Tratamiento con agonistas de la GnRH a largo plazo antes de la fertilización in vitro (FIV) para mejorar los resultados de la fertilidad en pacientes con endometriosis
Resumen
Antecedentes
Se sabe que la endometriosis tiene un impacto en la fertilidad y es frecuente que las pacientes afectadas necesiten tratamientos de fertilidad, incluida la fertilización in vitro (FIV) o la inyección intracitoplasmática de espermatozoides (ICSI), para mejorar la probabilidad de embarazo. Se ha postulado que el tratamiento a largo plazo con agonistas de la hormona liberadora de gonadotrofina (GnRH) antes de la FIV o ICSI puede mejorar los resultados del embarazo. Esta revisión sistemática sustituye a la revisión Cochrane anterior sobre este tema (Sallam 2006).
Objetivos
Determinar la efectividad y la seguridad del tratamiento con agonistas de la hormona liberadora de gonadotrofina (GnRH) a largo plazo (mínimo 3 meses) versus ningún tratamiento previo u otras modalidades de tratamiento previo, como la anticoncepción oral combinada continua a largo plazo (AOC) o el tratamiento quirúrgico del endometrioma, antes de la fertilización in vitro estándar (FIV) o la inyección intracitoplásmica de espermatozoides (ICSI) en pacientes con endometriosis.
Métodos de búsqueda
Se hicieron búsquedas en las siguientes bases de datos electrónicas, desde su inicio hasta el 8 de enero 2019: Registro Especializado Cochrane de Ensayos Controlados de Ginecología y Fertilidad, CENTRAL a través del Registro Cochrane CENTRAL de Estudios ONLINE (CRSO), MEDLINE, Embase, PsycINFO, Cumulative Index to Nursing and Allied Health Literature (CINAHL). Se realizaron búsquedas en los registros de ensayos para identificar ensayos no publicados y en curso. También se realizaron búsquedas de otros ensayos pertinentes en DARE (Database of Abstracts of Reviews of Effects), Web of Knowledge, OpenGrey, Latin American and Caribbean Health Science Information Database (LILACS), PubMed, Google y en las listas de referencias de los artículos relevantes.
Criterios de selección
Ensayos controlados aleatorizados (ECA) que incorporaron a pacientes con endometriosis diagnosticada quirúrgicamente y que compararon el uso de cualquier tipo de agonista de la GnRH durante al menos tres meses antes de un protocolo de FIV/ICSI con ningún tratamiento previo u otras modalidades de tratamiento previo, específicamente el uso de AOC continuo a largo plazo (mínimo de 6 semanas) o la escisión quirúrgica del endometrioma dentro de los seis meses anteriores a la FIV/ICSI estándar. Los resultados primarios fueron la tasa de nacidos vivos y la tasa de embarazos por paciente asignada al azar.
Obtención y análisis de los datos
Dos autores de revisión independientes evaluaron los estudios según los criterios de inclusión, extrajeron los datos y evaluaron el riesgo de sesgo. En caso necesario, se consultó a un tercer autor. Se estableció contacto con los autores de los estudios, según se requirió. Los resultados dicotómicos se analizaron mediante los riesgos relativos (RR) de Mantel‐Haenszel, los intervalos de confianza (IC) del 95% y un modelo de efectos fijos. Para un número pequeño de eventos, se utilizó un odds‐ratio de Peto (OR) con un IC del 95% en su lugar. Se analizaron los resultados continuos mediante la diferencia de medias (DM) entre los grupos y se presentaron con IC del 95%. Se estudió la heterogeneidad de los estudios mediante la estadística I2. Se evaluó la calidad de la evidencia utilizando los criterios GRADE.
Resultados principales
Se incluyeron ocho ECA de diseño paralelo, con un total de 640 participantes. No se evaluó ninguno de los estudios como de riesgo de sesgo bajo en todos los dominios y la principal limitación fue la falta de cegamiento. Con el uso de los criterios GRADE, la calidad de la evidencia varió de muy baja a baja.
Tratamiento con agonistas de la GnRH a largo plazo versus ningún tratamiento previo
No se sabe si el tratamiento con agonistas de la GnRH a largo plazo afecta la tasa de nacidos vivos (RR 0,48; IC del 95%: 0,26 a 0,87; 1 ECA, n = 147; I2 no calculable; evidencia de calidad muy baja) o la tasa de complicaciones generales (OR de Peto 1,23; IC del 95%: 0,37; a 4,14; 3 ECA, n = 318; I2 = 73%; evidencia de calidad muy baja) en comparación con la FIV/ICSI estándar. Además, no se sabe si esta intervención afecta la tasa de embarazo clínico (RR 1,13; IC del 95%: 0,91 a 1,41; 6 ECA; n = 552; I2 = 66%; evidencia de calidad muy baja), la tasa de embarazo múltiple (OR de Peto 0,14; IC del 95%: 0,03 a 0,56; 2 ECA; n = 208; I2 = 0%; evidencia de calidad muy baja), la tasa de aborto espontáneo (OR de Peto 0.45, IC del 95%: 0,10 a 2,00; 2 ECA, n = 208; I2 = 0%; evidencia de calidad muy baja), el número medio de ovocitos (DM 0,72; IC del 95%: 0,06 a 1,38; 4 ECA, n = 385; I2 = 81%; evidencia de calidad muy baja) o el número medio de embriones (DM ‐0,76; IC del 95%: ‐1,33 a ‐0,19; 2 ECA, n = 267; I2: 0%; evidencia de calidad muy baja).
Tratamiento con agonistas de la GnRH a largo plazo versus AOC continuo a largo plazo
Ningún estudio informó de esta comparación.
Tratamiento con agonistas de la GnRH a largo plazo versus tratamiento quirúrgico del endometrioma
Ningún estudio informó de esta comparación.
Conclusiones de los autores
Esta revisión plantea preguntas importantes con respecto al mérito del tratamiento con agonistas de la GnRH a largo plazo en comparación con ningún tratamiento previo antes de la FIV/ICSI estándar en pacientes con endometriosis. En contraposición a los hallazgos anteriores, no se sabe si el tratamiento con agonistas de la GnRH a largo plazo repercute en la tasa de nacidos vivos o en la tasa de complicaciones en comparación con la FIV/ICSI estándar. Además, no se sabe si esta intervención repercute en la tasa de embarazo clínico, la tasa de embarazo múltiple, la tasa de aborto espontáneo, el número medio de ovocitos y el número medio de embriones. En vista de la escasez y la calidad muy baja de los datos existentes, particularmente para los resultados primarios examinados, se necesitan ensayos adicionales de alta calidad para determinar de modo definitivo la repercusión del tratamiento con agonistas de GnRH a largo plazo sobre los resultados de la FIV/ICSI, no solo en comparación con ningún tratamiento previo, sino también en comparación con otras alternativas propuestas para el tratamiento de la endometriosis.
PICOs
Resumen en términos sencillos
Retroalimentación negativa a largo plazo de la hipófisis antes de la fertilización in vitro (FIV) para pacientes con endometriosis
Pregunta de la revisión
Se revisó la eficacia y la seguridad del tratamiento de pacientes con endometriosis conocida (una enfermedad caracterizada por la presencia de tejido endometrial fuera de la cavidad uterina) con agonistas de la hormona liberadora de gonadotrofina (GnRH) durante un período de tres a seis meses antes de la fecundación in vitro (FIV) y la inyección intracitoplásmica de espermatozoides (ICSI). El objetivo fue comparar esta intervención con ningún tratamiento antes de la FIV/ICSI, el tratamiento previo con anticoncepción oral combinada continua a largo plazo (AOC) o el tratamiento quirúrgico para extirpar endometriomas (quistes que se forman en los ovarios como consecuencia de la endometriosis).
Antecedentes
Muchas pacientes afectadas por endometriosis sufren de infertilidad y, como resultado, pueden buscar tratamiento de FIV/ICSI. Se sabe que la FIV/ICSI tiene menos éxito en pacientes con endometriosis y se ha propuesto una variedad de intervenciones antes de la FIV/ICSI para intentar mejorar los resultados. Estos incluyen el tratamiento con agonistas de la GnRH a largo plazo, el tratamiento con AOC continua a largo plazo o la cirugía para extirpar los endometriomas.
Características de los estudios
Se encontraron ocho ensayos controlados aleatorizados que compararon el tratamiento con agonistas de la GnRH a largo plazo con ningún tratamiento previo y que incluyeron un total de 640 pacientes con endometriosis antes de la FIV/ICSI. La evidencia está actualizada hasta enero de 2019.
Resultados clave
En comparación con ningún tratamiento previo, no se sabe si el tratamiento con agonistas de la GnRH a largo plazo antes de la FIV/ICSI en pacientes con endometriosis afecta la tasa de nacidos vivos. La evidencia indica que si se supone que la probabilidad de que la tasa de nacidos vivos sea del 36% sin tratamiento previo, la probabilidad después del tratamiento con agonistas de la GnRH a largo plazo sería de entre un 9% y un 31%. Tampoco se sabe si esta intervención afecta la tasa de complicaciones, la tasa de embarazo clínico, la tasa de embarazo múltiple, la tasa de aborto espontáneo, el número medio de ovocitos y el número medio de embriones. Ningún estudio comparó el tratamiento con agonistas de la GnRH a largo plazo con el tratamiento continuo a largo plazo con AOC o la cirugía para extirpar los endometriomas.
Calidad de la evidencia
La evidencia fue de muy baja calidad. Las principales limitaciones de la evidencia fueron la falta de cegamiento (el proceso en que las mujeres que participan en el ensayo, así como el personal de investigación, no son conscientes de la intervención utilizada), la inconsistencia (diferencias entre distintos estudios) y la imprecisión (error aleatorio y tamaño pequeño de cada estudio).
Authors' conclusions
Summary of findings
| Long‐term GnRH agonist therapy compared to no additional therapy in patients with endometriosis prior to standard IVF/ICSI | ||||||
| Patient or population: patients with endometriosis prior to standard IVF/ICSI | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Quality of the evidence | Comments | |
| Risk with no additional therapy | Risk with long‐term GnRH agonist therapy | |||||
| Live birth rate | Study population | RR 0.48 | 147 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of long‐term GnRH agonist therapy on live birth rate. | |
| 355 per 1,000 | 171 per 1000 | |||||
| Complication rate | Study population | Peto OR 1.23 | 318 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of long‐term GnRH agonist therapy on complication rate. | |
| 31 per 1,000 | 38 per 1000 | |||||
| Clinical pregnancy rate | Study population | RR 1.13 | 552 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of long‐term GnRH agonist therapy on clinical pregnancy rate. | |
| 331 per 1,000 | 374 per 1000 | |||||
| Multiple pregnancy rate | Study population | Peto OR 0.14 | 208 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of long‐term GnRH agonist therapy on multiple pregnancy rate. | |
| 74 per 1,000 | 10 per 1000 | |||||
| Miscarriage rate | Study population | Peto OR 0.45 | 208 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of long‐term GnRH agonist therapy on miscarriage rate. | |
| 46 per 1,000 | 21 per 1000 | |||||
| Mean number of oocytes | The mean number of oocytes was 8.43 | MD 0.72 higher | ‐ | 385 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of long‐term GnRH agonist therapy on mean number of oocytes. |
| Mean number of embryos | The mean number of embryos was 4.74 | MD 0.76 lower | ‐ | 267 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of long‐term GnRH agonist therapy on mean number of embryos. |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| a data arising exclusively from an unpublished trial; downgraded one level. b small number of events; downgraded two levels. c incorporates at least one open‐label study; downgraded one level. d high degree of heterogeneity; downgraded one level. e wide confidence intervals; downgraded one level. f data arising from an unpublished trial and a conference abstract; downgraded one level. g very high degree of heterogeneity; downgraded two levels. h significant contribution by an unpublished trial; downgraded one level. | ||||||
Background
Description of the condition
Endometriosis is a challenging disease observed in 16% to 50% of infertile women (Hart 2003; Meuleman 2009; Prescott 2016; Zondervan 2018). A cause and effect relationship between endometriosis and infertility has not been established, but it has been postulated that a combination of distorted pelvic anatomy, altered peritoneal function, alteration of immunological milieu within the peritoneal cavity, diminished ovarian reserve and altered endometrial receptivity could be the cause of infertility (ASRM 2006; Giudice 2010). Women with endometriosis often require in vitro fertilisation (IVF) to improve the chance of pregnancy, and more than one‐third of women undergoing IVF have endometriosis (Ozkan 2008). The benefits of in vitro fertilisation‐embryo transfer (IVF‐ET) in endometriosis‐associated infertility are well established (Barnhart 2002; Olivennes 2003; Opoein 2012). A meta‐analysis of 22 non‐randomised studies on the effect of endometriosis on IVF outcomes concluded that endometriosis interferes with all aspects of the reproductive process and is associated with lower success rates than other indications for IVF (Barnhart 2002). The overall chance of achieving pregnancy with IVF in these 22 studies was 25%. Nevertheless, a fairly recent large retrospective cohort study of 2245 women reported comparable pregnancy rates and live birth rates of IVF/intracytoplasmic sperm injection (ICSI) treatment in women with endometriosis‐related infertility as in women with tubal factors, except for those with ovarian endometrioma (Opoein 2012).
The severity of endometriosis has been shown to have a direct impact on the outcomes of IVF/ICSI treatment (Azem 1999; Coccia 2011; Olivennes 2003). A recent meta‐analysis of 27 non‐randomised studies reported a slight reduction in fertilisation rates among women with ASRM (American Society of Reproductive Medicine) stage I/II endometriosis undergoing IVF/ICSI treatment (risk ratio (RR) 0.93, 95% confidence interval (CI) 0.87 to 0.99, P = 0.03; (Harb 2013). In women with severe pelvic endometriosis and stage III/IV disease, implantation rates were significantly reduced (RR 0.79, 95% CI 0.67 to 0.93, P = 0.006), as were clinical pregnancy rates (RR 0.76, 95% CI 0.69 to 0.91, P = 0.0008). In stage I/II disease, a hostile pelvic environment due to alteration of immunological milieu within the peritoneal cavity may be responsible for impaired gamete and early embryo development (Ryan 1997; Surrey 1998). On the other hand, in stage III/IV endometriosis, lower oocyte yield and poor oocyte quality (Norenstedt 2001; Pal 1998), resulting in embryos of lesser quality have been reported to be responsible for lower pregnancy and implantation rates post‐IVF/ICSI (Arici 1996; Barnhart 2002).
Description of the intervention
The surgical management of asymptomatic ovarian endometrioma in women before IVF is debatable. A systematic review and meta‐analysis of 21 cohort studies (Raffi 2012), and recent non‐randomised studies (Tang 2013; Urman 2013), reported a negative impact of excision of endometriomas on ovarian reserve. Another systematic review of 20 non‐randomised studies (Tsoumpou 2009), concluded that excision of endometriomas has no significant effect on IVF pregnancy rates and ovarian response to stimulation compared with no treatment, nor does it improve fertility outcomes (Garcia‐Velasco 2004; Geber 2002; Suzuki 2005). On the other hand, non‐randomised studies have reported that resection of mild endometriosis and/or restoration of the pelvic anatomy (Harkki 2010; Marconi 2002), especially if done laparoscopically, may enhance the efficacy of assisted reproductive techniques, but that aggressive ovarian surgery should be avoided. Because of the negative effect of surgical management of endometriosis on the ovarian reserve, pituitary suppression therapy with gonadotrophin‐releasing hormone (GnRH) agonists before IVF has become a popular alternative chosen by many clinicians.
Long‐term therapy with GnRH agonists, synthetic peptide analogues of GnRH, has been shown to be effective in treating symptomatic endometriosis, including a reduction in the size of some endometriomas (Surrey 2010a). However, whether this improves fecundity is questionable. Several non‐randomised studies (Ma 2008; Marcus 1994; Nakamura 1992; Van der Hauven 2013), suggest that in women with endometriosis‐related infertility, long‐term treatment with GnRH agonists for at least three months (and up to six months) before IVF cycles are initiated may improve implantation and clinical pregnancy rates and reduce preclinical miscarriages, compared with a conventional IVF protocol. GnRH agonists (leuprorelin, goserelin, buserelin, gonadorelin, triptorelin, nafarelin), depending on their formulation, can be given intramuscularly, subcutaneously or intranasally. Upon completion of GnRH treatment, approximately two to four weeks from the last administration, an ovarian stimulation protocol with gonadotrophins is initiated according to standard IVF/ICSI procedures.
How the intervention might work
Continuous hypophyseal exposure to GnRH agonists leads to down‐regulation of GnRH receptors, which desensitises the pituitary gland. The hypogonadotrophic‐hypogonadal state results in prolonged amenorrhoea and a low estradiol (E2) level, thus depriving existing endometriotic lesions of their main growth stimulus. This may improve the symptoms of endometriosis, and it is plausible that it may also reverse the negative effects of endometriosis on IVF cycles, including poor folliculogenesis resulting in oocytes of reduced quality; a hostile peritoneal environment from macrophages, cytokines or vasoactive substances in the peritoneal fluid; mechanical interference with oocyte pickup and transportation; and anatomical dysfunction of the fallopian tubes and ovaries (Cahill 1996; Harb 2013; Harlow 1996). It has been suggested by several non‐randomised studies that poor oocyte quality results in decreased fertilisation rates (Bergendal 1998; Garrido 2000; Pal 1998). This in turn leads to embryos of lesser quality, along with a reduced implantation rate, especially in severe endometriosis (Arici 1996; Simon 1994). A higher rate of implantation and multiple pregnancy after ovarian stimulation has been reported in women in a hypogonadotrophic‐hypogonadal state (Edwards 1995); this might be due to improved endometrial responsiveness following amenorrhoea. However, other studies have reported contradictory findings (Haouzi 2010; Hickman 2002).
Why it is important to do this review
The previous version of this review suggested that long‐term GnRH agonist therapy before IVF/ICSI in women with endometriosis is associated with higher live birth and clinical pregnancy rates (Sallam 2006). However, the live birth data were extrapolated from a single study reporting on pregnancies that "reached viability" (Dicker 1992). Furthermore, the clinical pregnancy rate was calculated from meta‐analysis of three studies, of which two reported no significant difference (Rickes 2002; Surrey 2002), but with the data reaching statistical significance due to the size of the effect reported by Dicker 1992.
Based on the findings of Sallam 2006, IVF/ICSI with long‐term pituitary down‐regulation has been suggested as the first choice treatment for women with endometriosis‐associated infertility (ESHRE 2013). However, many clinicians still have doubts about its effectiveness. Concerns include the possibility that it may lower ovarian response to ovarian stimulation, especially among poor responders. The beneficial effects of prolonged GnRH treatment before IVF are seen most often for severe endometriosis, and the benefits and harms of the intervention in mild to moderate disease have not been fully evaluated. Other concerns focus on the unpleasant side effects of GnRH agonist therapy, such as vasomotor and psychological irritability, which may seriously affect women's quality of life. Complications associated with treatment such as infection, pelvic abscess and ovarian hyperstimulation syndrome (OHSS) have not been fully assessed. Alternatively, continuous use of combined oral contraceptives (COCs) (which have the fewest side effects) for six to eight weeks before IVF or ICSI in women with endometriosis has shown favourable fertility outcomes (de Ziegler 2010), although no comparison with long‐term GnRH treatment has been made.
Objectives
To determine the effectiveness and safety of long‐term gonadotrophin‐releasing hormone (GnRH) agonist therapy (minimum 3 months) versus no pretreatment or other pretreatment modalities, such as long‐term continuous combined oral contraception (COC) or surgical therapy of endometrioma, before standard vitro fertilisation (IVF) or intracytoplasmic sperm injection (ICSI) in women with endometriosis.
Methods
Criteria for considering studies for this review
Types of studies
We included all randomised controlled trials (RCTs) that reported use of any type of GnRH agonist for at least three months before an IVF/ICSI protocol was started in women diagnosed with endometriosis. We included cross‐over trials only if we could extract phase one data. We excluded quasi‐randomised trials.
Types of participants
Infertile women diagnosed with endometriosis (regardless of the grade of the disease) and treated with IVF or ICSI. The diagnosis of endometriosis was exclusively based on laparoscopy or laparotomy. For severity of endometriosis, grades according to the American Society of Reproductive Medicine (ASRM) score at laparotomy or laparoscopy were used.
Types of interventions
Intervention
Any type of GnRH agonist preparation used to down‐regulate the hypothalamic‐pituitary complex for three months or longer before a standard IVF or ICSI protocol in women diagnosed with endometriosis.
Comparison
-
Standard IVF/ICSI protocol only
-
Continuous combined oral contraception (COC) therapy for a minimum of six weeks before a standard IVF/ICSI protocol
-
Surgical excision of endometrioma within six months before the start of a standard IVF/ICSI protocol
Studies with a cointervention (e.g. surgical treatment of endometriosis in both groups prior to the study intervention) were eligible for inclusion.
Types of outcome measures
We recorded the following outcomes if the information was available.
Primary outcomes
-
Live birth rate: delivery of a live foetus after 22 completed weeks of gestational age
-
Complication rate: adverse effects per woman (e.g. infection, ovarian hyperstimulation syndrome (OHSS), serious vasomotor instability, overall adverse effects)
Secondary outcomes
-
Clinical pregnancy rate per couple/woman: defined as pregnancy confirmed by visualisation of a foetal sac on ultrasound (Zegers‐Hochschild 2017)
-
Multiple pregnancy rate: number of twin, triplet or higher‐order pregnancies (specified if possible) per pregnancy and confirmed by ultrasound or delivery
-
Miscarriage rate: miscarriage defined as the spontaneous loss of an intrauterine pregnancy prior to 22 completed weeks of gestational age (Zegers‐Hochschild 2017)
-
Ectopic pregnancy rate: ectopic pregnancy defined as a pregnancy outside the uterine cavity, diagnosed by ultrasound, surgical visualisation or histopathology, per couple/woman (Zegers‐Hochschild 2017)
-
Foetal abnormalities: number of women giving birth to a child with a foetal abnormality
-
Mean number of oocytes retrieved per woman
-
Mean number of embryos obtained per woman
Search methods for identification of studies
We searched for all published and unpublished randomised controlled trials (RCTs) of women diagnosed with endometriosis and treated with GnRH agonists to down‐regulate the hypothalamic‐pituitary complex before IVF or ICSI, without language restriction and in consultation with the Cochrane Gynaecology and Fertility (CGF) Information Specialist.
Electronic searches
We searched the following electronic databases from their inception to 8 January 2019.
-
The CGF Specialised Register of Controlled Trials; searched 8 January 2019 (PROCITE platform) (Appendix 1).
-
CENTRAL via the Cochrane CENTRAL Register of Studies ONLINE (CRSO); searched 8 January 2019 (Web platform) (Appendix 2).
-
MEDLINE; searched from 1946 to 8 January 2019 (OVID platform) (Appendix 3).
-
Embase; searched from 1980 to 8 January 2019 (OVID platform) (Appendix 4).
-
PsycINFO; searched from 1806 to 8 January 2019 (OVID platform) (Appendix 5).
-
Cumulative Index to Nursing and Allied Health Literature (CINAHL); searched from 1961 to 8 January 2019 (EBSCO platform) (Appendix 6).
We combined the MEDLINE search with the Cochrane highly sensitive search strategy for identifying randomised trials, which appears in the Cochrane Handbook for Systematic Reviews of Interventions (Lefebvre 2008). EMBASE, PsycINFO and CINAHL searches were combined with trial filters developed by the Scottish Intercollegiate Guidelines Network (www.sign.ac.uk/methodology/filters).
Other electronic sources of trials included the following.
-
Trial registers for ongoing and registered trials.
-
ClinicalTrials.gov; a service of the US National Institutes of Health (www.clinicaltrials.gov).
-
World Health Organization International Trials Registry Platform search portal (www.who.int/trialsearch).
-
-
DARE (Database of Abstracts of Reviews of Effects) in the Cochrane Library at onlinelibrary.wiley.com (for reference lists from relevant non‐Cochrane reviews).
-
Web of Knowledge; another source of trials and conference abstracts (wokinfo.com).
-
OpenGrey; for unpublished literature from Europe (www.opengrey.eu).
-
LILACS (Latin American and Caribbean Health Science Information Database; for trials from the Portuguese and Spanish speaking world; regional.bvsalud.org).
-
PubMed and Google (for recent trials not yet indexed in MEDLINE).
Searching other resources
We handsearched reference lists of articles retrieved by the search and contacted experts in the field to obtain additional data. We also handsearched relevant journals and conference abstracts that are not covered in the CGF register, in liaison with the Information Specialist.
Data collection and analysis
Selection of studies
Two review authors (EG and PM) carried out an initial screen of titles and abstracts retrieved by the search and identified potentially eligible studies. The full texts of these studies were retrieved. Two review authors (EG and PM or PB) independently examined these full‐text articles for compliance with the inclusion criteria and selected studies eligible for inclusion in the review. We corresponded with study investigators as required to clarify study eligibility. Disagreements as to study eligibility were resolved by discussion or by discussion with a third review author (IG). The selection process was documented with a PRISMA flow chart (Moher 2009).
Data extraction and management
Two review authors (EG and PM or PB) independently extracted data from eligible studies. Any discrepancies were resolved by discussion or by discussion with a third review author (IG). If a study appeared to meet the eligibility criteria but had aspects of methodology that were unclear, or if the data were provided in a form that was unsuitable for meta‐analysis, the study authors were contacted, and additional information on trial methodology or actual original trial data, or both, was sought. When studies had multiple publications, the review authors collated all reports of the same study, so that each study rather than each report was the unit of interest in the review, and such studies were given a single study ID.
Assessment of risk of bias in included studies
Two review authors (EG and PM or PB) independently assessed the included studies for risk of bias using the Cochrane 'Risk of bias' assessment tool (Higgins 2011) to assess allocation (random sequence generation and allocation concealment); blinding of participants and personnel; blinding of outcome assessors; incomplete outcome data; selective reporting and other bias. Disagreements were resolved by discussion or by discussion with a third review author (IG). We fully described all judgements and presented the conclusions in the 'Risk of bias' table, which was incorporated into the interpretation of review findings by means of sensitivity analyses.
We took care to search for within‐trial selective reporting, such as trials failing to report obvious outcomes or reporting them in insufficient detail to allow inclusion. We sought published protocols and compared outcomes between the protocol and the final published study.
Measures of treatment effect
For dichotomous data (e.g. live birth rate) we used the numbers of events in the control and intervention groups of each study to calculate risk ratios (RRs). For continuous data we calculated mean differences (MDs) between treatment groups. We present 95% confidence intervals (CIs) for all outcomes. If appropriate, we used number needed to treat for an additional beneficial outcome (NNTB), number needed to treat for an additional harmful outcome (NNTH), or both, to describe significant findings.
Unit of analysis issues
The primary analysis was per woman randomised. A sensitivity analysis using per pregnancy data was also included for relevant outcomes (multiple pregnancy, miscarriage, ectopic pregnancy, foetal abnormalities). We planned to briefly summarise in an additional table data that did not allow valid analysis (e.g. 'per cycle' or 'per embryo' data), and these were not meta‐analysed.
If studies reported only 'per cycle' data, we contacted study authors to request 'per woman randomised' data.
We counted multiple live births (e.g. twins, triplets) as one live birth event.
Dealing with missing data
We analysed the data on an intention‐to‐treat basis as far as possible and made attempts to obtain missing data from the original trialists. When data on live birth or clinical pregnancy could not be obtained, we assumed that the outcome did not occur. For other outcomes, we analysed only available data.
Assessment of heterogeneity
We considered whether clinical and methodological characteristics of the included studies were sufficiently similar for meta‐analysis to provide a clinically meaningful summary. We assessed statistical heterogeneity by measuring the I2 statistic. We considered an I2 measurement greater than 50% to indicate substantial heterogeneity (Higgins 2011). Where substantial heterogeneity was detected, we explored possible explanations in sensitivity analyses (see below) and considered a subgroup analysis. We took any statistical heterogeneity into account when interpreting the results, especially if variation in the direction of effect was noted.
Assessment of reporting biases
The review authors aimed to minimise the potential impact of publication bias and other reporting biases by ensuring a comprehensive search for eligible studies. If 10 or more studies were included in an analysis, we planned to use a funnel plot to explore the possibility of small‐study effects (a tendency for estimates of the intervention effect to be more beneficial in smaller studies; Higgins 2011).
Data synthesis
Where studies were sufficiently similar, we combined the data using a fixed‐effect model in the following comparisons.
-
Long‐term GnRH agonist pretreatment followed by standard IVF/ICSI versus standard IVF/ICSI only.
-
Long‐term GnRH agonist pretreatment versus at least six weeks of COC pretreatment, both followed by standard IVF/ICSI.
-
Long‐term GnRH agonist treatment versus surgical excision of endometriomas, both followed by standard IVF/ICSI.
Analyses were stratified by months of GnRH agonist pretreatment: three months, six months, over six months.
We displayed an increase in the risk of a particular outcome which may be beneficial (e.g. live birth) or detrimental (e.g. adverse effects) graphically in the meta‐analyses to the right of the centre line, and a decrease in the odds of an outcome to the left of the centre line.
Subgroup analysis and investigation of heterogeneity
Where data were available, we conducted subgroup analyses to determine separate evidence within the following subgroups.
-
Severity of endometriosis; stage I/II and stage III/IV disease.
-
Previous history of surgery to treat endometriosis.
-
Type of embryo transfer: fresh or cryopreserved.
Sensitivity analysis
We conducted sensitivity analyses for the primary outcome measures to determine whether the conclusions were robust to arbitrary decisions made regarding eligibility and analysis. These analyses included consideration of whether the review conclusions would have differed if:
-
eligibility had been restricted to studies at low risk of bias (i.e. no high risk of selection bias);
-
the effect measure had been expressed as odds ratio (OR) rather than RR;
-
the unit of analysis had been per pregnancy rather than per woman, for relevant outcomes (multiple pregnancy, miscarriage, ectopic pregnancy, foetal abnormalities).
Where substantial heterogeneity was detected, we explored clinical or methodological differences between studies that might have accounted for the heterogeneity.
Overall quality of the body of evidence: 'Summary of findings' table
We generated a 'Summary of findings' table using GRADEpro software (GRADEpro GDT 2015). This table was prepared by two review authors (EG and PM) working independently, with disagreements resolved by consensus. It evaluates the overall quality of the body of evidence for the main review comparison (GnRH agonist pretreatment versus no pretreatment) and reports the main review outcomes (live birth rate, complication rate, clinical pregnancy rate, multiple pregnancy rate, miscarriage rate, mean number of oocytes, and mean number of embryos), using GRADE criteria (study limitations (i.e. risk of bias), consistency of effect, imprecision, indirectness and publication bias). We have justified, documented and incorporated into the report of results for each outcome our judgements about evidence quality (high, moderate, low or very low).
Results
Description of studies
Results of the search
We ran the updated electronic search on 8 January 2019, and this yielded 817 articles. We identified 59 additional articles via other sources: 23 via ClinicalTrials.gov (www.clinicaltrials.gov) and 36 via the World Health Organization International Trials Registry Platform search portal (www.who.int/trialsearch). After removal of duplicates, 809 were left for screening. Of these, we excluded 770 as they were clearly not relevant. We reviewed 39 full‐text articles of which we excluded 27 (Characteristics of excluded studies). One study is awaiting classification as it was not clear how endometriomas were diagnosed (Characteristics of studies awaiting classification), and two studies were ongoing trials that had not yet reported their results (Characteristics of ongoing studies).
We identified eight studies (nine articles) that met the inclusion criteria for this review (Decleer 2016; Dicker 1992; NCT01269125; NCT01581359; Rickes 2002; Surrey 2002; Surrey 2010; Totaro 2009). Of these, we included two in our qualitative analysis (Surrey 2010; Totaro 2009), and six in our quantitative analysis (Decleer 2016; Dicker 1992; NCT01269125; NCT01581359; Rickes 2002; Surrey 2002). We have presented the trial flow diagram in Figure 1. The previous Sallam 2006 version of this review incorporated Dicker 1992, Rickes 2002 and Surrey 2002.
Study flow diagram.
Included studies
Study design and setting
We included eight parallel‐design randomised controlled trials (RCTs), five of which have been published as full articles (Decleer 2016; Dicker 1992; Rickes 2002; Surrey 2002; Surrey 2010), one was published as a conference abstract (Totaro 2009), and two are unpublished studies (NCT01269125; NCT01581359). Five studies were single‐centre (Decleer 2016; NCT01269125; NCT01581359; Rickes 2002; Surrey 2010), one study was multiple‐centre (Surrey 2002), and for two studies it was unclear (Dicker 1992; Totaro 2009). Two of the studies were carried out in the USA (Surrey 2002; Surrey 2010), and one each in Belgium (Decleer 2016), Germany (Rickes 2002), Greece (NCT01269125), Israel (Dicker 1992), Italy (Totaro 2009), and Spain (NCT01581359).
Participants
The eight included studies had a total of 640 participants: 322 women in the intervention group and 318 in the control group. All included patients had a diagnosis of endometriosis at laparoscopy or laparotomy. Decleer 2016 and NCT01269125 included patients with stages I or II endometriosis, Rickes 2002 included patients with stages II‐IV endometriosis and Totaro 2009 included patients with stages III or IV endometriosis. NCT01581359 and Surrey 2002 included patients with any severity of endometriosis. Dicker 1992 described endometriosis as severe and Surrey 2010 included patients with laparoscopically diagnosed endometriosis but "no current sonographic evidence of an ovarian endometrioma > 2 cm in mean diameter".
The mean age of participants ranged from 30.3 to 33.9 years, where reported (Decleer 2016; Dicker 1992; NCT01269125; NCT01581359; Surrey 2002; Surrey 2010). Rickes 2002 reported that the age range in their study was 23 to 40 years old and Totaro 2009 did not comment on age. In five studies, patients underwent in vitro fertilisation (IVF) (Decleer 2016; Dicker 1992; NCT01269125; NCT01581359; Surrey 2010), in two studies they had IVF or intracytoplasmic sperm injection (ICSI) (Rickes 2002; Surrey 2002), and in one study it was not clear which type of assisted reproductive technique was employed (Totaro 2009).
Interventions
Surgical treatment of endometriosis was performed prior to randomisation in five studies (Decleer 2016; Dicker 1992; NCT01269125; NCT01581359; Rickes 2002), whereas in three studies it was not clear whether any treatment was performed (Surrey 2002; Surrey 2010; Totaro 2009).
Three studies used leuprolide (NCT01269125; Surrey 2002; Surrey 2010), two studies used goserelin (Decleer 2016; Rickes 2002), two studies used triptorelin (NCT01581359; Totaro 2009), and one used decapeptyl (Dicker 1992). Duration of treatment was for three months in five studies (Decleer 2016; NCT01269125; NCT01581359; Surrey 2002; Surrey 2010), six months in two studies (Dicker 1992; Totaro 2009), and five to six months in one study (Rickes 2002).
For all included studies, the control group consisted of standard IVF/ICSI (Decleer 2016; Dicker 1992; NCT01269125; NCT01581359; Rickes 2002; Surrey 2002; Surrey 2010; Totaro 2009). None of the studies had control groups including the control group of continuous combined oral contraception (COC) therapy for a minimum of six weeks before standard IVF/ICSI or surgical excision of endometrioma within six months before the start of standard IVF/ICSI.
Outcomes
Primary outcomes
One study reported on the primary outcome of live birth (NCT01581359). Three studies reported on the primary outcome of complication rate (NCT01269125; NCT01581359; Surrey 2002); specifically ovarian hyperstimulation syndrome (OHSS) (NCT01269125; NCT01581359), local reaction at injection site (NCT01581359), and premature luteinisation (NCT01269125).
Secondary outcomes
Eight studies reported on clinical pregnancy rate (Decleer 2016; Dicker 1992; NCT01269125; NCT01581359; Rickes 2002; Surrey 2002; Surrey 2010; Totaro 2009), and of these, six could be included for meta‐analysis (Decleer 2016; Dicker 1992; NCT01269125; NCT01581359; Rickes 2002; Surrey 2002). The reasons for exclusion of Surrey 2010 and Totaro 2009 are detailed in Characteristics of included studies.
Two studies reported on multiple pregnancy rate (NCT01581359; Totaro 2009), two studies reported on miscarriage rate (NCT01581359; Totaro 2009), two studies reported on ectopic pregnancy rate (NCT01581359; Surrey 2002), and two studies reported on foetal abnormality rate (NCT01581359; Surrey 2002).
Six studies reported on the mean number of oocytes (Decleer 2016; Dicker 1992; NCT01581359; Surrey 2002; Surrey 2010; Totaro 2009), of which four could be included for meta‐analysis as per woman data (Decleer 2016; Dicker 1992; NCT01581359; Surrey 2002). The reasons for exclusion of Surrey 2010 and Totaro 2009 are detailed in Characteristics of included studies.
Two studies reported on the mean number of embryos obtained per woman (Decleer 2016; NCT01581359).
Author correspondence
We contacted the authors of Decleer 2016, NCT01269125, NCT01581359, Rickes 2002, Surrey 2010 and Totaro 2009 to obtain and clarify data. To date, we have received a response from NCT01581359 and Surrey 2010.
Excluded studies
We excluded 27 studies for the following reasons.
-
14 wrong study design
-
1 duplicate study
-
6 no IVF/ICSI in either group
-
2 sub‐studies of already included patient population
-
2 wrong comparator
-
1 wrong intervention
-
1 wrong patient population
Risk of bias in included studies
We assessed risk of bias in all included studies as demonstrated in Figure 2 and Figure 3. Detailed information can be found in Characteristics of included studies.
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Allocation
Random sequence generation
Seven studies utilised adequate methods for random sequence generation, such as computer‐generated random number tables, and hence we deemed them to be at low risk of bias (Decleer 2016; NCT01269125; NCT01581359; Rickes 2002; Surrey 2002; Surrey 2010; Totaro 2009). Dicker 1992 states that patients were "randomly and sequentially allocated" and attempts to contact the study authors at the time of the previous version of this review (Sallam 2006), were unsuccessful.
Allocation concealment
One study reported an adequate method of allocation concealment using "opaque‐sealed envelopes, sequentially numbered' and hence we deemed it to be at low risk of bias (NCT01581359). Four studies provided no details of allocation concealment and hence we deemed them to be at unclear risk of bias (Decleer 2016; Rickes 2002; Surrey 2002; Totaro 2009). We also judged Dicker 1992 to be at unclear risk of bias, based on patients being "randomly and sequentially allocated", as was NCT01269125, in which the random allocation was made by the first author. In correspondence with Surrey 2010, they confirmed that no allocation concealment was undertaken and hence we labelled the study as high risk of bias.
Blinding
Blinding of participants and personnel (performance bias)
One study confirmed blinding of participants and personnel and hence we judged it to be at low risk of bias (NCT01581359). Four studies reported being open label (Decleer 2016; NCT01269125; Rickes 2002; Surrey 2010), and the previous version of this review (Sallam 2006), mentions no placebo injections in another two studies (Dicker 1992; Surrey 2002); we judged all of these to be at high risk of bias. One study did not report on blinding and we labelled it as unclear risk of bias (Totaro 2009).
Blinding of outcome assessment (detection bias)
We judged NCT01581359 to be at low risk of bias, as blinding was maintained during the study until follow‐up was completed. We judged most other studies to be at high risk as they were open label (Decleer 2016; NCT01269125; Rickes 2002; Surrey 2010), or previously reported no blinding (Dicker 1992 and Surrey 2002 in Sallam 2006). We labelled Totaro 2009 as unclear risk of bias as no information was supplied.
Incomplete outcome data
An intention‐to‐treat analysis was performed in two studies (Rickes 2002; Surrey 2002), and hence we judged them to be at low risk of bias. For Surrey 2002, a lack of study dropouts was confirmed in author correspondence at the time of the previous version of this review (Sallam 2006). We marked four studies at high risk of bias in light of a lack of an intention‐to‐treat analysis (Decleer 2016; NCT01269125; NCT01581359; Surrey 2010). Where data were available, we used the number of participants randomised for this meta‐analysis (Decleer 2016; NCT01269125; NCT01581359). Where the data for meta‐analysis were not available, we included the study in the qualitative synthesis (Surrey 2010). We judged one study to be at unclear risk of bias as no information was supplied (Totaro 2009).
Selective reporting
We deemed all studies to be at low risk of bias as they all reported on a priori outcomes (Decleer 2016; Dicker 1992; NCT01269125; NCT01581359; Rickes 2002; Surrey 2002; Surrey 2010; Totaro 2009). We did, however, note that in one study, there was a change in the secondary outcomes (NCT01269125), and in another study there was a change in the primary outcome (NCT01581359).
Other potential sources of bias
We deemed one study to be at low risk of bias on the basis of no difference in basal patient characteristics and no conflict of interest or relevant funding sources (Decleer 2016). We judged two studies to be at unclear risk of bias on the basis of a lack of information on conflict of interest or funding sources (Dicker 1992; Totaro 2009). We labelled five studies as high risk of bias; three due to the fact that they all received funding from pharmaceutical companies involved in making GnRH analogues (Rickes 2002; Surrey 2002; Surrey 2010); and two due to the fact that they are presently unpublished plus lack information on conflict of interest and funding sources (NCT01269125; NCT01581359). In addition, in one study there was a higher proportion of stage III or IV endometriosis in the intervention arm than the control arm (Surrey 2002).
Effects of interventions
See: summary of findings Table for the main comparison.
1. Comparison of gonadotrophin‐releasing hormone (GnRH) agonist to standard in vitro fertilisation/intracytoplasmic sperm injection (IVF/ICSI) protocol only
Primary outcomes
1.1 Live birth rate
We are uncertain whether long‐term GnRH agonist therapy affects the live birth rate compared to standard IVF/ICSI (risk ratio (RR) 0.48, 95% confidence interval (CI) 0.26 to 0.87; 1 RCT, n = 147; I2 not calculable; very low‐quality evidence; Analysis 1.1; Figure 4). This suggests that with a live birth rate of approximately 36% (355 per 1000) with standard IVF/ICSI, the equivalent live birth rate with long‐term GnRH agonist therapy prior to standard IVF/ICSI lies between 9% and 31% (92 to 309 per 1000).
Forest plot of comparison: 1 GnRH agonist therapy versus standard IVF/ICSI alone, outcome: 1.1 Live birth rate.
We did not perform sensitivity analysis as only one unpublished trial reported on the primary outcome of live birth.
1.1.1 Subgroup analysis: severity of endometriosis
We are uncertain whether long‐term GnRH agonist therapy affects the live birth rate compared to standard IVF/ICSI in patients with either stage I/II endometriosis (Peto OR 0.16, 95% CI 0.03 to 0.83; 1 RCT, n = 33; I2 not calculable; very low‐quality evidence; Analysis 1.15) or stage III/IV endometriosis . (Peto OR 0.49, 95% CI 0.22 to 1.10; 1 RCT, n = 114; I2 not calculable; low‐quality evidence; Analysis 1.16).
Subgroup analysis: previous history of surgery to treat endometriosis
None of the included studies reported on this comparison.
Subgroup analysis: type of embryo transfer
None of the included studies reported on this comparison.
1.2 Complication rate
We are uncertain whether long‐term GnRH agonist therapy affects the overall complication rate (Peto OR 1.23, 95% CI 0.37 to 4.14; 3 RCTs, n = 318; I2 = 73%; very low‐quality evidence; Analysis 1.10). This suggests that with a complication rate of approximately 3% (31 per 1000) with standard IVF/ICSI, the equivalent complication rate with long‐term GnRH agonist therapy prior to IVF/ICSI lies between 1% and 13% (11 to 128 per 1000).
We did not perform sensitivity analysis as data for this outcome originated exclusively from unpublished trials. Looking at specific complications following long‐term GnRH agonist therapy, we are uncertain whether this intervention affects the incidence of ovarian hyperstimulation syndrome (OHSS) (Peto OR 0.75, 95% CI 0.17 to 3.40; 2 RCTs, n = 267; I2 = 39%; very low‐quality evidence), local reaction at the injection site (Peto OR 8.04, 95% CI 0.50 to 130.02; 1 RCT, n = 147; I2 not calculable; very low‐quality evidence) or premature luteinisation (Peto OR 1.00, 95% CI 0.06 to 16.18; 1 RCT, n = 120; I2 not calculable; very low‐quality evidence). See Analysis 1.2 and Figure 5.
Forest plot of comparison: 1 GnRH agonist therapy versus standard IVF/ICSI alone, outcome: 1.2 Complication rate.
1.2.1 Subgroup analysis: severity of endometriosis
We are uncertain whether long‐term GnRH agonist treatment affects the complication rate compared to standard IVF/ICSI in patients with stage I/II endometriosis (no events in either group, see Analysis 1.17) or stage III/IV endometriosis (Peto OR 0.14, 95% CI 0.01 to 1.42; 1 RCT, n = 114; I2 not calculable; very low‐quality evidence; Analysis 1.18).
1.2.2 Subgroup analysis: previous history of surgery to treat endometriosis
None of the included studies reported on this comparison.
1.2.3 Subgroup analysis: type of embryo transfer
None of the included studies reported on this comparison.
Secondary outcomes
1.3 Clinical pregnancy rate
We are uncertain of the effect of long‐term GnRH agonist therapy on clinical pregnancy rate (RR 1.13, 95% CI 0.91 to 1.41; six RCTs, n = 552; I2 = 66%; very low‐quality evidence; Analysis 1.3; Figure 6).
Forest plot of comparison: 1 GnRH agonist therapy vs Standard IVF/ICSI alone, outcome: 1.3 Clinical Pregnancy Rate.
Due to the high degree of heterogeneity, we repeated the analysis using a random‐effects model yielding similar results (RR 1.20, 95% CI 0.81 to 1.78; six RCTs, n = 552; I2 = 66%; very low‐quality evidence).
We could not include data from two studies in the meta‐analysis and we have summarised these in Table 1 and explained the reasons for this in Characteristics of included studies.
| Study ID | Clinical pregnancy rate | Mean number of oocytes |
| GnRH agonist therapy: 9* out of 18 Standard IVF/ICSI: 9* out of 19 * these numbers are approximate, based on calculations done using % and assumption on number of patients per group | GnRH agonist therapy (mean + SD): 10.3* + 7.19* Standard IVF/ICSI (mean + SD): 18* + 13.76* | |
| GnRH agonist therapy: 7 out of 29 Standard IVF/ICSI: 7 out of 32 | GnRH agonist therapy: 4.3 + 2.5^ Standard IVF/ICSI: 5.4 + 4.7^ ^: unclear if SD |
GnRH: gonadotrophin‐releasing hormone; IVF/ICSI: in vitro fertilisation/intracytoplasmic sperm injection; SD: standard deviation
1.4. Multiple pregnancy rate
We are uncertain whether long‐term GnRH agonist therapy affects the multiple pregnancy rate per woman randomised (Peto OR 0.14, 95% CI 0.03 to 0.56; 2 RCTs, n = 208; I2 = 0%; very low‐quality evidence; Analysis 1.4). Similar results are obtained if the multiple pregnancy rate is expressed per clinical pregnancy (Peto OR 0.17, 95% CI 0.03 to 1.05; 1 RCT, n = 147; I2 not calculable; very low‐quality evidence; Analysis 1.11).
1.5 Miscarriage rate
We are uncertain of the effect of long‐term GnRH agonist therapy on the miscarriage rate (Peto OR 0.45, 95% CI 0.10 to 2.00; 2 RCTs, n = 208; I2 = 0%; very low‐quality evidence; Analysis 1.5). Similar results are obtained if the miscarriage rate is expressed per clinical pregnancy (Peto OR 0.93, 95% CI 0.15 to 5.54; 1 RCTs, n = 57; I2 not calculable; very low‐quality evidence; Analysis 1.12).
1.6 Ectopic pregnancy rate
We are uncertain of the effect of long‐term GnRH agonist therapy on the ectopic pregnancy rate (Peto OR 0.14, 95% CI 0.00 to 7.30; 2 RCTs, n = 198; I2 not calculable; very low‐quality evidence; Analysis 1.6).
Similar results are obtained if the ectopic pregnancy rate is expressed per clinical pregnancy (Peto OR 0.22, 95% CI 0.00 to 13.15; 2 RCTs, n = 198; I2 not calculable; very low‐quality evidence; Analysis 1.13).
1.7 Foetal abnormality rate
We are uncertain of the effect of long‐term GnRH agonist therapy on the foetal abnormality rate (Peto OR 7.93, 95% CI 0.16 to 400.46; 2 RCTs, n = 198; I2 not calculable; very low‐quality evidence; Analysis 1.7).
Similar results are obtained if the foetal abnormality rate is expressed per clinical pregnancy (Peto OR 17.58, 95% CI 0.29 to 1073.88; 2 RCTs, n = 198; I2 not calculable; very low‐quality evidence; Analysis 1.14).
1.8 Mean number of oocytes
We are uncertain of the effect of long‐term GnRH agonist therapy on the mean number of oocytes (mean difference (MD) 0.72, 95% CI 0.06 to 1.38; 4 RCTs, n = 385; I2 = 81%; very low‐quality evidence; Analysis 1.8).
Due to the high degree of heterogeneity, we repeated the analysis using a random‐effects model yielding similar results (MD 0.42, 95% CI ‐1.38 to 2.22; 4 RCTs, n = 385; I2 = 81%; very low‐quality evidence).
We could not include data from two studies in the meta‐analysis and we have summarised these in Table 1 and explained the reasons for this in Characteristics of included studies.
1.9 Mean number of embryos
We are uncertain of the effect of long‐term GnRH agonist therapy on the mean number of embryos (MD ‐0.76, 95% CI ‐1.33 to ‐0.19; 2 RCTs, n = 267; I2 = 0%; very low‐quality evidence; Analysis 1.9).
Comparison of GnRH agonist to continuous combined oral contraception (COC) therapy for a minimum of six weeks before a standard IVF/ICSI protocol
No studies reported on this comparison.
Comparison of GnRH agonist to surgical excision of endometrioma within six months before the start of a standard IVF/ICSI protocol
No studies reported on this comparison.
Discussion
Summary of main results
This Cochrane Review aimed to determine the effectiveness and safety of long‐term gonadotrophin‐releasing hormone (GnRH) agonist therapy versus no pretreatment or other pretreatment modalities, specifically long‐term continuous combined oral contraception (COC) or surgical therapy, before standard in vitro fertilisation/intracytoplasmic sperm injection (IVF/ICSI) in women with endometriosis. Our searches revealed no studies comparing long‐term GnRH agonist therapy to other pretreatment modalities, thus this review has focused entirely on long‐term GnRH agonist therapy versus no pretreatment prior to standard IVF/ICSI.
Our findings suggest that there is uncertainty about the effect of long‐term GnRH agonist therapy on the live birth rate compared to standard IVF/ICSI. Subgroup analysis by endometriosis severity further highlighted the uncertainty of the effect of long‐term GnRH agonist therapy on the live birth rate. Subgroup analysis by previous history of surgery to treat endometriosis was not possible due to a lack of data. Subgroup analysis by type of embryo transfer was not possible for the same reason.
The quality of the evidence for complication rate was very low and hence it is difficult to draw any conclusion on safety. The data available suggest that the overall complication rate as well as the complication rate by subtype are not associated with any safety concerns with long‐term GnRH agonist therapy.
Subgroup analysis by endometriosis severity, similarly had no impact on complication rate. We did not conduct an analysis by months of GnRH agonist pretreatment due to the paucity of data.
Very low‐quality evidence leaves us very uncertain on the impact of long‐term GnRH agonist therapy on clinical pregnancy rate, multiple pregnancy rate, miscarriage rate, mean number of oocytes and mean number of embryos.
Our findings are summarised in summary of findings Table for the main comparison.
Overall completeness and applicability of evidence
Based on our findings, it is evident that there is a paucity of data for the primary outcomes of live birth and complications. Of the included studies, only one reported on the outcome of live birth rate, which of course will be the outcome of most significance to women with endometriosis undertaking fertility treatment (NCT01581359). Furthermore, this was a relatively small study which at the time of writing of this systematic review was not published and hence had not been through the peer review process.
Only three studies looked at complication rate and hence it is difficult to definitively draw conclusions on safety (NCT01269125; NCT01581359; Surrey 2002). We noted a high degree of heterogeneity, and this could be a reflection of the small number of events identified on a background of small studies.
The majority of studies focused on clinical pregnancy rate (Decleer 2016; Dicker 1992; NCT01269125; NCT01581359; Rickes 2002; Surrey 2002; Surrey 2010; Totaro 2009), although only some of these data could be used for meta‐analysis (Decleer 2016; Dicker 1992; NCT01269125; NCT01581359; Rickes 2002; Surrey 2002). Once again, there was no change in this outcome with long‐term GnRH agonist therapy. We noted a significant degree of heterogeneity, mainly due to the contribution of the older study by Dicker 1992 and the, as yet, not published study NCT01581359.
Similarly, we noted a significant degree of heterogeneity with the mean number of oocytes; once again, this was mainly due to the contribution of the older study by Dicker 1992. We speculate that this could be a reflection of older, outdated IVF protocols and techniques.
Overall, even though the findings of the review draw into question the use of long‐term GnRH agonists prior to IVF/ICSI in women with endometriosis, we feel that the overall quality of the evidence is too low to support a move away from the use of GnRH agonists. Further high‐quality research is required to draw definitive conclusions.
Quality of the evidence
In this study we identified and included data originating from eight randomised controlled trials (RCTs), two of which are not published and one which was published solely as an abstract. The total number of participants involved was 640. The risk of bias for individual studies is summarised in Figure 2 and Figure 3.
We rated the quality of the evidence based on the GRADE criteria. The quality of evidence was very low with issues arising due to lack of blinding, imprecision and inconsistency. See summary of findings Table for the main comparison.
Potential biases in the review process
We aimed to reduce the risk of publication bias by conducting systematic searches of multiple databases as well as trial registries to identify unpublished and ongoing studies. We contacted trial authors for (more) information where necessary, although unfortunately data were not always available or we did not receive a response in all cases. It is possible that our searches did not identify all unpublished studies. We performed subgroup analysis by endometriosis severity but data were not available for subgroup analysis by status of previous surgery for endometriosis. As prespecified, where possible, we performed sensitivity analyses on the primary outcomes. We were unable to conduct a funnel plot due to the small number of included studies.
Agreements and disagreements with other studies or reviews
Our review findings are in sharp contrast to the previous version of this review which reported an association between long‐term GnRH agonist therapy and greater live birth and clinical pregnancy rates (Sallam 2006). Compared to the previous version of this review, the outcome of live birth includes one new RCT in the meta‐analysis and excludes a previously included RCT (Dicker 1992), as this paper does not truly report on live birth as per the Zegers‐Hochschild 2017 definition. Further, for the outcome of clinical pregnancy rate, this review includes three new RCTs in the meta‐analysis. The net effect of the addition of these RCTs, which in terms of bias are no worse in design, is to bring the results closer to the line of no effect. Indeed, in NCT01581359, which was the only adequately blinded study to date (i.e. patients in the control group received placebo injections rather than commencing standard IVF/ICSI immediately), both the live birth and clinical pregnancy rates favoured standard IVF/ICSI rather than pretreatment with long‐term GnRH agonists. Given that visible endometriosis was surgically treated in NCT01581359 prior to randomisation, the argument arises for a potential deleterious effect of long‐term GnRH therapy prior to IVF/ICSI. However, due to the very low quality of this evidence, it is difficult to reach any definitive conclusions at present. The trend for miscarriage rate did not differ from the previous version of this review, whereas in this review we found no trend towards a higher oocyte yield with GnRH agonist therapy.
Data from retrospective studies suggest that the live birth and clinical pregnancy rates are improved with cryopreserved embryo transfer after GnRH agonist therapy (Mohamed 2011; Xie 2018). Due to a lack of data we were unable to examine this outcome in this systematic review.
To our knowledge there are no other reviews that address long‐term GnRH agonist therapy compared no or to other pretreatments prior to standard IVF/ICSI. It is worth noting however that at least one other RCT, which did not meet our inclusion criteria and used continuous COC as the control group, also reported no difference in outcomes compared to long‐term GnRH agonist therapy (NCT02400801).
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Forest plot of comparison: 1 GnRH agonist therapy versus standard IVF/ICSI alone, outcome: 1.1 Live birth rate.
Forest plot of comparison: 1 GnRH agonist therapy versus standard IVF/ICSI alone, outcome: 1.2 Complication rate.
Forest plot of comparison: 1 GnRH agonist therapy vs Standard IVF/ICSI alone, outcome: 1.3 Clinical Pregnancy Rate.
Comparison 1 GnRH agonist therapy versus standard IVF/ICSI alone, Outcome 1 Live birth rate.
Comparison 1 GnRH agonist therapy versus standard IVF/ICSI alone, Outcome 2 Complication rate.
Comparison 1 GnRH agonist therapy versus standard IVF/ICSI alone, Outcome 3 Clinical Pregnancy Rate.
Comparison 1 GnRH agonist therapy versus standard IVF/ICSI alone, Outcome 4 Multiple pregnancy rate (per woman randomised).
Comparison 1 GnRH agonist therapy versus standard IVF/ICSI alone, Outcome 5 Miscarriage rate (per woman randomised).
Comparison 1 GnRH agonist therapy versus standard IVF/ICSI alone, Outcome 6 Ectopic pregnancy rate (per woman randomised).
Comparison 1 GnRH agonist therapy versus standard IVF/ICSI alone, Outcome 7 Foetal abnormality rate (per woman randomised).
Comparison 1 GnRH agonist therapy versus standard IVF/ICSI alone, Outcome 8 Mean number of oocytes.
Comparison 1 GnRH agonist therapy versus standard IVF/ICSI alone, Outcome 9 Mean number of embryos.
Comparison 1 GnRH agonist therapy versus standard IVF/ICSI alone, Outcome 10 Complication rate (totals).
Comparison 1 GnRH agonist therapy versus standard IVF/ICSI alone, Outcome 11 Multiple pregnancy rate (per clinical pregnancy).
Comparison 1 GnRH agonist therapy versus standard IVF/ICSI alone, Outcome 12 Miscarriage rate (per clinical pregnancy).
Comparison 1 GnRH agonist therapy versus standard IVF/ICSI alone, Outcome 13 Ectopic pregnancy (per clinical pregnancy).
Comparison 1 GnRH agonist therapy versus standard IVF/ICSI alone, Outcome 14 Foetal abnormality rate (per clinical pregnancy).
Comparison 1 GnRH agonist therapy versus standard IVF/ICSI alone, Outcome 15 Live birth rate (ASRM I/II endometriosis).
Comparison 1 GnRH agonist therapy versus standard IVF/ICSI alone, Outcome 16 Live birth rate (ARSM III/IV endometriosis).
Comparison 1 GnRH agonist therapy versus standard IVF/ICSI alone, Outcome 17 Complication rate (ARSM I/II endometriosis.
Comparison 1 GnRH agonist therapy versus standard IVF/ICSI alone, Outcome 18 Complication rate (ARSM III/IV endometriosis).
Comparison 1 GnRH agonist therapy versus standard IVF/ICSI alone, Outcome 19 Clinical pregnancy rate (ARSM I/II endometriosis).
Comparison 1 GnRH agonist therapy versus standard IVF/ICSI alone, Outcome 20 Clinical pregnancy rate (ARSM III/IV endometriosis).
| Long‐term GnRH agonist therapy compared to no additional therapy in patients with endometriosis prior to standard IVF/ICSI | ||||||
| Patient or population: patients with endometriosis prior to standard IVF/ICSI | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Quality of the evidence | Comments | |
| Risk with no additional therapy | Risk with long‐term GnRH agonist therapy | |||||
| Live birth rate | Study population | RR 0.48 | 147 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of long‐term GnRH agonist therapy on live birth rate. | |
| 355 per 1,000 | 171 per 1000 | |||||
| Complication rate | Study population | Peto OR 1.23 | 318 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of long‐term GnRH agonist therapy on complication rate. | |
| 31 per 1,000 | 38 per 1000 | |||||
| Clinical pregnancy rate | Study population | RR 1.13 | 552 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of long‐term GnRH agonist therapy on clinical pregnancy rate. | |
| 331 per 1,000 | 374 per 1000 | |||||
| Multiple pregnancy rate | Study population | Peto OR 0.14 | 208 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of long‐term GnRH agonist therapy on multiple pregnancy rate. | |
| 74 per 1,000 | 10 per 1000 | |||||
| Miscarriage rate | Study population | Peto OR 0.45 | 208 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of long‐term GnRH agonist therapy on miscarriage rate. | |
| 46 per 1,000 | 21 per 1000 | |||||
| Mean number of oocytes | The mean number of oocytes was 8.43 | MD 0.72 higher | ‐ | 385 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of long‐term GnRH agonist therapy on mean number of oocytes. |
| Mean number of embryos | The mean number of embryos was 4.74 | MD 0.76 lower | ‐ | 267 | ⊕⊝⊝⊝ | The evidence is very uncertain about the effect of long‐term GnRH agonist therapy on mean number of embryos. |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| a data arising exclusively from an unpublished trial; downgraded one level. b small number of events; downgraded two levels. c incorporates at least one open‐label study; downgraded one level. d high degree of heterogeneity; downgraded one level. e wide confidence intervals; downgraded one level. f data arising from an unpublished trial and a conference abstract; downgraded one level. g very high degree of heterogeneity; downgraded two levels. h significant contribution by an unpublished trial; downgraded one level. | ||||||
| Study ID | Clinical pregnancy rate | Mean number of oocytes |
| GnRH agonist therapy: 9* out of 18 Standard IVF/ICSI: 9* out of 19 * these numbers are approximate, based on calculations done using % and assumption on number of patients per group | GnRH agonist therapy (mean + SD): 10.3* + 7.19* Standard IVF/ICSI (mean + SD): 18* + 13.76* | |
| GnRH agonist therapy: 7 out of 29 Standard IVF/ICSI: 7 out of 32 | GnRH agonist therapy: 4.3 + 2.5^ Standard IVF/ICSI: 5.4 + 4.7^ ^: unclear if SD | |
| GnRH: gonadotrophin‐releasing hormone; IVF/ICSI: in vitro fertilisation/intracytoplasmic sperm injection; SD: standard deviation | ||
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Live birth rate Show forest plot | 1 | 147 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.48 [0.26, 0.87] |
| 2 Complication rate Show forest plot | 2 | Peto Odds Ratio (Peto, Fixed, 95% CI) | Subtotals only | |
| 2.1 OHSS | 2 | 267 | Peto Odds Ratio (Peto, Fixed, 95% CI) | 0.75 [0.17, 3.40] |
| 2.2 Local reaction at the injection site | 1 | 147 | Peto Odds Ratio (Peto, Fixed, 95% CI) | 8.04 [0.50, 130.02] |
| 2.3 Premature luteinisation | 1 | 120 | Peto Odds Ratio (Peto, Fixed, 95% CI) | 1.0 [0.06, 16.18] |
| 3 Clinical Pregnancy Rate Show forest plot | 6 | 552 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.13 [0.91, 1.41] |
| 4 Multiple pregnancy rate (per woman randomised) Show forest plot | 2 | 208 | Peto Odds Ratio (Peto, Fixed, 95% CI) | 0.14 [0.03, 0.56] |
| 5 Miscarriage rate (per woman randomised) Show forest plot | 2 | 208 | Peto Odds Ratio (Peto, Fixed, 95% CI) | 0.45 [0.10, 2.00] |
| 6 Ectopic pregnancy rate (per woman randomised) Show forest plot | 2 | 198 | Peto Odds Ratio (Peto, Fixed, 95% CI) | 0.14 [0.00, 7.30] |
| 7 Foetal abnormality rate (per woman randomised) Show forest plot | 2 | 198 | Peto Odds Ratio (Peto, Fixed, 95% CI) | 7.93 [0.16, 400.46] |
| 8 Mean number of oocytes Show forest plot | 4 | 385 | Mean Difference (IV, Fixed, 95% CI) | 0.72 [0.06, 1.38] |
| 9 Mean number of embryos Show forest plot | 2 | 267 | Mean Difference (IV, Fixed, 95% CI) | ‐0.76 [‐1.33, ‐0.19] |
| 10 Complication rate (totals) Show forest plot | 3 | 318 | Peto Odds Ratio (Peto, Fixed, 95% CI) | 1.23 [0.37, 4.14] |
| 11 Multiple pregnancy rate (per clinical pregnancy) Show forest plot | 1 | 43 | Peto Odds Ratio (Peto, Fixed, 95% CI) | 0.17 [0.03, 1.05] |
| 12 Miscarriage rate (per clinical pregnancy) Show forest plot | 1 | 43 | Peto Odds Ratio (Peto, Fixed, 95% CI) | 0.93 [0.15, 5.54] |
| 13 Ectopic pregnancy (per clinical pregnancy) Show forest plot | 2 | 77 | Peto Odds Ratio (Peto, Fixed, 95% CI) | 0.22 [0.00, 13.15] |
| 14 Foetal abnormality rate (per clinical pregnancy) Show forest plot | 2 | 77 | Peto Odds Ratio (Peto, Fixed, 95% CI) | 17.58 [0.29, 1073.88] |
| 15 Live birth rate (ASRM I/II endometriosis) Show forest plot | 1 | 33 | Peto Odds Ratio (Peto, Fixed, 95% CI) | 0.16 [0.03, 0.83] |
| 16 Live birth rate (ARSM III/IV endometriosis) Show forest plot | 1 | 114 | Peto Odds Ratio (Peto, Fixed, 95% CI) | 0.49 [0.22, 1.10] |
| 17 Complication rate (ARSM I/II endometriosis Show forest plot | 1 | 33 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.0 [0.0, 0.0] |
| 18 Complication rate (ARSM III/IV endometriosis) Show forest plot | 1 | 114 | Peto Odds Ratio (Peto, Fixed, 95% CI) | 0.14 [0.01, 1.42] |
| 19 Clinical pregnancy rate (ARSM I/II endometriosis) Show forest plot | 3 | 273 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.05 [0.74, 1.48] |
| 20 Clinical pregnancy rate (ARSM III/IV endometriosis) Show forest plot | 2 | 181 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.89 [0.53, 1.49] |
