Poor responders to ovulation induction: is proceeding to in-vitro fertilization worthwhile?

H. Lashen1, W. Ledger, A. Lopez-Bernal and D. Barlow

Nuffield Department of Obstetrics and Gynaecology, University of Oxford, John Radcliffe Hospital, Oxford OX3 9DU, UK


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
A `poor response' in the context of in-vitro fertilization (IVF) can be defined as failure to produce an adequate number of mature follicles, and/or a peak oestradiol concentration less than a defined minimum. The cut-off points implied in this definition vary between different centres. Many opt to cancel the IVF cycle when their defined minimum concentrations are not reached despite the lack of evidence of improved outcome in subsequent cycles. Patients attending the Oxford Fertility Unit who are `poor responders' have always been given the option of continuing with treatment. The first cycles of IVF in 124 patients, with normal day 3 follicle stimulating hormone (FSH), who produced less than five follicles within a 2 year period were studied. The patients were divided into three groups according to the number of follicles produced: A (one or two follicles; n = 33), B (three follicles; n = 33) and C (four follicles; n = 58). The three groups were similar in age, day 3 FSH, total gonadotrophin dose, duration of stimulation, peak oestradiol concentration, oocyte yield, fertilization rate and the clinical pregnancy rate. However, group A had a significantly higher oestradiol concentration per follicle (P < 0.001). The clinical pregnancy rate/cycle in the three groups was comparable to our overall rate in the study period (25.5%). This paper suggests that poor responders with a normal day 3 FSH may still achieve a pregnancy rate similar to that of normal responders.

Key words: IVF/ovulation induction/poor responders


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Poor responders to ovulation induction represent a challenge to those carrying out assisted reproduction techniques, especially in-vitro fertilization (IVF) where a large number of oocytes is desirable. The definition of `poor response' varies from one group to another. Serafini et al. (1988) defined poor responders as those who produced less than three follicles despite adequate ovarian stimulation. Others have considered four follicles as their cut-off point (Jenkins et al., 1991Go), while Land et al. (1996), abandoned those IVF cycles where four follicles or less were produced. Fenichel et al. (1989) used both the number of follicles and a peak oestradiol value of <300 pg/ml during ovarian stimulation in their definition of poor response.

Despite these differences in definition, `poor response' often leads to cycle cancellation, in the hope that a better response might be obtained in a subsequent cycle. This policy is based on the assumption that poor responders have a poor outcome in terms of clinical pregnancy rate. However, we believe that `poor responders' actually represent a heterogeneous group of patients who can be divided clinically into two main groups. The first group includes those with low ovarian reserve as reflected by their high basal follicle stimulating hormone (FSH), and the second group includes women with normal ovarian reserve who are inherently low responders to gonadotrophin stimulation. However, the latter group might be assumed to have the same chance of achieving a pregnancy as the normal responders, since the oocyte quality should be satisfactory. Therefore, we have allowed the patients who fall into this category (poor responders with normal day 3 FSH) the option to proceed to IVF regardless of the number of follicles they produce provided that their oestradiol concentration exceeds 300 pg/ml (1000 pmol/l). In this study we investigated the IVF outcome of the patients who produced four follicles or less after ovulation induction, allowing us to compare our results with the various definitions of `poor response' detailed in the literature.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
This was a retrospective comparative study of the prospectively collected database of the Oxford Fertility Unit. We reviewed data on all IVF cycles started from January 1, 1995 to July 31, 1997 inclusive. The protocol for pituitary down-regulation, ovulation induction and human chorionic gonadotrophin (HCG) administration remained consistent throughout the study period.

The IVF–embryo transfer protocol has been described in detail elsewhere (Lockwood et al., 1995Go). We employed intranasal GnRH agonists given daily from day 21 of the cycle (or 21 days after the start of a gestagen-induced bleed if the patient was severely oligo- or amenorrhoeic). Confirmation of pituitary down-regulation after the agonist induced withdrawal bleed required a serum oestradiol concentration of <20 pg/ml (74 pmol/l).

All the patients with normal day 3 FSH (3–12 IU/l), who produced less than five follicles were included in the study. They were divided into three groups according to the three following cut-off points cited in the literature. Group A: less than three follicles (Serafini et al., 1988Go); group B: less than four follicles (Jenkins et al., 1991Go); group C: less than five follicles (Land et al., 1996Go). We compared the three groups with regard to age, basal FSH (cycle day 3), body mass index (BMI), total gonadotrophin dose (IU), duration of gonadotrophin stimulation (days), peak oestradiol concentration (pmol/l), oocyte yield (no. oocytes/no. follicles), oestradiol per follicle (peak oestradiol/no. follicles), and clinical pregnancy rate per cycle and per embryo transfer. `Clinical pregnancy' was defined as the presence of fetal heart pulsation on ultrasound scan at 8 weeks gestation. We also compared the above-mentioned parameters between the study group (who produced less than five follicles) and patients who produced five or six follicles during the study period. The above-mentioned parameters as well as the number of follicles, oocytes and embryos were compared between the women aged >35 and those <=35 years in the study group.

Statistical analysis
This was carried out using Minitab for Windows statistical package (Minitab Inc., PA, USA). The non-parametric tests; Mood's median, and Mann–Whitney were used as appropriate to compare the different parameters in the different groups. The data are presented as mean ± SEM and 95% confidence interval (95% CI), and the median. P < 0.05 was considered significant.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
A total of 124 patients was included in the study. None of these patients had one ovary only. The mean age and basal FSH was 35 ± 0.3 (34.4–35.7), and 6.8 ± 0.3 (6.1–7.4) respectively. The range of the age was 24–45 years. Fifty-eight patients were >35 years of age and 66 patients were <=35 years of age. Groups A (one or two follicles, n = 33), B (three follicles, n = 33) and C (four follicles, n = 58) were similar with regard to all the compared parameters except for the oestradiol/follicle which was significantly higher in group A, and the number of embryos transferred which was lower in groups A and B compared to C (Table IGo). The number of patients proceeding to embryo transfer in the three groups was 16/33 (49%), 28/33 (84%) and 45/58 (78%) in groups A, B and C respectively. In group A significantly fewer patients had embryo transfer (P = 0.002); however, the clinical pregnancy rate per embryo transfer was similar in the three groups (Table IGo). The reasons for not proceeding to embryo transfer in the three groups were mainly failed oocyte retrieval and failure of fertilization (Table IGo). The outcome of those who produced one follicle compared with those who produced two follicles is summarized in Table IIGo.


View this table:
[in this window]
[in a new window]
 
Table I. The response and outcome at different cut-off points
 

View this table:
[in this window]
[in a new window]
 
Table II. One versus two follicles in group A
 
The patients in the study group (who produced less than five follicles) and those who produced five or six follicles (n = 142) were comparable with regard to age, basal FSH, BMI, duration of gonadotrophin stimulation, oocyte yield, and normal fertilisation rate (Table IIIGo). However, those who produced five or six follicles received a significantly lower total gonadotrophin dose (P = 0.0001), higher peak oestradiol (P = 0.03), lower oestradiol/follicle (P = 0.0008), lower incidence of failed oocyte retrieval (P < 0.001) and a larger number of embryos transferred (P < 0.0005). Despite the significantly higher number of embryos transferred in those who produced five or six follicles, the clinical pregnancy rate was similar to those who produced less than five follicles. The clinical pregnancy rate per embryo transfer was similar in the two groups (Table IIGo), but the number of patients proceeding to embryo transfer was significantly higher in those who produced five or six follicles [123/142 (87%)] compared to the study group [89/124 (72%)] (P = 0.003).


View this table:
[in this window]
[in a new window]
 
Table III. Patients who produced less than five follicles versus those who produced five or six follicles
 
Patients >35 years of age compared with those <35 years in the study group
The patients in the two age groups were similar with regard to all the compared parameters except for the total gonadotrophin dose, which was significantly higher (P < 0.0001) in the older women (Table IVGo).


View this table:
[in this window]
[in a new window]
 
Table IV. Older versus younger patients in the poor responder group
 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Patients undergoing ovulation induction can show both intra- and inter-individual variability in their ovarian response to the same stimulation protocol. Explaining such variability and the inherently low response in some patients has been hampered by the lack of understanding of the mechanism by which the follicles are mobilized from the `primordial' stage to the `growing cohort' stage. The trend has been to abandon those cycles in which a small number of follicles are recruited, assuming a poor outcome. In the subsequent cycles several strategies for improving ovarian response have been tried.

Before the introduction of gonadotrophin-releasing hormone agonist (GnRHa) into IVF treatment, the combination of clomiphene citrate and human menopausal gonadotrophins (HMG) in poor responders was suggested by Trounson and Leeton (1982). However, Pellicer et al. (1987), using this combination, reported partial improvement in ovarian response in subsequent cycles which depended on the oestradiol trend in the previous cycle where HMG alone was used. Serafini et al. (1988), using GnRHa in a short flare-up protocol in 27 patients previously cancelled for poor response, reported a pregnancy rate of 33.3% per cycle. Similar results were reported by Droesch et al. (1989) and Muasher et al. (1992) using the short flare-up protocol. However, Sathanandan et al. (1989) reported a 10% pregnancy rate and no live births using the same protocol.

Increasing the HMG dose during the course of treatment was advocated by Laufer et al. (1986). However, van-Hooff et al. (1993) in a prospective randomized study found that increasing the daily gonadotrophin dose above 225 IU failed to improve the ovarian response in low responders.

Ben Rafael et al. (1986) suggested that differences in the metabolism of the exogenously administered FSH were responsible for different gonadal reactions. They concluded that the FSH metabolic lag could be overcome by increasing the gonadotrophin dose. However, Benadiva et al. (1988) reported that the magnitude of ovarian response was altered in the high responders but not the low responders by changing the gonadotrophin dose, which suggests that temporal individuality of endocrine profile cannot be altered by varying the dose of gonadotrophin.

More recently a combination of GnRHa and a high gonadotrophin dose was initiated in a group of patients whose first cycle was abandoned for poor response. In this non-randomized study, cancellation followed by a second cycle with increased gonadotrophin did not appear to improve outcome (Land et al., 1996Go). We have found that exceeding a daily gonadotrophin dose of 300 IU is unrewarding in terms of number of oocytes and embryos, and have argued that a starting dose of 150 IU per day in the first IVF cycle represented the best compromise between safety and efficacy for the younger IVF patient (Lashen et al., 1998Go).

The assumption of low or even absent probability of pregnancy in the poor responder group stems from the early days of IVF before the introduction of GnRHa and ultrasound guided oocyte retrieval. Pituitary desensitization with GnRHa prior to gonadotrophin ovarian stimulation has enabled better synchronization of follicular development, has alleviated the risk of premature luteinization, and has resulted in better oocyte quality. Moreover, ultrasound-guided oocyte retrieval has improved the oocyte yield per oocyte retrieval. These technical improvements have led us to re-assess the relationship between `poor response' and IVF outcome. In this study the similarity in the cumulative pregnancy rate per cycle and per embryo transfer using different definitions of poor response, and between the poor responders and those who produced five or six follicles, indicates that the poor responders' potential of achieving a pregnancy is partially independent of the number of follicles produced, provided that basal FSH concentration does not reflect diminished ovarian reserve. However, normal day 3 FSH values may not always reflect a normal ovarian reserve. Farhi et al. (1998) reported on 12 patients with normal basal FSH who totally failed to respond to ovarian stimulation and developed ovarian failure within 3–19 months. Furthermore, Licciardi et al. (1995) and Smotrich et al. (1995) argued that a high oestradiol concentration in early follicular phase may suppress the FSH value to a false normal value. In both studies the authors reported a poor outcome when oestradiol concentration exceeded a certain cut-off point, the choice of which was not clarified by Smotrich et al. Licciardi et al. (1995) have also failed to show that, for all cycles, elevated oestradiol was associated with suppressed FSH concentrations on day 3. Moreover, considering that the measurement of oestradiol concentration is assay dependent, selecting a reproducible cut-off point may not be possible.

In view of the limitations of normal basal FSH values in predicting normal ovarian reserve and the findings by Farhi et al. (1998), it seems appropriate in some cases to use the outcome of ovulation induction as an indicator for success. The comparison between the older (>35 years) and the younger (<=35 years) patients in the poor responder group shows similarity in the cumulative pregnancy rate and ovarian response in terms of the number of follicles, oocytes and embryos, and peak oestradiol concentration. The total gonadotrophin dose administered, however, in the older women was significantly higher despite the similarity in the duration of stimulation, reflecting the clinicians impression that older IVF patients should receive a higher gonadotrophin dose.

The proportion of patients who proceeded to embryo transfer was significantly lower in those who produced one or two follicles than those who produced three or four follicles, and in the poor responders generally compared with those who produced five or six follicles. This occurred despite the similarity in oocyte yield and fertilisation rates between the compared groups, identifying an obvious disadvantage in the poor response group. Failure of oocyte retrieval was evidently the rate-limiting factor in reaching the stage of embryo transfer particularly when only one or two follicles were available. However, in our view, the similarity in the cumulative pregnancy rate per cycle and per embryo transfer in the compared groups, despite the significant difference in the number of embryos transferred, strengthens the argument against abandoning the cycles of poor responders provided that the couples are properly counselled about the increased chance of the cycle ending without embryo transfer. The situation for those who produced one follicle only is not entirely clear as the very small number in the study precludes meaningful statistical analysis. However, they did not appear to differ from those who produced two follicles. Therefore, we recommend counselling these patients individually, taking into consideration any previous failed cycles, as well as their financial circumstances, before reaching a decision.

The use of ICSI to improve the fertilization rate in the poor responders, enabling more embryos to be available for transfer, has been studied by Moreno et al. (1998). In a prospective randomized study the authors found that ICSI had no advantage over IVF with regard to the cumulative pregnancy rate or the number of embryos transferred. However, as they failed to discuss the method and timing of randomization in relation to ovarian stimulation and oocyte retrieval, inadvertent bias could not be excluded. Furthermore, although the 4% advantage in the pregnancy rate in the ICSI group was not significant in the small sample size in their study, a larger study with sufficient power may disagree with their findings.

The comparable fertilization rate and cumulative pregnancy rate per embryo transfer in the different groups shows the good quality of the produced oocytes and embryos and the potential for pregnancy in poor responders. This information is clearly important when counselling patients who produce only one to four follicles prior to proceeding to oocyte retrieval.

We conclude that the diverse causes of poor ovarian response to ovulation induction should be taken into consideration when deciding whether to cancel a cycle. Poor responders with normal ovarian reserve represent a group with reasonable prospect of pregnancy despite lower than average oocyte number. Therefore, proceeding to oocyte retrieval in a selected group of poor responders is justifiable, particularly in the absence of evidence that a better response can be achieved in subsequent cycles. The small numbers in the study prevent a firm conclusion and a prospective randomized study is needed to determine the best management of poor responders in their first and subsequent cycles.


    Notes
 
1 To whom correspondence should be addressed at: Department of Obstetrics and Gynaecology, Solihull Hospital, Lode Lane, Solihull B91 2JL, UK Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Ben Rafael, Z., Strauss, J.F., Mastroianni, L. and Flickinger, G.L. (1986) Differences in ovarian stimulation in human menopausal gonadotrophin treated woman may be related to follicle-stimulating hormone accumulation. Fertil. Steril., 46, 586–592.[ISI][Medline]

Benadiva, C.A., Ben-Rafael, Z., Strauss, J.F. et al. (1988) Ovarian response of individuals to different doses of human menopausal gonadotropin. Fertil. Steril., 49, 997–1001.[ISI][Medline]

Droesch, K., Muasher, S.J., Brzyski, R.G. et al. (1989) Value of suppression with a GnRH agonist prior to gonadotropin stimulation for in vitro fertilization. Fertil. Steril., 51, 292–297.[ISI][Medline]

Farhi, J., Homburg, R., Orvieto, R. and Ben Rafael, Z. (1998) Non-response to ovarian stimulation in normogonadotrophic, normogonadal women: a clinical sign of impending onset of ovarian failure pre-empting the rise in basal follicle stimulating hormone levels. Hum. Reprod., 13, 241–243.

Fenichel, P., Grimaldi, M., Olivero, J.F. et al. (1989) Predictive value of hormone profiles before stimulation for in vitro fertilization. Fertil. Steril., 51, 845–849.[ISI][Medline]

Jenkins, J.M., Davies, D.W., Devonport, H. et al. (1991) Comparison of `poor responders' with `good responders' using a standard buserelin/human menopausal gonadotrophin regime for in-vitro fertilisation. Hum. Reprod., 6, 918–921.[Abstract]

Land, J.A., Yarmolinskaya, M.I., Dumoulin, J.C.M. and Evans, J.H. (1996) High-dose human menopausal gonadotrophin stimulation in poor responders does not improve the in vitro fertilization outcome. Fertil. Steril., 65, 961–965.[ISI][Medline]

Lashen, H., Ledger, W., Lopez-Bernal, A. and Barlow, D. (1998) Superovulation with high gonadotrophin dose for in vitro fertilisation: is it effective? J. Assist. Reprod. Genet., 15, 438–443.[ISI][Medline]

Laufer, N., De Cherney, A.H., Tarlatzis, B.C. and Naftolin, F. (1986) The association between pre-ovulatory serum 18-ß estradiol pattern and conception in human menopausal gonadotropin-human chorionic gonadotropin stimulation. Fertil. Steril., 46, 73–76.[ISI][Medline]

Licciardi, F., Liu, H.-C. and Rosenwaks, Z. (1995) Day 3 estradiol serum concentrations as prognosticators of ovarian stimulation response and pregnancy outcome in patients undergoing in vitro fertilisation. Fertil. Steril., 64, 991–994.[ISI][Medline]

Lockwood, G.M., Pinkerton, S.M. and Barlow, D.H. (1995) A prospective randomized single-blind comparative trial of nafarelin acetate with buserelin in long-protocol gonadotrophin-releasing hormone analogue controlled in-vitro fertilization cycles. Hum. Reprod., 10, 293–298.[Abstract]

Moreno, C., Ruiz, A., Simón, C. et al. (1998) Intracytoplasmic sperm injection as a routine indication in low responder patients. Hum. Reprod., 13, 2126–2129.[Abstract]

Muasher, S.J. (1992) Use of gonadotropin-releasing hormone agonists in controlled ovarian hyperstimulation for in vitro fertilization. Clin. Ther., 14, 74–86.[ISI][Medline]

Pellicer, A., Lightman, A., Diamond, M.P. et al. (1987) Outcome of in vitro fertilization in women with low response to ovarian stimulation. Fertil. Steril., 47, 812–815.[ISI][Medline]

Sathanandan, M., Warnes, G.M., Kirby, C.D. et al. (1989) Adjuvant leuprolide in normal, abnormal and poor responders to controlled ovarian stimulation for IVF/GIFT. Fertil. Steril., 51, 998–1006.[ISI][Medline]

Serafini, P., Stone, B., Kerin, J. et al. (1988) An alternative approach to controlled ovarian hyperstimulation in `poor responders' with a gonadotrophin releasing hormone analog. Fertil. Steril., 49, 90–95.[ISI][Medline]

Smotrich, D.B., Widra, E.A., Gindoff, P.R. et al. (1995) Prognostic value of day 3 estradiol on in vitro fertilisation outcome. Fertil. Steril., 64, 1136–1140.[ISI][Medline]

Trounson, A.O. and Leeton, J.F. (1982) The endocrinology of clomiphene stimulation. In Edwards, R.G. and Purdy, J.M. (eds), Human Conception In Vitro. Academic Press, New York, pp. 51–58.

van Hooff, M.H., Alberda, A.T., Huisman, G.J. et al. (1993) Doubling the human menopausal gonadotrophin dose in the course of an in-vitro fertilisation treatment cycle in low responders: a randomized study. Hum. Reprod., 8, 369–373.[Abstract]

Submitted on September 15, 1998; accepted on December 17, 1998.