|
|
| |||||||||||||||||||||||||||||||||||||||||||||||
Background Induction of superovulation with gonadotropins and intrauterine insemination are frequently used to treat infertility. We conducted a large, randomized, controlled clinical trial of these treatments.
Methods We studied 932 couples in which the woman had no identifiable infertility factor and the man had motile sperm. The couples were randomly assigned to receive intracervical insemination, intrauterine insemination, superovulation and intracervical insemination, or superovulation and intrauterine insemination. Treatment continued for four cycles unless pregnancy was achieved.
Results The 231 couples in the group treated with superovulation and intrauterine insemination had a higher rate of pregnancy (33 percent) than the 234 couples in the intrauterine-insemination group (18 percent), the 234 couples in the group treated with superovulation and intracervical insemination (19 percent), or the 233 couples in the intracervical-insemination group (10 percent). Stratified, discrete-time Cox proportional-hazards analysis showed that the couples in the group treated with superovulation and intrauterine insemination were 3.2 times as likely to become pregnant as those in the intracervical-insemination group (95 percent confidence interval, 2.0 to 5.3) and 1.7 times as likely as those in the intrauterine-insemination group (95 percent confidence interval, 1.2 to 2.6). The couples in the intrauterine-insemination group and in the group treated with superovulation and intracervical insemination were nearly twice as likely to conceive as those in the intracervical-insemination group.
Conclusions Among infertile couples, treatment with induction of superovulation and intrauterine insemination is three times as likely to result in pregnancy as is intracervical insemination and twice as likely to result in pregnancy as is treatment with either superovulation and intracervical insemination or intrauterine insemination alone.
Superovulation and intrauterine insemination are used alone or in combination for the treatment of unexplained infertility, male-factor infertility, and other cases of infertility in which the woman has an unobstructed genital tract and some ovarian function and the man has motile sperm. Although national data on the cost of these procedures have not been compiled, costs of $1,300 per cycle for superovulation and of $500 per cycle for intrauterine insemination are typical.1 The risks of superovulation include ovarian hyperstimulation,2 multiple pregnancy,3 and possibly, an increase in the risk of ovarian cancer.4
We report the results of a large, randomized, controlled clinical trial of the efficacy of superovulation and intrauterine insemination.
Methods
Centers and Subjects
This study was conducted at 10 clinical sites and was approved by the appropriate institutional review committee at each site. All the couples gave informed consent to participate.
Before enrollment, each couple underwent a standard evaluation for infertility, including semen analysis in the man; endometrial biopsy, hysterosalpingography, and laparoscopy in the woman; and a test for serum antisperm antibodies in the couple. Semen analysis was performed according to standardized methods5,6 by trained technicians at each center, who followed common protocols and used the same types of supplies and equipment. The immunobead test used to detect antisperm antibodies was a modification of the test of the World Health Organization.6 The results of an endometrial biopsy were considered to be "in phase" if there was no more than a three-day lag in histologic dating according to the onset of the next menstrual period. Women who had received treatment for American Fertility Society stage I or II endometriosis7 were enrolled only if six months had elapsed after either surgical therapy or the return of ovulatory cycles after medical therapy. Detailed inclusion and exclusion criteria are listed in Table 1.
|
The overall hypothesis of the study was that induction of superovulation or intrauterine insemination would result in a higher pregnancy rate among infertile couples than would no treatment. No treatment was defined as intracervical insemination timed to a surge in the urinary excretion of luteinizing hormone to ensure that sperm were present in the cervix and vaginal fornix at the time of ovulation. Eligible couples were randomly assigned to one of four groups: intracervical insemination timed to the surge in urinary excretion of luteinizing hormone (the control group), intrauterine insemination timed to the surge in urinary excretion of luteinizing hormone, superovulation and intracervical insemination, or superovulation and intrauterine insemination. Purified follicle-stimulating hormone (urofollitropin [Metrodin], Serono Laboratories, Norwell, Mass.) was used to induce superovulation.
Each couple received four treatment cycles unless pregnancy occurred. Rest cycles sometimes intervened between treatment cycles for either personal or clinical reasons. In superovulation cycles, treatment was canceled after day 3 if the serum estradiol concentration exceeded 3000 pg per milliliter (11,010 pmol per liter). Cycles were canceled in the intracervical-insemination and intrauterine-insemination groups if no surge in urinary excretion of luteinizing hormone was detected.
Procedures
After eligibility and informed consent were confirmed for each couple, randomization was carried out with use of a permuted-block procedure, stratified according to center.
Superovulation Protocol
Women assigned to one of the two superovulation groups were treated according to a standard protocol. Base-line pelvic ultrasonography was performed on day 1, 2, or 3 of the menstrual cycle. Then, 150 IU of follicle-stimulating hormone was administered intramuscularly daily from day 3 through day 7. On day 8, ultrasonography was repeated and serum estradiol was measured. Daily administration of follicle-stimulating hormone was continued, with the dose adjusted if necessary, until at least two follicles reached 
18 mm (average of two dimensions) and the serum estradiol concentration ranged from 500 to 3000 pg per milliliter (1835 to 11,010 pmol per liter). Once these criteria were met, treatment with follicle-stimulating hormone was discontinued and 10,000 IU of human chorionic gonadotropin (Profasi, Serono Laboratories) was administered intramuscularly. A single insemination was performed 36 to 40 hours later.
Spontaneous-Ovulation Protocol
Women who were not assigned to superovulation underwent insemination timed to spontaneous ovulation. Four days before the expected time of ovulation, the women began daily testing of their second morning urine specimen for luteinizing hormone, using a qualitative kit (OvuQuick, Quidel, San Diego, Calif.). Intrauterine insemination or intracervical insemination was performed on the day after the surge in urinary excretion of luteinizing hormone.
Intrauterine-Insemination Protocol
Semen specimens were collected by masturbation on site, evaluated according to standardized methods,5,6 and prepared for intrauterine insemination within one hour after collection. Semen was diluted 1:2 (vol/vol) with HEPES-buffered Ham's F10 medium containing 1.5 percent serum albumin (Albuminar, Bayer, Elkhart, Ind.). After centrifugation at 250xg for 10 minutes, the pellets were resuspended and combined in 3 ml of the medium. The sperm suspension was centrifuged for 10 minutes, and the pellet was resuspended in 0.35 ml of medium. Approximately 0.05 ml was used to determine the concentration and motility of the sperm. The remaining sample was drawn into a Shepard catheter (Cook Ob/Gyn, Bloomington, Ind.) attached to a 1-ml syringe for intrauterine insemination, which was performed within 2 1/2 hours after semen collection.
Intracervical-Insemination Protocol
Semen specimens were collected and evaluated as in the intrauterine-insemination protocol. Intracervical insemination was performed within 1 1/2 hours after semen collection.
Definition of Pregnancy
Serum ß-human chorionic gonadotropin was measured 15 days after insemination (luteal day 15). If the value exceeded 10 mIU per milliliter, the measurement was repeated on luteal day 17. Pregnancy was indicated by an increase in the serum ß-human chorionic gonadotropin concentration between luteal days 15 and 17. We subsequently also recorded the outcome of pregnancy — spontaneous abortion, induced abortion, ectopic pregnancy, or live birth — and whether multiple pregnancy had occurred.
Data Management and Quality Control
Data were transmitted to the data-coordinating center before the enrollment of each couple and periodically thereafter. Research nurses and members of the data-entry staff attended two-day training sessions at the data-coordinating center. The technicians who performed semen analyses attended five-day training sessions at the University of California, Davis, and thereafter were tested for proficiency approximately twice each year with the use of unmarked specimens. All centers were audited on site by the study coordinator once a year for compliance with the protocol and the completeness and comparability of data with the master data base. Computerized checks of the validity of the data were performed. Quality-control analyses of serum ß-human chorionic gonadotropin and serum estradiol measurements were conducted annually by the central laboratory at Columbia University by having each clinical laboratory analyze unmarked samples containing various concentrations of each hormone.
Statistical Analysis
The pregnancy rates per couple were calculated for each of the four groups, and the results were compared with the use of chi-square tests with one degree of freedom. The Bonferroni correction8 for multiple comparisons was used to adjust the alpha level for statistical significance. These analyses were repeated after adjustment for clinical center. We also tested whether the fecundity rates at the centers were homogeneous.
The possible effects of rest cycles were accounted for by the use of stratified, discrete-time Cox proportional-hazards analysis.9 The results are presented for a model stratified according to center. However, results from an unstratified model, which included the center as a covariate, and from an unstratified model that did not include centers were similar, indicating that there was no interaction between treatments and centers.
The results of a planned interim analysis were reported to an independent data safety and monitoring committee but not to the investigators or other personnel. The alpha level was adjusted at the time of the interim analysis according to the method of O'Brien and Fleming.10
Results
Characteristics of the Couples
The number of couples enrolled in each of the four treatment groups ranged from 231 to 234. The base-line characteristics of the couples in each group were similar (Table 2). Overall, the 932 couples enrolled in the study were observed for 4676 cycles (Table 3), of which 57 percent were treatment cycles, 18 percent were clinically mandated rest cycles, 18 percent were rest cycles requested by the couples, and 6 percent were cycles in which treatment was canceled. Most clinically mandated rest cycles in the superovulation groups were the result of the detection of ovarian cysts at the beginning of a cycle. The number of cycles of canceled treatment was higher in the groups that did not involve hormone therapy, largely because of the failure of the women to detect a surge in urinary excretion of luteinizing hormone. The rate of withdrawal from the study was higher among the couples in the superovulation groups (Table 3).
|
|
|
36 months of infertility. The rates were higher among women with a previous pregnancy (23 percent) than among those who had never been pregnant (18 percent).
|
We also compared the number of liveborn infants in each of the treatment groups (Table 6). Liveborn infants were more common in the groups treated with superovulation and intracervical insemination (P=0.01) and superovulation and intrauterine insemination (P<0.001) than in the intracervical-insemination group and were more common (P=0.01) in the group treated with superovulation and intrauterine insemination than in the group treated with intrauterine insemination alone.
|
Discussion
Induction of superovulation and intrauterine insemination are common treatments for couples with unexplained infertility and those with infertility associated with specific diagnoses in which no pregnancy has occurred despite standard diagnosis-directed treatments. We studied only couples in which evaluation of the woman revealed no abnormalities and that therefore would be considered to have either unexplained infertility or male-factor infertility. In these couples, the combination of the induction of superovulation with follicle-stimulating hormone and intrauterine insemination was associated with a likelihood of pregnancy that was more than three times as high as that for intracervical insemination alone.
Although to our knowledge there have been no previous large-scale, randomized comparisons of these treatments in couples with unexplained or male-factor infertility, a recent randomized trial comparing superovulation and intrauterine insemination with no treatment among women with minimal or mild endometriosis revealed that the use of superovulation and intrauterine insemination was associated with a birth rate that was 5.6 times as high as the rate in the absence of treatment.11 A combined analysis of the literature on unexplained infertility12 yielded estimated pregnancy rates of 4 percent per cycle for control cycles and intrauterine-insemination cycles, 8 percent per cycle for superovulation cycles, and 18 percent per cycle for cycles of superovulation and intrauterine insemination. We expected somewhat lower rates in our study, because we included men who had any motile sperm in their ejaculates. Some provocative results from this study regarding characteristics of semen were that intrauterine insemination had more of an effect on the likelihood of pregnancy for couples in which the man's semen values were in the lower quartiles and that superovulation had less of an effect for these couples.
The apparent variations in the rate of miscarriage among the groups deserve comment. The miscarriage rate was highest in the group treated with superovulation and intrauterine insemination, but the numerator was small. The fact that the rates of miscarriage were not lower in the group treated with superovulation and intracervical insemination and the intrauterine-insemination group than in the intracervical-insemination group suggests that there is nothing about superovulation or intrauterine insemination itself that leads to an increased likelihood of miscarriage. Moreover, there is no evidence or theoretical basis to suggest that superovulation and intrauterine insemination interact synergistically with respect to miscarriage.
We conclude that for infertile couples in which the woman has no identifiable infertility factor and the man has motile sperm, the combination of superovulation and intrauterine insemination is an effective means of achieving pregnancy. Moreover, the effects of superovulation and intrauterine insemination on pregnancy appear to be independent and additive. In recommending treatment options to couples, physicians should weigh these results against those for in vitro fertilization; they should also consider the costs of the various procedures, the results of semen analyses, the woman's age, and the incidence of ovarian hyperstimulation and high-order multiple pregnancies. At present, no rigid algorithm can be constructed. We recommend that couples be informed of all their options, be given realistic information about the chances of success as well as the costs and complications, and be involved in the final decision about which treatment method to use.
Supported in part by Cooperative Agreements with the National Institute of Child Health and Human Development (U10 HD26975, U10 HD26981, U10 HD27006, U10 HD27009, U10 HD27001, U10 HD27049, U10 HD33172, and U10 HD33173) and by Serono Laboratories.
Drs. Carson and Coutifaris have served as consultants to Serono Laboratories.
We are indebted to Serono Laboratories for supplying urofollitropin and human chorionic gonadotropin; to Quidel for supplying ovulation-detection kits; to Brian Little, M.D., who chaired the National Cooperative Reproductive Medicine Network Steering Committee; to Joseph Fleiss, Ph.D., for critical help in the initial design of the study; to Patricia Kringas, R.N., M.A., for coordinating the study; to Ms. Rosemary Coslit for data management; to Ms. Charlene Brazil for coordinating the andrology laboratory; to Myunghee Cho Paik, Ph.D., and Bruce Levin, Ph.D., for statistical consultation; to Maureen Hatch, Ph.D., and Anne Golden, Ph.D., for early epidemiologic consultation; to Donald J. McMahon, M.S., for coordinating computer services; and to John F. O'Connor, Ph.D., for laboratory quality control.
Source Information
From the University of Rochester, Rochester, N.Y. (D.S.G.); Baylor College of Medicine, Houston (S.A.C., J.E.B.); University of Pennsylvania Medical Center, Philadelphia (C.C., L.M.); the University of California, Davis, Sacramento (J.W.O., S.T.N.); Columbia University, New York (P.F.-L., R.E.C.); the University of Alabama, Birmingham (M.P.S.); Harvard University, Boston (J.A.H.); and the National Institutes of Health, Bethesda, Md. (D.L.V.).
Address reprint requests to Dr. Carson at Baylor College of Medicine, Department of Obstetrics and Gynecology, 6550 Fannin #801, Houston, TX 77030, or at scarson{at}bcm.tmc.edu.
References
In addition to the authors, other investigators of the National Cooperative Reproductive Medicine Network were as follows: Baylor College of Medicine, Houston — P. Casson, S. Lindsey; Brigham and Women's Hospital, Boston — K. Walsh, M. Rein; Tufts University, Boston — A. DeCherney; University of Alabama, Birmingham — R. Blackwell, E. Knochenhauer, K. Hammond, V. Willis; University of California, Davis, Sacramento — S. Boyers, M. Colombo, J. D'Amico, J. Chang, R. Covell, K. Sweeney, L. Wisner; University of Pennsylvania, Philadelphia — K. Timbers, J. Stansberry, L. Blasco, K. Walsh; University of Pittsburgh, Pittsburgh — J. Albert, S. Berga, K. Baffone, M. Everson, M. McQueen; University of Rochester, Rochester, N.Y. — G. Centola, W. Phipps, G. Santoriello; and University of Tennessee, Memphis — A. Milem; Data Safety and Monitoring Committee — J. Schreiber, S. Fowler, G. Colditz, T.L. Bush.
| |||||||||||||||||||||||||||||||||||||||||||||||
Related Letters:
Efficacy of Superovulation and Intrauterine Insemination in the Treatment of Infertility
Dickey R., Sartor S. M., Pryzak R., Habbema J.D.F., Carson S. A., Guzick D. S., Vogel D. L., te Velde E. R., Cohlen B. J.
Extract |
Full Text
N Engl J Med 1999;
341:128-129, Jul 8, 1999.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
Previous
