account for the different results. The major differences between these trials were the number of women enrolled, subject risk of breast cancer, percent of young patients (less than 50 years), hormone replacement therapy and withdrawal from the study (see Table 4 on page CP31). The Italian study recruited less than half the number of women than were recruited in the NSABP trial. [4,12] The women in the Italian study also had a lower risk of developing breast cancer based on the placebo group incidence rate. The Royal Marsden study had one-fifth as many patients and allowed women to take hormone replacement therapy. Their patient population was also much younger on average than that in the NSABP trial. [4,11] Overall, the NSABP trial was the largest, enrolled women at higher risk, and did not allow hormone replacement therapy with estrogen. These factors make its results the most valid. After critical review of the three trials, a conclusion that tamoxifen does reduce the incidence of breast cancer in high-risk women is prudent. [16] The FDA approved tamoxifen to lower the risk of developing breast cancer.

Toxicity monitoring and management

Although breast cancer, endometrial cancer, thromboembolic events and cataracts occurred in less than 1 percent of patients, the nature of these toxicities warrants monitoring. Women taking tamoxifen to reduce their risk of developing breast cancer should perform monthly self breast exams and have annual mammograms. The risk of endometrial cancer with tamoxifen use is low, but it is recommended that women taking this agent have an annual gynecologic exam. [4,16] In the clinical trials, this monitoring allowed researchers to detect the majority of breast and endometrial tumors at an early stage when they were highly curable. [4,11,12] Perimenopausal and postmenopausal women should be counseled about the signs and symptoms of thrombosis and the risk of cataracts. There is not a specific guideline to screen for thrombosis, but patients should be counseled that unilateral limb swelling and pain or shortness of breath may be symptoms of a blood clot and reason to seek medical attention. Similarly, subjects should be counseled about cataracts and instructed to see an ophthalmologist before starting tamoxifen and again if they have any vision changes. [17]

Tamoxifen is associated with non-life-threatening but bothersome adverse effects including hot flashes, headaches, nausea, vaginal bleeding, vaginal itching and vaginal discharge. Hormone replacement therapy with estrogen can relieve hot flashes caused by tamoxifen, but it is rarely indicated due to an increased risk of breast cancer. [18] Clinical trials have shown that many non-estrogen agents may be used successfully to manage hot flashes including Bellergal-S, clonidine, low-dose megestrol or medroxyprogesterone or vitamin E (see Table 5 on page CP31). [19] Although most of these treatment options are effective to some degree, they are often associated with adverse effects, which must be considered when selecting an agent. Headaches usually dissipate over time and can typically be relieved with over-the-counter analgesics. Nausea also decreases over time, but severe cases can usually be managed with promethazine or prochlorperazine. Women who experience abnormal vaginal bleeding or discharge should have a full gynecologic exam because it can be an indicator of endometrial cancer. The only pharmacologic manipulation to decrease this side effect is discontinuation of tamoxifen therapy. Overall, less than 5 percent of patients receiving tamoxifen stop using the drug secondary to toxicity. [15]

Summary/conclusions

Tamoxifen is the first agent approved to reduce the risk of breast cancer. Careful patient selection is necessary in order to optimize the risk-to-benefit ratio in using tamoxifen for chemoprevention. Women with a high risk of developing breast cancer and a low risk of adverse effects are clearly the best candidates. Although all high-risk women are likely to benefit from tamoxifen use, younger women at high risk will have the most favorable risk-to-benefit ratio. If a woman is deemed an appropriate candidate for tamoxifen therapy, she should be counseled about the potential benefits and risks involved with chemoprevention. Ultimately, the woman should be involved in the chemoprevention decision-making process.

Many high-risk women are motivated to lower their risk of developing breast cancer by their fear of ultimately developing the disease. Many of these women maintain intense screening, while others undergo prophylactic radical mastectomy. Tamoxifen is another alternative for such motivated patients. There remains some concern about endometrial cancer with tamoxifen, so other agents are being tested as chemopreventive agents, namely raloxifene, a new selective estrogen receptor modulator.

The STAR (study of tamoxifen and raloxifene) trial recently began enrolling patients to compare tamoxifen and raloxifene as chemopreventive agents for breast cancer. Raloxifene, a new agent to prevent osteoporosis, is promising because its action is similar to tamoxifen, but it does not appear to increase the risk of endometrial cancer. Pooled data from placebo-controlled trials with raloxifene showed an approximate 50 percent reduction in breast cancer cases. These trials were not designed to evaluate breast cancer incidence, but justified the STAR trial. [20] Women who are motivated to reduce their risk of breast cancer should see if they qualify for the STAR trial or ask their physicians about taking tamoxifen.

For a complete list of references, contact the editor.

                   Personal factors influencing incidence
                     or relative risk of breast cancer

Factor          Condition                       Incidence   Relative risk
Age -- annual   35-39 year                        0.06%
incidence by    40-44                             0.12
5-year group
(Female only)   45-49                             0.20
                50-54                             0.25
                55-59                             0.29
                60-64                             0.35
                65-69                             0.41
                70-74                             0.46
                75-79                             0.48
                80-84                             0.47
                85+                               0.42
                lifetime                         12.0
Genetics        BRCA1 or BRCA2 (+)               56-80%
(lifetime risk) (found in  1/800 women)
Number of       0                                 12%
primary rela-
tives with      1 (dx [greater than]50 years)     12
breat cancer    1 (dx [less than]50 years)       13-21
(lifetime risk) 2 (dx [greater than]50 years)    11-24
                2 (dx [less than]50 years)       25-48
Radiation       age [less than]16             35% by age 40
to breast       (for Hodgkin's Disease)
(lifetime risk) Chest X-ray                        12
Age             [less than]12 years                              1.3
of menarche
Age of          [less than]55 years                            1.5-2.0
menopause       [less than]45 years                              0.77
Age of first    25-29 years                                      1.5
live birth      after 30 years                                   1.9
                after 35 years                                 2.0-3.0
                Nulliparous                                    1.4-3.0
Breast disease
Non-            Cysts, Fibrosis,                                 1.0
 proliferative  Fibroadenoma, etc.
Proliferative   Hyperplasia, Papilloma                         1.5-2.0
Atypia          Atypical Hyperplasia                           4.0-5.0
Carcinoma                                                     6.9-12.0
 in situ
Alcohol intake  3-9 drinks/week                                  1.3
                [greater than]9 drinks/week                      1.6
Fat intake      Change in estrogen levels                      unclear


          High-risk female populations who qualified for the BCPT

Age               Family history
Over 35 years old 1 1st degree relative w/breast CA &

or                2 1st degree relatives w/breast CA &
Over 40 years old 2 1st degree relatives w/breast CA &

or                1 1st degree relative w/breast CA &

Over 45 years old 2 1st degree relatives w/breast CA
or                1 1st degree relative w/breast CA &
Over 50 years old 1 1st degree relative w/breast CA &
or

Over 55 years old 1 1st degree relative w/breast CA
or

Over 60 years old


Age               Biopsy and estrogen exposure indicators
Over 35 years old [greater than or equal to] 2 benign
                    bx and hx atypical hyperplaisa
or                [greater than or equal to] 1 benign bx
Over 40 years old First child @ [greater than or equal to]
                    29 or nulliparous
or                [greater than or equal to] 1 benign
                    bx showing atypical hyerplaisa
Over 45 years old
or                [greater than or equal to] 1 benign bx
Over 50 years old First child @ [greater than or equal to] 20
or                [greater than or equal to] 1 bx w/ atypical
                    hyperplasia and menarch [less than or equal to] 11
Over 55 years old
or                Menarche [less than or equal to] 11 and
                    first child [greater than or equal to] 30
Over 60 years old



                     NSABP P-1 patient characteristics

Characteristic          Placebo (n) Tamoxifen (n)
Age in years
35-39                        185          159
40-49                      2,411        2,422
50-59                      2,017        2,031
60-69                      1,590        1,571
70+                          396          393
First-degree relatives
with breast cancer
0                          1,595        1,540
1                          3,731        3,754
2                          1,092        1,069
3+                           181          213
Carcinoma in situ            411          415
Atypical hyperplasia         614          579
5-year predicted risk %
1.66-2.00                  1,660        1,636
2.01-3.00                  2,031        2,057
3.01-5.00                  1,791        1,741
[greater than]5.00         1,117        1,169
Total                      6,599        6,576



             Comparison of the breast cancer prevention trials

Characteristic                    NSABP  Royal Marsden      Italian
Number of subjects               13,388      2,471           5,408
Median follow-up months           54.6         70              46
Annual incidence of breast CA
in placebo group per 1,000 women   6.7        5.5             2.3
Age less than 50 years             40%        62%             36%
Estrogen replacement therapy        0%        42%              8%
Number of subjects               19.7% P    30.5% P       26.3% total
 prematurely stopping therapy    23.7% T    39.8% T    14% in first year



                 Non-estrogen agents to manage hot flashes

Agent and dosing
Vitamin E          400 IU orally twice daily
Bellergal-S        1 table orally twice daily
Clonidine          0.1 - 0.5 mg orally at bedtime
Clonidine TTS      1 patch changed weekly
Megesterol Acetate 20 - 40 mg orally twice a day



               Invasive breast cancer in the NSABP P-1 study
Number of invasive breast cancer cases by year of participation in the
study

   Placebo Tamoxifen
1  24      24
2  46      21
3  39      24
4  31      16
5  26       8
6   9       4


                  Number of cases by predicted 5-year risk
                        of developing breast cancer

                     Placebo Tamoxifen
1.66-2.00            35      13
2.01-3.00            42      29
3.01-5.00            43      27
[greater than] 5.00  55      20

COPYRIGHT 1999 Lebhar-Friedman, Inc.
COPYRIGHT 2000 Gale Group