Hum. Reprod. Update-2006-Kolibianakis-3-12

© The Author 2005. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved. For Permissions, please email: journals.permissions@oupjournals.org 3 Human Reproduction Update, Vol.12, No.1 pp. 3–12, 2006 doi:10.1093/humupd/dmi030 Advance Access publication August 25, 2005 Are endogenous LH levels during ovarian stimulation for IVF using GnRH analogues associated with the probability of ongoing pregnancy? A systematic review E.M.Kolibianakis1,5, J.Collins2,3, B.Tarlatzis1, E.Papanikolaou4 and P.Devroey4 1Department of Obstetrics and Gynaecology, Unit for Human Reproduction, Aristotle University of Thessaloniki, Thessaloniki, Greece, 2McMaster University, Hamilton, Ontario, 3Dalhousie University, Halifax, Nova Scotia, Canada and 4Centre for Reproductive Medicine, Dutch-Speaking Brussels Free University, Brussels, Belgium 5To whom correspondence should be addressed at: Unit for Human Reproduction, Papageorgiou General Hospital, Nea Efkarpia Peripheral Road, Thessaloniki 54603, Greece. E-mail: stratis.kolibianakis@otenet.gr The aim of this systematic review was to evaluate, among women with normal ovulation or World Health Organiza- tion (WHO) II oligoanovulation who undergo ovarian stimulation for IVF using GnRH analogues, whether endog- enous LH levels predict the likelihood of ongoing pregnancy beyond 12 weeks. A literature search identified six studies that answered the research question, among which two were prospective studies (one in GnRH agonist and one in GnRH antagonist cycles). None of the retrospective studies suggest that low endogenous LH levels are associated with a significantly decreased probability of ongoing pregnancy beyond 12 weeks in such patients. In the two prospective studies high endogenous LH levels during down-regulation were associated with a decreased probability of ongoing pregnancy beyond 12 weeks. Until further prospective studies modify the existing evidence summarized here, an adverse effect of low endogenous LH levels on the probability of ongoing pregnancy beyond 12 weeks is not a sensible rationale for LH supplementation during ovarian stimulation for IVF using GnRH analogues. Key words: GnRH agonists/GnRH antagonists/luteinizing hormone/ongoing pregnancy rate Introduction The role of endogenous LH during ovarian stimulation for IVF has been controversial since the pre-GnRH analogue era (Stanger and Yovich, 1985; Howles et al., 1987; Thomas et al., 1989). Following the introduction of analogues, interest focused mainly on whether low LH levels were associated with impaired IVF outcome. LH levels may be deeply suppressed after pituitary down- regulation for IVF, simulating the endocrine environment encoun- tered in World Health Organization (WHO) I patients. The rationale for LH supplementation in WHO I patients is clear (The European Recombinant Human LH Study Group, 1998); however, it cannot serve as a basis for LH supplementation in normo-ovulatory or WHO II patients. In theory, if low endogenous LH levels are associ- ated with a decreased probability of ongoing pregnancy, then LH supplementation might be indicated and its value should be assessed in randomized-controlled trials. On the other hand, if there is no association between LH and the probability of ongoing pregnancy or if low endogenous LH levels are associated with increased ongoing pregnancy rates, then there is no justification for consider- ing LH supplementation in normo-ovulatory or WHO II patients. Studies that address the association between LH levels and IVF outcome have measured LH levels on various days during ovarian stimulation and made use of different outcome measures. The out- comes include dose of FSH (Humaidan et al., 2002), estradiol levels on the day of HCG (Fleming et al., 1998), the number of fertilized oocytes (Humaidan et al., 2002), fertilization rates (Esposito et al., 2001), the number of cryopreserved embryos (Fleming et al., 1998) and the developmental potential of cryopreserved embryos (Rekha et al., 1998). To assess whether LH is truly an effector or predictor of IVF outcome in normo-ovulatory or WHO II patients, however, the outcome measure should be live birth, or it’s near-equivalent, ongoing pregnancy. The surrogate outcome measures listed above are not securely associated with live birth. The purpose of this systematic review is to assess, among normo-ovulatory or WHO II patients undergoing ovarian stimula- tion in GnRH analogue IVF cycles, whether endogenous LH lev- els predict ongoing pregnancy beyond 12 weeks. Materials and methods Identification of studies A literature search was performed using MEDLINE and the Cochrane Library. Additionally, references of retrieved articles were hand-searched. The search strategy was based on the following by guest on O ctober 19, 2012 http://hum upd.oxfordjournals.org/ D ow nloaded from E.M.Kolibianakis et al. 4 clinical question: among normo-ovulatory or WHO II patients undergoing ovarian stimulation for IVF using GnRH analogues, do endogenous LH levels predict the likelihood of ongoing preg- nancy? Ongoing pregnancy was defined as a pregnancy proceed- ing beyond 12 weeks of gestation. This was considered the most relevant outcome measure, especially in view of existing evidence suggesting an association between endogenous LH levels and early pregnancy loss (Westergaard et al., 2000). The search terms used in both databases were LH and pregnancy/IVF/assisted reproductive technologies (ART)/ovarian stimulation/oocytes/ reproduction/embryos. Meeting proceedings were not considered, because unpublished studies cannot be adequately evaluated for their design and qual- ity. Moreover, it has been shown that although there is a consider- able publication deficit in reproductive medicine for RCT, there is no concomitant publication bias (Evers, 2000). Selection of studies All published studies addressing the research question were ini- tially considered for this review regardless of the direction of study (retrospective or prospective), the sample size or the day endogenous LH levels were assessed during ovarian stimulation. The next step selected only studies in which ovarian stimulation was performed for IVF using gonadotrophins containing no LH, regardless of the analogue used. Studies were excluded if no down-regulation was used for ovarian stimulation for IVF or if LH levels were assessed only before the initiation of stimulation. If the reported outcome measure (pregnancy rate) was not ongoing pregnancy rate beyond 12 weeks, the study was labelled as ‘preg- nancy not as defined’, and these studies are described separately in this report. For those studies identified in which the outcome measure was not clearly defined, the corresponding authors were contacted and the information was retrieved where this was possible. Studies identified The first study describing the use of agonists in IVF appeared in the literature in 1984, by Porter et al. Literature search, after 1984, and hand search of the retrieved articles resulted in 416 studies that were potentially able to answer the research question. Of these articles, 78 papers were identified for further evaluation based on their titles and abstracts. Finally, 34 manuscripts were selected for full text review of which 24 manuscripts were excluded from the current review because they did not address the research question (n = 9), were review articles (11) or letters to editors (1), because down-regulation was not used for ovarian stimulation (n = 2) or because the gonadotrophin used for stimulation contained LH (n = 1). Of the 10 articles which did evaluate the association between LH levels during ovarian stimulation for IVF using GnRH analogues and pregnancy achievement in normo-ovulatory or WHO II patients, six studies used as the outcome measure achievement of ongoing pregnancy beyond 12 weeks, and four did not define pregnancy as specified in the research question for the current review. Data extraction The following data were recorded from each of the 10 studies: type of study, citation data, country, period of enrolment, number of patients included, number of cycles performed, baseline charac- teristics of the patients included, type and protocol of ovarian stimulation, type of gonadotrophin administered, criteria for HCG administration, day of embryo transfer, type of luteal support administered and method of LH assessment. Quantitative data synthesis Study features and results were assembled in tabular form and a formal meta-analysis was not done. Results Six studies, incorporating 1103 patients, that answered the research question were identified (Westergaard et al., 2000; Balasch et al., 2001; Esposito et al., 2001; Humaidan et al., 2002; Kolibianakis et al., 2004; Merviel et al., 2004). Two were pro- spective (Humaidan et al., 2002, Kolibianakis et al., 2004), four involved GnRH agonist cycles (717 patients) and two GnRH antagonist cycles (386 patients) (Tables I, II and III). The median number of patients included in the above studies was 183 (range 116–270). Five of the studies were done in Europe, one in the United States of America and all were published within the last 5 years. Table IV summarizes baseline characteristics of the patients included in the studies that answered the research question in the current review. All six studies described inclusion criteria and analysed one cycle per patient. Three of the studies evaluated the association between endogenous LH levels and probability of ongoing preg- nancy on day 8 (Westergaard et al., 2000; Humaidan et al., 2002; Kolibianakis et al., 2004), one study on day 7 of stimulation (Balasch et al., 2001) and one study on the day of HCG adminis- tration (Merviel et al, 2004). One study reported on the mean peri- ovulatory endogenous LH levels based on 4–5 measurements performed after stimulation day 5 (Esposito et al., 2001) (Table V). Table I. Characteristics of the studies included in the current review (A) Authors Country of origin Journal Type of study Is power analysis present? Number of patients included Patient diagnosis in subgroups of LH Is study period mentioned? Westergaard et al. (2000) Denmark Hum Reprod Retrospective Not available 200 Yes Yes Balasch. et al. (2001) Spain Hum Reprod Retrospective Not available 144 Yes Yes Esposito et al. (2001) United States of America Fertil Steril Retrospective Not available 166 No Yes Humaidan et al. (2002) Denmark Hum Reprod Prospective No 207 Yes Yes Merviel et al. (2004) France Fertil Steril Retrospective Not available 270 Yes Yes Kolibianakis et al. (2004) Belgium Hum Reprod Prospective Yes 116 Yes Yes by guest on O ctober 19, 2012 http://hum upd.oxfordjournals.org/ D ow nloaded from LH and ongoing pregnancy achievement in IVF 5 The LH threshold used to characterize patients as having low endogenous LH levels was defined by percentile analysis of LH values on day 8 of stimulation (Kolibianakis et al., 2004), was an arbitrary choice (Westergaard et al., 2000; Esposito et al., 2001; Humaidan et al., 2002; Merviel et al., 2004), or was based on pre- viously reported arbitrarily defined LH thresholds (Balasch et al., 2001). Multivariate analysis was performed in two studies to further elucidate the association of endogenous LH levels and ongoing pregnancy achievement (Esposito et al., 2001; Kolibianakis et al., 2004). The studies by Balasch et al. (2001) and Esposito et al. (2001) incorporated estradiol (E2) in the criteria used for HCG administration. In the remaining studies, criteria used for trigger- ing final oocyte maturation were independent of E2 levels. With one exception, however (Kolibianakis et al., 2004), the criteria were loosely defined (e.g. two or more follicles of 16 or more mm diameter). This means that HCG might not be administered as soon as the follicular threshold used in that study was reached but could be postponed. As Table V summarizes, none of the six studies reporting on LH levels before HCG administration found that low endogenous LH levels were associated with a significantly decreased probabil- ity of ongoing pregnancy in normo-ovulatory or WHO II patients undergoing IVF using GnRH analogues for inhibition of prema- ture LH surge. In the prospective GnRH agonist study (Humaidan et al., 2002), the chance of ongoing pregnancy was significantly lower if endogenous LH levels were ≥1.5 IU/l on day 8 of stimula- tion, compared to <1.5 IU/l [24.3% (9/37) versus 42.9% (73/170), respectively; P = 0.04]. In the prospective GnRH antagonist study a significant trend towards decreasing ongoing pregnancy rates with increasing endogenous LH levels on day 8 of stimulation was observed in a recombinant FSH (rFSH) and day 6 start GnRH antagonist protocol (Kolibianakis et al., 2004). In that study not only were the low endogenous LH levels on day 8 of stimu- lation of no concern for achievement of ongoing pregnancy but they were associated with a higher probability of an ongoing pregnancy. Tables VI, VII and VIII list the studies in which pregnancy was not defined as specified in the research question for the current review (Fleming et al., 2000; Coppola et al., 2003; Ferrari et al., 2004) or no information could be obtained from the authors regarding definition of pregnancy rate (Loumaye et al., 1997). As Table IX summarizes, only one of four studies found an associ- ation between endogenous LH levels and the probability of preg- nancy in normo-ovulatory or WHO II patients undergoing IVF using GnRH analogues. In that study, pregnancy rates were higher with LH >0.5 mIU/ml (Ferrari et al., 2004). Discussion Six studies evaluated the association between endogenous LH lev- els during ovarian stimulation and the likelihood of ongoing preg- nancy beyond 12 weeks in normo-ovulatory or WHO II patients treated for IVF with GnRH analogues. Several conclusions can be drawn from their results. Although the six studies used different LH assay methods and LH standards may vary among laborato- ries, the trends in all eligible studies involved no risk of lower on- going pregnancy rates with low LH values and possible risk with higher values. Because the association has been assessed in only two prospective studies, one in GnRH agonists and one in GnRHTa bl e II . Ch ar ac te ris tic s o f t he st ud ie s in cl u de d in th e c ur re n t r ev ie w (B ) E 2 , es tr ad io l; N A , n ot a v ai la bl e. A ut ho rs Ty pe o f fe rt ili za tio n Ty pe o f an al og u e u se d A na lo gu e pr o to co l W as d ow n- re gu la tio n co n fir m ed ? St ar tin g do se o f g on ad ot ro ph in s W as g on ad ot ro ph in do se a dju ste d d ur in g st im ul at io n ? H CG d os e u se d (IU ) D ay o f e m br yo tr an sf er Lu te al su pp or t W es te rg aa rd e t a l. (20 00 ) IV F/ IC SI B us er el in a ce ta te Lo n g m id lu te al Y es , by E 2 , U SS 22 5 IU G on al -F Y es 10 00 0 D ay 3 M ic ro n iz ed pr o ge st er o n e B al as ch e t a l. (20 01 ) IV F/ IC SI Le u pr ol id e a ce ta te Lo n g m id lu te al Y es , by E 2, U SS St ep d ow n sta rti ng w ith 4 50 IU G on al -F Y es 50 00 D ay 2 H CG Es po sit o et a l. (20 01 ) IV F Le u pr o lid e ac et at e Lo n g m id lu te al Y es , by E 2 , U SS 30 0– 45 0 IU G on al -F Y es 10 00 0 D ay 3 Pr o ge ste ro n e IM o r v ag in al pr o ge st er o n e H um ai da n et a l. (20 02 ) IV F/ IC SI B us er el in a ce ta te Lo n g m id lu te al Y es 10 0– 37 5I U G on al -F Y es 10 00 0 D ay 2 , d ay 3 o r da y 5 Cy cl og es t o r cr in on e M er v ie l e t a l. (20 04 ) IV F/ IC SI Ce tro re lix Si ng le d os e da y 8/ 0. 25 m g s. c. da ily fr o m da y 12 if n ec es sa ry N A 22 5 IU P ur eg o n o r G on al -F Y es 10 00 0 D ay 2 o r d ay 3 M ic ro n iz ed pr o ge st er o n e H CG K ol ib ia na ki s e t a l. (20 04 ) IV F/ IC SI G an ire lix 0. 25 m g fro m d ay 6 o f st im ul at io n N A , b as al h or m o n al le ve ls w er e pr es en t a t in iti at io n of st im ul at io n 20 0 IU P ur eg o n N o 10 00 0 D ay 3 o r d ay 5 M ic ro n iz ed pr o ge st er o n e by guest on O ctober 19, 2012 http://hum upd.oxfordjournals.org/ D ow nloaded from E.M.Kolibianakis et al. 6 Ta bl e II I. Ch ar ac te ris tic s o f t he st u di es in cl u de d in th e cu rr en t r ev ie w (C ) E 2 , es tr ad io l; N A , n o t a v ai la bl e. A ut ho rs O u tc om e m ea su re as so ci at ed w ith L H Ty pe o f as so ci at io n be tw ee n LH an d on go in g pr eg n an cy LH k it us ed D ay L H w as a ss es se d W ay th e LH th re sh ol d w as d ef in ed LH th re sh ol d u se d Pr o po rt io n o f pa tie nt s i n th e ‘ lo w LH ’ gr o u p Cr ite ria u se d fo r H CG ad m in ist ra tio n Cr ite ria fo r E 2 Cr ite ria fo r f ol lic le s W es te rg aa rd et a l. (20 00 ) D el iv er y ra te N o as so ci at io n A ut oD EL FI A h LH (W al la c O y, Tu rk u, Fi nl an d), A x Sy m sy s- te m , C V s m en tio ne d St im ul at io n da y 8 A ss ay d et ec tio n lim it, lit er at ur e 0. 5 IU /l 49 % , < 0. 5 IU /l Fo lli cl es N A ≥4 fo lli cl es o f > 16 m m B al as ch et a l. (20 01 ) O ng oi ng pr eg na nc y at 2 0 w ee ks N o as so ci at io n Im m u n o 1 Te ch ni co n B ay er , (T ar ry to w n , N Y , U SA ), C V s m en tio ne d St im ul at io n da y 7 Th re sh ol ds pr o po se d in th e lit er at ur e 0. 5 IU /l, 0. 7 IU /l, 1 IU /l 6. 9% , < 0. 5 IU /l Fo lli cl es , E 2 Co n sis te nt ris e in E 2 ≥2 fo lli cl es o f > 18 m m an d ≥4 fo lli cl es o f ≥ 14 m m Es po sit o et a l. (20 01 ) O ng oi ng pr eg na nc y at 2 0 w ee ks N o as so ci at io n B ay er Te ch n ic on Im m u n o 1 Sy ste m (T ar ry to w n, N Y , U SA ), C V s m en tio ne d A v er ag e of 4 -5 as se ss m en ts st ar tin g fro m da y 5 of st im ul at io n Th e th re sh o ld u se d w as th e lo w er li m it o bs er v ed in a n at u ra l cy cl e 3 IU /l 69 .8 % , < 3 IU /l Fo lli cl es , E 2 M ea n E 2 o f ap pr o x im at el y 25 0 pg /m l p er le ad in g fo lli cl e ≥2 fo lli cl es o f ≥ 20 m m H um ai da n et a l. (20 02 ) O ng oi ng pr eg na nc y at 1 2 w ee ks N o as so ci at io n A ut oD EL FI A h LH (W al la c O y, Tu rk u, Fi nl an d), C Vs m en tio ne d St im ul at io n da y 8 A rb itr ar y 0. 5 IU /l 11 .6 % , < 0. 5 IU /l Fo lli cl es N A ≥3 fo lli cl es o f ≥ 17 m m M er v ie l et a l. (20 04 ) D el iv er y ra te N o as so ci at io n A CS :1 80 (B ay er Co rp o ra tio n , Ta rr yt o w n , N Y , U SA ), C V s m en tio ne d D ay o f H CG A rb itr ar y 0. 5 IU /l 44 .1 % , < 0. 5 IU /l Fo lli cl es N A ≥4 fo lli cl es o f 1 6 m m K ol ib ia na ki s et a l. (20 04 ) O ng oi ng pr eg na nc y ra te at 12 w ee ks N eg at iv e as so ci at io n El ec sy s im m un o an al ys er (R o ch e D ia gn o st ic s