Usefulness of expanding the indications of early rescue intracytoplasmic sperm injection

Abstract Purpose Early rescue intracytoplasmic sperm injection (ICSI) is often performed in cases in which not even a single oocyte has extruded a second polar body 6 h after insemination. We evaluated the usefulness of expanding the indications of early rescue ICSI to cases in which <80% of oocytes have extruded second polar bodies 6 h after insemination. Methods Early rescue ICSI was performed on oocytes that were denuded 2.5 h post‐insemination and whose extrusion of the second polar bodies had been examined 6 h post‐insemination with a PolScope. Results In vitro fertilization was performed on 24 496 oocytes of 4944 cycles, and 1438 cycles had <80% rate of the second polar body extrusion. Rescue ICSI was performed on 3933 oocytes. Three pronuclei (3PN) incidence of rescue ICSI was 3.0% in oocytes with ≥50% rate of the second polar body extrusion. With respect to the second polar body extrusion rate, no differences were observed in normal fertilization, blastocyst development, implantation, miscarriage, or live birth rates for rescue ICSI. Conclusion By expanding the indications of early rescue ICSI using the PolScope to cases in which <80% of oocytes have extruded the second polar bodies, many fertilized oocytes can be obtained without considerably increasing the 3PN rate.


| INTRODUC TI ON
In vitro fertilization (IVF) is the first-choice fertilization method; however, it has drawbacks such as 5%-20% total fertilization failure (TFF) 1 and lower fertilization rate than intracytoplasmic sperm injection (ICSI). 2 For improving the fertilization rate, rescue ICSI is performed on oocytes that have not been fertilized in IVF. The earliest attempt was reported by Nagy et al. in 1993 3 on rescue ICSI performed on the day after oocyte retrieval. Several follow-up studies concluded that the pregnancy rate was low owing to oocyte aging. 4,5 In 2003, Chen et al. performed rescue ICSI on oocytes from which the second polar bodies had not been extruded 6 h after insemination, achieved a high pregnancy rate, and reported the usefulness of early rescue ICSI. 6 Early rescue ICSI is indicated for cases suspected of TFF, in which not even a single oocyte from which a second polar body has been extruded 4-8 h after insemination is obtained. [6][7][8][9][10][11][12] Although studies reported that rescue ICSI was performed in cases where the second polar bodies had been extruded from some oocytes, [13][14][15][16] they targeted cases in which the ratio of oocytes from which the second polar bodies had been extruded was <50%. It is expected that many unfertilized oocytes can be saved, but a study reported a high polyspermy rate in rescue ICSI performed in cases with an IVF fertilization rate of >25%, 14 thus raising concerns that unnecessary rescue ICSI may be performed if the indications are expanded.
When performing rescue ICSI on the day after oocyte retrieval, determining whether to perform the procedure based on the presence or absence of pronuclei (PN) is relatively simple. Early rescue ICSI, however, requires the physician to make a decision based on the number of polar bodies. Determining the number of polar bodies for oocytes with fragmented polar bodies is difficult, and performing unnecessary rescue ICSI on oocytes from which the second polar bodies have been extruded is problematic because it results in the formation of 3PN. Therefore, rescue ICSI is often indicated for TFF. To overcome the difficulty in determining the number of polar bodies, the spindle is observed to assist in the determination of second polar body extrusion. 17,18 However, few studies have examined this aspect, and no study has examined a large number of cases.
The present study examined the usefulness of expanding the indications of rescue ICSI while suppressing the risk of polyspermy by observing the spindle using PolScope and accurately determining the extrusion of second polar bodies. As the expansion of the indications of rescue ICSI to all oocytes from which second polar bodies have not been extruded has a small effect considering the increased effort required, in order to obtain >80% of fertilized oocytes, equivalent to ICSI, the indications of rescue ICSI were expanded to cases in which <80% of oocytes had extruded second polar bodies.

| Patients
A retrospective cohort study was conducted on 4944 IVF cycles and 2302 ICSI cycles from which metaphase (M) II oocytes were collected following oocyte retrieval in 2010-2019. Written patient consent was obtained after providing a thorough explanation. This study was conducted with the approval of the ethics committee of Minoura Ladies Clinic (Approval number: H21.03).

| Stimulation, oocyte retrieval, and insemination
Controlled ovarian stimulation was performed mainly by the agonist/human menopausal gonadotropin (hMG) method, antagonist/ hMG method, and clomiphene citrate/antagonist/hMG method.
Oocyte retrieval was performed under the guidance of transvaginal ultrasound 36-38 h after the trigger by hCG or GnRHa agonist.
While examining oocytes, cumulus-oocyte complexes were stretched, and the presence or absence of the first polar body was confirmed using an inverted microscope (IX70, Olympus). The oocytes were divided into MI oocytes and MII oocytes. Among the MII oocytes, the oocytes with fragmented polar bodies that appeared to have divided into two parts were separated and cultured. After oocyte retrieval, a maximum of five oocytes were added to the Center Well Dish containing 1 ml of human tubal fluid medium (Irvine Scientific, Santa Ana, USA) and then cultured.
For IVF, insemination was performed at least 3 h after preculture following oocyte retrieval. Standard insemination sperm concentration was set at 200 000/ml and corrected based on normal sperm F I G U R E 1 Polar bodies and spindles observed by relief contrast and PolScope. a, b, c, and d show changes over time until the second polar body is extruded, with the (A) spindle in the cytoplasm, (B) spindle in the protruded part on the cell membrane surface, (C) spindle crosslinked between the second polar body and cytoplasm, and (D) spindle that is thinning and disappearing. (E) and (F) show the judgment of fertilization by spindle in cases where the number of polar body is difficult to determine, with (E) 2PB with cross-linked spindle and (F) 1PB with spindle in the cytoplasm despite fragmented polar body morphology ratio and previous fertilization rate. ICSI was selected for cases where the adjusted motile sperm count was insufficient for the sperm count for insemination, as well as for cases in which TFF had occurred in a previous round of IVF. ICSI was performed at least 3 h after preculture.

| Denudation in IVF, judgment of second polar bodies, and implementation of rescue ICSI
The denudation of cumulus-oocyte complexes began 2.  In the assessment performed immediately after denudation, apart from oocytes with one polar body (1PB), an oocyte in which the spindle was observed in the cytoplasm despite the presence of a fragmented polar body, was determined as 1PB. An oocyte with two polar bodies (2PB), as well as a cross-linked or disappearing spindle, was determined as 2PB. An oocyte without a spindle in which the first polar body was not fragmented during oocyte retrieval, clearly appearing to be two polar bodies, was determined as 2PB.
Oocytes in which polar bodies were observed were transferred to a microdrop medium prepared with 35 μl of global medium (Life Global, Brussels, Belgium) and then cultured in an incubator (APM-30D, Astec, Fukuoka, Japan) at 37°C, 5% CO 2 , 5% O 2 , and 90% N 2 .
Oocytes determined to be 1PB immediately after denudation were observed again at 6 h after insemination to count the number of polar bodies. Oocytes where fragmentation occurred after denudation and in which the number of polar bodies was difficult to determine were compared with images obtained immediately after denudation, and oocytes in which the number of polar bodies clearly increased were determined as 2PB. When the number of polar bodies was ambiguous, the spindle was observed, and if the spindle remained, they were determined to be 1PB. However, if the spindle was cross-linked or disappearing, they were determined as 2PB. For oocytes determined as 1PB, the spindle was confirmed again before performing rescue ICSI.
Indications of rescue ICSI were cycles with <80% of oocytes determined as 2PB 6 h after insemination. The extrusion rate of the second polar body was defined as the ratio of inseminated oocytes that were observed as having extruded the second polar body when determining whether or not to perform rescue ICSI. Even if the oocytes were determined as 1PB, oocytes that were 1PB during denudation but in which the spindle was not observed, those in which the spindle was in the central cytoplasm, and those with poor cytoplasmic conditions (such as atrophied cytoplasm and constricted cytoplasm) were considered poor quality oocytes and were excluded from the target of rescue ICSI in cases where at least 50% of oocytes had extruded second polar bodies. Blastocyst freezing was performed on embryos expanded to Gardner classification ≥2BB on the afternoon of Day 4 and ≥3 on Day 5 or 6. 19 Thawed embryo transfer was performed 4-20 h after thawing in the ovulation cycle or hormone regulation cycle. Transferred embryos ≥4AA were graded as good, ≥3BB as fair, and <3BB as poor. Embryos transferred on the sixth day from oocyte retrieval were graded as good for ≥5AA, fair for ≥4BB, and poor for <4BB. Evaluation for implantation rates only included single blastocyst transfers.

| Statistical analysis
The Kruskal-Wallis test and the Bonferroni method for multiple comparison were performed for age. Logistic regression analysis was performed for fertilization and blastocyst development rates with oocyte age added as the explanatory variable, and logistic regression analysis was performed for implantation rates, miscarriage rates, and live birth rates, with oocyte age and transferred blastocyst grading added as the explanatory variables. A p-value <0.05 was determined to be statistically significant. Software R version 3.4.1 was used for statistical analysis.  Table 1 shows the comparison by fertilization methods.

| RE SULTS
Compared with that of rescue ICSI (76.3%), the 2PN rate was higher in IVF (79.3%) (p < 0.001) and showed no difference from that of ICSI.
Compared with rescue ICSI (57.3%), Day 6 blastocyst development rate was higher in IVF (69.7%) (p < 0.001) but showed no difference from ICSI (58.6%). Compared to rescue ICSI, the grade of transferred embryos was good in IVF (p < 0.001), but when analysis was conducted by inputting grade as the explanatory variable, there were no differences in the implantation, miscarriage, or live birth rates. Table 2 shows the comparison by second polar body extrusion rate of rescue ICSI oocytes. 3PN rate showed no difference in cycles with a second polar body extrusion rate of <25% and 25%-50% but was significantly higher in that of ≥50% (p = 0.025). There was no difference in the blastocyst development rate, implantation rate, miscarriage rate, and live birth rate. Table 3 shows the comparison of non-rescue ICSI oocytes, which did not undergo rescue ICSI despite being judged as 1PB, after they had been divided into good oocytes that did not undergo rescue ICSI as the second polar body extrusion rate had reached the specified number of ≥80% 6 h after insemination, and poor oocytes that did not undergo rescue ICSI owing to abnormal spindle or poor cytoplasm quality. The 2PN rate of good oocytes was 4.0% and that of poor oocytes was 10.8%. Compared with good oocytes, the abnormal fertilization rates of 1PN (p = 0.019) and 3PN (p < 0.001), as well as 0PN (p < 0.001), were higher in poor oocytes.

| DISCUSS ION
The indications of rescue ICSI were expanded to cases with <80% of oocytes in the cycle having extruded the second polar body 6 h after insemination, but there was no difference in 2PN, blastocyst development, implantation, miscarriage, and live birth rates even in cases F I G U R E 2 Classification of oocytes by the second polar body extrusion rate for IVF cycles with a second polar body extrusion rate of 50% or more. According to Cao et al., the 3PN rate of rescue ICSI increased as the IVF fertilization rate increased, with 8.6% for an IVF fertilization rate of 0%, 11.3% for that of <25%, and 14.5% for that of 25%-50%. 14 In this study, there was no difference in the 3PN rate between a second polar body extrusion rate of <25% and 25%-50%, whereas the 3PN rate increased for that of ≥50% but remained low at 3.0%. Moreover, the 0PN rate for oocytes that did not undergo rescue ICSI despite being 1PB as the second polar body extrusion rate had reached the specified number of ≥80% was 91.9%; many oocytes determined as 1PB were not fertilized. Oocytes that underwent rescue ICSI with a second polar body extrusion rate of 25%-50% accounted for 4.1% of oocytes that underwent IVF, whereas oocytes with a second polar body extrusion rate of 50%-80% accounted for 5.2%. The expansion of rescue ICSI indications to <50% can rescue many oocytes without increasing the 3PN rate, and the expansion to <80% can also rescue many oocytes without increasing the 3PN rate significantly.
An issue with the fertilization of rescue ICSI is the difficulty in determining the number of polar bodies. The implementation of rescue ICSI following the misjudgment of 1PB as 2PB, which has extruded the second polar body, forms 3PN. The misjudgment of the fragmented 1PB as 2PB, even when the second polar body has not been extruded, results in failed fertilization. According to Nagy et al., 18.2% of oocytes with fragmented polar bodies that did not undergo rescue ICSI were fertilized, 20 which exemplifies how difficult it is to accurately determine the number of polar bodies for oocytes with fragmented polar bodies. The 3PN rate of early rescue ICSI has been reported to be 1.6%-9.4%. [6][7][8]17 The 3PN rate of rescue ICSI in this study was 2.2%, which is higher than 1.6% in ICSI but still low. The low 3PN rate is attributed in part to the accurate assessment of second polar body extrusion in oocytes with fragmented polar bodies by additionally assessing the spindle state. Guo et al. have reported the usefulness of observing the spindle in early rescue ICSI and that the 3PN rates were 5.9% when the spindle was observed, and 9.4% when it was not observed. 17 Moon et al. have also reported the usefulness, with a 3PN rate of 4.5% when the spindle is observed and 26.5% when the spindle is not observed on the day after insemination. 21 The second reason for the low 3PN rate was polar body observation and MI/MII oocyte classification during oocyte retrieval. It has been reported that the fertilization rate decreases for a preculture time <3 h, 22,23 and normal fertilization may not progress despite sperm penetration in MI oocytes or oocytes that have just matured into MII. In standard IVF, insemination is performed 3 h after preculture without observing polar body during F I G U R E 3 Classification of polar bodies in IVF cycles and the implementation of rescue ICSI. "Present in the Cytoplasm" corresponds to Figure 1A, "Protruded" corresponds to Figure 1B, "Cross-linked" corresponds to Figure 1C, and "Disappearing" corresponds to Figure 1D oocyte retrieval, but some oocytes in MI during oocyte retrieval do not mature into MII before insemination and mature into MII before denudation. Some of those oocytes do not undergo normal fertilization and do not extrude second polar body even if penetrated by the sperm and may become 3PN if rescue ICSI is performed. In this study, the observation of polar body during oocyte retrieval and exclusion of MI oocytes from the target of rescue ICSI may also have contributed to lower rates of 3PN.
However, the 3PN rate of rescue ICSI was higher than that of ICSI and even higher in cases with a high second polar body extrusion rate. If rescue ICSI is performed on oocytes with a late extrusion of second polar body even if penetrated by the sperm, 3PN may be formed. Among good oocytes that did not undergo rescue ICSI despite being 1PB as the ratio of oocytes having extruded second polar body had reached the specified number, 4.0% of the oocytes were fertilized normally the next day. There are also a few oocytes that are late extruders, extruding the second polar body later than 6 h after insemination; thus, it is considered difficult to completely eliminate unnecessary rescue ICSI. In addition, poor quality oocytes with poor cytoplasmic conditions are fertilized to a certain degree without rescue ICSI, but it is better to exclude them from the target of rescue ICSI because of the large number of abnormally fertilized oocytes.
In rescue ICSI, the decrease in embryo development owing to oocyte aging associated with the delayed timing of fertilization has been a concern. The fertilization and pregnancy rates have reportedly reduced when the time from oocyte retrieval to ICSI was prolonged, [24][25][26] but it has also been reported that no effects were observed up to 9 h 27

ACK N OWLED G M ENTS
None.

CO N FLI C T S O F I NTE R E S T
Takashi Shibahara, Yuu Fukasaku, Nozomi Miyazaki, Hiroaki Kawato, and Hiroyuki Minoura declare that they have no conflict of interest.

H U M A N R I G HT S S TATE M E NT S A N D I N FO R M E D CO N S E NT
All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation

TA B L E 3
Comparison of non-rescue ICSI oocytes by embryo quality (institutional and national) and with the Helsinki declaration of 1964 and its later amendments. Informed consent was obtained from all patients for being included in the study.

E TH I C S A PPROVA L
Approval was obtained from the ethics committee of Minoura Ladies Clinic (Approval number: H21.03). The procedures used in this study adhere to the tenets of the declaration of Helsinki.

I N FO R M E D CO N S E NT
Written patient consent was obtained after providing a thorough explanation.