The effect of transdermal oestradiol on bleeding pattern, hormonal ...

4 downloads 55 Views 2MB Size Report
University, Rama IV Road, Bangkok 10330, Thailand. The most .... University (Bangkok, Thailand). As far as ..... Cambridge University Press,. Cambridge, UK, pp ...
The effect of transdermal oestradiol on bleeding pattern, hormonal profiles and sex steroid receptor distribution hi the endometrium of Norplant users Wisut Boonkasemsanti1, Damrong Reinprayoon1, Kamthorn Pruksananonda1, Somchai Niruttisard1, Surang Triratanachat1, Smai Leepipatpaiboon1 and Pongsak Wannakrairot2 Departments of 'Obstetrics and Gynecology and 2Pathology, Faculty of Medicine, Chulalongkom University, Rama IV Road, Bangkok 10330, Thailand

The most common side-effect and reason for discontinuation with Norplant use is bleeding disturbance. The aim of this study was to investigate whether the 6 week application of a patch which released 100 ug/day oestradiol would reduce the number of abnormal bleeding days or eliminate the problem. Another objective was to find out the correlation between the bleeding pattern and endometrial concentrations of oestrogen receptor (ER) and progesterone receptor (PR). Of 98 Norplant users, 34 patients had normal bleeding patterns and 64 patients had abnormal bleeding patterns. An oestradiol patch or a placebo patch were randomly used to treat 33 and 31 women with abnormal bleeding respectively. There was a clinical improvement in the oestradiol group compared with the placebo group, although this was not statistically significant There were no correlations between PR and ER concentration and the serum oestradiol, progesterone, levonorgestrel and sex hormone-binding globulin concentrations. Significantly increased mean immunostaining scores of stromal PR were observed in those Norplant users whose endometrium had an atrophic histological appearance. The serum oestradiol concentration did not show a significant change after treatment with the oestradiol patch compared with the placebo patch. Key words: endometrium/immunohistochemistry/ Norplant/steroid receptors/transdermal oestradiol

Human Reproduction Volume 11 Supplement 2 1996

Introduction Progestogen-only methods of contraception were developed when certain side-effects and risks were identified with the oestrogen component of combined oral contraceptives. To improve the limited contraceptive efficacy with oral progestogen only pills (POP) without a considerable increase in dose, alternative routes of administration for progestogens were developed. Such routes include injectables, s.c. implants and vaginal rings. The Norplant contraceptive consists of s.c. placed silastic implants releasing a progestogen, levonorgestrel. The release rate is 30-80 H-g/24 h (WHO, 1985), which is equivalent to a daily POP. Norplant offers a high contraceptive efficacy and few side-effects. The most common side-effect and reason for discontinuation with Norplant is bleeding disturbance (Odlind and Fraser, 1990). Disruption of the normal menstrual pattern occurs initially in 7080% of recipients (Sivin, 1994). The most common pattern is characterized by frequent, irregular and/ or prolonged bleeding periods. All studies have reported improvement (normalization) of the bleeding pattern with increased duration of Norplant use (Shoupe et al., 1992). The mechanisms behind the bleeding disturbances are poorly understood. There are probably several factors that contribute to the bleeding pattern: the lack of any oestrogenic effects, the direct suppressive effects of progestogen on the endometrium and the effects of progestogens on the ovulatory cycle. From clinical experience there is little doubt that contraceptive methods containing both progestogen and oestrogen result in fewer

O European Society for Human Reproduction & Embryology

115

ft g^

days of bleeding/spotting than a progestogen-only method. There are some reports that the addition of an oestrogen to long-acting progestogenic methods could improve the bleeding pattern; however, there are few, if any, studies that have systematically addressed this problem (Diaz et al., 1990; Fraser, 1990). It is expected diat a significant contribution to the understanding of the mechanism of endometrial bleeding can be made in our study by investigating the concentrations of progesterone receptor (PR) and oestradiol receptor (ER), as well as the concentrations of circulating levonorgestrel, progesterone, oestradiol and sex hormone binding globulin (SHBG). Last, but not least, morphometric observations (Johannisson et al., 1982) can be of great significance. The first aim of our study was to evaluate whether the administration of a low-dose transdermal oestradiol via Estraderm (Ciba Geigy, Switzerland), administered to women who suffer from prolonged bleeding during the use of Norplant implants, could improve the bleeding patterns by reducing the number of bleeding days and the number of bleeding episodes or, with this continuous administration of oestradiol via Estraderm, whether the bleeding could be abolished completely. The second aim of our study was to investigate whether the occurrence and frequency of bleeding were related to endometrial concentrations of ER and PR on the one hand, and to circulating concentrations of levonorgestrel, progesterone, oestradiol and SHBG on the other. Materials and methods Subject selection Women who were currently using Norplant implants and willing to participate in the study were recruited. Inclusion criteria A total of 120 healthy, fertile, non-lactating women using Norplant as contraceptive for a maximum of 1 year were selected. Women with a 'normal' bleeding pattern (bleeding < 8 days, intervals >20 days) were included in the first (control) group. 116

Women with bleeding problems (bleeding > 8 days and/or intervals 8 days and/or intervals between bleeding episodes of 8 days and/or interval 20 days).

Immunohistochemical procedure Endometrial samples were mounted using Tissue Tec (OCT compound; Miles Diagnostics Division, Elkhart, IN, USA). Frozen tissue was cut at 6|X in a cryostat at -20°C and thaw-mounted on glass slides. The immunohistochemical assay for ER and PR was performed as described by the manufacturer (Abbott Laboratories). ER and PR antibody incubation was prolonged overnight to ambient temperature in a moist chamber. Diaminobenzidine (DAB) was used as a chromogen, resulting in a brown nuclear localized reaction product (van Agthoven et al., 1994). Prior to the receptor assay, each specimen was thawed, rinsed with ice-cold saline, snap-frozen again and lyophilized (Koenders et al., 1978; Ryan and Walker, 1980). Each specimen was minced with a scalpel and extracted by intermittent vortexing (every 5 min) with a buffer containing 0.6 M KC1 (Ryan and Walker, 1980; Thorpe et al., 1986; Press et al, 1989). In this way 'total' receptors were measured, i.e. both nuclear and cytosolic receptors were extracted from endometrial cells, and the receptor concentrations were expressed in pmol/g protein. Quantitative ER and PR determinations by immunocytochemical staining were performed by P.W. (Department of Pathology, Faculty of Medicine, Chulalongkorn Hospital, Bangkok, Thailand). The assays of circulating levonorgestrel, oestradiol and progesterone were performed using World Health Organization (WHO) matched reagents. The SHBG assays were carried out using ammonium sulphate precipitation (Cekan et al, 1985).

Laboratory and other investigations Each endometrial biopsy was divided into three pieces. The first piece was fixed in Bourn's solution and used for histological analysis by S.N. and S.T. The second and third pieces were frozen immediately and kept at -70°C. These pieces were used for the determination of ER and PR concentrations. Receptor measurements were carried out using immunocytochemical assay kits (Abbott Laboratories, USA; Abbott ER-ICA monoclonal lot 73759M3OO, PgR-ICA monoclonal lot 66909M100; labelled streptavidin-biotin method).

Data analysis The bleeding data were analysed according to WHO criteria for the computer analysis of menstrual bleeding data using the Menstrual Diary Analysis Program MDS. Other data were processed by computer programs available at Chulalongkorn University (Bangkok, Thailand). As far as possible, parametric methods were used for the calculation of means, SD, confidence limits, Wests and analysis of variance. In the cases of gross non-normality of the observations, the Wilcoxson signed rank test was employed. 117

W.Boonkasemsantl el aL

Table I. Histopathological findings of endometnal biopsies pre- and post-treatment in control, oestradiol patch and placebo patch groups Control

Atrophic Proliferative Secretory Shedding

Oestradiol patch

Placebo patch

Hrst

Second

Fust

Second

First

Second

24 9

25 7 2

27 6

25 8

24 6 1

25 6

1

Results The general characteristics of the subjects were not different for the control, oestradiol patch and placebo patch groups. Their mean ages were 26 ± 4, 27 ± 3 and 25 ± 4 years respectively. Their mean heights were 155 ± 5, 156 ± 4 and 157 ± 4 cm and their weights were 51 ± 8, 52 ± 7 and 54 ± 7 kg respectively. Of the 33 women undergoing oestradiol patch treatment, 23 showed a clinical improvement after treatment, while 13 of the 31 women with placebo patches improved clinically. The difference was not statistically significant for this number of cases. Histopathological assessments of the endometnal biopsies showed that the majority of the samples were atrophic, with some being proliferative and few secretory endometrium (Roger et al., 1993), as shown in Table I. These histopathological findings did not show any difference between the groups or a difference between pre-treatment and post-treatment. Serum hormonal concentrations There were individual variations, but the means of the normal bleeding pattern group were higher than those patients with an abnormal bleeding pattern; however, this difference was not statistically significant (Figure 1). When we analysed individuals, there was always cyclical ovarian activity in women with a normal bleeding pattern, as shown by the cyclical variations in oestradiol concentration in the twice weekly serum samples for the 6 week follow-up period. The variations in different individuals were not in synchrony. Subjects with abnormal bleeding patterns also showed some ovarian activity, but the pattern of cyclical variation was not as obvious as in the other group. 118

Serum progesterone concentrations were constantly low throughout the 10 week period. No individual showed ovulation activity, and these findings were similar in all three groups (Figure 2). Ovarian activities occurred in the majority of Norplant users, as shown by fluctuating high levels of oestrogen, but without ovulation as shown by constantly low levels of serum progesterone. Serum levonorgestrel concentrations varied between 1000 and 1500 pmol/1. There were no significant differences in the mean values among control, oestradiol patch and placebo patch groups. Again, when we looked at individuals, the levonorgestrel concentrations expressed more steady patterns for those who had a normal bleeding pattern, but at different levels. The serum levonorgestrel concentrations in the abnormal bleeding individuals fluctuated during the study period, but no statistically significant differences were detected (Figure 3). SHBG concentrations were low and within the range 20-50 nmol/1. No significant difference was determined when comparing the mean values of the three groups. As with levonorgestrel, the SHBG concentrations in the normal bleeding pattern individuals were more steady than in diose experiencing abnormal bleeding (Figure 4). Receptors Positive immunohistochemical staining for both ER and PR was confined to the nuclei of epithelial and stromal endometrial cells. Figure 5 shows the mean ± SEM immunostaining scores for endometrial sex steroid receptor in the stromal endometrial cells. There was significantly higher PR staining than ER staining in the stromal cells. The distribution of both PR and ER in these stromal endometrial cells was not

Steroid receptor distribution in Norplant endometrium 700—-] Control (N = 34)

=

100

Oestrogen (pmol/1)

700-—,

100

-

700

1

Treatment (N = 33)

Placebo (N = 31) 100

' Diy 0

28

42

45 49

52

56

59

63

66 70

Figure 1. Serum oestradiol concentration (mean ± SEM) of the control, oestradiol patch and placebo patch groups. 5

_

t

0

1

5

1

Progesterone

(nmol/l) 1 treatment

0

1

5

1

(N=33)

i i i H o—' Day

28

0

42

45 49 52 56 59

I

J

T

T

Placebo (N=31)

63 66 70

Figure 2. Serum progesterone concentration (mean ± SEM) of the control, oestradiol patch and placebo patch groups. 1500

1

i 1000—I 1500—,

Control

Levonorgestrel

(pmol/1) *

* *

|Treatment

looe—' 1500—, Placebo 1000—I Day

i * 0

28

42

45 49

52

56

59 63

66 70

Figure 3. Serum levonorgestrel concentration (mean i SEM) of the control, oestradiol patch and placebo patch groups.

119

WJJoonltflsemsanti el aL

50~ Control (N»34)

10. 50-

SHBG

(nmol/l)

TYntincnt (N-33)

1050"

H ' i l

I

I

63

66

Placebo (N-31)

10Day

0

28

42

45

49

52

56

59

70

Figure 4. Serum sex honnone-binding globulin (SHBG) concentration (mean ± SEM) of the control, oestradiol patch and placebo patch groups.

• Oestrogen Receptor • Progesterone Receptor Stroma

1st

2nd

control

1st

2nd

treatment

1st

2nd

placebo

Figure 5. Histograni of endometrial sex steroid receptor immunostaining scores (mean ± SEM) in stromal cells of the endometrium. Oestrogen receptor = first column. Progesterone receptor = second column.

significantly different when comparing the first and second endometrial biopsies within the same group and between different groups. Figure 6 shows the mean ± SEM immunostaining scores for endometrial sex steroid receptor in the glandular endometrial cells in these Norplant users. There was significantly higher PR staining than ER staining in these glandular cells. Again, the pattern of distribution of both PR and ER in the glandular cells was not significantly different when comparing the first and second biopsy specimens within the same group and between different groups. PR staining intensities in stromal cells were significant higher than those in glandular cells of the same group and between groups. The ER staining intensities in stromal cells were also higher

120

than ER staining in glandular cells (not statistically significant) within the same group and between different groups. Total tissue sex steroid receptor Figure 7 shows the endometrial ER and PR concentrations in the different treatment groups. More than half of the biopsy specimens were inadequate for analysis: 12 cases in the control group, eight cases in the oestradiol patch treatment group and seven cases in the placebo patch treatment group. There were no statistically significant differences between the first and second biopsies of each group or between the three groups. It appeared that PR was more dominant than ER in these endometrial tissues. The detection of PR and ER using both techniques (immunohisto-

Steroid receptor distribution in Norplant endometrium

I Oestrogen Receptor | Progesterone Receptor

Glands

IK

2ad

1st

treatment

2nd

placebo

Figure 6. Histogram of endometrial sex steroid receptor immunostaimng scores (mean ± SEM) in glandular cells of the endometrium.

• Oestrogen Receptor D Progesterone Receptor

300 -i

lit

2nd

control (N=12)

1st

2nd

treatment (N=8)

1st

2nd

placebo (N=7)

Figure 7. Histogram of total endometrial sex tteroid receptor concentrations (mean ± SEM) of the control, oestradiol patch and placebo patch groups.

chemical staining and tritiated ligand binding reaction) was not contradictory. Relationships between receptor immunohistochemical staining and serum hormonal concentrations We found no correlation between serurn oestradiol concentrations and PR or ER immunostaining scores on the day of biopsy or during the treatment period. There was no correlation between serum progesterone concentrations and PR or ER inununostaining scores on the day of biopsy or during the treatment period. Similarly, no correlation was found between serum levonorgestrel concentrations and PR or ER immunostaining scores.

Discussion In this study we found that there was ovarian activity, especially follicular growth, in the majority of Norplant users. Serum oestradiol concentrations varied cyclically in women with normal bleeding patterns but not in those with abnormal bleeding patterns. Serum progesterone concentrations showed a constantly low value (