Transdermal testosterone pretreatment in poor responders undergoing ICSI: a randomized clinical trial

Does pretreatment with transdermal testosterone increase the number of cumulus-oocyte complexes (COCs) retrieved by more than 1.5 in poor responders undergoing intracytoplasmic sperm injection (ICSI), using recombinant follicle stimulating hormone (FSH) and gonadotrophin releasing hormone agonists (GnRHa)?

Testosterone pretreatment failed to increase the number of COCs by more than 1.5 as compared with no pretreatment in poor responders undergoing ICSI (difference between medians: 0.0, 95% CI: -1.0 to +1.0).

Androgens are thought to play an important role in early follicular development by enhancing ovarian sensitivity to FSH. In a recent meta-analysis, testosterone pretreatment resulted in an increase of 1.5 COCs as compared with no pretreatment. However, this effect was based on the analysis of only two randomized controlled trials (RCTs) including 163 patients. Evidently, there is a need for additional RCTs that will allow firmer conclusions to be drawn.

The present RCT was designed to detect a difference of 1.5 COCs (sample size required = 48 patients). From 02/2014 until 04/2015, 50 poor responders fulfilling the Bologna criteria have been randomized (using a randomization list) to either testosterone pretreatment for 21 days ( ITALIC! n = 26) or no pretreatment ( ITALIC! n = 24).

All patients underwent a long follicular GnRHa protocol. Recombinant FSH stimulation was started on Day 22 following GnRHa initiation. In the testosterone pretreatment group, a daily dose of 10 mg of testosterone gel was applied transdermally for 21 days starting from GnRHa initiation. Results are expressed as median (interquartile range).

No differences in baseline characteristics were observed between the two groups compared. Testosterone levels [median (interquartile range)] were significantly higher in the testosterone pretreatment on the day of initiation of FSH stimulation [114 (99.5) ng/dl versus 20 (20) ng/dl, respectively, ITALIC! P < 0.001]. Duration of FSH stimulation [median (interquartile range)] was similar between the groups compared [12.5 (3.0) days versus 12 (3.0) days, respectively, ITALIC! P = 0.52]. The number of COCs retrieved [median (interquartile range)] was not different between the testosterone pretreatment and the no pretreatment groups [3.5 (4.0) versus 3.0 (3.0), 95% CI for the median: 2.0-5.0 versus 2.7-4.3, respectively; difference between medians: 0.0, 95% CI: +1.0 to -1.0). Similarly no differences were observed regarding fertilization rates [median (interquartile range)] [66.7% (32.5) versus 66.7% (42.9), respectively, ITALIC! P = 0.97] and live birth rates per randomized patient (7.7% versus 8.3%, respectively, rate difference: -0.6%, 95% CI: -19.0 to +16.9).

The study was not powered to detect differences less than 1.5 COCs, although it is doubtful whether these differences would be clinically relevant. Moreover, due to sample size restrictions, no conclusions can be drawn regarding the probability of live birth.

The results of this randomized clinical trial, suggesting that pretreatment with 10 mg of transdermal testosterone for 21 days does not improve ovarian response by more than 1.5 oocytes, could be used to more accurately consult patients with poor ovarian response. However, an improvement in IVF outcome using a higher dose of testosterone or a longer pretreatment period cannot be excluded.

The study was partially funded by a Scholarship from the Academy of Athens. C.A.V. reports personal fees and non-financial support from Merck, Sharp and Dome, personal fees and non-financial support from Merck Serono, personal fees and non-financial support from IPSEN Hellas S.A., outside the submitted work. B.C.T. reports grants from Merck Serono, grants from Merck Sharp & Dohme, personal fees from Merck Serono, personal fees from Merck Sharp & Dohme, personal fees from IBSA & Ferring, outside the submitted work.

NCT01961336.

10 October 2013.

02/2014.