Ovarian stimulation in intrauterine insemination with donor sperm: a randomized study comparing clomiphene citrate in fixed protocol versus highly purified urinary FSH

R. Matorras1,3, T. Diaz1, B. Corcostegui1, O. Ramón1, J.I. Pijoan2 and F.J. Rodriguez-Escudero1

1 Human Reproductive Unit, Department of Obstetrics and Gynecology and 2 Clinical Epidemiology Unit, País Vasco University, Hospital from Cruces, Baracaldo, Vizcaya, Spain


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
BACKGROUND: The study was conducted to compare the results of intrauterine donor insemination (DI) under ovarian stimulation with either clomiphene citrate (CC), in a fixed protocol, or FSH, with ovarian monitoring. METHODS: Forty-nine patients were randomized using a computer-generated list to receive highly purified urinary FSH (starting dose of 150 IU) and were subjected to periodic vaginal ultrasound and estradiol determinations. HCG was given when >=2 follicles (>=17 mm) were identified and estradiol reached >400 pg/ml. Intrauterine insemination (IUI) was performed 36 h later. The other 51 received CC on a fixed protocol (100 mg/day from the day 5–10 of the ovarian cycle) with HCG being administered on the day 12, and IUI performed 36 h later. Up to six IUI cycles were performed on all patients if pregnancy was not reached before. Women failing to conceive in the CC group underwent IUI with FSH. The main outcome measures were intrauterine gestational sac observed by transvaginal ultrasound, per cycle and per woman pregnancy rate (PR) and multiple PR. RESULTS: The per cycle PR was significantly higher in the FSH group, 14.4% (30/209) versus 6.1% (16/261), as well as the per woman PR, 61.2% (30/49) versus 31.4% (16/51). 12.5% (2/16) of pregnancies obtained in the CC group were multiple, compared with 20% (6/30) in the FSH group. There were no triplets or higher order pregnancies in CC versus two in FSH (6.7% of pregnancies). Patients failing to conceive with CC, who later underwent intrauterine DI with FSH, had similar results to the primary FSH group: 54.3% PR per patient (19/35) and 16.0% per cycle (19/118), with a multiple PR of 31.6% (6/19). The PR for women starting with CC cycles and, if pregnancy was not obtained, continuing with six FSH cycles, was 69.2%. CONCLUSIONS: The PR obtained with CC stimulation was approximately half that obtained with FSH. There was a trend to lower multiple PR with CC. It is recommended that each case should be considered on an individual basis and the treatment options discussed with patients. In our opinion, CC could be a reasonable approach for young women with good prognosis, whereas in the remaining cases FSH would be the preferable method.

Key words: clomiphene citrate stimulation/FSH stimulation/intrauterine donor insemination/pregnancy rate/prospective randomized trial


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Donor insemination (DI) is presently performed almost universally with frozen–thawed sperm. The principal reason is to prevent the transmission of human immunodeficiency virus (HIV), avoiding HIV transmission during the `window' period (American Fertility Society, 1990Go). Since the use of frozen–thawed sperm yields a lower pregnancy rate (PR) (Smith et al., 1981Go; Richter et al., 1984Go; Brown et al., 1988Go; Subak et al., 1992Go), a number of strategies have been designed to increase the PR. Intrauterine insemination (IUI) has been shown to increase PR (Byrd et al., 1990Go; Patton et al., 1992Go; Matorras et al., 1996Go; O'Brien and Vandekerckove, 2000Go). Concerning ovarian stimulation, the combination of gonadotrophin stimulation with IUI achieved high PR (Wainer et al., 1995Go; Matorras et al., 1996Go). However, there were some concerns regarding the high multiple PR, as well as regarding costs. It is well known that, in timed intercourse cycles, the multiple PR with clomiphene citrate (CC) stimulation is much lower than with gonadotrophin stimulation (Speroff et al., 1999Go). The same is true regarding costs; however, PR are lower. In IUI with husband's sperm, the PR obtained with ovarian stimulation with CC is half that with gonadotrophins (Guzick et al., 1998Go).

The aim of the present study is to compare the PR and the multiple PR on a cumulative basis in two different DI protocols: IUI under ovarian stimulation with FSH and under CC stimulation in a fixed protocol.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
The setting where the study was performed was the Human Reproductive Unit of our Hospital, it being the only public centre with DI in the North of Spain. At the time the study was performed, no ICSI was available in our Unit. At our Unit, DI in unstimulated cycles is no longer performed, due to low PR (6.7%), despite efforts to increase it.

During two consecutive years (September 1998 to September 2000) the women consulting for DI, fulfilling the inclusion criteria, were randomized into two groups: IUI under ovarian stimulation with FSH (49 women) or with CC (51 women). Randomization was performed according to a randomly computer-generated number list. Informed consent and Institutional Review Board approval were obtained. The women were randomized either to CC or FSH at the beginning of the treatment and remained in this group during the six programmed cycles, unless pregnancy was achieved. If miscarriage occurred, women did not continue in the study, and each case was managed individually, outside the study protocol. Later, the women who had not become pregnant during the six cycles with CC were subjected to six additional cycles with FSH. This group will be designated as `FSH secondary group'. Women failing to reach pregnancy with FSH were no longer subjected to any DI cycles. There were no drop-outs.

The inclusion criteria were: infertility history of >2 years, woman's age <40 years, at least one normal patent tube, no previous DI or other assisted reproduction techniques (in order to avoid a selection bias). Mean age of women was 31.17 ± 3.18 years and the infertility duration 5.02 ± 1.98 years. In 94% of cases there was a primary infertility. DI indications were: abnormal sperm (94%), single woman (2%), protected intercourse because of HIV-positive partner (4%). Abnormal sperm was defined as those cases in which the semen sample was abnormal according to published standards (World Health Organization, 1992Go), following the recovery of <3x106 motile sperm after Puresperm preparation (NidaCom Laboratories, Sweden). Forty-one per cent of patients had azoospermic partners.

All women were subjected to the same infertility investigation: endometrial biopsy (only during the beginning of the study), progesterone and prolactin determination, hysterosalpingography (when abnormal, laparoscopy) and semen analysis.

IUI with frozen–thawed sperm was performed as previously reported (Matorras et al., 1996Go). The only change made was the substitution of Percoll by Puresperm. The management of the CC group consisted of a fixed regime protocol, without ovarian monitoring, and without performing any ultrasound or estradiol monitoring. CC (Clomifen; Casen-Fleet, Spain), 100 mg/day was given during 5 consecutive days, starting on the day 5 of the cycle. On day 12, 5000 IU of HCG (Profasi; Serono Laboratories) was administered i.m.

The FSH group was stimulated and monitored as previously described (Matorras et al., 1996Go): starting on 2nd day with 150 IU/day s.c. highly purified urinary FSH (Metrodin-HP; Serono Laboratories). Ovarian response was monitored with vaginal ultrasound and plasma estradiol. The first control was performed on the day 7 of the cycle. FSH was then adjusted according to the response. HCG (5000 IU i.m.) was given the day on which there were >=2 follicles of >=17 mm, the estradiol level being >400 pg/ml (conversion factor to SI unit 3.671). If there were >6 follicles >=16 mm and/or estradiol level >2000 pg/ml, the cycle was cancelled due to risk of ovarian hyperstimulation syndrome (OHSS).

IUI was performed 36 h after HCG administration in both groups. The insemination was performed via a polyethylene catheter (Frydman catheter; Imesa, Barcelona, Spain), with the woman in lithotomy position, as previously reported (Matorras et al., 1996Go). After 20 min of resting in the supine position, the patients resumed normal activity.

In the FSH group, luteal phase was supplemented as follows: with HCG in cases with estradiol levels <1500 pg/ml (2500 IU on day +1, +3, +5 and +7 following insemination) or with vaginal micronized progesterone in cases with estradiol levels >1500 pg/ml (Utrogestan; Laboratoires Besins-Iscovesco, Paris, France) (100 mg/12 h during the 15 days following insemination). In the CC group, all the women received HCG (2500 IU on day +1, +3, +5 and +7 following the last insemination).

Six IUI cycles were performed on all patients if pregnancy was not reached before. In FSH patients, where two consecutive cycles or three non-consecutive cycles were cancelled because of low or high response or due to falling estradiol levels, leuprolide acetate (Procrin; Abbott) in a long protocol was used in the remaining cycles. Low response was defined as failure to obtain >=2 follicles of >=16 mm of diameter and to reach 400 pg/ml of estradiol. High response was defined as estradiol levels >2000 pg/ml or >=6 follicles >=16 mm. Cycles with low or high response were cancelled.

The donors were unselectively used in both groups. The laboratory staff managing the semen samples did not know the allocation of the patients.

In all cases, pregnancy was defined by the visualization of a gestational sac at week 6 of amenorrhoea.

The sample size was calculated as follows: assuming per patient PR of 27% with CC and of 59% with FSH, to demonstrate that such PR were significantly different with a power of 90% and an alpha of 0.05 in a one case–one control design would require 48 CC patients and 48 FSH patients.

Statistical analysis was performed by means of {chi}2, Fisher's exact test, Mann–Whitney and Student t-tests following the standard criteria of applicability. Each parameter was tested by means of the odds ratio (OR) and its 95% confidence interval (CI). Statistical significance limit was defined as alpha = 0.05.

The ratios increase in PR/increase in multiple PR ({Delta} PR/{Delta} multiple PR) and {Delta} PR/{Delta} in triplet rate were used to quantify the risk of multiple pregnancy for each additional pregnancy obtained.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Homogeneity of the groups
Both groups were similar regarding the main demographic characteristics (Table IGo).


View this table:
[in this window]
[in a new window]
 
Table I. Main characteristics of clomiphene citrate and FSH groups
 
Outcome of the DI cycles (Table IIGo)
All women underwent a maximum of six programmed DI cycles if pregnancy was not reached before. The main characteristics of FSH cycles were: 31.9 ± 9.8 FSH ampoules used per cycle, 974 ± 492 pg/ml of estradiol the day of HCG and 3.2 ± 1.68 follicles >=17 mm. The per started cycle PR was significantly higher in FSH group 14.4% (30/209) versus 6.1% (16/261) (P = 0.004, {chi}2 = 8; OR = 2.6, 95% CI = 1.3–5.1). The per-started cycle PR with FSH was about double that of CC during the 6 months of treatment (Table IIIGo). The per-woman PR was also significantly higher in FSH group: 61.2% (30/49) versus 31.4% (16/51) (P = 0.005, {chi}2 =7.8; OR = 3.4, 95% CI = 1.4–8.6).


View this table:
[in this window]
[in a new window]
 
Table II. Comparison of the outcome of intrauterine donor insemination with clomiphene citrate and FSH
 

View this table:
[in this window]
[in a new window]
 
Table III. Per started cycle pregnancy rate (PR) and cumulative PR
 
In the CC group, three cycles were cancelled: one because of OHSS and two for non-medical reasons (one because of vacation, one due to concomitant disease). In the FSH group, there were 29 cycles cancelled: 12 because of low response, three due to hyperresponse, three for falling levels of estradiol and 11 for other reasons (six because of vacations, two driving problems, two concomitant disease, one forgetting to perform the monitoring). The cancellation rate per started cycle was 12.2% (29/238) in the FSH group versus 1.1% (3/264) in the CC group (P < 0.0001, {chi}2 = 23.8; OR = 12.1, 95% CI = 3.5–32.9). The per-started cycle PR was significantly higher in the FSH group 12.6% (30/238) versus 6.1% (16/264) in the CC group (P =0.02, {chi}2 = 5.78; OR = 2.2, 95% CI = 1.1–4.4). There were four cases of OHSS (one mild and three moderate) in the FSH group, two of them requiring hospitalization (4 days each). There was one case of mild OHSS in the CC group. The per cycle rate of OHSS was 4/209 (1.9%) with FSH versus 1/261 (0.4%) with CC (P > 0.05; OR = 5.1, 95% CI = 0.5–79.8).

Of the total number of pregnancies, 20% (6/30) were multiple with FSH versus 12.5% (2/16) with CC (P > 0.05; OR = 1.7, 95% CI = 0.26–202). All these cases were twins in the CC group whereas there were two cases of triplets in the FSH group (6.7% of pregnancies). The ratio {Delta} PR/{Delta} multiple PR was 3.6 and the ratio {Delta} PR/{Delta} in triplet rate was 7.2.

The abortion rate was similar in both groups: 26.6% with FSH, 8 (one of them a hydatiform mole)/30 versus 31.3% (5/16) with CC (P > 0.05). Regarding deliveries, 9.1% (1/9) were multiple with CC (twins) and 20.8% (5/24) with FSH (one case of triplets, one case of triplets spontaneously reducing to twins and three additional twins).

Secondary FSH group
Patients failing to obtain pregnancy with CC received six additional cycles with FSH. The main characteristics of secondary FSH cycles were: 29.5 ± 8.9 FSH ampoules employed per cycle, 927 ± 514 pg/ml of estradiol on the day of HCG and 3.01 ± 2.1 follicles >=17 mm.

The PR was 54.2% per patient (19/35) and 16.0% per cycle (19/119), with a multiple PR of 31.6% (6/19). The 31.6% of the pregnancies were multiple (6/19). There was one quadruplet, two triplet and three twin pregnancies. The abortion rate was 15.8% (3/19).

The cumulative PR for the women starting with CC stimulation was 69.3% (including those achieving pregnancy with CC and those with FSH, after failing with CC). From the 35 pregnancies obtained, the 22.9% were multiple (14.3% double, 5.7% triplet, 2.8% quadruplet).


    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Fresh sperm is no longer used in DI, because of the HIV risk. Since frozen–thawed sperm has considerably lower PR than fresh sperm (Smith et al., 1981Go; Richter et al., 1984Go; Brown et al., 1988Go; Subak et al., 1992Go; Matorras et al., 1996Go), a number of strategies have been designed to compensate for this reduction in PR. IUI is associated with higher PR than cervical insemination both in stimulated and unstimulated cycles (Matorras et al., 1996Go; O'Brien and Vandekerckove, 2000Go). With the combination of IUI and gonadotrophins, higher PR have been obtained than with cervical insemination and gonadotrophins (Wainer et al., 1995Go; Matorras et al., 1996Go). In a previous report analysing IUI–DI under gonadotrophin stimulation, the PR were 24% per cycle and 66% per woman, the cumulative PR being 86% (Matorras et al., 1996Go). The frequency of multiple pregnancy was high (29% of pregnancies), especially the frequency of triplets or higher order pregnancies (12.9% of pregnancies). Since multiple pregnancy carries a number of medical and social complications, there is an increasing concern to reduce the multiple pregnancy in assisted reproductive techniques. Moreover, gonadotrophin stimulation means an increased cost (drug and ovarian monitoring).

DI results differ widely, depending on donor quality, selection of women (age, infertility study, associated conditions), male diagnosis, insemination technique, freezing method and ovarian management. Thus, whereas some groups report good PR in unstimulated cycles, others report poor results. In our study, we compare the standard DI strategy in our centre (DI–IUI under gonadotrophin stimulation, with ultrasound and estradiol monitoring) with a simpler and less expensive strategy (DI–IUI with CC and HCG, without ovarian monitoring).

The CC group had a per-woman PR almost 50% lower than the gonadotrophin group, the per-cycle PR being 60% lower. A similar result has been reported with IUI using husband's sperm (Guzik et al., 1998). Regarding multiple pregnancy, although higher rates were found in FSH (20 versus 12.5%), statistical significance was not reached, probably due to the limited size of the series. Concerning high order multiple pregnancy, 6.7% of FSH pregnancies were triplets but none were in the CC group.

Although one could speculate that PR in the CC group could be improved if the ovarian cycle were monitored (more accurate timing of HCG and insemination, cancellation in low responders), our study deliberately omitted ovarian monitoring (on grounds of costs) in CC groups.

As expected, the cancellation rate was much higher with FSH. However, the per-started cycle PR remained much higher in the FSH group, and there was a trend toward a higher rate of OHSS, although its frequency was low and there were no severe cases.

In a first assessment of our findings, the management of IUI–DI with ovarian stimulation with FSH would seem a more advisable strategy than with CC, if PR (approximately double that of CC) are analysed. However, the multiple PR, especially concerning high order multiple pregnancy, was higher in the FSH group. In our opinion, the main parameter to be considered by both medical staff and patients is the ratio {Delta} PR/{Delta} multiple PR. This ratio is 3.6 for twins or higher order pregnancies. That means that for each additional multiple pregnancy associated with FSH therapy, 3.6 additional pregnancies were achieved (2.6 single, 1 multiple). However, double pregnancy risks are not excessive and twins are not seen by many couples as an adverse outcome, and can even be desired by some. Perhaps the most valuable parameter in taking a practical decision would be the ratio {Delta} PR/{Delta} high order multiple PR, this being 7.2 in our study. That means that for each triplet pregnancy obtained we would achieve 7.2 pregnancies (1 triplet, 6.2 single or twin).

For patients failing to obtain pregnancy, six additional FSH cycles were performed. The PR in this `secondary FSH' group were almost identical to those of primary FSH. Thus previous CC treatment does not imply that patients not reaching pregnancy will have a poor prognosis when receiving FSH later. On the other hand, there was a trend toward a higher multiple PR in secondary FSH. Since ovarian stimulation data were similar, we think this corresponds to the random fluctuation rates in small number series.

One of the main objectives of our study was to evaluate whether previous performance of DI–IUI under ovarian stimulation with CC could decrease the final multiple PR. However, our results were disappointing. In fact, with CC there was a trend toward a lower multiple PR than with FSH, but the PR was much lower. Thus, when patients failing to become pregnant with CC underwent FSH stimulation, they finally achieved a similar PR with a similar multiple PR.

In conclusion, each strategy has its advantages and disadvantages. Couples undergoing DI should be informed of the aforementioned differences and make their own choice. We would suggest starting with clomiphene and IUI in young women with good prognostic factors, whereas the remaining cases, as well as those failing to reach pregnancy with clomiphene, should receive IUI and gonadotrophins. However, in groups with high PR in unstimulated DI cycles, and in normal young patients, unstimulated DI could also be a reasonable alternative for the first three to six cycles.


    Notes
 
3 To whom correspondence should be addressed at: María Diaz de Haro 7, 6iz. 48013 Bilbao, Spain. E-mail: rmatorras{at}hcru.osakidetza.net Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
American Fertility Society (1990) New guidelines for the use of semen donor insemination. Fertil. Steril., 53 (Suppl.), 1–13.[ISI][Medline]

Brown, C.A., Boone, W.R. and Shapiro, S.S. (1988) Improved cryopreserved semen fecundability in an alternating fresh–frozen artificial insemination donor program. Fertil. Steril., 50, 825–827.[ISI][Medline]

Byrd, W., Bradshaw, K., Carr, B., Edman, C., Odom, J. and Ackerman, G. (1990) A prospective randomized study of pregnancy rates following intrauterine insemination using frozen donor sperm. Fertil. Steril., 53, 521–527.[ISI][Medline]

Guzick, D.S., Sullivan, M.W., Adamson, G.D., Cedars, M.I., Falk, R.J., Peterson, E.P. and Steinkampf, M.P. (1998) Efficacy of treatment for unexplained infertility. Fertil. Steril., 70, 207–213.[ISI][Medline]

Matorras, R., Gorostiaga, A., Diez, J., Corcóstegui, B., Pijoan, J.I., Ramón, O. and Rodríguez-Escudero, F.J. (1996) Intrauterine insemination with frozen sperm increases pregnancy rates in donor insemination cycles under gonadotropin stimulation. Fertil. Steril., 65, 620–625.[ISI][Medline]

O'Brien, P. and Vandekerckove, P. (2000) Intra-uterine versus cervical insemination of donor sperm for subfertility (Cochrane Review). In The Cochrane Library, 2. Update Software, Oxford.

Patton, P.E., Burry, K.A., Thurmond, A., Novy, M.J. and Wolf, D.P. (1992) Intrauterine insemination outperforms intracervical insemination in a randomized controlled study with frozen donor semen. Fertil. Steril., 57, 559–564.[ISI][Medline]

Richter, M.A., Haning, R.V., Jr and Shapiro, S.S. (1984) Artificial donor insemination: fresh versus frozen semen; the patient as her own control. Fertil. Steril., 41, 277–280.[ISI][Medline]

Smith, K.D., Rodriguez-Rigau, L.J. and Steinberger, E. (1981) The influence of ovulatory dysfunction and timing of insemination on the success of artificial insemination donor (AID) with fresh or cryopreserved semen. Fertil. Steril., 36, 496–502.[ISI][Medline]

Speroff, L., Glass, R.H. and Kase, N.G. (1999) Clinical Gynecologic Endocrinology and Infertility, 6th edn, in Spanish. Lippincott–Williams and Wilkins, pp. 1097–1132.

Subak, L.L., Adamson, G.D. and Boltz, N.L. (1992) Therapeutic donor semen insemination: a prospective trial of fresh versus frozen semen. Am. J. Obstet. Gynecol., 166, 1597–1604.[ISI][Medline]

Wainer, R., Merlet, F., Ducot, B., Bailly, M., Tribalt, S. and Lombroso, R. (1995) Prospective randomized comparison of intrauterine and intracervical insemination with donor spermatozoa. Hum. Reprod., 10, 2912–2922.

World Health Organization (1992) Laboratory Manual for the Examination of Human Semen–Cervical Mucus Interaction, 3rd edn. Cambridge University Press, Cambridge, pp. 43–45.

Submitted on December 1, 2001; resubmitted on March 27, 2002; accepted on April 11, 2002.