Cohort Profile: CITIC-Xiangya Assisted Reproductive Technology Cohort (CXART Cohort)

doi:10.21203/rs.3.rs-2205266/v1

Download PDF

Research Article

Cohort Profile: CITIC-Xiangya Assisted Reproductive Technology Cohort (CXART Cohort)

https://doi.org/10.21203/rs.3.rs-2205266/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

To address the data source gap between the wide use of assisted reproductive technology (ART) and deficiencies in clinical and biological research, the CITIC-Xiangya Assisted Reproductive Technology (CXART) Cohort began to enroll infertile patients who attended the Reproductive and Genetic Hospital of CITIC-Xiangya for ART treatment since January 2016, and enrollment is expected to be completed in January 2026. By November 30, 2020, 223,494 ART cycles were recorded from 119,590 infertile couples. In addition to the detailed medical, treatment and laboratory records, examination and detection results, billing information captured from the hospital’s pre-designed and ART-specialized electronic medical records system; and follow-up data at four time points (14 and 28 days after transfer, 42 days and 1 year after delivery) collected by the active follow-up team, we also investigated lifestyle habits, environmental exposure, dietary nutrition and psychological emergencies of 26,747 infertile couples by proactive investigation. To date, a total of 71,689 biological samples have been collected from infertile couples, including peripheral blood, follicular fluid, cumulus cells, and semen. The cohort has the capacity to examine the causes and elucidate the mechanisms of infertility, novel therapeutic targets, and better reproductive outcomes, and provides a theoretical basis for the exploration of personalized precision medicine, especially in Chinese populations.

infertility

clinical cohort study

registry

record linkage

biobank

Infertility affects one in six couples [1], representing an important and increasingly prevalent public health problem [2]. Assisted reproductive technology (ART) has become an effective procedure for infertile patients since resulting in its first live birth in 1978 [3], with over 8 million children born worldwide through ART [4]. The short- and long-term health of children conceived following ART and their mothers are of substantial public health interest [5]. Although the reported estimated cumulative delivery rate increased from 27.1% in 2010 to 32.1% in 2014 [6], owing to inherent imperfections of human reproduction and possible methodological issues, ART has not yet reached the desired level of efficacy, which also constitutes the grounds for the need for continuous monitoring and research [7].

To improve clinical practice and optimize the chances of infertile patients achieving parenthood, more evidence is needed to meet the priorities for future infertility research [8]. This includes clinical research and mechanistic exploration, such as risk factors and etiology of unexplained infertility [9], optimization of clinical outcomes in ART [10], fertility preservation [11], and intervention strategies for reproductive aging [12]. These questions were barriers to improving care for those with fertility problems. Although prospective randomized trials are useful tools for evaluating treatment effects, they often face ethical and resource challenges, and are underpowered for subpopulations or inaccessible for evaluations of long-term adverse outcomes [13]. Real-world data are a valuable adjunct to traditional clinical trials as they have the potential to demonstrate ongoing changes under real-life conditions, and provide more evidence on effectiveness, safety, adherence, and disease burden in real-world settings. In addition, high-quality traceable specialist biobanks are key resources for exploring pathogenic mechanisms in the field of reproductive health. Comprehensive case resources combined with the integrated analysis of clinical data provide a reliable research basis for future personalized diagnosis and treatment.

ART in mainland China has developed rapidly and dramatically. A cross-sectional nationwide survey indicated that in 2016, there were 906,840 in vitro fertilization (IVF) cycles performed in mainland China, of which 311,309 infants were born, accounting for 1.7% of the total number of infants born in the same year [14]. However, China does not have a nationwide cycle-based ART registry, and almost no large-scale cohorts or biobanks based on the national infertility-specific population have been reported. The CITIC-Xiangya Assisted Reproductive Technology (CXART) Cohort is a longitudinal, prospective, ongoing cohort with healthcare data and extensive biomaterial collection at the Reproductive and Genetic Hospital of CITIC-Xiangya, a regional ART center in South China. In 2016, the CXART cohort carried out 41,000 ART procedures, representing approximately a quarter of the annual number for the entire United States [15]. To better understand the processes and outcomes of healthcare for patients with infertility, we initiated this cohort study in 2016, enrolling infertile patients with their clinical and follow-up data, as well as biological samples. The cohort aims to develop a large data source within the field of reproduction. It could have the capacity to examine the causes and elucidate mechanisms of infertility, novel therapeutic targets, and better reproductive outcomes, and provide a theoretical basis for the exploration of personalized precision medicine.

Overview of cohort

This longitudinal cohort study was conducted at the CITIC-Xiangya Reproductive and Genetics Hospital (Figure 1). The study was approved by the Institutional Review Board of the hospital (LL-SC-2022014) and registered at ClinicalTrial.gov (NCT05404464). A multidisciplinary research team was developed for the cohort, including epidemiologists, statisticians, reproductive and genetic clinicians, follow-up working groups, sample managers, and information engineers.

All couples with infertility who visited our infertility clinic between January 2016 and January 2026 were and will continue to be regularly screened for registration by well-trained nurses. Participants who completed the preoperative examinations and were eligible for ART were and will be included in the cohort after full informed consent.

Assisted reproductive treatment techniques were limited to IVF, intracytoplasmic sperm injection (ICSI), and preimplantation genetic testing (PGT). In addition, CITIC-Xiangya participated in the China National Birth Cohort (CNBC) study, in which we developed a cohort of 2,745 families receiving ART in 2017 [16]. Accordingly, we carried out an additional questionnaire in the CXART cohort, thus constituting a CNBC sub-cohort. The collection plan of biological samples began in 2016; however, the CITIC-Xiangya biobank with complete procedures and facilities was constructed since 2019, participants with biomaterial formed Biobank sub-cohort based on the main cohort (Figure 1).

ART process and follow up

Figure 1 illustrates the overall process of the CXART cohort development. Couples with infertility at the first visit entered the fresh ET cycle first, and non-first-visit couples may proceed directly to a frozen embryo transfer (ET) cycle in accordance with a standard protocol. Patients with a fresh cycle generally needed to undergo preoperative examination, ovulation induction, triggering, oocyte retrieval, sperm retrieval, insemination, embryo/blastocyst culture, and embryo transfer. Patients who cancelled fresh ET or used PGT underwent frozen ET. Those who failed to achieve live birth in the last transfer considered entering a frozen or fresh ET cycle, according to the number and quality of frozen embryos.

After the transfer, in addition to patient self-reports, four rounds of follow-ups were conducted by the active follow-up team at 14 and 28 days after transfer, and 42 days and 1 year after delivery, respectively. Human chorionic gonadotropin (HCG) and the first B-ultrasound testing were mainly completed in the hospital, and subsequent examinations, as well as delivery records, were required to uploading to the hospital WeChat official account or mini grogram. All patients were encouraged to report any abnormalities during pregnancy, and these details and any subsequent medical records were documented in the follow-up system (FUS), a module built into the electronic medical records (EMR) system. Unreported pregnancy and delivery information was obtained by the follow-up staff via telephone or online at corresponding time points. The growth and development data of the offspring at 1 year of age were obtained through telephone follow-up and recorded in the FUS.

Biological sample collection

We designed a biological sample collection protocol for this cohort with reference to the bioethical principle of non-maleficence. The remaining samples from the couples with infertility were collected after informed consent was obtained during ART treatment on the day of oocyte retrieval, including anticoagulated peripheral blood, follicular fluid, cumulus cells, and semen (Figure 1). Samples were collected by clinical departments and sent to the biobank for standardized processing as follows: peripheral blood from the couples was separated into blood cells and plasma by centrifugation; transparent follicular fluid was collected within 30 minutes after oocyte retrieval; the cumulus cells were collected within 30 minutes after degranulation and frozen in the TRIZOL (ambion, Life Technologies); and the semen was frozen directly after insemination. All samples were processed by professionals and stored at -80°C in the CITIC-Xiangya biobank.

The patient and cycle information were automatically entered into the RuRo sample-management system by scanning of the bar code on each processed sample, and the sample-related information was manually entered by sample managers. A unique identification code was generated and marked on the corresponding sample, which was then stored according to the storage location assigned by the system. During the storage period of the samples, information and biological quality controls were carried out regularly, and the later use or destruction of the samples was supervised and their records were updated over time.

Data collection

To establish the cohort, managed by the Clinical Data Center and Scientific Research Department since 2016, the Reproductive and Genetic Hospital of CITIC-Xiangya has successively cooperated with Qifeng Technology Co., Ltd. and Hanyun Medical Information Technology Co., Ltd. to build a pre-designed and ART-specialized EMR system. Each couple is bound with a unique patient ID (PID) number through the resident ID card. All the data are recorded in cycles, and each cycle is assigned a unique cycle number based on the course of treatment.

Figure 1 shows the process of longitudinal data collection, and the details of the collected variables are provided in Supplementary Table S1. During the infertility clinic phase, the patients’ socio-demographics, personal history, physical examination, and infertility diagnosis were collected from the health information system (HIS). In addition, CNBC sub-cohort participants also completed questionnaires through the hospital WeChat official account or mini grogram, including multiple standard scales, such as the Pittsburgh Sleep Quality Index (PSQI), Perceived Stress Scale-10 (PSS-10), Center for Epidemiologic Studies Depression Scale (CESD), and Self-rating Anxiety Scale (SAS).

During the cycle, data on cycle information, preprocessing, stimulation, trigger, embryo transfer, and luteal support were collected from the HIS. In terms of examination and detection, we actively collected test data on biochemistry, immunology, endocrinology, semen, blastula biopsy, genetics, and chromosomes from the laboratory information system (LIS) as well as detection data on B-ultrasound, salpingography, and hysteroscopy from the picture archiving and communication system (PACS).

We collected laboratory records from the HIS, including sperm acquisition and processing, oocyte retrieval, fertilization and embryo culture, blastocyst culture, and embryo freezing and thawing. Expense information on the ART process was collected from enterprise resource planning (EPR). Moreover, the Biobank sub-cohort also collected data from the RuRo sample management system, including patient and cycle information, sampling date, quantity, source, location and usage. Follow-up data on implantation, clinical pregnancy, perinatal, neonatal outcomes after ET, as well as offspring growth and development information were extracted from the FUS.

Data extraction, linkage and validation

According to the requirements of the research team, anonymized target data from the EMR, questionnaires, and biobank were extracted and organized into pre-designed structured format by experienced medical informatics experts. We randomly selected 500 cycles to review and validate the consistency of the extracted data with the raw data. The data extraction codes were adjusted after the assessment, and iterative data extraction processes were performed until all the extracted data were 100 % consistent and complete for any given sample.

The EMR system automatically linked the HIS, LIS, PACS, EPR and FUS modules through the unique PID and ART cycle numbers. There were 162 couples who submitted the questionnaire twice, and we only retained the most complete questionnaire data from them. Questionnaire data and EMR data were linked in two steps. First, accurate matching was performed according to PID, further fuzzy matching was performed according to the automatically generated questionnaire survey date and cycle start date, so as to assign the data of each questionnaire to the closest cycle of the corresponding couple. Data from the biobank were linked to the EMR data by the PID and ART cycle numbers. For 122 cycles with missing or conflicting ART cycle numbers, we went back to the original records and verified them based on other variables, including abbreviated name, archive number, progress notes, dates of cycle start, oocyte retrieval, ET, and delivery. To further identify a unique cycle in the cohort, we developed a cohort ID based on information regarding the PID, cycle number, and cycle start date.

Logical checks were automatically performed to ensure the accuracy of the data. Manual checks were conducted by the quality control practitioners through monthly chart reviews. Furthermore, data were cleaned using transparent and prespecified rules, including the standardization of medical texts, construction of variable dictionaries, and approaches for missing data and outliers. The diagnostic criteria for clinical diseases, such as infertility and endometriosis, and the calculation of ART laboratory performance indicators, such as the proportion of oocytes recovered, cleavage rate, blastocyst development rate, were based on the corresponding international or Chinese standards/consensus [17-19]. The definition of the infertility core result set adhered to standardized definitions and reporting guidelines [20].

Baseline characteristics of study population

By November 30, 2020, 119,590 couples with infertility were enrolled, of whom 73,644 and 26,747 couples were simultaneously recruited to the Biobank and CNBC sub-cohorts, respectively. Moreover, from 15,144 couples, 32,523 peripheral blood, 17,412 follicular fluid, 6557 cumulus cell, and 15,197 semen samples were collected (Table 1). The registered patients were distributed across all 34 province-level administrative divisions in China (Fig. 2), the top five represented provinces were Hunan (57.21%, where the hospital located), Jiangxi (12.57%), Guizhou (12.33%), Hubei (5.07%) and Fujian (2.46%).

Table 1. Type and number of samples collected.

Types of samples	Types of statistics	Number
Peripheral blood	Samples	32,523
	Patients (female)	15,073
	Patients (male)	14,283
Follicular Fluid	Samples	17,412
	Patients	14,585
Cumulus cells	Samples	6,557
	Patients	5,637
Seminal fluid	Samples	15,197
	Patients	13,064

Table 2. Baseline characteristics of participants from the CXART cohort and two sub-cohorts.

Variables	CXART cohort		CNBC sub-cohort		Biobank sub-cohort
Variables	Total number	Median (Q_L-Q_U) / No. (%)	Total number	Median (Q_L-Q_U) / No. (%)	Total number	Median (Q_L-Q_U) / No. (%)
Recruitment	119,590		26,747		15,073
Female age (years)	119,590	31 (28, 36)	26,747	31 (28, 35)	15,073	31 (29, 35)
Male age (years)	112,651	33 (30, 38)	26,714	33 (30, 38)	15,049	33 (30, 37)
Female BMI (kg/m²)	119,588	21.78 (20.03, 23.53)	26,747	21.64 (19.98, 23.42)	15,073	21.77 (20.00, 23.51)
Male BMI (kg/m²)	108,610	24.04 (21.80, 26.26)	26,572	24.19 (21.80, 26.30)	14,764	24.21 (21.97, 26.45)
Female education level (years)	110,260		26,670		14,808
<6		4588 (4.2%)		945 (3.5%)		696 (4.7%)
7–9		39,134 (35.5%)		8701 (32.6%)		4721 (31.9%)
10–12		23,814 (21.6%)		6701 (25.1%)		3126 (21.1%)
>12		42,395 (38.5%)		10,315 (38.7%)		6259 (42.3%)
Unclear		329 (0.3%)		8 (< 0.1%)		6 (< 0.1%)
Male education level (years)	110,200		26,662		14,808
<6		2969 (2.7%)		650 (2.4%)		379 (2.6%)
7–9		37,049 (33.6%)		8601 (32.3%)		4624 (31.2%)
10–12		23,979 (21.8%)		6281 (23.6%)		3234 (21.8%)
>12		45,868 (41.6%)		11,123 (41.7%)		6565 (44.3%)
Unclear		335 (0.3%)		7 (< 0.1%)		6 (< 0.1%)
Current smoking (Female)	110,429		26,672		14,802
Yes		1605 (1.5%)		450 (1.7%)		321 (2.2%)
No		108,824 (98.5%)		26,222 (98.3%)		14,481 (97.8%)
Current smoking (Male)	109,836		26,684		14,801
Yes		34,578 (31.5%)		10,439 (39.1%)		6061 (40.9%)
No		75,258 (68.5%)		16,245 (60.9%)		8785 (59.1%)
Infertility years (years)	118,515	3 (2, 5)	26,545	3 (1, 5)	14,658	3 (2, 5)
Infertility type	113,992		26,733		15,019
Primary		44,034 (38.6%)		10,295 (38.5%)		5976 (39.8%)
Secondary		69,958 (61.4%)		16,438 (61.5%)		9043 (60.2%)
Infertility related diseases	119,590		26,747		15,073
Fallopian tube abnormalities		90,530 (75.7%)		21,772 (81.4%)		12,326 (81.8%)
Severe oligoasthenozoospermia		22,912 (19.2%)		4847 (18.1%)		3131 (20.8%)
Polycystic ovary syndrome		18,895 (15.8%)		4968 (18.6%)		2802 (18.6%)
Endometriosis		12,415 (10.4%)		2970 (11.1%)		1905 (12.6%)
Monogenic/chromosomal abnormalities		10,388 (8.7%)		2768 (10.4%)		1384 (9.2%)
Moderate to severe uterine mucosa		7534 (6.3%)		2375 (8.9%)		1070 (7.1%)
Recurrent miscarriage		6458 (5.4%)		1620 (6.1%)		917 (6.1%)
AFC	116,942	17 (10, 29)	26,747	18 (10, 30)	15,073	19 (11, 30)
Basic AMH (ng/ml)	116,179	3.59 (1.76, 6.57)	26,699	3.86 (1.88, 7.03)	14,933	3.58 (1.79, 6.59)
Basic FSH (mIU/mL)	116,387	5.89 (4.97, 7.11)	26,698	5.82 (4.93, 6.99)	14,722	6.07 (5.07, 7.42)
Basic LH (mIU/mL)	116,761	3.52 (2.57, 4.91)	26,682	3.41 (2.51, 4.71)	14793	3.75 (2.68, 5.30)
Basic E2 (pg/mL)	116,756	35 (27, 47)	26,683	35 (27, 45)	14,792	37 (28, 51)
Frequency of oocyte retrieval	119,590		26,747		15,073
0		8641 (7.2%)		39 (0.1%)		83 (0.5%)
1		86,862 (72.6%)		20,344 (76.1%)		12,409 (82.3%)
≧ 2		24,087 (20.1%)		6364 (23.8%)		2581 (17.1%)
Frequency of embryo transplantation	119,590		26,747		15,073
0		14,869 (12.4%)		3065 (11.5%)		3678 (24.4%)
1		72,059 (60.3%)		16,248 (60.7%)		8730 (57.9%)
≧ 2		32,662 (27.3%)		7434 (27.8%)		2665 (17.7%)
Abbreviations: AFC, antral follicle count; AMH, anti- Müllerian hormone; BMI, body mass index; CNBC, China national birth cohort; CXART, CITIC-Xiangya Assisted Reproductive Technology Cohort; E2, estradiol; LH, luteinizing hormone; FSH, follicle stimulating hormone.

Table 3. Baseline characteristics of cycles from the CXART cohort and two sub-cohorts.

Variables	CXART cohort		CNBC sub-cohort		Biobank sub-cohort
Variables	Total number	Median (Q_L-Q_U) / No. (%)	Total number	Median (Q_L-Q_U) / No. (%)	Total number	Median (Q_L-Q_U) / No. (%)
Oocyte retrieval cycle	144,157		35,316		18,226
Ovulation inducing scheme	144,148		35,316		18,226
Agonist regimens		84,270 (58.5%)		19,975 (56.6%)		9624 (52.8%)
Antagonist regimens		37,044 (25.7%)		8617 (24.4%)		5259 (28.9%)
Others		22,834 (15.8%)		6724 (19.0%)		3343 (18.3%)
OHSS	144,157		35,316		18,226
Yes		4090 (2.8%)		1010 (2.9%)		786 (4.3%)
No		140,067 (97.2%)		34,306 (97.1%)		17,440 (95.7%)
Types of ART	143,892		35,252		18,164
IVF		85,714 (59.6%)		20,126 (57.1%)		10,696 (58.9%)
ICSI		32,319 (22.5%)		7669 (21.8%)		3091 (17.0%)
IVF + ICSI		6372 (4.4%)		2567 (7.3%)		1580 (8.7%)
PGT		19,487 (13.5%)		4890 (13.9%)		2797 (15.4%)
Source of sperm	144,157		35,316		18,226
Autologous		134,611 (93.4%)		33,591 (95.1%)		17,428 (96.6%)
Donor		9546 (6.6%)		1,725 (4.9%)		798 (4.4%)
Number of oocytes recovered	144,157	9 (5, 14)	35,316	9 (5, 14)	18,226	10 (6, 15)
Number of cleavage embryos	144,157	7 (3, 11)	35,316	7 (3, 11)	18,226	8 (4, 12)
Blastocyst culture	144,157		35,316		18,226
Yes		64,696 (44.9%)		15,976 (45.2%)		8436 (46.3%)
No		79,461 (55.1%)		19,340 (54.8%)		9790 (53.7%)
Embryo transfer cycle	148,658		33,453		14,710
Type of cycle	148,658		33,453		14,710
Fresh embryo transfer		72,689 (48.9%)		16,494 (49.3%)		7095 (48.2%)
Frozen embryo transfer		75,969 (51.1%)		16,959 (50.7%)		7615 (51.8%)
Type of transplant	148,658		33,453		14,710
Cleavage stage		79,133 (53.2%)		18,445 (55.1%)		7839 (53.3%)
Blastocyst stage		65,847 (44.3%)		14,284 (42.7%)		6574 (44.7%)
Cleavage + Blastocyst		3678 (2.5%)		724 (2.2%)		297 (2.0%)
Number of embryos transferred	148,658		33,453		14,710
1		49,158 (33.1%)		10,825 (32.4%)		4864 (33.1%)
2		99,500 (66.9%)		22,628 (67.6%)		9846 (66.9%)
Abbreviations: ART, assisted reproductive technology; CNBC, China national birth cohort; CXART, CITIC-Xiangya Assisted Reproductive Technology Cohort; ICSI, intracytoplasmic sperm injection; IVF, in vitro fertilization; OHSS, ovarian hyper-stimulation syndrome; PGT, preimplantation genetic testing.

Table 4. Preliminary outcomes of the CXART cohort.

Variables	No.	Median (Q_L-Q_U) / No. (%)
Pregnancy outcomes of transplant cycles	148,658
Singleton pregnancy		62,145 (41.8%)
Twin pregnancy		25,697 (17.3%)
Multiple pregnancy		902 (0.6%)
No pregnancy		59,914 (40.3%)
Outcomes of clinical pregnancy cycles	88,553
Singleton live birth		54,952 (62.1%)
Twin live births		17,967 (20.3%)
Multiple live births		90 (0.001%)
Ectopic pregnancy		1511 (1.7%)
Early miscarriage		10,838 (12.2%)
Middle and late abortion		3523 (4.0%)
Induce labor		39 (< 0.001%)
Stillbirth		46 (< 0.001%)
Outcomes of live birth cycles	73,009
Gestational diabetes	55,029	11,839 (21.5%)
Gestational hypertension	54,588	3568 (6.5%)
Mode of delivery	72,728
Natural childbirth		18,208 (25.0%)
Cesarean section		54,520 (75.0%)
Preterm birth	72,870	12,058 (16.5%)
Gestational week of delivery (weeks)	68,467	38.5 (37.3, 39.4)
Outcomes of live births	91,156
Birth weight	89,805
Singleton birth weight (kg)	54,326	3.3 (3.0, 3.6)
Twin birth weight (kg)	35,227	2.5 (2.2, 2.8)
Multiple birth weight (kg)	252	1.8 (1.6, 2.0)
Follow-up at 42 days postpartum
Feeding method	88,924
Breastfeeding		47,026 (52.9%)
Formula feeding		23,304 (26.2%)
Mixed		18,594 (20.9%)
Follow-up of offspring at 1 year postpartum
Weight (kg)	30,857	10.0 (9.0, 10.5)
Height (cm)	18,946	75.0 (73.0, 78.0)

Table 2 presents the baseline characteristics of all participants. The median female and male age at recruitment were 31 (28, 36) and 33 (30, 38) years in the main cohort, respectively. The body mass index (BMI) and educational background of the couples were comparable across cohorts. Women rarely smoked, and the proportion of male smokers were 31.5%, 39.1%, and 40.9% in the cohorts, respectively. The median infertility year was 3, with > 60% of the women with infertility having secondary infertility. In terms of infertility-related diseases, the most common diagnosis was fallopian tube abnormalities, which accounted for > 76% across all the cohorts. The medians for antral follicle count (AFC) and basic endocrine measures, including levels of anti-Müllerian hormone (AMH), follicle stimulating hormone (FSH), and luteinizing hormone (LH), among the three cohorts were similar. In the CXART cohort, 7.2% of participants had no documented oocyte retrieval and used donor oocytes or previously frozen embryos, compared with < 0.5% in the two sub-cohorts. Most patients had only one oocyte retrieval, the proportions of those having ≥ 2 oocyte retrieval records were 20.1%, 23.8%, and 17.1% in the main- and sub-cohorts, respectively. Overall, 12.4% of the patients had no record of embryo transfer in the main cohort, and the proportion was higher in the Biobank sub-cohort at 24.4%.

Cycle characteristics

The baseline cycle characteristics of the main cohort and two sub-cohorts are shown in Table 3. Data for 223,494 cycles of 119,590 couples with infertility were recorded. The final study cohort included 144,157 oocyte retrieval cycles from 110,949 patients, and 148,658 transfer cycles from 10,4721 patients. For oocyte retrieval cycles of the main and sub-cohorts, agonist stimulation was the most commonly used regimen, ranging from 52.8–58.5%, the reporting rate of ovarian hyper stimulation syndrome (OHSS) ranged from 2.8–4.3%. IVF was the most commonly used ART, followed by ICSI and PGT. The proportion of donor sperm was lowest in the Biobank sub-cohort (4.4%). The median number of oocytes retrieved and cleaved were 9 (5, 14) and 7 (3, 11), respectively, for the main cohort. The proportion of blastocyst cultures in each cohort was similar. For transfer cycles, the proportion of fresh or frozen embryo transfer in each cohort was close to 50%. Cleavage ET was slightly more common than blastocyst ET in all cohorts. Approximately one-third of the cycles in each cohort were single ETs.

Preliminary outcomes

Through a combination of self-reporting and active follow-up, we tracked and recorded pregnancies across 148,658 transfer cycles. As shown in Table 4, clinical pregnancies were achieved in 59.7% of transfers, with singletons, twins, and multiples accounting for 41.8%, 17.3%, and 0.6%, respectively. Among these clinical pregnancy cycles, pregnancy loss occurred in 17.6%, with an early miscarriage rate of 12.2%, and a mid-to-late pregnancy loss rate of 4.0%. Live births were achieved in 73,009 (82.4%) clinical gestations, resulting in a total of 91,156 births, mostly by cesarean section (75.0%). Of these, 21.5% and 6.5% of pregnant women experienced gestational diabetes and gestational hypertension, respectively. The median gestational age at delivery was 38.5 (37.3, 39.4) weeks, with 16.5% of preterm births. Data of birth weight showed the median birth weights were 3.3 (3.0, 3.6), 2.5 (2.2, 2.8), and 1.8 (1.6, 2.0) kg for singletons, twins, and multiples, respectively. Follow-up at approximately 42 days postpartum showed that 52.9% of newborns were breastfeeding and 26.2% were formula feeding. The weight of 30,857 offspring and the height of 18,946 offspring were successfully followed up at 1 year postpartum, and the medians were 10.0 (9.0, 10.5) kg and 75.0 (73.0, 78.0) cm, respectively.

Summary of cohort

The CXART cohort is an ongoing longitudinal cohort that accesses a large and comprehensive list of healthcare information from patients with infertility, spanning from assisted reproductive therapy, pregnancy, postpartum, and childhood follow-up periods. In the past five years (January 2016 to November 2020), we collected information on 223,494 cycles from 119,590 couples with infertility, and the data volume was substantial, with over 900 variables being documented. Similar to national registry databases, such as the USA Society for Assisted Reproductive Technology (SART) and European IVF-monitoring Consortium (EIM) [13], CXART provides a more complete set of information about personal histories, descriptions, physical examinations, laboratory tests, imaging results, diagnoses, medication usage, and treatment processes documented according to consistent workflow and formats. The complete tracking of transfer outcomes by the cohort also bridges gaps in data from reproductive and obstetric institutions in China. Moreover, the CNBC sub-cohort collected preoperative survey data of 26,747 couples with infertility, and the Biobank sub-cohort consisted of 71,689 biological samples collected from registered couples with infertility, and further enriching the resources and improving the academic value of the main cohort.

Strengths and weaknesses

The CXART cohort has great potential for answering a wide range of clinical and basic research questions in human reproduction. In addition to our previously published studies [21–24], several aspects have been considered. First, due to the larger infertility population and high level of detail documented, CXART supports a comprehensive exploration of exposure-outcome association studies in reproductive health. Exposure is multi-dimensional, and could include socio-demographic characteristics, lifestyle behaviors, environmental exposure, dietary nutrition, emotional and psychological stress, or baseline biochemical, immunity, endocrinology, semen, and genetic tests. Exposure in these studies can also comprise any intervention during the treatment process. Meanwhile, the possible outcomes are various, which can be the laboratory performance indicators or clinical outcomes after ET. Based on these data, we can utilize full potential of epidemiological and statistical methods to systematically conduct association studies and causal inference analyses.

Second, Big Data in human reproduction allows us to focus on therapeutic evaluations in real-world settings. For instance, in response to the coexistence of controversial effectiveness and increasing clinical utilization of preimplantation genetic testing for aneuploidy (PGT-A) [25], we are assessing its efficacy using this real-world data. With the help of the complete data chain, we can start from the patient's intention at the beginning of the cycle and consider the complete impact of oocyte retrieval or blastocyst formation on the comparison. Propensity score matching was employed to balance the baseline [26]. Sensitivity analyses with propensity score stratification, inverse-probability weighting, and traditional multivariable logistic regression were applied to assess the robustness of the findings. Moreover, due to the diverse population registered in the CXART cohort, it is feasible to explore the benefit of PGT-A in different subpopulations, including couples with unexplained infertility, advanced maternal age, recurrent miscarriage, recurrent implantation failure, and severe male factors. Using real-world datasets aids in the development of more patient-tailored approaches for infertility care. A new set of evidence will then be generated to guide clinical practice.

Third, with a substantial number of biological samples combined with precious and comprehensive case resources, it is feasible to conduct basic exploratory research. Based on the clues found by macro-association analyses, we can detect, predict, and validate biomarkers in reproductive health and further develop new diagnostic and prognostic biomarkers of fecundity and treatment success. In addition, genomics, exposomes, metabolomics, and other multi-omics detection and analyses can be performed to provide important evidence for mechanistic research. We have identified more than 30 genes and nearly 200 potential pathogenic variants, explaining the genetic etiology of gametogenesis disorders, fertilization failure, and early embryonic development arrest, such as CFAP47 [27], DNAH10 [28], ACTL9 [29], BTG4 [30] and MOS [31]. More population genetics studies are being conducted to explore potential genetic causes. Standardized phenotypes of patients (possibly using contemporary biomarkers and genetic markers) are critical for advancing personalized infertility care [32].

Fourth, China is still at an exploratory stage regarding the inclusion of ART projects into the payment scope of the National Medical Insurance Fund [33]. CXART has documented a high level of detail about costs over the full course of ART, enabling analysis of the structure of and differences in disease economic burden between and within infertile populations with different ART indications or methods. Furthermore, in patients with different characteristics, we can also conduct cost-effectiveness analyses of different ovulation-stimulating drugs, ovulation-stimulating programs, fertilization methods, and ET strategies. As an important provincial capital city in central China [34], the medical practice findings from Changsha are pertinent for enhancing the inclusion of ART in the medical insurance policy, providing data support for commercial insurance to participate in the assisted reproductive market, and further improving ART accessibility.

However, several limitations in this cohort study must be considered. First, the cohort might not be representative of the Chinese infertile population, given that our patient recruitment was based at a single center. However, the patients at the Reproductive and Genetic Hospital of CITIC-Xiangya are from all over China and have a certain diversity of population characteristics. Second, the Biobank sub-cohort did not cover the entire cohort, because the cohort recruitment started before the biobank was put into use, and another important reason was that the collection of samples was based on fully informed consent of the patients. Awareness and education should be further increased to improve the coverage of sample collection, especially for patients with special or rare indications. Third, there was some loss to follow-up with online or telephone follow-up, owing to a common difficulty in the field of ART in China. Assisted reproductive therapy and obstetric delivery are administered in separate institutions, and the registered patients were distributed across the country. The delivery hospitals were also scattered, which made it difficult to follow-up for ART outcomes. Despite our dedicated follow-up team, we could not rule out potential information bias. A more comprehensive follow-up plan should be designed and implemented to expand the content, extend the time node, and improve the follow-up rate to further investigate the long-term health of ART offspring.

Funding

This study was supported by Hunan Provincial Grant for Innovative Province Construction (2019SK4012), National Key Research and Development Plan of China (2016YFC1000206) and Research Grant of CITIC-Xiangya (YNXM-202107).

Competing Interests

The authors have no relevant interests to disclose.

Acknowledgements

We thank all the participants for contribution to our study.

Author Contributions

Jing Tan and Ge Lin designed the study. Liang Hu, Yangqin Peng, Xiaojuan Wangand Lu Tan collected the data. Yiquan Xiong, Shujuan Ma and Yangqin Peng were involved in data cleaning and statistics. Shujuan Ma wrote the first draft of the manuscript, Berthold Hocher and Xin Sun contributed to the critical revision of the manuscript for important intellectual content. Jing Tan and Ge Lin were the study guarantors. All authors commented on previous versions of the manuscript, read and approved the final manuscript.

Ethics Approval

The study and biobank were approved by the Institutional Review Board of the Reproductive and Genetic Hospital of CITIC-Xiangya (LL-SC-2022014 and LL-SC-2017012). All the participants gave written informed consent. The study was performed in line with the principles of the Declaration of Helsinki.

Data availability

This is an ongoing cohort study and currently data are not yet openly available. Researchers interested in CXART can contact the corresponding authors to discuss potential collaborations. Researchers can apply to access the registry data by submitting a formal study protocol. After the application is approved by the CITIC-Xiangya Academic Committee, anonymized data could be shared with collaborators for research purposes.

Inhorn MC, Patrizio P. Infertility around the globe: new thinking on gender, reproductive technologies and global movements in the 21st century. Hum Reprod Update. 2015;21(4):411–26. doi:10.1093/humupd/dmv016.
Sun H, Gong TT, Jiang YT, Zhang S, Zhao YH, Wu QJ. Global, regional, and national prevalence and disability-adjusted life-years for infertility in 195 countries and territories, 1990–2017: results from a global burden of disease study, 2017. Aging. 2019;11(23):10952–91. doi:10.18632/aging.102497.
Steptoe PC, Edwards RG. Birth after the reimplantation of a human embryo. Lancet (London England). 1978;2(8085):366. doi:10.1016/s0140-6736(78)92957-4.
Fauser BC. Towards the global coverage of a unified registry of IVF outcomes. Reprod Biomed Online. 2019;38(2):133–7. doi:10.1016/j.rbmo.2018.12.001.
Penova-Veselinovic B, Wijs LA, Yovich JL, Burton P, Hart RJ. Cohort profile: The Growing Up Healthy Study (GUHS)-A prospective and observational cohort study investigating the long-term health outcomes of offspring conceived after assisted reproductive technologies. PLoS ONE. 2022;17(7):e0272064. doi:10.1371/journal.pone.0272064.
Chambers GM, Dyer S, Zegers-Hochschild F, et al. International Committee for Monitoring Assisted Reproductive Technologies world report: assisted reproductive technology, 2014†. Human reproduction. (Oxford England). 2021;36(11):2921–34. doi:10.1093/humrep/deab198.
Pirtea P, de Ziegler D, Toner JP. Dreaming up fertility treatment coverage and musing on simple and meaningful assisted reproductive technology metrics: the global assisted reproductive technology report by the International Committee for Monitoring Assisted Reproductive Technologies. Fertil Steril. 2021;116(3):662–3. doi:10.1016/j.fertnstert.2021.07.008.
Duffy JMN, Adamson GD, Benson E, et al. Top 10 priorities for future infertility research: an international consensus development study. Fertil Steril. 2021;115(1):180–90. doi:10.1016/j.fertnstert.2020.11.014.
Klimczak AM, Patel DP, Hotaling JM, Scott RT Jr. Role of the sperm, oocyte, and embryo in recurrent pregnancy loss. Fertil Steril. 2021;115(3):533–7. doi:10.1016/j.fertnstert.2020.12.005.
Orvieto R, Venetis CA, Fatemi HM, et al. Optimising Follicular Development, Pituitary Suppression, Triggering and Luteal Phase Support During Assisted Reproductive Technology: A Delphi Consensus. Front Endocrinol. 2021;12:675670. doi:10.3389/fendo.2021.675670.
Brannigan RE, Fantus RJ, Halpern JA. Fertility preservation in men: a contemporary overview and a look toward emerging technologies. Fertil Steril. 2021;115(5):1126–39. doi:10.1016/j.fertnstert.2021.03.026.
Chico-Sordo L, Córdova-Oriz I, Polonio AM, et al. Reproductive aging and telomeres: Are women and men equally affected? Mech Ageing Dev. 2021;198:111541. doi:10.1016/j.mad.2021.111541.
De Geyter C, Wyns C, Calhaz-Jorge C, et al. 20 years of the European IVF-monitoring Consortium registry: what have we learned? A comparison with registries from two other regions. Hum Reprod (Oxford England). 2020;35(12):2832–49. doi:10.1093/humrep/deaa250.
Bai F, Wang DY, Fan YJ, et al. Assisted reproductive technology service availability, efficacy and safety in mainland China: 2016. Human reproduction (Oxford, England). 2020;35(2):446–52. doi:10.1093/humrep/dez245.
Cyranoski D. China's embrace of embryo selection raises thorny questions. Nature. 2017;548(7667):272–4. doi:10.1038/548272a.
Hu ZB, Du JB, Xu X, et al. [Profile of China National Birth Cohort]. Zhonghua liu xing bing xue za zhi = Zhonghua liuxingbingxue zazhi. 2021;42(4):569–74. doi:10.3760/cma.j.cn112338-20201211-01402.
Endometriosis CGo. [Guideline for the diagnosis and treatment of endometriosis]. Zhonghua fu chan ke za zhi. 2015;50(3):161–9.
Zegers-Hochschild F, Adamson GD, de Mouzon J, et al. International Committee for Monitoring Assisted Reproductive Technology (ICMART) and the World Health Organization (WHO) revised glossary of ART terminology, 2009. Fertility and sterility. 2009;92(5):1520-4. doi:10.1016/j.fertnstert.2009.09.009.
Medicine, ESIGoEaASiR. The Vienna consensus: report of an expert meeting on the development of ART laboratory performance indicators. Reprod Biomed Online. 2017;35(5):494–510. doi:10.1016/j.rbmo.2017.06.015.
Duffy JMN, Bhattacharya S, Bhattacharya S, et al. Standardizing definitions and reporting guidelines for the infertility core outcome set: an international consensus development study. Fertil Steril. 2021;115(1):201–12. doi:10.1016/j.fertnstert.2020.11.013.
Chen H, Li J, Cai S, et al. Blastocyst Transfer: A Risk Factor for Gestational Diabetes Mellitus in Women Undergoing In Vitro Fertilization. J Clin Endocrinol Metab. 2022;107(1):e143-e52. doi:10.1210/clinem/dgab594.
Cai S, Li J, Zeng S, et al. Impact of vitamin D on human embryo implantation-a prospective cohort study in women undergoing fresh embryo transfer. Fertil Steril. 2021;115(3):655–64. doi:10.1016/j.fertnstert.2020.09.005.
Li Y, Wen Q, Liao J, et al. Trophectoderm Biopsy Differentially Influences the Level of Serum β-Human Chorionic Gonadotropin With Different Embryonic Trophectoderm Scores in Early Pregnancy From 7847 Single-Blastocyst Transfer Cycles. Front Endocrinol. 2022;13:794720. doi:10.3389/fendo.2022.794720.
Xie P, Hu L, Peng Y, et al. Risk Factors Affecting Alternate Segregation in Blastocysts From Preimplantation Genetic Testing Cycles of Autosomal Reciprocal Translocations. Front Genet. 2022;13:880208. doi:10.3389/fgene.2022.880208.
Gleicher N, Patrizio P, Brivanlou A. Preimplantation Genetic Testing for Aneuploidy - a Castle Built on Sand. Trends Mol Med. 2021;27(8):731–42. doi:10.1016/j.molmed.2020.11.009.
Garrido MM, Kelley AS, Paris J, et al. Methods for constructing and assessing propensity scores. Health Serv Res. 2014;49(5):1701–20. doi:10.1111/1475-6773.12182.
Liu C, Tu C, Wang L, et al. Deleterious variants in X-linked CFAP47 induce asthenoteratozoospermia and primary male infertility. Am J Hum Genet. 2021;108(2):309–23. doi:10.1016/j.ajhg.2021.01.002.
Tu C, Cong J, Zhang Q, et al. Bi-allelic mutations of DNAH10 cause primary male infertility with asthenoteratozoospermia in humans and mice. Am J Hum Genet. 2021;108(8):1466–77. doi:10.1016/j.ajhg.2021.06.010.
Dai J, Zhang T, Guo J, et al. Homozygous pathogenic variants in ACTL9 cause fertilization failure and male infertility in humans and mice. Am J Hum Genet. 2021;108(3):469–81. doi:10.1016/j.ajhg.2021.02.004.
Zheng W, Zhou Z, Sha Q, et al. Homozygous Mutations in BTG4 Cause Zygotic Cleavage Failure and Female Infertility. Am J Hum Genet. 2020;107(1):24–33. doi:10.1016/j.ajhg.2020.05.010.
Zhang YL, Zheng W, Ren P, et al. Biallelic mutations in MOS cause female infertility characterized by human early embryonic arrest and fragmentation. EMBO Mol Med. 2021;13(12):e14887. doi:10.15252/emmm.202114887.
Fauser B. The use of big data to inform individualized ovarian stimulation for infertility care is still in its infancy. Fertil Steril. 2022;117(2):419–20. doi:10.1016/j.fertnstert.2021.12.011.
Wu SM, Lin XZ. Fertility support measures are further improved Assisted reproductive technology will be gradually included in medical insurance. First Financial Daily. 2022 2022-08-17;Sect. A06.
Zhou Y, Wang T, Fu J, Chen M, Meng Y, Luo Y. Access to reproductive health services among the female floating population of childbearing age: a cross-sectional study in Changsha, China. BMC Health Serv Res. 2019;19(1):540. doi:10.1186/s12913-019-4334-4.

Supplementaryinformation.xlsx

Download PDF

Version 1

posted

You are reading this latest preprint version

Cohort Profile: CITIC-Xiangya Assisted Reproductive Technology Cohort (CXART Cohort)

Status:

Version 1

Abstract

Figures

Introduction

Participants And Methods

Preliminary Results

Baseline characteristics of study population

Cycle characteristics

Preliminary outcomes

Discussion

Summary of cohort

Strengths and weaknesses

Declarations

References

Supplementary Files

Status:

Version 1