Chemoprevention

Michelle I. Savage

Powel H. Brown

INTRODUCTION

Breast cancers represent the highest proportion of noncutaneous cancer in women in the United States. Current estimates predict more than 226,000 diagnoses of breast cancer in these women, resulting in 39,000 deaths in 2012 alone (1). Critical insights have been and continue to be established in breast carcinogenesis, resulting in strategies enabling more effective screening, risk assessment, risk reduction, and intervention.

Screening methods, such as digital mammography and breast MRI scans, are now routine, with standard practice including annual mammographic screening for all women beginning at age 40 (2) and annual MRI scans gaining in use for women at very high risk. Other tests not yet routine but gaining in use include breast ultrasound and tomography analyses.

Effective prevention of breast cancer is critically dependent upon the identification of high-risk patients and has been more thoroughly developed than other cancers. Classification of individuals at high risk based upon known risk factors and biomarkers specific for breast cancer (e.g., obesity, alcohol consumption, BRCA1/2, TP53, PTEN, mammographic breast density, family history, and endocrine-related risks, such as lack of children and early menarche) facilitates the identification of women most likely to benefit from early intervention. This enables targeted chemoprevention specifically within higher risk populations, maximizing the potential for effective prevention. To this end, models aimed at assessing risk have been developed and continue to be the focus of studies seeking to further improve their effectiveness at predicting breast cancer risk. These include the Gail, Tyrer-Cuzick (“IBIS”), Berry-Parmigiani-Aguilar (BRCA-Pro), Claus, and Couch models as well as the Breast and Ovarian Analysis of Disease Incidence and Carrier Estimation Algorithm (BOADICEA) (3, 4). Each of these models was developed to facilitate stratification of the population into categories of predictable risk, defining individuals at low, average, moderate, high and very high risk, thereby enabling the identification of those most likely to benefit from preventive therapies.

In addition to screening and risk assessment, recent advances in breast carcinogenesis research have led to more effective strategies for risk reduction. Prophylactic bilateral mastectomies, which have demonstrated a 90% reduction of breast cancer risk, are currently used as an aggressive strategy for primary breast cancer prevention among extremely high-risk women (e.g., BRCA1/2 carriers) (5). However, breast cancer prevention incorporates both less aggressive strategies designed for the general population (e.g., behavioral approaches reducing exposure to carcinogens such as medical radiation, limiting alcohol consumption, and maintaining a healthy weight) as well as the more aggressive risk-based approaches (e.g., preventive therapeutics, bilateral mastectomy, and possibly vaccines). Today, primary breast cancer prevention efforts integrate these strategies, centering on reducing exposure to known carcinogens and exogenous estrogen in combination with behavioral strategies to reduce risk. The remaining prevention strategies are used in high-risk groups, including preventive therapies approved by the FDA (e.g., tamoxifen and raloxifene), surgical strategies (bilateral mastectomies), and additional interventions, such as novel drugs and vaccines currently being tested in clinical trials.

Chemoprevention provides the means to reduce breast cancer incidence and is the current focus of a broad range of studies investigating the therapeutic potential of natural and synthetic agents for the prevention of breast cancer. However, as breast cancer encompasses both estrogen receptor (ER)-positive and ER-negative subtypes, distinct chemopreventive strategies may be required for effective intervention. This requires the evaluation of both short- and long-term toxicities of preventive agents to establish the individualized risk-benefit ratios.

In this chapter, we will outline the results of landmark clinical studies testing the selective estrogen receptor modulators (SERMs) tamoxifen and raloxifene that demonstrated the effectiveness of chemoprevention in breast cancer. We will also review additional studies targeting the ER for breast cancer prevention as well as strategies focused
on decreasing risk of ER-negative breast cancer. Finally, we will summarize current recommendations for management of women at increased risk of both ER-positive and ER-negative breast cancer. Strategies incorporating multiple aspects of prevention carry the highest potential for effective reduction of breast cancer incidence and mortality. Furthermore, those strategies addressing the individual patient as a whole, combining risk assessment, screening, and preventive strategies, will lay the foundation for breast cancer prevention in the future.

BREAST CANCER CHEMOPREVENTION

Antiestrogen drugs are highly effective for the treatment of breast cancer and have been shown to reduce the incidence of second primary breast cancers in women with early stage breast cancer (6, 7). These results led to testing selective estrogen receptor modulators (SERMs), as well as other hormonal agents, for primary prevention of breast cancer in high-risk women.

Selective Estrogen Receptor Modulators (SERMs)

Five antiestrogen SERMs have been tested in clinical trials over the past two decades, including tamoxifen, raloxifene, idoxifene, arzoxifene, and lasofoxifene. The first-generation SERM, tamoxifen, was the first FDA-approved endocrine preventive therapy in high-risk women; the four large Phase III cancer prevention trials testing tamoxifen are outlined in Table 20-1. In addition, three Phase III studies have been conducted to determine the preventive effects of the second generation SERM, raloxifene, on breast cancer, followed by a fourth study comparing treatment with raloxifene versus tamoxifen. More recently, two Phase III studies have tested the third generation SERMs lasofoxifene and arzoxifene.

Tamoxifen

Endocrine treatment has been shown to reduce recurrence and mortality rates of ER-positive breast cancer and is able to do so in a manner independent of chemotherapy (6). The Early Breast Cancer Trialists’ Collaborative Group (EBCTCG) recently reported an updated meta-analyses for women in trials that examined the 10- to 15-year follow-up effects of 5 years of adjuvant tamoxifen (6, 7). This report demonstrated that second primary breast cancers were reduced by 50% in women using tamoxifen. These studies laid the foundation for investigations focused on the development of breast cancer preventive drugs and resulted in a series of Phase III tamoxifen prevention trials in moderate-to-highrisk women with no diagnosis of breast cancer (Table 20-1). These four Phase III studies follow in chronological order and form the initial timeline for SERM-based therapeutic breast cancer prevention.

Royal Marsden Trial: Recruitment for the Royal Marsden Tamoxifen Breast Cancer Prevention Trial extended from 1986 to 1996 (8). Initially designed as a pilot trial, the primary goal was to determine the preventive effects of tamoxifen in 2,494 high-risk undiagnosed women. 20-year follow-up results identified a non-statistically significant decrease in overall (hazard ratio [HR] = 0.84, CI 0.64-1.10, p = .2) and invasive (HR = 0.78, CI 0.58-1.04, p = .005) breast cancer incidence following tamoxifen treatment (8). In addition, this study demonstrated a significant effect of tamoxifen on ER-positive (HR = 0.61, CI 0.43-0.864) but not ER-negative breast cancers.

NSABP P-1 (BCPT) Trial: The largest of the SERM breast cancer prevention trials, the National Surgical Adjuvant Breast and Bowel Project (NSABP) Breast Cancer Prevention Trial (BCPT) P-1, recruited 13,388 women from 1992-1996 (9). The effect of 5 years of tamoxifen treatment (20 mg/day, n = 6,681) on the incidence of invasive breast cancer was investigated in pre- and postmenopausal women at increased risk of breast cancer. A 49% decrease in invasive breast cancer was seen with tamoxifen versus placebo (relative risk [RR] = 0.51, CI 0.39-0.66). These initial results were reported in 1998 and led to early termination of the trial, followed by FDA approval of tamoxifen for treatment of women at high risk of breast cancer. The 7-year follow-up results confirmed the initial study findings, demonstrating reductions in invasive (RR = 0.57, CI 0.46-0.70), noninvasive (RR = 0.63, CI 0.45-0.89), ER-positive (RR = 0.38, CI 0.28-0.50), and ductal carcinoma in situ (DCIS) (HR = 0.54, CI 0.36-0.80) breast cancers, but no decrease in ER-negative tumor incidence (9). Although the NSABP P-1 trial also identified toxicity associated with tamoxifen, including increased hot flushes, vaginal discharge, and increased risk of endometrial cancer and thromboembolic events, the study established tamoxifen as the most effective treatment strategy for the prevention of breast cancer, particularly, ER-positive breast cancer.

Italian Tamoxifen Prevention Study: The Italian Randomized Tamoxifen Prevention Trial recruited 5,408 women at normal risk of breast cancer (10). However, to avoid the undesired side effect of increased incidence of endometrial cancer, the Italian trial limited the study population to healthy women who had previously undergone a hysterectomy. Importantly, many of these women took postmenopausal hormones after their hysterectomy. An unanticipated high subject dropout rate (26%) resulted in early termination of recruitment, but follow-up was continued with the previously enrolled subjects. While no statistically significant reduction in breast cancer incidence was initially observed, the 11-year follow-up report demonstrated a 76% reduction in hormone receptor (HR)-positive breast cancers in high-risk patients previously treated with a bilateral oophorectomy (RR = 0.24, CI 0.10-0.59), suggesting a preventive effect associated with tamoxifen (10).

IBIS-I Trial: Recruitment for the International Breast Cancer Intervention Study I breast cancer prevention trial spanned from 1992 to 2001 and resulted in the accrual of 7,154 women at high risk for breast cancer (11). With the primary objective of identifying whether the risk-benefit ratio associated with tamoxifen treatment was sufficient to support its use for the prevention of breast cancer, this study measured breast cancer incidence following 5 years of treatment with tamoxifen (20 mg/day, n = 3,578) versus placebo (n = 3,566). The 96-month follow-up report demonstrated a reduction of risk in subjects not receiving hormone replacement therapy (HRT) while on treatment that was limited to all (RR = 0.62, CI 0.46-0.83), ER-positive (RR = 0.49, CI 0.32-0.74), and DCIS (HR = 0.52, CI 0.27-0.99) breast cancer, although a non-statistically significant decrease in ER-negative breast cancer was reported. In addition, 5 years of tamoxifen treatment was associated with an improved long-term risk-benefit ratio characterized by a cancer preventive benefit persisting for 10 years and a reduction in toxicity after stopping treatment (11).

Adverse Events Associated with Tamoxifen Treatment: Across the four large-scale Phase III tamoxifen cancer prevention trials (all treating with 20 mg/day for 5 years, except the Royal Marsden Trial, which included 8 years of treatment), endometrial/uterine cancer and thromboembolic and cardiovascular events constituted the most common adverse events (8, 9, 10 and 11). Other negative side effects most consistently reported within these studies included cerebrovascular events, vaginal symptoms, and hot flushes or cold/night sweats (Table 20-2). In addition, an increase in risk of cataracts was observed in subjects receiving tamoxifen versus placebo in the NSABP P-1 (21%) and IBIS-I (0.4%, not statistically significant) trials. Most negative effects related to tamoxifen treatment did not persist beyond the active treatment period, and while long-term follow-up reports demonstrate no significant increase in total or cause-specific death, neither do they demonstrate a significant improvement in survival among subjects taking tamoxifen versus placebo (Tables 20-2 and 20-3).

TABLE 20-1 Selective Estrogen-Receptor Modulator (SERM) Breast Cancer Prevention Studies

Trial

Participants

Study Design

Reduction in Incidence

RR/HR (95%CI)

TAMOXIFEN

Royal Marsden¹, ²

(Recruitment: 1986-1996)

• 2,494 high-risk women

• 30-70 y of age

1,238 tamoxifen (20 mg/d) and 1,233 placebo

Treatment time: for 8 y

All breast cancers: 16%

Invasive breast cancer: 22%

Invasive ER-positive breast cancer: 39%

HR 0.84 (0.64-1.10)

HR 0.78 (0.58-1.04)

HR 0.61 (0.43-0.86)

NOTE: No difference initially; longer follow-up showed reductions in invasive and ER-positive breast cancer incidence, but no significant change in ER-negative or overall breast cancer incidence

NSABP-P1(BCPT)³, ⁴

(Recruitment: 1992-1997)

• 13,388 high-risk women

• >35 y of age

6,681 tamoxifen (20 mg/d) and 6,707 placebo

Treatment time: 5 y

Follow-up: 7 y

Invasive breast cancer: 43%

Invasive ER-positive breast cancer: 62%

Noninvasive DCIS/LCIS breast cancer: 37%

RR 0.57 (0.46-0.70)

RR 0.38 (0.28-0.50)

RR 0.63 (0.45-0.89)

Italian⁵, ⁶

(Recruitment: 1992-1997)

• 5,408 normal-risk women with a hysterectomy

2,700 tamoxifen (20 mg/d) and 2,708 placebo

Treatment time: 5 y

All breast cancers: 16%

Invasive breast cancer: 20%

Invasive ER-positive breast cancer: -10%

RR 0.84 (0.60-1.17)

RR 0.80 (0.56-1.15)

RR 1.10 (0.59-2.05)

• 35-70 y of age

Follow-up: 11 y

Noninvasive breast cancer: -50%

RR 1.50 (0.53-4.20)

IBIS-I⁷, ⁸

(Recruitment: 1992-2001)

• 7,154 high-risk women

• 35-70 y of age

3,579 tamoxifen (20 mg/d) and 3,575 placebo

Treatment time: 5 y

Follow-up: 96-mo

All breast cancers: 27%

Invasive breast cancer: 26%

Invasive ER-positive breast cancer: 34%

Noninvasive DCIS breast cancer: 37%

RR 0.73 (0.58-0.91)

RR 0.74 (0.58-0.94)

RR 0.66 (0.50-0.87)

RR 0.63 (0.32-1.20)

RALOXIFENE

MORE⁹

(Recruitment: 1994-1998)

• 7,705 postmenopausal women with low BMD

• <80 y of age

2,557 raloxifene (60 mg/d), 2,572 raloxifene (120 mg/d) and 2,576 placebo

Treatment time: 4 y

Follow-up: 40 mos avg.

All breast cancers: 65%

Invasive breast cancer: 76%

Invasive ER-positive breast cancer: 90%

RR 0.35 (0.21-0.58)

RR 0.24 (0.13-0.44)

RR 0.10 (0.04-0.24)

CORE¹⁰

(Recruitment: 1998-2002)

• 5,213 postmenopausal women with low BMD

(reconsented from MORE trial)

• <80 y of age

3,510 raloxifene (60 mg/d) and 1,703 placebo

Treatment time: an additional 4 years after 4 y of raloxifene on MORE trial

All breast cancers: 50%

Invasive breast cancer: 59%

Invasive ER-positive breast cancer: 66%

Noninvasive breast cancer: -78%

HR 0.50 (0.30-0.82)

HR 0.41 (0.24-0.71)

HR 0.34 (0.18-0.66)

HR 1.78 (0.37-8.61)

RUTH¹¹

(Recruitment: 1998-2000)

• 10,101 postmenopausal women with CHD

• >35 y of age

5,044 raloxifene (60 mg/d) and 5,057 placebo

Treatment time: median 5.6 y

All breast cancers: 33%

Invasive breast cancer: 44%

Invasive ER-positive breast cancer: 55%

Noninvasive breast cancer: -117%

Reduction in invasive ER-positive tumors

HR 0.67 (0.47-0.96)

HR 0.56 (0.38-0.83)

HR 0.45 (0.28-0.72)

HR 2.17 (0.75-6.24)

TAM VS RALOX

NSABP-P2¹², ¹³

(STAR)

(Recruitment: 1999-2004)

• 19,490 high-risk, postmenopausal women

• >35 y of age

9,736 tamoxifen (20 mg/d) and 9,754 raloxifene (60 mg/d)

Treatment time: 5 y

Follow-up: median 81-mo

Raloxifene vs. tamoxifen:

Invasive breast cancer: -24%

Noninvasive DCIS breast cancer: -22%

Noninvasive DCIS/LCIS breast cancer: -2%

RR 1.24 (1.05-1.47)

RR 1.22 (0.88-1.69)

RR 1.02 (0.61-1.70)

NOTE: Accrual figures as represented on 81-mo follow-up report

Extrapolated data for raloxifene vs. placebo:

Invasive breast cancer: 38%

Noninvasive breast cancer: 39%

LASOFOXIFENE

PEARL¹⁴, ¹⁵

(Recruitment: 2001-2007)

• 8,556 women with low BMD

• 59-80 y of age

2,852 lasofoxifene (0.25 mg/d), 2,852 lasofoxifene (0.5 mg/d) and 2,852 placebo

0.5 mg lasofoxifene vs. placebo:

All breast cancers: 79%

Invasive ER-positive breast cancer: 83%

HR 0.21 (0.08-0.55)

HR 0.17 (0.05-0.57)

Treatment time: 5 y

Noninvasive DCIS breast cancer: 50%

HR 0.50 (0.09-2.73)

ARZOXIFENE

GENERATIONS¹⁶, ¹⁷

(Recruitment: 2004-2009)

• 9,354 women with low

BMD

• 60-85 y of age

4,676 arzoxifene (20 mg/d) and 4,678 placebo

Treatment time: ≤60 mo (Results reported are at 48 mo follow-up)

All breast cancers: 59%

Invasive ER-positive breast cancer: 70%

Noninvasive breast cancer: 70%

Noninvasive DCIS breast cancer: 62%

HR 0.41 (0.25-0.68)

HR 0.30 (0.14-0.63)

HR 0.30 (0.08-1.09)

HR 0.38 (0.10-1.42)

bone mineral density (BMD); continued outcomes of raloxifene evaluation (CORE); coronary heart disease (CHD); International Breast Intervention Study (IBIS-I); Italian Randomized Tamoxifen Prevention Trial (Italian); Multiple Outcomes of Raloxifene Evaluation (MORE); National Surgical Adjuvant Breast and Bowel Project Breast Cancer Prevention Trial (BCPT) P1 (NSABP-P1); National Surgical Adjuvant Breast and Bowel Project Study of Tamoxifen and Raloxifene (STAR) P2 (NSABP-P2); Postmenopausal Evaluation and Risk-Reduction with Lasofoxifene (PEARL) Trial; Raloxifene Use for the Heart (RUTH) Trial; Royal Marsden Tamoxifen Prevention Trial (Royal Marsden).

Data from:¹, ²Powles TJ, et al. J Natl Cancer Inst 2007; 99(4):283-290 and Br J Cancer 1989; 60(1):126-131;

³, ⁴Fisher B, et al. J Natl Cancer Inst 2005; 97(22):1652-1662 and 1998; 90(18): 1371-1388;

⁵, ⁶Veronesi U, et al. J Natl Cancer Inst 2007; 99(9):727-737 and 2003; 95(2):160-5;

⁷, ⁸Cuzick J, et al. Lancet 2002; 360(9336):817-824 and J Natl Cancer Inst 2007; 99(4):272-282;

⁹Cauley JA, et al. Breast Cancer Res Treat 2001; 65(2):125-134;

¹⁰Vogel VG, et al. Clin Breast Cancer 2009; 9(1):45-50;

¹¹Grady D, et al. J Natl Cancer Inst 2008; 100(12):854-61;

¹², ¹³Vogel VG, et al. JAMA 2006; 295(23):2727-2741 and Cancer Prev Res (Phila) 2010; 3(6):696-706;

¹⁴, ¹⁵Cummings SR, et al. N Engl J Med 2010; 362(8):686-696 and LaCroix AZ, et al. J Natl Cancer Inst 2010; 102(22):1706-1715;

¹⁶, ¹⁷Powles TJ, et al. Breast Cancer Res Treat 2012; 134(1):299-306 and Cummings SR, et al. J Bone Miner Res 2011; 26(2):397-404.

Endometrial/Uterine Cancer Increased risk of endometrial cancer following treatment with tamoxifen was reported in the Royal Marsden, NSABP P-1, and IBIS-1 Trials. The first published evidence of this was in the 1998 interim analysis of the Royal Marsden study that, by the 20-year followup, identified a 2.5-fold increase in endometrial cancer for subjects who had received tamoxifen versus placebo (8). The NSABP P-1 and IBIS-1 Trials have also reported a 1.5- to 3.4-fold elevated risk of endometrial cancer associated with tamoxifen therapy in participants ≥50 years of age (9). The majority of all endometrial cancer cases (53 in the tamoxifen arm, 17 in the placebo arm) were classified as International Federation of Gynecology and Obstetrics (FIGO) (12) stage I.

Thromboembolic and Cardiovascular Events Primary incidences of thromboembolic and cardiovascular events consistently reported in the four trials were in increased rates of stroke and venous thromboembolic events. Strokes were 25% lower for subjects in the tamoxifen arm of the Royal Marsden Trial (8); however, the NSABP P-1 (9), IBIS-I (11), and Italian (10) trials reported non-statistically significant increased rates of stroke/cerebrovascular accidents in the tamoxifen study arm. Increased incidence of deep vein thrombosis and overall thromboembolic events was reported in the tamoxifen arms of the Royal Marsden (8), IBIS-I (11), and NSABP P-1 (9) trials.

Risk vs. Benefit: In an analysis of the risks and benefits associated with tamoxifen treatment for the prevention of breast cancer, Gail and coworkers concluded that these are dependent on the age, race, and breast cancer risk level of the individual (13). They described the increased risk for deep vein thrombosis, endometrial cancer, pulmonary embolism, and stroke, predicting 15, 16, 15, and 13, respectively, additional cases per 1,000 women following 5 years of tamoxifen treatment. In addition, differential efficacy of tamoxifen was age and race dependent, with the overall benefit defined as 97, 53, and 1 fewer cases for invasive breast cancer, in situ breast cancer, and hip fractures, respectively, per 1,000 women treated with tamoxifen for 5 years.

Due to the variation in risk-benefit effects and ratios for African American women versus white women, the Gail model, originally developed to predict risk in white women using the NCI’s Breast Cancer Risk Assessment Tool (www. cancer.gov/bcrisktool), has been updated specifically for African American women. This risk assessment model, known as the Women’s Contraceptive and Reproductive Experiences (CARE) model, more accurately predicts risk of breast cancer within this population. Finally, while tamoxifen has been shown to significantly reduce risk of invasive breast cancer by 49%, a greatly increased benefit is seen in women with prior atypical ductal hyperplasia (ADH) and lobular carcinoma in situ (LCIS) lesions, which exhibit relative reductions of 86% and 56%, respectively (14).

Preventive Therapy is Now Standard of Care: Collectively, the results of the tamoxifen trials, particularly the dramatic results of the NSABP P-1 breast cancer prevention trial, led to FDA approval of the SERM tamoxifen as a viable therapeutic strategy for breast cancer risk reduction. This decision, representing the first approval of a preventive agent for breast cancer risk reduction by the FDA, has led to the acceptance of tamoxifen as the standard premenopausal endocrine therapy for the prevention of breast cancer, especially ER-positive breast cancer.

In 1999, based upon the collective results of these four Phase III randomized clinical trials investigating the effects of tamoxifen in the prevention of breast cancer, the American Society of Clinical Oncology (ASCO) published a review of the data with recommendations for its use as a cancer preventive therapy (15). The most recent ASCO guidelines include tamoxifen therapy (20 mg/d for 5 years) as a treatment option for long-term (≥10 years) risk reduction of invasive ER-positive breast cancer in premenopausal women ≥35 years of age with LCIS or a 5-year projected absolute risk of breast cancer ≥1.66% (16). However, tamoxifen should not be used in women with a history or at high risk of deep venous thrombosis, pulmonary embolism, or cerebral vascular accidents. In addition, ASCO guidelines recommend avoiding tamoxifen use in individuals who are immobilized, pregnant, nursing, or receiving HRT. Despite FDA approval and consensus recommendations by ASCO and other professional organizations, tamoxifen is very rarely used for breast cancer prevention due to patient concerns about side effects and a lack of demonstrated survival benefit from tamoxifen use in this setting.

Raloxifene

Raloxifene was initially developed as a potential therapeutic agent for osteoporosis for the prevention of bone fractures. Three large-scale Phase III raloxifene clinical trials have since investigated the effects of the drug versus placebo as a preventive therapy for breast cancer, bone fractures, and heart disease (the Multiple Outcomes of Raloxifene Evaluation [MORE], Continuing Outcomes Relevant to Evista [CORE], and Raloxifene Use for the Heart [RUTH] trials) (Table 20-1). In addition, the NSABP Study of Tamoxifen and Raloxifene (STAR) P-2 Trial compared raloxifene to tamoxifen. The consecutive nature of the information provided by the MORE/CORE trials has enabled the comparison of incidence rates among women treated for 4 years (duration of treatment in the MORE trial) versus 8 years (duration of treatment in both the MORE and CORE trials).

Multiple Outcomes of Raloxifene Evaluation: In the 1990s, the MORE trial was conducted to determine the effectiveness of raloxifene in reducing bone fractures in 7,705 postmenopausal women 80 years of age or younger with osteoporosis, but also investigated the secondary endpoints of breast cancer and heart disease (17). Participants were treated for 3 years with low-dose raloxifene (60 mg/day), high-dose raloxifene (120 mg/day), or placebo, and all participants received both calcium and Vitamin D (cholecalciferol) supplements. This trial demonstrated that postmenopausal raloxifene treatment was associated with a dose-dependent reduction in vertebral bone fractures (60 mg/d: RR = 0.7, CI 0.5-0.8; 120 mg/d: RR = 0.5, CI 0.4-0.7) (17). In addition, dose-independent reductions were observed in all (RR = 0.35, CI 0.21-0.58) and invasive breast cancers (RR = 0.24, CI 0.13-0.44). However, as with tamoxifen, the study results demonstrated a decrease in risk of ER-positive (RR = 0.10, CI 0.04-0.24), but not ER-negative breast cancer.

TABLE 20-2 Select Adverse Events and Side Effects Associated with Tamoxifen

		Royal Marsden Trial¹, ² (events on treatment or for entire follow-up; participant figures at 20-year follow-up)			NSABP P-1 Trial (BCPT)³, ⁴ (participant figures at 7-year follow-up)				Italian Study⁵, ⁶ (events during active treatment; participant figures at baseline)				IBIS-1 Trial⁷, ⁸ (entire period; participant figures at 96-month follow-up)
Reported Event		Tamoxifen n = 1079	Placebo n = 1034	p-value	Tamoxifen n = 6681, rate^b	Placebo n = 6707, rate^b	RR^a	95% CI	Tamoxifen n = 2700, rate^b	Placebo n = 2708, rate^b	RR^a	95% CI	Tamoxifen %; n = 3579	Placebo %, n = 3575	RR^a	95% CI
Osteoporotic fractures		19	22	0.6	80, 1.97	116, 2.88	0.68	0.51-0.92	—	—	—	—	91, 2.5	76, 2.1	1.19	0.89-1.62
Cancer:		—	—	—	—	—	—	—	—	—	—	—	—	—	—	—
Endometrial/uterine		13	5	0.06	—	—	—	—	—	—	—	—	17	11	1.55	0.68-3.65
	Invasive in situ	— —	— —	— —	53, 2.24 1, 0.04	17, 0.68 3, 0.12	3.28 0.35	1.87-6.03 0.01-13.17	— —	— —	— —	— —	— —	— —	— —	— —
Other than breast and/or endometrial/uterine		64^e	70^e	0.8^e	178^e	155^e	—	—	106, 4.37^d	91, 3.73^d	1.17^d	0.88-1.55^d	—	—	—	—
Cerebrovascular event		—	—	—	—	—	—	—	12, 1.19	7, 0.67	1.78	0.70-4.52	32, 1.12	34, 1.19	0.94	0.56-1.57
Cardiovascular problems (vascular-related):		10	12	0.2	—	—	—	—	5, 0.49	5, 0.48	1.04	0.30-3.58	—	—	—	—
Stroke		7	9	0.6	71, 1.75	50, 1.23	1.24	0.97-2.08	6, 0.59	2, 0.19	3.11	0.63-15.4	—	—	—	—
Stroke/cerebrovascular accident		—	—	—	—	—	—	—	—	—	—	—	15, 0.53	12, 0.42	1.25	0.55-2.93
Transient ischemic attack (TIA)		—	—	—	31, 0.76	34, 0.84	0.91	0.54-1.52	6, 0.59	5, 0.48	1.24	0.38-4.08	17, 0.60	22, 0.77	0.77	0.39-1.52
Pulmonary embolism (PE)		—	—	—	28, 0.69	13, 0.32	2.15	1.08-4.51	—	—	—	—	—	—	—	—
Venous thromboembolic event		8	3	0.2	—	—	—	—	44, 4.45	28, 2.72	1.63	1.02-2.62	—	—	—	—
Superficial thrombophlebitis		—		—	—				—	—	— —	23, 0.81	8, 0.28	2.88	1.24-7.44
Deep vein thrombosis (DVT)		4^c	2^c	—	49, 1.21	34, 0.84	1.44	0.91-2.30	—	—	—	—	—	—	—
DVT/PE		—	—	—	—	—	—	—	—	—	—	—	23, 0.81	8, 0.28	1.84	1.21-2.82
Gynecological problems		37	13	0.001	—	—	—	—	—	—	—	—	—	—	—	—
Hysterectomy		177	96	<0.001	—	—	—	—	—	—	—	—	—	—	—	—
Vaginal symptoms:		37	17	0.008	—	—	—	—	—	—	—	—	—	—	—	—
Discharge		321	167	<0.001	54.77	34.13	1.60	—	505, 60.60	173, 17.59	3.44	2.90-4.09	—	—	—	—
Dryness		—	—	—	—	—	—	—	295, 34.09	269, 29.93	1.14	0.97-1.34	—	—	—	—
Genital itching		—	—	—	47.13	38.29	1.23	—	—	—	—	—	—	—	—	—
Bleeding		—	—	—	21.96	21.26	1.03	—	—	—	—	—	—	—	—	—
Hot flashes		598	394	<0.001	77.66	65.04	1.19	—	635, 119.29	446, 67.20	1.78	1.57-2.00	81.8	67.7		p value <.0001
Cold sweats		—	—	—	21.40	14.77	1.45	—	—	—	—	—	—	—	—	—
Night sweats		—	—	—	66.80	54.92	1.22	—	—	—	—	—	—	—	—	—
^aRelative risk (tamoxifen vs. placebo); ^bPer 1,000 women; ^cResults identified in interim report; ^dOther than breast cancer; ^eOther than breast or endometrial/uterine cancer.
Data from:¹, ²Powles TJ, et al. J Natl Cancer Inst 2007;99(4):283-290 and Br J Cancer 1989;60(1):126-131; ³, ⁴Fisher B, et al. J Natl Cancer Inst 2005;97(22):1652-1662 and 1998;90(18):1371-1388;⁵, ⁶Veronesi U, et al. J Natl Cancer Inst 2007;99(9):727-737 and 2003;95(2):160-165; ⁷, ⁸Cuzick J, et al. Lancet 2002;360(9336):817-824 and J Natl Cancer Inst 2007;99(4):272-282.

TABLE 20-3 Select Causes of Death in Tamoxifen Trials

	Royal Marsden Trial¹, ² (events on treatment or for entire follow-up; participant figures at 20-year follow-up)				NSABP P-1 Trial (BCPT)³, ⁴ (participant figures at 7-year follow-up)				Italian Study⁵, ⁶ (events during active treatment; participant figures at baseline)				IBIS-1 Trial⁷, ⁸ (entire period; participant figures at 96-month follow-up)
Cause of Death	Tamoxifen n = 1079	Placebo n = 1034	HR^a	95% CI; p-value	Tamoxifen n = 6681, rate^c	Placebo n = 6707, rate^c	RR^b	95% CI	Tamoxifen n = 2700, rate^c	Placebo n = 2708, rate^c	RR^b	95% CI	Tamoxifen %; n = 3579	Placebo %, n = 3575	RR^b	95% CI
Total Deaths:	54	54	0.99	0.95	126, 3.08	114, 2.80	1.10	0.85-1.43	36, 1.46	38, 1.54	0.95	0.60-1.49	65	55	—	—
Cancer:	42	33	—	—	57	71	—	—	22	25	—	—	36	34	—	—
Breast	—	—	—	—	12	11	—	—	2	2	—	—	11	13	—	—
Endometrial/uterine	—	—	—	—	0	1	—	—	—	—	—	—	1	0	—	—
Other	—	—	—	—	45	59	—	—	20	23	—	—	53	42	—	—
Cardiac & vascular disease	—	—	—	—
Stroke	1	2	—	—	35	22	—	—	3	5	—	—	6	2	—	—
Heart Condition	6	2	—	—
Other	5	17	—	—	33	21	—	—	11	8	—	—	19	16	—	—
^aHazard ratio; ^bRelative risk (tamoxifen vs. placebo); ^cPer 1000 women)
Data from:¹, ²Powles TJ, et al. J Natl Cancer Inst 2007;99(4):283-290 and Br J Cancer 1989;60(1):126-131; ³, ⁴Fisher B, et al. J Natl Cancer Inst 2005;97(22):1652-1662 and 1998;90(18): 1371-1388; ⁵, ⁶Veronesi U, et al. J Natl Cancer Inst 2007;99(9):727-737 and 2003;95(2):160-165; ⁷, ⁸Cuzick J, et al. Lancet 2002;360(9336):817-824 and J Natl Cancer Inst 2007;99(4):272-282.