11.1 Relevant Studies in Milder Approaches in Ovarian Stimulation in Women with Advanced Reproductive and/or Diminished Ovarian Reserve

Mild and minimal stimulation protocols have been favored more for women with expected poor ovarian response (POR) due to diminished ovarian reserve (DOR) and/or advanced reproductive age (ARA). A review from 2017 identified five randomized studies in poor responders comparing conventional high-dose FSH protocols to mild stimulation IVF protocols [1]. One of these was with gonadotropin-releasing hormone (GnRH) antagonist high-dose (500 international units (IU)/day of total FSH) protocol (n:33) versus letrozole plus FSH (150 IU/day) (n:31). This study showed comparable clinical pregnancy rates with much less gonadotropin use [2]. Another trial randomized patients with history of one or more failed cycles to mild stimulation with letrozole 5 mg/day for 5 days between cycle days 2–6 and human menopausal gonadotropin (HMG) 150 IU daily from cycle day 7 with flexible GnRH antagonist protocol (n:30) and conventional high-dose (purified HMG 300 IU daily) GnRH agonist short flare protocol (n:30). This study showed comparable clinical pregnancy rates per cycle with no differences in mean number of oocytes/embryos ratio [3].

Then there were three larger studies. In one of these prospective randomized studies, the authors treated women with DOR and POR (between ages of 18 and 42) with either clomiphene citrate 150 mg daily for 5 days between cycle days 3 and 7 (n:148) or GnRH agonist short flare protocol with daily use of recombinant FSH at 450 IU (n:156). Live birth rate per cycle (3% versus 5%, respectively) and live birth rate per embryo transfer (9% versus 9%, respectively) were comparable among groups [4] .

In one other trial, women with ARA (age ≥ 35 years) or DOR or history of POR were randomized to one cycle of a mild stimulation protocol with a fixed dose of FSH at 150 IU daily started on cycle day 5 after oral contraceptive pre-treatment, which was combined with fixed dose antagonist start on day 6 of stimulation (n:195). The authors had a set cancellation criteria. If the patients showed less than 2 follicles 15 mm or above after day 7, the cycles were canceled. Other group was randomized to one cycle of GnRH agonist long protocol with daily use of HMG at 450 IU (n:199). The clinical (15.3% versus 15.5%, respectively) and ongoing pregnancy (12.8% versus 13.6%, respectively) rates per patient were comparable. There were no data concerning the cryopreservation of surplus embryos, and therefore cumulative pregnancy rates could not be calculated [5].

Lastly, patients <43 years of age with DOR were randomized to mild stimulation with clomiphene citrate 100 mg daily for 5 days between cycle days 2 and 6 and recombinant follitropin alfa 150 IU/lutropin alfa 75 IU (Pergoveris) started daily on day 5. GnRH antagonist was started on day 8 (n:309). Other randomized groups received long GnRH agonist protocol with HMG with starting dose of 300 IU daily and with a maximum dose of 450 IU daily (n:331). Cycle cancellation rate was significantly higher in the mild stimulation arm (13% versus 2.7%). However, clinical pregnancy rate per cycle (13.2% versus 15.3%, respectively) and ongoing pregnancy rate per embryo transfer (17.8% versus 16.8%, respectively) were comparable among groups even if significantly more metaphase II oocytes were retrieved in conventional high-dose stimulation group as compared to the mild stimulation arm. There were no differences in top-grade embryos [6].

A meta-analysis published in 2016 investigated the efficiency of mild stimulation protocols with clomiphene citrate in women with POR [7]. There were only four studies, two of which were by Ragni et al. and Revelli et al. as described above. Other two studies were from 2004 and 2011, and the earlier study was quasi-randomized since protocols were assigned per weekday of the stimulation start [8, 9]. In these two studies, the description of the mild stimulation protocols with clomiphene citrate actually included high gonadotropin doses. D’Amato et al. described the stimulation protocol as recombinant FSH 300 IU twice daily with clomiphene citrate 100 mg for 5 days, and Karimzadeh et al. gave recombinant FSH or HMG at 225–300 IU daily doses along with 100 mg daily and 5-day regimen of clomiphene citrate. Therefore, relevance of this meta-analysis to the defined mild stimulation protocols may not be relevant. Still the authors did not show any difference between the protocols with and without clomiphene citrate in terms of live birth rate and clinical pregnancy rate [7].

Recently the American Society for Reproductive Medicine (ASRM) came up with a practice committee opinion on mild stimulation as compared to conventional stimulation for IVF in women POR [10]. After reviewing two publications [5, 11], it was concluded that although live birth rates cannot be estimated, in women with POR, there is fair evidence that clinical pregnancy rates after IVF are not substantially different when mild stimulation protocols using low-dose gonadotropins (defined as ≤150 IU/day) are compared to conventional stimulation protocols. A similar conclusion was reached in terms of similar clinical pregnancy rates with mild stimulation protocols with oral superovulating agents like clomiphene citrate and letrozole when these mild stimulation protocols are compared to conventional stimulation protocols for IVF in poor responders [2, 3, 6, 12, 13]. However, the evidence was found insufficient to recommend for or against for using oral agents alone for IVF versus using conventional IVF stimulation protocols for poor responders while assessing one trial [4], but that trial also suggested that mild ovarian stimulation with clomiphene citrate alone may be cost effective.

The summary conclusion of ASRM was in POR patients for IVF; strong consideration should be given for mild ovarian stimulation protocols due to lower cost and comparable pregnancy rates to conventional IVF stimulation protocols [10].

11.2 Issues with Prospective Randomized Controlled Trials and Their Interpretation

Certainly, the practice of IVF is a more outcome-based field more than any other field in medicine. The short-term accepted outcomes for IVF are clinical pregnancy rates and ongoing pregnancy rates. These are readily available from prospective randomized controlled trials within a short time period. The midterm outcome and the desired outcome include live birth rates, which may be accounted as live births anywhere between 24th and 40th weeks of pregnancy. Perhaps, even more important outcomes are take-home baby rates and general developmental outcomes of the offspring conceived with IVF. These outcomes are least reported in the literature for any specific IVF approach for POR and DOR patients.

One other point to consider is how the success is defined in IVF. Per cycle pregnancy rate paradigm was created by government agencies when the mandatory IVF data registries were set up in Europe and in the United States. This rate has been typically used for marketing while ignoring the impact of frozen-thawed embryo transfers on IVF treatment outcomes. Therefore, per fresh cycle pregnancy rates are not reflecting the true outcome. Currently the paradigm is shifting toward pregnancy rate per patient. This definition opens venues for more widespread acceptance of mild stimulation approaches for patients with DOR and/or ARA. Such protocols cannot be properly implemented without the presence of a highly successful embryo-freezing program, because especially in protocols involving minimal stimulation or luteal phase mild stimulation, fresh embryo transfers are not considered as explained in subsequent sections.

The studies reviewed above are set protocols without patient-tailored modifications during the protocol implementation although we are living in the era of personalized medicine or precision medicine. The patient-specific approaches may be expected to achieve better outcomes rather than firmly set and “cookie-cutter” protocols.

It is difficult to tailor many parameters at once within the context of a prospective randomized controlled trial. On the other hand, it is observed that DOR and POR patients are profoundly heterogeneous groups. We and others noted that DOR patients present with different basal FSH and estradiol levels and different antral follicle counts with each menstrual cycle. Hence, their oocyte and embryo yield may change over repetitive cycles. In that respect, treatment cycle preparation and choosing the right cycle to start the stimulation are the important aspects of treatment regimens for women with DOR and/or ARA or with history of POR. Some protocols for DOR patients may be associated with a sub-receptive endometrium as well. Considering these and many other factors, achieving conformity in management of DOR patients can easily be deemed impossible. Therefore, customization of the approaches for each such patient should be the norm rather than the exception.

11.3 Conclusion

Milder approaches of ovarian stimulation for IVF in women with DOR and/or ARA are finally met with some acceptance. In the upcoming sections on minimal and mild stimulation protocols for women with DOR and/or ARA and those with predicted POR, we will cover the customization aspects of such protocols. We will go-over stimulation cycle preparation, which minimal or mild stimulation protocol to use, endogenous or exogenous LH activity support while preventing premature LH surge and others. Most importantly, like many others, in order to reach higher pregnancy and live birth rates than those reported in prospective randomized trials, such protocols cannot be performed properly without having an excellent embryo freezing and successful frozen-thawed embryo transfer program as we also discuss in subsequent sections.