F.Zayed 1 ' 3, E.A.Lenton 1 ' 2 and I.D.Cooke 2

Transcription

1 Human Reproduction vol.12 no. 11 pp , 1997 Comparison between stimulated in-vitro fertilization and stimulated intrauterine insemination for the treatment of unexplained and mild male factor infertility F.Zayed 1 ' 3, E.A.Lenton 1 ' 2 and I.D.Cooke 2 'Sheffield Fertility Centre, 26 Glen Road, Sheffield S7 IRA and 2 University Department of Obstetrics and Gynaecology, Jessop Hospital for Women, Sheffield S3 7RE, UK 3 To whom correspondence should be addressed at: Department of Obstetrics and Gynaecology, Jordan University of Science and Technology, P.O. Box , Amman Jordan A prospective trial was undertaken to evaluate the efficacy of stimulated in-vitro fertilization (SIVF) and stimulated intrauterine insemination (SIUI) in couples with unexplained and mild male factor infertility. In all, 80 couples were allocated to treatment with SIVF or SIUI, both treatments following the same protocol [clomiphene citrate and follicle stimulating hormone (FSH) injection], except that higher doses of FSH were used in the SIVF treatment cycles. Initially, 41 couples were allocated to and started treatment with SIVF but eight cases were eventually converted to SIUI because of under-response. Similarly, although 39 couples were initially allocated to SIUI treatment, five of these converted to SIVF because of over-response. The treatment cycles that were converted either to SIUI or to SIVF were not considered as treatment failures but as treatment changes and so were included in the analyses. Of the final 38 SIVF cycles, four were cancelled (dysfunctional response), failed fertilization occurred in five cycles and 29 subjects reached embryo transfer. There were two biochemical pregnancies [positive human chorionic gonadotrophin (HCG) only], two clinical abortions and seven live births. Of the final 42 SIUI cycles, only two were cancelled, insemination being performed in the remaining 40 cases. The result was one clinical abortion, three ectopics and eight live births. The proportion of cycles with positive HCG was identical (28.9% per cycle treated for SIVF and 28.6% for SIUI) and the livebirth rates were also not different (18.4% per cycle treated for SIVF and 19.0% for SIUI). The cost per maternity of SIUI was approximately half that of SIVF ( 1923 versus 4611) and so we conclude that, as SIUI had an efficacy that was not significantly different from SIVF (using similar protocols) but was more cost-effective, it must be considered the more appropriate form of management for the treatment of unexplained and mild male factor infertility. Indeed, it is hard to justify the routine use of IVF, as a first approach, in unexplained infertility. Key words: intrauterine insemination/in-vitro fertilization/ mild male factor/unexplained infertility Introduction In the early days of stimulated in-vitro fertilization (SIVF), one of the common ovarian stimulation regimes utilized clomiphene citrate and human menopausal gonadotrophin (HMG). Clomiphene citrate, an oestrogen antagonist, in conjunction with HMG stimulates and sustains the growth of the spontaneously recruited follicles and good results were obtained provided monitoring of endogenous luteinizing hormone (LH) secretion during the late follicular phase was employed. Macnamee et al. (1988) showed that if an LH surge was detected and supported by human chorionic gonadotrophin (HCG) administration, then treatment outcome was not compromised. The main difficulty was that a proportion of patients had a premature or atypical LH surge. If this went undetected, ovulation may have already occurred prior to the time set for egg retrieval, or only poor quality eggs that did not fertilize well were obtained (Stanger and Yovich, 1985; Homburg et al, 1988). Significant simplification of ovarian stimulation came about with the introduction of gonadotrophin-releasing hormone analogues (GnRHa), which effectively prevented the LH surge. There were disadvantages, such as a higher incidence of ovarian hyperstimulation syndrome (OHSS), and, more generally, the requirement for greater amounts of exogenous gonadotrophins to bring about an equivalent follicular response. The consequence was that down-regulated IVF treatment cycles took longer and were more expensive than the earlier clomiphene citrate/hmg regimens but were easier to deliver in a clinical setting. However, it is quite possible that such an aggressive approach to the management of patients with unexplained infertility (patent tubes, regular ovulation and a normal male partner) is strictly unnecessary (Dodson et al, 1991; Sengoku et al, 1994). It has been reported that stimulated intrauterine insemination (SIUI) yields improved pregnancy rates compared with ovarian stimulation and timed intercourse (Chung et al, 1995) and is of value as a first line of therapy in patients with unexplained infertility, but patients who have not conceived after several attempts at SIUI may still become pregnant following SIVF (Ranieri et al, 1995). This prospective study set out to examine the hypothesis that, in patients with unexplained or mild male factor infertility, SIUI treatment, while not necessarily more efficient (higher pregnancy rates), is at least more cost-effective than SIVF treatment. For the comparison to be valid, similar protocols [using clomiphene citrate and follicle stimulating hormone (FSH)] were used for both treatments European Society for Human Reproduction and Embryology

2 SIVF versus SIUI in unexplained infertility Table I. Comparison of the semen analyses of patients described as having unexplained infertility and those with mild male factor infertility. Mild male factor was diagnosed when one semen parameter only was below the normal limits for our laboratory. Isolated defects in each variable (concentration, motility and morphology) were detected. Results are expressed as geometric means with 68% confidence intervals in parentheses Diagnostic group and treatment plan (n) Concentration (X10 6 /ml) Motility (%) Morphology (%) Unexplained infertility SIVF (33) SIUI(27) Mild male factor SIVF (8) SIUI(12) 55 (21-132) 42 (20-141) 12 (8-18) 11 (7-19) 45 (31-67) 41 (30-70) 18 (6-29) 16 (2-28) 43 (26-64) 39 (25-65) 16 (9-24) 15 (6-24) SIVF = stimulated in-vitro fertilization; SIUI = stimulated intrauterine insemination. Materials and methods Subjects Between September 1994 and September 1995, all couples with unexplained or mild male factor infertility who fulfilled the criteria were recruited. None of the women had had previous infertility treatment, all were <40 years old, with a minimum duration of infertility of 2 years. They had ovulatory cycles confirmed by a recent ovulatory mid-luteal phase progesterone measurement and bilateral tubal patency, which had been confirmed by either a recent laparoscopy or hysterosalpingogram. Couples with unexplained infertility had semen analyses with a sperm concentration 5=20 XlO 6 /ml, motility (grade I+II) 2*30% and normal forms 3*25%. Couples with mild male factor infertility had semen analyses where one only of the above parameters was below the normal range. In no case was any variable <5 X10 6 spermatozoa/ml (concentration) or <5% (motility or morphology) (Table I). These mild male factor patients were included in the study because we have shown that their rate of failed fertilization is no different from couples with unexplained infertility (using either natural cycle IVF or conventional down-regulated stimulated IVF to assess fertilization; unpublished observations). Couples were randomly allocated to one of two treatment groups prior to their first consultation. At the initial consultation (all performed by F.Z.) it was explained to patients that their treatment would be some form of assisted conception, either SIVF or SIUI, and the treatment they ultimately received would to some extent be influenced by their response to stimulation. Patients were prepared for either treatment and were given implications counselling for both IUI and IVF. Three patients were excluded as they declined to take part in a study. A few couples indicated a preference to be in a different group from the one allocated, generally for financial reasons as SIUI is delivered at a lower cost than SIVF. The wishes of four patients from each arm of the study who wanted to change their allocated treatment were respected, particularly as it was anticipated that a number of other patients would also have their groups changed once treatment was underway, as a consequence of their response to stimulation. Thus, this study represents only pseudorandomization in the allocation of treatments. The drugs used were the same for both treatments, except that lower amounts of FSH (Metrodin HP; Serono Laboratories Ltd, Hertfordshire, UK) were used for SIUI than for SIVF because the intention was to recruit fewer follicles in the SIUI group. Treatment protocol Blood was obtained on day 2 of the cycle for the measurement of FSH, LH and oestradiol (IMX, Abbott Diagnostics Division, Maidenhead, Berks, UK). Ovarian stimulation was commenced provided both FSH and LH concentrations were <10 IU/1 and oestradiol was <200 pmol/1. Clomiphene citrate (50 mg daily) was given from day 4 to day 8 of the cycle followed by i.m. FSH (Metrodin HP) at a dose of 75 IU daily (SIUI) or 150 IU daily (SIVF) from cycle day 7. On day 10, patients attended for monitoring, which consisted of plasma LH and oestradiol determination and an ultrasound scan to count the number of follicles. FSH administration and monitoring were continued daily until either the leading follicle reached 17.5 mm diameter or a spontaneous LH surge was detected. In addition, the SIVF group collected a second blood sample in the evening for more accurate detection of a possible LH surge and hence timing of egg recovery. The criteria for giving an ovulatory dose of HCG (5000 IU) were =s3 follicles of >16 mm diameter and an oestradiol concentration measurement of >1000 pmol/1 per follicle for SIUI or 2*3 similar follicles for SIVF. If the start of a spontaneous LH surge was detected (previously low LH concentrations rising to >10 IU/1, HCG was still given to 'top-up' the spontaneous surge. Intrauterine insemination with prepared semen was planned for 24^18 h following HCG administration or the spontaneous LH surge, whichever came first. Egg recovery and IVF were more carefully timed for h after HCG administration or the start of the LH surge in the group destined to have SIVF. Egg collection was achieved using transvaginal ultrasound (Combison 310; Kretztechnik, AG, Tiffenbach, Austria) as a simple outpatient procedure. No more than three embryos (as required by the Human Fertilisation and Embryology Authority) were replaced h after egg recovery. Patients in both groups were given luteal support with dydrogesterone (Duphaston, 10 mg tds; Solvay Healthcare, Southampton, UK) until a pregnancy test was performed 14 days postovulation. The data were analysed to include all pregnancies (live births, clinical abortions, ectopics and biochemical pregnancies). Implantation was considered to have occurred if the plasma HCG concentration rose above 10 IU/1 and a clinical pregnancy was described following identification of an intrauterine gestational sac with positive fetal heart. Results The mean age of women allocated to SIVF was 32.9 years compared with 33.2 years for the SIUI treatment group. Cycle day 2 basal plasma oestradiol, LH and FSH concentrations were also not different (geometric means: oestradiol 244 and 245 pmol/1, LH 4.9 and 5.3 IU/1, FSH 7.5 and 7.7 IU/1 in the SIVF and SIUI treatment groups respectively). Stimulated IVF treatment group Forty-one patients were initially allocated to SIVF treatment. Eight of these patients eventually had their treatment converted 2409

3 F.Zayed, E.A.Lenton and LD.Cooke Table II. Results obtained following initial randomly allocated treatment of unexplained (U) and mild male factor (M) patients by either stimulated invitro fertilization (SIVF) or stimulated intrauterine insemination (SIUI). The protocol used for the SIVF patients differed only from the SIUI group in that more Metrodin was administered (150 IU versus 75 IU daily) and egg collection was performed. Both groups received clomiphene citrate and neither was down-regulated with luteinizing hormone-releasing hormone (LHRH) agonist. Conversion between allocated treatments was implemented for over-response (SIUI to SIVF) or under-response (SIVF to SIUI). Cycles were cancelled if there was an absent or discordant response No. of cycles assigned to treatment (U+M) No. remaining after conversion (U+M) No. of cycles cancelled No. of cycles with failed fertilization No. of cycles with embryo transfer or IUI No. of cycles with positive HCG No. of live births Positive HCG/cycle started (%) Live births/cycle started (%) Positive HCG/cycle completed 2 (%) Live births/cycle completed 8 (%) SIVF SIUI 41(33+8) 39( ) 38(31+7) 42(29+13) NA a Completed = IUI or embryo transfer; NA = not applicable; HCG = human chorionic gonadotrophin. to SIUI (for poor response) but a further five patients from the SIUI group over-responded and so were converted to SIVF. Thus 38 couples actually ended up in the SIVF treatment group (Table II). Ultimately, four cycles were cancelled due to unclear LH profiles. A total of 34 cycles had between them 125 follicles >16 mm diameter (3.7 follicles per subject) and so proceeded to egg collection, yielding 189 oocytes (5.6 oocytes per subject), of which 120 eggs were considered mature (3.5 'good' oocytes per subject); 89 oocytes fertilized (47% of all eggs, 74% of 'good' eggs) and 55 embryos were replaced in 29 women (1.9 embryos per subject). Unexpected total fertilization failure occurred in five cases (four unexplained and one mild male factor). There were 11 cycles with a positive HCG which gave a positive HCG per cycle started rate of 28.9% and a positive HCG per cycle with embryo transfer of 37.9%. There were seven singleton live births, two clinical abortions and two biochemical pregnancies. The live birth rate per cycle started was 7/38 (18.4%) and per cycle completed 7/29 (24.1%). The mean dose of FSH required per patient was 900 IU, or 12 ampoules. Stimulated IUI treatment group Initially, 39 patients were allocated to SIUI treatment; then eight patients were added from the SIVF group and five patients were removed and converted to SIVF (due to overresponse). Thus, the SIUI group finally consisted of 42 couples (Table II). Only two cycles were cancelled, again because of anomalous LH profiles. On the day of HCG there were 92 follicles >16 mm diameter (2.3 follicles per subject). A positive HCG was detected in 12 cycles, giving a positive HCG per cycle started rate of 28.6% and per cycle with insemination of 30.0%. There were eight live births (including one twin and one triplet), one clinical abortion, three ectopics, but no biochemical pregnancies. The live birth rate per cycle started was 8/41 (19.0%) and per cycle completed 8/40 (20.0%). The mean dose of FSH required per patient was 500 IU, or 6.7 ampoules. Unexplained versus mild male factor Contrary to expectations, there was a trend towards better outcomes in the mild male factor compared with couples with unexplained infertility, although the numbers were too small for this to be significant. Of the seven mild male factor couples undergoing SIVF, six reached embryo transfer, resulting in three with positive HCG (50%). The equivalent figure for the unexplained couples undergoing SIVF was seven cycles with positive HCG from the 23 cycles reaching embryo transfer (30.4%). Similarly, for SIUI treatment, there were five cycles with positive HCG in the 12 inseminated mild male factor patients (41.0%) compared with 7/28 (25.0%) in the unexplained group. Oestradiol and LH concentrations There were no significant differences in the LH concentrations in the 4 days prior to administration of HCG in either treatment group (Table III). About half the patients in both groups had started their spontaneous LH surge by the day of HCG. Oestradiol concentrations were higher in the patients who received the larger dose of FSH (SIVF). This was not surprising, since there were more follicles >16 mm on the day of HCG in the SIVF group (3.3 per subject against 2.3 in the SIUI group). Cost-effectiveness A simple breakdown of the component costs of the two treatments, based on current costs in our clinic, is shown as Table IV. The cost per maternity, which is the amount that must be spent to achieve one live birth was more than double, at approximately 4600 for the SIVF treatment arm, compared with 1900 for SIUI treatment. Discussion Many treatments have been suggested for couples with unexplained infertility. Treatment options range from expectant management to a variety of assisted reproductive techniques such as SIUI, gamete intra-fallopian transfer (GIFT) and SIVF, as described in a multicentre trial (Crosignani et al, 1991). Undoubtedly, assisted reproduction has a place, but the mechanisms whereby infertility is overcome are not clear. Probably the main benefit arises directly from ovarian stimulation and the resultant increase in the numbers of follicles, oocytes and hence embryos available, thus increasing the probability of implantation. Possible secondary benefits may arise through improved folliculogenesis, which in turn could be correcting a subtle ovulatory problem, or by increased production of progesterone, which might improve endometrial receptivity (Pepperell and McBain, 1985; Welner et al., 1988). Potential advantages of IVF over IUI could be that oocyte retrieval overcomes defective ovulation (luteinized unruptured follicle syndrome), or that by observing fertilization in vitro, there is confidence that cleaving embryos are present in the uterus. SIVF, when there are surplus embryos, may also have some 2410

4 SIVF versus SIUI in unexplained infertility Table III. Median (range) concentrations of oestradiol and luteinizing hormone in the stimulated in-vitro fertilization (SIVF) and stimulated intrauterine insemination (SIUI) groups relative to the day human chorionic gonadotrophin was administered (day 0). The only difference between the two treatment groups was that SIVF patients were empirically given 150 IU Metrodin daily compared with 75 IU for SIUI patients. Both groups received clomiphene citrate, 50 mg daily, for 5 days Oestradiol (pmol/1) Luteinizing hormone (IU/1) SIVF SIUI SIVF SIUI Day-3 Day-2 Day-1 Day ( ) 4500 ( ) 6700 ( ) 9600 ( ) 2200 ( ) 3200 ( ) 4400 ( ) 5900 ( ) 6.9 ( ) 6.7 ( ) 7.2 ( ) 12.4 ( ) 8.1 ( ) 7.1 ( ) 8.5 ( ) 15.1 ( ) Table IV. Simple calculation of the cost per maternity of treatment using either stimulated in-vitro fertilization (SIVF) or stimulated intrauterine insemination (SIUI) in equivalent groups of patients with patent Fallopian tubes Treatment component Metrodin ( 16 for 75 IU) Monitoring and scans ( 100) Sperm preparation ( 50) Egg collection and culture ( 500) Embryo transfer or IUI ( 100) Total cost No. of live births Cost per maternity SIVF (n) 6880 (38) 3800 (38) 1700 (34) (34) 2900 (29) SIUI (n) (42) (42) (40) nil (40) advantage in that an element of embryo selection can be introduced. However, if as a result of applying more aggressive stimulation overall embryo potential is reduced, then that advantage is lost. In this study, the same ovarian stimulation regime (clomiphene citrate from day 4 followed by FSH from day 7) was used for both groups. This was partly to ensure that the two arms of the study were directly comparable but also because we wanted to re-assess IVF performed without prior downregulation. With the facility for egg collections from Monday through to Saturday and considerable experience with natural cycle IVF (Lenton et al, 1992), we experienced no problems in this small series in predicting and therefore avoiding egg collection on Sundays. Timing is far less critical for IUI than for IVF and is one of the major advantages of SIUI for the treatment of unexplained infertility. Even so, sperm preparation and insemination facilities are still required 6 days a week in order to provide a high quality service. The subjects in the SIVF treatment group received 150 IU of FSH per day unless they had been converted from SIUI, in which case they were routinely treated with 75 IU per day. Thus, patients in the SIVF group received on average 12 ampoules of FSH per cycle compared with 6.7 ampoules in the SIUI group (P < 0.05). Consequently, SIVF patients produced larger numbers of mature follicles (>16 mm on day of HCG) than SIUI patients (3.7 against 2.3 follicles per subject) and these were associated with high circulating oestradiol concentrations (Table III). In fact if only follicles >16 mm were considered, the amount of oestradiol produced per mature follicle 1 was identical (2600 and 2570 pmol/1) for both the SIVF and SIUI groups. The number of eggs collected in the SIVF group (189) was greater than the number of mature follicles (125), which presumably resulted from attempts to maximize egg yields by emptying all follicles, even the smaller ones. The number of 'good' eggs (120), however, correlated well with the number of mature follicles. Thus, in the SIVF group, although there were 3.7 follicles and 5.6 eggs/cycle, after fertilization only 1.9 embryos/cycle on average were replaced, which was probably equivalent to the 2.3 available mature follicles present in the SIUI patients. Certainly outcomes, however expressed (implantation and live birth rates per cycle started or cycle completed), were identical (Table II). The observation that SIVF is relatively less efficient (3.7 follicles required to produce the same outcome as 2.3 follicles with SIUI) is interesting and worthy of further comment. Firstly, while stimulation with clomiphene citrate and FSH undoubtedly develops several follicles, these follicles will be of varying sizes and maturity. Collecting eggs from all follicles, no matter what their size, gives only an illusion of success. Overall fertilization rates (47% of all oocytes) were poor and the proportion of couples with total failed fertilization was higher (13%) than expected from equivalent down-regulated IVF cycles (8%, unpublished observations for all cases of unexplained and mild male factor infertility). It seemed that many of the eggs were compromised and unable to fertilize normally. Fertilization rates per 'good' egg only, were normal for IVF (74%), and it is likely, although impossible to assess directly, that only the larger follicles (>16 mm) produced eggs with real potential. Alternative explanations of the unexpectedly high total fertilization failure rate (13%) in some subjects could be a consequence of inappropriate timing of egg recovery in a small number of cycles or merely a reflection of low numbers of subjects. It is possible that in a small number of is marked but generally unappreciated variation between oestradiol concentrations measured using different commercially available assay systems. These differences can be as much as 50% and will be amplified under stimulated conditions (IVF and IUI), so investigators are well advised to determine normal concentrations and ranges appropriate to their local system. 2411

5 F.Zayed, E.A.Lenton and LD.Cooke the cases of unexplained infertility there may be real defects in the spermatozoa or eggs (rather than defects acquired as a result of the IVF process itself) which result in consistent failure of fertilization (Molloy et al, 1991). To diagnose these cases accurately may, however, require several IVF attempts, as fertilization often occurs, even after total failed fertilization, in subsequent IVF cycles (Rowlands et ah, 1994). Physiologically, fertilized eggs/embryos reach the uterine cavity about 5 days after ovulation, and treatment with SIUI mimics this natural state, while SIVF interposes a number of artificial steps. Our data do not support the notion that these artificial steps are deleterious since, if allowance is made for eggs which may not have fertilized in vivo in the SIUI subjects, the result is an estimated 1.7 embryos per SIUI subject compared with 1.9 per subject in the SIVF group. It is therefore not surprising that the pregnancy outcomes were identical (Table II). Rather more disturbing was the high number (3/ 12) of ectopic pregnancies in the SIUI group. Larger numbers of pregnancies following SIUI need to be evaluated before the real relevance of this observation can be determined. From a practical viewpoint, it was a great advantage to be able to carry out these treatments in the context of an IVF clinic because of the simplicity of converting treatments between SIUI and SIVF. Without this facility, a further 13/80 (16%) of cycles would have required cancellation, bringing the total cycle cancellations across both treatments to 24%. The efficacy of this policy is demonstrated by the low incidence of complications, no cases of OHSS and only one triplet pregnancy. It seems unlikely that SIUI provided independently of SIVF will ever safely deliver results of the order reported here because of the necessity for a more conservative approach to cycle management. The decision to include mild male factor couples in this trial was justified by their performance, although the numbers of patients were too small for statistical consideration. There was no greater prevalence of fertilization failure, and pregnancy outcomes were, if anything, better than in the couples with genuine unexplained infertility. This suggests that if the 'mild' sperm defect is in any way responsible for the couple's infertility, it is through failure of sperm transport rather than failure of sperm function (i.e. ability to fertilize an egg). In this context, SIUI remains the most appropriate treatment for mild male factor. Considering the simplified estimates of cost per maternity (Table IV), it is clear that SIVF maternities, with the greater number of skilled procedures involved, will inevitably cost more than SIUI maternities ( 4600 versus 1900). Although drug costs per patient were higher with SIVF, these did not contribute significantly to the overall costs. If, however, conventional long-protocol SIVF had been performed (using initial down-regulation with LHRH agonists), experience in our clinic indicates that ~30 ampoules of FSH would have been required per couple. Assuming that the live birth rates remained unchanged (24%), the impact of the extra drugs alone would increase the cost per maternity to These simple calculations do not take into account the increased length of a down-regulated treatment cycle and the acceptability of this to patients, nor do they consider the cost impact of 2412 complications such as OHSS, which is more prevalent following conventional down-regulated IVF (Neumann et al, 1994) than following simple SIUI. Of course, conventional downregulated IVF with its higher egg yields will remain the treatment of choice if additional procedures such as intracytoplasmic sperm injection (ICSI) are required but, conversely, IUI using down-regulation prior to stimulation should be avoided because drug usage will remain high and many patients will inevitably develop too many follicles for IUI to be performed with safety (Sengoku et al, 1994). We conclude that SIUI is the first choice treatment for unexplained and mild male factor infertility (Chung et al, 1995) and that it will show the same efficacy as SIVF (using clomiphene and FSH) provided the treatment is delivered to the same standard (careful monitoring, detection and management of spontaneous LH surges). Treatment delivered within the context of a specialized IVF centre will be more efficient and safer because of the potential to salvage over-responding cycles by conversion to IVF. From the cost-benefit viewpoint, SIUI, with its lower cost per maternity and therefore greater opportunity for repeat treatments, should be available routinely in all assisted conception units. Indeed, the data presented here make it difficult to support the current and widespread practice of using SIVF, particularly involving down-regulation, as a first line of treatment for any patient with unexplained and mild male factor infertility (Ranieri et al., 1995). Acknowledgements We would like to express our gratitude to all the staff employed at the Sheffield Fertility Centre during the time period when these patients were treated, in particular Drs Anju Kumar, Imad Al- Ghazzawi, Amina Shafik, Abdalla Salih, Padmasari Poonati and Karen Turner (senior biologist). References Chung, C.C., Fleming, R., Jamieson, M.E. et al. (1995) Randomized comparison of ovulation induction with and without intrauterine insemination in the treatment of unexplained infertility. Hum. Reprod., 10, Crosignani, P.G., Walters, D.E. and Soliani, A. (1991) The ESHRE multicentre trial on the treatment of unexplained infertility: a preliminary report. Hum. Reprod., 6, Dodson, W.C., Walmer, D.K., Hughes, C.L. et al. (1991) Adjunctive leuprolide therapy does not improve cycle fecundity in controlled ovarian hyperstimulation and intrauterine insemination of sub-fertile women. Obstet. Gynecol, 78, Homburg, R., Armar, N.A., Eshel, A. et al. (1988) Influence of serum luteinising hormone concentrations on ovulation, conception and early pregnancy loss in polycystic ovary syndrome. Br. Med. J., 297, Lenton, E.A., Cooke, I.D., Hooper, M. et al. (1992) In-vitro fertilization in the natural cycle. Balliere's Clin. Obstet. Gynaecol, 6, Macnamee, M.C., Edwards, R.G. and Howies, CM. (1988) The influence of stimulation regimes and luteal phase support on the outcome of IVF. Hum. Reprod., 3 (Suppl. 2), Molloy, D., Harrison, K., Breen, T. and Hennessey, J. (1991) The predictive value of idiopathic failure to fertilize on the first in vitro fertilization attempt. Fertil. Steril, 56, Neumann, P.J., Gharib, S.D. and Weinstein, M.C. (1994) The cost of a successful delivery with in vitro fertilisation. N. Engl. J. Med., 331, Pepperell, RJ. and McBain, J.C. (1985) Unexplained fertility: a review. Br. J. Obstet. Gynaecol, 92,

6 SIVF versus SIUI in unexplained infertility Ranieri, M., Beckett, V.A., Marchant, S. et al. (1995) Gamete intra-fallopian transfer or in-vitro fertilization after failed ovarian stimulation and intrauterine insemination in unexplained infertility? Hum. Reprod., 10, Rowlands, D.J., McDermott, A. and Hull, M.G.R. (1994) The prognosis for assisted conception treatment after unexpected failure of fertilization in vitro: a comparative study. Hum. Reprod., 9, Sengoku, K., Tamate, K., Takaoka, Y. et al. (1994) A randomized prospective study of gonadotrophin with or without gonadotrophin-releasing hormone agonist for treatment of unexplained infertility. Hum. Reprod., 9, Stanger, J.D. and Yovich, J.L. (1985) Reduced in vitro fertilization of human oocytes from patients with raised basal luteinizing hormone levels during the follicular phase. Br. J. Obstet. Gynaecoi, 92, Welner, S., DeCherney, A.H. and Polan, M.L. (1988) Human menopausal gonadotrophins: a justifiable therapy in ovulatory women with long-standing idiopathic infertility. Am. J. Obstet. Gynecol., 158, Received on September 17, 1996; accepted on August 5,