- Email: [email protected]
Effects of estradiol supplementation during the luteal phase of in vitro fertilization cycles: a meta-analysis
Effects of estradiol supplementation during the luteal phase of in vitro fertilization cycles: a meta-analysis Byung Chul Jee, M.D.,a Chang Suk Suh, M.D.,a,b,c Seok Hyun Kim, M.D.,b,c Yong Beom Kim, M.D.,a and Shin Yong Moon, M.D.b,c a Department of Obstetrics and Gynecology, Seoul National University Bundang Hospital, Seongnam; and b Department of Obstetrics and Gynecology, College of Medicine, and c Institute of Reproductive Medicine and Population, Seoul National University, Seoul, South Korea
Objective: To clarify whether adding E2 to standard luteal P supplementation is beneficial both in GnRH agonist and antagonist IVF cycles. Design: Meta-analysis of nine randomized controlled trials. Setting: University hospital center for reproductive medicine and IVF. Intervention(s): None. Main Outcome Measure(s): Clinical pregnancy rate (PR) per patient, clinical PR per embryo transfer (ET), implantation rate, ongoing PR per patient, clinical abortion rate, and ectopic PR. Result(s): There were no statistically significant differences between E2þP versus P-only group regarding overall IVF outcomes. From seven studies including GnRH agonist cycles, no statistical significant differences were found between the two groups in clinical PR per patient (relative risk [RR] 1.32, 95% confidence interval [CI] 0.79–2.19), clinical PR per ET (RR 1.83, 95% CI 0.96–3.49), implantation rate (RR 1.20, 95% CI 0.34–4.21), ongoing PR per patient (RR 1.34, 95% CI 0.37–4.82), clinical abortion rate (RR 1.05, 95% CI 0.48–2.28), and ectopic PR (RR 0.53, 95% CI 0.07–4.10). Clinical PR per patient (RR 0.94, 95% CI 0.62–1.42) and ongoing PR per patient (RR 1.09, 95% CI 0.79–1.50) from three studies including GnRH antagonist cycles only were all similar between the two groups. Conclusion(s): The combined data presented in this meta-analysis suggest that the addition of E2 to P for luteal phase support does not improve IVF outcomes in GnRH agonist and antagonist cycles. However, the authors feel that there is an obvious need for further large-scale studies regarding GnRH antagonist cycles. (Fertil Steril 2010;93:428–36. 2010 by American Society for Reproductive Medicine.) Key Words: Estradiol, luteal phase, in vitro fertilization, GnRH agonist, GnRH antagonist
Luteal phase support has been a routine practice in IVF-ET because stimulated IVF cycles are associated with a defective luteal phase in almost all patients (1, 2). Three recent metaanalyses demonstrated that hCG is equally effective (3, 4) or superior to P (5) for luteal phase support with respect to clinical pregnancy rate (PR). Nonetheless, P is often favored, because hCG has a potential for increasing rates of ovarian hyperstimulation syndrome (4, 6). Various formulations of P are currently available, including oral, vaginal, rectal, and IM forms. Vaginal P gel and IM P were found to have similar clinical and ongoing PR (7). However, in a subsequent meta-analysis, clinical PR and delivery rate were significantly higher when IM P was used compared with vaginal application (3). Intramuscular Received December 15, 2008; revised January 29, 2009; accepted February 11, 2009; published onilne April 1, 2009. B.C.J. has nothing to disclose. C.S.S. has nothing to disclose. S.H.K. has nothing to disclose. Y.B.K. has nothing to disclose. S.Y.M. has nothing to disclose. Presented at the 64th Annual Meeting of the American Society for Reproductive Medicine, San Francisco, November 8–12, 2008. Reprint requests: Chang Suk Suh, M.D., Department of Obstetrics and Gynecology, Seoul National University Bundang Hospital, 300 Gumi, Bundang, Seongnam, Gyeonggi 463-707, South Korea (FAX: 82-31-787-4054; E-mail: [email protected]).
428
P has been widely used for luteal phase support but is often associated with a number of side effects, including painful injections with a rash, inflammatory reactions, and abscess formations (8). In these circumstances, vaginal administration of P can be a viable alternative. Moreover, intravaginal route of P supplementation is regarded as a first-choice luteal support regimen in stimulated IVF cycles (9). An earlier report indicated that serum E2 concentrations severely drop at the end of the luteal phase (10); therefore, a concern has been raised about an additional supply of E2 during the luteal phase in IVF cycles. In the first half of 1990, two prospective randomized studies were performed to evaluate the possible benefit of adding 6 mg (11) and 2 mg (12) E2 valerate daily in women treated with a GnRH agonist long protocol and gonadotropins for IVF. In those studies, the clinical PRs were almost equal between the groups with and without E2 cotreatment. During the luteal phase, ovarian E2 has experienced a sharp fall after its preovulatory peak and starts to rise again. The differences in luteal E2 in conception and nonconception cycles start to appear on day 9 to 10 with respect to the LH peak (13). Based on these observations, two studies were conducted in which patients with a precipitous drop of luteal phase serum E2 (14) or serum E2 concentration <100
Fertility and Sterility Vol. 93, No. 2, January 15, 2010 Copyright ª2010 American Society for Reproductive Medicine, Published by Elsevier Inc.
0015-0282/10/$36.00 doi:10.1016/j.fertnstert.2009.02.033
pg/mL at 11 days after ET (15) were selected. In those studies, oral E2 was additionally administered within 10 days after ovulation triggering (14) or 11 days after ET (15) and significantly higher clinical (14) or ongoing (15) PR per patient was observed. Farhi et al. (16) compared IVF outcomes with and without 2 mg E2 valerate as a luteal phase support in patients with serum E2 concentration at triggering day >2,500 pg/mL. In that study, E2 was started 11 days after ET and a significantly higher clinical PR per ET and implantation rate was noted in GnRH agonist long protocol cycles but not in short protocol. Although E2 was started several days later after ET and the patients were selected according to serum E2 concentration at specific points, three studies consistently demonstrated the beneficial effect of adding E2 as a luteal phase support. Pritts and Atwood (3) performed a meta-analysis addressing this issue based on three randomized trials (11, 12, 16). Although E2 doses and duration of treatment varied, they concluded that addition of oral E2 to standard P treatment improved implantation rate. However, that conclusion was derived from only one study (16); moreover, clinical PRs from three studies showed no difference between study and control group (relative risk [RR] 1.15, 95% confidence interval [CI] 0.83–1.58). Through a Cochrane review, Daya and Gunby (4) reported that there were no significant differences in clinical or ongoing PR, miscarriage, or live birth rate when P combined with E2 was compared to P alone; however, this conclusion was drawn only from two randomized trials (12, 17). The results from four subsequent studies were quite inconsistent and controversial. Almost equal clinical PR per patient was reported in two studies (18, 19), but significantly increased PR per patient and per ET by adding E2 was noted from other studies (20, 21). Interestingly, Lukaszuk et al. (21) reported a significantly higher clinical PR per patient in a 6 mg E2 supplement group, but not in a 2 mg E2 group, compared with no E2 (56.5% vs. 48.9% vs. 36.0%, respectively). Since 2007, four reports have been available; Drakakis et al. (22) reported a significantly higher clinical PR in an E2 supplement group in patients using GnRH agonist long protocol, but this same effect was not observed in subsequent studies (23, 24). In GnRH antagonist cycles, there were no differences in clinical and ongoing PR between E2 supplement vs. no supplement group (23, 25). Although two recent meta-analyses confirmed that the addition of E2 to P for luteal phase support in IVF/ICSI cycles has no beneficial effect on PR (26, 27), it is unclear whether the effect of E2 supplementation is different between GnRH agonist and antagonist cycles. No meta-analysis addressing this issue has been reported. A recent review (9) indicates that the addition of E2 seems to be beneficial in long GnRH agonist protocol but not in short GnRH agonist and GnRH antagonist protocol. The conclusion about long GnRH agonist protocol stemmed from Fertility and Sterility
four studies (11, 12, 16, 21) but they did not attempt to perform a meta-analysis. Moreover, the conclusion about GnRH antagonist protocol was based on only one study (28). In the present meta-analysis, we attempted to clarify the effect of E2 supplement in a luteal phase of stimulated IVF cycles. Furthermore, analysis of subgroups was performed to demonstrate whether the effect of E2 supplementation is different between GnRH agonist and antagonist cycles.
MATERIALS AND METHODS A literature search of the National Library of Medicine and the National Institutes of Health (PubMed), Medline and Cochrane Controlled Trials Register (CENTRAL; Cochrane Library) was performed using the key words ‘‘estradiol,’’ ‘‘estrogen,’’ ‘‘luteal phase support,’’ ‘‘luteal supplementation,’’ and ‘‘in vitro fertilization.’’ The last search date was October 2008. The inclusion criteria for selecting an article were defined as follows: 1. Fresh IVF-ET cycles using autologous oocyte. 2. The patients underwent ovarian hyperstimulation with pituitary suppression by GnRH agonist or antagonist; the data could be separable according to the method of pituitary suppression. 3. E2 starting on at least ET day, supplemented by oral, vaginal, or transdermal route. 4. P supplemented either vaginally or IM during an entire luteal phase. 5. Only prospective, randomized, controlled studies. 6. Published in English. We initially found 15 original articles comparing IVF-ET outcomes after administration of E2þP versus P only as a luteal supplementation. Finally, nine studies met the inclusion criteria and were reviewed in this meta-analysis (12, 17, 18, 20–23, 25, 28). All of the included studies were a prospective and randomized. We used the published data only. In all except two studies (20, 21), patients could enter the study only once. Cycle characteristics and IVF-ET outcomes from nine studies are summarized in Table 1. With one study including both GnRH agonist and antagonist cycle separately (23), IVF outcomes of GnRH agonist cycles (long protocol) could be extracted from seven studies and outcomes of GnRH antagonist cycles could be obtained from three studies. Six studies were excluded in this meta-analysis. In the study by Smitz et al. (11), 28% of the subjects underwent zygote intrafallopian transfer, which could not be separated from IVF-ET results. Estradiol valerate was administered from 4 days after retrieval. Gleicher et al. (14) recruited 330 consecutive ovarian stimulation cycles; however, most patients underwent intrauterine insemination (IUI), which could not be separated from IVF-ET results. In the study by Kaider and Coulam (15), oral E2 was started when serum E2 concentration was <100 pg/mL and serum hCG concentration >5 mIU/mL 11 days after ET. Farhi et al. (16) started oral E2 7 days after ET in patients with serum E2 429
430
TABLE 1 Summary of nine prospective randomized studies comparing IVF-ET outcomes after E2DP versus P only as a luteal supplementation.
Jee et al.
Authors, year (ref.)
Estradiol as a luteal phase support
Ceyhan et al., 2008 (25)
Patient characteristics Normal responders; exclude endometriosis, PCOS, severe male factor; mean age 31.4 vs. 30.9
Engmann et al., First IVF cycle; low and high responders excluded; 2008 (23),a mean age 34.9 vs. 35.7
Regimen and duration of luteal support (starting date ending date)
No. of patients
IVF outcomes (E2DP vs. P)
E2: transdermal E2 100 mg/d 2/wk (OPU 8 wks GA) P: vaginal MP 600 mg/d (OPU 8 wk GA)
Antagonist (fixed)
30 vs. 29
CPR/PA 43.3 vs. 44.8 CPR/ET 56.5 vs. 61.9 OG/PA 36.6 vs. 34.4
E2; vaginal ME (Estrace) 2 mg 2/d (ET fetal heartbeat) P: IM P 50 mg/d (OPU fetal heartbeat)
Long
55 vs. 59
CPR/PA 49.1 vs. 71.2c IR 26.9 vs. 39.6 CAR 7.4 vs. 11.9 OG/PA 45.5 vs. 62.7
Antagonist (flexible)
25 vs. 19
CPR/PA 48.0 vs. 52.6, IR 23.5 vs. 18.9 OG/PA 48.0 vs. 47.4 CAR 0.0 vs. 10.0 CPR/PA 43.6c vs. 15.8 IR 9.7 vs. 4.0 CAR 23.5 vs. 16.7 EP 5.6 vs. 0.0
Long E2: E2V 2 mg þ 0.5 mg norgestrel 3/d for 15 d and transdermal E2 50 mg q 4d (OPU pregnancy test) P: oral MP 100 mg 3 þ vaginal MP 200 mg 3 (ET-1 pregnancy test)
Vol. 93, No. 2, January 15, 2010
Drakakis et al., 2007 (22)
Not specified; mean age 35.4 vs. 35.8
Fatemi et al., 2006 (28)
Age %39, <3 prior cycles; E2: E2V 2 mg 2/d (OPUþ1 7 weeks GA) exclude PCOS, endometriosis III/IV; mean age: 32.1 vs. 32.0 P: vaginal MP 300 mg 2/d (same)
Jee. Estradiol as a luteal phase support. Fertil Steril 2010.
Protocol for suppression
Antagonist (fixed)
39 vs. 38
101 vs. 100 OG/PA 29.7 vs. 26.0 IR 42.4% vs. 37.8% Early pregnancy loss (including chemical, EP, and first trimester abortion) 23.1 vs. 23.4
Fertility and Sterility
TABLE 1 Continued. Authors, year (ref.)
Patient characteristics
Regimen and duration of luteal support (starting date ending date)
Protocol for suppression
No. of patients
IVF outcomes (E2DP vs. P) CPR/PA 56.5c vs. 48.9 vs. 36.0 CPR/ET 51.3c vs. 32.9 vs. 23.1 IR 29.9c vs. 17.8 vs. 9.8 CAR 12.8 vs. 17.4 vs. 22.2 EP 0.0 vs. 4.2 vs. 5.3 CPR/PA 62.9c vs. 18.9 CPR/ET 31.4c vs. 9.5 OG/PA 48.6c vs. 18.9 CAR 22.7 vs. 0.0
Lukaszuk et al., 2005 (21)
Age <40, mean age 31.1 vs. 31.7 vs. 32.1
E2: E2V 6 mg/d vs. 2 mg/d (OPU not specified) P: vaginal MP 200 mg 3/d (OPU not specified)
Long, pretreated by OC
69 vs. 47 vs. 50
Gorkemli et al., 2004 (20),b
Age %41, mean age 30.8 vs. 30.8
E2: transdermal E2 100 mg/d (OPU pregnancy test) P: vaginal MP 200 mg 3/d (OPU 10 wk GA)
Long
70 vs. 74
Rashidi et al., 2004 (18)
Not specified, mean age 29.6 vs. 30.0
Long
41 vs. 37
CPR/PA 31.7 vs. 35.1 CAR 17.6 vs. 7.7
Tay and Lenton, 2003 (17)
Basal FSH <10 mIU/mL, mean age 35 vs. 33
Long
28 vs. 35
CPR/PA 17.9 vs. 20.0
Lewin et al., 1994 (12)
Not specified, mean age 33.1 vs. 32.7
E2: E2V 4 mg/d (ET fetal heartbeat) P: IM P 100 mg/d (ET 12 wk GA) E2: E2V 2 mg/d (OPUþ1 not specified) P: vaginal MP 200 mg 2/d (OPUþ1 not specified) E2: E2V 2 mg/d (ET ETþ20) P: IM P 50 mg/d (same)
Long
50 vs. 50
CPR/PA 26.0 vs. 28.0 CPR/ET26.5 vs. 28.0 LBR 76.1 vs. 78.6
Note: CAR ¼ clinical abortion rate; CPR/ET ¼ clinical pregnancy rate per ET; CPR/PA ¼ clinical pregnancy rate per patient; E2V ¼ estradiol valerate; ET ¼ embryo transfer day; ETþ1 ¼ 1 day after ET; ET-1 ¼ 1 day before ET; EP ¼ ectopic pregnancy rate; IM ¼ intramuscular; IR ¼ implantation rate; LBR ¼ live birth rate; ME ¼ micronized estradiol; MP ¼ micronized progesterone; OC ¼ oral contraceptives; OG/PA ¼ ongoing pregnancy per patient; OPU ¼ ovum pick-up day. a Excluding eight women who were assigned to OCþmicrodose GnRH agonist group. b 16 mg/d methylprednisolone was given orally to all patients for 5 days after ET. c Significant increase compared with control. Jee. Estradiol as a luteal phase support. Fertil Steril 2010.
431
concentration at triggering day >2,500 pg/mL. One study written in Spanish was excluded (19). Serna et al. (24) performed an RCT, but the method of pituitary suppression could not be separated.
’’no difference’’ between study and control group (RR 1.48, 95% CI 0.99–2.23). When including data from the 6 mg E2 supplement group only, the RR still did not reach statistical significance (RR 1.28, 95% CI 0.83–1.99).
We also found seven studies published as abstracts (29–35) but all of them were excluded due to following reasons;
Clinical PR per ET in E2 groups was almost two times higher than in no E2 groups; however, the difference did not reach statistical significance. Implantation rates, ongoing PRs per patient, clinical abortion rates, and ectopic PRs were similar between study and control groups. Live birth rate was reported to be similar in both groups, but the rate was reported in only one study (12).
Estradiol was started on day 7 after retrieval in the study by Pouly et al. (33). The method of pituitary suppression could not be identified in the study by Karlikaya et al. (34). The remaining five were published later as full-length articles. In the present meta-analysis, primary outcome was defined as clinical PR per patient. Secondary outcomes were clinical PR per ET, implantation rate, ongoing PR per patient, clinical abortion rate, and ectopic PR. Clinical pregnancy was defined as seeing a gestational sac (GS) on transvaginal ultrasound. Biochemical pregnancy was completely ignored. The other parameters were defined as follows: Implantation rate: total number of GS/total number of embryos transfered. Ongoing pregnancy: clinical pregnancy developing beyond 12 weeks or the second trimester. Clinical abortion rate: number of first-trimester abortions/ number of clinical pregnancies. Ectopic PR: number of ectopic pregnancies/number of clinical and ectopic pregnancies A meta-analysis was performed using Review Manager (RevMan ver. 4.2 for Windows; Nordic Cochrane Center, Copenhagen, Denmark). Random effect model was always used, because the included studies had a substantial heterogeneity regarding dosage or route of the medication and treatment duration. A publication bias was assessed by the Begg-Mazumdar test regarding clinical PR per patient. Using the Begg-Mazumdar test, publication biases were not detected in both situations: Kendall tau ¼ 0.14 (P¼.72) in analysis including either GnRH agonist or antagonist cycles; Kendall tau ¼ 0.24 (P¼.56) in analysis including GnRH agonist cycles only. RESULTS First, IVF outcomes from nine studies including either GnRH agonist or antagonist cycles were assessed. Clinical PR per patient from eight studies was higher in E2 supplement groups compared to no E2 groups (45.4% vs. 35.0%); however, this difference was not statistically significant when using random effect model (RR 1.27, 95% CI 0.83–1.96; Table 2). Lukaszuk et al. (21) reported a significantly higher clinical PR per patient in a 6 mg E2 supplement group (RR 1.57, 95% CI 1.03–2.40) but not in a 2 mg E2 group (RR 1.36, 95% CI 0.85–2.18) compared with no E2 group. In the present metaanalysis, data from the 2 mg and 6 mg E2 supplement groups were combined. This combination of data resulted in overall 432
Jee et al.
Estradiol as a luteal phase support
Second, IVF outcomes from GnRH agonist cycles only were assessed separately. Again, clinical PR per patient (from seven studies) in E2 supplement groups was higher but not statistically different compared with no E2 groups (45.4% vs. 33.2%; RR 1.32, 95% CI 0.79–2.19; Fig. 1). There were no differences in clinical PR per ET, implantation rates, ongoing PRs per patient, clinical abortion rates, and ectopic PRs between study and control groups. A high RR (1.83) in clinical PR per ET was mainly attributed to two studies, in which a significantly higher clinical PR per ET was reported in E2 supplement groups: RR 3.32, 95% CI 1.91–5.79 (20); and RR 1.84, 95% CI 1.18–2.88 (21). The IVF outcomes of GnRH antagonist cycles could be extracted from three studies. Clinical PRs per patient from two studies and ongoing PRs per patient from three studies were all similar between study and control groups. DISCUSSION The results of the present meta-analysis indicate that the addition of E2 to P for luteal phase support does not improve IVF outcomes in GnRH agonist cycles. This finding is consistent with three earlier meta-analyses (4, 26, 27). In addition, we found that adding E2 to P for luteal phase support has no benefit of enhancing IVF outcomes in GnRH antagonist cycles. It is well established that luteal function is compromised in IVF cycles using GnRH agonist, and this phenomenon could potentially be attributed to ovarian stimulation and the resultant altered hormone levels, the process of oocyte retrieval, and direct effect of the GnRH agonist on the corpus luteum (36). Use of GnRH agonists may be associated with decreased production of E2 and P in the luteal phase, a reduction in the length of the luteal phase, and impairment of endogenous gonadotropin secretion caused by persistent pituitary suppression. Although GnRH antagonist causes less sustained pituitary suppression than GnRH agonist, several studies suggest that IVF cycles using GnRH antagonist protocol also need luteal phase support (37–39). However, adequate regimens for luteal phase support have not been precisely studied in GnRH antagonist cycles. Because the luteal phase characteristics appear to be similar between GnRH agonist and antagonist cycles (40), it could be assumed that supplementation regimen would be the same in both GnRH agonist and antagonist protocols. In this context, it is likely that the addition of E2 to P for luteal phase Vol. 93, No. 2, January 15, 2010
TABLE 2 Results of meta-analysis comparing IVF-ET outcomes after E2DP versus P only as a luteal supplementation. No. of included studies (ref.)
IVF-ET outcomes E2DP group
P group
RR (95% CI)
Including either GnRH agonist or antagonist cycles Clinical PR per patienta 8 (12, 17, 18, 20–23, 25) Clinical PR per ET 4 (12, 20, 21, 25) Implantation rate 2 (22, 23) Ongoing PR per patient 4 (20, 23, 25, 28) Clinical abortion rate 5 (18, 20–23) Ectopic PR 2 (21, 22)
45.4% (206/454) 36.9% (132/358) 18.3% (69/378) 39.9% (112/281) 16.0% (28/175) 2.5% (2/81)
35.0% (137/391) 19.9% (59/297) 20.5% (70/342) 34.2% (96/281) 11.7% (12/103) 4.0% (1/25)
1.27 (0.83–1.96)h 1.53 (0.84–2.78)h 1.26 (0.41–3.86)h 1.23 (0.73–2.08)h 1.02 (0.43–2.38) 0.53 (0.07–4.10)
Including GnRH agonist cycles only Clinical PR per patientb 7 (12, 17, 18, 20–23) Clinical PR per ET 3 (12, 20, 21) Implantation rate 2 (22, 23) Ongoing PR per patient 2 (20, 23) Clinical abortion rate 5 (18, 20–23) Ectopic PR 2 (21, 22)
45.4% (181/399) 35.5% (119/335) 17.1% (53/310) 47.2% (59/125) 17.2% (28/163) 2.5% (2/81)
33.2% (114/343) 16.7% (46/276) 20.8% (59/284) 38.3% (51/133) 11.8% (11/93) 4.0% (1/25)
1.32 (0.79–2.19)h 1.83 (0.96–3.49)h 1.20 (0.34–4.21)h 1.34 (0.37–4.82)h 1.05 (0.48–2.28) 0.53 (0.07–4.10)
Including GnRH antagonist cycles only Clinical PR per patient 2 (23, 25) Ongoing PR patient 3 (23, 25, 28)
45.5% (25/55) 34.0% (53/156)
47.9% (23/48) 30.4% (45/148)
0.94 (0.62–1.42) 1.09 (0.79–1.50)
Parameters
Note: Relative risk (RR) calculated by random effect model. CI ¼ confidence interval; other abbreviations as in Table 1. a Begg-Mazumdar test revealed that there was no publication bias (Kendall tau ¼ 0.14; P¼ .72). b Begg-Mazumdar test revealed that there was no publication bias (Kendall tau ¼ 0.24; P¼ .56). h There was a significant heterogeneity. Jee. Estradiol as a luteal phase support. Fertil Steril 2010.
support does not improve IVF outcomes in GnRH antagonist cycles, as similarly found in GnRH agonist cycles. In an earlier study, E2 and P receptors in the endometrium as well as endometrial histology did not differ in the luteal phase of ovariectomized women supplemented with or without exogenous E2 (41). A subsequent study suggests that luteal E2 is not fundamental in the preparation of the endometrium for implantation (13, 42). In a more recent trial, in which a GnRH antagonist was used, the endocrine profile (FSH, E2, and P) was similar in both groups receiving P alone vs. PþE2 during the luteal phase (43). In GnRH antagonist cycles, no significant differences were noted in estrogen receptor expression from endometrial biopsies obtained on the day of oocyte retrieval and 3, 5, and 10 days after retrieval among three study groups: no luteal support, vaginal micronized P, and micronized P þ oral E2 (44). In that study, only PR expression in the luminal endometrium was significantly increased 3 days after retrieval in women receiving P alone as compared with the other groups. Although we and others (4, 26, 27) confirmed that the addition of E2 to P for luteal phase support in IVF cycles has no beneficial effect on PR, it is still unclear whether the effect of E2 supplementation is dependent on the dosage or route of administration. Earlier reports indicate that Fertility and Sterility
clinical PRs were consistently similar between groups with added 6 mg (11) or 2 mg (12) E2. However, one recent randomized trial demonstrated a significantly higher clinical PR in 6 mg E2 supplement group but not in 2 mg group compared with no E2 group (21). In the nine studies included in this meta-analysis, E2 and even P was supplemented by various routes of administration; therefore more sophisticated study would be needed to clarify the effect of E2 supplementation. The debate on E2 supplementation mainly stems from midluteal decline of serum E2 and its impact on endometrial receptivity. Based on the hypothesis that midluteal decline of serum E2 is deleterious to successful conception, several studies were conducted among selected patients showing midluteal drop of serum E2; they consistently reported a beneficial effect of midluteal E2 supplementation, although participating patients included those undergoing IUI as well as IVF-ET (14, 15). Despite the insignificant role proven by several meta-analyses of routine E2 supplementation during the entire luteal phase in IVF cycles, it remains to be determined whether the supplementation is beneficial to a specific group of IVF patients. The clinical impact of absolute midluteal level of serum E2 or E2 decline on success after ET is also still controversial 433
FIGURE 1 A forest plot demonstrates no benefit of estrogen supplement during luteal phase of IVF cycles using GnRH agonist (top) or antagonist (bottom) in terms of clinical pregnancy rate (PR) per patient. Random effect model was used, because the included studies had a substantial heterogeneity regarding dosage or route of medication and treatment duration. No publication bias was detected by the Begg-Mazumdar test in the studies using GnRH agonist. CI ¼ confidence interval; RR ¼ relative risk.
Jee. Estradiol as a luteal phase support. Fertil Steril 2010.
(45–51). One recent report noted that occurrence of conception did not correlate with the absolute midluteal level of serum E2 or with percent E2 decline (50), but other investigators reported that midluteal E2 level was significantly higher in a pregnant group (51). The profile of patients that may benefit from additional E2 supplementation should be further clarified, along with the optimal dose of E2, when it should start, and when it should end, as mentioned in a recent review (52). In conclusion, the combined data presented in the present meta-analysis suggest that supplementary administration of E2 to P for luteal phase support appears unnecessary in IVF cycles using either GnRH agonist or antagonist. However, the authors feel that there is an obvious need for further large-scale studies regarding GnRH antagonist cycles.
REFERENCES 1. Kolibianakis EM, Devroey P. The luteal phase after ovarian stimulation. Reprod Biomed Online 2002;5:26–35.
434
Jee et al.
Estradiol as a luteal phase support
2. Pabuccu R, Akar ME. Luteal phase support in assisted reproductive technology. Curr Opin Obstet Gynecol 2005;17:277–81. 3. Pritts EA, Atwood AK. Luteal phase support in infertility treatment: a meta-analysis of the randomized trials. Hum Reprod 2002;17:2287–99. 4. Daya S, Gunby J. Luteal phase support in assisted reproduction cycles. Cochrane Database Syst Rev 2004;3:CD004830. 5. Nosarka S, Kruger T, Siebert I, Grove D. Luteal phase support in in vitro fertilization: meta-analysis of randomized trials. Gynecol Obstet Invest 2005;60:67–74. 6. Ludwig M, Diedrich K. Evaluation of an optimal luteal phase support protocol in IVF. Acta Obstet Gynecol Scand 2001;80:452–66. 7. Levine H. Luteal support in IVF using the novel vaginal progesterone gel Crinone 8%: results of an open-label trial in 1184 women from 16 US centers. Fertil Steril 2000;74:836–7. 8. Penzias A. Luteal phase support. Fertil Steril 2002;77:318–23. 9. Fatemi HM, Popovic-Todorovic B, Papanikolaou E, Donoso P, Devroey P. An update of luteal phase support in stimulated IVF cycles. Hum Reprod Update 2007;13:581–90. 10. Smitz J, Devroey P, Braeckmans P, Camus M, Khan L, Staessen C, et al. Management of failed cycles in an IVF/GIFT programme with the combination of a GnRH analogue and hMG. Hum Reprod 1987;2:309–14. 11. Smitz J, Bourgain C, Van Waesberghe L, Camus M, Devroey P, Van Steirteghem AC. A prospective randomized study on oestradiol valerate supplementation in addition to intravaginal micronized progesterone in buserelin and hMG induced superovulation. Hum Reprod 1993;8:40–5.
Vol. 93, No. 2, January 15, 2010
12. Lewin A, Benshushan A, Mezker E, Yanai N, Schenker JG, Goshen R. The role of estrogen support during the luteal phase of in vitro fertilization–embryo transplant cycles: a comparative study between progesterone alone and estrogen and progesterone support. Fertil Steril 1994;62: 121–5. 13. Zegers-Hochschild F, Altieri E. Luteal estrogen is not required for the establishment of pregnancy in the human. J Assist Reprod Genet 1995;12: 224–8. 14. Gleicher N, Brown T, Dudkiewicz A, Karande V, Rao R, Balin M, et al. Estradiol/progesterone substitution in the luteal phase improves pregnancy rates in stimulated cycles—but only in younger women. Early Pregnancy 2000;4:64–73. 15. Kaider AS, Coulam CB. Luteal estrogen supplementation in pregnancies associated with low serum estradiol concentrations. Early Pregnancy 2000;4:191–9. 16. Farhi J, Weissman A, Steinfeld Z, Shorer M, Nahum H, Levran D. Estradiol supplementation during the luteal phase may improve the pregnancy rate in patients undergoing in vitro fertilization–embryo transfer cycles. Fertil Steril 2000;73:761–6. 17. Tay PY, Lenton EA. Inhibition of progesterone secretion by oestradiol administered in the luteal phase of assisted conception cycles. Med J Malaysia 2003;58:187–95. 18. Rashidi BH, Asheghan H, Jafarabadi M, Tehrani Nejad E. The role of estrogen supplementation during the luteal phase in in vitro fertilization–embryo transfer cycles: a comparative study between progesterone alone and estrogen plus progesterone support. Int Congress Ser 2004;1271:120–3. 19. Paredes Chavez FC, Barros Delgadillo JC, Ochoa Rueda SS, Barroso Villa G, Villalobos Acosta S, Sanchez Solıs V, et al. Estrogen role in the luteal phase support in in vitro fertilization with embryo transfer cycles. Ginecol Obstet Mex 2004;72:645–55. 20. Gorkemli H, Ak D, Akyurek C, Aktan M, Duman S. Comparison of pregnancy outcomes of progesterone or progesterone þ estradiol for luteal phase support in ICSI-ET cycles. Gynecol Obstet Invest 2004;58:140–4. 21. Lukaszuk K, Liss J, Lukaszuk M, Maj B. Optimization of estradiol supplementation during the luteal phase improves the pregnancy rate in women undergoing in vitro fertilization–embryo transfer cycles. Fertil Steril 2005;83:1372–6. 22. Drakakis P, Loutradis D, Vomvolaki E, Stefanidis K, Kiapekou E, Anagnostou E, et al. Luteal estrogen supplementation in stimulated cycles may improve the pregnancy rate in patients undergoing in vitro fertilization/intracytoplasmic sperm injection–embryo transfer. Gynecol Endocrinol 2007;23:645–52. 23. Engmann L, DiLuigi A, Schmidt D, Benadiva C, Maier D, Nulsen J. The effect of luteal phase vaginal estradiol supplementation on the success of in vitro fertilization treatment: a prospective randomized study. Fertil Steril 2008;89:554–61. 24. Serna J, Cholquevilque JL, Cela V, Martınez-Salazar J, Requena A, GarciaVelasco JA. Estradiol supplementation during the luteal phase of IVF-ICSI patients: a randomized, controlled trial. Fertil Steril 2008;90:2190–5. 25. Ceyhan ST, Basaran M, Kemal Duru N, Yilmaz A, G€oktolga U, Baser I. Use of luteal estrogen supplementation in normal responder patients treated with fixed multidose GnRH antagonist: a prospective randomized controlled study. Fertil Steril 2008;89:1827–30. 26. Gelbaya TA, Kyrgiou M, Tsoumpou I, Nardo LG. The use of estradiol for luteal phase support in invitro fertilization/intracytoplasmic sperm injection cycles: a systematic review and meta-analysis. Fertil Steril 2008;90:2116–25. 27. Kolibianakis EM, Venetis CA, Papanikolaou EG, Diedrich K, Tarlatzis BC, Griesinger G. Estrogen addition to progesterone for luteal phase support in cycles stimulated with GnRH analogues and gonadotrophins for IVF: a systematic review and meta-analysis. Hum Reprod 2008;23:1346–54. 28. Fatemi HM, Kolibianakis EM, Camus M, Tournaye H, Donoso P, Papanikolaou E, et al. Addition of estradiol to progesterone for luteal supplementation in patients stimulated with GnRH antagonist/rFSH for IVF: a randomized controlled trial. Hum Reprod 2006;21:2628–32. 29. Farhi J, Steinfeld Z, Shorer M, Nahum H, Weissman A, Glezerman M, et al. Estrogen supplementation during luteal phase improves pregnancy rate in IVF–ET. Fertil Steril 1999;72:S161.
Fertility and Sterility
30. Engmann L, Diluigi A, Schmidt D, Benadiva C, Maier D, Nulsen J. A prospective randomized trial evaluating the effect of luteal phase estradiol supplementation on the success of IVF treatment. Fertil Steril 2005;84:S88. 31. Fatemi HM, Kolibianakis EM, Camus M, Tournaye H, Van Steirteghem A, Devroey P. Progesterone versus progesterone combined with estradiol as luteal support in cycles stimulated with GnRH antagonist/rec-FSH for IVF: a randomized clinical trial. Fertil Steril 2005;84:S322. 32. Liss J, Lukaszuk K, Lukaszuk M, Maj B, Lukaszuk K. Increase of the pregnancy rate by the estradiol supplementation during the luteal phase in IVF-ET cycles. Hum Reprod 2005;20:i20. 33. Pouly JL, Piekrishvili R, Beaume-Brugon F, Schubert B, Janny L. Luteal phase supplementation with oestrogens does not improve the IVF pregnancy rate: a randomized study. Hum Reprod 2005;20:i72. 34. Karlikaya G, Karagozoglu H, Kumbak B, Lacin S, Guney A, Kahraman S. Adding estradiol to progesteron for luteal phase support does not improve the clinical outcome in assisted reproductive technologies (ART) cycles. Fertil Steril 2006;86:S112. 35. Serna J, Cholquevilque JL, Villasante A, Oriol B, Requena A, GarcıaVelasco J. Adding estradiol patches to the luteal phase of IVI/ICSI cycles did not improve pregnancy nor miscarriage rates. Fertil Steril 2006;86: S73. 36. Fauser BC, Devroey P. Reproductive biology and IVF: ovarian stimulation and luteal phase consequences. Trends Endocrinol Metab 2003;14: 236–42. 37. Beckers NG, Macklon NS, Eijkemans MJ, Ludwig M, Felberbaum RE, Diedrich K, et al. Nonsupplemented luteal phase characteristics after the administration of recombinant human chorionic gonadotropin, recombinant luteinizing hormone, or gonadotropin-releasing hormone (GnRH) agonist to induce final oocyte maturation in in vitro fertilization patients after ovarian stimulation with recombinant follicle stimulating hormone and GnRH antagonist cotreatment. J Clin Endocrinol Metab 2003;88: 4186–92. 38. Kolibianakis EM, Bourgain C, Platteau P, Albano C, Van Steirteghem AC, Devroey P. Abnormal endometrial development occurs during the luteal phase of nonsupplemented donor cycles treated with recombinant follicle-stimulating hormone and gonadotropin-releasing hormone antagonists. Fertil Steril 2003;80:464–6. 39. Tavaniotou A, Devroey P. Luteal hormone profile of oocyte donors stimulated with a GnRH antagonist compared with natural cycles. Reprod Biomed Online 2006;13:326–30. 40. Friedler S, Gilboa S, Schachter M, Raziel A, Strassburger D, Ron El R. Luteal phase characteristics following GnRH antagonist or agonist treatment—a comparative study. Reprod Biomed Online 2006;12:27–32. 41. De Ziegler D, Bergeron C, Cornel C, Medalie DA, Massai MR, Milgrom E, et al. Effects of luteal estradiol on the secretory transformation of human endometrium and plasma gonadotropins. J Clin Endocrinol Metab 1992;74:322–31. 42. Younis S, Ezra Y, Sherman Y, Simon A, Schencker G, Laufer N. The effect of estradiol depletion during the luteal phase on endometrium development. Fertil Steril 1994;62:103–7. 43. Fatemi HM, Camus M, Kolibianakis EM, Tournaye H, Papanikolaou EG, Donoso P, et al. The luteal phase of recombinant follicle-stimulating hormone/gonadotropin-releasing hormone antagonist in vitro fertilization cycles during supplementation with progesterone or progesterone and estradiol. Fertil Steril 2007;87:504–8. 44. Hubayter ZR, Vlahos NF, Lai TH, Garcia JA, Zhao Y. The impact of luteal phase support on endometrial estrogen and progesterone receptor expression: a prospective randomized study. Fertil Steril 2008;90: S65. 45. Sharara FI, McClamrock HD. Ratio of estradiol concentration on the day of human chorionic gonadotropin administration to mid-luteal oestradiol concentration is predictive of in-vitro fertilization outcome. Hum Reprod 1999;14:2777–82. 46. Ng EHY, Yeung WSB, Lau EYL, So WWK, Ho PC. A rapid decline in serum estradiol concentrations around the mid-luteal phase had no adverse effect on outcome in 763 assisted reproduction cycles. Hum Reprod 2000;15:1903–8.
435
47. Maximovich A, Stahler MS, Keye WR Jr, Brinton DA. In vitro fertilization (IVF) pregnancy rate and luteal phase estradiol level. Fertil Steril 2000;74:S156. 48. Ben-Ozer S, Eisenkop S, Vermesh M. Luteal phase estradiol (E2) and progesterone (P) profiles differ between conception and nonconception in-vitro fertilization–embryo transfer (IVF-ET) cycles utilizing pituitary down-regulation. Fertil Steril 2001;76:S252–3. 49. Akman MA, Erden HF, Bener F, Liu JE, Bahceci M. Can luteal phase estradiol levels predict the pregnancy outcome in in vitro fertilization cycles of good responders whose excess embryos yield blastocysts? Fertil Steril 2002;77:638–9.
436
Jee et al.
Estradiol as a luteal phase support
50. Friedler S, Zimerman A, Schachter M, Raziel A, Strassburger D, Ron El R. The midluteal decline in serum estradiol levels is drastic but not deleterious for implantation after in vitro fertilization and embryo transfer in patients with normal or high responses. Fertil Steril 2005;83:54–60. 51. Ganesh A, Goswami S, Chattopadhyay R, Chakraborty C, Chaudhury K, Chakravarty BN. Luteal phase estradiol level: a potential predictive marker for successful pregnancy in in vitro fertilization/intracytoplasmic sperm injection. Fertil Steril. Published online June 18, 2008 [Epub ahead of print]. 52. Hubayter ZR, Muasher SJ. Luteal supplementation in in vitro fertilization: more questions than answers. Fertil Steril 2008;89:749–58.
Vol. 93, No. 2, January 15, 2010
Objective: To clarify whether adding E2 to standard luteal P supplementation is beneficial both in GnRH agonist and antagonist IVF cycles. Design: Meta-analysis of nine randomized controlled trials. Setting: University hospital center for reproductive medicine and IVF. Intervention(s): None. Main Outcome Measure(s): Clinical pregnancy rate (PR) per patient, clinical PR per embryo transfer (ET), implantation rate, ongoing PR per patient, clinical abortion rate, and ectopic PR. Result(s): There were no statistically significant differences between E2þP versus P-only group regarding overall IVF outcomes. From seven studies including GnRH agonist cycles, no statistical significant differences were found between the two groups in clinical PR per patient (relative risk [RR] 1.32, 95% confidence interval [CI] 0.79–2.19), clinical PR per ET (RR 1.83, 95% CI 0.96–3.49), implantation rate (RR 1.20, 95% CI 0.34–4.21), ongoing PR per patient (RR 1.34, 95% CI 0.37–4.82), clinical abortion rate (RR 1.05, 95% CI 0.48–2.28), and ectopic PR (RR 0.53, 95% CI 0.07–4.10). Clinical PR per patient (RR 0.94, 95% CI 0.62–1.42) and ongoing PR per patient (RR 1.09, 95% CI 0.79–1.50) from three studies including GnRH antagonist cycles only were all similar between the two groups. Conclusion(s): The combined data presented in this meta-analysis suggest that the addition of E2 to P for luteal phase support does not improve IVF outcomes in GnRH agonist and antagonist cycles. However, the authors feel that there is an obvious need for further large-scale studies regarding GnRH antagonist cycles. (Fertil Steril 2010;93:428–36. 2010 by American Society for Reproductive Medicine.) Key Words: Estradiol, luteal phase, in vitro fertilization, GnRH agonist, GnRH antagonist
Luteal phase support has been a routine practice in IVF-ET because stimulated IVF cycles are associated with a defective luteal phase in almost all patients (1, 2). Three recent metaanalyses demonstrated that hCG is equally effective (3, 4) or superior to P (5) for luteal phase support with respect to clinical pregnancy rate (PR). Nonetheless, P is often favored, because hCG has a potential for increasing rates of ovarian hyperstimulation syndrome (4, 6). Various formulations of P are currently available, including oral, vaginal, rectal, and IM forms. Vaginal P gel and IM P were found to have similar clinical and ongoing PR (7). However, in a subsequent meta-analysis, clinical PR and delivery rate were significantly higher when IM P was used compared with vaginal application (3). Intramuscular Received December 15, 2008; revised January 29, 2009; accepted February 11, 2009; published onilne April 1, 2009. B.C.J. has nothing to disclose. C.S.S. has nothing to disclose. S.H.K. has nothing to disclose. Y.B.K. has nothing to disclose. S.Y.M. has nothing to disclose. Presented at the 64th Annual Meeting of the American Society for Reproductive Medicine, San Francisco, November 8–12, 2008. Reprint requests: Chang Suk Suh, M.D., Department of Obstetrics and Gynecology, Seoul National University Bundang Hospital, 300 Gumi, Bundang, Seongnam, Gyeonggi 463-707, South Korea (FAX: 82-31-787-4054; E-mail: [email protected]).
428
P has been widely used for luteal phase support but is often associated with a number of side effects, including painful injections with a rash, inflammatory reactions, and abscess formations (8). In these circumstances, vaginal administration of P can be a viable alternative. Moreover, intravaginal route of P supplementation is regarded as a first-choice luteal support regimen in stimulated IVF cycles (9). An earlier report indicated that serum E2 concentrations severely drop at the end of the luteal phase (10); therefore, a concern has been raised about an additional supply of E2 during the luteal phase in IVF cycles. In the first half of 1990, two prospective randomized studies were performed to evaluate the possible benefit of adding 6 mg (11) and 2 mg (12) E2 valerate daily in women treated with a GnRH agonist long protocol and gonadotropins for IVF. In those studies, the clinical PRs were almost equal between the groups with and without E2 cotreatment. During the luteal phase, ovarian E2 has experienced a sharp fall after its preovulatory peak and starts to rise again. The differences in luteal E2 in conception and nonconception cycles start to appear on day 9 to 10 with respect to the LH peak (13). Based on these observations, two studies were conducted in which patients with a precipitous drop of luteal phase serum E2 (14) or serum E2 concentration <100
Fertility and Sterility Vol. 93, No. 2, January 15, 2010 Copyright ª2010 American Society for Reproductive Medicine, Published by Elsevier Inc.
0015-0282/10/$36.00 doi:10.1016/j.fertnstert.2009.02.033
pg/mL at 11 days after ET (15) were selected. In those studies, oral E2 was additionally administered within 10 days after ovulation triggering (14) or 11 days after ET (15) and significantly higher clinical (14) or ongoing (15) PR per patient was observed. Farhi et al. (16) compared IVF outcomes with and without 2 mg E2 valerate as a luteal phase support in patients with serum E2 concentration at triggering day >2,500 pg/mL. In that study, E2 was started 11 days after ET and a significantly higher clinical PR per ET and implantation rate was noted in GnRH agonist long protocol cycles but not in short protocol. Although E2 was started several days later after ET and the patients were selected according to serum E2 concentration at specific points, three studies consistently demonstrated the beneficial effect of adding E2 as a luteal phase support. Pritts and Atwood (3) performed a meta-analysis addressing this issue based on three randomized trials (11, 12, 16). Although E2 doses and duration of treatment varied, they concluded that addition of oral E2 to standard P treatment improved implantation rate. However, that conclusion was derived from only one study (16); moreover, clinical PRs from three studies showed no difference between study and control group (relative risk [RR] 1.15, 95% confidence interval [CI] 0.83–1.58). Through a Cochrane review, Daya and Gunby (4) reported that there were no significant differences in clinical or ongoing PR, miscarriage, or live birth rate when P combined with E2 was compared to P alone; however, this conclusion was drawn only from two randomized trials (12, 17). The results from four subsequent studies were quite inconsistent and controversial. Almost equal clinical PR per patient was reported in two studies (18, 19), but significantly increased PR per patient and per ET by adding E2 was noted from other studies (20, 21). Interestingly, Lukaszuk et al. (21) reported a significantly higher clinical PR per patient in a 6 mg E2 supplement group, but not in a 2 mg E2 group, compared with no E2 (56.5% vs. 48.9% vs. 36.0%, respectively). Since 2007, four reports have been available; Drakakis et al. (22) reported a significantly higher clinical PR in an E2 supplement group in patients using GnRH agonist long protocol, but this same effect was not observed in subsequent studies (23, 24). In GnRH antagonist cycles, there were no differences in clinical and ongoing PR between E2 supplement vs. no supplement group (23, 25). Although two recent meta-analyses confirmed that the addition of E2 to P for luteal phase support in IVF/ICSI cycles has no beneficial effect on PR (26, 27), it is unclear whether the effect of E2 supplementation is different between GnRH agonist and antagonist cycles. No meta-analysis addressing this issue has been reported. A recent review (9) indicates that the addition of E2 seems to be beneficial in long GnRH agonist protocol but not in short GnRH agonist and GnRH antagonist protocol. The conclusion about long GnRH agonist protocol stemmed from Fertility and Sterility
four studies (11, 12, 16, 21) but they did not attempt to perform a meta-analysis. Moreover, the conclusion about GnRH antagonist protocol was based on only one study (28). In the present meta-analysis, we attempted to clarify the effect of E2 supplement in a luteal phase of stimulated IVF cycles. Furthermore, analysis of subgroups was performed to demonstrate whether the effect of E2 supplementation is different between GnRH agonist and antagonist cycles.
MATERIALS AND METHODS A literature search of the National Library of Medicine and the National Institutes of Health (PubMed), Medline and Cochrane Controlled Trials Register (CENTRAL; Cochrane Library) was performed using the key words ‘‘estradiol,’’ ‘‘estrogen,’’ ‘‘luteal phase support,’’ ‘‘luteal supplementation,’’ and ‘‘in vitro fertilization.’’ The last search date was October 2008. The inclusion criteria for selecting an article were defined as follows: 1. Fresh IVF-ET cycles using autologous oocyte. 2. The patients underwent ovarian hyperstimulation with pituitary suppression by GnRH agonist or antagonist; the data could be separable according to the method of pituitary suppression. 3. E2 starting on at least ET day, supplemented by oral, vaginal, or transdermal route. 4. P supplemented either vaginally or IM during an entire luteal phase. 5. Only prospective, randomized, controlled studies. 6. Published in English. We initially found 15 original articles comparing IVF-ET outcomes after administration of E2þP versus P only as a luteal supplementation. Finally, nine studies met the inclusion criteria and were reviewed in this meta-analysis (12, 17, 18, 20–23, 25, 28). All of the included studies were a prospective and randomized. We used the published data only. In all except two studies (20, 21), patients could enter the study only once. Cycle characteristics and IVF-ET outcomes from nine studies are summarized in Table 1. With one study including both GnRH agonist and antagonist cycle separately (23), IVF outcomes of GnRH agonist cycles (long protocol) could be extracted from seven studies and outcomes of GnRH antagonist cycles could be obtained from three studies. Six studies were excluded in this meta-analysis. In the study by Smitz et al. (11), 28% of the subjects underwent zygote intrafallopian transfer, which could not be separated from IVF-ET results. Estradiol valerate was administered from 4 days after retrieval. Gleicher et al. (14) recruited 330 consecutive ovarian stimulation cycles; however, most patients underwent intrauterine insemination (IUI), which could not be separated from IVF-ET results. In the study by Kaider and Coulam (15), oral E2 was started when serum E2 concentration was <100 pg/mL and serum hCG concentration >5 mIU/mL 11 days after ET. Farhi et al. (16) started oral E2 7 days after ET in patients with serum E2 429
430
TABLE 1 Summary of nine prospective randomized studies comparing IVF-ET outcomes after E2DP versus P only as a luteal supplementation.
Jee et al.
Authors, year (ref.)
Estradiol as a luteal phase support
Ceyhan et al., 2008 (25)
Patient characteristics Normal responders; exclude endometriosis, PCOS, severe male factor; mean age 31.4 vs. 30.9
Engmann et al., First IVF cycle; low and high responders excluded; 2008 (23),a mean age 34.9 vs. 35.7
Regimen and duration of luteal support (starting date ending date)
No. of patients
IVF outcomes (E2DP vs. P)
E2: transdermal E2 100 mg/d 2/wk (OPU 8 wks GA) P: vaginal MP 600 mg/d (OPU 8 wk GA)
Antagonist (fixed)
30 vs. 29
CPR/PA 43.3 vs. 44.8 CPR/ET 56.5 vs. 61.9 OG/PA 36.6 vs. 34.4
E2; vaginal ME (Estrace) 2 mg 2/d (ET fetal heartbeat) P: IM P 50 mg/d (OPU fetal heartbeat)
Long
55 vs. 59
CPR/PA 49.1 vs. 71.2c IR 26.9 vs. 39.6 CAR 7.4 vs. 11.9 OG/PA 45.5 vs. 62.7
Antagonist (flexible)
25 vs. 19
CPR/PA 48.0 vs. 52.6, IR 23.5 vs. 18.9 OG/PA 48.0 vs. 47.4 CAR 0.0 vs. 10.0 CPR/PA 43.6c vs. 15.8 IR 9.7 vs. 4.0 CAR 23.5 vs. 16.7 EP 5.6 vs. 0.0
Long E2: E2V 2 mg þ 0.5 mg norgestrel 3/d for 15 d and transdermal E2 50 mg q 4d (OPU pregnancy test) P: oral MP 100 mg 3 þ vaginal MP 200 mg 3 (ET-1 pregnancy test)
Vol. 93, No. 2, January 15, 2010
Drakakis et al., 2007 (22)
Not specified; mean age 35.4 vs. 35.8
Fatemi et al., 2006 (28)
Age %39, <3 prior cycles; E2: E2V 2 mg 2/d (OPUþ1 7 weeks GA) exclude PCOS, endometriosis III/IV; mean age: 32.1 vs. 32.0 P: vaginal MP 300 mg 2/d (same)
Jee. Estradiol as a luteal phase support. Fertil Steril 2010.
Protocol for suppression
Antagonist (fixed)
39 vs. 38
101 vs. 100 OG/PA 29.7 vs. 26.0 IR 42.4% vs. 37.8% Early pregnancy loss (including chemical, EP, and first trimester abortion) 23.1 vs. 23.4
Fertility and Sterility
TABLE 1 Continued. Authors, year (ref.)
Patient characteristics
Regimen and duration of luteal support (starting date ending date)
Protocol for suppression
No. of patients
IVF outcomes (E2DP vs. P) CPR/PA 56.5c vs. 48.9 vs. 36.0 CPR/ET 51.3c vs. 32.9 vs. 23.1 IR 29.9c vs. 17.8 vs. 9.8 CAR 12.8 vs. 17.4 vs. 22.2 EP 0.0 vs. 4.2 vs. 5.3 CPR/PA 62.9c vs. 18.9 CPR/ET 31.4c vs. 9.5 OG/PA 48.6c vs. 18.9 CAR 22.7 vs. 0.0
Lukaszuk et al., 2005 (21)
Age <40, mean age 31.1 vs. 31.7 vs. 32.1
E2: E2V 6 mg/d vs. 2 mg/d (OPU not specified) P: vaginal MP 200 mg 3/d (OPU not specified)
Long, pretreated by OC
69 vs. 47 vs. 50
Gorkemli et al., 2004 (20),b
Age %41, mean age 30.8 vs. 30.8
E2: transdermal E2 100 mg/d (OPU pregnancy test) P: vaginal MP 200 mg 3/d (OPU 10 wk GA)
Long
70 vs. 74
Rashidi et al., 2004 (18)
Not specified, mean age 29.6 vs. 30.0
Long
41 vs. 37
CPR/PA 31.7 vs. 35.1 CAR 17.6 vs. 7.7
Tay and Lenton, 2003 (17)
Basal FSH <10 mIU/mL, mean age 35 vs. 33
Long
28 vs. 35
CPR/PA 17.9 vs. 20.0
Lewin et al., 1994 (12)
Not specified, mean age 33.1 vs. 32.7
E2: E2V 4 mg/d (ET fetal heartbeat) P: IM P 100 mg/d (ET 12 wk GA) E2: E2V 2 mg/d (OPUþ1 not specified) P: vaginal MP 200 mg 2/d (OPUþ1 not specified) E2: E2V 2 mg/d (ET ETþ20) P: IM P 50 mg/d (same)
Long
50 vs. 50
CPR/PA 26.0 vs. 28.0 CPR/ET26.5 vs. 28.0 LBR 76.1 vs. 78.6
Note: CAR ¼ clinical abortion rate; CPR/ET ¼ clinical pregnancy rate per ET; CPR/PA ¼ clinical pregnancy rate per patient; E2V ¼ estradiol valerate; ET ¼ embryo transfer day; ETþ1 ¼ 1 day after ET; ET-1 ¼ 1 day before ET; EP ¼ ectopic pregnancy rate; IM ¼ intramuscular; IR ¼ implantation rate; LBR ¼ live birth rate; ME ¼ micronized estradiol; MP ¼ micronized progesterone; OC ¼ oral contraceptives; OG/PA ¼ ongoing pregnancy per patient; OPU ¼ ovum pick-up day. a Excluding eight women who were assigned to OCþmicrodose GnRH agonist group. b 16 mg/d methylprednisolone was given orally to all patients for 5 days after ET. c Significant increase compared with control. Jee. Estradiol as a luteal phase support. Fertil Steril 2010.
431
concentration at triggering day >2,500 pg/mL. One study written in Spanish was excluded (19). Serna et al. (24) performed an RCT, but the method of pituitary suppression could not be separated.
’’no difference’’ between study and control group (RR 1.48, 95% CI 0.99–2.23). When including data from the 6 mg E2 supplement group only, the RR still did not reach statistical significance (RR 1.28, 95% CI 0.83–1.99).
We also found seven studies published as abstracts (29–35) but all of them were excluded due to following reasons;
Clinical PR per ET in E2 groups was almost two times higher than in no E2 groups; however, the difference did not reach statistical significance. Implantation rates, ongoing PRs per patient, clinical abortion rates, and ectopic PRs were similar between study and control groups. Live birth rate was reported to be similar in both groups, but the rate was reported in only one study (12).
Estradiol was started on day 7 after retrieval in the study by Pouly et al. (33). The method of pituitary suppression could not be identified in the study by Karlikaya et al. (34). The remaining five were published later as full-length articles. In the present meta-analysis, primary outcome was defined as clinical PR per patient. Secondary outcomes were clinical PR per ET, implantation rate, ongoing PR per patient, clinical abortion rate, and ectopic PR. Clinical pregnancy was defined as seeing a gestational sac (GS) on transvaginal ultrasound. Biochemical pregnancy was completely ignored. The other parameters were defined as follows: Implantation rate: total number of GS/total number of embryos transfered. Ongoing pregnancy: clinical pregnancy developing beyond 12 weeks or the second trimester. Clinical abortion rate: number of first-trimester abortions/ number of clinical pregnancies. Ectopic PR: number of ectopic pregnancies/number of clinical and ectopic pregnancies A meta-analysis was performed using Review Manager (RevMan ver. 4.2 for Windows; Nordic Cochrane Center, Copenhagen, Denmark). Random effect model was always used, because the included studies had a substantial heterogeneity regarding dosage or route of the medication and treatment duration. A publication bias was assessed by the Begg-Mazumdar test regarding clinical PR per patient. Using the Begg-Mazumdar test, publication biases were not detected in both situations: Kendall tau ¼ 0.14 (P¼.72) in analysis including either GnRH agonist or antagonist cycles; Kendall tau ¼ 0.24 (P¼.56) in analysis including GnRH agonist cycles only. RESULTS First, IVF outcomes from nine studies including either GnRH agonist or antagonist cycles were assessed. Clinical PR per patient from eight studies was higher in E2 supplement groups compared to no E2 groups (45.4% vs. 35.0%); however, this difference was not statistically significant when using random effect model (RR 1.27, 95% CI 0.83–1.96; Table 2). Lukaszuk et al. (21) reported a significantly higher clinical PR per patient in a 6 mg E2 supplement group (RR 1.57, 95% CI 1.03–2.40) but not in a 2 mg E2 group (RR 1.36, 95% CI 0.85–2.18) compared with no E2 group. In the present metaanalysis, data from the 2 mg and 6 mg E2 supplement groups were combined. This combination of data resulted in overall 432
Jee et al.
Estradiol as a luteal phase support
Second, IVF outcomes from GnRH agonist cycles only were assessed separately. Again, clinical PR per patient (from seven studies) in E2 supplement groups was higher but not statistically different compared with no E2 groups (45.4% vs. 33.2%; RR 1.32, 95% CI 0.79–2.19; Fig. 1). There were no differences in clinical PR per ET, implantation rates, ongoing PRs per patient, clinical abortion rates, and ectopic PRs between study and control groups. A high RR (1.83) in clinical PR per ET was mainly attributed to two studies, in which a significantly higher clinical PR per ET was reported in E2 supplement groups: RR 3.32, 95% CI 1.91–5.79 (20); and RR 1.84, 95% CI 1.18–2.88 (21). The IVF outcomes of GnRH antagonist cycles could be extracted from three studies. Clinical PRs per patient from two studies and ongoing PRs per patient from three studies were all similar between study and control groups. DISCUSSION The results of the present meta-analysis indicate that the addition of E2 to P for luteal phase support does not improve IVF outcomes in GnRH agonist cycles. This finding is consistent with three earlier meta-analyses (4, 26, 27). In addition, we found that adding E2 to P for luteal phase support has no benefit of enhancing IVF outcomes in GnRH antagonist cycles. It is well established that luteal function is compromised in IVF cycles using GnRH agonist, and this phenomenon could potentially be attributed to ovarian stimulation and the resultant altered hormone levels, the process of oocyte retrieval, and direct effect of the GnRH agonist on the corpus luteum (36). Use of GnRH agonists may be associated with decreased production of E2 and P in the luteal phase, a reduction in the length of the luteal phase, and impairment of endogenous gonadotropin secretion caused by persistent pituitary suppression. Although GnRH antagonist causes less sustained pituitary suppression than GnRH agonist, several studies suggest that IVF cycles using GnRH antagonist protocol also need luteal phase support (37–39). However, adequate regimens for luteal phase support have not been precisely studied in GnRH antagonist cycles. Because the luteal phase characteristics appear to be similar between GnRH agonist and antagonist cycles (40), it could be assumed that supplementation regimen would be the same in both GnRH agonist and antagonist protocols. In this context, it is likely that the addition of E2 to P for luteal phase Vol. 93, No. 2, January 15, 2010
TABLE 2 Results of meta-analysis comparing IVF-ET outcomes after E2DP versus P only as a luteal supplementation. No. of included studies (ref.)
IVF-ET outcomes E2DP group
P group
RR (95% CI)
Including either GnRH agonist or antagonist cycles Clinical PR per patienta 8 (12, 17, 18, 20–23, 25) Clinical PR per ET 4 (12, 20, 21, 25) Implantation rate 2 (22, 23) Ongoing PR per patient 4 (20, 23, 25, 28) Clinical abortion rate 5 (18, 20–23) Ectopic PR 2 (21, 22)
45.4% (206/454) 36.9% (132/358) 18.3% (69/378) 39.9% (112/281) 16.0% (28/175) 2.5% (2/81)
35.0% (137/391) 19.9% (59/297) 20.5% (70/342) 34.2% (96/281) 11.7% (12/103) 4.0% (1/25)
1.27 (0.83–1.96)h 1.53 (0.84–2.78)h 1.26 (0.41–3.86)h 1.23 (0.73–2.08)h 1.02 (0.43–2.38) 0.53 (0.07–4.10)
Including GnRH agonist cycles only Clinical PR per patientb 7 (12, 17, 18, 20–23) Clinical PR per ET 3 (12, 20, 21) Implantation rate 2 (22, 23) Ongoing PR per patient 2 (20, 23) Clinical abortion rate 5 (18, 20–23) Ectopic PR 2 (21, 22)
45.4% (181/399) 35.5% (119/335) 17.1% (53/310) 47.2% (59/125) 17.2% (28/163) 2.5% (2/81)
33.2% (114/343) 16.7% (46/276) 20.8% (59/284) 38.3% (51/133) 11.8% (11/93) 4.0% (1/25)
1.32 (0.79–2.19)h 1.83 (0.96–3.49)h 1.20 (0.34–4.21)h 1.34 (0.37–4.82)h 1.05 (0.48–2.28) 0.53 (0.07–4.10)
Including GnRH antagonist cycles only Clinical PR per patient 2 (23, 25) Ongoing PR patient 3 (23, 25, 28)
45.5% (25/55) 34.0% (53/156)
47.9% (23/48) 30.4% (45/148)
0.94 (0.62–1.42) 1.09 (0.79–1.50)
Parameters
Note: Relative risk (RR) calculated by random effect model. CI ¼ confidence interval; other abbreviations as in Table 1. a Begg-Mazumdar test revealed that there was no publication bias (Kendall tau ¼ 0.14; P¼ .72). b Begg-Mazumdar test revealed that there was no publication bias (Kendall tau ¼ 0.24; P¼ .56). h There was a significant heterogeneity. Jee. Estradiol as a luteal phase support. Fertil Steril 2010.
support does not improve IVF outcomes in GnRH antagonist cycles, as similarly found in GnRH agonist cycles. In an earlier study, E2 and P receptors in the endometrium as well as endometrial histology did not differ in the luteal phase of ovariectomized women supplemented with or without exogenous E2 (41). A subsequent study suggests that luteal E2 is not fundamental in the preparation of the endometrium for implantation (13, 42). In a more recent trial, in which a GnRH antagonist was used, the endocrine profile (FSH, E2, and P) was similar in both groups receiving P alone vs. PþE2 during the luteal phase (43). In GnRH antagonist cycles, no significant differences were noted in estrogen receptor expression from endometrial biopsies obtained on the day of oocyte retrieval and 3, 5, and 10 days after retrieval among three study groups: no luteal support, vaginal micronized P, and micronized P þ oral E2 (44). In that study, only PR expression in the luminal endometrium was significantly increased 3 days after retrieval in women receiving P alone as compared with the other groups. Although we and others (4, 26, 27) confirmed that the addition of E2 to P for luteal phase support in IVF cycles has no beneficial effect on PR, it is still unclear whether the effect of E2 supplementation is dependent on the dosage or route of administration. Earlier reports indicate that Fertility and Sterility
clinical PRs were consistently similar between groups with added 6 mg (11) or 2 mg (12) E2. However, one recent randomized trial demonstrated a significantly higher clinical PR in 6 mg E2 supplement group but not in 2 mg group compared with no E2 group (21). In the nine studies included in this meta-analysis, E2 and even P was supplemented by various routes of administration; therefore more sophisticated study would be needed to clarify the effect of E2 supplementation. The debate on E2 supplementation mainly stems from midluteal decline of serum E2 and its impact on endometrial receptivity. Based on the hypothesis that midluteal decline of serum E2 is deleterious to successful conception, several studies were conducted among selected patients showing midluteal drop of serum E2; they consistently reported a beneficial effect of midluteal E2 supplementation, although participating patients included those undergoing IUI as well as IVF-ET (14, 15). Despite the insignificant role proven by several meta-analyses of routine E2 supplementation during the entire luteal phase in IVF cycles, it remains to be determined whether the supplementation is beneficial to a specific group of IVF patients. The clinical impact of absolute midluteal level of serum E2 or E2 decline on success after ET is also still controversial 433
FIGURE 1 A forest plot demonstrates no benefit of estrogen supplement during luteal phase of IVF cycles using GnRH agonist (top) or antagonist (bottom) in terms of clinical pregnancy rate (PR) per patient. Random effect model was used, because the included studies had a substantial heterogeneity regarding dosage or route of medication and treatment duration. No publication bias was detected by the Begg-Mazumdar test in the studies using GnRH agonist. CI ¼ confidence interval; RR ¼ relative risk.
Jee. Estradiol as a luteal phase support. Fertil Steril 2010.
(45–51). One recent report noted that occurrence of conception did not correlate with the absolute midluteal level of serum E2 or with percent E2 decline (50), but other investigators reported that midluteal E2 level was significantly higher in a pregnant group (51). The profile of patients that may benefit from additional E2 supplementation should be further clarified, along with the optimal dose of E2, when it should start, and when it should end, as mentioned in a recent review (52). In conclusion, the combined data presented in the present meta-analysis suggest that supplementary administration of E2 to P for luteal phase support appears unnecessary in IVF cycles using either GnRH agonist or antagonist. However, the authors feel that there is an obvious need for further large-scale studies regarding GnRH antagonist cycles.
REFERENCES 1. Kolibianakis EM, Devroey P. The luteal phase after ovarian stimulation. Reprod Biomed Online 2002;5:26–35.
434
Jee et al.
Estradiol as a luteal phase support
2. Pabuccu R, Akar ME. Luteal phase support in assisted reproductive technology. Curr Opin Obstet Gynecol 2005;17:277–81. 3. Pritts EA, Atwood AK. Luteal phase support in infertility treatment: a meta-analysis of the randomized trials. Hum Reprod 2002;17:2287–99. 4. Daya S, Gunby J. Luteal phase support in assisted reproduction cycles. Cochrane Database Syst Rev 2004;3:CD004830. 5. Nosarka S, Kruger T, Siebert I, Grove D. Luteal phase support in in vitro fertilization: meta-analysis of randomized trials. Gynecol Obstet Invest 2005;60:67–74. 6. Ludwig M, Diedrich K. Evaluation of an optimal luteal phase support protocol in IVF. Acta Obstet Gynecol Scand 2001;80:452–66. 7. Levine H. Luteal support in IVF using the novel vaginal progesterone gel Crinone 8%: results of an open-label trial in 1184 women from 16 US centers. Fertil Steril 2000;74:836–7. 8. Penzias A. Luteal phase support. Fertil Steril 2002;77:318–23. 9. Fatemi HM, Popovic-Todorovic B, Papanikolaou E, Donoso P, Devroey P. An update of luteal phase support in stimulated IVF cycles. Hum Reprod Update 2007;13:581–90. 10. Smitz J, Devroey P, Braeckmans P, Camus M, Khan L, Staessen C, et al. Management of failed cycles in an IVF/GIFT programme with the combination of a GnRH analogue and hMG. Hum Reprod 1987;2:309–14. 11. Smitz J, Bourgain C, Van Waesberghe L, Camus M, Devroey P, Van Steirteghem AC. A prospective randomized study on oestradiol valerate supplementation in addition to intravaginal micronized progesterone in buserelin and hMG induced superovulation. Hum Reprod 1993;8:40–5.
Vol. 93, No. 2, January 15, 2010
12. Lewin A, Benshushan A, Mezker E, Yanai N, Schenker JG, Goshen R. The role of estrogen support during the luteal phase of in vitro fertilization–embryo transplant cycles: a comparative study between progesterone alone and estrogen and progesterone support. Fertil Steril 1994;62: 121–5. 13. Zegers-Hochschild F, Altieri E. Luteal estrogen is not required for the establishment of pregnancy in the human. J Assist Reprod Genet 1995;12: 224–8. 14. Gleicher N, Brown T, Dudkiewicz A, Karande V, Rao R, Balin M, et al. Estradiol/progesterone substitution in the luteal phase improves pregnancy rates in stimulated cycles—but only in younger women. Early Pregnancy 2000;4:64–73. 15. Kaider AS, Coulam CB. Luteal estrogen supplementation in pregnancies associated with low serum estradiol concentrations. Early Pregnancy 2000;4:191–9. 16. Farhi J, Weissman A, Steinfeld Z, Shorer M, Nahum H, Levran D. Estradiol supplementation during the luteal phase may improve the pregnancy rate in patients undergoing in vitro fertilization–embryo transfer cycles. Fertil Steril 2000;73:761–6. 17. Tay PY, Lenton EA. Inhibition of progesterone secretion by oestradiol administered in the luteal phase of assisted conception cycles. Med J Malaysia 2003;58:187–95. 18. Rashidi BH, Asheghan H, Jafarabadi M, Tehrani Nejad E. The role of estrogen supplementation during the luteal phase in in vitro fertilization–embryo transfer cycles: a comparative study between progesterone alone and estrogen plus progesterone support. Int Congress Ser 2004;1271:120–3. 19. Paredes Chavez FC, Barros Delgadillo JC, Ochoa Rueda SS, Barroso Villa G, Villalobos Acosta S, Sanchez Solıs V, et al. Estrogen role in the luteal phase support in in vitro fertilization with embryo transfer cycles. Ginecol Obstet Mex 2004;72:645–55. 20. Gorkemli H, Ak D, Akyurek C, Aktan M, Duman S. Comparison of pregnancy outcomes of progesterone or progesterone þ estradiol for luteal phase support in ICSI-ET cycles. Gynecol Obstet Invest 2004;58:140–4. 21. Lukaszuk K, Liss J, Lukaszuk M, Maj B. Optimization of estradiol supplementation during the luteal phase improves the pregnancy rate in women undergoing in vitro fertilization–embryo transfer cycles. Fertil Steril 2005;83:1372–6. 22. Drakakis P, Loutradis D, Vomvolaki E, Stefanidis K, Kiapekou E, Anagnostou E, et al. Luteal estrogen supplementation in stimulated cycles may improve the pregnancy rate in patients undergoing in vitro fertilization/intracytoplasmic sperm injection–embryo transfer. Gynecol Endocrinol 2007;23:645–52. 23. Engmann L, DiLuigi A, Schmidt D, Benadiva C, Maier D, Nulsen J. The effect of luteal phase vaginal estradiol supplementation on the success of in vitro fertilization treatment: a prospective randomized study. Fertil Steril 2008;89:554–61. 24. Serna J, Cholquevilque JL, Cela V, Martınez-Salazar J, Requena A, GarciaVelasco JA. Estradiol supplementation during the luteal phase of IVF-ICSI patients: a randomized, controlled trial. Fertil Steril 2008;90:2190–5. 25. Ceyhan ST, Basaran M, Kemal Duru N, Yilmaz A, G€oktolga U, Baser I. Use of luteal estrogen supplementation in normal responder patients treated with fixed multidose GnRH antagonist: a prospective randomized controlled study. Fertil Steril 2008;89:1827–30. 26. Gelbaya TA, Kyrgiou M, Tsoumpou I, Nardo LG. The use of estradiol for luteal phase support in invitro fertilization/intracytoplasmic sperm injection cycles: a systematic review and meta-analysis. Fertil Steril 2008;90:2116–25. 27. Kolibianakis EM, Venetis CA, Papanikolaou EG, Diedrich K, Tarlatzis BC, Griesinger G. Estrogen addition to progesterone for luteal phase support in cycles stimulated with GnRH analogues and gonadotrophins for IVF: a systematic review and meta-analysis. Hum Reprod 2008;23:1346–54. 28. Fatemi HM, Kolibianakis EM, Camus M, Tournaye H, Donoso P, Papanikolaou E, et al. Addition of estradiol to progesterone for luteal supplementation in patients stimulated with GnRH antagonist/rFSH for IVF: a randomized controlled trial. Hum Reprod 2006;21:2628–32. 29. Farhi J, Steinfeld Z, Shorer M, Nahum H, Weissman A, Glezerman M, et al. Estrogen supplementation during luteal phase improves pregnancy rate in IVF–ET. Fertil Steril 1999;72:S161.
Fertility and Sterility
30. Engmann L, Diluigi A, Schmidt D, Benadiva C, Maier D, Nulsen J. A prospective randomized trial evaluating the effect of luteal phase estradiol supplementation on the success of IVF treatment. Fertil Steril 2005;84:S88. 31. Fatemi HM, Kolibianakis EM, Camus M, Tournaye H, Van Steirteghem A, Devroey P. Progesterone versus progesterone combined with estradiol as luteal support in cycles stimulated with GnRH antagonist/rec-FSH for IVF: a randomized clinical trial. Fertil Steril 2005;84:S322. 32. Liss J, Lukaszuk K, Lukaszuk M, Maj B, Lukaszuk K. Increase of the pregnancy rate by the estradiol supplementation during the luteal phase in IVF-ET cycles. Hum Reprod 2005;20:i20. 33. Pouly JL, Piekrishvili R, Beaume-Brugon F, Schubert B, Janny L. Luteal phase supplementation with oestrogens does not improve the IVF pregnancy rate: a randomized study. Hum Reprod 2005;20:i72. 34. Karlikaya G, Karagozoglu H, Kumbak B, Lacin S, Guney A, Kahraman S. Adding estradiol to progesteron for luteal phase support does not improve the clinical outcome in assisted reproductive technologies (ART) cycles. Fertil Steril 2006;86:S112. 35. Serna J, Cholquevilque JL, Villasante A, Oriol B, Requena A, GarcıaVelasco J. Adding estradiol patches to the luteal phase of IVI/ICSI cycles did not improve pregnancy nor miscarriage rates. Fertil Steril 2006;86: S73. 36. Fauser BC, Devroey P. Reproductive biology and IVF: ovarian stimulation and luteal phase consequences. Trends Endocrinol Metab 2003;14: 236–42. 37. Beckers NG, Macklon NS, Eijkemans MJ, Ludwig M, Felberbaum RE, Diedrich K, et al. Nonsupplemented luteal phase characteristics after the administration of recombinant human chorionic gonadotropin, recombinant luteinizing hormone, or gonadotropin-releasing hormone (GnRH) agonist to induce final oocyte maturation in in vitro fertilization patients after ovarian stimulation with recombinant follicle stimulating hormone and GnRH antagonist cotreatment. J Clin Endocrinol Metab 2003;88: 4186–92. 38. Kolibianakis EM, Bourgain C, Platteau P, Albano C, Van Steirteghem AC, Devroey P. Abnormal endometrial development occurs during the luteal phase of nonsupplemented donor cycles treated with recombinant follicle-stimulating hormone and gonadotropin-releasing hormone antagonists. Fertil Steril 2003;80:464–6. 39. Tavaniotou A, Devroey P. Luteal hormone profile of oocyte donors stimulated with a GnRH antagonist compared with natural cycles. Reprod Biomed Online 2006;13:326–30. 40. Friedler S, Gilboa S, Schachter M, Raziel A, Strassburger D, Ron El R. Luteal phase characteristics following GnRH antagonist or agonist treatment—a comparative study. Reprod Biomed Online 2006;12:27–32. 41. De Ziegler D, Bergeron C, Cornel C, Medalie DA, Massai MR, Milgrom E, et al. Effects of luteal estradiol on the secretory transformation of human endometrium and plasma gonadotropins. J Clin Endocrinol Metab 1992;74:322–31. 42. Younis S, Ezra Y, Sherman Y, Simon A, Schencker G, Laufer N. The effect of estradiol depletion during the luteal phase on endometrium development. Fertil Steril 1994;62:103–7. 43. Fatemi HM, Camus M, Kolibianakis EM, Tournaye H, Papanikolaou EG, Donoso P, et al. The luteal phase of recombinant follicle-stimulating hormone/gonadotropin-releasing hormone antagonist in vitro fertilization cycles during supplementation with progesterone or progesterone and estradiol. Fertil Steril 2007;87:504–8. 44. Hubayter ZR, Vlahos NF, Lai TH, Garcia JA, Zhao Y. The impact of luteal phase support on endometrial estrogen and progesterone receptor expression: a prospective randomized study. Fertil Steril 2008;90: S65. 45. Sharara FI, McClamrock HD. Ratio of estradiol concentration on the day of human chorionic gonadotropin administration to mid-luteal oestradiol concentration is predictive of in-vitro fertilization outcome. Hum Reprod 1999;14:2777–82. 46. Ng EHY, Yeung WSB, Lau EYL, So WWK, Ho PC. A rapid decline in serum estradiol concentrations around the mid-luteal phase had no adverse effect on outcome in 763 assisted reproduction cycles. Hum Reprod 2000;15:1903–8.
435
47. Maximovich A, Stahler MS, Keye WR Jr, Brinton DA. In vitro fertilization (IVF) pregnancy rate and luteal phase estradiol level. Fertil Steril 2000;74:S156. 48. Ben-Ozer S, Eisenkop S, Vermesh M. Luteal phase estradiol (E2) and progesterone (P) profiles differ between conception and nonconception in-vitro fertilization–embryo transfer (IVF-ET) cycles utilizing pituitary down-regulation. Fertil Steril 2001;76:S252–3. 49. Akman MA, Erden HF, Bener F, Liu JE, Bahceci M. Can luteal phase estradiol levels predict the pregnancy outcome in in vitro fertilization cycles of good responders whose excess embryos yield blastocysts? Fertil Steril 2002;77:638–9.
436
Jee et al.
Estradiol as a luteal phase support
50. Friedler S, Zimerman A, Schachter M, Raziel A, Strassburger D, Ron El R. The midluteal decline in serum estradiol levels is drastic but not deleterious for implantation after in vitro fertilization and embryo transfer in patients with normal or high responses. Fertil Steril 2005;83:54–60. 51. Ganesh A, Goswami S, Chattopadhyay R, Chakraborty C, Chaudhury K, Chakravarty BN. Luteal phase estradiol level: a potential predictive marker for successful pregnancy in in vitro fertilization/intracytoplasmic sperm injection. Fertil Steril. Published online June 18, 2008 [Epub ahead of print]. 52. Hubayter ZR, Muasher SJ. Luteal supplementation in in vitro fertilization: more questions than answers. Fertil Steril 2008;89:749–58.
Vol. 93, No. 2, January 15, 2010