pronuclear stage transfer*

"

FERTILITY AND STERILITY Copyright

([J

Vol. 57, No.3, March 1992

1992 The American Fertility Society

Printed on acid~free paper in U.S.A.

After superovulation-intrauterine insemination fails: the prognosis for treatment by gamete intrafallopian transfer/pronuclear stage transfer*

Deirdre Robinson, M.D.t Craig H. Syrop, M.D.:!: Diane G. Hammitt, Ph.D. Department of Obstetrics and Gynecology, The University of Iowa Hospitals and Clinics, Iowa City, Iowa

Objective: To determine the prognosis for gamete intrafallopian transfer (GIFT)/pronuclear stage transfer (PROST) treatment after prior superovulation-intrauterine insemination (lUI). Design: Matched, retrospective. Setting: Outpatient university endocrine-infertility program. Patients, Participants: One hundred forty-four women matched for infertility factors and age were studied according to the following three treatment groups: superovulation-lUI only, GIFT/ PROST only, or GIFT/PROST after superovulation-lUI. Main Outcome Measures: Per cycle and cumulative pregnancy rates (PRs) were compared utilizing life table analysis. Results: Cumulative PRs (0.408) for superovulation-lUI only were lower than initial (0.469) and cumulative (0.802) cycle fecundity of GIFT/PROST (P = 0.002). Per cycle and cumulative PRs did not differ between GIFT/PROST only versus GIFT/PROST after superovulation-lUI. Conclusions: Gamete intrafallopian transfer/PROST may be cost-effective when compared with superovulation-lUI. The prognosis for GIFT/PROST success is not negatively affected by earlier Fertil Steril1992;57:606-12 superovulation-lUI treatment failure. Key Words: Superovulation-intrauterine insemination failure, prognosis for gamete intrafallopian transfer

The optimal method of achieving pregnancy in patients with nontubal infertility is controversial. Interest in ovarian superovulation and intrauterine insemination (lUI) and the availability of gamete intrafallopian transfer (GIFT) and pronuclear stage transfer (PROST) have created a need to evaluate the efficacy and cost associated with Received June 6, 1991; revised and accepted November 15, 1991. * Presented at the 38th Annual Meeting of the Society for Gynecologic Investigation, San Antonio, Texas, March 20 to 23, 1991. t Present address: Department of Obstetrics and Gynaecology, St. George's Hospital Medical School, London, United Kingdom. :j: Reprint requests: Craig H. Syrop, M.D., Department of Obstetrics and Gynecology, University of Iowa Hospitals and Clinics, 200 Hawkins Drive, Iowa City, Iowa 52242-1009.

606

Robinson et al.

The prognosis for GIFT/PROST

each treatment method. Pregnancy rates (PRs) for superovulation-lUI are reputed to approach that of GIFT. DespIte the advocated use of superovulation-lUI as a less expensive, near equivalent, less invasive alternative to GIFT, randomized-controlled trials comparing the fecundity of superovulation-lUI with GIFT have not been done. The ramifications of failure to achieve pregnancy with superovulation-lUI on future success of GIFT or PROST are unknown. This study, using matched contemporary patient groups with a broad spectrum of infertility factors, compares the PRs and fecundity of superovulation-lUI with GIFT or PROST and explores the influence of prior superovulation-lUI treatment failure on future GIFT/PROST success.

Fertility and Sterility

MATERIALS AND METHODS

Records of patients participating in the University of Iowa Reproductive Endocrinology/Infertility Clinic from January 1988 through June 1990 were reviewed for all cycles of superovulation -lUI, GIFT, or PROST. Infertility was identified as primary or secondary. Infertility factors included male, cervical, female antisperm antibodies (binding> 17% by indirect immunobead testing), mild tubal disease, endometriosis, and unexplained infertility. Male factor infertility included oligospermia «20 X 106 sperm/mL), asthenospermia «40% motility), autosperm antibodies (immunobead direct binding > 17% to any region except tail tip) , and/or infertile hamster egg sperm penetration assay «14% of 00cytes penetrated). Mild tubal disease was such that GIFT or PROST would not be precluded. Endometriosis was documented by laparoscopy or laparotomy and staged according to the revised American Fertility Society guidelines (1). Patients receiving donor oocytes were excluded. If more than one infertility factor was present, each was tabulated individually. Each cycle was reviewed for the occurrence of pregnancy, defined as rising serum human chorionic gonadotropin levels with a gestational sac documented by ultrasound. Biochemical pregnancies were not included in PRs. A spontaneous abortion was defined as a pregnancy loss before 20 weeks' gestation, excluding biochemical pregnancies. Patient groups consisted of patients completing GIFT and/or PROST cycles after treatment failure with superovulation-lUI and patients completing GIFT and/or PROST cycles without pretreatment with superovulation-lUI. A large third group ofpatients undergoing only superovulation-lUI was identified. From these, a control group was matched for infertility factors, including severity of endometriosis and male factor, and age within 3 years to the GIFT/PROST only group. Only patients receiving superovulation-lUI during a concurrent treatment time period were eligible for analysis. Pronuclear stage transfer was used to document in vitro fertilization (IVF) before tubal transfer in patients after failed GIFT, with male factor infertility, or the presence of antisperm antibodies. Life table analysis was used for evaluation of cycle and cumulative fecundity data. Ovarian stimulation and laboratory protocols for GIFT and PROST have been reported previously (2). Stimulations for superovulation-lUI cycles used

Vol. 57, No.3, March 1992

human menopausal gonadotropins (1 to 2 ampules/d) initiated between cycle days 3 and 5 and continued for 7 to 12 days until at least one but less than four follicles reached 18 mm in largest diameter. Human chorionic gonadotropin (10,000 U intramuscularly) was used in all cases to mimic a luteinizing hormone surge. Semen samples for superovulation-lUI were collected into sterile urine specimen containers and placed in a 37°C water bath for 20 minutes to allow liquefaction. Two parts of Biggers, Whitten, and Whittingham medium were mixed with one part semen and centrifuged at 300 X g for 8 minutes. The supernatant was removed and the pellet resuspended. Three parts medium were added to one part semen and centrifuged at 300 X g for 6 minutes. The pellet was resuspended for insemination in 0.5 to 0.7mL. RESULTS

A total of 144 study patients completed 257 total cycles of superovulation -lUI, 54 cycles of GIFT, and 88 cycles of PROST. Fifty-two patients completed GIFT and/or PROST cycles after failure to achieve a pregnancy with superovulation-lUI. Forty-six patients completed one or more GIFT and/or PROST cycles without pretreatment with superovulationlUI. Forty-six control patients receiving only superovulation-lUI were matched to those patients undergoing GIFT/PROST only cycles. All patient groups were mutually exclusive. The mean patient age for the GIFT/PROST after superovulation-lUI, GIFT/PROST only, and superovulation-lUI only groups was 33.5, 32.6, and 32.5 years, respectively. No significant difference was present. Patients receiving superovulation-lUI only treatment completed a mean of 2.07 (range 1 to 6) cycles. Couples who had failed to achieve pregnancy with superovulation-lUI and proceeded to GIFT or PROST cycles had completed a mean of 3.06 (range 1 to 10) cycles of superovulation-lUI. No statistical difference was present between the number of cycles completed in either group. The number of completed cycles of GIFT/PROST by patients pretreated with superovulation-lUI and those not previously treated by superovulation-lUI (1.47 versus 1.39) was not statistically different. Patients treated with superovulation-lUI only had a higher percentage of primary infertility (71.7%) than those treated with G1FT/PROST only (58.7%) or GIFT/PROST after superovulation-lUI (54.7%) (P < 0.05). No statistical difference was present between the latter two

Robinson et al.

The prognosis for GIFT/PROST

607

Table 1

Patient Characteristics

Mean age (y) Primary infertility (%) Mean cycles superovulations-lUI Mean cycles GIFT/ PROST

GIFT/PROST after superovulations-lUI (n = 52)

GIFT/PROST only (n = 46)

Superovulations-IUlonly (n = 46)

Probability

33.5 54.7

32.6 58.7

32.5 71.7t

<0.05

2.07 (1 to 6)

NS

3.06 (1 to 10):1: 1.47 (1 to 5)

NS*

1.39 (1 to 4)

* NS, not significant. t Significantly more superovulation-lUI only patients had pri-

NS

:I: Values in parentheses are ranges.

mary infertility than those treated by GIFT/PROST.

groups (Table 1). Although duration of infertility was not determined, it has been reported that outcome of superovulation-lUI is not related to length of infertility (3). Each infertility factor was equally represented in all study groups with the exception of female serum antisperm antibodies (n = 6), which were not present in the superovulation-lUI only group (Table 2). The most commonly treated infertility factors were male factor, endometriosis, and unexplained infertility. Of couples with male factor infertility, the same proportion of patients in all study groups elected to use husband sperm in treatment cycles (76.1% to 86.8%). Patient use of procedures is seen in Table 3. The use of either GIFT or PROST among patients pretreated with superovulation-lUI and those using GIFT/PROST only was not statistically different in each of the largest infertility groups. Pronuclear stage transfer was more commonly used for couples with male factor infertility than GIFT. Patients with unexplained infertility were significantly more likely to be treated with GIFT. Pregnancy rates per patient for male factor, endometriosis, and unexplained infertility were deterTable 2

Infertility Factors

Male factor Endometriosis Unexplained Antisperm antibodies Tubal disease Cervical

GIFT/PROST after superovulation-lUI

GIFT/PROST only

Superovulation-lUI only

Probability

27 (51.9)* (86.8)t 23 (44.2) 15 (28.8) 2 (3.8) 5 (9.6) 4 (7.7)

25 (54.3) (76.1)t 23 (50.0) 10 (21.7) 4 (8.7) 4 (8.7) 1 (2.2)

25 (54.3) (76.1)t 16 (34.8) 6 (13.0) 0(0.0) 4 (8.7) 2 (4.3)

NS:I: NS NS NS NS NS NS

* Values in parentheses are percents.

t Percent of cycles with husband's sperm.

608

mined for each study group. Couples with male factor infertility had a much higher PR using GIFT or PROST than superovulation-lUI alone. Pretreatment with superovulation-lUI did not alter PRs with GIFT/PROST. The patients with endometriosis or unexplained infertility tended toward higher PRs with GIFT/PROST, although no statistical difference was present. The remaining infertility factor groups were too small for statistical comparison, but a trend toward higher PRs with GIFT/PROST was present in all groups (Table 4). The spontaneous pregnancy loss rate for the GIFT/PROST only group and the GIFT/PROST after superovulation-lUI was not statistically different (12.5% versus 10.3%). No spontaneous abortions occurred in the superovulation-lUI only group. Life table analysis was performed for each study group (Table 5). Matched patients undergoing superovulation-lUI only therapy had a significantly lower initial cycle and cumulative fecundity as compared with the remaining two study groups using the Mantel-Haenszellife table test. The superovulation-lUI initial cycle fecundity was 0.213 as compared with 0.593 for patients who previously had failed superovulation-lUI and had undergone GIFT /

Robinson et aI.

The prognosis for GIFT/PROST

:I: NS, not significant.

Fertility and Sterility

Table 3

Use of GIFT or PROST* GIFT/PROST after superovulation -lUI

Male factor Endometriosis Unexplained

GIFT/PROST only

GIFT

PROST

GIFT

PROST

Probability

17.1 34.3 65.4

82.9 65.7 34.6

28.6 44.4 70.0

71.4 55.6 30.0

NSt NS NS

* Values are percents.

t NS, not significant.

PROST and 0.469 for the GIFT/PROST only group. The cumulative probability of achieving pregnancy after four or more cycles of superovulation-lUI was 0.408. This remains less than the initial cycle fecundity for either of the two remaining groups receiving GIFT/PROST. The cumulative probability of pregnancy for the GIFT/PROST only group and the patients using GIFT/PROST after superovulation-lUI was 0.744 and 0.802, respectively, after three or more completed treatment cycles. In Figure 1, the plots of cumulative PRs are not significantly different between these two study groups (P = 0.47) but are significantly higher than the superovulation-lUI only group (P = 0.002).

in patients with oligospermia, cervical factor, unexplained infertility, ovulatory dysfunction, and endometriosis. The best success has been found with cervical factor infertility and the least improvement with male factor infertility (5-10). Serhal et al. (5) reported a 26.4% per cycle PR for unexplained infertility patients treated with superovulation-lUI. Dodson et al. (6) evaluated the efficacy of superovulation-lUI in a broader patient population reporting cycle fecundities for endometriosis (0.17), cervical factor (0.29), and unexplained infertility (0.19). The overall initial cycle fecundity was 0.14 with a cumulative fecundity of 0.35 that plateaued at three cycles. In general, cumulative PRs for superovulation-lUI have not improved after four treatment cycles (11-12). By comparing superovulation-lUI outcomes with reported GIFT PRs of27% to 30%, both Serhal et al. (5) and Dodson et al. (6) advocated the use of superovulation -lUI before proceeding to more costly and invasive procedures such as GIFT. Neither of these studies included patients with male infertility, control study groups, or compared superovulation-lUI PRs with their own center's matched GIFT patient populations. No randomized controlled trials comparing superovulation-lUI and GIFT have been published. Comparisons between studies and institutions are

DISCUSSION

Superovulation -lUI has gained popularity as a method of treating nontubal causes of infertility. The development of GIFT, PROST, and IVF-embryo transfer (ET) has led to advances in sperm preparation and ovulation induction. Intrauterine insemination alone has been associated with PRs ranging from 0% to 62% when used for a variety of treatment indications (4). The use of superovulation in combination with lUI has led to improved PRs Table 4

Pregnancy Rates by Diagnosis GIFT/PROST after superovulation -lUI

Male factor Endometriosis Unexplained Antisperm antibodies Tubal disease Cervical

23 18 10 2 4 1

GIFT/PROST only

(82.1)* (80.0) (66.7) (100.0) (80.0) (25.0)

* Values in parentheses are percents. t Significantly more GIFT/PROST male factor patients conceived compared with those receiving superovulation-lUI only.

Vol. 57, No.3, March 1992

22 16 5 1 1 2

Superovulation-lUI only

Probability

8 (30.8)t 7 (41.2) 4 (40.0) o (0.0) o (0.0) o (0.0)

<0.001 NS:j: NS NS NS NS

(78.6) (69.2) (83.3) (25.0) (25.0) (100.0) :j: NS, not significant.

Robinson et al.

The prognosis for GIFT/PROST

609

Table 5

Life Table Analysis GIFT/PROST after superovulation-lUI

GIFT/PROST only

Per cycle fecundity

Cumulative fecundity

Per cycle fecundity

Cumulative fecundity

Per cycle fecundity

Cumulative fecundity

0.593 0.222 0.375

0.593 0.683 0.802

0.469 0.324 0.286

0.469 0.642 0.744

0.213 0.123 0.000 0.143

0.213 0.310 0.310 0.408

Cycle 1 2 >3 >4

difficult because of potential differences in patient selection, sperm preparation, laboratory techniques, and ovulation induction methods. Although the superovulation-lUI only group has a lower (but not significantly) percentage of unexplained infertility patients, our G1FT/PROST study group is otherwise well matched to control superovulation-lUI patients receiving concurrent treatment using the same laboratory and clinical protocols. The 1989 IVF-ET Registry reported an overall 30% clinical PR per GIFT cycle for 133 participating clinics (13). Gamete intrafallopian transfer has been used in patients with oligospermic infertility, although success, especially with severe oligospermia using standard methods, has been poor. Improved success is seen with modifications to increase the number of sperm transferred (14-16). Pronuclear stage transfer has been developed, in response to lower PRs with standard GIFT techniques, to demonstrate IVF before tubal transfer. Pregnancy rates

1.0 . , - - - - - , - - - - , - - - - - ,

Cumulative

O.B

Probability of Conception

0.6

0.4

0.2

0.0

...L-_ _ _--'-_ _ _ _--'-_ _ _- - '

2

3

>4

Number of Cycles

Figure 1 Probability of conception by treatment group. 0, superovulation-lUI only; e, GIFT/PROST only; T, GIFT/PROST after superovulation-lUI.

610

Superovulation-lUI only

Robinson et al.

The prognosis for GIFT/PROST

of 25.7% to 32% per ET have been reported in patients with oligospermia undergoing PROST procedures (17, 18). Pronuclear stage transfer was preferred in our study patients with male factor infertility. Although primary infertility is noted to be more highly represented in the superovulation-lUI group, the difference in outcome between the use of superovulation -lUI and G 1FT/PROST in similar patient populations is remarkable. The use of GIFT/ PROST produced higher PRs per patient for all infertility factors with a significant improvement for male factor infertility. The use of life table analysis (19, 20) effectively demonstrates the superiority of GIFT/PROST per cycle and cumulatively within the confines of this study design. Our study population reveals a plateau in cumulative fecundity for superovulation-lUI after two cycles, with a slight increase after four or more treatment cycles. The fecundity remains less than that of a single GIFT/PROST cycle. In our institution, the financial cost of four cycles of superovulation-lUI is essentially equivalent to one cycle of GIFT. Therefore, the cost of achieving a 40% chance of pregnancy with either method is similar. Continuation with GIFT/PROST cycles led to a 0.744 to 0.802 cumulative fecundity after three or more cycles. Even if a higher spontaneous abortion rate is present with GIFT/PROST, the overall outcome remains better than with superovulation-lUI. The reason for treatment success or failure for each method is difficult to discern. The success of superovulation-lUI may be related to increasing numbers of gametes available at the site of fertilization, improving ovulation defects, and improving the timing of insemination (5, 6). Gamete intrafallop ian transfer and PROST can also increase the numbers of gametes for fertilization, as well as allowing for evaluation of morphology and function.

Fertility and Sterility

Gamete intrafallopian transfer and PROST may overcome oocyte abnormalities, defective ovum pickup, unruptured follicle syndrome, unrecognized tubal defects, or oocyte entrapment after follicle rupture. Replacing controlled numbers of gametes into the fallopian tubes decreases the chance of multiple gestations (5, 21, 22). A retrospective multivariate analysis comparing PRs for superovulation-lUI and GIFT was performed by Kaplan (22). Although retrospective and without control groups, an increased chance of pregnancy with either treatment method was present with the absence of endometriosis, a total motile sperm count of greater than 30 X 106 , and a duration of infertility of <3 years. The use of G1FT was also associated with a greater chance of pregnancy. Because reasons for achieving pregnancy with any treatment method remain speculative, it is important to discern if success with GIFT/PROST is merely because of patient selection bias. Is the higher GIFT/PROST fecundity because of inclusion of patients whose infertility is relatively easy to overcome and who may have had success with another method such as superovulation-lUI? Specifically, the inclusion of a group of study patients who represent treatment failures with superovulation-lUI provides insight to this question. The distribution of infertility factors for the GIFT/PROST after superovu1ation -lUI group is similar to that of the GIFT / PROST only group. Despite prior failure with superovulation -lUI, the cycle and cumulative fecundity' as well as PRs by infertility factor, are identical to those patients undergoing GIFT/PROST without pretreatment with superovulation-lUI. Therefore, GIFT /PROST fecundity was not inflated by eliminating pretreatment with superovulation-lUI. Conversely, the chance of GIFT/PROST success was not negatively affected by earlier superovulation -lUI treatment failure. This proposes that there is (are) underlying infertility factor(s) that cannot be corrected with superovulation-lUI that can be overcome with GIFT/PROST. Our results indicate that the use of G1FT/PROST provides an efficient, cost-effective means of achieving pregnancy. An important finding is that pretreatment with superovulation-lUI before GIFT/ PROST does not adversely affect outcome and allows for an equal chance of success with GIFT / PROST. The use of G1FT/PROST requires surgical expertise and a high caliber gamete laboratory. The surgical and anesthetic risks associated with laparoscopy cannot be discounted. However, the use of

Vol. 57, No.3, March 1992

superovulation-lUI should not be mandated before using GIFT/PROST. Careful consideration of all options, including the initial use of GIFT or PROST, should be made when establishing a treatment plan. Importantly, patients should not be discouraged that prior superovulation-lUI treatment failure will adversely affect subsequent GIFT/PROST outcome. Acknowledgment. The authors thank Catherine C. Hood, B.S., M.S., (Department of Obstetrics and Gynecology, Reproductive Testing Laboratories, University of Iowa Hospitals and Clinics, Iowa City, Iowa) for statistical support.

REFERENCES 1. The American Fertility Society. Revised American Fertility Society classification of endometriosis. 1985. Fertil Steril 1985;43:351-2. 2. Hammitt DG, Syrop CH, Hahn SJ, Walker DL, Butkowski CL, Donovan JF. Comparison of concurrent pregnancy rates for in vitro fertilization-embryo transfer, pronuclear stage embryo transfer and gamete intrafallopian transfer. Hum Reprod 1990;5:947-54. 3. Dodson WC, Haney AF. Controlled ovarian hyperstimulation and intrauterine insemination for treatment of infertility. Fertil Steril 1991;55:457-67. 4. Allen NC, Herbert CM, Maxson WS, Rogers BJ, Diamond MP, Wentz AC. Intrauterine insemination: a critical review. Fertil Steril 1985;44:569-80. 5. Serhal PF, Katz M, Little V, Woronowski H. Unexplained infertility-the value of Pergonal superovulation combined with intrauterine insemination. Fertil Steril 1988;49:602-6. 6. Dodson WC, Whitesides DB, Hughes CL Jr, Easley HA III, Haney AF. Superovulation with intrauterine insemination in the treatment of infertility: a possible alternative to gamete intrafallopian transfer and in vitro fertilization. Fertil Steril 1987;48:441-5. 7. Corson SL, Batzer FR, Gocial B, Maislin G. Intrauterine insemination and ovulation stimulation as treatment of infertility. J Reprod Med 1989;34:397-406. 8. Kemmann E, Bohrer M, Shelden R, Fiasconaro G, Beardsley L. Active ovulation management increases the monthly probability of pregnancy occurrence in ovulatory women who receive intrauterine insemination. Fertil Steril 1987;48:91620. 9. Melis GB, Paoletti AM, Strigini F, Fabris FM, Canale D, Fioretti P. Pharmacologic induction of multiple follicular development improves the success rate of artificial insemination with husband's semen in couples with male-related or unexplained infertility. Fertil SteriI1987;47:441-5. 10. Margalloth E, Sauter E, Bronson RA, Rosenfeld DL, Scholl GM, Cooper GW. Intrauterine insemination as treatment for antisperm antibodies in the female. Fertil Steril 1988;50: 441-6. 11. Chaffkin LM, Nulsen JC, Luciano AA, Metzger DA. A comparative analysis of the cycle fecundity rates associated with combined human menopausal gonadotropin (hMG) and intrauterine insemination (lUI) versus either hMG or lUI alone. Fertil Steril 1991;55:252-7.

Robinson et al.

The prognosis for GIFT/PROST

611

12. Remohi J, Gastaldi C, Patrizio P, Gerli S, Ord T, Asch RH, et al. Intrauterine insemination and controlled ovarian hyperstimulation in cycles before GIFT. Hum Reprod 1989;4: 918-20. 13. Medical Research International, Society for Assisted Reproductive Technology, The American Fertility Society. In vitro fertilization-embryo transfer (IVF -ET) in the United States: 1989 results from the IVF-ET registry. Fertil Steril1991;55: 14-23. 14. Yovich JL, Matson PL. The influence of infertility etiology on the outcome ofIVF -ET and GIFT treatments. Int J Fertil 1990;35:26-32. 15. Matson PL, Blackledge DG, Richardson PA, Turner SR, Yovich JM, Yovich JL. The role of gamete intrafallopian transfer (GIFT) in the treatment of oligospermic infertility. Fertil Steril1987;48:608-12. 16. Wiedemann R, Noss U, Hepp H. Gamete intrafallopian transfer in male subfertility. Hum Reprod 1989;4:408-11. 17. Balmaceda JP, Gastaldi C, Remohi J, Borrero C, Ord T, Asch

612

Robinson et aI.

The progrwsis for GIFT/PROST

18.

19.

20.

21.

22.

RH. Tubal embryo transfer as a treatment for infertility due to male factor. Fertil Steril1989;50:476-9. Yovich JL, Blackledge DG, Richardson PA, Matson PL, Turner SR, Draper R. Pregnancies following pronuclear stage tubal transfer. Fertil Steril1987;48:851-7. Lalich RA, Marut EL, Prins GS, Scommegna A. Life table analysis of intrauterine insemination pregnancy rates. Am J Obstet Gynecol 1988;158:980-4. Cramer DW, Walker AM, Schiff 1. Statistical methods in evaluating the outcome of infertility therapy. Fertil Steril 1979;32:80-6. Cefalu E, Cittadini E, Balmeceda J, Guastella G, Ord T, Rojas FJ, et al. Successful gamete intrafallopian transfer following failed artificial insemination by donor: evidence for a defect in gamete transport? Fertil Steril 1988;50:279-82. Kaplan CR, Olive DL, Sabella V, Asch RM, Balmaceda JP, Riehl RM, et al. Gamete intrafallopian transfer vs superovulation with intrauterine insemination for the treatment of infertility. J In Vitro Fert Embryo Transfer 1989;6:298-304.

Fertility and Sterility