Morphokinetic analysis of early human embryonic development and its relationship to endometriosis resection: a retrospective time-lapse study using the KIDScore™ D3 and D5 implantation data algorithm

237 embryos (endometriosis group: n = 126; non-endometriosis control group: n = 111) undergoing infertility treatment at our clinic from 2014 until 2017 were included in this retrospective study. The time interval was chosen as the same core teams for performing the surgery and the assisted reproductive therapies were responsible in this period. Inclusion criteria were female patients aged between 18 and 45 years undergoing IVF and/ or ICSI treatment. The endometriosis group was divided into two subgroups. It was differentiated between embryos from patients with endometriosis that underwent complete resection (d3 n = 37; d5 n = 36) and embryos from patients with endometriosis that did not underwent complete resection (d3 n = 15; d5 n = 20). If this information was missing, we excluded these embryos.

Endometriosis was confirmed laparoscopically with histological examination of a biopsy. When patients were scheduled for complete resection of their endometriosis the surgery could be performed laparoscopically in all cases. The primary trocar was inserted subumbilically after insufflating the abdominal cavity with CO₂ via a Veress needle. Further 5 mm-trocars were placed in the lower belly appropriate for reaching the individually observed endometriotic lesions. After systematic evaluation of the whole abdominal situs a stage was determined. “Complete resection” was defined as the complete removal of every endoscopically visible endometriotic lesion but did not include systematic excision of an eventual adenomyosis uteri. Describing the stage of endometriosis rASRM (reversed American Society for Reproductive Medicine) score was used [5]. This score was developed by the reversed American Society for Reproductive Medicine. A common score for endometriosis. The rASRM is a scoring system. The score is using points corresponding the size of peritoneal and ovarian endometriosis. The points are also assigned for adhesions. The sum of the points results in a score of four grades.

The in vitro culture was performed in a closed time-lapse incubator (EmbryoScope^®, Vitrolife) up to day 3 or up to day 5.

In addition, we provide a small case series of four patients. They underwent IVF/ ICSI before and after complete endometriosis resection.

Ethical approval was obtained from the Ethics Committee of Würzburg University (file number 20191007 01).

Ovarian hyperstimulation was performed according to the agonist or antagonist protocol as previously described e. g. by Diedrich et al. [17]. Oocytes were aspirated by transvaginal follicle aspiration 36 h after an ovulation induction with 5000 or 10,000 IU human chorionic gonadotropin (HCG, Predalon^® MSD). Aspiration was performed by experienced surgeons specialised in reproduction. In case of endometriosis the aspirate did not have any contact to endometrioma. If necessary, aspiration system was changed. Cumulus-oocyte complexes were collected in SynVitro Flush media (Cooper Surgical, Ref. 15,760,125) at 37 °C and then transferred into Continous Single Culture medium (CSC-C, Irvine Scientific, Ref. 90,165) at 37 °C and 6.8% CO₂. The embryo cultures were covered with oil for embryo culture (Irvine Scientific, Ref. 9305). For IVF oocytes were inseminated with a progressively motile sperm concentration ranging between 0.1 and 0.5 × 10⁶/ml in CSC. Prior to the ICSI procedure cumulus cells were removed from the oocytes with hyaluronidase enzyme (Cumulase^®, Cooper Surgical, Ref. 16,125,000) and denuding pipettes with appropriate lumen sizes (175–135 µm, Vitromed, Jena, Germany, Ref. V-Den-135, − 150, − 175). Introcytoplasmatic sperm injection was performed in multipurpose handling medium (MHM, Irvine Scientific, Ref. 90,166) and polyvinylpyrolidone (PVP, Cooper Surgical, Ref. 10,905,000). After ICSI the oocytes were cultured in CSC-C medium in an Embryo Slide culture dish (Vitrolife, Ref. FR-S-ES-D) for 3 to 5 days without media change. In the case of IVF, fertilized oocytes were transferred to the Embryo Slide culture dishes on day 1 after fertilization check. Embryos were cultured in an EmbryoScope (Vitrolife) at 6.8% CO₂ and 5% O₂ at 37 °C. During embryo culture embryo evaluation was constantly performed according to Instanbul consensus [4]. Briefly, embryo quality was calculated in terms of number of blastomeres, cell fragmentation and symmetry. For the evaluation of KIDScore™ D3 and D5 only the best embryos, which were choosen for transfer or cryopreservation, according to the embryologist’s view were annotated on D3 or D5, depending on the day of embryo transfer. KIDScore analysis was performed using the Embryoviewer software (Vitrolife) and evaluated by applying KIDScore™ D3 (continuous scale from 1 to 5) or KIDScore™ D5 (continuous scale running from 1 to 9.9) according to the day of embryo transfer. Using the morphokinetic KIDScore™ D3 the model assigns a low score to those with the statistically lowest probability of implantation and a higher score to those with a statistically higher probability of implantation. KIDScore™ D5 considers morphology and morphokinetic traits. The higher the score the higher the chance of implantation.

KIDScore™ D3 is based on six annotations: the number of pronuclei equals 2 (2PN), time from insemination to pronuclei fading(tPNf), time from insemination to the 2-cell stage (t2), time from insemination to the 3-cell stage (t3), time from insemination to the 5-cell stage (t5), time from insemination to the 8-cell stage (t8). KIDScore™ D5 is based on 8 annotations: 2PN, t2, t3, t4, t5, timing of full blastocyst (tB), quality of Inner Cell Mass (ICM) (a, b, c) and the quality of the Trophectoderm (a, b, c).

Pregnancy was assessed using vaginal ultrasound. All patients with positive pregnancy test at home get an appointment for ultrasound in the 7th pregnancy week. Life birth rate is defined as number of deliveries per fresh embryo transfer. Pregnancy rate is defined as number of pregnancies per fresh embryo transfer. Abortion rate is defined as number of abortions out of all pregnancies.

The University medical Centre Würzburg is an endometriosis centre level III. Surgery is performed by specialized surgeons. Patients obtained complete resection. Not all hospitals fulfil the qualifications for adequate treatment of endometriosis. Therefore, endometriosis centres were established. In this process, the hospitals are audited and certified by an independent body, EuroEndoCert. This institution uses strict guidelines for certification. There are three stages of centres: endometriosis centre (level I), clinical endometriosis centre (level II), clinical and scientific endometriosis centre (level III) (all: [35]).

The data handling and the statistical operations were performed using Microsoft Excel 2019 (Microsoft, USA) and SPSS 25 (IBM, USA). For non-parametric comparisons of the ordinally scaled KIDScore ™ measurements the Mann–Whitney U-test was used to check for statistical significance. Fisher’s exact test was performed to determine the shown significance levels of the pregnancy rate comparisons. The Computation of the biological effect size “r” according to Cohen was performed as described by Fritz et al., 2012 the computed test statistic of the mentioned statistical test in SPSS and the respective case number. p-values < 0.05 were considered statistically significant (Fig. 1).

In this study, we enrolled 84 patients, 43 with histologically proven endometriosis and 41 without endometriosis (control group, laparoscopically confirmed). In total we had data from 261 embryos. We analysed 237 cultured embryos, 126 from patients with endometriosis and 111 from patients without endometriosis. A total of 128 single or double embryo transfers were performed. For 97 embryos we annotated a D3 score, and for 140 embryos we annotated a D5 score. The difference is statistically significant. The embryos were assigned to one of the four groups depending on the day of transfer and on the confirmation of endometriosis. Age, BMI, stimulation cycles, oocyte retrieval and transferred embryos per cycle of these four groups are displayed in Table 1.

We observed a significant difference concerning body mass index (BMI) of endometriosis group D5 and control group D5 (p = 0.005) (Table 1). In addition, significantly fewer oocytes were aspirated per cycle in endometriosis group D3 than in endometriosis group D5 and control group D5 (p = 0.001, p = 0.009, respectively) and fewer embryos were transferred per cycle than in endometriosis group D5 (p = 0.028). The rate of live births per performed IVF/ICSI cycle showed no significant difference in the four groups (Table 1).

Figure 2 shows the distribution of KIDScore™ measured on day 3 and day 5 for embryos from patients with and without endometriosis (regardless of whether the endometriosis was completely resected or not). On day 3 the median for endometriosis patients was 4 (on a scale from 1 to 5) as well as the median for non-endometriosis patients (control group). However, on day 5 a trend towards a lower embryo quality measured by time lapse monitoring was observed between the mentioned groups (p = 0.175): embryos from patients without endometriosis showed a higher KIDScore™ D5 (median 6.8) (on a scale from 1 to 9.9) than embryos from patients with endometriosis (median 6.5).

Figure 3 also covers KIDScore™ day 3 and KIDScore™ day 5, but here the endometriosis group was divided into two subgroups. One consisted of embryos from patients with endometriosis who underwent complete resection of their endometric lesions. The other group contained only embryos from patients with histologically confirmed endometriosis and not being treated by complete resection. Of note, for analyses of day 3 11 embryos had to be excluded from further computations as it was not clearly documented that the complete surgery was performed as described in material and methods; for day 5 10 embryos had to be excluded. For KIDScore™ day 3 no significant differences in the embryo quality could be observed. In contrast to that KIDScore™ D5 showed a statistical highly significant difference: the median KIDScore™ D5 of embryos from patients without complete resection of their endometriosis was 2.6. In comparison to the control group which showed a median of 6.8 a significant decrease in embryo quality was recorded (p = 0.003). When the endometriosis was completely removed surgically, a median score of 7.2 was achieved. This indicates a highly significant increase (p = 0.002) in embryo quality compared to the KIDScore™ D5 patients without complete removal and is not differing significantly from the not-endometriosis bearing control group (p = 0.911). Computation of the biological effect size according to Cohen showed a moderate up to strong biological effect (r = 0.4) for endometriosis complete resection vs. no resection and a moderate biological effect (r = 0.3) for no endometriosis vs. endometriosis without resection.

Interestingly, complete resection also had a small impact on pregnancy and abortion rate.

Patients that underwent complete resection showed a similar pregnancy rate (32.6%) compared to patients without endometriosis (32.8%), (p = 1.000). Patients with endometriosis and no resection showed a trend towards a lower pregnancy rate (17.6%) compared to patients that underwent complete resection (p = 0.346) and also compared to patients without endometriosis (p = 0.370). Statistical difference was not observed.

After complete resection, the abortion rate was 44.4%. In contrast, in case of endometriosis without resection—although only two patients met this criterion—the abortion rate was 100%. The difference did not reach statistical significance (p = 0.455). The abortion rate in the group of women without endometriosis (42.9%) compared to women with non-resected endometriosis (p = 0.217) and compared to women with resected endometriosis (p = 1.000) did not differ significantly from each other.

All: (Fig. 4).

Figure 5 shows a small case series of four patients suffering from endometriosis and undergoing IVF-treatment before and after the complete resection of their endometriosis.

Patient 1 was 41 years old when she underwent her first ICSI treatment. Her BMI was 21.3 kg/m². She had her first ICSI in March 2014 and was stimulated under protection of a GnRH analogon. Seven oocytes were retrieved, and two embryos were transferred. KIDScore™ day 3 was 4 and 5. There was no pregnancy achieved in this cycle. She started her next cycle in August 2014 with the same protocol. 11 oocytes were aspirated, two embryos were transferred. KIDScore™ day 5 was 8.2 and 2.6. ß-HCG was negative again. In November 2014 she underwent laparoscopy. Endometriosis rASRM II was diagnosed. Lesions were resected completely. Afterwards she had another ICSI in October 2015 stimulated with GnRH antagonist protocol. 5 oocytes were aspirated, 2 embryos were transferred. KIDScore™ day 3 was 5 and 2. Unfortunately, ß-HCG was negative again.

Patient 2 was 32 years old when she underwent her first ART treatment. Her BMI was 20.9 kg/m². She already had her first diagnostic surgery because of endometriosis in 2004. From 2012 to 2014 she underwent many cycles of IVF and ICSI. The first ART in which the embryos were evaluated with time-lapse monitoring was in April 2014. She was stimulated with GnRH agonist protocol, 15 oocytes were aspirated, and 2 embryos transferred. KIDScore™ day 3 was 2 and 4. ß-HCG was negative. In September 2014 she underwent complete resection of endometriosis rASRM II. Two months later she went for another ICSI cycle, stimulated with GnRH antagonist protocol. 8 oocytes were aspirated, and 2 embryos transferred. KIDScore™ day 5 was 8.4 and 7.7. ß-HCG was positive, and she delivered one child.

Patient 3 was 31 years old when she started ART at our hospital. She already has been pregnant but interrupted the pregnancy because of a trisomia. Her BMI was 25.3 kg/m². She did not lose weight through the years of treatment. She underwent diagnostic surgery because of endometriosis in March 2014. In May 2014 she had an ICSI, stimulated with GnRH agonist protocol. 12 oocytes were retrieved, and 2 embryos transferred. KIDScore™ day 5 was 7.2 and 6.8. ß-HCG was positive. Unfortunately, she had a spontaneous abortion. In November 2014 she started her next ICSI with the same stimulation protocol. 14 oocytes were aspirated, 2 embryos transferred, KIDScore™ day 5 was 1.6 for both embryos but ß-HCG was negative. The following ICSI-treatment was in February 2015. With the use of the same protocol for stimulation again 14 oocytes were retrieved, and 2 embryos transferred. KIDScore™ day 5 was 1.7 and 1.6, ß-HCG was negative again. In July 2015 she had another ICSI. With GnRH agonist protocol 9 oocytes were retrieved, and 2 embryos transferred. KIDScore™ day 5 was 1.7 and 6.5. ß-HCG was positive, but pregnancy ended in a spontaneous abortion. In September 2015 she underwent complete resection of deep infiltrating endometriosis ENZIAN A3 B2 C1. Shortly afterwards she started a new cycle of ICSI, but unfortunately we do not have the KIDScore™ of the two transferred embryos. The following ICSI treatment was in September 2016. 7 oocytes were retrieved, 2 embryos transferred, KIDScore™ day 3 was 4 for both embryos. ß-HCG was negative. She had another ICSI in December 2017. 8 oocytes were aspirated, 2 embryos transferred, KIDScore™ day 5 was 7.2 and 1.7 and ß-HCG negative.

Patient 4 was 39 years old when she started her first ICSI treatment. Her BMI was 24.2 kg/m². In August 2014 she was stimulated with GnRH agonist protocol, only 5 oocytes were retrieved, KIDScore™ day 3 was 4 and ß-HCG was negative. In October 2014 she underwent complete resection of endometriosis rASRM II. Afterwards she was stimulated with the ultralong protocol in January 2015. Again only 5 oocytes were retrieved, KIDScore™ day 3 was 5 and ß-HCG was negative.

Patient 2, 3 and 4 showed an increasing KIDScore™/ increasing embryo quality after complete resection, whereas patient 1 showed a decreasing KIDScore™/ embryo quality after complete resection (Fig. 5).