Vaginal misoprostol enhances intrauterine insemination

Samuel E. Brown1,3,5, James P. Toner1,4, John A. Schnorr1, Shaun C. Williams1, William E. Gibbons1, Dominique de Ziegler2 and Sergio Oehninger1

1 Department of Obstetrics and Gynecology, The Jones Institute for Reproductive Medicine, Eastern Virginia Medical School, Norfolk, VA 23507, USA and 2 Hôpital de Nyon, Nyon and University Hospital Geneva, Geneva, Switzerland


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
This study examined whether the prostaglandin E1 analogue misoprostol (400 µg), when placed vaginally at the time of intrauterine insemination (IUI) improves pregnancy rates. A prospective, placebo-controlled, randomized and double-blind study involving 274 women in 494 IUI cycles resulted in a total of 64 pregnancies (13% per cycle). Misoprostol cycles totalled 253, with 43 pregnancies (17% per cycle), whereas placebo cycles totalled 241, with 21 pregnancies (9% per cycle). The cumulative pregnancy rate with misoprostol treatment was significantly greater than with placebo (P = 0.004, Cox proportional hazards regression). The benefit of misoprostol was seen in clomiphene cycles (14 versus 4%, P = 0.006), and was indicated in FSH cycles (33 versus 15%, borderline significance) and natural cycles (15.6 versus 7.7%, not significant), but was not seen in clomiphene/FSH cycles (18.2 versus 23.5%, not significant). Misoprostol treatment did not increase pain score on the day of IUI (1.1 versus 1.4) and at 1 day post IUI (0.6 versus 0.8). Complications were rare in both groups [six (2%) subject cycles in the misoprostol cycles compared with two (1%) in the placebo group]. It is concluded that the use of vaginal misoprostol may improve the chance for pregnancy in women having IUI in a wide variety of cycle types.

Key words: intrauterine insemination/misoprostol/pregnancy/prostaglandin/vaginal


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Human semen contains large amounts of prostaglandins (PG) (Templeton et al., 1978Go; Bendvold et al., 1984Go), the main source being the seminal vesicles (Gottlieb et al., 1991Go). The most dominant PG in the human ejaculate are PGE and PGF and 19-hydroxylated PG (Bendvold et al., 1986Go). It has been suggested (Asplund, 1947Go) that there is a link between PG and fertility, since lower concentrations of PG were found in the seminal fluid of men from infertile couples (Asplund, 1947Go). A further study showed that PGE specifically was decreased in men from infertile couples (Bygdeman et al., 1970Go). PG increase myometrial contractility and isthmic tubal relaxation, affect luteal maintenance, immunosuppression and enhance spermatozoon–oocyte binding—all properties that may account for their fertility-enhancing effects (Karlsson, 1959Go; Coutinho and Maia, 1971Go; Henzl et al., 1972Go; Aitken and Kelly, 1985Go; Skibinski et al., 1992Go; Thibodeaux et al., 1992Go).

Intrauterine insemination (IUI) has an important role in the treatment of infertile couples. Although intracytoplasmic sperm injection (ICSI) has become a formidable treatment for patients with severe oligoasthenoteratozoospermia, there is a real role of IUI as a clinical treatment modality (Oehninger, 1996Go; Oehninger et al., 1997Go). IUI involves the direct transfer of washed spermatozoa to the intrauterine space, thus by-passing potential vaginal–cervical sperm barriers and increasing sperm concentration near the site of fertilization in the Fallopian tubes. It is known that painful uterine cramping occurs when PG from semen are placed in the intrauterine space. Consequently, semen preparation for IUI deliberately eliminates PG from the inseminate. In natural reproduction, spermatozoa migrate from the vagina into the cervical mucus, and thus leave naturally produced PG and other substances in the vagina. During standard IUI, these other seminal constituents (perhaps PG) are not present and therefore their positive effects on fertility are absent.

Misoprostol is a commercially available synthetic PG that is structurally related to PGE1 (Collins et al., 1985Go) and was used initially to prevent peptic ulcer disease induced by chronic ingestion of non-steroidal anti-inflammatory drugs (NSAID) (Dajani and Nissen, 1985Go). Misoprostol has also been used extensively as an oral and intravaginal abortifacient, and most recently as a medical aid to evacuate early pregnancy failures in humans (Creinin and Darney, 1993GoPeyron et al., 1993Go). Misoprostol has also been used as a safe and inexpensive medication for cervical ripening and labour induction in human pregnancies (Sanchez-Ramos et al., 1997Go).

Due to the reported safety, widespread commercial availability and relative similarity of misoprostol to PGE, the usefulness of vaginally placed misoprostol as adjunctive therapy at the time of IUI was investigated, and its tolerability and effects on clinical pregnancy rates assessed.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Protocol
All women presenting to the Jones Institute for Reproductive Medicine for IUI between February 1998 and March 2000 were offered study entry. Subjects were excluded if they had any history of allergic response or sensitivity to misoprostol or PG, or any known history of kidney or liver disorders. Subjects were instructed to avoid NSAID, antihistamines and coitus from 72 h before expected insemination until 72 h after insemination. Subjects not willing to abstain from coital activity were excluded from the study participation. This study was approved by the Institutional Review Board of Eastern Virginia Medical School (#11-11-97-0111).

Infertility diagnoses were classified as: (i) male factor, if semen analyses revealed oligozoospermia (sperm concentration <20x106 spermatozoa/ml), asthenozoospermia (<50% progressively motile spermatozoa), and/or teratozoospermia [normal strict morphology 4–14%, thereby eliminating `poor prognosis' patients with <4% normal forms (Acosta and Kruger, 1996Go)]; (ii) endometriosis (minimal–mild) confirmed by laparoscopic visualization and ablated by laser or electrocautery; (iii) polycystic ovarian syndrome (PCOS), when anovulatory cycles were coupled with any manifestation of hyperandrogenism with the exclusion of those with signs/symptoms of congenital adrenal hyperplasia or other endocrinopathies (Dunaif et al., 1987Go; ACOG, 1995); (iv) oligo-ovulation or anovulatory cycles without signs or symptoms of hyperandrogenism and hypergonadotrophism; (v) tubal disease, when laparoscopy or hysterosalpingography (HSG) defined at least one abnormal Fallopian tube (in such cases, the disease was corrected by laparoscopy or laparotomy to the minimum end point of having at least one patent tube); (vi) uterine disease, abnormalities found on HSG (i.e. polyps or submucosal fibroids) with further hysteroscopic surgical correction; (vii) decreased ovarian reserve, those patients with cycle day 3 serum FSH concentrations >=10 mIU/ml (MEIA; Abbott IMx-System®, Abbott Park, IL, USA) or a high oestradiol concentration (>90 pg/ml; MEIA); and (viii) unexplained infertility, as defined by the absence of abnormal results in the above diagnostic evaluations.

The decision to initiate IUI as a treatment for infertility was made by the attending physician. Subjects underwent one of the following treatments: natural cycle, clomiphene citrate, gonadotrophin or clomiphene/gonadotrophin stimulation. Natural cycles were used in a minority of cycles when mild spermatozoa or spermatozoa–cervical mucus problems were identified. Clomiphene citrate was used empirically or when anovulation was part of the infertility diagnosis, and was started at 50 mg/day and given on cycle days 5–9 or 3–7. The dosage was increased in subsequent cycles until ovulation was achieved, and was then repeated for up to four to six ovulatory cycles. IUI was offered only to those with ovulatory cycles while taking clomiphene. Ovulatory cycles were those defined as having either a positive urine luteinizing hormone (LH) kit test or those having ovulation trigger with human chorionic gonadotrophin (HCG) injection (as defined below).

The decision to initiate gonadotrophins or clomiphene/gonadotrophin combined ovarian stimulation with IUI treatment was usually made after previous treatments with clomiphene citrate coupled with IUI had failed to produce pregnancy. When the decision was made to start ovarian stimulation/IUI, purified FSH (Metrodin® or Fertinex®; Serono Labs Inc., Norwell, MA, USA) or recombinant FSH (Gonal-F®; Serono) was initiated on day 3 of a spontaneous menstrual cycle or induced menstrual bleeding. Cycles were monitored with serial serum oestradiol and LH concentrations, as well as ovarian follicular size determinations by ultrasound. No cycles employed pituitary desensitization with gonadotrophin-releasing hormone agonist (Dodson et al., 1991Go). Cycles began with 1–2 ampoules of gonadotrophin and continued on a step-down fashion, depending on patient response. Patients without PCOS were started with 2–3 ampoules [human menopausal gonadotrophin (HMG), FSH or combined] and followed in a step-down fashion via individualized protocols. Some cycles in patients with PCOS (n = 67) utilized the combination of clomiphene and gonadotrophins, and in these cases clomiphene 100 mg was taken daily from cycle day 3 to day 7; initiation of daily gonadotrophin injections was on cycle day 9. The starting dose of clomiphene would vary depending on three differing predicted ovarian responses to stimulation (high responders with 50 mg/day, normal responders with 100 mg/day, and low responders with 150 mg/day). Ovulation was triggered by i.m. injection of 10 000 IU of HCG when the lead follicle(s) reached a diameter of 18–20 mm for clomiphene/FSH cycles. IUI was then performed ~36 h later. Patients were counselled on the risks of multiple gestation when more than one mature follicle was triggered, and our defined stimulation technique generally resulted in one to three mature follicles per cycle. Although rare, cycles with more than three mature follicles were not cancelled.

Semen samples for IUI were collected and prepared by sperm washing (human tubal fluid and 0.2% human serum albumin; Irvine Scientific, Santa Ana, CA, USA), and with further processing by the use of gradient centrifugation (Percoll or ISolate®) in cases of leukospermia (white blood cell concentration >1x106/ml with negative semen cultures) or positive antisperm antibodies (via direct immunobead testing) (Irianni et al., 1993Go; Ombelet et al., 1997Go). Donor sperm cases were included (see Table IVGo). IUI was performed using a UnisemTM intrauterine cannula (Unimar®; Cooper Surgical, Shelton, CT, USA) or ShepardTM intrauterine insemination catheter (Cook®, Spencer, IN, USA) in cases of stenosis. All patients utilising gonadotrophins (including clomiphene/gonadotrophin cycles) had luteal support with vaginal progesterone suppositories (50 mg daily) starting 48 h post insemination. Luteal support was continued until serum or urinary pregnancy evaluation ~16 days after IUI, and when pregnant, through gestational week 10.


View this table:
[in this window]
[in a new window]
 
Table IV. Semen parameters and insemination catheter use in misoprostol and placebo groups
 
Assignment/masking/participant flow/follow-up
Prior to speculum removal after IUI, an opaque white study suppository was placed in the posterior vaginal fornix. Each suppository contained either placebo or 400 µg of misoprostol; there was no visible difference between active and placebo suppositories. The suppository base was inert triglyceride. Two different labelled suppository batches were made at a time (n = 10), and content information was retained by the pharmacy (pharmacy blinding). In addition to the batch randomization, coin-flip randomization was performed by the clinic nurse immediately before suppository use. Repeat insemination cycles per subject were re-randomized. Patients remained supine for 15 min after IUI, and then resumed normal activities. Subjects were contacted by telephone for follow-up data concerning side effects, bleeding and analogue pain scale determinations (0–10) (Salomaki et al., 1996Go; Klopfenstein et al., 2000Go).

Pregnancy determinations were made by HCG titre evaluation at ~16 days post insemination. Positive pregnancy tests were further followed by transvaginal ultrasonography at 4–6 weeks from insemination. Those with gestational sacs with cardiac activity were defined as clinical pregnancies.

Statistical analysis
Pregnancy rates in PGE1 and placebo cycles were compared using Cox proportional hazards regression with the Anderson–Gill counting process notation to account for repeated cycles of some patients (Anderson and Gill, 1982Go; Fleming and Harrington, 1991). Patients may have multiple events as well as cycle-dependent covariates. Differences in patient demographic and clinical variables between PGE1 and placebo cycles were analysed using a generalized estimating equation (GEE) approach to clustered data (Diggle et al., 1994Go).


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
In total, 274 women consented for study participation, entailing 494 IUI cycles that resulted in a total of 64 pregnancies. Misoprostol cycles totalled 253 with 43 pregnancies (17% per cycle), whereas placebo cycles totalled 241 with 21 pregnancies (9% per cycle). The cumulative pregnancy rate of misoprostol patients was significantly greater than that of placebo patients (P = 0.004, by Cox proportional hazard regression). The estimated cumulative rates from the Cox model for both groups are shown in Figure 1Go. The following serial IUI attempts were included in the analysis: 269 first; 127 second; 59 third; 27 fourth; seven fifth; three sixth; two seventh; and one eighth.



View larger version (12K):
[in this window]
[in a new window]
 
Figure 1. PGE1 and placebo cumulative pregnancy rates for all cycles.

 
Pregnancy rates per cycle for misoprostol and placebo cycles by cycle type are listed in Table IGo. Although the pregnancy rate of the misoprostol treatment with natural cycle was greater than that for placebo, the difference was not statistically significant (Figure 2Go). In the natural cycles, essentially all of the pregnancies for both the misoprostol and placebo groups occurred in the first cycle. The pregnancy rate in clomiphene-stimulated misoprostol-treated cycles was significantly greater than that of placebo (P = 0.006, Figure 3Go). The pregnancy rate of the gonadotrophin (FSH)-stimulated misoprostol cycles was greater than that of placebo, but the difference was of borderline significance (Figure 4Go). For FSH/clomiphene-stimulated cycles, there was no significant difference between misoprostol-treated and placebo pregnancy rates (Figure 5Go). Stimulation data could not be determined accurately for 10 cycles due to recording/chart error.


View this table:
[in this window]
[in a new window]
 
Table I. Pregnancy rate (%) by treatment and cycle type
 


View larger version (12K):
[in this window]
[in a new window]
 
Figure 2. PGE1 and placebo cumulative pregnancy rates for natural cycles.

 


View larger version (13K):
[in this window]
[in a new window]
 
Figure 3. PGE1 and placebo cumulative pregnancy rates for clomiphene-stimulated cycles.

 


View larger version (12K):
[in this window]
[in a new window]
 
Figure 4. PGE1 and placebo cumulative pregnancy rates for FSH-stimulated cycles.

 


View larger version (13K):
[in this window]
[in a new window]
 
Figure 5. PGE1 and placebo cumulative pregnancy rates for FSH/clomiphene-stimulated cycles.

 
Comparisons of demographic variables between misoprostol and placebo cycles are shown in Table IIGo. No significant differences were noted between the active treatment (misoprostol) and placebo groups regarding age, gravidity, parity, body mass index (BMI), duration of infertility, cycle day of insemination, previous number of inseminations, and use of luteal progesterone support. Some 83% of subjects were Caucasian, 12% were African American, and 5% were Asian. None of the above-mentioned variables was significantly associated with pregnancy status.


View this table:
[in this window]
[in a new window]
 
Table II. Subject demographic data
 
The distribution of the differing infertility diagnoses between the misoprostol and placebo groups is shown in Table IIIGo. With regard to infertility diagnoses, there was a significant difference between misoprostol and placebo cycle groups with respect to uterine factor, unexplained and pelvic adhesive disease. Some 4% (10/244) of misoprostol cycles were uterine factor, compared with 1% (2/240) of placebo (P = 0.025). In addition, 12% (30/244) of misoprostol cycles had unexplained infertility diagnosis compared with 20% (49/240) of placebo cycles (P = 0.016). Less than 1% (1/244) of misoprostol cycles had a diagnosis of pelvic adhesive disease, compared with 4% (9/240) of the placebo cycles (P = 0.006). In order to determine if any of the diagnoses influenced pregnancy along with misoprostol, Cox regression models with treatment and each variable as a covariate were calculated. For all diagnoses, treatment (PGE1 versus placebo) remained statistically significant. None of the infertility diagnoses had a significant effect on pregnancy.


View this table:
[in this window]
[in a new window]
 
Table III. Infertility diagnoses in misoprostol and placebo groups (%)
 
The association of pregnancy and the semen parameters, IUI catheters and cervical tenaculum use is shown in Table IVGo. Tenaculum use during IUI was utilized significantly more in the misoprostol group [17% (41/243) as compared with the placebo group (10% (23/230), P = 0.013], but overall, tenaculum use was not significantly associated with pregnancy.

No significant differences were noted between the active treatment and placebo cycles regarding subjective side effects, complications and pain variables (Table VGo), and none of these variables was associated with pregnancy. The placebo group showed a non-significant increased pain score on the day of IUI (1.4 versus 1.1) and day post IUI (0.8 versus 0.6). Complications occurred in six (2%) cycles in the active treatment regimen [severe pelvic pain (n = 4) and fever >100.4°F (n = 2)] and in two (1%) cycles in the placebo group (severe pain). All but one of the pain episodes occurred >2 h post IUI, and were completely resolved within 48 h. One case occurred immediately from the time of IUI, even before suppository placement. All fevers resolved within 24 h, and none required antibiotic use. Vaginal spotting after IUI occurred in 23 (9%) active treatment cycles versus 15 (6%) in the placebo group. Pregnancy information was not available from three IUI cycles due to lack of follow-up and inability to contact the subjects, and therefore were not included in the analysis.


View this table:
[in this window]
[in a new window]
 
Table V. Subjective side effects and complications
 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Of all body fluids, human seminal plasma possesses the highest concentration of PG (Mann and Lutwak-Mann, 1981Go). The total PG content of the average human male ejaculate is 1 mg, and comprises PGE and PGF together with their 19-hydroxylated derivatives. The most dominant active prostaglandin, PGE, has a mean semen concentration of 73.2 µg/ml, and notably high inter-individual variation (range 2–272 µg/ml) (Templeton et al., 1978Go; Bendvold et al., 1984Go, 1986Go). PG were first described in 1947 (Asplund, 1947Go), with lower concentrations of PGE being found in couples with unexplained infertility. However, the precise physiological effect of a decrease in PGE mediating infertility in these couples has never been determined (Bygdeman et al., 1970Go).

PG are known to enhance sperm transport and increase the fertilization rate in rabbits (Mandl, 1972Go). Novel animal studies involving the use of intrauterine PGE infusion have resulted in the maintenance of corpora luteal function and stimulation of progesterone production, ensuring uterine receptivity for pregnancy (Thibodeaux et al., 1992Go).

Human in-vitro research has shown that PGE induces a relaxation response on the non-pregnant human uterine and Fallopian tube smooth muscle, whereas PGF has been shown in vitro to create a contractile response. The in-vivo effects from both PG are stimulatory on the myometrium (Ingelman-Sundburg, 1968Go; Coutinho and Maia, 1971Go). Moreover, it has also been shown that PGE is more potent than PGF on myometrial response and that both PG inhibit tubal motility, thus suggesting that the relaxation of the tubal isthmus is a prerequisite for sperm penetration into the Fallopian tube (Coutinho and Maia, 1971Go). Additional effects such as immunosuppression afforded by seminal PG has been shown both in vivo and in vitro, suggesting an attenuated female immunological response to spermatozoa (Skibinski et al., 1992Go). Strikingly PGE, but not PGF, has been shown to improve significantly the ability of human spermatozoa to penetrate zona-free hamster oocytes (Aitken and Kelly, 1985Go).

Human in-vivo research has shown the potential benefit of vaginally placed PG in the assistance of reproductive success. In 1959, it was noted that the most common reaction to intravaginal administration of seminal fluid was contraction of the uterine corpus, with relaxation of the cervical canal (Karlsson, 1959Go). Enhanced passage of spermatozoa through cervical mucus has been noted in the in-vivo use of PG (Eskin et al., 1973Go). Additionally, intravenous injection of PGE1 has been shown to stimulate contractility of the non-pregnant human uterus, assisting sperm movement to the Fallopian tubes (Roth-Brandel et al., 1970Go). Cervical relaxation, improved spermatozoa–mucus penetration, enhanced uterine contractility, tubal relaxation, and improved spermatozoa–zona binding all appear to be related to the normal deposition of PGE into the female reproductive tract.

Misoprostol [(±)-methyl-11{alpha}, 16-dihydroxy-16-methyl-9-oxoprost-13-E-en-1-oate] is a synthetic PG structurally related to PGE1 (Collins et al., 1985Go). The compound was used initially to prevent peptic ulcer disease induced by the chronic ingestion of NSAID. Since its commercial introduction, misoprostol has been reported as an effective oral and intravaginal abortifacient, and most recently as a medical aid to evacuate early pregnancy failures in humans (Creinin and Darney, 1993Go; Peyron et al., 1993Go). Misoprostol has also been shown to be an effective agent for cervical ripening and labour induction (Sanchez-Ramos et al., 1997Go).

Due to these favourable effects of PG in regard to the potentiation of fertilization, the utility of augmenting IUI with vaginally placed PGE1 analogue (misoprostol) was investigated with regard to pregnancy success and complications. A significant increase was found in the pregnancy rates of women undergoing ovarian stimulation with clomiphene citrate, and there was a trend toward increasing pregnancy rates both in natural cycles and in women where gonadotrophins were used for stimulation, when vaginal misoprostol was placed at the time of insemination.

The positive effect of vaginal misoprostol leading to an improved clinical pregnancy rate is not clearly understood. As mentioned previously, the known effects of PGE on increasing myometrial contractility, potential relaxation of tubal isthmus, improved spermatozoon–oocyte binding/penetration and attenuation of the female immune response to spermatozoa may all facilitate fertilization potential. It is not understood why a statistically significant increased pregnancy rate was found in the clomiphene/gonadotrophin-only stimulated cycles, and not in the natural, gonadotrophin-only cycles or combination clomiphene/gonadotrophin cycles. Due to this large study cohort involving a predominance of clomiphene cycles, it is speculated that the improved clinical pregnancy rate would eventually be seen in all groups (power limiting) if numbers of cycles were increased. The precise physiological mechanism(s) of how misoprostol leads to improved clinical pregnancy rates requires further research.

The findings presented here support the hypothesis that vaginal PG are important in assisting the process of human fertilization and establishment of pregnancy. However, a causal link remains to be established and further studies are needed to determine the precise role of PG in early human reproductive events.


    Notes
 
3 Present address: Florida Institute for Reproductive Medicine, 836 Baptist Medical Center Pavilion, Jacksonville, FL 32207, USA Back

4 Present address: Atlanta Center for Reproductive Medicine, Woodstock, GA 30189, USA Back

5 To whom correspondence should be addressed at: Florida Institute for Reproductive Medicine, 836 Baptist Medical Center Pavilion, Jacksonville, FL 32207, USA. E-mail: sambrown99{at}yahoo.com Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
ACOG Technical Bulletin (1995) Polycystic ovarian syndrome.

Acosta, A.A. and Kruger, T.F. (eds) (1996) Human Spermatozoa in Assisted Reproduction. The Parthenon Publishing Group, New York, pp. 489–513.

Aitken, R.J. and Kelly, R.W. (1985) Analysis of the direct effects of prostaglandins on human sperm function. J. Reprod. Fertil., 73, 139–146.[Abstract]

Anderson, P.K. and Gill, R.D. (1982) Cox's regression model for counting processes: a large sample study. Ann. Statist., 10, 1100–1120.[ISI]

Asplund, J. (1947) A quantitative determination of the content of contractive substances in human sperm and their significance for the motility and vitality of the spermatozoa. Acta Physiol. Scand., 13, 103.[ISI]

Bendvold, E., Svanborg, K., Eneroth, P. et al. (1984) The natural variations in prostaglandin concentration in human seminal fluid and its relation to sperm quality. Fertil. Steril., 41, 743–747.[ISI][Medline]

Bendvold, E., Gottlieb, C., Svanborg, K. et al. (1986) Absence of prostaglandins in semen of men with cystic fibrosis is an indication of the contribution of the seminal vesicles. J. Reprod. Fertil., 78, 311–314.[Abstract]

Bygdeman, M., Fredricsson, B., Svanborg, K. et al. (1970) The relation between fertility and prostaglandin content of seminal fluid in man. Fertil. Steril., 21, 622–629.[ISI][Medline]

Collins, P.W., Pappo, R. and Dajani, E.Z. (1985) Chemistry and synthetic development of misoprostol. Dig. Dis. Sci., 30, 114S–117S.

Coutinho, E.M. and Maia, H.S. (1971) The contractile response of the human uterus, fallopian tubes, and ovary to prostaglandins in vivo. Fertil. Steril., 22, 539–543.[ISI][Medline]

Creinin, M.D. and Darney, P.D. (1993) Methotrexate and misoprostol for early abortion [published erratum appears in Contraception, 49, 99 (1994)]]. Contraception, 48, 339–348.

Dajani, E.Z. and Nissen, C.H. (1985) Gastrointestinal cytoprotective effects of misoprostol. Clinical efficacy overview. Dig. Dis. Sci., 30, 194S–200S.[Medline]

Diggle, P.J., Liang, K.-Y. and Zeger, S.L. (1994) Analysis of Longitudinal Data. Oxford University Press, New York.

Dodson, W.C., Walmer, D.K., Hughes, C.L., Jr et al. (1991) Adjunctive leuprolide therapy does not improve cycle fecundity in controlled ovarian hyperstimulation and intrauterine insemination of subfertile women. Obstet. Gynecol., 78, 187–190.[Abstract]

Dunaif, A., Graf, M., Mandeli, J. et al. (1987) Characterization of groups of hyperandrogenic women with acanthosis nigricans, impaired glucose tolerance, and/or hyperinsulinemia. J. Clin. Endocrinol. Metab., 65, 499–507.[Abstract]

Eskin, B.A., Azarbal, S., Sepic, R. et al. (1973) In vitro responses of the spermatozoa-cervical mucus system treated with prostaglandin (F2a). Obstet. Gynecol., 41, 436–439.[ISI][Medline]

Fleming, T. and Harington, D. (1991) Counting Processes and Survival Analysis. John Wiley and Sons, New York.

Gottlieb, C., Ploen, L., Kvist, U. et al. (1991) The fertility potential of male cystic fibrosis patients. Int. J. Androl., 14, 437–440.[ISI][Medline]

Henzl, M.R., Noriega, L., Aznar, R. et al. (1972) The uterine effects of vaginally administered prostaglandin E2. Prostaglandins, 1, 205–215.[Medline]

Ingelman-Sundburg, M. (1968) The Prostaglandins. Little, Boston.

Irianni, F.M., Ramey, J., Vaintraub, M.T. et al. (1993) Therapeutic intrauterine insemination improves with gonadotropin ovarian stimulation. Arch. Androl., 31, 55–62.[ISI][Medline]

Karlsson, T. (1959) The influence of seminal fluid on the motility of the nonpregnant human uterus. Acta Obstet. Gynecol. Scand., 38, 503–521.[Medline]

Klopfenstein, C.E., Herrmann, F.R., Mamie, C. et al. (2000) Pain intensity and pain relief after surgery. A comparison between patients' reported assessments and nurses' and physicians' observations. Acta Anaesthesiol. Scand., 44, 58–62.[ISI][Medline]

Mandl, J.P. (1972) The effect of prostaglandin E on rabbit sperm transport in vivo. J. Reprod. Fertil., 31, 263–269.[Medline]

Mann, T. and Lutwak-Mann, C. (1981) Biochemistry of seminal plasma and male accessory fluid. In Male Reproductive Function and Semen. Springer-Verlag, New York, pp. 312–463.

Oehninger, S. (1996) Intracytoplasmic sperm injection: results from Norfolk, USA. Hum. Reprod., 11, (Suppl. 1)73–75.[Medline]

Oehninger, S., Franken, D. and Kruger, T. (1997) Approaching the next millennium: how should we manage andrology diagnosis in the intracytoplasmic sperm injection era? (editorial). Fertil. Steril., 67, 434–436.[ISI][Medline]

Ombelet, W.C.M. et al. (1997) Human Spermatozoa in Assisted Reproduction. The Parthenon Publishing Group, New York.

Peyron, R., Aubeny, E., Targosz, V. et al. (1993) Early termination of pregnancy with mifepristone (RU 486) and the orally active prostaglandin misoprostol (see comments). N. Engl. J. Med., 328, 1509–1513.[Abstract/Free Full Text]

Roth-Brandel, U., Bygdeman, M. and Wiqvist, N. (1970) Effect of intravenous administration of prostaglandin E1, and F2 on the contractility of the non-pregnant human uterus in vivo. Acta Obstet. Gynecol. Scand. Suppl., 5, 19–25.[Medline]

Salomaki, T.E., Kokki, H., Turunen, M. et al. (1996) Introducing epidural fentanyl for on-ward pain relief after major surgery. Acta Anaesthesiol. Scand., 40, 704–709.[ISI][Medline]

Sanchez-Ramos, L., Kaunitz, A.M., Wears, R.L. et al. (1997) Misoprostol for cervical ripening and labor induction: a meta-analysis. Obstet. Gynecol., 89, 633–642.[Abstract/Free Full Text]

Skibinski, G., Kelly, R.W., Harrison, C.M. et al. (1992) Relative immunosuppressive activity of human seminal prostaglandins. J. Reprod. Immunol., 22, 185–195.[ISI][Medline]

Templeton, A.A., Cooper, I. and Kelly, R.W. (1978) Prostaglandin concentrations in the semen of fertile men. J. Reprod. Fertil., 52, 147–150.[Abstract]

Thibodeaux, J.K., Myers, M.W., Roussel, J.D. et al. (1992) Intrauterine infusion of prostaglandin E2 and subsequent luteal function in cattle. Prostaglandins, 44, 531–541.[Medline]

Submitted on June 19, 2000; accepted on September 28, 2000.