Acceptability and patterns of uterine bleeding in sequential trimegestone-based hormone replacement therapy: a dose-ranging study

Farook Al-Azzawi1,5, May Wahab1, John Thompson2, Malcolm Whitehead3 and William Thompson4

1 Gynaecology Research Group, Department of Obstetrics and Gynaecology, and 2 Epidemiology, Robert Kilpatrick Clinical Sciences Building, Leicester Royal Infirmary, Leicester LE2 7LX, and Department of Obstetrics and Gynaecology, 3 King's College Hospital, London and 4 Queen's University, Belfast, UK


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Trimegestone is a norpregnane progestogen which is being developed in combination with oral oestradiol as postmenopausal hormone replacement therapy (HRT). In this multicentre dose-ranging study using randomized parallel groups, four doses of trimegestone were used to compare data on the patterns of uterine bleeding, the endometrial histology, and the control of menopausal symptoms in 203 women who completed treatment for 6 months. The treatment consisted of micronized oestradiol (2 mg/day) and one of four doses of trimegestone, which was administered sequentially for days 15–28 of the treatment cycle. Higher doses of trimegestone were associated with later onset of bleeding, which was lighter and of shorter duration than that observed with lower doses. The variability of the day of onset of bleeding in individual women was greater when bleeding occurred before the end of the progestogen phase (early bleeders) than when it occurred afterwards (late bleeders). All women enrolled in the study experienced good control of menopausal symptoms, with minimal progestogenic adverse effects, there being no statistically significant difference between the four dose groups.

Key words: bleeding/endometrium/HRT/menopause/trimegestone


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Oestrogen replacement therapy relieves symptoms of oestrogen deficiency, prevents bone loss and osteoporotic fractures, and confers protection against atherosclerotic cardiovascular disease (Akkad et al., 1996Go; Daly et al., 1996Go). A progestogen is needed to protect the endometrium against hyperplasia and carcinoma, though when added sequentially to the oestrogen this usually results in withdrawal bleeding. This may dissuade women from continuing with the treatment, particularly if the bleeding is irregular or heavy (Al-Azzawi and Habiba, 1994Go). Indeed, up to 50% of women stop hormone replacement therapy (HRT) because of bleeding problems (Nachtigall, 1990Go).

Progestogens can also cause premenstrual tension (PMT) symptoms, such as irritability, bloatedness, fluid retention and cyclical mastalgia. These PMT symptoms appear dose-related (Magos et al., 1986Go). However, optimizing the progestogen dose to below minimum effective dose may jeopardize endometrial protection and may also cause poor cycle control. Because the duration of progestogen administration appears to be an important factor in protecting the endometrium (Sturdee et al., 1978Go; Whitehead, 1978Go), shortening the exposure to the progestogen may increase the risk of hyperplasia and cancer.

Trimegestone is a novel norpregnane progestogen which in binding studies with human recombinant steroid receptor has been shown to have potent progesterone and a very low androgen receptor affinity, and no detectable affinity for the oestrogen receptor (Bouchoux et al., 1995Go). Such a profile makes this progestogen suitable for evaluation in combination with oestradiol in HRT. We now report a dose-ranging study of four doses of trimegestone added sequentially to oral oestradiol. Data were collected on the patterns of vaginal bleeding, the endometrial histology, and the control of menopausal symptoms in women receiving treatment over 6 months.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
This was a prospective, multicentre study involving a parallel-group, double-blind, randomized design. The protocol was approved by all local ethics committees, and all patients provided their signed, informed consent.

We recruited apparently healthy women, aged 45–65 years, and with an intact uterus who were at least 6 months postmenopausal. All had either taken HRT for more than 2 years, had had HRT for at least 1 year with pretreatment follicle-stimulating hormone (FSH) and oestradiol concentrations within the postmenopausal range, or had been amenorrhoeic for at least 6 months with FSH and oestradiol concentrations within the menopausal range. Patients who had previous use of an oestradiol implant were excluded from the study. Tests of liver and renal function were performed and those women with abnormalities were excluded. All women aged over 50 years had to have undergone mammography within 3 years before entry, and all women were required have had a normal cervical smear within 6 months before entry. At baseline, general, breast and pelvic examinations were required to be normal. An endometrial biopsy was obtained using the Vabra curette at the screening visit and fixed in formol saline. Patients were excluded if the histology showed any evidence of hyperplasia or cancer.

During the study, additional sex steroid treatments were not allowed and women who used treatments known to interfere with steroid metabolism were withdrawn. At the final visit, general, breast and pelvic examinations were repeated and a further endometrial biopsy was obtained using the Vabra curette on day 23 or 24 of the last treatment cycle and again fixed in formol saline. If the biopsy material at this visit was inadequate for histological assessment, then hysteroscopy under local anaesthesia was performed to confirm that the endometrial surface appeared normal and atrophic. All endometrial biopsies were evaluated by two independent pathologists (Dr J.Pryse-Davies and Dr C.Bergeron) who were blinded to the trimegestone dose.

Treatment
Allocation to treatment which lasted for six 28-day cycles was random. Each cycle consisted of oral micronized oestradiol (RU 3499: Hoechst Marion Roussel, Romainville, France) 2 mg/day for 28 days together with trimegestone tablets (Hoechst Marion Roussel) (either 0.05, 0.1, 0.25 or 0.5 mg per day) for days 15–28 of each cycle. The first visit was a pre-study assessment and took place 2–3 weeks before commencement of the 168-day treatment. All participants were assessed clinically at the end of the third cycle and then on the 23rd or 24th day of the sixth treatment cycle. Failure to attend on that day meant that a seventh cycle of treatment was prescribed.

Assessment of symptom control
Women completed the Greene Climacteric Scale (Greene, 1976Go) and answered questions about breast tenderness, bloating, nausea, abdominal cramps, irritability, leg cramps, acne, seborrhoea and headaches. The scales were completed at baseline and the end of the third and sixth treatment cycles.

Daily diaries
`Bleeding' was defined as vaginal blood loss requiring a sanitary pad or tampon: `spotting' was defined as that which did not require sanitary protection. We defined the first day of the oestrogen phase as a fixed reference time, `day 1'. Diaries were kept for each treatment cycle; bleeding was scored as 0 = no bleeding, 1 = spotting, 2 = slight bleeding, 3 = moderate bleeding and 4 = heavy bleeding. The total bleeding score was the sum of the daily scores.

Definitions
The following definitions were applied: a `bleeding episode' was defined as bleeding for one or more days with at least two bleed-free days before and after; progestogen-associated bleeding was vaginal bleeding with any day of that episode starting between day 22 of one cycle and day 7 of the next cycle, inclusive (all other bleeding was defined as intermenstrual bleeding); the `bleeding interval' was the number of days between the first day of bleeding of two consecutive progestogen-associated bleeding episodes.

At the end of the third and sixth treatment cycles, the diaries were collected and checked for accuracy and completeness. They also served as an additional check of compliance because a box had to be ticked when every tablet was taken. Used medication packs were collected and all remaining tablets were counted. If <80% of the medications had been used, then the patient's data were excluded from the analyses. All adverse events and concomitant medications were also recorded. Those cycles in which data were incomplete (incomplete diaries) were excluded from the bleeding analysis. Bleeding data from the cycles in which the endometrial biopsy was performed were also excluded.

Statistical analysis
Data were collected from up to seven cycles for each woman and these repeated measures were analysed using mixed models. For continuous data, the residual maximum likelihood (REML) was used (Patterson and Thompson, 1971Go) to fit the model. For binary data, a generalized linear mixed model with binomial errors was used (Brestlow and Clayton, 1993Go). Calculations were made using GENSTAT (GENSTAT 5 Committee, 1993Go). Patient data that were not collected separately for each cycle were analysed using analysis of variance (ANOVA) or the chi-square test. The transition probabilities in Table VGo were calculated by the method of Harlow and Zeger (1991).


View this table:
[in this window]
[in a new window]
 
Table V. Percentage probability of the day of onset of the progestogen-associated bleeding (PAB) in the current cycle predicting this in the next cycle. All doses of trimegestone combined
 

    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Between November 1994 and December 1995, 266 women were randomized to one of the four trimegestone dose groups. Of these women, 203 completed the study and 202 of these kept diaries. A total of 195 women provided endometrial tissue sufficient for reliable histological examination. There were no statistically significant differences between the four dose groups in terms of age, height, weight, body mass index (BMI), time since menopause, previous use of HRT or smoking habits (data not shown).

The number of women starting and continuing treatment at each dose is shown in Table IGo; there was no statistically significant difference between the four dose groups. Table IIGo lists the number of women reporting adverse events and severe adverse events, and the number in whom study medication was discontinued. It is possible for the same patient to appear in all three categories, e.g. in the 0.05 mg/day group, 36 women reported breast pain as an adverse event; six described it as severe and five withdrew as a result. There were no statistically significant differences between the numbers of patients from each trimegestone dose group for adverse events, severe adverse events or discontinuation of medication. The following adverse events occurred and led to the discontinuation of study medications. Two cases of thrombophlebitis were reported, one in each of the 0.05 mg and 0.5 mg trimegestone groups. One case each of raised blood pressure, allergic reaction and weight loss was reported in the 0.05, 0.1, 0.25 and 0.5 mg dose groups. Irritability and depression were contributory factors in the withdrawal of seven women, headache in eight and gastrointestinal symptoms in 12, though none of these adverse events was related specifically to the trimegestone phase of treatment. One case of breast carcinoma was diagnosed during the study and thus the woman was withdrawn; a subsequent review revealed that the cancer had been present but not identified in the pretreatment mammogram.


View this table:
[in this window]
[in a new window]
 
Table I. Numbers of women starting and continuing treatment, by dose of trimegestone (mg) and at defined times
 

View this table:
[in this window]
[in a new window]
 
Table II. Numbers of women reporting adverse events (AER), severe adverse events (SAE), and in whom study medication was withdrawn (AED), by dose of trimegestone
 
Direct questioning on specific symptoms that might have been related to progestogen administration, namely mastalgia, acne, nausea, irritability, leg cramps, headaches, abdominal cramps, seborrhoea and bloatedness, revealed no statistically significant differences between the four dose groups

Effects on climacteric symptoms
There was a significant improvement in the anxiety, depression, somatic, vasomotor and sexual components of the Greene Climacteric Scale in all treatment groups (ANOVA, P < 0.01) after 3 months of treatment, which was maintained at the end of the study. There was no statistically significant difference in the improvement of scale components between those women who did not have HRT before starting the study and those who did; neither was there any statistically significant difference between the treatment groups (data not shown).

Frequency of progestogen-associated- and intermenstrual bleeding and amenorrhoeapa
These data are presented in Table IIIa (women) and IIIbGo (cycles). Data are also included in these tables on the number of cycles and women with missing data due to an incomplete diary. It should be noted that data for all women in the last treatment cycle were excluded because of the on-treatment biopsy. There were no significant differences between the four doses for any of these variables.


View this table:
[in this window]
[in a new window]
 
Table III. (a) Number of women and (b) number of cycles with biopsies, missing data, progestogen-associated bleeding (PAB), intermenstrual bleeding (IMB), PAB + IMB and amenorrhoea in subjects completing the study
 
Length of bleeding interval and day of onset of bleeding
Data on the mean length of bleeding interval by dose of trimegestone and at defined times are shown in Table IVaGo. There were no statistically significant differences between treatment groups (either for all women who completed the study or for all women, including those who dropped out or were withdrawn).


View this table:
[in this window]
[in a new window]
 
Table IV. (a) Mean length of bleeding interval (days) and (b) mean day of onset of progestogen-associated bleeding (PAB), at defined times and by dose of trimegestone
 
The mean day of onset of progestogen-associated bleeding occurred progressively later as the dose of trimegestone increased (P < 0.0001) (Table IVbGo). The effects of age, time since menopause, bodyweight, BMI, smoking and dose of trimegestone on the mean day of onset of progestogen-associated bleeding was also examined. Only the dose of trimegestone was a significant determinant of the differences between the groups (P < 0.0001). There was no change over time for either the mean length of the bleeding interval or mean day of onset of progestogen-associated bleeding within each dose group in those women who bled. Analysis of the diaries on an intention-to-treat basis, including partial information from women who discontinued the study, showed similar results.

The bleeding pattern in individual women
The mean (range) day of onset of bleeding in relation to the trimegestone dose for each woman is shown in Figure 1Go. The mean day of onset of progestogen-associated bleeding occurred progressively later with higher doses (P = 0.0001). These data suggest that the variability in the day of onset of bleeding is reduced, with a mean day of bleeding onset of day 29 or later. This was more noticeable with the highest dose of trimegestone as more women in this dose group bled on or after day 29. The within-subject variance of the day of onset of progestogen-associated bleeding for women whose day of onset was <29 days was significantly greater than that of women whose mean day of onset was >=29 days (P <0.0001).



View larger version (27K):
[in this window]
[in a new window]
 
Figure 1. Mean and range of day of onset of bleeding by trimegestone dose. A line drawn at day 29 separates `early' bleeders with wide variability of the day of onset of progestogen-associated bleeding (PAB) from `late' bleeders with narrow variability.

 
Bleeding on or before day 28
The percentages of women who had complete diaries and who had at least one progestogen-associated bleeding episode starting on or before day 28 were 94%, 89%, 52% and 15% with respect to increasing doses of trimegestone. Similarly, 85%, 67%, 32% and 8% of all cycles started on or before day 28. These differences were highly significant (P < 0.0001), the result being the same when the data were analysed on an intention-to-treat basis.

Prediction of the day of onset of progestogen-associated bleeding
An investigation was made to determine whether knowing the day of onset of progestogen-associated bleeding in any cycle might predict the day of onset of bleeding in the next cycle (Table VGo). For example, a day of onset of <=28 days in the current cycle was associated with a 74% chance of this day of onset of progestogen-associated bleeding occurring in the next cycle. Absence of such bleeding in the current cycle was associated with a 15% chance of no progestogen-associated bleeding in the next cycle.

Number of days of progestogen-associated- and intermenstrual bleeding per cycle
The duration of progestogen-associated bleeding (in cycles with such bleeding and excluding those cycles without) was 7.7, 7.2, 5.8 and 4.9 days with trimegestone dose levels of 0.05, 0.1, 0.25 and 0.5 mg, respectively. These differences were statistically highly significant (P <0.0001). Comparable data for the number of days of intermenstrual bleeding (in cycles with such bleeding and excluding those cycles without) were 3.4, 2.7, 1.5 and 3.4, respectively.

Total bleeding score
The severity of vaginal bleeding, expressed as a total bleeding score, progressively declined both with the increase in trimegestone dose (P < 0.0001) and over time (P < 0.0001) (Table VIGo).


View this table:
[in this window]
[in a new window]
 
Table VI. Mean of total bleeding score during progestogen-associated bleeding (PAB) by trimegestone dose and at defined times during the study
 
Endometrial histology
The results of the histological assessment of the endometrium at the end of the study are shown in Table VIIGo, in relation to the trimegestone dose. There were no statistically significant differences between treatment groups. One sample from the 0.5 mg trimegestone group showed simple hyperplasia with secretory features, while three proliferative endometria were obtained from one patient in each of the other three dose groups.


View this table:
[in this window]
[in a new window]
 
Table VII. Histological findings at the end of the study by dose of trimegestone
 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
This dose-ranging study was designed to obtain efficacy and safety data with a sequential combination of trimegestone and micronized oestradiol. The efficacy data will be published separately: in summary, climacteric symptoms were relieved and the incidence of androgenic and progestogenic adverse effects was low.

The endometrial safety data were reassuring, with only one biopsy showing simple hyperplasia. This is below the incidence seen in an untreated population of postmenopausal women. The 98% procurement rate of adequate endometrial biopsies for histological analysis compared favourably with the published data on Vabra curettage (Grimes, 1982Go) and the sampling rate in large multicentre studies (Sturdee et al., 1994Go). There were only four patients from whom an endometrial biopsy could not be obtained, and in all four the uterine cavity was shown by hysteroscopy to be atrophic. There were no statistically significant differences between the histological results from the four dose groups. The finding of atrophic and/or inactive endometrium in a woman experiencing regular withdrawal bleeding is intriguing, and either reflects regional differences within the endometrium or suggests a different mechanism for bleeding as compared with that seen in women with spontaneous postmenopausal bleeding in the absence of a specific pathology.

There was no clear dose response with respect to secretory changes in endometrial histology (Table VIIGo). Therefore, one interpretation of the data would be to recommend the 0.05 mg/day dose of trimegestone because this is the lowest effective dose in the present study which provided endometrial safety. However, data on the patterns of bleeding might lead to a different conclusion.

As the trimegestone dose increased, the mean day of onset of progestogen-associated bleeding occurred later (Table IVbGo), and the variability in the onset of such bleeding was also reduced (Figure 1Go). This pattern, with `early' bleeders having a greater variability in the length of bleeding interval compared with `late' bleeders, confirms our previous report of 103 women using oral oestradiol and norethisterone (Habiba et al., 1996Go). The current study also shows that both the duration and severity of progestogen-associated bleeding (Table VIGo) are reduced as the trimegestone dose was increased. The severity of bleeding was also reduced over time.

This interesting association of a later onset and shorter duration of bleeding suggests that a fuller differentiation of the endometrium before shedding occurs with higher progestogen doses. `Early-onset' bleeding and the association with a longer duration of blood loss may reflect regional differences within the uterus in maturation and shedding of the endometrium with a lower dose of progestogen.

The chance of transition from a particular day of onset of bleeding into another pattern was calculated. Women who bleed while still in the progestogen phase had a higher incidence of early bleeding in the next cycle; conversely `later-onset' bleeding in one treatment cycle was more likely to be associated with late-onset bleeding in the next cycle (Table VGo). The finding of a later onset of bleeding in the treatment cycle being associated with a more predictable pattern of bleeding is in accordance with our previous report (Habiba et al., 1996Go), and suggests that this method of predicting the day of onset of bleeding might be used both by the clinician and patient to anticipate the pattern of bleeding when using a particular preparation of HRT.

The increased number of episodes of intermenstrual bleeding, which appeared to be of shorter duration as the dose of trimegestone increased, may be due to the progressive delay in the onset of bleeding with higher doses, which may include an extension of the bleeding outside the progestogen-associated bleeding window. There is a tendency for more bleeding days in the intermenstrual bleeding window in the 0.5 mg dose group, which may be due to the progressively later onset of bleeding. However, the frequency of intermenstrual bleeding was too low to permit further interpretation of this increase.

Although 12 women experienced adverse events related to withdrawal bleeding for which they discontinued the medication, the effect of the trimegestone dose was so dominant that their exclusion had no significant impact on the data.

In summary, we have shown that the dose of trimegestone is the major determinant of the pattern of uterine bleeding. Higher doses were associated with a later onset of bleeding, a shorter duration, and less heavy bleeding. Variability in the duration of bleeding also decreased. We believe that all of these factors will improve patient continuance and therefore consider that for routine addition to oral micronized oestradiol in postmenopausal women, the highest dose of trimegestone (0.5 mg) is to be preferred over the lowest (0.05 mg).


    Acknowledgments
 
This study was supported by Hoechst Marion Roussel, Romainville, France, and Wyeth-Ayerst International, USA, in a joint development programme.


    Notes
 
5 To whom correspondence should be addressed Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Akkad, A., Hartshorne, T., Bell, P.R.F. and Al-Azzawi, F. (1996) Carotid plaque regression on oestrogen replacement: a pilot study: Eur. J. Vasc. Endovasc. Surg., 11, 347–348.[ISI][Medline]

Al-Azzawi, F. and Habiba, M. (1994) Regular bleeding on hormone replacement therapy: a myth? Br. J. Obstet. Gynaecol., 101, 661–662.[ISI][Medline]

Bouchoux, F., Cerede, E. and Philibert, D. (1995) Measurement of the relative binding affinity of RU27987 to recombinant human steroid receptors: progestogen, glucocorticoid, androgen and estrogen. Determination of the binding parameters of RU27987 and progesterone for the recombinant human progesterone receptor. 93/5353/PH. Data on file at Roussel-Uclaf, Romainville, France.

Brestlow, N.E. and Clayton, D.G. (1993) Approximate inference in generalized linear mixed models. J. Am. Stat. Assoc., 88, 9–25.[ISI]

Daly, E., Vessey, M.P., Barlow, D. et al. (1996) Hormone replacement therapy in a risk-benefit perspective. Maturitas, 23, 247–259.[ISI][Medline]

GENSTAT 5 Committee (1993) GENSTAT 5 Release 3 Reference Manual, Clarendon Press, Oxford.

Greene, J. (1976) A factor analytic study of climacteric symptoms. J. Psychosom. Res., 20, 425–430.[ISI][Medline]

Grimes, D.A. (1982) Diagnostic dilation and curettage: a reappraisal. Am. J. Obstet. Gynecol., 142, 1–6.[ISI][Medline]

Habiba, M., Bell, S.C., Abrams, K. and Al-Azzawi, F. (1996) Endometrial responses to hormone replacement therapy: the bleeding pattern. Hum. Reprod., 11, 503–508.[Abstract]

Harlow, S.D. and Zeger, S.L. (1991) An application of longitudinal methods to the analysis of menstrual diary data. J. Clin. Epidemiol., 44, 1015–1025.[ISI][Medline]

Magos, A.L., Brewster, E., Singh, R. et al. (1986) The effects of norethisterone in postmenopausal women on oestrogen replacement therapy: a model for the premenstrual syndrome. Br. J. Obstet. Gynaecol., 93, 1290–1296.[ISI][Medline]

Nachtigall, L.E. (1990) Enhancing patient compliance with hormone replacement therapy at menopause. Obstet. Gynecol., 75 (Suppl. 4), 77S–80S; discussion 81S–83S.[Medline]

Patterson, H.D. and Thompson, R. (1971) Recovery of inter-block information when block sizes are unequal. Biometrika, 58, 545–554.[ISI]

Sturdee, D.W., Wade-Evans, T., Paterson, M.E. et al. (1978) Relations between bleeding pattern, endometrial histology, and oestrogen treatment in menopausal women. Br. Med. J., 1, 1575–1577.[Medline]

Sturdee, D.W., Barlow, D.H., Ulrich, L.G. et al. (1994) Is the timing of withdrawal bleeding a guide to endometrial safety during sequential oestrogen-progestogen replacement therapy? Lancet, 343, 979–982.[ISI][Medline]

Whitehead, M.I. (1978) The effects of oestrogens and progestogens on the postmenopausal endometrium. Maturitas, 1, 287–298.[ISI]

Submitted on August 12, 1998; accepted on November 23, 1998.