Adjuvant Endocrine Therapy in Premenopausal Women with Breast Cancer

Abstract

Breast cancer remains the leading cause of cancer related mortality in premenopausal women. Multiple advances in local and systemic therapies have dramatically improved outcomes in women with HR+ early stage breast cancer. Despite these advances, early and late relapses occur. Therefore multiple adjuvant endocrine therapy trials have been conducted with the goal of decreasing breast cancer recurrence and mortality. Recently, large international trials evaluating extended endocrine therapy as well as ovarian suppression with and without tamoxifen or exemestane have been reported. These studies add to the large body of existing data related to adjuvant endocrine therapy in premenopausal women with breast cancer and provide additional therapeutic options in those at high risk of disease recurrence. This review will synthesize the most recent data and provide an evidenced based approach, highlighting quality-of-life concerns, when considering adjuvant endocrine therapies in premenopausal women.

Introduction

In the United States, breast cancer remains the most common cancer in women with over 230,000 new cases and 40,000 deaths per year¹. The vast majority of new cases represent early stage disease (i.e. Stage I–II) with approximately one-quarter diagnosed in premenopausal women. Hormone-receptor positive (HR+) breast cancer is the most common subtype² and decades of clinical trials optimizing adjuvant endocrine therapies have led to significant improvements in outcomes³. Most recently, large international trials have shown decreased breast cancer recurrence rates with extended endocrine therapy⁴ and adjuvant ovarian suppression^5,6. Despite these advances, the optimal strategy for endocrine therapy in premenopausal early stage HR+ breast cancer remains challenging given the good prognosis of many patients and the inherent risk of overtreatment as well as short- and long-term toxicities associated with such therapies.

This review focuses on the current state of evidence related to adjuvant endocrine therapy for HR+ breast cancer in premenopausal women, primarily the most recent data related to extended endocrine therapy and the role of ovarian suppression. Additionally, these data will be reviewed in the context of quality-of-life (QOL) and survivorship concerns as they relate to premenopausal women.

Endocrine Therapies

Adjuvant endocrine options for premenopausal women in the contemporary era include tamoxifen with or without ovarian suppression (OS)/ovarian ablation (OA), an aromatase inhibitor (AI) with OS/OA, or OS/OA alone. Treatment with endocrine therapy is indicated only for breast cancers that have estrogen receptor (ER) expression measured by clinically validated techniques⁷.

Tamoxifen is a selective estrogen receptor modulator (SERM) that can be used to treat both pre- and postmenopausal women with breast cancer. When administered for 5 years, it reduces the risk of disease recurrence in early stage breast cancer by approximately 40% and the risk of death by approximately 30%⁸. Its therapeutic effect is independent of plasma estradiol levels.

Since the goal of therapy is to reduce estrogen receptor signaling and the ovaries produce the vast majority of estrogen in premenopausal women, one alternative to tamoxifen monotherapy is ovarian ablation (OA) or ovarian suppression (OS), either alone or in combination with tamoxifen. OA is the most effective modality to suppress circulating estrogen and is achieved either via bilateral oophorectomy or radiation, both of which lead to permanent cessation of menses. Alternatively, ovarian function can be suppressed temporarily with the use of luteinizing hormone-releasing hormone (LHRH) agonists such as triptorelin, goserelin, or leuprolide. Consideration should be given to administering these intramuscular or subcutaneous depot agents every 28 days (rather than every 84 days) since most clinical trials used monthly administration and there is concern that the efficacy of the medication could wane before the end of the dosing period⁹. However, due to lack of efficacy data, use of OS/OA as the sole therapy for breast cancer treatment is not recommended unless a patient is unable or unwilling to receive treatment with another appropriate systemic therapy⁹.

Another alternative to tamoxifen is an aromatase inhibitor (AI). In postmenopausal women with early stage breast cancer the AIs, including the nonsteroidal agents anastrozole and letrozole and the steroidal agent exemestane, appear to be equally effective¹⁰ and have consistently been shown to be more effective than tamoxifen in postmenopausal women¹¹. However, AI medications alone are not useful in premenopausal women as these drugs act peripherally by blocking conversion of androgens to estrogen and have no impact in high estrogen states¹². As such, it is imperative to determine menopausal status prior to considering endocrine therapy in any patient with HR+ breast cancer, and combine therapy with OS/OA if treatment with an AI is desired in a woman who is not definitely postmenopausal.

Thus, for a premenopausal woman endocrine therapy options include tamoxifen alone, OA/OS alone, or OA/OS in combination with either tamoxifen or an AI. As is described in detail below, the determination of which premenopausal women should receive endocrine therapy in combination with OS/OA is complex because of issues related to both efficacy and tolerability (see below, section Contemporary Trials of Ovarian Suppression).

Menopausal Status and Chemotherapy-Induced Ovarian Failure

Menopause has been broadly defined as age >60, having undergone bilateral oophorectomy (BSO), or having amenorrhea for at least 12 months in the absence of factors potentially influencing menstruation (e.g. chemotherapy, tamoxifen, or OS)¹³. Confirming menopausal status can pose a challenge in women either who have undergone hysterectomy without BSO or who develop chemotherapy-induced ovarian failure (CIOF). Women under age 60 with HR+ breast cancer who have undergone hysterectomy without BSO and who will be receiving chemotherapy should have ovarian function assessed prior to chemotherapy initiation in order to determine pre-chemotherapy menopausal status. This information helps inform choice of endocrine therapy and potential need for monitoring of ovarian function.

Women with CIOF can experience reactivation of ovarian function during AI therapy despite having estradiol concentrations in the postmenopausal range at the time of AI initiation^14,15. Importantly, estradiol levels can increase even though menses do not resume. Notably, younger age at the time of chemotherapy is independently associated with a higher chance of ovarian function recovery following AI therapy, and no upper age limit has yet been identified¹⁵.

One small study of 58 women with CIOF with mean age 48 demonstrated lower disease-free survival (DFS) in the cohort that experienced ovarian function recovery based on resumption of menses or elevated estradiol levels^14,15. In contrast to these data, subgroup analyses of 105 patients with CIOF in BIG 1-98, which compared letrozole with tamoxifen as front-line therapy in postmenopausal women, and women who were premenopausal at the time of tamoxifen initiation in MA.17, which compared letrozole versus placebo after five years of tamoxifen, identified no difference in breast cancer outcomes with AI therapy in these cohorts^16,17. The reason for the discrepancy between these findings is uncertain, and may be due to differences in patient population.

In the absence of meeting age criteria (>60 years) or having undergone BSO, consensus guidelines suggest considering testing of plasma follicle-stimulating hormone (FSH) and estradiol levels to confirm a postmenopausal state¹³. However, a recommended monitoring interval has not been established and the sensitivity of available estradiol assays in most laboratories to detect very low estradiol concentrations is poor^18–20. Ultrasensitive estradiol assays utilizing mass spectroscopy are commercially available and can be considered in women at risk for recovery of ovarian function. Women with CIOF under age 50 are at high risk of ovarian function recovery. In addition, providers should not assume that all women over 50 who become postmenopausal from chemotherapy will remain postmenopausal during AI therapy. Either tamoxifen or OS/OA plus either tamoxifen or AI therapy should be used, or ovarian function should be closely monitored using ultrasensitive estradiol assays.

Efficacy of 5 Years of Adjuvant Tamoxifen

Initially convened in 1985, the Early Breast Cancer Trialists’ Collaborative Group (EBCTCG) utilizes individual patient level data to conduct meta-analyses on multiple aspects of early breast cancer therapy. Until recently, the standard adjuvant endocrine therapy for patients with HR+ premenopausal breast cancer was five years of tamoxifen²¹. This recommendation was largely based on the 2011 EBCTCG analysis comparing tamoxifen to no tamoxifen in ER-positive (ER+) (n=10,645) early stage breast cancer⁸. For the entire ER+ cohort, tamoxifen for five years reduced breast cancer recurrence risk by nearly one-half through 10 years (recurrence rate ratio [RR] 0.53 during years 0–4; RR 0.68 during years 6–9; both p<0.00001) with a stable recurrence rate observed during years 10–14 (RR 0.97). Breast cancer mortality was reduced by nearly one-third through 15 years (death RR 0.71 during years 0–4, 0.66 during years 5–9, and 0.68 during years 10–14; all p<0.0001). Most striking, the absolute mortality difference with five years of tamoxifen compared to no therapy was three times higher at year 15 (24% vs. 33%) compared to year 5 (9% vs. 12%) suggesting ongoing benefit despite discontinuation of tamoxifen (carryover effect). All findings were independent of PR status, use of chemotherapy, and nodal status. Young, presumably premenopausal, women (aged <45) compared to those aged 55–69, showed similar reductions in recurrence, breast cancer death, and all-cause mortality.

Even in patients with low-level ER+ expression compared to higher levels of expression, no statistically significant difference in efficacy was observed. In the EBCTCG meta-analysis, determination of ER overexpression was based on historical ligand-binding assays and utilized a cutoff of 10 femtomoles of receptor protein per mg cytosol protein. Although modern assessment of ER status is based on immunohistochemistry (percentage of tumor cells stained by antibody against ER), concordance with historical ligand-binding assays is good²². Current guidelines recommend that a breast cancer that has at least 1% of tumor cells positive for ER be considered ER-positive and treated accordingly, so as to avoid omitting a potentially beneficial therapy from the patient’s treatment regimen¹³.

Risk of Late Relapse in HR+ Breast Cancer

The risk of relapse in breast cancer varies over time and is dependent on a number of prognostic features including stage, histopathology, and HR status. Distinct from HR-negative (HR−) breast cancer that typically has the highest relapse rates early after diagnosis, HR+ breast cancer has a unique predilection for late relapses²³. Based on Surveillance, Epidemiology, and End Results data, HR+ breast cancer has an estimated annual hazard rate for relapse of approximately 1–2% that plateaus and persists through years 10–15²⁴.

Extended Tamoxifen Therapy (10 versus 5 years)

Despite the carryover effect of five years of tamoxifen therapy, late relapses occur. Therefore multiple trials have examined prolonged endocrine therapy to determine if continued treatment beyond 5 years will improve breast cancer outcomes. NSABP B-14, the largest historical trial, only included patients who were lymph node negative (LN−), and surprisingly observed an advantage for those who discontinued tamoxifen at five years compared to extended therapy (DFS 82% vs. 78%, respectively; p=0.03)²⁵.

In contrast to the above data, two recently reported large trials demonstrated improvements in breast cancer outcomes with extended tamoxifen therapy. The Adjuvant Tamoxifen: Longer Against Shorter (ATLAS)⁴ trial randomized women with ER+ or ER-unknown early stage breast cancer who had completed five years of tamoxifen to five additional years or to observation. The Adjuvant Tamoxifen-To Offer More? (aTTom)²⁶ trial randomized women with ER+ or ER-unknown early stage breast cancer to 5 or 10 years of tamoxifen. Ten years following randomization, ATLAS identified an absolute reduction in breast cancer recurrence in ER+ patients of approximately 4% (21.4% who continued tamoxifen vs. 25.1% on observation; p=0.002). There was also a very modest absolute reduction in breast cancer mortality and a non-significant difference in overall mortality in the entire cohort (Table 1). As was observed in the 2011 EBCTCG meta-analysis, the effect of extended tamoxifen was time-dependent with a relatively minor decrease in recurrence rates during years 5–9 (RR, 0.90; 95% CI, 0.79–1.02, p=0.10) and a nearly 25% reduction after year 10 (RR, 0.75; 95% CI, 0.62–0.90, p=0.002). These findings were independent of age, stage, or menopausal status.

Table 1. Adjuvant tamoxifen trials testing treatment beyond 5 years.

F/U-follow up; mo-months; ATLAS-Adjuvant Tamoxifen: Longer Against Shorter; aTTom-Adjuvant Tamoxifen-To Offer More?; Tam-tamoxifen; ER+-estrogen receptor positive; NR-not reported; Yr-year.

Study	Number of patients (%)	Study Arm	Recurrence Rate, N (%)	p-value	Breast Cancer Mortality, N (%)	p-value	Overall Mortality, N (%)	p-value
ATLAS4, Total	12894, ER+ 6846	Tam Control	617 (21) 711 (25)	0.002	331 (12) 397 (15)	0.01	679 (10.5) 691 (10.7)	0.84
ER+ Premenopausal	326 (10) 304 (9)	Tam Control	64 (20) 73 (24)	0.79	NR		NR
ER+ Postmenopausal	3035 (89) 3044 (89)	Tam Control	553 (18) 638 (20)
aTTom26	6953, ER+ 2755	10 yr Tam 5 yr Tam	580 (28) 672 (32)	0.003	392 (NR) 443 (NR)	0.05	849 (NR) 910 (NR)	0.10
2013 Meta-Analysis27	29138, ER+ or unknown	Tam Control	828 (11) 1018 (13)	0.01	454 (10) 563 (12)	0.0003	823 (11) 911 (12)	0.03

Mirroring the ATLAS data, at 15 years following diagnosis the aTTom trial demonstrated a similar absolute reduction in breast cancer recurrence of approximately 4% (28% for 10 years of tamoxifen vs. 32% for 5 years of tamoxifen; RR 0.85; 95% CI, 0.76–0.95; p=0.003) with a modest reduction in breast cancer mortality and no difference in overall mortality. However, combined analyses of both trials showed that after year 10, extended tamoxifen significantly reduced breast cancer mortality (RR 0.75; 95% CI, 0.65–0.86; p=0.002) and overall survival (RR 0.84; 95% CI, 0.77–0.93; p=0.005)²⁶. A recent meta-analysis combining these data along with prior extended tamoxifen trials have confirmed these findings²⁷.

How Might Extended Therapy Be Applied to Premenopausal Women?

As seen in Table 1, fewer than 10% of patients in ATLAS were premenopausal at the time of randomization, which limits generalizations. However, there is a suggestion of a greater absolute reduction in recurrence rate with extended tamoxifen therapy in premenopausal women as compared to postmenopausal women (4.4 vs. 2.7%, chi-square p=0.79)⁴. In the NSABP B-14 study described above, analysis conducted by age (age <49 vs. >50) failed to demonstrate a difference in DFS between groups²⁵.

At present there are no validated assays available to estimate benefit from extended endocrine therapy. Recent data suggest that the Breast Cancer Index, an 11-gene assay that combines Molecular Grade Index to assess tumor grade and proliferation, the HOXB13/IL17BR index, and 4 reference genes, may be helpful for predicting risk of late recurrence, but this tool is not yet recommended for routine clinical use²⁸. Development of such a tool could optimize treatment-decision making, and lead to use of extended treatment in those patients at increased risk of delayed recurrence and to avoidance of treatment and its associated toxicity for those with minimal risk. Current guidelines recommend that patients who remain premenopausal or perimenopausal following five years of adjuvant tamoxifen should be offered extended therapy up to 10 years^13,21,29.

Historical Data Supporting Ovarian Suppression or Ovarian Ablation

Multiple historical trials comparing a variety of combinations of adjuvant therapies with OS/OA have been conducted³⁰. In 2007, 11906 ER+ premenopausal women were included in the LHRH-Agonists in Early Breast Cancer Overview meta-analysis³¹. When used alone, LHRH-agonists did not significantly reduce recurrence or breast cancer mortality. However, the addition of LHRH agonists to tamoxifen, chemotherapy, or both reduced breast cancer recurrence by 13% (95% CI, 2.4–21.9; p=0.02) and breast cancer mortality by 15% (95% CI 1.80–26.7; p=0.03) while LHRH agonists showed similar efficacy to chemotherapy alone.

Notably, this meta-analysis observed a differential benefit in those that were youngest (age <40). It was hypothesized that younger women were more likely to recover ovarian function following chemotherapy and therefore most likely to gain from OS/OA. Despite these signals of activity, a number of unanswered issues remained as the analyzed trials were largely conducted prior to the contemporary use of anthracyclines and taxanes (which are associated with lower rates of CIOF) and the standard use of tamoxifen and AIs.

Contemporary Trials of Ovarian Suppression

Based on the consistent but relatively small improvements in breast cancer recurrence for AIs over tamoxifen in the postmenopausal setting¹¹, the question of whether premenopausal women who undergo OS/OA would benefit from concurrent AI therapy compared to tamoxifen remained to be answered. Three large contemporary trials have reported the role of OS/OA in combination with tamoxifen or an AI in premenopausal women. The main outcomes of the Austrian Breast and Colorectal Study Group (ABCSG)-12^32,33, Suppression of Ovarian Function Trial (SOFT)⁵, and the Tamoxifen and Exemestane Trial (TEXT)/SOFT joint analysis⁶ are shown in Table 2.

Table 2. Clinical Trials Evaluating Ovarian Suppression in Combination or Sequential Therapy for Premenopausal ER+ Breast cancer.

F/U-follow up; ABCSG-Austrian Breast and Colorectal Study Group; SOFT-Suppression of Ovarian Function Trial; TEXT-Tamoxifen and Exemestane Trial; HR-hazard ratio; CI-confidence interval; NR-not reported; OFS-ovarian function suppression; Tam-tamoxifen; Anas-anastrazole; Exem-exemestane; LN-lymph node negative.

Trial	Patients	Median F/U	Treatment	Disease Free Survival (DFS)			Overall Survival (OS)
				DFS (%)	HR; 95% CI	p-value	OS (%)	HR; 95%CI	p-value
INT-014249	N=345; 100% LN-, 0% Chemo	9.9 years	Tam + OFS	89.7	1.17; 0.64–2.12	0.62	97.6	1.19; 0.53–2.65	0.67
			Tam	87.9			95.2
ABCSG-1233	N=1803; 70% LN-, 5% Chemo	62 months	Anas + OFS	89.3	1.08; 0.81–1.44	0.59	94.9	1.75; 1.08–2.83	0.02
			Tam + OFS	90.2			97
SOFT/TEXT6 Joint Analysis	N=4690; 58% LN-, 57% Chemo	68 months	Exem + OFS	91.1	0.72; 0.6–0.85	<0.001	95.9	1.14; 0.86–1.51	0.37
			Tam +OFS	87.3			96.9
SOFT5 (Entire cohort)	N=3066; 65% LN-, 53% Chemo	67 months	Exem +OFS	90.9	0.64; 0.49–0.83	NR	NR
			Tam + OFS	88.4	0.81; 0.63–1.03	0.09	96.7	0.74; 0.51–1.09	0.13
			Tam	86.4			95.1
SOFT (Prior Chemo)	N=1628 100% Chemo		Exem+ OFS	83.8	0.7; 0.53–0.92	NR	92.3	0.87; 0.59–1.27	NR
			Tam + OFS	80.7	0.82; 0.64–1.07	0.96	94.5	0.64; 0.42–0.96	0.03
			Tam	77.1			90.9
SOFT (<35 years)	N=233, 94% Chemo		Exem +OFS	83.4	NR		NR
			Tam + OFS	78.9	NR		NR
			Tam	67.7	NR		NR

The ABCSG-12 trial randomized 1803 HR+ premenopausal women with early stage breast cancer to goserelin plus tamoxifen or goserelin plus anastrozole with or without zoledronic acid. Most patients did not receive chemotherapy. At 62 months, there was no difference in DFS between those who received anastrozole versus tamoxifen (HR 1.08, 95% CI 0.81–1.44; p=0.591). Surprisingly, overall survival appeared worse in those randomized to anastrozole compared to tamoxifen (46 vs. 27 deaths; HR 1.75, 95% CI, 1.08–2.83; p=0.02). This finding initially suggested a possible distinct synergism between tamoxifen and the LHRH agonist compared to anastrozole. However, in an unplanned subgroup analysis it appeared that obesity was associated with a three-fold increase in the risk of death in women on anastrozole versus tamoxifen (HR, 3.03; 95% CI, 1.35 to 6.82; P = .004), possibly because of inadequate ovarian suppression by LHRH agonists in obese patients³⁴.

The specific design and treatment allocations for the SOFT and TEXT trials as well as the joint analysis are shown in Figure 1. In TEXT, all women received OS and were randomized 1:1 to either tamoxifen or exemestane. In contrast, SOFT consisted of three arms and patients were randomized 1:1:1 to tamoxifen alone or OS/OA with either tamoxifen or exemestane. Due to lower than expected event rates, a joint analysis of SOFT and TEXT was performed⁶. As shown in Table 2, there was a 3.8% absolute improvement in DFS (the primary end point in each trial as well as the joint analysis) in women receiving exemestane plus OS (E/OS) compared to tamoxifen plus OS (T/OS). As the primary goal of adjuvant therapy is to prevent distant recurrence (and thereby death), it is notable that there was only a statistically significant 1.8% difference in recurrence at a distant site (E/OS: 93.8% vs. T/OS: 92%; HR 0.78; 95% CI, 0.62–0.97; p<0.02) and no significant difference in overall survival (E/OS: 95.9% vs. T/OS: 96.9%).

Figure 1. Design and Treatment Allocation for SOFT, TEXT, and Joint Analysis.

SOFT-Suppression of Ovarian Function Trial; TEXT-Tamoxifen and Exemestane Trial; OFS-ovarian function suppression.

Based on subgroup analyses, it appeared that women who received chemotherapy were most likely to benefit from E/OS. Of the 57.4% that received chemotherapy, there was an improvement in distant recurrence-free survival in favor of E/OS compared to T/OS in TEXT (91.8 vs. 89.2%; HR 0.77; 95% CI, 0.56–1.06) and in SOFT (88 vs. 84.6%; HR 0.81; 95% CI, 0.58–1.13). Similarly, those with LN+ disease differentially benefited from E/OS compared to T/OS.

The primary analysis of DFS between tamoxifen or exemestane plus OS compared with tamoxifen alone in the SOFT trial was reported separately. As shown in Table 2, no significant difference in DFS or overall survival was observed between groups. However, two distinct subgroups did appear to derive benefit from OS and included those who received chemotherapy and those aged less than 35 years. The preplanned subgroup analysis of the women aged <35 revealed that OS plus tamoxifen yielded an approximately 11% absolute increase in DFS compared to tamoxifen alone. Exemestane plus OS was associated with an even greater absolute improvement in DFS (≈16%) compared with tamoxifen alone. Overall, the SOFT and TEXT data provide compelling evidence that AI or tamoxifen plus OS can significantly reduce the risk of recurrence compared to tamoxifen alone in premenopausal women at the highest risk for relapse (age <35 and those who maintain premenopausal status following chemotherapy).

Overall Survival and the Role of SOFT/TEXT in the Contemporary Era

In the SOFT and TEXT trials, no improvement in overall survival was seen with the addition of OS. However, given the few events observed in this cohort, the overall excellent prognosis (96–97% 5-year survival), and the long natural history of HR+ breast cancer, the absence of an overall survival difference is likely due to insufficient follow up and long-term results (10–15 years) are eagerly awaited. The worse overall survival with AI plus OS in ABCSG-12 has not yet been replicated in SOFT/TEXT; follow-up is ongoing.

Symptom Burden and Quality-of-Life During Endocrine Therapy

In order to appreciate the relative toxicities observed in SOFT/TEXT, it is important to consider the safety profiles of single agent tamoxifen and AI. Common adverse effects of endocrine therapies can be divided into 1) symptoms that are bothersome but impact QOL and compliance and 2) serious toxicities that can impact morbidity and mortality but do not necessarily cause symptoms. It is important to recognize and attempt to treat the bothersome symptoms, as early discontinuation of these medications is common, is often attributed to adverse effects, and likely leads to worse outcomes^35–37.

Tamoxifen

Subjectively, tamoxifen most often leads to vasomotor dysfunction such as hot flashes and night sweats as well as sexual dysfunction and menstrual irregularities. Vasomotor symptoms occur in >75% of women and are felt to be severe in up to 30%^38,39. Multiple pharmacologic agents (e.g. venlafaxine, citalopram, gabapentin) have been tested in randomized placebo-controlled trials that confirmed significant improvements over placebo and therefore these drugs should be considered first-line treatment in women with burdensome hot flashes⁴⁰.

Tamoxifen is also associated with rare but serious risk of venous thromboembolism (VTE) and uterine cancer. The risk of uterine cancer is increased with obesity and prior use of estrogen replacement therapy^4,26,41. Risk factors for VTE while on tamoxifen include older age, concurrent tobacco use, personal or family history of VTE, and the presence of the factor V Leiden mutation⁴².

Importantly, although the incidence of uterine cancer was higher in tamoxifen-treated patients, no significant difference was observed in women aged <45 (15 year incidence, 0.4 vs. 0.3%, p=0.97). Similarly, thromboembolism was observed in greater frequency in tamoxifen-treated patients versus control, all of which occurred in women aged 55–69. Taken together, these data support that young women (aged <45) treated with tamoxifen have the least risk for uterine cancer and thromboembolic disease and benefit similarly to older patients.

It is important to acknowledge that treatment with ten years of tamoxifen (as compared to five years) is associated with a further increase in uterine cancer and thromboembolic disease. In ATLAS, an increased cumulative risk of uterine cancer during years 5–14 was observed (incidence 3.1% vs. 1.6%, RR 1.74; 95% CI, 1.30–2.34; p=0.0002). This resulted in an absolute mortality increase of 0.2% in the extended therapy group. The incidence of pulmonary embolism was also increased (RR 1.87; 95% CI, 1.13–3.07; p=0.01). However, there was no increased risk of stroke and tamoxifen appeared to have a protective effect on ischemic heart disease. Similarly, an increase in uterine cancer was observed in aTTom with a slightly higher associated death rate (1.1% vs. 0.6%; p=0.02). These analyses were not separated by menopausal status or age so it is unknown if the risks are increased in premenopausal women.

Aromatase Inhibitors

AI therapy in postmenopausal women is associated with more pronounced vulvovaginal symptoms (vaginal dryness and dyspareunia) and can lead to greater sexual dysfunction compared to tamoxifen⁴³. Unique to this class of medications, the AI-associated musculoskeletal symptoms (AIMSS) can be a significant cause of morbidity in up to 25% and lead to early discontinuation⁴⁴. These symptoms can manifest in a variety of ways including generalized arthralgias, carpel tunnel syndrome, tendonitis, and myalgias. Greater bone loss (osteoporosis), more fractures, and increased cardiovascular morbidity have been observed in trials of AI compared to tamoxifen. In a meta-analysis of seven trials comparing AIs to tamoxifen in postmenopausal women, AIs significantly increased the risk of fractures (Odds Ratio [OR] 1.47; 95% CI, 1.34–1.61) and cardiovascular disease (OR 1.26; 95% CI, 1.10–1.43)⁴⁵. AI-associated bone loss is more pronounced in premenopausal women^6,33.

Strategies to minimize AI-associated bone loss involve promoting regular weight bearing exercise, limiting bone offending medications (e.g. corticosteroids) or behaviors (e.g. tobacco use), and maintaining adequate vitamin D and calcium intake. For pre- and postmenopausal women who develop osteoporosis based on a DEXA scan (T-score of −2.5 or lower) or are at high risk for fracture based on the World Health Organization Fracture Risk Assessment Tool (FRAX®), the use of bisphosphonates or denosumab is indicated⁴⁶.

Tamoxifen or AI plus Ovarian Suppression

In the SOFT/TEXT joint analysis⁶, significant toxicities were observed with both E/OS and T/OS, leading to high rates of hot flashes (91.7 vs. 93.3%), musculoskeletal symptoms (88.7 vs. 76%), fatigue (61.3 vs. 62.9%), insomnia (58.2 vs. 58.5%), depression (50.3 vs. 50.1%), and dyspareunia (30.5 vs. 25.8%). Acknowledging the limitations of cross-trial comparisons, compared to the symptom burden observed in the Arimidex, Tamoxifen, Alone or in Combination (ATAC) trial, patients treated on SOFT/TEXT had increased toxicity compared to AI or tamoxifen alone in older postmenopausal women. For example, in ATAC, anastrozole or tamoxifen alone, as compared to the observed toxicity in SOFT/TEXT, led to lower reported rates of hot flashes (30–32% with AI or tamoxifen alone in ATAC vs. 91–93% in SOFT/TEXT), fatigue (16–20% vs. 61–62%) and dyspareunia (7–17% vs. 25–30%). This increase in toxicity is most likely due to the ovarian suppression rather than to the endocrine therapy medications themselves. These symptoms can lead to psychological distress, depression, worse self-image, and poor emotional, physical, and functional wellbeing⁴⁷.

Changes in global QOL were similar between groups, although the individual symptoms differed between study cohorts. As expected, the E/OS group reported more vaginal dryness, greater loss of sexual interest, and bone/joint pain whereas the T/OS group reported more hot flashes. Not surprisingly, more women on E/OS discontinued therapy compared to T/OS (16 vs. 11%). This finding is especially troubling as it is well established that rates of early discontinuation are far higher outside of clinical trials⁴⁸.

In the SOFT analysis, as expected, T/OS was less well tolerated than tamoxifen alone with higher rates of grade 3 or 4 adverse events (31.3 vs. 23.7%)⁵. Consistent with this finding, INT 0142, a phase III study of T/OS compared to tamoxifen alone conducted in the United States, also observed significantly increased toxicity with combination therapy compared to tamoxifen alone (See Table 2 for details). Despite early closure which precluded sufficient power to detect a difference in DFS, the patient-reported outcomes data for this secondary analysis met its accrual goal confirming the findings of more menopausal symptoms, lower sexual activity, and inferior quality of life in the subset randomized to T/OS⁴⁹. Taken together, based on the lack of disease outcomes benefit combined with the increased adverse effects in SOFT, tamoxifen alone should remain standard of care for low-risk HR+ premenopausal women.

CONCLUSIONS

Over 30 years of well-designed randomized controlled trials in early stage breast cancer have led to markedly improved breast cancer outcomes. The recent reports of extended endocrine therapy and OS/OA further this progress and provide new therapeutic options for high-risk HR+ premenopausal early stage breast cancer.

How might we translate these findings to the clinic? Figure 2 provides an algorithm when considering adjuvant endocrine therapy in HR+ premenopausal women and is consistent with the 2015 St. Gallen International Expert Consensus²⁹. For high-risk patients, which can be defined as those aged <35 or those with sufficient risk to warrant treatment with chemotherapy and who remain premenopausal following treatment, AI plus OS is associated with a significant improvement in DFS over tamoxifen plus OS or tamoxifen alone. Given the increased mortality with surgical menopause at an early age⁵⁰, for very young women temporary ovarian suppression with a LHRH agonist should be considered unless there is a strong indication for permanent ovarian ablation (e.g. BRCA carrier). If OS is planned, a LHRH agonist can be given concurrently with or following chemotherapy as was done in TEXT and SOFT, respectively. Prior to initiating concurrent AI therapy, suppression of ovarian function should be confirmed biochemically using an ultrasensitive estradiol assay. It will be important to aggressively manage toxicity in order to maximize compliance with therapy, or to consider switch to an alternative therapy such as tamoxifen monotherapy if the combination regimen is detrimental to QOL.

Figure 2. Adjuvant Endocrine Options for HR+ Premenopausal Breast Cancer.

* To confirm postmenopausal state, use follicle-stimulating hormone and plasma estradiol level to confirm postmenopausal levels. If uncertain if definitely permanently postmenopausal at time of switch to AI, continue to monitor for breakthrough ovarian function. **Unless clear indication for permanent menopause (e.g. BRCA carrier), consider temporary ovarian suppression. Should confirm suppression of ovarian function biochemically. *** Concurrent OS was utilized in TEXT which appeared to have slightly improved distant recurrence-free survival compared to SOFT (see text). HR-hormone receptor; BC-breast cancer; AI-aromatase inhibitor; OS-ovarian suppression.

Following five years of such combination therapy, the role of extended duration endocrine therapy with or without OS/OA remains unknown. Potential short- and long-term toxicities cannot be understated and physical, sexual, and psychological symptoms should be aggressively treated or referred to appropriate services (psychiatry, counseling etc.). Long-term follow up will also be needed to clarify for whom the potentially increased cardiovascular and bone risks associated with AI plus OS outweigh the improvements in breast cancer outcomes.

For lower-risk HR+ premenopausal women, tamoxifen for 5–10 years or switch therapy is reasonable. It is noteworthy that fewer than one-fifth of participants in ATLAS had low-risk (LN− or < 2cm) HR+ breast cancers, which limits our ability to estimate the benefit of extended therapy in such patients. In low-risk patients, based on the SOFT data treatment with OS plus AI or tamoxifen may not be necessary; however, long-term follow up will be necessary to confirm the negative results.

Overall, the data from these recently reported trials highlight the excellent prognosis for many premenopausal women with breast cancer. For those with higher risk disease, it is now evident that there are numerous effective treatment options. However, it remains challenging to determine exactly which patients have high enough risk of recurrence to be offered ovarian suppression. In addition, optimization of therapy will require a continued dialogue between providers and patients, in order to maximize benefit while minimizing the negative impact on QOL for younger breast cancer survivors.

KEY POINTS.

1. Tamoxifen, a selective estrogen receptor modulator, is effective in both pre- and postmenopausal women.

2. Aromatase inhibitors can reactivate ovarian function women with chemotherapy-induced ovarian failure, including women over age 50. Therefore AI monotherapy should not be initiated unless postmenopausal status is confirmed biochemically. Monitoring of ovarian function during AI therapy is also recommended in patients who develop amenorrhea following chemotherapy.

KEY POINTS.

1. Unlike HR negative breast cancer, the risk of relapse in HR+ breast cancer persists >10–15 years after diagnosis.

2. Tamoxifen for 5 years compared to no therapy in HR+ breast cancer is associated with a nearly one-half (40–45%) relative reduction in recurrence and a one-third (25–30%) relative reduction in breast cancer mortality, independent of age, stage, menopausal status, and chemotherapy use.

3. Extended tamoxifen (10 versus 5 years) is associated with a relative reduction in breast cancer recurrence by nearly one-quarter (15–25%).