Breast cancer chemoprevention: A service in need of menopause specialist support

Introduction

In June 2013, NICE (National Institute of Health and Clinical Excellence) recommended women at moderate or high risk of breast cancer due to their family history, have opportunity to discuss the use of the Selective Oestrogen Receptor Modulators (SERMs) tamoxifen or raloxifene if risk-reducing surgery and surveillance was unsuitable or unacceptable. ¹ Despite the estimated significant impact in reducing the risk of breast cancer diagnosis in this population cohort, uptake and adherence to chemoprevention in routine clinical practice remain very low. ² An important impediment is patient and healthcare professional concern about treatment side effects, which are all associated with changes to estrogen activity. ² An added consideration, not fully explored by the 2013 NICE guidance and still unresolved, is which clinical specialities should have responsibility for counselling, prescribing and monitoring chemoprevention when neither tamoxifen nor raloxifene are licensed for this indication in the UK. GPs and clinicians working in family history or clinical genetics settings appear to have many issues relating to this. ³ Noticeable in its absence from the NICE guidance and debate surrounding improving implementation of chemoprevention is the potential role menopause specialists could have given they are best placed to advise about the management of estrogen-related side effects. This review summarises current recommendations surrounding breast cancer chemoprevention in the UK with the intention of highlighting aspects of care relevant to gynae-endocrine practice.

SERMS and breast cancer chemoprevention trials

In the 1980s, the results from large randomised adjuvant treatment trials of tamoxifen in women with early breast cancer reported a significant reduction (∼35%) in the risk of diagnosis of contralateral breast cancer. ⁴ The main adverse effects of an increased risk of diagnosis of endometrial cancer and thromboembolic events were uncommon and considered sufficiently low to support initiation of four placebo-controlled randomised tamoxifen chemoprevention trials in higher risk healthy women. ⁵ To determine whether it was a safer alternative to tamoxifen in the chemoprevention setting due to a known anti-estrogenic effect in the breast and lack of an estrogenic effect on the endometrium, breast cancer events were evaluated with exposure to raloxifene in placebo-controlled trials in postmenopausal women who had either osteoporosis, risk factors for coronary heart disease and in a direct head-to-head randomised chemoprevention trial with tamoxifen.^6–8 Initial overview of the main outcomes from these trials confirmed a significant reduction in the risk of diagnosis of estrogen receptor positive (ER+ve) but not estrogen receptor negative (ER−ve) breast cancer, and tamoxifen was confirmed to be equally effective in both pre and postmenopausal women. ⁵ Both tamoxifen and raloxifene were associated with an increased risk of venous thromboembolic events. ⁵ Subsequent meta-analysis of all randomised chemoprevention trials utilising SERMs showed the beneficial reduction in diagnosis of ER+ve breast during the active treatment phase (i.e. 5 years) continued following treatment cessation and that the risk reduction with tamoxifen in postmenopausal women was greater than that observed in women allocated to receive raloxifene (38% vs. 27%, respectively). ⁹ Longer term follow-up has shown the ‘carry-over’ benefit with tamoxifen to be associated with a reduction in the risk of breast cancer diagnosis for at least 20 years. ¹⁰ As most side effects stop with tamoxifen cessation, this has favourable implications for tamoxifen’s overall risk-benefit ratio. Other SERMs (i.e. lasofoxifene, arzoxifene) are still subject to evaluation as potential chemoprevention agents as are aromatase inhibitors. The latter are associated with a greater reduction in the risk of diagnosis of contralateral, ER+ve breast cancer risk compared with tamoxifen in adjuvant treatment trials but at the expense of arthralgia and increased fractures. ¹¹

The recommendations about chemoprevention in the 2013 NICE familial breast cancer guidance (NG164) was based on review of nine studies (two systemic reviews of randomised trials, five randomised controlled trials, one case series and one qualitative study) out of a total of 132 assessed to be relevant and of sufficient clinical quality using GRADE methodology and NICE checklists. The Guidance Development Group (GDG) concluded overall, the harm from iatrogenic side effects with tamoxifen and raloxifene (i.e. endometrial cancer, thromboembolic events, fractures and stroke) was outweighed by the potential benefit in reducing the risk of breast cancer diagnosis. ¹

Aromatase inhibitors were not recommended for chemoprevention as at the time of the NICE review, the only randomised study to differentiate between the different aromatase inhibitors was considered too low quality for clinical recommendations to be made despite a significant (65%) reduction in the diagnosis of ER+ve cancer with exposure to exemestame.^1,12 Subsequent to the NICE guidance, anastrozole has been shown to reduce the diagnosis of ER+ve breast cancer by 58% after a median follow-up of five years compared with placebo. ¹³ In both trials, women experienced significant vasomotor symptoms and arthralgia. Anastrozole was also associated with significant vaginal dryness, joint stiffness, carpel tunnel syndrome, hypertension and dry eyes. ¹³

NICE CG164: Recommendations for clinical practice ¹

The following summarises the key recommendations of the GCG:

Women at high^a and moderate^b risk are considered eligible for discussion of chemoprevention for a maximum of five year’s duration.

For high-risk women, the evidence of benefit is sufficiently strong to outweigh potential harms or side effects so ‘chemoprevention should be offered’.

^bWomen at moderate risk have between a 17% and 30% lifetime risk of breast cancer based on their family history. ¹

Presenting the risks and benefits from chemoprevention

NICE recommends discussion and provision of written information including absolute risk estimates associated with breast cancer and non-breast cancer outcomes to ensure appropriate individualised advice. However, neither the 2013 guidance nor a further up-date in August 2015 provided any such information, instead presenting risk ratios and percentage changes, which are open to misinterpretation.

1. Breast cancer diagnosis and breast cancer mortality

The absolute impact of tamoxifen and raloxifene on risk of breast cancer diagnosis is dependent on the baseline risk of breast cancer and the likelihood of developing ER+ve cancer. In BRCA2 mutation carriers and women at moderate risk, the majority of cancers (∼75%) diagnosed are ER+ve, whereas only a minority of cancers (∼25%) diagnosed in BRCA1 mutation carriers are ER−ve . ¹⁴ Table 1 summarises the potential absolute reduction in breast cancer diagnoses in these different risk groups by age if chemoprevention if taken for five years. The greatest impact will be in BRCA2 mutation carriers. For all risk groups, however, negative and positive framing demonstrates many will be exposed to risks and side effects without benefit. As yet, identification of women within these risk groups who would specifically benefit from current chemoprevention strategies is not possible. There are no interventions targeting ER−ve disease, which is of particular relevance for BRCA1 mutation carriers.

OPEN IN VIEWER

Table 1.

Absolute reduction in breast cancer diagnoses with five years’ treatment with tamoxifen.

Absolute reduction in the number of breast cancers diagnosed per 1000 women over 10 years (duration of chemoprevention five years) a
	Breast cancer risk		No tamoxifen		Tamoxifen		Absolute reduction in diagnosis with tamoxifen
	During the initial five years’ tamoxifen treatment, %	In five years after tamoxifen cessation, %	Negative framing (risk of diagnosis)	Positive framing (risk of no diagnosis)	Negative framing (risk of diagnosis)	Positive framing (risk of no diagnosis)
40–44 years
Moderate risk	2.1	3.1	52	948	41	959	−11
BRCA1	10	15	250	750	232	768	−18
BRCA2	10	10	200	800	155	845	−45
45–49 years
Moderate risk	3.1	2.7	58	942	45	955	−13
BRCA1	15	15	300	700	278	722	−22
BRCA2	10	10	200	800	155	845	−45
50–54 years
Moderate risk	2.7	3.3	60	940	47	953	−13
BRCA1	15	10	250	750	230	770	−20
BRCA2	10	15	200	800	155	845	−45
55–59 years
Moderate risk	3.3	4.0	73	927	57	943	−16
BRCA1	10	10	250	750	232	768	−18
BRCA2	15	15	250	750	197	803	−53

a

Estimates are based on following assumptions. (1) ∼75% of breast cancers diagnosed in women at moderate risk and BRCA2 mutation carriers are ER+ve and 25% of cancers in BRCA1 mutation carriers are ER−ve. (2) The risk reduction with five years’ tamoxifen treatment is 38% and the risk reduction in the five years following cessation of tamoxifen is 22%. (3) Five-year breast cancer risk for moderate risk women is based on a fourfold increase in baseline risk and calculated using the Gail risk model. ¹⁵

For women at high personal risk, irrespective of whether they are proven gene mutation carriers, NICE recommends discussion of chemoprevention and prophylactic surgery. ¹ Whereas chemoprevention reduces the risk of diagnosis by 38% clinical trial data about impact on specific or all-cause mortality is lacking, probably due to studies being underpowered to detect significant differences in such outcomes. It has been estimated tamoxifen may reduce breast cancer mortality by 18% ⁹ and using this estimate, absolute reduction in breast cancer deaths can be calculated although it must be emphasised this is not definitive evidence, only an assumption of effect (Table 2). In contrast, prophylactic surgery is associated with evidence of mortality reductions. Bilateral prophylactic mastectomy reduces the risk of breast cancer diagnosis by 85% to 100%, breast cancer mortality by 81% to 100% and prophylactic bilateral salphino-oophorectomy reduces the risk of breast cancer in premenopausal women by 37% to 100%, ovarian cancer by 69% to 100% and all-cause mortality by 55% to 100%. ¹⁶ For women estimated to be at moderate increased risk, prophylactic surgery is not routinely discussed but if it is requested by an individual, it is recommended multi-disciplinary agreement is sought before proceeding. The absolute risk reduction with chemoprevention is small (Table 1).

OPEN IN VIEWER

Table 2.

Estimated absolute reduction in breast cancer deaths with five years’ treatment with tamoxifen.

Absolute excess reduction in breast cancer deaths per 1000 women if tamoxifen taken for 5 years
	Moderate risk	BRCA1 mutation carriers	BRCA2 mutation carriers
40–44 years	−1.4	−2	−6
45–49 years	−1.8	−3	−6
50–54 years	−1.7	−3	−6
55–59 years	−2	−2	−7

Note: Estimates are based on the reduction in number of diagnoses of breast cancer in Table 1, and 10-year breast cancer survival in treated ER+ve breast cancer of 70% and treated ER−ve breast cancer of 60%. ⁹

2. Non-breast cancer side effects and outcomes with tamoxifen and raloxifene

Both tamoxifen and raloxifene have mixed estrogen agonist and antagonist activity. Their mutual benefit in reducing the diagnosis of ER+ve breast cancer is mediated via their anti-estrogenic activity but otherwise, they share some outcomes and differ in others (Table 3).^7–9 Furthermore, the activity of tamoxifen varies by menopausal status. ¹⁷ In postmenopausal women with the exception of its anti-estrogenic effect on the breast, tamoxifen has predominantly estrogenic activity but in premenopausal women anti-estrogenic effects prevail. Whilst the latter results in lack of an adverse effect on the endometrium or blood coagulation, bone density loss does occur and may not be regained following its cessation.^18–20 The recommendation of NICE that there be no lower age limit for commencing tamoxifen chemoprevention raises questions about whether bone density assessment should be considered in younger women who use it.

OPEN IN VIEWER

Table 3.

Tamoxifen and raloxifene; potential risks and benefits.

	Benefits	Risks	Null effect
Tamoxifen (premenopausal)	• Reduction in ER+ve breast cancer diagnosis	• Vasomotor symptoms • ? venous thrombosis • Bone loss (? future risk of fracture)	• Stroke • Endometrial cancer
Tamoxifen (postmenopausal)	• Reduction in ER+ve breast cancer diagnosis • Maintains bone density	• Vasomotor symptoms • Venous thrombosis • Endometrial cancer	• Stroke
Raloxifene (postmenopausal)	• Reduction in ER+ve breast cancer diagnosis • Reduction in osteoporotic fractures	• Vasomotor symptoms • Venous thrombosis • Vaginal dryness • Death from stroke (incidence not increased)	• Heart disease

Vasomotor symptoms and vaginal dryness are common side effects with both drugs. Hot flushes are worse with tamoxifen and vaginal dryness with raloxifene. ²¹ Randomised tamoxifen chemoprevention trials that permitted the concomitant use of hormone replacement therapy (HRT) for symptom relief have not reported an increased risk of breast cancer diagnosis but as the observed reduction was less compared with the use of tamoxifen alone, it would be counter-intuitive to prescribe HRT in this context.^10,22

Taking into consideration the outcomes of breast cancer, endometrial cancer, osteoporotic fracture and venous thromboembolism diagnoses, an approximate estimation of the overall number of associated health events can be made with exposure to tamoxifen and raloxifene. In premenopausal women, the absolute reduction in the number of health events with use of tamoxifen for five years is in favour of chemoprevention and there is little difference comparing women who may or may not have had a hysterectomy as endometrial cancer risk is not elevated (Table 4). In postmenopausal women, the favourable effect of tamoxifen is reduced in women with a uterus due to an increased risk of endometrial cancer, and in hysterectomised women, tamoxifen has a more favourable profile than raloxifene (Table 5). Although there is growing evidence for prolonged benefit of tamoxifen on reducing risk of breast cancer diagnosis, the assumption that the risk of adverse events falls following cessation does not account for the possibility, for example that bone loss in younger women may result in an increased fracture risk later in life.

OPEN IN VIEWER

Table 4.

Tamoxifen and absolute number of health events in pre and postmenopausal women.

Chemoprevention with tamoxifen use for five years: overall absolute risks (+) and benefits (−) per 1000 women a
	Moderate risk	BRCA1 mutation carriers	BRCA2 mutation carriers
Tamoxifen – intact uterus
40 to 44 years	−7	−15	−42
45 to 49 years	−9	−18	−41
Tamoxifen – hysterectomy
40 to 44 years	−7	−15	−42
45 to 49 years	−10	−19	−42

a

Estimates based on age-specific rates of diagnosis for breast cancer (Table 1), fractures, ²³ endometrial cancer ²⁴ and deep vein thrombosis ²⁵ and risk estimates for tamoxifen for these outcomes from chemoprevention trials.^9,17,18

OPEN IN VIEWER

Table 5.

Tamoxifen and raloxifene and absolute number of health events in postmenopausal women.

Chemoprevention with tamoxifen and raloxifene if taken for five years: overall absolute excess risks (+) and benefits (−) per 1000 women a
	Moderate risk	BRCA1 mutation carriers	BRCA2 mutation carriers
Tamoxifen – intact uterus
50 to 54 years	−2.8	−9.8	−34.8
55 to 59 years	+2.4	+4.4	−32.6
Tamoxifen – hysterectomy
50 to 54 years	−15.8	−22.8	−47.8
55 to 59 years	−19.6	−21.6	−56.6
Raloxifene
50 to 54 years	−10.7	−17.5	−31.8
55 to 59 years	−13.6	−14.3	−33.3

a

Estimates based on age-specific rates of diagnosis for breast cancer (Table 1), fractures, endometrial cancer and deep vein thrombosis ²⁵ risk estimates for tamoxifen for these outcomes from chemoprevention trials.^9,17,18

Patient uptake of and adherence to chemoprevention

Both uptake and treatment adherence need to be high for chemoprevention to impact on breast cancer diagnosis. NICE assumed their recommendation for chemoprevention would have a significant impact on clinical practice with an estimated an uptake of 25%. ¹ However, this has not been the case. In the UK and elsewhere, less than 10% of eligible women agree to do so. ² The most consistent reason identified appears related to unfavourable perception of risks and benefits and side effect management in the absence of access to HRT. ² Adherence data from women participating in clinical trials, where continuance rates are anticipated to be higher than in a non-clinical trial setting, show a decline over time. ² In only one study did more than 80% of women persist with treatment by five years. Poor treatment continuance is associated with use of tamoxifen rather than raloxifene, smoking, depression and older age. ²

Who should have responsibility for counselling, prescribing and monitoring chemoprevention?

It is recommended that only healthcare professionals within a specialist genetic clinic should discuss and provide written information about chemoprevention and its absolute risks and benefits to women at moderate and high familial risk. ¹ However, the current organisation of and demand for UK specialist genetics services means generally only patients at high familial risk are offered an appointment in such clinics. This raises the issue of where moderate risk women should be counselled and by whom. Even in the specialist clinic setting, clinicians admit to difficulty in interpreting the NICE guidance, particularly if their perception of chemoprevention benefit is low and in primary care, GPs appear to have less awareness of chemoprevention and the issues surrounding it use. ³ Regardless of where counselling takes place, the complexity of discussion required potentially will involve significant time commitment. At the time of publication of the NICE guidance, neither tamoxifen nor raloxifene had a UK license for chemoprevention of breast cancer and to date, they still do not. Therefore, the healthcare professional who discusses and prescribes chemoprevention has full responsibility for this and has to obtain documented informed consent. It is unsurprising therefore GPs appear reluctant to initiate prescriptions, preferring to continue a prescription originating in secondary or tertiary care. ³

Little clinical guidance was provided by NICE about monitoring treatment other than promoting attendance for breast cancer surveillance and referral of women in the event of development of post-menopausal bleeding. ¹ As GPs admit to a low level of awareness and knowledge about chemoprevention, it can be questioned whether follow-up in primary care is optimal as things stand. However, other options are limited in that specialist genetic clinics are not set up to provide on-going review of medication and breast clinics need to focus their increasingly limited resource on the diagnosis and treatment of women with cancer. Given the side effect profile of SERMS, pathways should be developed that utilise expertise of and support from menopause specialists.

Summary

There are still outstanding questions regarding the risk-benefit implications of current chemoprevention strategies for hormone-responsive breast cancer. Even if all were resolved, in clinical practice, initiation and monitoring treatment requires ability of involved health professionals to provide reliable information about breast and non-breast cancer outcomes and strategies for their management as appropriate. Most issues are related to changes in estrogen activity and are beyond the expertise of many GPs, health professionals in family history or clinical genetics clinics or surgeons in breast clinics and require menopause expertise. Given the recent NICE menopause guidance recognised the importance of menopause specialists in the management of symptomatic breast cancer patients, ²⁶ it is hoped that similar recommendation will be made in the next familial breast cancer guidance up-date, which is scheduled for publication in March 2017. If not, health professionals involved in providing gynae-endocrine services should take the initiative. The author will provide a summary of the up-dated NICE guidance for the Journal of Post Reproductive Health following its publication.