Human Reproduction vol.5 no.7 pp 862-869, 1990

Oocyte donation: the role of endometrial receptivity

M.C.Davies1, M.C.Anderson2, B.A.Mason3 and H.S.Jacobs Department of Reproductive Endocrinology, Cobbold Laboratories, University College and Middlesex School of Medicine, Mortimer Street, London Wl, 2Department of Histopathology, St Mary's Hospital Medical School, Praed Street, London W2, UK and 3 Hallam Medical Centre, 112 Harley Street, London Wl

a clinical pregnancy (Edmonds et al., 1982; Miller et al., 1980). In-vitro fertilization (IVF) results in even greater wastage, with only 10-15% of embryos implanting (Rogers et al., 1986; Osborn and Moor, 1988). The study of ovum donation cycles may throw light on the reasons for this wastage and thus offer guidance for improving IVF outcome.

'To whom correspondence should be addressed

Materials and methods Sixteen patients, 15 with primary ovarian failure and one carrier of chromosomal abnormality, received 21 embryo transfers following in-vitro fertilization of donated oocytes. Hormone replacement therapy was given to mimic a natural menstrual cycle. Five pregnancies were achieved (four infants delivered and one early pregnancy loss). All patients had endometrial biopsies taken in a preceding cycle of hormone replacement therapy. The majority (61%) showed delayed maturation compared with the expected appearances for the day of the cycle as assessed on light microscopy. Patients who conceived had a significantly better endometrial response than those who did not. Five out of seven patients (71%) conceived when the endometrium was 'in phase' ( < 2 days' delay). This suggests that endometrial receptivity is a key factor in conception. The implications for improving outcome from in-vitro fertilization treatment are discussed. Key words: donor gametes/endometrium/implantation

Introduction The first successful pregnancy achieved by ovum donation and embryo transfer in a patient with ovarian failure was reported in 1984 by Lutjen and colleagues. They also described a regimen of hormone replacement therapy (HRT) designed to mimic the natural menstrual cycle. Subsequently, other workers have introduced alternative techniques of ovum transfer (Asch et al., 1987; Yovich et al., 1987) and differing HRT regimens (Simon et al., 1986; Serhal and Craft, 1987). There have been attempts to define die 'window' of endometrial receptivity for implantation (Navot et al., 1986; Rosenwaks, 1987; Salat-Baroux et al., 1988; Cameron et al., 1988); the majority of pregnancies have followed embryo transfer on day 17-19 of an artificial 28-day cycle. However, we have little information about the uterine environment during HRT administration. Patients with ovarian failure provide a model to study the relative contributions of endometrial status and embryo quality in achieving a pregnancy. In natural conception cycles, it is estimated that at least one third of embryos fail to give rise to 862

Patients From November 1987 to date, 40 couples have enrolled in our ovum donation programme. This report describes the results obtained in the 33 patients who have completed at least one monitored cycle of hormone replacement therapy. Their mean age was 33.1 years (SD 5.0, range 24—41 years) and mean duration of infertility 5.4 years (SD 4.0, range 1 — 16 years). Their diagnoses are listed in Table I. Seventeen patients had amenorrhoea caused by premature ovarian failure. In two patients this followed surgery for bilateral ovarian cysts. In one patient ovarian auto-antibodies were present, and in six, a variety of other auto-antibodies were detected. Four patients required thyroid replacement and one had hypothyroidism, insulin-dependent diabetes and adrenal auto-antibodies. Five patients had primary amenorrhoea with gonadal dysgenesis and three of these had Turner's syndrome with karyotypes 45XO; 46,X,del (X)(pll.l-pter) and 40%:60% mosaicism 45X/46X psudic (X)(qter-p22:22:pll-qter). None had cardiac or renal anomalies. Nine patients entered the ovum donation programme after repeated failure of IVF attempts with a very poor response to ovarian stimulation; three of them had previously undergone ovarian surgery. One patient had presented with hypogonadotrophic hypogonadism but showed no response to LHRH or gonadotrophin therapy and, aged 39 years, was presumed to have ovarian failure. One patient had suffered recurrent miscarriage, losing four pregnancies because of non-dysjunction of chromosome 13. Of the 16 patients who have so far received embryo transfers, 11 had primary and five secondary infertility (only one had a living child). Their mean age was 33.7 years. All but one of the male partners were normally fertile. In one case donor semen was used. Preparation of recipients Couples entering the programme were all counselled by a doctor (MCD) and had baseline investigations repeated, namely, gonadotrophin and oestradiol concentrations in the female partner, with karyotype, thyroid function and auto-antibody screen, © Oxford University Press

Endometrium in egg donation cycles

Table I. Patient characteristics Diagnosis

Number of patients

Mean age (years)

Premature ovarian failure (P.O.F.) Gonadal dysgenesis Failure of IVF stimulation Hypogonadotrophic hypogonadism + P.O.F. Carrier of chromosomal abnormality

17

32.9 ± 3.8 26.6 ± 2.2 37.0 ± 4.3 39 28

5 S l l

Number pregnant

Number receiving ET

HORMONE REPLACEMENT THERAPY

75 cS CD CO

50 5T §

s

25

i CD

O

8

12 16 20 Days of cycle

24

28

32

Fig. 1. Hormone replacement therapy.

E 1000 a. T3

ra

500

Day 12

Day 16

Day 21

1BU

160



140



o

120



CD c

100



80



o

60

*

» •

*

st • ••

40 20



n Day 8

Day 12

1

Fig. 3. Histological assessment of endometrial biopsies (n = 33; score = discrepancy in days from expected appearances; six biopsies unclassifiable).

1500

Day8

- 2 - 1 0 Endometrial score

Day 16

Day 21

Fig. 2. Serum hormone measurements in monitored cycles (median values indicated).

cervical smear and pelvic ultrasound, rubella antibody status and blood grouping. Both partners were screened for HTV-1 and hepatitis B and a semen analysis was performed. Recipients without normal menstrual cycles received hormone replacement therapy in the form of oral oestradiol valerate (Progynova; Schering) and intramuscular progesterone (Gestone; Paines and Byrne) in a regimen based on that of Lutjen and colleagues (1984; Figure 1). Vaginal administration of progesterone was substituted in 11 cases (Cyclogest; Hoechst; 200 mg D15-16, 400 mg D17-26). The first cycle of hormone replacement therapy was monitored by serial measurements of serum gonadotrophin, 17-/3-oestradiol and progesterone concentrations. Blood samples were taken on days 8, 12, 16 and 21 of the cycle; sampling times were not standardized. All hormone estimations were performed by radioimmunoassay (RIA) as follows: LH and FSH by RIA with solid phase separation using reagents from the Chelsea Hospital for Women, London; oestradiol by direct RIA with l25I-labelled oestradiol (Baxter Healthcare, UK); progesterone with kits from Immunodiagnostic Systems Ltd, UK. An endometrial biopsy was obtained on Day 21 by aspiration curettage (3-mm Vabra curette; Cory Bros.) and fixed in Bouin's solution, paraffin embedded and stained with haematoxylin and eosin. All biopsies were assessed using light microscopy by the same observer (MCA) according to the criteria of Noyes et al. (1950). Cyclical therapy was continued and embryo transfer performed in the next or subsequent cycle. 863

M.C.Davies et al.

Donors Ovum donors were recruited from patients undergoing oocyte recovery for IVF or gamete intra-Fallopian transfer (GIFT). All

donors were volunteers and donation was anonymous. Potential donors were excluded if they were aged ^ 3 7 years, if they had any significant personal or family history of illness or heritable

V .

B

mA*

V

864

Endometrium in egg donation cycles

Fig. 4. (A) Endometrial biopsy, day 21, i.m. progesterone, showing normal secretory development (score = 0). Patient conceived in subsequent cycle (Bouin, HES, X150). (B) Endometrial biopsy, day 21, i.m. progesterone, showed delayed maturation with subnuclear vacuolation (score = —4). Patient failed to conceive. (Bouin, HES, X150). (C) Endometrial biopsy, day 21, i.m. progesterone, showing poor response (score = unclassifiable) (Bouin, HES, X150). disease, or if they had fewer than eight oocytes collected. Blood was taken for chromosome analysis, blood grouping and screening for HIV-1, hepatitis B and syphilis serology, thalassaemia and sickle cell trait as appropriate. The physical characteristics of donor and recipient were matched as far as possible. The number of oocytes donated depended on the donor's wishes and was finalized at the time of collection. Donated oocytes were fertilized in vitro with the spermatozoa of the recipient's husband. A fresh or, if necessary, frozen semen sample was prepared by 'swim-up' techniques. Oocytes were incubated at 37°C in supplemented Earles culture medium and inseminated after 3 —6 h. Embryo transfer was performed by the transcervical route as an outpatient procedure. Fresh embryos were transferred at the 2- or 4-cell stage, ~48 h after oocyte recovery, provided the recipient's cycle was in synchrony with the donor. The optimal timing for transfer was considered to be day 18 of the simulated cycle, i.e. the fourth day of progesterone administration. The recipient's cycle could be adjusted by introducing progesterone 1 - 2 days earlier to synchronize the timing of transfer. However, the majority of embryos were frozen at the pronuclear stage using propanediol as a cryoprotectant and cryopreserved until day 18 of the recipient's next cycle. In three treatment cycles, minor individual adjustments were made in the HRT regimen in the light of results from those patients' monitoring cycles. Pregnancy was diagnosed by elevated serum /3-HCG concentra-

tions 14 days after embryo transfer and confirmed by pelvic ultrasound scan at 7 weeks' amenorrhoea. On diagnosis of pregnancy, the dosage of oestradiol valerate was increased to 8 mg/day and progesterone to 100 mg/day. Hormone assays were monitored weekly and hormonal support adjusted accordingly, aiming to reduce and discontinue therapy by 16 weeks. Patients were referred to their local obstetrician for antenatal care and delivery. Statistical analysis of the data was performed by univariate and multivariate analysis of variance on raw data and logarithmically transformed values, and by non-parametric analysis (Fisher's randomization test) as appropriate. All oocyte donors and recipients and their partners gave informed consent for these procedures. The programme received ethical approval from the Middlesex Hospital Clinical Investigations Panel and was approved by the Voluntary Licensing Authority. Results Thirty-three patients completed at least one monitored cycle of hormone replacement therapy. In their first cycle, 22 patients received intramuscular progesterone and 11 received vaginal progesterone in the regimens described above. Results of the hormonal profiles of these cycles are summarized in Figure 2. Adequate serum levels of 17-/3-oestradiol and progesterone were 865

M.C.Davtes et at.

Table II. Patients receiving embryo transfer No.

Age

Diagnosis

Endometrial score

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

39 25 39 27 24 41 33 37 38 37 36 34 33 39 29 26

Failed IVF P.O.F. Failed IVF Recurrent miscarriage Turner's Failed IVF P.O.F. P.O.F. P.O.F. P.O.F. P.O.F. P.O.F. P.O.F Failed IVF P.O.F. Turner's

+2 0 0

o

-l -l -l -2 -2 -2 -4 -41 -4 -4 Unclass* Unclass*

Treatment cycles

; ; ;

: :

Embryos transferred

Outcome

3 2 3 1 3 1 3 2 1 2 2 1 2 1 2 1

Pregnant Pregnant Pregnant

x x x x x x x x x x x x x x x x

PN F 4-cell F PN F PN F; 2 PN F; 2 PN F; 3 PN F PN F; 2 2-cell; 2 PN F PN F 2-cell F PN F PN F PN F PN F

x PN F x PN F x PN F

Pregnant Pregnant

x PN F x 4-cell

P.O.F., premature ovarian failure, unclass, biopsy could not be classified; PN, pronuclcar embryo; F, frozen—thawed embryo. •Vaginal administration of progesterone.

attained, the median values lying within the reference ranges for the same days of a normal menstrual cycle (Landgren et al., 1980); however, the range of oestradiol measurements was wide. Serum levels of progesterone equivalent to normal luteal values were reached in the majority of patients [peak value > 32 nmol/1 in 25/33 (76%) of subjects]. Values were significantly higher in the 'intramuscular progesterone' group than the 'vaginal progesterone' group (day 21 mean = 60 ± 8.0 cf. 29.8 ± 5.4 nmol/1; P = 0.014). Median values of serum FSH concentrations on days 8, 12, 16 and 21 of the cycle were, respectively, 12.1, 15.1, 15.4 and 8.0IU/1; the corresponding values of LH were 12.3, 23.3, 25.4 and 11.3 IU/1. In the group of patients with premature ovarian failure, elevated gonadotrophins were not suppressed by HRT until late in the cycle, and in some individuals remained elevated throughout [8/17 (47%) had FSH consistently > 15 IU/1]. One patient had low gonadotrophins caused by hypopituitarism. The majority of endometrial biopsies (27/33; 82%) showed a discrepancy with the expected appearances for the appropriate day of the menstrual cycle. From the histological assessment an endometrial 'score' was derived, indicating the degree of discrepancy, which was assessed to the nearest day. For example, a biopsy taken on day 21 of the simulated cycle but showing appearances equivalent to day 18 of a normal menstrual cycle was scored ' - 3 ' . The results are shown in Figure 3. Delayed maturation was noted in 20/33 (61%) of endometrial samples. Only one patient had a biopsy in advance of the expected appearance. She had not responded to IVF stimulation but still had spontaneous menstrual cycles. Biopsies taken during cycles with intramuscular progesterone showed less discrepancy with the chronological date (i.e. better maturation) than those with vaginal progesterone. The mean endometrial scores were —1.5 and —2.9, respectively (P = 0.03). Six biopsies could not be classified, showing neither adequate secretory nor proUferative activity. Four were from patients with gonadal dysgenesis (two with Turner's syndrome) and thus four 866

of the five biopsies in this group of patients showed a very poor response. The remaining two unsatisfactory biopsies were from patients receiving vaginal progesterone (Figure 4). Subsequently, four patients have undergone repeated monitored cycles of hormone replacement therapy. Two patients who had previously received vaginal progesterone underwent a further cycle with intramuscular administration. Both achieved higher serum progesterone values and one showed a better endometrial response (scores - 4 , —2). The other two patients had gonadal dysgenesis and had had 'unclassifiable' biopsies. One showed an improved response, but the other has undergone three monitored cycles during 9 months of cyclical HRT and has shown no endometrial response despite consistently 'normal' serum hormone concentrations. There was no detectable difference in endometrial response between patients who were amenorrhoeic on entering the programme, patients already receiving HRT (usually an oral contraceptive pill) and those with spontaneous menses. The mean endometrial scores were respectively, - 1 . 6 (n = 11), - 2 . 1 (n = 13) and - 1 . 6 (n = 9); (N/S). Sixteen patients have undergone at least one treatment cycle with embryo transfer. The details of these cycles are summarized in Table II. Thirteen patients had used oral oestrogen with intramuscular progesterone and three with vaginal progesterone. Five patients underwent a second treatment cycle. Thus a total of 21 treatment cycles on HRT were analysed. A total of 41 embryos were transferred: three fresh and 38 (93%) frozen thawed. The majority of embryos (35; 85%) was transferred at the pronuclear stage, and six had cleaved (2 x 2-cell, 4 x 4-cell) which were all graded 1-2. Five pregnancies resulted (31% of patients treated), two in patients with premature ovarian failure, two in patients with failure of ovarian response for IVF and one in a patient with gonadal dysgenesis (Turner's mosaicism). This pregnancy rate was equivalent to 24% per embryo transfer and the implantation rate was 12% per embryo transferred. All the pregnancies arose from frozen—thawed embryos and in only one case had cleavage

Endometrium in egg donation cycles

TaWe IH. Factors affecting outcome of treatment

CO CD O O

cn CD O

c LLJ

Conceived (n = 5)

1 0

• o*

-1 -2

o o o

Embryos transferred (n)

-3 o o

-4

Age (years) 17-|3-oestradiol (pmol/1) Day 8 Day 12 Day 16 Day 21 Progesterone (nmol/1) Day 16 Day 21 Endometrial score (see text)

32.2 ± 3.4 610.8 637.2 560.4 459.8

± ± ± ±

127.8 85.2 136 2 85.0

43.2 ± 9.1 57.6 ± 16.2 0.0

Not conceived (n = 11) 34.4 ± 1.5 459.8 625.2 487.3 621.0

± ± ± ±

39.7 49.9 59.5 98.7

25.8 ± 4 . 6 P = 0.07 62.2 ± 13.5 -2.6 P = 0.009

Unsuccessful Conception cycles cycles 2.60 ± 0.24 1.75 ± 0 . 1 7

P = 0.04

o o

n = 16 2 unclassifiable • = conceived Fig. 5. Relation of treatment outcome to endometrial score.

occurred after thawing. No patient became pregnant after the transfer of only one embryo. No multiple pregnancies occurred. Figure 5 shows the endometrial score in the patients' preceding monitored cycles. No patient became pregnant when the endometrium lagged two or more days behind expected appearances. In four cases a singleton pregnancy with fetal heart activity was confirmed by ultrasound at 7 weeks amenorrhoea. The fifth patient (case no. 7) had an early pregnancy loss, with slight vaginal bleeding from 5'A weeks; although serum /3-HCG, oestradiol and progesterone concentrations were satisfactory, no fetal pole was identified on ultrasound at 7 weeks. The patient subsequently had further bleeding and /3-HCG levels declined. In the four ongoing pregnancies, hormonal support was given as described. The dosage of HRT was reduced from 12 weeks and all medication was discontinued at 16-17 weeks. All patients have delivered normal infants, two male and two female, three by Caesarean section and one spontaneous vaginal delivery. Discussion Treatment by ovum donation in patients with primary ovarian failure presents the clinician with the challenge of artificially creating a 'physiological' menstrual cycle. Data presented here and elsewhere (Navot et al., 1986; Rosenwaks, 1987; Cameron et al., 1988) show that despite wide individual variation in metabolism of administered sex steroids (Nillius and Johansson, 1971; Dusterberg and Nishino, 1982), a fairly simple regimen of hormone replacement therapy can achieve apparently adequate serum levels of 17-j3-oestradiol and progesterone. However, the interpretation of peripheral hormone measurements must be limited both by the conditions of sampling and by the poor

relationship between serum hormone values and end-organ response (Campbell and Whitehead, 1982). We have used the histological appearance of the endometrium as an indicator of end-organ response to therapy. In only 6/33 (18%) of patients was the endometrium completely consistent with a normal cycle and 10/33 (30%) were 'in phase' as defined by

Oocyte donation: the role of endometrial receptivity.

Sixteen patients, 15 with primary ovarian failure and one carrier of chromosomal abnormality, received 21 embryo transfers following in-vitro fertiliz...
4MB Sizes 0 Downloads 0 Views