Progesterone level and progesterone/estradiol ratio on the day of hcg administration: detrimental cutoff levels and new treatment strategy

Progesterone level and progesterone/estradiol ratio onthe day of hCG administration: detrimental cutofflevels and new treatment strategy Eman A. Elgindy, M.D.
Department of Obstetrics and Gynecology, Zagazig University School of Medicine; and Al-Banoon Fertility Center, Zagazig,Egypt Objective: To identify if there are certain cutoff levels for P and or the P/E2 ratio on the day of hCG that would bedefined as detrimental for occurrence of pregnancy in women with normal ovarian reserve undergoing cleavage-stage embryo transfer (ET). Secondarily, to determine if these same cutoffs might have the same potentialnegative effect in women undergoing blastocyst ET.
Design: Prospective cohort study including two randomized cohorts.
Setting: Private and university fertility centers.
Participant(s): A total of 240 women undergoing long agonist protocol with at least four grade 1 day 3 embryos.
Intervention(s): Women were randomized in a 1:1 ratio to undergo day 3 or day 5 embryo transfer.
Main Outcome Measure(s): Clinical pregnancy rate (CPR) was the primary outcome.
Result(s): Using receiver operator characteristics, cutoffs for P and P/E2 ratio were 1.5 ng/mL and 0.55, respec-tively. Patients with P %1.5 ng/mL and P/E % 0.55 undergoing cleavage-stage ET had higher CPR. Using multiple regression, P/E2 ratio was the only independent predictor for pregnancy. The P and P/E2 cutoffs werenot correlated with CPR in blastocyst transfers.
Conclusion(s): Progesterone >1.5 ng/mL and P/E > 0.55 affect the CPR in women undergoing cleavage-stage, but not blastocyst ET. P/E2 ratio is the only independent prognosticator for cycle outcome in women undergoingcleavage-stage ET. (Fertil Steril
2011;95:1639–44. 2011 by American Society for Reproductive Medicine.) Key Words: Premature luteinization, intracytoplasmic sperm injection, GnRH agonist, clinical pregnancy rate The introduction of GnRH analogues in ovulation induction has review are confounded by the different GnRH analogue protocols significantly decreased the incidence of premature LH surges administered. Moreover, the majority of studies that failed to dem- However, despite pituitary down-regulation, several researchers onstrate the negative association used an arbitrarily defined thresh- have described a phenomenon reported as premature luteinization old value of 0.9 ng/mL. The investigators further emphasized that (PL) It has been described as a subtle premature increase the use of receiver operator characteristic (ROC) curve analysis might be a more appropriate method to analyze any possible det- administration above a threshold level, which varied among rimental effect of P.
different studies and was usually arbitrarily defined It has Importantly, it has been reported that, in the presence of an ade- been suggested that elevated P levels reflect the total amount of quate response to controlled ovarian stimulation (COS), elevated P P secreted by maturing follicles, and these levels have been levels were not associated with a lower CPR, but when the response found to correlate positively with the number of mature follicles to COS was weak, premature P elevation led to lower CPR and with E2 levels on hCG day So, it has been Therefore, whether the unfavorable pregnancy outcome is caused suggested that P/E2 ratio more accurately reflects PL than by low ovarian reserve or high P levels can be studied more clearly a single hormone level. Even so, different cutoff levels have by excluding women with poor ovarian reserve. To our knowledge, been suggested among different populations with different investigations regarding the effect of elevation of both P and P/E2 on ovarian responses .
the day of hCG on pregnancy outcomes in women with normal ovar- The question of whether the presence of these increased serum P ian reserve are lacking.
levels or P/E2 ratios on the day of hCG administration have any Moreover, all of the embryo transfers (ETs) in earlier studies had detrimental effects on embryo implantation is a subject of much been performed on the second or third day of embryo culture. The debate . A recent meta-analysis suggested that the best available evidence suggests that CPR is significantly higher af- increase in P levels does not correlate with cycle outcome in terms ter blastocyst-stage ET (BET) compared with cleavage-stage ET of the clinical pregnancy rate (CPR) However, results of that (CET). Importantly, this should be applied in patients with highnumbers of 8-cell embryos on day 3 Many studies have shown that supraphysiologic levels of steroid Received September 20, 2010; revised December 6, 2010; accepted hormones alter endometrial P/E2 ratios with subsequent impaired December 8, 2010; published online February 16, 2011.
endometrial receptivity . On the other hand, Bourgain and E.A.E. has nothing to disclose.
Devroey demonstrated the absence of endometrial advanced Reprint requests: Eman A. Elgindy, M.D., Department of Obstetrics and maturation in the midluteal phase of stimulated cycles. Therefore, Gynecology, Zagazig University School of Medicine, Zagazig, Egypt(E-mail: ).
we theorized that even if there is a certain detrimental cutoff level Fertility and Sterility Vol. 95, No. 5, April 2011 Copyright ª2011 American Society for Reproductive Medicine, Published by Elsevier Inc.
for P and/or P/E2 ratio in women undergoing CET, it might not affect the CPR among women undergoing BET. To test this theory, a studywould need two comparable (e.g., randomized) cohorts undergoing Accuracy of P and P/E2 cutoffs for day 3 embryo transfers.
either day 3 or day 5 ETs.
The objective of the present study was to identify if there are cer- tain cutoff levels for P and or the P/E2 ratio on the day of hCG that would be defined as detrimental for occurrence of pregnancy in women with normal ovarian reserve undergoing CET. As a second- Positive predictive value ary objective, we wished to determine if these same cutoffs might Negative predictive value have the same potential negative effect in women undergoing BET.
Positive likelihood ratio MATERIALS AND METHODS Negative likelihood ratio 0.67 (0.58–0.77) 0.66 (0.56–0.76) This study was conducted from August 2008 to June 2010 in private and uni-versity IVF units after approval of the Institutional Review Board at Zagazig Elgindy. P and P/E ratio on day of hCG. Fertil Steril 2011.
University. The inclusion criteria were age %35 years, regular cycles (24–35days), day 3 FSH <9.5 IU/L, and antral follicle count >6. All participantswere also required to have an endometrial thickness R8 mm on hCG dayand at least four grade 1 embryos (i.e., regular symmetric blastomeres with Statistical Analysis no fragmentation) on day 3 after retrieval . An informed consent was Data were statistically described as mean  SD or frequencies and percent- obtained from each of the couples.
ages. Comparison of quantitative variables was done using Student t test forindependent samples. For comparing categoric data, c2 test was performed, Stimulation Protocol except when the expected frequency of events was less than five, in whichcase the Fisher exact test was used.
The long luteal down-regulation protocol was used, as described previously ROC analysis was conducted to search for the most efficient cutoff When at least three follicles were R17 mm, 10,000 IU hCG (Chorio- values for P and P/E2 which could discriminate between successful and mon; Ibsa, Switzerland) was administered. Oocytes were retrieved
35 unsuccessful ICSI outcomes in women undergoing day 3 ET. The best hours after hCG. Intracytoplasmic sperm injection (ICSI) was performed cutoff values were determined based on an equivalent sensitivity and spec- in a standard way. The fertilized oocytes were observed for cleavage on ificity and the highest value of the area under the ROC curve (AUC). Uni- variate and multivariate analysis models were also used to test for thepreferential effect of all independent variables on CPR in each cohort.
A probability value (P value) of < .05 was considered to be statistically On day 3, participants who had at least four grade 1 embryos were random- ized into two groups (A and B) by using block randomization. Allocationconcealment was performed using 240 identical dark sealed envelopes, pre- pared by a statistician and kept in the unit pharmacy. When a woman was A total of 550 women were potentially eligible for recruitment. 240 eligible and agreed to participate, she was instructed to select one envelopeonly once to determine the group to which she was assigned. The randomi- did not meet inclusion criteria, 50 refused to participate, and 20 were zation key was kept with the pharmacy director and was not opened until excluded for other reasons. Therefore, on day 3 after retrieval, 240 after statistical analysis.
women were randomized.
For CET, the optimal cutoff values for P and P/E2 ratio were >1.5 ng/mL and >0.55, respectively, for not achieving pregnancy ). Day 3 and day 5 ET patients were classified Participants in group A (n ¼ 120) underwent CET on day 3, and participantsin group B (n ¼ 120) underwent BET on day 5. In both groups, serum hCG according to the defined cutoff values. In both groups, no tests were performed 2 weeks after ET, and transvaginal ultrasound was differences were found between patients with P >1.5 or P/E2 scheduled 3 weeks later to confirm pregnancy.
>0.55 and those with P %1.5 or P/E % 0.55 regarding the baseline characteristics (However, in both groups, women Hormone Measurements with P >1.5 ng/mL and P/E > 0.55 had higher E2 level, P/E2 ratio, On hCG day, serum LH, E and number of retrieved oocytes. Meanwhile, numbers of grade 2, and P levels were measured (Elecsys 2010; Roche, Germany). The analytical sensitivities and total precisions of all stud- 1 and transfered embryos were similar. For day 3 ET, women with ied hormones have been described previously . The P/E P %1.5 or P/E %0.55 had significantly higher CPR than those 2 ratio was calcu- lated as [P (ng/mL)  1,000]/E with higher values (57.8% vs. 24.3% [P¼.001] and 57.3% vs.
26.3% [P¼.002], respectively). In contrast, for day 5 ET there was no significant difference in CPR between patients with P %1.5 CPR was the primary outcome. E 0.55 and those with higher values (59.5% vs.
2 levels on hCG day, number of oocytes re- trieved, and fertilization rates were secondary outcomes.
50% [P¼.33], 58.6% vs. 47.6% [P¼.36], respectively).
For day 3 ET, pregnant women and those who didn't get pregnant Sample Size Calculation were balanced regarding baseline and ICSI cycle characteristics.
Still, pregnant patients had statistically significant lower P levels Sample size calculation was performed a priori using clinical pregnancy as the base outcome. Prior data indicated that CPR among women with and 2 ratios on hCG day (P¼.0001). Multivariate regression without PL was 26% and 54%, respectively Using uncorrected chi- analysis was performed with adjustment for age, duration of infertil- squared (c2) statistic (power 80%, a-error 0.05), 47 women were needed ity, basal FSH, basal antral follicle count, duration of stimulation, to be included in each cohort. Because rate of PL was reported to be gonadotropin dose, E2 on hCG day, and numbers of retrieved and
38% , a total of 120 women were included in each randomized cohort.
fertilized oocytes, grade 1 embryos, and transfered embryos. P/E2 P and P/E2 ratio on day of hCG Vol. 95, No. 5, April 2011

Receiver operating characteristic (ROC) curve for defining optimal Receiver operating characteristic (ROC) curve for defining optimal detrimental cutoff value for P on hCG day.
detrimental cutoff value for P/E2 ratio on hCG day.
Elgindy. P and P/E Elgindy. P and P/E 2 ratio on day of hCG. Fertil Steril 2011.
2 ratio on day of hCG. Fertil Steril 2011.
ratio was the only independent predictor for pregnancy (odds ratio patients, further studies are still required to determine whether the 0.062, 95% confidence interval 0.008–0.48; P¼.008).
same threshold level applies to more specific populations. In the For day 5 ET, pregnant women and those who didn't get pregnant present study, patients undergoing CET with P %1.5 ng/mL and were balanced regarding baseline and ICSI cycle characteristics. P P/E2 ratio %0.55 had significantly higher CPR than those with levels and P/E2 ratios were similar between them (P values .13 higher values.
and .06, respectively). Multivariate regression analysis revealed no Many investigators have questioned the value of using abso- independent predictor for pregnancy (P>.05).
lute P levels on hCG day to predict pregnancy outcomes. Instead,they have proposed the use of the P/E2 ratio, taking into account the total number of follicles developing after COS The results of the present study suggest that an increased P level of Women with higher P and P/E2 had higher E2 levels and larger >1.5 ng/mL and a P/E2 ratio >0.55, on hCG day, are associated with number of retrieved oocytes, as demonstrated in the present lower CPR in women with normal ovarian reserve using long ago- study and others . A P/E2 ratio >1 was suggested in nist protocol and undergoing CET. In contrast, neither of these cut- some studies as a definition of PL, and this ratio was offs has a detrimental effect on pregnancy outcome in women associated with poor pregnancy outcomes However, in undergoing BET.
those studies, patients with variable ovarian reserves were ROC analysis was used to define these detrimental cutoffs, be- included and P/E > 1 was associated with low ovarian reserve.
cause this method is able to identify optimal thresholds on the basis On the other hand, Lai et al. evaluated in a retrospective of which patients can be classified into pregnant or not pregnant with study the relation between P/E2 ratio and CPR in women with a certain probability. Importantly, these thresholds characterize the normal ovarian reserves by using ROC analysis. Optimal P/E2 specific population analyzed and the protocol of treatment used cutoff value was R1.2 for not achieving pregnancy, which had Progesterone >1.5 ng/mL and P/E > 0.55 on the day of hCG ad- a sensitivity of 75.0%, specificity of 32.0%, and AUC of 0.53.
ministration were identified as the most appropriate thresholds for Strangely, when they used this cutoff to form two groups, they not achieving pregnancy among women undergoing CET. These re- found no difference in CPR between them. Although patients sults are relatively in line with a recent study including 4,037 pa- with low ovarian reserve were excluded, it was apparent that tients, in which P levels >1.5 ng/mL were associated with lower women with higher ratios had significantly lower E2 level and CPR after IVF/ICSI cycles . However, that was a retrospective numbers of retrieved oocytes.
study in women with variable ovarian reserves using different In the present study, an effort was made to establish firm inclusion COS protocols. Moreover, although the large sample size in that criteria for women with normal ovarian reserve using the long ago- study would be expected to compensate for any misclassified nist protocol so that reliable conclusions could be drawn. The Fertility and Sterility Baseline characteristics, ovarian stimulation, and pregnancy outcomes in women with or without P and P/E2 ratio elevation on hCG day, mean ± SD.
Day 3 ET (n [ 120) Day 5 ET (n [ 120) Day 3 ET (n [ 120) Day 5 ET (n [ 120) P >1.5 ng/mL P %1.5 ng/mL P >1.5 ng/mL P %1.5 ng/mL P/E2 >0.55 P/E2 %0.55 Cause of infertility < .001 2.28  0.79 1.11  0.31 < .001 2.89  0.73 1.13  0.39 3105.3  972.6 2468.4  374.2 < .001 3622.2  995.4 2619.9  420.7 < .001 2788  682 2545  579 3037  582 2744  615 < .001 0.87  0.26 0.42  0.06 < .001 1.06  0.13 0.38  0.08 Oocytes retrieved < .001 15.6  6 Fertilized oocytes < .001 11.4  5.04 10.8  3.6 Embryos transferred Clinical pregnancy Note: 2PN ¼ two pronuclei; AFC ¼ antral follicle count; CI ¼ confidence interval; ET ¼ embryo transfer; Gn ¼ gonadotropin; RR ¼ relative risk.
Elgindy. P and P/E2 ratio on day of hCG. Fertil Steril 2011.
optimal cutoff value for the P/E2 ratio was >0.55 for not achieving Still, defective receptivity can not be exclusively the pregnancy, which had a sensitivity of 82.5%, specificity of 44.4%, proposed mechanism, because day 5 embryos had excellent and AUC of 0.661. Importantly, the P/E2 ratio was the only indepen- implantation. It could be that, high follicular P and E2 levels have dent predictor for pregnancy among women undergoing CET.
the potential to advance the endometrium without influencing the Therefore, it appears that using P alone to predict pregnancy out- embryos. Therefore, the replacement of day 3 embryos occurs in come is confounding and the influence of both E2 and P should be taken into consideration.
establishing an embryo-endometrium cross-dialogue and failure of For patients undergoing BET, there was no significant difference implantation On the other hand, on the fifth luteal day, the en- in CPR between patients with P %1.5 ng/mL or P/E % dometrium might have significantly recovered from these detrimen- those with higher values. Recently, Papanikolaou et al. reported tal effects. As a result, the embryo-endometrium cross-dialogue that P >1.5 ng/mL on hCG day has a detrimental effect on CPR in appears to be sufficiently established. In accordance with this expla- women undergoing CET but not BET. However, the P value was nation, Bourgain et al. reported that early luteal severe antipro- chosen arbitrarily, and the study only included women receiving liferative effects of the stimulation protocol were observed in both the GnRH antagonist protocol. Regarding P/E2 ratio, to the best of glandular and stromal cells compared with natural cycle controls.
our knowledge, there have been no previous studies investigating However, this difference was no longer present on later cycle the possible relation between P/E2 ratio and CPR in women under- days. So, it could be assumed that the detrimental effects of P and going BET. It appears that extending culture and transfering em- E2 on pregnancy outcome are attributed to temporally defective bryos on day 5 is a new treatment strategy for women with high P endometrial receptivity which recovers a few days later.
and P/E2 ratio on hCG day.
The potential weakness of the present study includes the inability Possible impaired endometrial receptivity and/or poor oocyte to blind either patients or physicians. Importantly, the study popula- quality were suggested for unfavorable effect of elevated P and tion included potentially good responders using the long agonist P/E2 on pregnancy outcome . However, the lack of protocol, which limits the generalizability of the study findings.
association between P or P/E2 ratio elevation and fertilizationrates or grade 1 embryos in the present study and others mightbe indicative of the absence of a detrimental effect on oocyte quality.
Regarding endometrial receptivity, other studies have shown that In conclusion, P >1.5 ng/mL and P/E > 0.55 affect the CPR in supraphysiologic levels of steroid hormones not only induce women undergoing CET but not BET. The P/E2 ratio is the only in- morphologic alterations in endometrium, but also alter endometrial dependent prognosticator for cycle outcome in women undergoing P/E2 ratios, with subsequent impaired endometrial receptivity 1. Smitz J, Ron-El R, Tarlatzis BC. The use of gonado- acetate-down-regulated GIFT cycles are associated to-estradiol ratio on the day of human chorionic go- trophin releasing hormone agonists for in vitro fertil- with decreased clinical pregnancy rates. J Assist Re- nadotropin administration does not have a negative ization and other assisted procreation techniques: prod Genet 1996;13:459–63.
impact on clinical pregnancy rate in women with nor- experience from three centres. Hum Reprod 1992;7 mal ovarian reserve treated with a long gonadotropin (Suppl 1):49–66.
Harada T, Terakawa N. Subtle rise in serum pro- releasing hormone agonist protocol. Fertil Steril 2. Hofmann GE, Bentzien F, Bergh PA, Garrisi GJ, gesterone during the follicular phase as a predictor Williams MC, Guzman I, et al. Premature luteiniza- of the outcome of in vitro fertilization. Fertil Steril 15. Venetis CA, Kolibianakis EM, Papanikolaou E, tion in controlled ovarian hyperstimulation has no ad- Bontis J, Devroey P, Tarlatzis BC. Is progesterone el- verse effect on oocyte and embryo quality. Fertil 9. Edelstein MC, Seltman HJ, Cox BJ, Robinson SM, evation on the day of human chorionic gonadotrophin Shaw RA, Muasher SJ. Progesterone levels on the administration associated with the probability of 3. Ubaldi F, Camus M, Smitz J, Bennink HC, van day of human chorionic gonadotropin administration pregnancy in in vitro fertilization? A systematic re- Steirteghem A, Devroey P. Premature luteinization in cycles with gonadotropin-releasing hormone ago- view and meta-analysis. Hum Reprod Update in in vitro fertilization cycles using gonadotropin- nist suppression are not predictive of pregnancy out- releasing hormone agonist (GnRH-a) and recombi- come. Fertil Steril 1990;54:853–7.
16. Fanchin R, Righini C, Olivennes F, Ferreira AL, de 10. Abuzeid MI, Sasy MA. Elevated progesterone levels Ziegler D, Frydman R. Consequences of premature in the late follicular phase do not predict success of progesterone elevation on the outcome of in vitro fer- in vitro fertilization–embryo transfer. Fertil Steril tilization: insights into a controversy. Fertil Steril 4. Bosch E, Valencia I, Escudero E, Crespo J, Simon C, Remohi J, et al. Premature luteinization during 11. Fanchin R, de Ziegler D, Castracane VD, Taieb J, 17. Papanikolaou EG, Kolibianakis EM, Tournaye H, gonadotropin-releasing hormone antagonist cycles Olivennes F, Frydman R. Physiopathology of prema- Venetis CA, Fatemi H, Tarlatzis B, et al. Live and its relationship with in vitro fertilization out- birth rates after transfer of equal number of blasto- come. Fertil Steril 2003;80:1444–9.
cysts or cleavage-stage embryos in IVF. A system- 5. Schoolcraft W, Sinton E, Schlenker T, Huynh D, 12. Younis JS, Matilsky M, Radin O, Ben-Ami M. In- Hamilton F, Meldrum DR. Lower pregnancy rate creased progesterone/estradiol ratio in the late follic- with premature luteinization during pituitary sup- ular phase could be related to low ovarian reserve in in 18. Blake DA, Farquhar CM, Johnson N, Proctor M.
vitro fertilization–embryo transfer cycles with a long Cleavage stage versus blastocyst stage embryo trans- gonadotropin-releasing hormone agonist. Fertil Steril fer in assisted conception.Cochrane Database Syst 6. Givens CR, Schriock ED, Dandekar PV, Martin MC.
Rev; 2007:CD002118.
Elevated serum progesterone levels on the day of hu- 13. Ozcakir HT, Levi R, Tavmergen E, Goker EN. Prema- man chorionic gonadotropin administration do not ture luteinization defined as progesterone estradiol ra- Porter RN, Pike IL, Saunders DM. Failure of im- predict outcome in assisted reproduction cycles.
tio >1 on hCG administration day seems to adversely plantation in human in vitro fertilization and em- Fertil Steril 1994;62:1011–7.
affect clinical outcome in long gonadotropin- 7. Randall GW, Gantt PA, Gantt D, Kirk MJ, releasing hormone agonist cycles. J Obstet Gynaecol progesterone/estrogen ratios in humans and mice.
Romines N. Elevated serum progesterone values at Fertil Steril 1986;45:69–74.
the time of ovulation induction in luteal leuprolide 14. Lai TH, Lee FK, Lin TK, Horng SG, Chen SC, Chen YH, et al. An increased serum progesterone- Fertility and Sterility 20. Hadi FH, Chantler E, Anderson E, Nicholson R, ongoing pregnancy rates in controlled ovarian stimu- 27. Ezra Y, Simon A, Sherman Y, Benshushan A, McClelland RA, Seif MW. Ovulation induction and lation cycles for in vitro fertilization: analysis of over Younis JS, Laufer N. The effect of progesterone ad- 4000 cycles. Hum Reprod 2010;25:2092–100.
ministration in the follicular phase of an artificial cy- 25. Papanikolaou EG, Kolibianakis EM, Pozzobon C, cle on endometrial morphology: a model of 21. Bourgain C, Devroey P. The endometrium in stimu- Tank P, Tournaye H, Bourgain C, et al. Progesterone premature luteinization. Fertil Steril 1994;62:108–12.
rise on the day of human chorionic gonadotropin ad- 28. Fanchin R, Righini C, Olivennes F, de Ziegler D, ministration impairs pregnancy outcome in day 3 Selva J, Frydman R. Premature progesterone eleva- 22. Racowsky C, Vernon M, Mayer J, Ball GD, Behr B, single-embryo transfer, while has no effect on day 5 sin- tion does not alter oocyte quality in in vitro fertiliza- Pomeroy KO, et al. Standardization of grading em- gle blastocyst transfer. Fertil Steril 2009;91:949–52.
tion. Fertil Steril 1996;65:1178–83.
bryo morphology. Fertil Steril 2009;94:1152–3.
26. Silverberg KM, Burns WN, Olive DL, Riehl RM, 29. Bourgain C, Ubaldi F, Tavaniotou A, Smitz J, Van 23. Elgindy EA, El-Haieg DO, Mostafa MI, Shafiek M.
Schenken RS. Serum progesterone levels predict suc- Steirteghem AC, Devroey P. Endometrial hormone Does luteal estradiol supplementation have a role in cess of in vitro fertilization/embryo transfer in pa- receptors and proliferation index in the periovulatory long agonist cycles? Fertil Steril 2010;93:2182–8.
tients stimulated with leuprolide acetate and human phase of stimulated embryo transfer cycles in com- 24. Bosch E, Labarta E, Crespo J, Simon C, Remohi J, menopausal gonadotropins. J Clin Endocrinol Metab parison with natural cycles and relation to clinical Jenkins J, et al. Circulating progesterone levels and pregnancy outcome. Fertil Steril 2002;78:237–44.
P and P/E2 ratio on day of hCG Vol. 95, No. 5, April 2011

Source: http://www.royanaward.com/files13/Elgindy%20P%20paper%20F%20AS.pdf