Methods: We studied the interaction between HRT and tamoxifen on serum cholesterol, fibrinogen, antithrombin HI. (AT HI) and bone mineral density (BMD) in ...
Annals of Oncology 7: 671-675, 1996. © 1996 Kluwer Academic Publishers. Printed in the Netherlands.
Original article The efFect of tamoxifen and hormone replacement therapy on serum cholesterol, bone mineral density and coagulation factors in healthy postmenopausal women participating in a randomised, controlled tamoxifen prevention study J. Chang, T. J. Powles, S. E. Ashley, R. K. Gregory, V. A. Tidy, J. G. Treleaven & R. Singh Breast Unit, Royal Marsden Hospital, Sutton, Surrey, UK
Summary
Significant reductions of plasma fibrinogen by 14% and AT HI by 8% were seen in women who received tamoxifen. Background: The role of hormone replacement therapy There were no further significant changes in these coagula(HRT) in women who have been treated for breast cancer re- tion factors in women on tamoxifen/HRT combinations. mains controversial. The addition of tamoxifen may protect Tamoxifen resulted in an annual increase in BMD of the these women from any proliferative effect of exogenous femur and spine by 2% and 1.5%, respectively, when comoestrogen on the breast. The aim of this analysis was to deter- pared to placebo (n - 38). The addition of HRT to tamoxifen mine if tamoxifen and HRT may be safely administered to- resulted in further 2% annual increase in BMD of the femur. gether. Conclusions: We conclude that there were no significant Methods: We studied the interaction between HRT and adverse intereactions with tamoxifen and HRT in this small tamoxifen on serum cholesterol, fibrinogen, antithrombin HI series of patients. The combination results in a reduction of (AT HI) and bone mineral density (BMD) in postmenopausal serum cholesterol, increase in BMD especially in the femur healthy women enrolled in a randomised tamoxifen chemo- and appears not to have any adverse effect on coagulation prevention trial. factors. Multicentre studies should be conducted to evaluate Results: Tamoxifen decreased serum cholesterol by a the effect of this combination on relieving menopausal sympmean of 13% from pretreatment values (n — 153). The addi- toms, disease relapse and overall survival in women who have tion of HRT to tamoxifen did not result in further reduction received treatment for breast cancer. in serum cholesterol (n •» 20). HRT alone led to a reduction of serum cholesterol by a mean of 5% in 14 women on placebo. The addition of tamoxifen to HRT resulted in further Key words: antithrombin HI, bone mineral density, cholesreduction in serum cholesterol by a mean of 7% (n — 44). terol, fibrinogen, HRT, tamoxifen
Introduction
Postmenopausal women may experience a variety of debilitating menopausal symptoms: hot flushes, dyspareunia, atrophic vaginitis, sleep and mood disturbances. Hormone replacement therapy (HRT) ameliorates these symptoms and decreases morbidity and mortality from coronary heart disease and osteoporosis [1]. The absolute prohibition of HRT in postmenopausal breast cancer survivors may diminish their quality of life. Of greater concern is the induction of early menopause in women who have received adjuvant chemotherapy. Premature ovarian failure is both age and drug dependent With standard adjuvant CMF chemotherapy (cyclophosphamide, methotrexate and 5-FU), more than 80% of women aged over 35 years will become amenorrheic and in the majority, this would translate into permanent ovarian failure [2]. Premature menopause carries the attendant risks of heart disease and osteoporosis and these women may survive
breast cancer only to succumb to other afflictions at a later stage. The concern of prescribing HRT to breast cancer survivors is that oestrogen replacement may either induce new primary breast cancers or result in earlier disease relapse. The association between HRT and increased breast cancer incidence remains controversial [3]. Meta-analyses of breast cancer incidence and HRT do not collectively show a consistent increased risk of breast cancer in women who have ever used HRT [46]. Importantly, the use of low dose conjugated HRT (0.625 mg/day) does not appear to appreciably increase the risk of breast cancer [4]. Although in the general population HRT does not increase the incidence of breast cancer significantly, it may be possible that women with a personal history of breast cancer are more susceptible to any tumour promoting effects of HRT. Adjuvant tamoxifen reduces the incidence of contralateral breast cancer by 40% to 50% [7] and it thus may be possible that tamoxifen, when prescribed
672 together with HRT may eliminate or reduce any potential cancer promoting effect of oestrogen on the breast [8]. Tamoxifen has both oestrogenic and anti-oestrogenic properties. It reduces the incidence of fatal myocardial infarction [9], lowers serum cholesterol [10, 11], increases bone mineral density in postmenopausal women by virtue of its oestrogenic properties [12, 13]. The effects on the lipid profile and BMD of combined tamoxifen and HRT are unknown. Tamoxifen is reported to slightly increase the risk of venous thrombosis in women with breast cancer [7]. HRT has not been shown epidemiologically to increase the risk of thromboembolic events [14]. Thus, HRT and tamoxifen combinations would not be expected to exacerbate hypercoaguability but monitoring of coagulation factors would be important in these women. The aim of this analysis was to determine if HRT could be safely administered with tamoxifen. We studied the interaction between HRT and tamoxifen or placebo in healthy postmenopausal women at risk of breast cancer, with particular attention to changes in serum cholesterol, clotting factors (plasma fibrinogen, antithrombin m) and bone mineral density. Any adverse interaction, if found, would need to be addressed before clinical trials of HRT and tamoxifen are commenced in women with breast cancer. Methods Subjects Since 1986, 2450 healthy women at increased risk of breast cancer have been recruited into a double blind, placebo controlled prevention study of tamoxifen versus placebo; details of this study design have been previously reported [15, 16). This trial has ethical approval by the Hospital Ethical Committee and all participants have given written informed consent Participants were prescribed 'Tamoplac' by the physician and randomised to receive either tamoxifen or an identical placebo 20 mg orally (Farmos Ltd, Finland). The planned duration of medication was 8 years. Women with a past history of venous thrombosis, any previous malignancy, or an estimated life expectancy of less than 10 years were excluded from this trial. The analysis for interaction between HRT and tamoxifen/placebo was restricted to postmenopausal women. Postmenopausal status was defined as the cessation of menstrua] cycles for at least 12 months or age of more than 54 years if they had previously undergone hysterectomy. Natural conjugated oestrogens (Prempak C or Premarin 0.625 mgs) or Estraderm patches 25 micrograms were the most commonly prescribed forms of HRT. Measurements for total serum cholesterol, plasma fibrinogen and antithrombin Dl (AT HI) were obtained at time of randomisation, at 6 months and annually thereafter. Bone mineral density (BMD) of the total hip and lumbar spine L1-L4 were measured at randomisation and annually thereafter. Women who had received HRT for at least 6 months prior to randomisation and women who had been treated with tamoxifen/placebo for at least 6 months prior to starting HRT were included in this analysis. Repeat measurements of serum cholesterol, clotting factors and BMD were performed after at least 6 months of HRT and tamoxifen/placebo combination.
Cholesterol
Total serum cholesterol was measured using a fully enzymatic oxidase-aminophenazone method. This measurements were performed on non-fasting serum as fasting status has not been shown to influence results significantly [17].
Clotting factors
Plasma fibrinogen and AT III was analysed using photometric assays (Boehringer Mannheim, Kit 524484 for fibrinogen; Kit 759376 for AT in). The methodologies and principles were described for the fibrinogen and AT HI assays by Hesse and Becker et al., respectively [18,19).
Bone mineral density
Bone mineral density (g/cm2) of the femur and lumbar spine L1-L4 was performed using dual energy X-ray absorptiometry (DXA; Hologic QDR 1000 densitometer). Although BMD measurements were made by several observers, reanalysis of randomly selected subgroup of scans gave a correlation coefficient of greater than 0.999 [13]. Statistical analysis Baseline readings for serum cholesterol, fibrinogen, AT III and BMD were performed within 3 months of starting tamoxifen/placebo or HRT. Repeat measurements were conducted at a variable time at least 6 months after commencement of treatment. Changes in values were calculated as a percentage of the pretreatment reading. Statistical comparisons with baseline values were made with paired student Mest. An overall measure of change in BMD was derived by calculating the area under the curve and expressed as a mean annual % change.
Results The median time of follow-up of this analysis was 4 years. Repeat measurements for serum cholesterol and clotting factors were performed at a median of 6 months from baseline readings (range 6 to 36 months). Repeat measurements for BMD were conducted at a median of 11 months (range 9 to 27 months). There were 6 categories of participants in this analysis: women who were treated with tamoxifen (group A) or placebo (group B); women in whom HRT was subsequently added to tamoxifen (group C) or placebo (group D); women who were on HRT before randomisation to tamoxifen (group E) or placebo (group F). Cholesterol There was a significant decrease in serum cholesterol by a mean of 13% from pretreatment values (P < 0.001) in 153 women receiving tamoxifen (group A) and a small mean reduction of 2% in 149 women on placebo (group B, P < 0.05). This lowered cholesterol level was maintained with no further reduction in 20 women on tamoxifen who were subsequently treated with HRT (group C). HRT alone led to a smaller reduc-
673 tion of 5% in serum cholesterol (group D, P = 0.1). The addition of tamoxifen to HRT further significantly lowered the cholesterol level by a mean of 7% in 44 women (group E, P < 0.02). These results are summarised in Table 1 and Figure 1.
significant changes in women receiving HRT or tamoxifen and HRT combinations (groups B,C, D, E and F). These results are summarised in Table 1 and Figure 3. There have been no reported cases of deep venous thrombosis or pulmonary emboli in any participant in this chemoprevention study.
Clotting factors Bone mineral density Significant reduction in plasma fibrinogen by 14% of pretreatment level was noted in 90 women receiving tamoxifen (group A, P < 0.001). There were no significant changes in fibrinogen values in women receiving HRT or tamoxifen and HRT combinations (groups B, C, D, E and F). These results are summarised in Table 1 and Figure 2. Significant reduction of AT in by 8% of pretreatment values was observed in 93 women receiving tamoxifen (group A, P< 0.001). There were no other
There was a significant annual increase in BMD in the femur by 2% and in the spine by 1% in women receiving 38 tamoxifen (group A; P < 0.001 and P < 0.1, respectively). There was a small annual decrease in BMD in both the femur and spine in 26 women receiving placebo (group B). Therefore, the net annual increase in BMD in postmenopausal women on tamoxifen when compared to placebo was 2% and 1.5% in the femur and spine, respectively. Ten women on tamoxifen who were started on HRT (group C) had a further 2% annual increase in BMD of Table 1. Effect of tamoxifen and HRT on serum cholesterol, fithe femur (P < 0.05) but no additional effect on the brinogen, AT HI, BMD. spine was observed (group C). The number of women on placebo (n = 7) who subsequently received HRT CholesFibrinoAT ID Bone mineral density terol (N), gen (N), (N), were insufficient to analyse the effects of HRT on BMD P value /"value P value Spine (N), Femur (group D). Likewise, the number of participants (n = 5) P value (N) P value Group A
-13% (153) P
