Prematurely ruptured dominant follicles often retain competent oocytes in infertile women

Ovulation consists of a follicle’s rupture and subsequent oocyte extrusion, although there is a paucity of evidence regarding whether every follicle’s rupture is associated with extrusion of its oocyte. We examined this issue in a large-scale window-of-opportunity study by attempting aspiration of single dominant follicles that were found to have ruptured before a scheduled oocyte retrieval during in vitro fertilisation and embryo transfer treatment of infertile women. We were able to aspirate 587 of 1,071 ultrasonographically confirmed post-rupture dominant follicles from 1,071 women (i.e. one dominant follicle per woman) and retrieved 225 oocytes (oocyte recovery ratio: 43.4% of aspirated follicles), which yielded 28 live births (live birth ratio: 11.0% of retrieved oocytes). Interestingly, the live birth ratio for post-rupture dominant follicles was not statistically different from that achieved using regular pre-rupture aspiration of dominant follicles (1,085/8,977, 12.1%). These findings suggest that oocyte extrusion frequently does not occur after follicle rupture in infertile women undergoing in vitro fertilisation treatment, although the oocyte retained in the follicle can remain competent for use during that treatment.


Discussion
This study demonstrated that oocytes were retrievable from >40% of ultrasonographically confirmed post-rupture dominant follicles that could be aspirated during natural or minimally stimulated IVF-ET. Furthermore, we found that those oocytes were as competent as oocytes retrieved from pre-rupture follicles. These findings reinforce the concept that follicle rupture (fluid release) and ovulation (extrusion of the COC) are separate processes in a sequence of events.
The present study also confirmed that oocyte retention can occur in a large number of patients, which has previously been reported in controlled ovarian hyperstimulation cycles 1,2 , and extended this possibility to the rupture of a single dominant follicle. The oocyte recovery rate from the present study (43.4%) is lower than the previously reported recovery rate (13/22, 60%), which could be attributed to the different aspiration modalities  www.nature.com/scientificreports www.nature.com/scientificreports/ (ultrasound-guided transvaginal aspiration vs. laparoscopically visualised direct aspiration). The most notable implication of the present study is that oocyte retention (extrusion loss) may occur even in spontaneous ovulatory cycles (natural cycles) that generate a single dominant follicle.
There are several possible explanations for the extrusion loss that we observed in the post-rupture dominant follicles. The first explanation involves a true failure of the ovulatory process, whereby oocytes are never extruded from the follicle. In this context, cumulus cell expansion is considered essential for COC release from the mural granulosa cell layer, although the precise mechanism of this release remains unknown 3,4 . After the LH surge, cumulus cells start to produce and secrete an extracellular matrix that consists of hyaluronan, proteoglycans, and proteoglycan-binding proteins, which facilitates cell detachment from the COC and release of the COC from the mural granulosa cell layer 5 . The association between cumulus cell expansion and COC release is also supported by findings from mouse models, as genetic disruptions of the extracellular matrix reduce ovulation frequency and increase entrapment of oocytes in the luteinised follicles, which suggests that cumulus cell expansion and COC release are inseparable co-dependent processes 3,6-10 . Our results revealed that a non-negligible number of post-rupture dominant follicles had less expanded cumulus cells (i.e. immature COCs) despite the accelerated maturation of the oocyte (i.e. a higher proportion of metaphase II). Therefore, we assume that limited expansion of cumulus cells, combined with accelerated oocyte maturation, may contribute to extrusion loss in prematurely ruptured dominant follicles. However, this cannot be a major mechanism, as most post-rupture COCs were composed of mature oocytes and mature cumulus cells. Intra-follicular ovarian pregnancy, in which a conceptus is located deep within the corpus luteum, might be a consequence of true failure of oocyte extrusion. In this scenario, the spermatozoon may enter the follicle through the stigma and fertilise the oocyte that is entrapped in the follicle, leading to in situ implantation [11][12][13] .
A second explanation for extrusion loss is a delay in the ovulatory process. In this context, the oocyte extrusion could proceed more slowly than we have assumed, or the extrusion process may resume long after the follicle's rupture. However, a transvaginal ultrasonographic observation study of human chorionic gonadotropin-triggered follicles demonstrated that follicular fluid loss is a rapid process, with average times of 0.9 min for >70% follicle fluid loss and 6.1 min for complete fluid loss 14 . Furthermore, laparoscopic observations favour a similarly rapid process, as oocytes appear at the surface of the ovary or fallopian tube within 2-3 min after follicular rupture 15 . Other laparoscopic observations have indicated that follicular fluid release was more comparable to an explosive event than a slow oozing event 2 , and the authors speculated that the forceful outflow of follicular fluid contributed to the extrusion of the free-floating COC 2 . Thus, it is difficult to imagine that a diminished follicle, after follicular fluid release, resumes the process of extruding an oocyte that has been trapped in blood-containing fluids 2 . In addition, our observations revealed that the oocytes remained inside post-ruptured follicles despite the compressive pressure from a transvaginal ultrasonographic probe and needle puncture during the oocyte retrieval. This suggests that ruptured follicles are not vulnerable to external compressive pressure during oocyte extrusion. Therefore, based on previous reports and our experience, we do not believe that what we observed is simply a snapshot in an ongoing process of oocyte extrusion.
A third explanation for extrusion loss is the effects of exogenous manipulation, including the use of drugs for triggering the LH surge, which modulates some biochemical sequence(s) during the ovulatory process. Interestingly, extrusion loss occurred more often in cycles with lower LH levels at triggering than in cycles with higher LH levels at triggering (Table 3). This may suggest that inappropriate premature triggering may increase  www.nature.com/scientificreports www.nature.com/scientificreports/ the chance of oocyte retention. Nevertheless, the odds ratio attributed to the LH level at triggering was as low as 0.99, which suggests that the timing of the triggering only had a minor effect on the likelihood of oocyte retention. In addition to its timing, the pharmacological actions of buserelin might increase the likelihood of extrusion loss, although we cannot comment on this possibility, as all women received buserelin in this study.
There are several limitations in this study. First, the study included IVF-ET cycles that were heterogeneous in terms of any ovarian stimulation. Another limitation is the possible selection bias that may have been introduced by only considering follicles that had prematurely ruptured before the scheduled oocyte retrieval. Moreover, our experience was limited to infertile women for whom IVF-ET was indicated. Therefore, our findings may not be replicated during cycles in which follicular rupture occurs later than 33-35 h after LH triggering, or in the cycles of fertile women. Moreover, the oocyte recovery ratio of post-rupture follicle (43.4%) might be overestimated, as quite a few cases were judged impossible to needle puncture, and the lowest possible estimate of the oocyte recovery ratio based on our data is 23.8% (255/1,071 of total post-ruptured follicles).
In conclusion, this study revealed that mature oocytes could routinely be retrieved from follicles that had prematurely ruptured in infertile women, which suggests that extrusion loss is not uncommon among these women. Future studies should determine whether extrusion loss only occurs during LH-surge manipulated cycles with premature follicle rupture in infertile women, or whether it occurs during the spontaneous ovulatory cycles of normally fertile women.

Materials and Methods
ethical approval. The clinical application of oocyte retrieval from ultrasonographically identified ruptured follicles was initially approved by The Ethical Committee of Towako Clinics in 2010 (#2010-1). This study's retrospective analysis of the related outcomes was also approved at a later date (#2014-3). Only patients who provided informed consent underwent the aspiration of post-rupture follicles with subsequent use of any retrieved oocytes in their IVF-ET treatment, in accordance with the Declaration of Helsinki. Furthermore, written informed consent was obtained from the patients for the use of the videos. patients and study design. This opportunistic retrospective study included patients who underwent IVF-ET treatment in Shimbashi Yume Clinic between February 2011 and January 2016. The patients were eligible for natural-cycle or minimally stimulated IVF-ET, and completed 1-3 treatment cycles at the clinic. Most patients (76%) had previously experienced at least one IVF-ET treatment failure at other institutions. Thus, the patients desired an alternative method, and we offered them natural or minimally-stimulated cycles with re-assessment of their ovarian function. Couples with azoospermia were excluded, but couples with oligozoospermia were included. The patients' clinical characteristics are shown in Table 4.
We identified 11,942 cycles (5,261 natural cycles and 6,681 minimally-stimulated cycles) in which a single dominant follicle had grown. Cycles in which ≥2 follicles grew to ≥11 mm on the day of triggering were not included in the analysis (Fig. 1). Among the 11,942 cycles, we identified 1,071 cycles with 1,071 dominant follicles (i.e. 1 follicle per cycle) that had unexpectedly ruptured, based on ultrasonographic confirmation, before the scheduled oocyte retrieval. Despite the follicle's rupture, we performed post-rupture follicle aspiration and any resulting oocytes were used for IVF-ET. The IVF outcomes were compared to the regular IVF outcomes using oocytes that were normally aspirated from pre-ruptured follicles. Cycles in which the dominant follicle started to rupture at the oocyte retrieval (i.e. an ongoing decrease in follicular diameter was observed) were assigned to the pre-rupture follicle group.
follicle stimulation. For natural-cycle IVF, the patients only received buserelin acetate (300 µg) for LH triggering at 33-35 h before the scheduled oocyte retrieval. For minimally stimulated IVF, the patients received clomiphene citrate (25-50 mg per day) or letrozole (1.25 mg per day) for days 3-5 of the cycle, and also received buserelin acetate (300 µg) for the LH triggering. oocyte retrieval. Follicle monitoring and triggering of the LH surge were performed as previously described 16 . The follicles were monitored using transvaginal ultrasonography (HI VISION Avius ® ; Hitachi Medical Corporation, Tokyo, Japan) and levels of oestradiol (E2) and LH were measured every day or every other day after day 10. Once a dominant follicle had reached a diameter of 16-18 mm and E2 levels of ≥190 pg/mL were detected, the buserelin acetate was administered nasally to trigger the LH surge and oocyte retrieval was scheduled for 33-35 h later. On the morning of the planned oocyte retrieval, follicles were examined using transvaginal ultrasonography to determine whether the dominant follicle remained intact (pre-rupture) or had prematurely ruptured (post-rupture). Ruptured follicles were identified based on the total or subtotal loss of echo-free follicular fluid contents, relative to the most recent examination.
If the ruptured follicle could be located within the ovary, based on the presence of an irregularly shaped central hypoechoic cavity, the physician (ST) punctured it using a 23-G needle while taking care not to puncture the small follicles around the ruptured dominant follicle. Representative needle aspirations are shown in Supplementary Videos 1 and 2. Most post-rupture dominant follicles had a 1-2-mm fluid-containing compartment that could still be punctured. After the puncture, the internal fluid was aspirated as thoroughly as possible by gently moving the needle tip around inside the space while repeatedly rotating the needle around its axis. Aspiration was performed using an automatic aspirator and negative pressure (−160 to −200 mmHg). Once back-flow had ceased, the needle was gently removed (without any additional puncture) and then flushed with a collection medium that contained heparin. The oocyte retrieval procedures were recorded and stored as ultrasonographic movies, and were later reviewed to exclude the possibility of inadvertent puncture of nearby small follicles 16 . The median time evaluation of cocs. Photographs of 178 COCs from post-rupture dominant follicles were available to evaluate the maturation status of the cumulus cells and intercellular substances. Five hundred photographs of COCs from the pre-rupture follicle were randomly selected as controls, and we found that 467 of those photographs were of sufficient quality to perform the evaluation.