Mild stimulation for in vitro fertilization: making progress downward.

The prevalence of infertility is 9% of married or co-habiting couples in all countries of the world (Boivin et al., 2007). In many cases, the infertility can be treated effectively only by means of assisted reproductive techniques (ART). In some developing countries, however, ART services are too expensive to be offered. Even in developed countries, ART is only available to women who can afford to pay and they often find that the time commitment, discomfort and complications are unbearable. Innovative approaches are needed to dislodge the chief barrier to availability, which is the cost of in vitro fertilization (IVF). In a recent comment to Nature, Alan Trounson said ‘If you remove all the expensive stuff and use low-cost drugs (such as clomiphene) and remove just one or two oocytes, and only transfer one embryo, it can be done for ,US$100’ (Pearson, 2008). Low cost drugs and less potent ovarian stimulation are one way to reduce the cost of IVF. Sensible as drastic cost reduction may be, some degree of ovarian stimulation is required, as shown by the difficult history of natural cycle IVF. Early pregnancy rates were disappointing to patients and clinicians (Lenton et al., 1992). Even now pregnancy rates are only 7% per cycle (or one in 14 cycles) and 44% overall in women who were able to complete an average of four cycles (Pelinck et al., 2007). Clearly, broad application of natural cycle IVF is a distant goal, as patients expect higher success rates and the time they must invest in a cycle is not negligible. Although natural cycle IVF has not gained widespread acceptance, even a small reduction in the degree of ovarian stimulation could be an important step on the road to low cost treatment (Edwards et al., 1996). Mild ovarian stimulation, with the aim of reducing the number of developing follicles, could reduce cost and adverse events and would be consistent with the transfer of fewer embryos, resulting in fewer multiple births (Fauser et al., 1999). The most common protocol for IVF involves up to 3 weeks of gonadotrophin releasing hormone (GnRH) agonist treatment, up to 2 weeks of FSH injections and frequent visits for ultrasound and hormone evaluation, as well as the essential luteal support after embryo transfer. The long GnRH agonist protocol is effective and its management has become a comfortable routine over the years, but it is costly and difficult for patients. The development of many follicles may lead to ovarian hyperstimulation syndrome and the transfer of multiple embryos increases multiple birth rates (van Voorhis, 2007; Andersen et al., 2008). Clearly, a milder stimulation protocol would reduce the complexity of IVF cycles, but can it do so without a significant reduction in success rates? Two reviews in this issue evaluate the current literature on mild stimulation protocols. One explores broadly the value of different approaches to mild stimulation in IVF cycles (Verberg et al. 2008a). These authors find that (i) GnRH antagonist co-treatment to suppress a premature luteinizing hormone (LH) rise reduces FSH dosage and treatment days compared with the GnRH agonist long protocol, (ii) limited gonadotrophin dosage is associated with better pregnancy rates than anti-estrogens or aromatase inhibitors and (iii) the use of LH late in the follicular phase reduces total FSH dosage. In vitro maturation of oocytes also avoids ovarian stimulation, although its efficacy with respect to standard IVF has not been demonstrated in level I studies (Rao and Tan, 2005). The review found, however, that many mild stimulation studies involved poor prognosis patients, while it is mainly young patients with good ovarian responses who have the highest risk of complications from ovarian stimulation. The authors conclude that mild stimulation should yield more than one dominant follicle for the success rates to compete with those of standard stimulation protocols. The second review has a narrower focus and a more in-depth analysis (Verberg et al., 2008b). This review asked whether retrieving a small number of oocytes after mild ovarian stimulation reflected a poor ovarian response and poor implantation rates, as it would with conventional stimulation. Only three studies met the search criteria for randomized controlled trials which compared milder stimulation protocols involving use of GnRH antagonist and lower dosage of