Analysis of parental abnormal chromosomal karyotype and subsequent live births in Chinese couples with recurrent pregnancy loss

The frequency and distribution of chromosomal abnormalities and the impact of parental chromosomal aberration on the pregnancy outcomes of couples with recurrent pregnancy loss remains controversial. 3235 RPL couples who experienced two or more miscarriages before 20 weeks were diagnosed in our tertiary referral hospital during 2008–2018 and included in the single-center retrospective cohort study covering a 10-year period. Chromosome aberration was detected in 121 (3.74%) among 3235 RPL couples which included 75 female and 46 male cases at an individual level. 101 cases were structural aberrations including balanced translocations in 46(38.0%) cases, Robertsonian translocations in 13(10.7%) cases, inversions in 42(34.7%) cases and 20(16.5%) cases were numerical aberrations. 121 carriers and 428 non-carriers were followed up for two years, 55 carriers and 229 non-carriers were subsequent pregnant after diagnosis by natural conception or intrauterine insemination. The frequency of carriers to have a health newborn was not significantly different with non-carriers (72.7% vs. 71.2%, adjusted P = 0.968). This study described the majority of carriers were balanced translocations and chromosome aberrations had a limited influence on live birth rate from the present data. The results of the study also remind us that natural conception may be also a good alternative rather than PGD (Pre-implantation Genetic Diagnosis) which is common in many other reproductive centers for such patients.

Peripheral blood karyotype analysis. Chromosome analysis was performed on routinely cultured peripheral blood lymphocytes as described previously 12 . Briefly, the sections were treated with trypsin using standard techniques, the slides were Giemsa stained and then G-banding analysis was performed. Add colchicine 4 h before cytology preparation. For each sample, at least 20 cells from two independent cultures were used for microscopic observation and analysis in metaphase. Follow-up. All patients had been followed up for at least two years to get the subsequent first pregnancy outcomes. The details of each individual were entered into a computerized database with clinical features and miscarriage history recorded. Data of the present study was retrieved from medical records and telephone interviews.

Statistical methods. t test was used for measurement data between the two groups, and chi-square test or
Fisher's exact test was used for count data. Binary logistic regression analysis was used to evaluate risk factors for pregnancy outcomes. P < 0.05 was considered statistically significant. The statistical software of SPSS 20.0 was applied in the study.

Results
The frequency and distribution of aberrant chromosomal RPL couples. The First Affiliated Hospital of Xi'an Jiaotong University is a tertiary referral teaching hospital. 5680 recurrent miscarriage couples came to the Reproductive Medicine Center from January 2008 to December 2018 as shown in Fig. 1. The flow chart shows that 954 couples had not completed the etiology screening evaluation and 1491 couples had not peripheral karyotype test of both female and male, so the remaining 3235 couples had complete karyotype analysis results. There are 121 couples of abnormal chromosomal karyotypes in 3235 couples with complete results (including abnormalities of either the female or the male and excluding chromosomal normal polymorphisms) with the abnormality rate of 3.74% (121/3235). Among 121 couples with abnormal chromosomal karyotypes, 101 cases were structural abnormalities (3.12%, 101/3235), and 20 cases were abnormal numbers (0.62%, 101/3235).
As shown in Fig. 2A www.nature.com/scientificreports/   www.nature.com/scientificreports/ couples, and the distribution of abnormal chromosome types in female and male respectively can be shown in the Fig. 2B. During the following-up of 121 chromosomal abnormal couples with recurrent miscarriage, 55 couples were pregnant and 66 couples were not pregnant merely by medical expectant management through natural conception or intrauterine insemination without IVF/PGD as in the flowchart of Fig. 1. The proportions of the four types of chromosomal abnormalities among pregnant and non-pregnant carriers was shown in Fig. 2C and D.
The two groups had no statistical difference in the four-type abnormal distribution by chi-square test (P = 0.31).

Comparison of the etiological results and live birth rates between 55 carriers and 229 non-carriers.
Because it is difficult to achieve the complete the pregnancy results from thousands of patients from 2008 to 2018, 428 RPL couples with normal chromosomes who came to our outpatient department in the whole year of 2018 were selected and followed up for 2 years. They completed all etiological screenings and 229 of them were pregnant in the follow-up period as in the Fig. 1. Comparison of 55 carriers and 229 non-carriers showed that female carriers were younger at the time of consultation, while other clinical characteristics and combined pathological factors were not statistically different in Table 1. The outcome after the pregnancy, namely the live birth rate (LBR), was also not statistical different (P = 0.87). Among the 55 carriers, 51 carriers with primary recurrent miscarriage (no previous live birth history), and 4 were secondary recurrent miscarriages. 6 females were diagnosed with polycystic ovary syndrome (PCOS, according to the Rotterdam criteria 13 ), and 3 females were diagnosed with decreased ovarian reserve (DOR, according to the hormonal markers and ultrasound parameters 14 ). To further analyze of other etiological screening results in the 55 carriers, 8 cases  www.nature.com/scientificreports/ (14.6%) were positive for infection factors (including male or female genital tract Mycoplasma and Chlamydia infection), 2 cases (3.64%) were with abnormal uterine anatomical structure, 14 cases (25.5%) were with imbalance of peripheral blood lymphocyte subsets, 9 cases (16.4%) were with endocrine disorders (including ovarian hormone abnormalities, thyroid hormone abnormalities and hyperprolactinemia), 7 cases (12.7%) were with nutritional element deficiency (including folic acid and vitamin B12). Among the combined autoimmune antibodies, 13 cases (23.64%) were positive for anti-phospholipid antibodies, and 9 cases (16.4%) were positive for connective tissue antibodies. During the follow-up period 40 in 55 pregnant RPL carriers gave birth to healthy babies in the way of natural conception or intrauterine insemination without IVF/PGD, the live birth rate (LBR) in the carriers (72.7%) was similar to that in the non-carriers (71.2%). It could be seen that, apart from age, the above-mentioned combined pathological factors and the final LBR was not statistically different between carriers and non-carriers in the Table 1. The results were still consistent after using binary logistic analysis to adjust the age factor. Table 2 showed the details of every patient number, the age of the female, the karyotypes of the female and the male, the number of miscarriages and the outcome of pregnancy of 55 carriers. 40 of 55 carriers gave birth to healthy newborns at the end with the LBR of 72.73%. Among the 40 cases, 7 cases were numerical abnormalities (LBR of 87.5%) and 33 cases were structural abnormalities (LBR of 70.21%). The structural abnormalities included 14 cases with balanced ectopic (LBR of 60.87%), 17 cases with inverted position (LBR of 80.95%), and 2 cases with Robertsonian translocations (LBR of 66.67%). As shown in Fig. 4, there was no statistical difference in the LBR in the four types of chromosomal abnormalities (P = 0.35).
Among the 55 pregnant couples, 34 were female and 21 were male carriers. In the Table 3 we analyzed the women's age, number of miscarriages, distribution of karyotype abnormalities and the total LBR in female and male carriers respectively. There was no statistical difference in all items and showed gender of carrier had no effect on the pregnancy outcome (P = 0.428).
In order to rule out the influence of other etiological factors on the pregnancy outcomes, we analyzed the female age, the number of abortions, infection factors, anatomical uterine abnormalities, autoimmune antibodies positive rate, blocking antibody deficiency, peripheral blood lymphocyte subset disorders, endocrine disorders and nutritional elements deficiency between 40 carriers with live birth and 15 carriers with miscarriage again in the Table 4. All the differences in above items between the two groups were not statistically significant and no one showed huge influence to alter pregnant outcomes.

Discussion
The cause of spontaneous abortion is generally attributed to two sources, namely seed or gamete problems and environmental problems. Gamete problems are often considered to be abnormal parental chromosomes or abnormal fetal chromosomes. The results of this study showed that the incidence of chromosomal abnormalities in couples with recurrent miscarriage was 3.74% (Fig. 1). The present results consistent with previous studies have shown that the incidence of chromosomal abnormalities in the general population is less than 1% 15,16 and RPL population is 2-5% 17,18 , indicating that parental chromosomal abnormalities rate increased assuredly in the miscarriage couples.
Balanced translocation was the most common type, accounting for 38.02% and consistent with other findings 11 . A meta-analysis from Zouhair reported that frequency of chromosomal abnormalities in couples with RPL was 5.16% and the most common reciprocal translocation accounts 48.4% in the worldwide literature review 12 . The balanced translocations and inversions will not affect the parents themselves in phenotype, but their unbalanced gametes during meiosis may indeed be part of the cause of miscarriage. Similarly, Robertsonian translocation of parental chromosomes can also cause miscarriage, birth defects or mental retardation of offspring 19 . However, all these studies could not demonstrate the explicit causality between aberrant chromosome and abortions.
Additionally, the LBR of RPL carriers in our reproductive center have reached more than 70%, indicating that the proportion of miscarriage caused by chromosomal abnormalities in RPL couples was very slight. A retrospective study from Howard et al. concluded that no statistically significant was found in the LBR between RPL couples with chromosomal abnormality (45.2%, 33/73) and the normal couples was (55.3%, 325/588), regardless of number of miscarriages and rearrangement types of chromosomal abnormalities 20 . In Goddijn's study the screening results of 1324 RPL couples showed that all the 41 couples with abnormal structure chromosomes did not yield an unbalanced fetal chromosome pattern 21 . It is also consistent with Franssen's study, the LBR of RPL carriers was equivalent to the normal couples after six accumulated gestations, and had no relevance with the type of abnormal chromosome (83% vs. 84%) 22 . However, Sugiura's study showed pregnancy prognosis was worsened with either maternal or paternal reciprocal translocations than normal couples (63% vs. 78.7% of LBR) 23 . Pregnancy outcomes for RPL couples with chromosomal abnormalities were still very satisfactory generally, although the decrease in the live birth rate may not have been detected due to insufficient sample size in our study. In addition, inversion of chromosome 9(inv(9)) is also considered as normal polymorphism in other reports 24,25 . The live birth rate of inv(9) in our study is 77.8%(14/18) which is not significantly different with the LBR in the other groups.
Preimplantation genetic testing (PGD) has been proposed as a controversial method in the worldwide for selecting normal chromosome embryos in the IVF to lower risk of miscarriage for patients with unexplained RPL and balanced translocations carriers. However, well-designed trials comparing EM (expectant management) to PGD have not been performed. Several previous cases indicated benefits of PGD including fewer miscarriages and shorter time to successful pregnancy without taking into account the emotional and financial cost of a failed or canceled cycle. More recent reports suggested clinical outcomes including pregnancy rate, live birth rate (53% vs. 67%) and clinical miscarriage rate were similar between PGD and EM among recurrent miscarriage patients 26  www.nature.com/scientificreports/ Even in the parental carriers of structural chromosomal rearrangement and history of RPL, no significant difference with regards to reproductive outcomes such as miscarriage rate, time to live birth, or live birth rate was observed between couples who pursued PGD compared with EM 27,28 . These data combined with our results allow us to reflect on the actual benefits of PGD to these patients, so clinicians can be more cautious when making an alternative of PGD in clinical work. Natural conception is also recommended as a good alternative for these aberrant chromosomal carriers. 50-60% of spontaneously aborted product of conception have been detected with chromosomal abnormality 29 . The abnormal chromosomes of the fetus are derived from the parental abnormal chromosomes or produced in the process of gamete meiosis and mitosis of the fertilized egg by mistake randomly. The types of fetus abnormality were often mainly manifested as trisomies of chromosome 13, 18, 21 and X monosomy (45, X) 30 , but not consisted with the translocation chromosomes of the parents showed as in the Fig. 3. According to Howard Carp's study, parental karyotyping was not particularly predictive of a subsequent miscarriage, 43.5% of abortus   www.nature.com/scientificreports/ from parental carriers were euploidic and the parental aberration was passed on to the abortus in only 10% of cases 31 . The phenotypes are inconsistent that parental karyotyping prefers balanced translocations (No. 8, 2, 6) and inversions (No. 9, 1, 6) rather than the more common numerical aberrations such as trisomies (No. 13,18,21) and polyploidy in fetus. Most aberrant chromosomes in the fetus are generated randomly and only a small percent derives from their parents. One of the most important results in our data is the influence of parental chromosomes on live birth rate (LBR). In our study, the LBR of both carriers and non-carriers can reach about 70% without relationship of gender, female age, chromosome abnormal type, number of previous abortions and other pathological factors. Amounts of non-genetic pathological causes related to endocrine, infection, immune and nutrition were detected not only in aberrant chromosomal carriers but also in non-carriers, while these factors have a strong impact on the pregnancy outcomes. After effective treatments such as anticoagulation and immunotherapy, the LBR of re-pregnancy after two recurrent miscarriages has reached more than 70% internationally. A prospective study showed that closely following management and treatment of other high-risk factors can increase the LBR of RPL couples with chromosomal abnormalities from 25 to 70% 32 or from 20 to 71% without the addition of assisted reproductive technology 11 . The differences of LBR in RPL carriers in previous reports may due to the different management of non-genetic pathological factors that are usually more important in fetal survival.
The formation frequency of abnormal gametes theoretically is not equal to the birth rate of abnormal babies in practice. We still recommend that the chromosome test or next-generation sequencing analysis of the amniotic fluid through puncture should be performed around 18 weeks of gestation in the natural pregnancy of RPL patients with chromosomal abnormalities, even so the deletion or duplication of smaller fragments still cannot be detected.
We strongly recommend that RPL carriers should still undergo comprehensive and systemic etiological screening. It is necessary to actively deal with other causes of miscarriage in order to improve the chances of successful pregnancy for RPL patients with chromosomal abnormalities. In order to improve the live birth rate, our treatment included surgical correction of the anatomically abnormal uterus, paternal lymphocyte treatment, anticoagulation aimed at anti-phospholipid antibody and immunosuppressive therapy were strongly recommended besides chromosome abnormality in our opinions. Paternal lymphocyte treatment and immunosuppressive therapy were done according to our experience and suggestions from some published reports [33][34][35] . However, many of the treatments offered to patients with recurrent pregnancy loss especially unexplained cause are not based on good evidence. A comprehensive reviews showed there was no role for immunotherapy in improving the LBR in women in the prevention of idiopathic RPL 36 .
The present study did not detect the karyotype of aborted fetuses and not achieve complete amniocentesis results from the pregnancy carriers. So we could not assess the impact of fetal chromosomal problems came from parents. In the study of embryo chromosome analysis of abortion tissue, trisomy and polyploidy are the majority which account for 65% and 17% respectively, a considerable proportion of fetus with aberrations include trisomy, structural abnormality and low-frequency mosaic could survive after birth 30 . The phenotypes are inconsistent that parental karyotyping prefers balanced translocations and inversions rather than the more common numerical aberrations such as trisomy and polyploidy in fetus. In addition, G-banding karyotype analysis used in this study can only detect a part of patients with abnormal numerical and structural chromosomes. Conventional karyotype analysis identifies balanced and unbalanced chromosomal rearrangements and copy number variants (CNVs) to a ∼ 5 Mb resolution. Due to the limitations of the detection method itself, it could not exclude some other types of genes or chromosome abnormalities related to miscarriage problems, such as deletion, insertion, duplication and point mutation of some gene fragments. In 2019, Chen et al. used low pass genome sequencing (GS) to detect the chromosomes of RPL couples with abnormalities rate increased to 11.7% compared to traditional karyotyping with 5.7%. However, inversions and copy-number variants detected by GS additionally had not been confirmed to directly related with miscarriage. 10 carriers observed in follow-up observations and five of them miscarried again (miscarriage rate of 50%). The small sample size did not indicate that the risk of miscarriage of abnormal chromosome couples was higher than that of couples with normal chromosomes 37 .
Finally, the lack of samples even in this 11-year study and other combined known and unknown non-genetic factors are shortcomings in the present data. The etiology of recurrent miscarriage is complicated and there are many controversies in the treatment. The coexistence of these other pathological factors and chromosomal abnormalities makes the results confused and controversial.

Conclusions
In conclusion, balanced translocation is the most common phenotype in RPL carriers, and LBR of subsequent first pregnancy is similar to the non-carriers. The present studies can help to provide more scientific clinical consultation, such as more accurate diagnosis and the prognostic outcome of subsequent pregnancy, and help doctors to raise awareness of miscarriage-related chromosome problems and foster a theoretical basis for reasonable treatment.