- Email: [email protected]
Effects of transdermal testosterone in poor responders undergoing IVF: systematic review and meta-analysis
Reproductive BioMedicine Online (2012) 25, 450– 459
www.sciencedirect.com www.rbmonline.com
REVIEW
Effects of transdermal testosterone in poor responders undergoing IVF: systematic review and meta-analysis ´lez-Comadran a,b, Montserrat Dura ´n b, Ivan Sola ` c,d, Mireia Gonza ´n Carreras a,f, Miguel A Checa a,f,g,* ´bregues e, Ramo Francisco Fa a Department of Obstetrics and Gynecology, Hospital del Mar, Barcelona, Spain; b International Master in `noma de Barcelona, Barcelona, Spain; c Iberoamerican Cochrane Centre, Reproductive Medicine, Universitat Auto ´blica Institute of Biomedical Research, IIB Sant Pau, Barcelona, Spain; d CIBER de Epidemiologı´a y Salud Pu (CIBERESP), Spain; e Institut Clı´nic of Gynecology, Obstetrics and Neonatology, Hospital Clı´nic-Institut d’Investigacions `diques August Pi i Sunyer (IDIBAPS), Faculty of Medicine, University of Barcelona, Barcelona, Spain; f Universitat Biome `noma de Barcelona, Barcelona, Spain; g Centro de Infertilidad y Reproduccio ´n Humana, Barcelona, Spain Auto
* Corresponding author. E-mail address: [email protected] (M.A. Checa). Miguel A Checa was awarded a Bachelor of medicine and surgery at the University of Barcelona in 2008. She is a specialist in Obstetrics and Gynecology at Hospital del Mar de Barcelona. She was studied for an International Master in reproductive medicine at Universitat Auto `noma de Barcelona in 2011–2012.
Abstract A systematic review and meta-analysis was performed to evaluate the effect of transdermal testosterone preceding ovar-
ian stimulation in women with poor ovarian response undergoing IVF. Studies comparing pretreatment with transdermal testosterone versus standard ovarian stimulation among poor responders were included. The main outcome assessed was live birth. Three trials were included (113 women in the testosterone group, 112 in the control group). Testosterone-treated women achieved significantly higher live birth rate (risk ratio, RR, 1.91, 95% CI 1.01 to 3.63), clinical pregnancy rate (RR 2.07, 95% CI 1.13 to 3.78) and required significantly lower doses of FSH (RR 461.96, 95% CI 611.82 to 312.09). However, differences observed in clinical pregnancy per embryo transferred were not statistically significant (RR 1.72, 95% CI 0.91 to 3.26). No differences were observed regarding number and quality of the oocytes retrieved. In conclusion, transdermal testosterone significantly increases live birth and reduces the doses of FSH required. These findings support the theoretical synergistic role of androgens and FSH on folliculogenesis. The present data should be interpreted with caution because of the small number of trials and clinical heterogeneity. The identification of poor responders that could especially benefit from testosterone treatment should be addressed in further studies. RBMOnline ª 2012, Reproductive Healthcare Ltd. Published by Elsevier Ltd. All rights reserved. KEYWORDS: IVF, live birth, poor responder, randomized controlled trial, transdermal testosterone
1472-6483/$ - see front matter ª 2012, Reproductive Healthcare Ltd. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.rbmo.2012.07.011
Transdermal testosterone in poor responders undergoing IVF
Introduction The poor response to ovarian stimulation among women undergoing IVF is of great concern in reproductive medicine. The reported incidence of poor response varies from 9% to 26% (Vollenhoven et al., 2008). For years, several authors have attempted to define ‘poor response’ in assisted reproduction (Kailasan et al., 2004; Arslan et al., 2005; Sallam et al., 2005; Turhan, 2006). In July 2011, The European Society for Human Reproduction and Embryology, developed a consensus on the definition of ‘poor response’ to ovarian stimulation for IVF, known as Bologna criteria to standardize the definition in a simple and reproducible manner (Ferraretti et al., 2011). In this context, Polyzos and Devroey (2011) among others concluded that the applicability of this definition needs to be tested through clinical trials and discouraged the performance of a meta-analysis with the currently available trials since their results may lead to recommendations of ambiguous value. Certain modalities have been tested through randomized trials to improve the response to gonadotrophin stimulation and thus reproductive outcomes. In daily practice, the most commonly used agents before or during ovarian stimulation are transdermal testosterone, dehydroepiandrosterone (DHEA), aromatase inhibitors, recombinant LH and recombinant human chorionic gonadotrophin (HCG). The results from these trials that aim to increase intraovarian androgen concentrations in poor responders have shown conflicting results (Fanchin et al., 2011). Moreover, some meta-analyses have evaluated the effect of different types of androgen supplementation, such as transdermal testosterone and DHEA, and the duration of the treatment; however, these studies are markedly heterogeneous in their inclusion criteria (Sunkara et al., 2011; Bosdou et al., 2012). The present meta-analysis focused on the intervention of transdermal testosterone administration, which achieves powerful systemic androgenization and subsequently, a greater action of FSH compared with other androgen-modulating agents, such as aromatase inhibitors (Mitwally and Casper, 2002). Live birth should be the most relevant outcome of interest in the treatment of infertility and human reproductive disorders. Therefore, the aim of this meta-analysis is to assess the effect of transdermal testosterone pretreatment on live birth rate among poor responders undergoing ovarian stimulation for IVF.
Materials and methods The study was exempt from Institutional Review Board approval because this was a systematic review and meta-analysis. The preferred reporting items for systematic reviews and meta-analysis (PRISMA statement) were endorsed to report the results of this systematic review (Moher et al., 2009).
Search strategy
451 Cochrane Central Register of Controlled Trials (CENTRAL). The search combined terms and descriptors related to testosterone, poor ovarian response and IVF outcomes. The search strategy was modified to comply with the requirements of each database consulted. Validated filters were added to that strategy to retrieve clinical trials (Higgins and Green, 2008). Moreover, this study searched for ongoing trials in the main clinical trials registers, including www.controlled-trials.com, www.clinicaltrials.gov and the WHO International Clinical Trials Registry Platform (www.who.int/trialsearch). No language limits were used. The reference lists were screened for all relevant articles and overviews. The complete search strategy is available upon request from the authors.
Eligibility criteria The review included randomized controlled clinical trials of women with poor ovarian response undergoing IVF and receiving embryo transfers independently of whether the cause of infertility or subfertility was due to male or female factors. The type of intervention evaluated was the administration of transdermal testosterone preceding gonadotrophin treatment compared with standard gonadotrophin ovarian stimulation protocols without the administration of transdermal testosterone during the period of follicular stimulation.
Outcome measures The main outcome of interest for the review was the live birth rate. Secondary outcomes were clinical pregnancy rate and clinical pregnancy adjusted per embryo transferred, miscarriage, the number of fertilized oocytes, the number of oocytes retrieved, the number of metaphase-II oocytes retrieved, the total dose of FSH administered and oestradiol concentrations on day of HCG injection. All of the outcomes of interest were considered per woman randomized (Supplementary Table 1, available online only). The outcomes were defined according to the terminology recommended in the International Committee Monitoring Assisted Reproductive Technologies glossary (Zegers-Hochschild et al., 2009) and the updated and revised nomenclature for the description of early pregnancy events (Farquharson et al., 2005).
Data extraction The data were collected using standard forms, in which the characteristics of the study design, participants, interventions, comparisons and main results were recorded. Two independent authors (MGC and MDR) judged study eligibility, assessed the risk of bias and extracted data solving discrepancies by agreement, and if needed, reaching consensus with a third author (MAC). The agreement between reviewers was analysed using the weighted kappa for each inclusion criterion (Fleiss, 1993).
Assessment of risk of bias An exhaustive electronic search was performed in the following databases (from their inception until December 2011): MEDLINE, EMBASE, Science Citation Index and the
The risk of bias in the included studies was assessed following the guidance suggested by the Cochrane Collaboration
452
M Gonza ´lez-Comadran et al.
(Higgins and Green, 2008). This study addressed six specific domains (explicit eligibility criteria, sequence generation, allocation concealment, patient blinding, outcome assessor blinding and patients lost to follow up) and assessed the adequacy of the study report in relation to each domain. A judgment of ‘yes’ for all domains indicates a low risk of bias and a judgement of ‘no’ for one or more domains indicates a high risk of bias. The risk of bias in the specific domains was interpreted as unclear when the information was not available. Patients lost to follow up were described in percentages. The risk of bias for the included trials is detailed in Table 1.
Statistical analysis
eligible to be included in the review by one or both reviewers and 16 were excluded because of non-randomized comparisons (n = 3) or because the results were irrelevant for the purpose of the present meta-analysis (n = 13). During the second phase of the inclusion process, out of the four studies, one was excluded because of non-randomized comparisons (Balasch et al., 2006). Finally, three randomized controlled trials met the inclusion criteria and were included (Massin et al., 2006; Fa ´bregues et al., 2009; Kim et al., 2011). A flow chart of the included trials is shown in Figure 1. The two reviewers achieved good agreement on the selection of the trials (weighted kappa 0.694, P < 0.001, standard error 0.195).
Description of the studies For each study, the treatment effect was measured with risk ratios (RR) for dichotomous outcomes and mean differences for continuous outcomes, which were presented with their corresponding 95% confidence intervals (CI). Event data were extracted following the intention-to-treat principle. When possible, outcome data from each study were pooled using a Mantel–Haenszel model, applying the fixed-effects model. A post-hoc analysis calculated the risk for clinical pregnancy per embryo transferred using the estimates of the intervention effects from original trials and their standard errors using the inverse variance approach (Deeks et al., 2011). Statistical heterogeneity was quantified using the I2 statistic, which describes the percentage of total variation across studies that is due to heterogeneity rather than sampling error (Higgins et al., 2003). All analyses were conducted using Review Manager version 5.1 (2011, The Nordic Cochrane Centre, Copenhagen, The Cochrane Collaboration).
Results Out of a total of 20 documents identified in the initial electronic search, four studies were considered potentially Table 1
A total of 225 women were randomized to an intervention group (n = 113) and a control group (n = 112). The mean age reported across the studies was comparable (Table 2) and each study reported no statistical difference regarding this parameter among the intervention and the control group. The inclusion criteria were relatively heterogeneous among the studies, especially regarding the main inclusion criteria, which was ‘poor response to ovarian stimulation during an IVF/intracytoplasmic sperm injection cycle’. This item was defined by Fa ´bregues et al. (2009) as the failure to produce 3 follicles with a mean diameter 14 mm or the collection of 3 follicles at retrieval, whereas Kim et al. (2011) set the limit of poor response the production of 3 follicles with a mean diameter 16 mm or the collection of 3 follicles at retrieval. In contrast, Massin et al. (2006) defined poor response as a plasma oestradiol value below 1.200 pg/ml on HCG day and the collection of 5 follicles at retrieval, adding as a necessary criteria for enrolment the evidence of a decreased ovarian reserve at day 3 of a spontaneous cycle, as determined with plasma hormonal values outside the normal range (FSH >12 IU/l, oestradiol >70 pg/ml and inhibin B <45 pg/ml).
Methodological data of the clinical trials included in the review.
Criterion
Massin et al. (2006)
´bregues et al. Fa (2009)
Kim et al. (2011)
Explicit eligibility criteria Sequence generation Allocation concealed Patient blinding Outcome assessor blinding Patients lost to follow up (%)
Yes
Yes
Yes
Yes (random permutation table)
Yes (computergenerated list) Yes (sealed envelopes) No Unclear
Yes (computergenerated list) Yes (sealed envelopes) No Unclear
0%
0%
Unclear Yes Unclear
3.7% (2/27 women from the testosterone group discontinued the intervention) Sample size calculation: 56 Women randomized: 53 Two women did not receive the assigned intervention due to personal reasons (one per group). The results from these two women were not included in the analysis.
Transdermal testosterone in poor responders undergoing IVF
453 (Table 2) (Massin et al., 2006). Neither the studies of Fa ´bregues et al. (2009) nor Kim et al. (2011) had any patient lost to follow up.
Outcomes of interest
Figure 1 process.
Flow chart for the trial identification and selection
With regard to the stimulation protocol, there were also notable differences across the three studies. Massin et al. (2006) used a gonadotrophin-releasing hormone (GnRH) analogue protocol, whereas Kim et al. (2011) used a multiple-dose GnRH antagonist protocol, and both studies applied the same protocol to the intervention and the control group. In contrast, Fa ´bregues et al. (2009) used a GnRH analogue protocol, with different regimens for the two groups with regard to the gonadotrophin stimulation, providing LH supplementation to women in the control group, as shown in Table 2. With reference to the transdermal treatment with testosterone across the studies, in the intervention group, Massin et al. (2006) and Kim et al. (2011) used transdermal testosterone in gel (10 mg of transdermal testosterone for 15–20 days and 12.5 mg for 21 days during pituitary desensitization, respectively), while Fa ´bregues et al. (2009) used patches of 2.5 mg per day for 5 days. Massin et al. (2006) was the only study that used an identical placebo gel in the control group, whereas the other two studies did not blind the patient assignment.
Internal validity of the trials In general, the trials provided complete data regarding methodological aspects of the randomized controlled trial included in this meta-analysis and did not show major biases in their design or execution. All of the trials performed randomization through computer-generated sequences and two of them (Fa ´bregues et al., 2009; Kim et al., 2011) concealed the allocation through sealed envelopes. However, the outcome assessment was not blinded in the included trials and only Massin et al. (2006) blinded patient intervention with an identical placebo gel in the control group. The methodological data regarding the included trials are shown in Table 1. Additionally, only one trial had two patients allocated in the testosterone group who discontinued intervention
Live birth Only two trials, Massin et al. (2006) and Kim et al. (2011), reported this outcome in their published data. However, Fa ´bregues et al. (2009) responded to this study’s request and provided additional information regarding this outcome. The three studies achieved a higher number of live births among women receiving transdermal testosterone treatment (23 events out of 113 patients), compared with the control group (12 events out of 112 patients). Fa ´bregues et al. (2009) observed six live births in the intervention group (n = 31) compared with four live births in the control group (n = 31), whereas Kim et al. (2011) reported 15 (n = 55) and seven (n = 55), respectively, and Massin et al. (2006) reported two and one live births, respectively. The pooled analysis of the data from the three trials showed that women receiving transdermal treatment with testosterone achieved significantly higher number of live births as compared with women undergoing standard ovarian stimulation (RR 1.91, 95% CI 1.01 to 3.63, I2 = 0%; Figure 2A). Clinical pregnancy The three trials achieved a higher number of clinical pregnancies in the transdermal testosterone group (27 events out of 113 patients) compared with the control group (13 events out of 112 patients). The pooled analysis showed a significant higher number of clinical pregnancies among women receiving transdermal testosterone as compared with the control group (RR 2.07, 95% CI 1.13 to 3.78, I2 = 0%; Figure 2B). Clinical pregnancy per embryo transferred None of the trial reported this outcome. However the total number of embryos transferred was calculated from the data provided. Only two trials, Kim et al. (2011) and Massin et al. (2006), achieved higher clinical pregnancies adjusted per embryo transferred among women in the intervention group compared with the control group. A post-hoc analysis showed higher clinical pregnancy per embryo transferred within the intervention group (RR 1.72, 95% CI 0.91 to 3.26, I2 = 0%), although results did not reach statistical significance (Figure 2C). Miscarriage Two trials (n = 172) reported data regarding miscarriage rate, Fa ´bregues et al. (2009) and Kim et al. (2011), with three events out of 86 patients in the intervention group and two events out of 86 patients in the control group. The analysis did not detect significant differences between the two groups (RR 1.50, 95% CI 0.26 to 8.76, I2 = 0%; Supplementary Figure 1A, available online only). Number of fertilized oocytes Two trials (n = 172) provided data regarding this outcome, Fa ´bregues et al. (2009) and Kim et al. (2011). The analysis did not detect significant differences in the mean number
454
Table 2
Description of the included studies.
Study
Massin et al. (2006)
Fa ´bregues et al. (2009)
Design
Double-blind RCT
RCT
Intervention group
Control group
Outcomes
n
Age (years)
Ovarian stimulation protocol
n
Age (years)
Ovarian stimulation protocol
27
36.9 ± 3.8
Transdermal testosterone gel 10 mg daily for 15–20 days during pituitary desensitization GnRH analogue protocol rhFSH (Gonal-F) with the same starting dose and adjustments as the control cycle
26
37.3 ± 4.0
Identical placebo gel daily for 15–20 days during pituitary desensitization
Live birth
GnRH analogue protocol rhFSH (Gonal-F) with the same starting dose and adjustments as the control cycle
Clinical pregnancy Biochemical pregnancy
31
36.5 ± 4.1
Testosterone patches of 2.5 mg daily for 5 days during pituitary desensitization GnRH analogue protocol
36.4 ± 2.0
GnRH analogue protocol
Days 1 and 2, 300 IU rhFSH (Gonal-F) and four ampoules of HMG (Menopur); days 3 and 4, four ampoules of HMG; individualized treatment with HMG on day 5 onwards
Implantation rate
Miscarriage
Oocytes retrieved Metaphase-II oocytes Fertilized oocytes Patients reaching ovum retrieval Dose of rhFSH administered Oestradiol concentrations on HCG day Endometrial thickness on HCG day
M Gonza ´lez-Comadran et al.
Days 1 and 2, 400 and 350 IU rhFSH (Gonal-F); days 3 and 4, 150 IU rhFSH; individualized treatment with rhFSH on day 5 onwards
31
Oocytes retrieved Metaphase-II oocytes Fertilized oocytes Cancellation rate Oestradiol concentrations on HCG day Dose of rhFSH administered Clinical pregnancy
Miscarriage Oocytes retrieved Metaphase-II oocytes Fertilized oocytes Dose of rhFSH administered
Biochemical pregnancy
Clinical pregnancy Day 1, 300 IU rhFSH and adjustments every 3–4 days
Age values are mean ± SD. GnRH = gonadotrophin-releasing hormone; HCG = human chorionic gonadotrophin; HMG = human menopausal gonadotrophin; RCT = randomized controlled trial; rhFSH = recombinant human FSH.
Kim et al. (2011)
RCT
55
37.8 ± 3.0
Transdermal testosterone gel 12.5 mg daily for 21 days during pituitary desensitization GnRH antagonist multiple dose protocol Day 1, 300 IU rhFSH and adjustments every 3–4 days
55
37.9 ± 2.9
GnRH antagonist multiple dose protocol
Live birth
Transdermal testosterone in poor responders undergoing IVF
455 of fertilized oocytes among women receiving transdermal treatment with testosterone compared with women in the control group (RR 1.03, 95% CI 0.60 to 1.46), although the results of this analysis showed moderate inconsistency (I2 = 58%; Supplementary Figure 1B). Number of oocytes retrieved The three trials showed a similar mean number of oocytes retrieved among women receiving transdermal treatment with testosterone compared with women in the control group (RR 1.28, 95% CI 0.77 to 1.78), although the results were moderately inconsistent (I2 = 40%; Supplementary Figure 1C). Number of metaphase-II oocytes retrieved With regard to the mean number of metaphase-II oocytes, the analysis of the three trials did not show significant differences between women receiving transdermal treatment with testosterone compared with the control group (RR 1.11, 95% CI 0.65 to 1.56), although the results were moderately inconsistent (I2 = 42%; Figure 2D). Total dose of FSH administered The three studies provided data regarding this outcome. Women in the control group required higher doses of FSH during ovarian stimulation compared with women receiving transdermal testosterone. The analysis showed that women in the control group required significantly higher amounts of FSH to stimulate follicular growth compared with the intervention group (RR 461.96, 95% CI 611.82 to 312.09, I2 = 0%; Figure 2E). Oestradiol concentrations on HCG day Only two studies reported the mean concentration of oestradiol on the day of HCG administration, Massin et al. (2006) and Fa ´bregues et al. (2009). The analysis did not show significant differences in the mean concentration of oestradiol between the two groups (RR 23.41, 95% CI 218.52 to 171.69), although the results showed high inconsistency (I2 = 83%; Supplementary Figure 1D).
Discussion The present systematic review gathers published evidence and data obtained from original authors to provide pooled estimates regarding the use of transdermal testosterone prior to ovarian stimulation among poor responders undergoing an IVF cycle. The results from this meta-analysis show that pretreatment with transdermal testosterone significantly improved live birth rates, as well as other pregnancy outcomes in poor responders undergoing IVF. A potential role of androgens in folliculogenesis has been suggested by several authors (Fanchin et al., 2011; Gleicher et al., 2011). However, the positive effect of increasing the androgen availability in the ovary and its precise biological role, particularly among women with poor ovarian reserve remain to be clarified (Hillier et al., 1988). This potential stimulatory role of androgens in follicle cell proliferation and basal follicular growth has been investigated in both animal and human models. In rodents,
456
Figure 2
M Gonza ´lez-Comadran et al.
Effectiveness of transdermal testosterone priming versus standard treatment.
treatment with testosterone or dihydrotestosterone led to alterations in the cell cycle of granulosa cells and follicular atresia (Pradeep et al., 2002). Conversely, studies in primates showed that treatment with testosterone resulted in a marked increase in the number of growing follicles and the proliferation of granulosa and thecal cells, as well as a
reduction in granulosa cell apoptosis (Vendola et al., 1999; Weil et al., 1999). Interestingly, this latter group revealed the selective colocalization of the mRNA of androgen receptor and FSH receptor within the growing follicles. They demonstrated a significant positive correlation between these two
Transdermal testosterone in poor responders undergoing IVF receptors at the mRNA level in individual follicles (Weil et al., 1999) and observed increased levels of FSH receptor mRNA in granulosa cells after androgen supplementation. Recently, Nielsen et al. (2011) observed a response in women that was consistent with these findings. These data suggest that androgens enhance follicle responsiveness to FSH, particularly in early antral stages of folliculogenesis, and could to some extent explain the hypersensitivity to gonadotrophin stimulation observed in women with polycystic ovarian syndrome (Fanchin et al., 2011). Taking into consideration that granulosa cells express androgen receptors at early stages of folliculogenesis and that these stages of follicle maturation occur weeks to months before ovulation, it would be reasonable to assume that prolonged androgen supplementation would be associated with an increased pool of follicles available for gonadotrophin stimulation and thus a larger pool of the oocytes available for retrieval, although side effects should also be considered. This assumption would involve particularly the two drugs that exert their effects through systemic androgenization, i.e. testosterone and DHEA, and that have been tested among poor responders and analysed in previous meta-analyses, which appear to improve reproductive outcomes (Sunkara et al., 2011; Bosdou et al., 2012). Apart from this, if it is assumed that androgens have this synergistic effect with FSH on folliculogenesis, the pretreatment of poor responders with androgens during an IVF cycle could not only enhance the number of growing follicles but could also reduce the dose of exogenous FSH required. Interestingly, the results from this meta-analysis show that women pretreated with transdermal testosterone required significantly lower amounts of FSH compared with the control group (RR 461.96, 95% CI 611.82 to 312.09), which in turn, could facilitate adherence to gonadotrophin administration and thus improve reproductive outcomes. Despite this possible increased sensitivity to FSH observed among women in the intervention group, no differences were detected in the number or quality of oocytes retrieved, the fertilization rate or the concentrations of oestradiol reached during gonadotrophin stimulation, although these results showed some level of heterogeneity (I2 25%). Conversely, the clinical pregnancy rate and clinical pregnancy per embryo transferred were greater among women that were pretreated with transdermal testosterone compared with the control group (RR 2.07, 95% CI 1.13 to 3.78 and RR 1.72, 95% CI 0.91 to 3.26, respectively). However, when the clinical pregnancy rate was adjusted per embryo transferred, differences among the two groups were not statistically significant. These results must nonetheless be interpreted with caution, since original data regarding the total number of embryos transferred was not published and thus clinical pregnancy per embryo transferred was calculated from secondary variables. The current study has several potential limitations that need to be considered. First, the number of events included in the meta-analysis is still relatively small: 23 live births in the testosterone group (n = 113) versus 12 in the control group (n = 112). This low rate of events available in the scientific literature leads to treatment estimates being affected by some imprecision, decreasing confidence in the results.
457 In addition, it is necessary to highlight the heterogeneity of the included trials regarding the inclusion criteria, particularly the study by Massin et al. (2006), whose definition of poor response was, to some extent, broad. Recently, in Bologna, a consensus was reached to standardize criteria for defining a woman as a poor responder (Ferraretti et al., 2011). Hence, the application of Bologna criteria in upcoming trials could improve the external validity of their results. There is also a marked heterogeneity with regard to the protocols for ovarian stimulation used throughout the studies, as well as the dose and time of exposure to transdermal testosterone. With reference to the latter, there is limited evidence regarding the dose of testosterone required and what the intraovarian androgen availability would be when testosterone is administered systemically at non-virilizing concentrations (Fanchin et al., 2011). Yet, if several weeks are required for a growing follicle with functional androgen receptors, to reach the antral stage, it is unknown how long this exposure would need to last in order to increase the functional ovarian reserve (Fanchin et al., 2011). In spite of the clinical heterogeneity in the inclusion criteria and the ovarian stimulation protocols in the included trials, the results from the pooled analysis showed a strong consistency for the main review outcomes. Furthermore, the included trials present certain methodological concerns. Most significantly, Massin et al. (2006) and Fa ´bregues et al. (2009) reported a higher average of embryos transferred per woman in the testosterone group compared with the control group, although the potential bias was minimized by the intention-to-treat analysis performed in this meta-analysis. Additionally, the control group in the trial by Fa ´bregues et al. (2009) received higher doses of FSH for ovarian stimulation compared with the testosterone group. Interestingly, only one trial provided information regarding the initial assessment of anti-Mu ¨llerian hormone (AMH) (Massin et al., 2006). In this context, the determination of AMH concentrations could be useful to define a potential subgroup that could benefit to a greater extent from the testosterone treatment. Thus, if the theory of the synergistic role of androgens and FSH on folliculogenesis is assumed plausible (Gleicher et al., 2011), the analysis of possible AMH variations after testosterone treatment could have significance if a correlation was found with a potential increase in the functional ovarian reserve. Despite the limitations described, this study also has some strengths. The primary outcome was live birth, since it should be the most relevant outcome in reproductive disorders. In this context, Fa ´bregues et al. (2009) replied to this study’s request and provided data that were not initially included in their published report. The need to access all of the data from clinical trials is increasingly recognized (Gøtzsche, 2011) as an essential step in the development of rigorous evidence syntheses and to provide clinicians and policy makers with confident treatment effect estimates to make their decisions (Lehman and Loder, 2012), even more when the outcomes are centred on the patient interests (Liberati, 2012). Moreover, all of the included studies in the present meta-analysis were randomized controlled trials. Recently, two other meta-analyses were published to address the effect of androgen supplementation and
458 androgen-modulating agents in poor responders undergoing IVF (Sunkara et al., 2011; Bosdou et al., 2012). Compared with the present study, Sunkara et al. (2011), which included two trials (Massin et al., 2006; Fa ´bregues et al., 2009), concluded that there was no evidence to support androgen supplementation or androgen-modulating agents among poor responders. In contrast, the meta-analysis by Bosdou et al. (2012), based on a pooled analysis of two studies included in the current review (Massin et al., 2006; Kim et al., 2011), reported that live birth rates increased by 11% among women in the testosterone group (risk difference 0.11, 95% CI 0.00 to 0.22). Considering the broad range of values expressed by this confidence interval and expecting some variability between the basal risks through the patients allocated to the control groups, this analysis preferred to calculate the results in relative terms in the pooled analysis. Furthermore, as mentioned before, since androgen receptors are expressed in granulosa cells at early stages of follicle maturation, it is surprising that such a short treatment up to 20 days of testosterone supplementation could achieve significantly higher live birth rates. Hence, extending testosterone supplementation for a longer period could enhance the pool of follicles sensitive to gonadotrophins and therefore increase the number of oocytes available for retrieval. However, this assumption still needs to be tested through properly designed trials, and side effects of long periods of androgenization must be considered. In conclusion, there is evidence from randomized controlled trials to support the use of transdermal testosterone prior to ovarian stimulation in women who are considered poor responders, and this treatment has shown to significantly improve live birth rates and reduce the doses of FSH required for ovarian stimulation. The exact subgroup of poor responders who would benefit from this treatment still needs to be identified. However, the result should be interpreted with caution because of the small number of trials and their clinical heterogeneity. Although trends in all parameters appear to favour testosterone supplementation, further investigations are needed to confirm these findings.
Acknowledgements The authors thank Marta Pulido, MD, for editing the manuscript and editorial assistance.
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10. 1016/j.rbmo.2012.07.011.
References Arslan, M., Bocca, S., Mirkin, S., Barroso, G., Stadtmauer, L., Oehninger, S., 2005. Controlled ovarian hyperstimulation protocols for in vitro fertilization: two decades of experience after the birth of Elizabeth Carr. Fertil. Steril. 84, 555–569. Balasch, J., Fa ´bregues, F., Pen ˜arrubia, J., Carmona, F., Casamitjana, R., Creus, M., Manau, D., Casals, G., Vanrell, J.A., 2006. Pretreatment with transdermal testosterone may improve
M Gonza ´lez-Comadran et al. ovarian response to gonadotrophins in poor-responder IVF patients with normal basal concentrations of FSH. Hum. Reprod. 21, 1884–1893. Bosdou, J.K., Venetis, C.A., Kolibianakis, E.M., Toulis, K.A., Goulis, D.G., Zepiridis, L., et al., 2012. The use of androgens or androgen-modulating agents in poor responders undergoing in vitro fertilization: a systematic review and meta-analysis. Hum. Reprod. Update 18, 127–145. Deeks, J.J., Higgins, J.P.T., Altman, D.G. Chapter 9: Analysing data and undertaking meta-analyses. In: Higgins, J.P.T., Green, S. (Eds.), Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. (accessed June 12.06.2012). Fa ´bregues, F., Pen ˜arrubia, J., Creus, M., Manau, D., Casals, G., Carmona, F., et al., 2009. Transdermal testosterone may improve ovarian response to gonadotrophins in low-responder IVF patients: a randomized, clinical trial. Hum. Reprod. 24, 349–359. Fanchin, R., Frydman, N., Even, M., Berwanger da Silva, A.L., Grynberg, M., Ayoubi, J.M., 2011. Androgens and poor responders: are we ready to take the plunge into clinical therapy? Fertil. Steril. 96, 1062–1065. Farquharson, R.G., Jauniaux, E., Exalto, N., 2005. ESHRE Special Interest Group for Early Pregnancy (SIGEP). Updated and revised nomenclature for description of early pregnancy events. Hum. Reprod. 20, 3008–3011. Ferraretti, A.P., La Marca, A., Fauser, B.C., Tarlatzis, B., Nargund, G., Gianaroli, L., 2011. ESHRE working group on Poor Ovarian Response Definition. ESHRE consensus on the definition of ‘poor response’ to ovarian stimulation for in vitro fertilization: the Bologna criteria. Hum. Reprod. 26, 1616–1624. Fleiss, J.L., 1993. The statistical basis of meta-analysis. Stat. Methods Med. Res. 2, 121–145. Gleicher, N., Weghofer, A., Barad, D.H., 2011. The role of androgens in follicle maturation and ovulation induction: friend or foe of infertility treatment? Reprod. Biol. Endocrinol. 9, 116. Gøtzsche, P.C., 2011. We need access to all data from all clinical trials [editorial]. The Cochrane Library. www.thecochranelibrary. com/details/editorial/1359903/We-need-access-to-all-datafrom-all-clinical-trials.html/ (accessed 8.03.2012). Higgins, J.P., Thompson, S.G., Deeks, J.J., Altman, D.G., 2003. Measuring inconsistency in meta-analysis. B.M.J. 327, 557–560. Higgins, J.P.T., Green, S., 2008. Cochrane Handbook for Systematic Reviews of Interventions Version 5.0.0 [updated February 2008]. The Cochrane Collaboration. (accessed 8.03.2012). Hillier, S.G., Harlow, C.R., Shaw, H.J., Wickings, E.J., Dixson, A.F., Hodges, J.K., 1988. Cellular aspects of pre-ovulatory folliculogenesis in primate ovaries. Hum. Reprod. 3, 507–511. Kailasan, C., Keay, S.D., Wilson, P., Ford, W.C., Jenkins, J.M., 2004. Defining poor ovarian response during IVF cycles, in women aged <40 years and its relationship with treatment outcome. Hum. Reprod. 19, 1544–1547. Kim, C.H., Howles, C.M., Lee, H.A., 2011. The effect of transdermal testosterone gel pretreatment on controlled ovarian stimulation and IVF outcome in low responders. Fertil. Steril. 95, 679–683. Lehman, R., Loder, E., 2012, in press. Missing clinical trial data. B.M.J. doi: http://dx.doi.org/10.1136/bmj.d8158. Liberati, A., 2012. Need to re-align patient-oriented and commercial and academic research [editorial]. The Cochrane Library.
Transdermal testosterone in poor responders undergoing IVF responders undergoing assisted reproduction technique–a prospective, randomized, double-blind study. Hum. Reprod. 21, 1204–1211. Mitwally, M.F., Casper, R.F., 2002. Aromatase inhibition improves ovarian response to follicle-stimulating hormone in poor responders. Fertil. Steril. 77, 776–780. Moher, D., Liberati, A., Tetzlaff, J., Altman, D.G.PRISMA Group, 2009. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS. Med. 6, 1000097. Nielsen, M.E., Rasmussen, I.A., Kristensen, S.G., Christensen, S.T., Møllga ˚rd, K., Wreford Andersen, E., 2011. In human granulosa cells from small antral follicles, androgen receptor mRNA and androgen levels in follicular fluid correlate with FSH receptor mRNA. Mol. Hum. Reprod. 17, 63–70. Polyzos, N.P., Devroey, P., 2011. A systematic review of randomized trials for the treatment of poor ovarian responders: is there any light at the end of the tunnel? Fertil. Steril. 96, 1058–1061. Pradeep, P.K., Li, X., Peegel, H., Menon, K.M., 2002. Dihydrotestosterone inhibits granulosa cell proliferation by decreasing the cyclin D2 mRNA expression and cell cycle arrest at G1 phase. Endocrinology 143, 2930–2935. Sallam, H.N., Ezzeldin, F., Agameya, A.F., Rahman, A.F., El-Garem, Y., 2005. Defining poor responders in assisted reproduction. Int. J. Fertil. Womens Med. 50, 115–120. Sunkara, S.K., Pundir, J., Khalaf, Y., 2011. Effect of androgen supplementation or modulation on ovarian stimulation outcome in poor responders: a meta-analysis. Reprod. Biomed. Online 22, 545–555.
459 Turhan, N.O., 2006. Poor response–the devil is in the definition. Fertil. Steril. 86, 777. Vendola, K., Zhou, J., Wang, J., Famuyiwa, O.A., Bievre, M., Bondy, C.A., 1999. Androgens promote oocyte insulin-like growth factor I expression and initiation of follicle development in the primate ovary. Biol. Reprod. 61, 353–357. Vollenhoven, B., Osianlis, T., Catt, J., 2008. Is there an ideal stimulation regimen for IVF for poor responders and does it change with age? J. Assist. Reprod. Genet. 25, 523–529. Weil, S., Vendola, K., Zhou, J., Bondy, C.A., 1999. Androgen and follicle-stimulating hormone interactions in primate ovarian follicle development. J. Clin. Endocrinol. Metab. 84, 2951–2956. Zegers-Hochschild, F., Adamson, G.D., de Mouzon, J., Ishihara, O., Mansour, R., Nygren, K., et al., 2009. International Committee for Monitoring Assisted Reproductive Technology; World Health Organization. The International Committee for Monitoring Assisted Reproductive Technology (ICMART) and the World Health Organization (WHO) Revised Glossary on ART Terminology. Fertil. Steril. 92, 1520–1524. Declaration: The authors report no financial or commercial conflicts of interest. Received 28 March 2012; refereed 13 July 2012; accepted 18 July 2012.
www.sciencedirect.com www.rbmonline.com
REVIEW
Effects of transdermal testosterone in poor responders undergoing IVF: systematic review and meta-analysis ´lez-Comadran a,b, Montserrat Dura ´n b, Ivan Sola ` c,d, Mireia Gonza ´n Carreras a,f, Miguel A Checa a,f,g,* ´bregues e, Ramo Francisco Fa a Department of Obstetrics and Gynecology, Hospital del Mar, Barcelona, Spain; b International Master in `noma de Barcelona, Barcelona, Spain; c Iberoamerican Cochrane Centre, Reproductive Medicine, Universitat Auto ´blica Institute of Biomedical Research, IIB Sant Pau, Barcelona, Spain; d CIBER de Epidemiologı´a y Salud Pu (CIBERESP), Spain; e Institut Clı´nic of Gynecology, Obstetrics and Neonatology, Hospital Clı´nic-Institut d’Investigacions `diques August Pi i Sunyer (IDIBAPS), Faculty of Medicine, University of Barcelona, Barcelona, Spain; f Universitat Biome `noma de Barcelona, Barcelona, Spain; g Centro de Infertilidad y Reproduccio ´n Humana, Barcelona, Spain Auto
* Corresponding author. E-mail address: [email protected] (M.A. Checa). Miguel A Checa was awarded a Bachelor of medicine and surgery at the University of Barcelona in 2008. She is a specialist in Obstetrics and Gynecology at Hospital del Mar de Barcelona. She was studied for an International Master in reproductive medicine at Universitat Auto `noma de Barcelona in 2011–2012.
Abstract A systematic review and meta-analysis was performed to evaluate the effect of transdermal testosterone preceding ovar-
ian stimulation in women with poor ovarian response undergoing IVF. Studies comparing pretreatment with transdermal testosterone versus standard ovarian stimulation among poor responders were included. The main outcome assessed was live birth. Three trials were included (113 women in the testosterone group, 112 in the control group). Testosterone-treated women achieved significantly higher live birth rate (risk ratio, RR, 1.91, 95% CI 1.01 to 3.63), clinical pregnancy rate (RR 2.07, 95% CI 1.13 to 3.78) and required significantly lower doses of FSH (RR 461.96, 95% CI 611.82 to 312.09). However, differences observed in clinical pregnancy per embryo transferred were not statistically significant (RR 1.72, 95% CI 0.91 to 3.26). No differences were observed regarding number and quality of the oocytes retrieved. In conclusion, transdermal testosterone significantly increases live birth and reduces the doses of FSH required. These findings support the theoretical synergistic role of androgens and FSH on folliculogenesis. The present data should be interpreted with caution because of the small number of trials and clinical heterogeneity. The identification of poor responders that could especially benefit from testosterone treatment should be addressed in further studies. RBMOnline ª 2012, Reproductive Healthcare Ltd. Published by Elsevier Ltd. All rights reserved. KEYWORDS: IVF, live birth, poor responder, randomized controlled trial, transdermal testosterone
1472-6483/$ - see front matter ª 2012, Reproductive Healthcare Ltd. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.rbmo.2012.07.011
Transdermal testosterone in poor responders undergoing IVF
Introduction The poor response to ovarian stimulation among women undergoing IVF is of great concern in reproductive medicine. The reported incidence of poor response varies from 9% to 26% (Vollenhoven et al., 2008). For years, several authors have attempted to define ‘poor response’ in assisted reproduction (Kailasan et al., 2004; Arslan et al., 2005; Sallam et al., 2005; Turhan, 2006). In July 2011, The European Society for Human Reproduction and Embryology, developed a consensus on the definition of ‘poor response’ to ovarian stimulation for IVF, known as Bologna criteria to standardize the definition in a simple and reproducible manner (Ferraretti et al., 2011). In this context, Polyzos and Devroey (2011) among others concluded that the applicability of this definition needs to be tested through clinical trials and discouraged the performance of a meta-analysis with the currently available trials since their results may lead to recommendations of ambiguous value. Certain modalities have been tested through randomized trials to improve the response to gonadotrophin stimulation and thus reproductive outcomes. In daily practice, the most commonly used agents before or during ovarian stimulation are transdermal testosterone, dehydroepiandrosterone (DHEA), aromatase inhibitors, recombinant LH and recombinant human chorionic gonadotrophin (HCG). The results from these trials that aim to increase intraovarian androgen concentrations in poor responders have shown conflicting results (Fanchin et al., 2011). Moreover, some meta-analyses have evaluated the effect of different types of androgen supplementation, such as transdermal testosterone and DHEA, and the duration of the treatment; however, these studies are markedly heterogeneous in their inclusion criteria (Sunkara et al., 2011; Bosdou et al., 2012). The present meta-analysis focused on the intervention of transdermal testosterone administration, which achieves powerful systemic androgenization and subsequently, a greater action of FSH compared with other androgen-modulating agents, such as aromatase inhibitors (Mitwally and Casper, 2002). Live birth should be the most relevant outcome of interest in the treatment of infertility and human reproductive disorders. Therefore, the aim of this meta-analysis is to assess the effect of transdermal testosterone pretreatment on live birth rate among poor responders undergoing ovarian stimulation for IVF.
Materials and methods The study was exempt from Institutional Review Board approval because this was a systematic review and meta-analysis. The preferred reporting items for systematic reviews and meta-analysis (PRISMA statement) were endorsed to report the results of this systematic review (Moher et al., 2009).
Search strategy
451 Cochrane Central Register of Controlled Trials (CENTRAL). The search combined terms and descriptors related to testosterone, poor ovarian response and IVF outcomes. The search strategy was modified to comply with the requirements of each database consulted. Validated filters were added to that strategy to retrieve clinical trials (Higgins and Green, 2008). Moreover, this study searched for ongoing trials in the main clinical trials registers, including www.controlled-trials.com, www.clinicaltrials.gov and the WHO International Clinical Trials Registry Platform (www.who.int/trialsearch). No language limits were used. The reference lists were screened for all relevant articles and overviews. The complete search strategy is available upon request from the authors.
Eligibility criteria The review included randomized controlled clinical trials of women with poor ovarian response undergoing IVF and receiving embryo transfers independently of whether the cause of infertility or subfertility was due to male or female factors. The type of intervention evaluated was the administration of transdermal testosterone preceding gonadotrophin treatment compared with standard gonadotrophin ovarian stimulation protocols without the administration of transdermal testosterone during the period of follicular stimulation.
Outcome measures The main outcome of interest for the review was the live birth rate. Secondary outcomes were clinical pregnancy rate and clinical pregnancy adjusted per embryo transferred, miscarriage, the number of fertilized oocytes, the number of oocytes retrieved, the number of metaphase-II oocytes retrieved, the total dose of FSH administered and oestradiol concentrations on day of HCG injection. All of the outcomes of interest were considered per woman randomized (Supplementary Table 1, available online only). The outcomes were defined according to the terminology recommended in the International Committee Monitoring Assisted Reproductive Technologies glossary (Zegers-Hochschild et al., 2009) and the updated and revised nomenclature for the description of early pregnancy events (Farquharson et al., 2005).
Data extraction The data were collected using standard forms, in which the characteristics of the study design, participants, interventions, comparisons and main results were recorded. Two independent authors (MGC and MDR) judged study eligibility, assessed the risk of bias and extracted data solving discrepancies by agreement, and if needed, reaching consensus with a third author (MAC). The agreement between reviewers was analysed using the weighted kappa for each inclusion criterion (Fleiss, 1993).
Assessment of risk of bias An exhaustive electronic search was performed in the following databases (from their inception until December 2011): MEDLINE, EMBASE, Science Citation Index and the
The risk of bias in the included studies was assessed following the guidance suggested by the Cochrane Collaboration
452
M Gonza ´lez-Comadran et al.
(Higgins and Green, 2008). This study addressed six specific domains (explicit eligibility criteria, sequence generation, allocation concealment, patient blinding, outcome assessor blinding and patients lost to follow up) and assessed the adequacy of the study report in relation to each domain. A judgment of ‘yes’ for all domains indicates a low risk of bias and a judgement of ‘no’ for one or more domains indicates a high risk of bias. The risk of bias in the specific domains was interpreted as unclear when the information was not available. Patients lost to follow up were described in percentages. The risk of bias for the included trials is detailed in Table 1.
Statistical analysis
eligible to be included in the review by one or both reviewers and 16 were excluded because of non-randomized comparisons (n = 3) or because the results were irrelevant for the purpose of the present meta-analysis (n = 13). During the second phase of the inclusion process, out of the four studies, one was excluded because of non-randomized comparisons (Balasch et al., 2006). Finally, three randomized controlled trials met the inclusion criteria and were included (Massin et al., 2006; Fa ´bregues et al., 2009; Kim et al., 2011). A flow chart of the included trials is shown in Figure 1. The two reviewers achieved good agreement on the selection of the trials (weighted kappa 0.694, P < 0.001, standard error 0.195).
Description of the studies For each study, the treatment effect was measured with risk ratios (RR) for dichotomous outcomes and mean differences for continuous outcomes, which were presented with their corresponding 95% confidence intervals (CI). Event data were extracted following the intention-to-treat principle. When possible, outcome data from each study were pooled using a Mantel–Haenszel model, applying the fixed-effects model. A post-hoc analysis calculated the risk for clinical pregnancy per embryo transferred using the estimates of the intervention effects from original trials and their standard errors using the inverse variance approach (Deeks et al., 2011). Statistical heterogeneity was quantified using the I2 statistic, which describes the percentage of total variation across studies that is due to heterogeneity rather than sampling error (Higgins et al., 2003). All analyses were conducted using Review Manager version 5.1 (2011, The Nordic Cochrane Centre, Copenhagen, The Cochrane Collaboration).
Results Out of a total of 20 documents identified in the initial electronic search, four studies were considered potentially Table 1
A total of 225 women were randomized to an intervention group (n = 113) and a control group (n = 112). The mean age reported across the studies was comparable (Table 2) and each study reported no statistical difference regarding this parameter among the intervention and the control group. The inclusion criteria were relatively heterogeneous among the studies, especially regarding the main inclusion criteria, which was ‘poor response to ovarian stimulation during an IVF/intracytoplasmic sperm injection cycle’. This item was defined by Fa ´bregues et al. (2009) as the failure to produce 3 follicles with a mean diameter 14 mm or the collection of 3 follicles at retrieval, whereas Kim et al. (2011) set the limit of poor response the production of 3 follicles with a mean diameter 16 mm or the collection of 3 follicles at retrieval. In contrast, Massin et al. (2006) defined poor response as a plasma oestradiol value below 1.200 pg/ml on HCG day and the collection of 5 follicles at retrieval, adding as a necessary criteria for enrolment the evidence of a decreased ovarian reserve at day 3 of a spontaneous cycle, as determined with plasma hormonal values outside the normal range (FSH >12 IU/l, oestradiol >70 pg/ml and inhibin B <45 pg/ml).
Methodological data of the clinical trials included in the review.
Criterion
Massin et al. (2006)
´bregues et al. Fa (2009)
Kim et al. (2011)
Explicit eligibility criteria Sequence generation Allocation concealed Patient blinding Outcome assessor blinding Patients lost to follow up (%)
Yes
Yes
Yes
Yes (random permutation table)
Yes (computergenerated list) Yes (sealed envelopes) No Unclear
Yes (computergenerated list) Yes (sealed envelopes) No Unclear
0%
0%
Unclear Yes Unclear
3.7% (2/27 women from the testosterone group discontinued the intervention) Sample size calculation: 56 Women randomized: 53 Two women did not receive the assigned intervention due to personal reasons (one per group). The results from these two women were not included in the analysis.
Transdermal testosterone in poor responders undergoing IVF
453 (Table 2) (Massin et al., 2006). Neither the studies of Fa ´bregues et al. (2009) nor Kim et al. (2011) had any patient lost to follow up.
Outcomes of interest
Figure 1 process.
Flow chart for the trial identification and selection
With regard to the stimulation protocol, there were also notable differences across the three studies. Massin et al. (2006) used a gonadotrophin-releasing hormone (GnRH) analogue protocol, whereas Kim et al. (2011) used a multiple-dose GnRH antagonist protocol, and both studies applied the same protocol to the intervention and the control group. In contrast, Fa ´bregues et al. (2009) used a GnRH analogue protocol, with different regimens for the two groups with regard to the gonadotrophin stimulation, providing LH supplementation to women in the control group, as shown in Table 2. With reference to the transdermal treatment with testosterone across the studies, in the intervention group, Massin et al. (2006) and Kim et al. (2011) used transdermal testosterone in gel (10 mg of transdermal testosterone for 15–20 days and 12.5 mg for 21 days during pituitary desensitization, respectively), while Fa ´bregues et al. (2009) used patches of 2.5 mg per day for 5 days. Massin et al. (2006) was the only study that used an identical placebo gel in the control group, whereas the other two studies did not blind the patient assignment.
Internal validity of the trials In general, the trials provided complete data regarding methodological aspects of the randomized controlled trial included in this meta-analysis and did not show major biases in their design or execution. All of the trials performed randomization through computer-generated sequences and two of them (Fa ´bregues et al., 2009; Kim et al., 2011) concealed the allocation through sealed envelopes. However, the outcome assessment was not blinded in the included trials and only Massin et al. (2006) blinded patient intervention with an identical placebo gel in the control group. The methodological data regarding the included trials are shown in Table 1. Additionally, only one trial had two patients allocated in the testosterone group who discontinued intervention
Live birth Only two trials, Massin et al. (2006) and Kim et al. (2011), reported this outcome in their published data. However, Fa ´bregues et al. (2009) responded to this study’s request and provided additional information regarding this outcome. The three studies achieved a higher number of live births among women receiving transdermal testosterone treatment (23 events out of 113 patients), compared with the control group (12 events out of 112 patients). Fa ´bregues et al. (2009) observed six live births in the intervention group (n = 31) compared with four live births in the control group (n = 31), whereas Kim et al. (2011) reported 15 (n = 55) and seven (n = 55), respectively, and Massin et al. (2006) reported two and one live births, respectively. The pooled analysis of the data from the three trials showed that women receiving transdermal treatment with testosterone achieved significantly higher number of live births as compared with women undergoing standard ovarian stimulation (RR 1.91, 95% CI 1.01 to 3.63, I2 = 0%; Figure 2A). Clinical pregnancy The three trials achieved a higher number of clinical pregnancies in the transdermal testosterone group (27 events out of 113 patients) compared with the control group (13 events out of 112 patients). The pooled analysis showed a significant higher number of clinical pregnancies among women receiving transdermal testosterone as compared with the control group (RR 2.07, 95% CI 1.13 to 3.78, I2 = 0%; Figure 2B). Clinical pregnancy per embryo transferred None of the trial reported this outcome. However the total number of embryos transferred was calculated from the data provided. Only two trials, Kim et al. (2011) and Massin et al. (2006), achieved higher clinical pregnancies adjusted per embryo transferred among women in the intervention group compared with the control group. A post-hoc analysis showed higher clinical pregnancy per embryo transferred within the intervention group (RR 1.72, 95% CI 0.91 to 3.26, I2 = 0%), although results did not reach statistical significance (Figure 2C). Miscarriage Two trials (n = 172) reported data regarding miscarriage rate, Fa ´bregues et al. (2009) and Kim et al. (2011), with three events out of 86 patients in the intervention group and two events out of 86 patients in the control group. The analysis did not detect significant differences between the two groups (RR 1.50, 95% CI 0.26 to 8.76, I2 = 0%; Supplementary Figure 1A, available online only). Number of fertilized oocytes Two trials (n = 172) provided data regarding this outcome, Fa ´bregues et al. (2009) and Kim et al. (2011). The analysis did not detect significant differences in the mean number
454
Table 2
Description of the included studies.
Study
Massin et al. (2006)
Fa ´bregues et al. (2009)
Design
Double-blind RCT
RCT
Intervention group
Control group
Outcomes
n
Age (years)
Ovarian stimulation protocol
n
Age (years)
Ovarian stimulation protocol
27
36.9 ± 3.8
Transdermal testosterone gel 10 mg daily for 15–20 days during pituitary desensitization GnRH analogue protocol rhFSH (Gonal-F) with the same starting dose and adjustments as the control cycle
26
37.3 ± 4.0
Identical placebo gel daily for 15–20 days during pituitary desensitization
Live birth
GnRH analogue protocol rhFSH (Gonal-F) with the same starting dose and adjustments as the control cycle
Clinical pregnancy Biochemical pregnancy
31
36.5 ± 4.1
Testosterone patches of 2.5 mg daily for 5 days during pituitary desensitization GnRH analogue protocol
36.4 ± 2.0
GnRH analogue protocol
Days 1 and 2, 300 IU rhFSH (Gonal-F) and four ampoules of HMG (Menopur); days 3 and 4, four ampoules of HMG; individualized treatment with HMG on day 5 onwards
Implantation rate
Miscarriage
Oocytes retrieved Metaphase-II oocytes Fertilized oocytes Patients reaching ovum retrieval Dose of rhFSH administered Oestradiol concentrations on HCG day Endometrial thickness on HCG day
M Gonza ´lez-Comadran et al.
Days 1 and 2, 400 and 350 IU rhFSH (Gonal-F); days 3 and 4, 150 IU rhFSH; individualized treatment with rhFSH on day 5 onwards
31
Oocytes retrieved Metaphase-II oocytes Fertilized oocytes Cancellation rate Oestradiol concentrations on HCG day Dose of rhFSH administered Clinical pregnancy
Miscarriage Oocytes retrieved Metaphase-II oocytes Fertilized oocytes Dose of rhFSH administered
Biochemical pregnancy
Clinical pregnancy Day 1, 300 IU rhFSH and adjustments every 3–4 days
Age values are mean ± SD. GnRH = gonadotrophin-releasing hormone; HCG = human chorionic gonadotrophin; HMG = human menopausal gonadotrophin; RCT = randomized controlled trial; rhFSH = recombinant human FSH.
Kim et al. (2011)
RCT
55
37.8 ± 3.0
Transdermal testosterone gel 12.5 mg daily for 21 days during pituitary desensitization GnRH antagonist multiple dose protocol Day 1, 300 IU rhFSH and adjustments every 3–4 days
55
37.9 ± 2.9
GnRH antagonist multiple dose protocol
Live birth
Transdermal testosterone in poor responders undergoing IVF
455 of fertilized oocytes among women receiving transdermal treatment with testosterone compared with women in the control group (RR 1.03, 95% CI 0.60 to 1.46), although the results of this analysis showed moderate inconsistency (I2 = 58%; Supplementary Figure 1B). Number of oocytes retrieved The three trials showed a similar mean number of oocytes retrieved among women receiving transdermal treatment with testosterone compared with women in the control group (RR 1.28, 95% CI 0.77 to 1.78), although the results were moderately inconsistent (I2 = 40%; Supplementary Figure 1C). Number of metaphase-II oocytes retrieved With regard to the mean number of metaphase-II oocytes, the analysis of the three trials did not show significant differences between women receiving transdermal treatment with testosterone compared with the control group (RR 1.11, 95% CI 0.65 to 1.56), although the results were moderately inconsistent (I2 = 42%; Figure 2D). Total dose of FSH administered The three studies provided data regarding this outcome. Women in the control group required higher doses of FSH during ovarian stimulation compared with women receiving transdermal testosterone. The analysis showed that women in the control group required significantly higher amounts of FSH to stimulate follicular growth compared with the intervention group (RR 461.96, 95% CI 611.82 to 312.09, I2 = 0%; Figure 2E). Oestradiol concentrations on HCG day Only two studies reported the mean concentration of oestradiol on the day of HCG administration, Massin et al. (2006) and Fa ´bregues et al. (2009). The analysis did not show significant differences in the mean concentration of oestradiol between the two groups (RR 23.41, 95% CI 218.52 to 171.69), although the results showed high inconsistency (I2 = 83%; Supplementary Figure 1D).
Discussion The present systematic review gathers published evidence and data obtained from original authors to provide pooled estimates regarding the use of transdermal testosterone prior to ovarian stimulation among poor responders undergoing an IVF cycle. The results from this meta-analysis show that pretreatment with transdermal testosterone significantly improved live birth rates, as well as other pregnancy outcomes in poor responders undergoing IVF. A potential role of androgens in folliculogenesis has been suggested by several authors (Fanchin et al., 2011; Gleicher et al., 2011). However, the positive effect of increasing the androgen availability in the ovary and its precise biological role, particularly among women with poor ovarian reserve remain to be clarified (Hillier et al., 1988). This potential stimulatory role of androgens in follicle cell proliferation and basal follicular growth has been investigated in both animal and human models. In rodents,
456
Figure 2
M Gonza ´lez-Comadran et al.
Effectiveness of transdermal testosterone priming versus standard treatment.
treatment with testosterone or dihydrotestosterone led to alterations in the cell cycle of granulosa cells and follicular atresia (Pradeep et al., 2002). Conversely, studies in primates showed that treatment with testosterone resulted in a marked increase in the number of growing follicles and the proliferation of granulosa and thecal cells, as well as a
reduction in granulosa cell apoptosis (Vendola et al., 1999; Weil et al., 1999). Interestingly, this latter group revealed the selective colocalization of the mRNA of androgen receptor and FSH receptor within the growing follicles. They demonstrated a significant positive correlation between these two
Transdermal testosterone in poor responders undergoing IVF receptors at the mRNA level in individual follicles (Weil et al., 1999) and observed increased levels of FSH receptor mRNA in granulosa cells after androgen supplementation. Recently, Nielsen et al. (2011) observed a response in women that was consistent with these findings. These data suggest that androgens enhance follicle responsiveness to FSH, particularly in early antral stages of folliculogenesis, and could to some extent explain the hypersensitivity to gonadotrophin stimulation observed in women with polycystic ovarian syndrome (Fanchin et al., 2011). Taking into consideration that granulosa cells express androgen receptors at early stages of folliculogenesis and that these stages of follicle maturation occur weeks to months before ovulation, it would be reasonable to assume that prolonged androgen supplementation would be associated with an increased pool of follicles available for gonadotrophin stimulation and thus a larger pool of the oocytes available for retrieval, although side effects should also be considered. This assumption would involve particularly the two drugs that exert their effects through systemic androgenization, i.e. testosterone and DHEA, and that have been tested among poor responders and analysed in previous meta-analyses, which appear to improve reproductive outcomes (Sunkara et al., 2011; Bosdou et al., 2012). Apart from this, if it is assumed that androgens have this synergistic effect with FSH on folliculogenesis, the pretreatment of poor responders with androgens during an IVF cycle could not only enhance the number of growing follicles but could also reduce the dose of exogenous FSH required. Interestingly, the results from this meta-analysis show that women pretreated with transdermal testosterone required significantly lower amounts of FSH compared with the control group (RR 461.96, 95% CI 611.82 to 312.09), which in turn, could facilitate adherence to gonadotrophin administration and thus improve reproductive outcomes. Despite this possible increased sensitivity to FSH observed among women in the intervention group, no differences were detected in the number or quality of oocytes retrieved, the fertilization rate or the concentrations of oestradiol reached during gonadotrophin stimulation, although these results showed some level of heterogeneity (I2 25%). Conversely, the clinical pregnancy rate and clinical pregnancy per embryo transferred were greater among women that were pretreated with transdermal testosterone compared with the control group (RR 2.07, 95% CI 1.13 to 3.78 and RR 1.72, 95% CI 0.91 to 3.26, respectively). However, when the clinical pregnancy rate was adjusted per embryo transferred, differences among the two groups were not statistically significant. These results must nonetheless be interpreted with caution, since original data regarding the total number of embryos transferred was not published and thus clinical pregnancy per embryo transferred was calculated from secondary variables. The current study has several potential limitations that need to be considered. First, the number of events included in the meta-analysis is still relatively small: 23 live births in the testosterone group (n = 113) versus 12 in the control group (n = 112). This low rate of events available in the scientific literature leads to treatment estimates being affected by some imprecision, decreasing confidence in the results.
457 In addition, it is necessary to highlight the heterogeneity of the included trials regarding the inclusion criteria, particularly the study by Massin et al. (2006), whose definition of poor response was, to some extent, broad. Recently, in Bologna, a consensus was reached to standardize criteria for defining a woman as a poor responder (Ferraretti et al., 2011). Hence, the application of Bologna criteria in upcoming trials could improve the external validity of their results. There is also a marked heterogeneity with regard to the protocols for ovarian stimulation used throughout the studies, as well as the dose and time of exposure to transdermal testosterone. With reference to the latter, there is limited evidence regarding the dose of testosterone required and what the intraovarian androgen availability would be when testosterone is administered systemically at non-virilizing concentrations (Fanchin et al., 2011). Yet, if several weeks are required for a growing follicle with functional androgen receptors, to reach the antral stage, it is unknown how long this exposure would need to last in order to increase the functional ovarian reserve (Fanchin et al., 2011). In spite of the clinical heterogeneity in the inclusion criteria and the ovarian stimulation protocols in the included trials, the results from the pooled analysis showed a strong consistency for the main review outcomes. Furthermore, the included trials present certain methodological concerns. Most significantly, Massin et al. (2006) and Fa ´bregues et al. (2009) reported a higher average of embryos transferred per woman in the testosterone group compared with the control group, although the potential bias was minimized by the intention-to-treat analysis performed in this meta-analysis. Additionally, the control group in the trial by Fa ´bregues et al. (2009) received higher doses of FSH for ovarian stimulation compared with the testosterone group. Interestingly, only one trial provided information regarding the initial assessment of anti-Mu ¨llerian hormone (AMH) (Massin et al., 2006). In this context, the determination of AMH concentrations could be useful to define a potential subgroup that could benefit to a greater extent from the testosterone treatment. Thus, if the theory of the synergistic role of androgens and FSH on folliculogenesis is assumed plausible (Gleicher et al., 2011), the analysis of possible AMH variations after testosterone treatment could have significance if a correlation was found with a potential increase in the functional ovarian reserve. Despite the limitations described, this study also has some strengths. The primary outcome was live birth, since it should be the most relevant outcome in reproductive disorders. In this context, Fa ´bregues et al. (2009) replied to this study’s request and provided data that were not initially included in their published report. The need to access all of the data from clinical trials is increasingly recognized (Gøtzsche, 2011) as an essential step in the development of rigorous evidence syntheses and to provide clinicians and policy makers with confident treatment effect estimates to make their decisions (Lehman and Loder, 2012), even more when the outcomes are centred on the patient interests (Liberati, 2012). Moreover, all of the included studies in the present meta-analysis were randomized controlled trials. Recently, two other meta-analyses were published to address the effect of androgen supplementation and
458 androgen-modulating agents in poor responders undergoing IVF (Sunkara et al., 2011; Bosdou et al., 2012). Compared with the present study, Sunkara et al. (2011), which included two trials (Massin et al., 2006; Fa ´bregues et al., 2009), concluded that there was no evidence to support androgen supplementation or androgen-modulating agents among poor responders. In contrast, the meta-analysis by Bosdou et al. (2012), based on a pooled analysis of two studies included in the current review (Massin et al., 2006; Kim et al., 2011), reported that live birth rates increased by 11% among women in the testosterone group (risk difference 0.11, 95% CI 0.00 to 0.22). Considering the broad range of values expressed by this confidence interval and expecting some variability between the basal risks through the patients allocated to the control groups, this analysis preferred to calculate the results in relative terms in the pooled analysis. Furthermore, as mentioned before, since androgen receptors are expressed in granulosa cells at early stages of follicle maturation, it is surprising that such a short treatment up to 20 days of testosterone supplementation could achieve significantly higher live birth rates. Hence, extending testosterone supplementation for a longer period could enhance the pool of follicles sensitive to gonadotrophins and therefore increase the number of oocytes available for retrieval. However, this assumption still needs to be tested through properly designed trials, and side effects of long periods of androgenization must be considered. In conclusion, there is evidence from randomized controlled trials to support the use of transdermal testosterone prior to ovarian stimulation in women who are considered poor responders, and this treatment has shown to significantly improve live birth rates and reduce the doses of FSH required for ovarian stimulation. The exact subgroup of poor responders who would benefit from this treatment still needs to be identified. However, the result should be interpreted with caution because of the small number of trials and their clinical heterogeneity. Although trends in all parameters appear to favour testosterone supplementation, further investigations are needed to confirm these findings.
Acknowledgements The authors thank Marta Pulido, MD, for editing the manuscript and editorial assistance.
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10. 1016/j.rbmo.2012.07.011.
References Arslan, M., Bocca, S., Mirkin, S., Barroso, G., Stadtmauer, L., Oehninger, S., 2005. Controlled ovarian hyperstimulation protocols for in vitro fertilization: two decades of experience after the birth of Elizabeth Carr. Fertil. Steril. 84, 555–569. Balasch, J., Fa ´bregues, F., Pen ˜arrubia, J., Carmona, F., Casamitjana, R., Creus, M., Manau, D., Casals, G., Vanrell, J.A., 2006. Pretreatment with transdermal testosterone may improve
M Gonza ´lez-Comadran et al. ovarian response to gonadotrophins in poor-responder IVF patients with normal basal concentrations of FSH. Hum. Reprod. 21, 1884–1893. Bosdou, J.K., Venetis, C.A., Kolibianakis, E.M., Toulis, K.A., Goulis, D.G., Zepiridis, L., et al., 2012. The use of androgens or androgen-modulating agents in poor responders undergoing in vitro fertilization: a systematic review and meta-analysis. Hum. Reprod. Update 18, 127–145. Deeks, J.J., Higgins, J.P.T., Altman, D.G. Chapter 9: Analysing data and undertaking meta-analyses. In: Higgins, J.P.T., Green, S. (Eds.), Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011.
Transdermal testosterone in poor responders undergoing IVF responders undergoing assisted reproduction technique–a prospective, randomized, double-blind study. Hum. Reprod. 21, 1204–1211. Mitwally, M.F., Casper, R.F., 2002. Aromatase inhibition improves ovarian response to follicle-stimulating hormone in poor responders. Fertil. Steril. 77, 776–780. Moher, D., Liberati, A., Tetzlaff, J., Altman, D.G.PRISMA Group, 2009. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS. Med. 6, 1000097. Nielsen, M.E., Rasmussen, I.A., Kristensen, S.G., Christensen, S.T., Møllga ˚rd, K., Wreford Andersen, E., 2011. In human granulosa cells from small antral follicles, androgen receptor mRNA and androgen levels in follicular fluid correlate with FSH receptor mRNA. Mol. Hum. Reprod. 17, 63–70. Polyzos, N.P., Devroey, P., 2011. A systematic review of randomized trials for the treatment of poor ovarian responders: is there any light at the end of the tunnel? Fertil. Steril. 96, 1058–1061. Pradeep, P.K., Li, X., Peegel, H., Menon, K.M., 2002. Dihydrotestosterone inhibits granulosa cell proliferation by decreasing the cyclin D2 mRNA expression and cell cycle arrest at G1 phase. Endocrinology 143, 2930–2935. Sallam, H.N., Ezzeldin, F., Agameya, A.F., Rahman, A.F., El-Garem, Y., 2005. Defining poor responders in assisted reproduction. Int. J. Fertil. Womens Med. 50, 115–120. Sunkara, S.K., Pundir, J., Khalaf, Y., 2011. Effect of androgen supplementation or modulation on ovarian stimulation outcome in poor responders: a meta-analysis. Reprod. Biomed. Online 22, 545–555.
459 Turhan, N.O., 2006. Poor response–the devil is in the definition. Fertil. Steril. 86, 777. Vendola, K., Zhou, J., Wang, J., Famuyiwa, O.A., Bievre, M., Bondy, C.A., 1999. Androgens promote oocyte insulin-like growth factor I expression and initiation of follicle development in the primate ovary. Biol. Reprod. 61, 353–357. Vollenhoven, B., Osianlis, T., Catt, J., 2008. Is there an ideal stimulation regimen for IVF for poor responders and does it change with age? J. Assist. Reprod. Genet. 25, 523–529. Weil, S., Vendola, K., Zhou, J., Bondy, C.A., 1999. Androgen and follicle-stimulating hormone interactions in primate ovarian follicle development. J. Clin. Endocrinol. Metab. 84, 2951–2956. Zegers-Hochschild, F., Adamson, G.D., de Mouzon, J., Ishihara, O., Mansour, R., Nygren, K., et al., 2009. International Committee for Monitoring Assisted Reproductive Technology; World Health Organization. The International Committee for Monitoring Assisted Reproductive Technology (ICMART) and the World Health Organization (WHO) Revised Glossary on ART Terminology. Fertil. Steril. 92, 1520–1524. Declaration: The authors report no financial or commercial conflicts of interest. Received 28 March 2012; refereed 13 July 2012; accepted 18 July 2012.