Polycystic ovary syndrome (PCOS) typically manifests with hyperandrogenism, oligo-anovulation and polycystic ovarian morphology. Studies have shown that PCOS is the main cause of anovulatory infertility, affecting about 8-13% of childbearing age women (1–3). For PCOS patients, in vitro fertilization (IVF) treatment is an assisted reproductive option, where controlled ovarian stimulation (COS) is the main step. However, COS in these patients frequently leads to large quantities of poor-quality oocytes and increased incidence of ovarian hyperstimulation syndrome (OHSS).

Anti-Mullerian hormone (AMH), antral follicle count (AFC), and basal follicle stimulating hormone (FSH) levels have been used to adjust ovarian stimulation (OS), minimize risks and optimize assisted reproductive technology (ART) outcomes, but all of them have certain limitations (4–7). They do not reflect the dynamic nature of follicular growth in response to exogenous gonadotrophins (Gn) and their ability to predict clinical outcomes after ART is limited (7, 8).

In 2011, Genro et al. (9) proposed the concept of the follicular output rate (FORT), the ratio of pre-ovulatory follicle count (PFC) on the trigger day to the AFC, to quantify the follicular development potential. Gallot et al. (10) further explored the correlation between pregnancy rate and FORT in patients who underwent IVF treatment with regular menstrual cycles. They found that better pregnancy outcomes were related to high FORT values. In addition, Hassan et al. (11) studied women with unexplained infertility and found that FORT had an independent effect on clinical pregnancy rate.

Several studies have suggested that FORT may reflect clinical outcomes after ART in PCOS patients, but consensus has not been reached (12–14). In an early study of 140 patients with PCOS, the fertilization rate and high quality embryo rate were highest in the middle-FORT group, although the FORT groups did not differ significantly in the clinical pregnancy rate (12). However, Tan et al. (13) found that the clinical pregnancy rate and high quality embryo rate increased with FORT in PCOS patients. A recent study of PCOS patients performed by Yang et al. (14) showed that the cumulative live birth rate (CLBR) was highest in the high-FORT group but lowest in the middle-FORT group.

Although Yang et al. (14) investigated the association of FORT with cumulative ART outcomes, the impact of clinical characteristics of PCOS was not taken into account in this study. A previous research showed that PCOS patients with different phenotypic characteristics reflected varied ovarian responses to COS (15). Furthermore, AFC and CLBR were significantly different in PCOS patients with different phenotypic features (15–17). To date, conclusive and definite data about the role of clinical characteristics of PCOS on the relationship between FORT and cumulative ART outcomes are still lacking.

In this study, we investigated the relationship between FORT and cumulative ART outcomes in PCOS patients with different characteristics after one IVF cycle including all fresh and subsequent frozen-thaw embryos, in order to guide COS medication and help PCOS patients get better reproductive outcomes.

In this retrospective study of 454 PCOS patients, we found that FORT was an independent factor affecting the cumulative IVF outcomes. The CLBR and CCPR were positively correlated with FORT when the FORT was less than 70%.

FORT, as a simple and noninvasive tool in our clinical practice, could objectively reflect dynamic changes of follicular growth in response to exogenous Gn. Genro et al. found a negative association between FORT and AMH levels in peripheral blood, which might be explained by the hypothesis that AMH inhibited the sensitivity of antral follicles to Gn (9). Hassan et al. showed that no significant difference in the serum AMH levels was found among the FORT groups (11). We found that FORT was positively associated with AMH levels in peripheral blood. The difference between our results and other studies may be due to the disparities in the studied populations. We studied patients with PCOS, whereas Genro et al. and Hassan et al. studied patients with different infertility causes (9, 11). In the studies conducted on PCOS cases only, the serum AMH levels can be used as a marker of ovarian responsiveness and there was a positive association between AMH levels and assisted reproductive outcomes (21). In our study, basal testosterone levels in the high-FORT group were dramatically higher than that in the low-FORT and middle-FORT groups. This finding supports the previous studies which showed that basal testosterone level positively correlated with ovarian response and follicular count on trigger day (≥ 14 mm) (22, 23). The possible mechanism is that androgens could enhance FSH receptor expression in granulosa cells and are considered to promote follicular development by amplifying the effects of FSH (24). Additionally, androgens also augment the expression of insulin-like growth factor 1 (IGF-1) in the primate ovary, which is crucial for regulating follicular growth (22, 25).

This study found that the CCPR, cleavage embryos, 2PN zygotes, number of retrieved oocytes and PFC increased progressively from the low to high FORT groups. And the numbers of available embryos and high quality embryos were significantly lower in the low FORT group. These results are in agreement with the earlier reported results (10–12). In the high-FORT group, patients have better ovarian responsiveness to exogenous gonadotrophins, resulting in increased mature and retrieved oocytes, and consequently better clinical outcomes. In our study, the fertilization rate was significantly higher in the high FORT group. Our findings are in agreement with those of Hassan et al. (11), but they differ from that obtained in other studies (10, 12, 14), which did not show any difference in fertilization rate among the FORT groups. The difference between other studies and our findings may be attributed to the disparities in the number of cases investigated and study populations. We also found that the available embryo rate was higher in the low FORT group. This may be due to the relatively low number of retrieved oocytes in the low-FORT group. Despite these contradictions, the findings revealed that FORT can be used as a qualitative reflector of the follicular responsiveness to FSH, oocyte competence, and embryo quality.

The central finding of our study is the positive correlation between FORT and CCPR and CLBR in PCOS patients. The CCPR and CLBR in the high-FORT group were significantly higher than that in the low-FORT group. With the widespread use of embryo cryo-resuscitation technology, the CCPR and CLBR, defined as the pregnancy and live birth after using up all fresh and frozen embryos derived from one single COS cycle, appear to be more accurate and comprehensive measures to reflect the effectiveness and safety of IVF treatment (26–29). After reviewing the published literatures regarding PCOS and FORT, we found that there were few researches used CCPR and CLBR as clinical outcome indicators. Yang et al. (14) investigated the relationship between FORT and CLBR and showed that the CLBR was highest in the high-FORT group and lowest in the middle-FORT group. Differences between this study and our results may be due to the disparities in the study populations and the COS protocols. Also, in their study, there were significant differences in Gn dosage and stimulation days among three groups. This may affect the outcome because PCOS patients are usually high ovarian responders.

With a threshold effect model, our results showed that when the FORT was lower than 70%, the CCPR increased by 3.9% and the CLBR increased by 2.4% with each additional unit of FORT. When the FORT was greater than 70%, the CCPR and CLBR did not increase significantly even if the FORT increased. The positive association between the FORT and the cumulative ART outcomes in the first segment of the curve suggested the importance of the high FORT values for the success of IVF. Patients with high FORT values may have more oocytes and produce more euploid embryos that can be used for embryo transfer, thereby increasing the chance of pregnancy and live birth. However, when the FORT reached a certain value, there was no longer a significant beneficial relationship. This may be due to the fact that pregnancy outcomes are influenced by many other factors, such as obesity, environmental exposure (including smoking and alcohol), stress and antiphospholipid syndrome (30, 31). Therefore, for PCOS patients, we should also pay attention to the general conditions and lifestyle rather than simply increasing the FORT.

According to our results, patients with hyperandrogenemia (HA) can increase the CCPR and CLBR by increasing the FORT. HA is the core etiology and primary endocrine characteristic of PCOS. HA leads to premature granulosa cell luteinization and abnormal oocyte maturation by altering follicular fluid microenvironment and the feedback of ovarian hormones to hypothalamic-pituitary-ovarian (HPO) axis (32). In addition, due to the expression of androgen receptors in pancreas and hepatocytes, high testosterone levels could lead to hyperinsulinemia, which seriously impairs ovarian function resulting in premature arrest of follicular development and oligo-anovulation (33). Furthermore, high testosterone levels in PCOS patients also influence glucose metabolism of endometrium, which leads to local insulin resistance and subsequently endometrial lesion (34). Evidence show that the CLBR of patients with hyperandrogenemia is significantly lower than that of individuals without hyperandrogenemia (16). Therefore, improving FORT is a good choice for these poor-prognosis patients with HA to improve the CCPR and CLBR. Furthermore, we also found that patients with polycystic ovarian morphology or ovulation disorder had better cumulative IVF outcomes in the high-FORT group. These results together suggest that FORT, as a noninvasive and simple tool in our clinical practice, may contribute to improving the CCPR and CLBR for PCOS patients with typical clinical characteristics.

Our results showed that the CCPR and CLBR increased with FORT in PCOS patients over 30 years old, whereas the relationship was not statistically significant in patients under the age of 30. This indicated that the older the age, the more positive correlation between the FORT and cumulative ART outcomes. We know that oocyte quality gradually declines with women aging and the competence of women’s oocytes begins to deteriorate around their third decade (35). Multiple potential mechanisms may be responsible for this, such as meiotic spindle abnormalities, mitochondrial dysfunction and chronic exposure to oxidative stress, which usually lead to aneuploidy of the embryo and a higher incidence of adverse pregnancy outcomes (36–39). In general, younger women have better oocyte quality, which may somewhat attenuate the impact of FORT on pregnancy outcomes. Therefore, for the older age group of PCOS patients, due to the decline of oocyte quality, higher FORT values and more oocytes retrieved are needed to achieve better cumulative ART outcomes.

In this study, we found that the FORT was significantly related with the CCPR and CLBR in PCOS patients without overweight and obesity. By boosting FORT, we can increase the CCPR and CLBR in these individuals. Obesity and overweight are known risk factors for cumulative ART outcomes (40), and PCOS patients are more likely to be obese and overweight, which contributes to diminished response to ART, adverse pregnancy outcomes and higher incidence of other complications (41, 42). Obesity could impair endometrial function through inflammation, oxidative stress or other ways, which could cause decidual formation abnormalities and embryo implantation failure (43). Additionally, obesity affects oocyte function by inducing abnormal chromosome pairing and altering follicles’ liquid microenvironment (32). Therefore, weight control and alleviating metabolism disorders are more beneficial to the prognosis of overweight and obese patients than increasing the FORT.

The main strengths of our study rest on the following aspects. First, it was the first study to investigate the role of clinical characteristics of PCOS on the relationship between FORT and cumulative IVF outcomes. Second, the present study first uncovered a curvilinear relationship between FORT and CCPR and CLBR, and this relationship might be useful in establishing an optimal treatment strategy for PCOS patients to obtain better reproductive outcomes. Additionally, our research used CCPR and CLBR as main outcome measures, which is an important advantage over other metrics.

Despite the strengths, this study has some limitations. One disadvantage of the index FORT is related to its operator-dependent characteristic. It is unlikely to rule out the variation in marking AFC and PFC by different sonographers. Secondly, the association between FORT and the incidence of OHSS was not investigated in our study. Other limitations of our study lie in the retrospective and monocentric character, as well as the small study population. Prospective and multicentric investigations with larger simple size and longer duration of observation would be necessary to further validate the findings.

In summary, the present findings indicate that cumulative IVF outcomes have a positive correlation with FORT in PCOS patients when the FORT was less than 70%. For PCOS patients with polycystic ovarian morphology, ovulation disorder or hyperandrogenic manifestations, a high FORT could be conductive to achieving better pregnancy outcomes.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by the research ethics committee of the Second Hospital of Hebei Medical University. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

Study design: ZZ. Data acquisition: RJ, PC, and YL. Data analysis: RLJ, MC, and HH. Writing manuscript: Rulan Jiang. All authors read the article and agreed to the submitted version.