Accurate prediction of ovarian response is crucial for most optimal and individualized ovarian stimulation. Clinicians can also provide better advice to patients and predict the risk of adverse events after ovarian stimulation, such as prolonged cycle time, poor ovarian response and cancellation of cycle, or OHSS. Biological and biochemical markers such as AFC and AMH have been proven to predict both the poor and hyper ovarian response with fairly good accuracy [2]. However, previous studies demonstrated that these biomarkers represent a “static” snapshot of the individual ovarian reserve which do not properly reflect the “dynamic” nature of follicular growth in response to exogenous ovarian stimulation [4].
FORT was first introduced by Genro et al. in 2011 to evaluate ovarian response during stimulation [9]. It is calculated as the ratio between the number of pre-ovulatory follicles obtained in response to FSH administration and the preexisting pool of small antral follicles. Thereafter, some studies suggested that FORT < 0.30 indicated low ovarian sensitivity [9, 10]. It was one of the classic indexes to evaluate ovarian response, and has been studied in different populations and different ovarian stimulation protocols. Hassan A et. al concluded that FORT is an independent variable affecting the clinical pregnancy rate in IVF/ICSI cycles. Higher FORT values had better oocyte yield and clinical pregnancy rates in women with unexplained infertility undergoing IVF/ICSI with potentially normal ovarian response [16]. For patients with polycystic ovary syndrome (PCOS) undergoing IVF-ET, FORT was also a powerful tool for measuring ovarian reactivity. A high FORT to obtain high-quality embryos and perform FET could achieve good pregnancy outcome [17]. For the care and management of hypo-responders in ART, FORT proved to be a relevant and crucial quantitative, and qualitative index. Impaired sensitivity to FSH revealed by FORT should be considered in the decision of treatment protocol, gonadotropin, and stimulation doses to be used for hypo-responders [18]. However, this index had some drawbacks that should not be ignored. FORT did not assess the actual number of oocytes retrieved, which was strongly associated with live birth rates [1].
In 2018, the FOI was proposed to address the ovarian sensitivity. It was calculated as the ratio between the total number of oocytes collected at the end of ovarian stimulation, and the number of antral follicles available at the start of stimulation. FOI ≤ 0.50 indicated low ovarian sensitivity and FOI > 0.50 for normal ovarian sensitivity. Hypo-responsiveness and suboptimal/poor response were not synonymous. FOI might be used alone or combined with FORT to most optimally reflect the ovarian response to Gn. However, FOI could be influenced by the initial Gn dosage, genetic or environmental factors, asynchronous follicular development, and technical issues during triggering and OPU [4].
Another marker of the ovarian potential to produce oocytes in response to hormonal stimulation was OSI, which was calculated by dividing the total administered FSH dose and the number of retrieved oocytes. It was first developed by Biasoni et al. and the definition was modified later [11, 19, 20]. The OSI was an interesting tool to assess the ovarian sensitivity to exogenous Gn and could be used to adjust the stimulation regimen in subsequent IVF cycles. A retrospective comparative cohort study including a total of 2150 women who underwent the first IVF cycle using long-agonist protocol validated the use of OSI as a highly reliable index of ovarian responsiveness to recombinant FSH and could be useful to estimate the FSH dose [21]. Another retrospective cohort study, with patients ≥ 39 years who performed their first ART cycle with an antagonist protocol, suggested that OSI was the best index to predict cumulative implantation rate and CLBR. Both OSI and FOI predicted embryo culture with success, but OSI was more accurate [22]. However, The OSI does not consider the Gn type (e.g., whether LH or LH analog was added) or the AFC.
Follicular sensitivity index (FSI) as a new tool for objective assessment of follicular responsiveness to exogenous gonadotropins was proposed in 2017. It was calculated as preovulatory follicle count (PFC) × 100,000/ (AFC × total received FSH doses). Hassan AMA et al. demonstrated that FSI could predict the clinical pregnancy rate in women with unexplained infertility or tubal factor undergoing IVF/ICSI using GnRH agonist protocol. Higher FSI values had significantly higher oocyte yield and fertilization and clinical pregnancy rates [23]. However, the use of FSI may have some limitations in practice. For example, in women with PCOS, the use of FSI will be limited by the high AFC. Till now, no further large-scale studies have been performed to validate the utilization the FSI as a predictor of ovarian response to exogenous gonadotropins.
Taken together, there is no single perfect indicator for ovarian response, which can be affected by many factors. Polymorphism of FSHR, LHR, exons of the LH gene, and sequence variants in the genes that participate in estrogen synthesis such as CYP19A1, were all related to the ovarian response to exogenous Gn, even CLB [24–27]. Besides, other factor, such as oxidative stress in the follicular environment, was also associated with ovarian response [28]. In clinical practice, more indexes are needed to be explored or combined with other indexes, to better predict ovarian response and further reproductive outcomes.
The new marker in this study, named average Gn dosage per follicle, links the number of pre-ovulatory follicles on the trigger day to the degree of hormonal stimulation, expressing how many units of exogenous Gn are needed to obtain each follicle. It can allow doctors to avoid the influence of OPU techniques, inappropriate trigger timing and other factors. Thus, it may be used singularly or combined with other indexes to better assess ovarian response. In this retrospective study, patients with the lowest value of this index (Group A) had the lowest age, and the "static" indexes of ovarian reserve such as basic FSH and AFC were the best. Meanwhile, with the least total Gn dose and shortest Gn duration, the most oocytes and good quality embryos were retrieved. Other indexes of ovarian response, FORT, FOI, and OSI were also the best. As to the cumulative pregnancy outcomes of an entire ART cycle, patient with low (Group A) and medium (Group B) Gn dosage per follicle had more chances of getting pregnant or live birth, compared with patients with high value of this index (Group C). In a multivariate regression model, the average Gn per follicle was also shown to represent an independent predictor of CLB.
In this study, we also carried on another comparison and analysis. Patients were divided into two groups based on whether CLB was achieved in an entire ART cycle. Then baseline characteristics and demographic data, cycle parameters and indexes for ovarian response were compared. The results showed that patients who achieved CLB had better "static" markers of ovarian reserve, such as basal FSH and AFC, and higher "dynamic" markers of ovarian response, such as FOI and OSI, and lower average Gn dosage per follicle. Although FORT was not statistically different between the two groups, it was also observed that the FORT was slightly higher in patients with CLB, which could not be ruled out as being related to the small sample size.