Background: It has recently been suggested that recombinant FSH administration may result in an increased risk of venous thrombosis. An open-label, randomized, controlled trial was carried out to compare the impact of urinary and recombinant FSH on haemostasis.
Methods: Fifty infertile women were randomized, using a random number generator on a personal computer, to receive either highly purified urinary FSH (u-hFSH) or recombinant human FSH (r-hFSH); a starting dose of 150 IU. Human chorionic gonadotrophin 10000 IU was administered once there was at least one follicle > or =18 mm. The luteal phase was supported with progesterone 50 mg/day for at least 15 days. Fifty normally menstruating women were recruited as controls. Repeated measurements of estradiol, progesterone, prothrombin time (PT) expressed as INR, activated partial thromboplastin time (APTT) ratio, fibrinogen (FBG), factor VIII (FVIII), normalized activated protein C ratio (nAPC ratio), antithrombin III activity (AT), protein C activity (PC), protein S activity (PS), tissue-type plasminogen activator antigen (t-PA), type 1 plasminogen activator inhibitor (PAI), prothrombin fragments 1+2 (F1+2), were performed during both hyperstimulated and natural cycles, and at onset of the following menstruation or at 8 weeks of pregnancy.
Results: At the end of gonadotrophin administration PT INR increased in the u-hFSH group, while AT and t-PA significantly decreased. In the patients treated with r-hFSH, only F1+2 significantly decreased. No significant changes were observed in the control group. In the luteal phase FBG increased significantly in all groups. In the u-hFSH group no other significant changes were noted compared to pre-ovulatory values, while compared to baseline values AT, PS and t-PA significantly decreased. In the r-hFSH group during the luteal phase PT INR significantly decreased, but did not differ from baseline levels. Other parameters such as FBG, FVIII, t-PA, rose significantly, but only FVIII and FBG values were significantly higher than baseline levels. In the women who became pregnant a significant increase in t-PA and a significant decrease in PS at the midluteal phase were observed. After one month all the haemostatic parameters returned to baseline value if pregnancy failed to occur, while in the pregnant women a significant increase in FVIII and a significant decrease in PS were observed.
Conclusions: Ovarian stimulation with recombinant FSH does not influence coagulation and fibrinolysis significantly, as already reported for urinary gonadotrophins. The moderate changes induced by both treatments are no longer detectable after 4 weeks.