I read with interest the article published in your journal by Bakas et.al. (1).
The authors said that in the GnRH antagonist group, when the leading follicle had reached a mean diameter of ≥ 16 mm, 0.25 mg ganirelix was started. Why did authors choose a mean diameter of 18 mm to start of antagonist? It can be argued that when GnRH antagonists were first administered, premature elevation in LH had already occurred. In actuality we could not measure serum progesterone and LH serially. Therefore, the intervention may have been too late. In Lamback’s study, when GnRH anatgonist started at a dominant follicle, premature luteinization was detected in 3.4% of the cycles (2). Furthermore, it has been reported that almost 50% of the cycles that showed a premature LH elevation had dominant follicles that were 16.5 mm or smaller (3).
In this study, there is no placebo arm. To exclude possible bias, a placebo should ideally be used. Only Lamback et al. used a placebo and found no significant effect of using GnRH antagonists (2).
The authors did not calculate power analysis. I think this trial was underpowered for assessing the effect on clinical pregnancy rates.
In this article, clinical pregnancy was defined as the presence of heartbeat on transvaginal ultrasound performed 2 weeks after hCG administration. It is clear that this time too early for detecting a heartbeat (4).
Sefa Kelekci, Prof. Dr.
Bezmialem Vakif University
School of Medicine
Department of Obstetrics and Gynecology
Fatih, Istanbul, Turkey
1. Bakas P, Konidaris S, Liapis A, Gregoriou O, TzanakakiD, Creatsas G. Role of
gonadotropin-releasing hormone antagonist in the management of subfertile couples with
intrauterine insemination. Fert Ster 2011; 95: 2024-7.
2. Lambalk CB, Leader A, Olivennes F, Fluker MR, Andersen AN, Ingerslev J, Khalaf Y, Avril C, Belaisch-Allart J, Roulier R et al. Treatment with the GnRH antagonist ganirelix prevents premature LH rises and luteinisation in stimulated intrauterine insemination: results of a double-blind, placebo-controlled, multicentre trial. Hum Reprod 2006; 21: 632–9.
3. Cunha-Filho JS, Kadoch J, Righini C, Fanchin R, Frydman R, Olivennes F. Premature LH and progesterone rise in intrauterine insemination cycles: analysis of related factors. Reprod Biomed Online 2003; 7: 194–9.
4. Neilson JP. Ultrasound for fetal assessment in early pregnancy. Cochrane Database System Review 2000;2:CD000182.
Published online in Fertility and Sterility doi:10.1016/j.fertnstert.2011.05.046
The Authors Respond
We would like to thank Dr. Kelekci for the interest in our paper and the questions submitted.
The GnRH antagonist was not started with a dominant follicle size of 18 mm as you mentioned, but at a size of >16 mm mean diameter. The symbol ‘>’ was used to denote that follicular size was not exactly 16 mm but it could be 16.1 mm,16.2mm or 16.3 mm , because it is extremely difficult for all patients to be at a dominant follicular size of 16 mm exactly at the time they start the GnRH antagonist administration.
Regarding the timing of LH measurement, in this study LH was measured the day of HCG administration. We do agree with your comment that premature LH secretion could happen in up to 50% of cycles with a size of the dominant follicle 16.5 mm or less (1). We could not perform serial measurement of LH during the stimulation cycle for practical reasons, and therefore it could be possible that the incidence of premature LH secretion has been underestimated. But even so, the present study shows that if the clinician is concerned about possible premature LH surge during a stimulation cycle for intrauterine insemination, the use of GnRH antagonist when the dominant follicle is >16mm will reduce significantly the incidence of premature LH surge in patients taking GnRH antagonist compared to patients using gonadotropins only.
On the other hand, in the study by Lamback et al.(2), despite a significant reduction seen in premature LH surge in patients taking gonadotropins and GnRH antagonist compared to patients taking the placebo, there was no significant difference in pregnancy rates.
Ideally, the use of placebo could improve the study design by excluding possible unrecognized bias, but it is very difficult to perform because it requires design and manufacturing of placebo medication in a manner similar to GnRH antagonist. But, it is unknown if the use of placebo group instead of control group has an impact on study results, and even if it does, then the use of a control group is necessary in order to see what is happening during stimulation only with gonadotropins for intrauterine insemination.
Power analysis was performed and the power of the study for a type I error of 0.05 was less than 80% for clinical pregnancy rates. The power of the study for the cumulative pregnancy rates was 90% with a type I error of 0.05.
Pregnancy was assessed with the performance of a b-HCG measurement 14 days after IUI. Clinical pregnancy rate was defined as the presence of heartbeat on transvaginal ultrasound performed 2 weeks after HCG measurement and no administration (6 weeks).
Panagiotis Bakas, M.D.
Odyseas Gregoriou, M.D.
Angelos Liapis, M.D.
Sokratis Konidaris, M.D.
Georgios Creatsas, M.D.
Second Department of Obstetrics and Gynecology
University of Athens
1. Cunha-Filho JS, Kadoch J, Righini C, Fanchin R, Frydman R, Olivennes F. PrematureLH and progesterone rise in intrauterine insemination cycles: analysis of related factors. Reprod Biomed Online 2003; 7: 194–9.
2. Lambalk CB, Leader A, Olivennes F, Fluker MR, Andersen AN, Ingerslev J, Khalaf Y,Avril C, Belaisch-Allart J, Roulier R et al. Treatment with the GnRH antagonist ganirelixprevents premature LH rises and luteinisation in stimulated intrauterine insemination:results of a double-blind, placebo-controlled, multicenter trial. Hum Reprod 2006; 21:632–9.
Published online in Fertility and Sterility doi:10.1016/j.fertnstert.2011.05.045