This study was approved by the Ethics Committee of Shiga University of Medical Science (registration number 2014-090). All clinical studies were conducted according to the Declaration of Helsinki for Medical Research involving Human Subjects. Informed consent was obtained from the participants. The subjects were patients under 41 years of age who agreed and underwent the in vitro fertilization-embryo transfer (IVF-ET) protocol of our department, which includes routine hysteroscopy and local endometrial curettage (injury) before first frozen thawed embryo transfer. The patients were enrolled from June 2014 to September 2017. Patients who had a history of RIF, recurrent pregnancy loss (RPL) or suffering from genetic disorders, endocrine and autoimmune diseases, submucosal myoma, adenomyosis, uterine malformation, or endometrial thinning (< 8 mm at implantation phase) were excluded. RIF was defined as the failure of clinical pregnancy after 4 good quality embryo transfers, with at least three fresh or frozen IVF cycles, as per Coughlan et al [19]. RPL was defined as the patient with 3 or more miscarriage [20].
As shown in Fig. 1, IVF or intracytoplasmic sperm injection was performed with a gonadotropin releasing hormone (GnRH) agonist protocol or a GnRH antagonist protocol, and the blastocysts were frozen. The ovulation day was then identified using a urine ovulation test and examination with ultrasonography, and hysteroscopy and endometrial tissue collection were performed 5–9 days after ovulation. Whether endometrial macropolyps and uterine malformations were present was determined by hysteroscopy, and patients with these diseases were excluded from this study at that moment. Immediately after hysteroscopy, the tissue around the center of the anterior endometrium was collected with 4.5 J. A. M. W Type Uterine Curettes, and immunostaining of this sample for CD138 was performed, as reported previously [21, 22]. One pathologist examined this tissue to determine how many plasma cells there were in 10 random HPFs (Olympus BX-41; hpf diameter = 0.55 mm; hpf area = 0.24 mm2) of the endometrial stromal area.
The first aim of the present study was to determine if the pregnancy rate and ongoing pregnancy rate were different when CE was diagnosed with one or more plasma cells found in 10 HPFs of the endometrial stromal tissue at the implantation phase, which exactly matches the histological diagnosis, compared with the patients without any plasma cells found. The next research aim was to examine whether the pregnancy rate, live birth rate, and miscarriage rate differ between Non-CE and CE when the diagnostic criteria for CE are changed. That is, four different criteria for CE were used according to the number of plasma cells: one or more plasma cells in 10 HPFs; 2 or more plasma cells in 10 HPFs; 3 or more plasma cells in 10 HPFs; and 5 or more plasma cells in 10 HPFs. Pregnancy rates, live birth rates, and miscarriage rates were calculated in the Non-CE and CE groups for each criterion.
Blastocysts were transferred within 90 days of endometrial tissue collection. An estradiol patch was started on days 2–3 of menstruation and increased gradually (Fig. 3). When the endometrial thickness reached 8 or more mm, oral administration of dydrorgesterone (DYD) was started (40 mg/day for patients weighing less than 65 kg or 60 mg for those weighing 65 kg or more). Single frozen-thawed blastocyst transfer was performed 6 days after DYD administration, and a blood human chorionic gonadotropin (hCG) test was performed 2 weeks after blastocyst transfer. When hCG was detected, ultrasonography was performed within one week from that day, and those with a gestational sac in utero were defined as pregnant. When pregnancy was recognized, administration of these hormones was continued until 13 to 15 weeks of pregnancy. When the patients did not conceive or miscarriage resulted, their administration was discontinued as appropriate.
The target number of participants in the present study was calculated based on a retrospective study reported by Cicinelli et al [9]. According to their report, when CE was cured with antibiotics, the ongoing pregnancy rate was 60.8% (28/46), but when it persisted, the rate was 13.3% (2/15); there was a significant difference between them. Based on these results, the number of patients required for enrollment was calculated using software provided by the Department of Biostatistics, Vanderbilt University (http://biostat.mc.vanderbilt.edu/wiki/Main/PowerSampleSize). In the section of Dichotomous, independent, prospective, two proportion, and Fisher’s exact test were selected to measure the sample size. An α of 0.05 was selected as the probability of falsely rejecting the null hypothesis, with 0.8 for power (the probability of always rejecting the null hypothesis if the null hypothesis is false in the statistical hypothesis test), 0.605 for P0 (the probability of the outcome for a control patient in prospective studies), and 0.133 for P1 (the probability of the outcome in an experimental subject in prospective studies). When a value of 1 was selected for m (the ratio of control to experimental subjects for independent prospective studies), it was calculated that enrollment of 19 cases was necessary for each group. When m was chosen as 0.56, 0.36, 0.26 and 0.55, the number of cases required for control (Non-CE) and CE became 26 and 15, 33 and 12, 42 and 11, and 26 and 15, respectively.
Statistical analysis was performed using Graph Pad Prism 5 (GraphPad Software Inc., La Jolla, CA). The normality of the distribution of each dataset was analyzed using the Kolmogorov-Smirnov test, and then Student’s t-test or the non-parametric Mann-Whitney U test was used depending on the distribution pattern. The significance of differences in pregnancy, live birth, and abortion rates between the Non-CE and CE groups was examined using Fisher’s test. A significant difference was defined as a P value less than 0.05.