Population
This retrospective cohort study included PGT cycles with ET in the Department of Reproductive Medicine of the Third Affiliated Hospital, Guangzhou Medical University (Guangzhou, China) between January 1, 2018, and April 30, 2020. The control group involved the non-PGT cycles with other inclusion criteria identical to those for the study group. This study was approved by the ethics committee of the Third Affiliated Hospital of Guangzhou Medical University. The inclusion criteria were as follows: (1) serum β-hCG levels ≥25 mIU/mL 14 days after transfer; and (2) single vitrified-warmed blastocyst transfer (Fig. 1). Cycles with donor sperms or oocytes were excluded from the study (Fig. 1).
Ovarian stimulation and oocyte retrieval
Controlled ovarian stimulation was performed using either long protocol or antagonist protocol. On days 2–4 of the cycle, 150–300 IU follicle-stimulating hormone (Gonal-F, Merck Serono, Italy) was used for ovarian stimulation. Oocyte maturation was triggered with 250 µg recombinant human chorionic gonadotropin (hCG) (Ovidrel, Merck Serono, Italy), and oocyte retrieval was performed 36 h later.
Embryo culture and blastocyst biopsy
All the embryos were cultured to the blastocyte stage (5–6 days after oocyte retrieval), when the TE biopsy was performed on good- and fair-quality embryos. Good-quality embryos were defined as blastocysts of 3–6 AA/AB/BA/BB using the Gardner scoring system. Fair-quality embryos included blastocysts of 4–6 BC/CB [5]. The perforation of zona pellucida was induced with three to five laser pulses (Satum Active Laser System, RI, England). Five to 10 TE cells were aspirated with the biopsy pipet. After washing, the TE biopsy samples were placed in the polymerase chain reaction (PCR) tube with 2 μL of PBS, centrifuged immediately, and stored in a refrigerator at –80℃ for further processing. The biopsied blastocysts were vitrified using a kit (Kitazato Biopharma Co., Ltd., Shizuoka, Japan).
Whole gene amplification and PGT
Whole gene amplification (WGA) of the biopsied samples and parental reference DNA was performed with amplification equipment (GeneQ, Hangzhou Bioer Technology, China) using a kit (SurePLEX WGA, Basecare, China). Next-generation sequencing and single nucleotide polymorphism (SNP) microarray were used for genome testing, with SNP microarray mainly applied for monogenetic diseases. Data were processed using the Genome Studio program (iScan, Illumina, America).
Frozen-thawed blastocyst transfer
Three protocols were used for endometrial preparation: natural cycle, artificial cycle, and ovarian stimulation cycle. Vaginal progesterone (Crinone, Merck Serono, England), 90 mg once a day, was applied for luteal-phase support. The serum β-hCG test was performed 14 days after ET, and luteal-phase support was continued to the tenth week in the case of intrauterine pregnancy.
Hormonal assays
The immunochemiluminometric assay was performed for testing β-hCG (Architech i2000SR; Abbott Laboratories Inc., IL, USA). The range of detection was between 1.2 and 225,000 mIU/mL. The sensitivity of the assay was 1.2 mIU/mL, and the intraassay coefficient of variation was 7%.
Definitions of pregnancy outcomes
Clinical pregnancy was defined as an intrauterine/extrauterine gestational sac detected by ultrasound with positive serum β-hCG. Biochemical pregnancy loss was defined as serum β-HCG level > 25 mIU/mL 14 days after transferring the embryo, which declined to < 5 mIU/mL at the end without any visible gestational sac on ultrasound. Early miscarriage was defined as fetal growth arrest or no cardiac activity detected in the gestational sac during the first 12 weeks of pregnancy. Ongoing pregnancy was defined as the pregnancy continued beyond 12 weeks with a live fetus. Live birth indicated pregnancy continued after 28 weeks of gestation with a live fetus.
Statistical analysis
Since this was a retrospective study, the baseline characteristics were different between the PGT and non-PGT groups. Therefore, propensity score matching (PSM) was used to screen a group of patients so that the baseline characteristics were similar between the two groups. The multiple logistic regression model was used to calculate the propensity score, with PGT as the dependent variable and the variables in Table 1 as independent variables (female age, male age, duration of infertility, number of previous gestations and transfers, AMH level, BMI, endometrial thickness 5 days before transfer, and days of ET). The PSM was performed with 1:2 matching by the nearest neighbor matching, with the caliper width equal to 0.2 of the standard deviation (SD) of the logit of the propensity score (PS). SD was calculated for baseline variables before and after PSM; an absolute value lesser than 0.1 indicated a negligible imbalance.
Statistical analysis was performed using SPSS 22.0 software (IBM, NY, USA). Quantitative variables with homogenous variance were expressed as `X ± SD, and the means were compared using the Student’s t test. A chi-squared test was used to compare rates. Multiple linear regression was applied to analyze the influence of PGT on serum β-hCG levels, while receiver operating characteristic curves (ROC curves) were plotted to predict pregnancy outcomes using serum β-hCG levels. A P value <0.05 was considered statistically significant.