Intensive FSH stimulation is a key tool of assisted reproduction technology but can cause severe complications through the development of an excessive number of small ovarian follicles. We tested the hypothesis that, in the late stages of ovulation induction, LH activity in the form of low-dose human CG (hCG) can stimulate and selectively modulate ovarian follicle function and growth, independently of FSH administration. Four groups of GnRH agonist-suppressed normoovulatory women (10 each

more »

... received recombinant human FSH (r-hFSH) (150 IU/d) for 7 d followed by: group A, r-hFSH 150 IU/d alone; group B, r-hFSH 50 IU/d and hCG 50 IU/d; group C, r-hFSH 25 IU/d and hCG 100 IU/d; group D, hCG 200 IU/d alone. Despite several days of lowered or absent r-hFSH administration, 70% of hCG-treated patients successfully com-pleted treatment. In these subjects, preovulatory E2 levels and large (>14 mm diameter) ovarian follicle development were not reduced; conversely, the number of small (<10 mm diameter) ovarian follicles was significantly decreased in groups B-D vs. group A. Low-dose hCG administration did not cause follicle luteinization. We conclude that, following FSH priming, LH activity administration can: 1) stimulate folliculogenesis for several days, in spite of rapidly declining FSH levels; and 2) hasten small follicle demise. Therefore, LH activity administration could be used to design radically novel ovulation induction regimens that, by partly or completely replacing mid-/late follicular phase FSH administration, may reduce costs and improve safety of assisted reproduction technology. (J Clin Endocrinol Metab 87: 1156 -1161, 2002) C ONTROLLED OVARIAN STIMULATION (COS) is a critical component of assisted reproduction technology (ART), consisting of the administration of pharmacological doses of exogenous gonadotropins to achieve the development of multiple ovarian follicles and oocytes. Gonadotropin preparations used in COS contain FSH alone or in combination with variable amounts of LH activity (1). While FSH is considered the fundamental driver of folliculogenesis, the role of LH in this process is more controversial (2, 3). During the normal menstrual cycle, elevated FSH levels in the early follicular phase stimulate recruitment and growth of preantral and small antral follicles; in the mid-and late follicular phase, however, the decline of FSH concentrations and a progressive rise of LH levels are associated with the selection and growth of the dominant follicle destined for ovulation, and the demise of cohorts of smaller follicles. These events appear to be related to the expression of LH receptors on the granulosa cells (GC) of the dominant follicle that render it responsive to LH stimulation and lower dependence on FSH (4, 5). Exogenous gonadotropin administration in COS causes a rise of FSH concentrations throughout the follicular phase; as endogenous LH secretion is usually suppressed with GnRH analogs and exogenous LH activity supplementation is either not provided (as with recombinant human FSH, r-hFSH) or modest (as with human menopausal gonadotropin, hMG), folliculogenesis in COS is mostly driven by FSH. The phys-iologic control mechanisms of the normal menstrual cycle are thus overridden, and the net result is the development of numerous ovarian follicles of all sizes. Although multiple folliculogenesis is the goal of COS, excessive follicle development is also the key factor involved in the pathogenesis of severe ovulation induction complications such as high order multiple gestations and the ovarian hyperstimulation syndrome (OHSS). In this study, we tested the hypothesis that LH activity, administered as low-dose hCG, can be used to stimulate the growth of large follicles and hasten the demise of small follicles when administered in the mid-/late follicular phase of COS treatment for ART. The application of the concepts derived from this study could be used to design ART ovulation regimens that combine efficacy with lower drug-related costs and increased safety through the potential reduction of complications. Materials and Methods Patient population A total of 40 female patients diagnosed as having unexplained or mild male related infertility were studied. All subjects had regular menstrual cycles of 26 -34 d duration, a normal body mass index of 20 -25 kg/m 2 , a pelvic ultrasound showing uterus and ovaries of normal size and structure, a hysterosalpingogram and/or laparoscopy demonstrating tubal patency, normal plasma and urinary chemistry and hematological values, thyroid and reproductive hormones within the normal range. Although ovulation induction had been previously performed in some of the subjects, no patient had received any hormone therapy (including gonadotropins) for at least 3 months preceding the study. Protocol Our Institutional Review Board approved the protocol and all patients provided informed consent. Patients underwent early follicular phase reproductive hormone determinations. The incidence of patients