Approximately 2–4% of RPL is associated with a parental balanced structural chromosome rearrangement, most commonly balanced reciprocal or Robertsonian translocations. These structural rearrangement disorders are already approved to be treated by preimplantation genetic diagnosis (PGD) through assisted reproductive technology (ART) to avoid subsequent miscarriage. There are some other structural abnormalities associated with RPL such as chromosomal inversions, insertions, and mosaicism. There might be some association with RPL including some single gene defects such as cystic fibrosis or sickle cell anemia. Parental karyotyping should be performed to evaluate RPL with genetic counseling. Appropriate genetic counseling is indicated in all cases of RPL associated with parental chromosomal abnormalities. Definitive treatment may need ART with PGD depending on the particular diagnosis. Preimplantation genetic screening (PGS) is explained in the following explanations in details. Application of donor gametes on treatment may be suggested in cases involving genetic anomalies that always result in embryonic aneuploidy such as Robertsonian translocations involving homologous chromosomes.

Approximately 10–15% of causes of RPL are derived from anatomical malformation of female reproductive organs and are generally thought to induce miscarriage by insufficient vasculature of the endometrium, prompting abnormal and inadequate placentation. There are some suggested causes which are congenital uterine anomalies, intrauterine adhesions, and uterine fibroids or polyps. These abnormalities are thought to be potential causes of RPL through interruption of vascular supply of the endometrium. The most closely linked to RPL is uterine septum, congenital uterine anomaly, with as much as a 76% risk of spontaneous pregnancy loss among affected patients. Other Müllerian anomalies, including unicornuate, didelphic, and bicornuate uteri, have been associated with smaller increases in the risk for RPL. Arcuate uterus may or may not be causing RPL. The presence of intrauterine adhesions, sometimes associated with Asherman syndrome, may significantly impact placentation and result in early pregnancy loss. Intramural fibroids larger than 5 cm, as well as submucosal fibroids of any size, may be associated with RPL. Myomectomy should be considered in cases of submucosal fibroids or any type of fibroids larger than 5 cm, especially in the patient under infertility treatment. Significantly, improvement of live birth rates has been shown by myomectomy from 57 to 93% [4]. Myomectomy could be performed via open laparotomy, laparoscopy, or hysteroscopy. Congenital anomalies caused by prenatal exposure to diethylstilbestrol (DES) are well known with reference to RPL. However, influence of DES administration is becoming less clinically significant since most affected patients move beyond their reproductive ages. Uterine anatomic anomalies should be evaluated by either office hysteroscopy or hysterosalpingography (HSG). If these abnormalities are determined, resection of intrauterine adhesions or intrauterine septa should be performed hysteroscopically. Successful hysteroscopic septum resection brings the patients nearly normal pregnancy outcomes, with term delivery rates of around 75% and live birth rates of around 85% [5].

Endocrine disorders such as luteal insufficiency, polycystic ovary syndrome (PCOS), diabetes mellitus, thyroid gland disease, and hyperprolactinemia might be associated with RPL in approximately 17–20% [6]. Luteal insufficiency has been identified to be an inadequate progesterone production by the corpus luteum and insufficient endometrial maturation for implantation. However, definite influence of luteal insufficiency on RPL is controversial, and endometrial biopsies for diagnosis of luteal insufficiency are getting less performed. Insulin resistance with resultant hyperinsulinemia may play a role on RPL of the patients complicated with PCOS as well as type II diabetes mellitus, because treatment of those patients with insulin-sensitizing drug, metformin, decreased the rate of spontaneous pregnancy loss [7]. There is an evidence of PCOS in at least 40% of women with RPL. Poorly controlled type I diabetes mellitus is also associated with an increased risk of spontaneous abortion. Untreated hypothyroidism is clearly associated with spontaneous miscarriage and RPL, but the relation between antithyroid antibodies and RPL in euthyroid patients is currently under investigation. There are data to suggest that euthyroid women with antithyroid antibodies, especially those undergoing infertility treatment, are likely to become clinically hypothyroid when they achieved pregnancy. Because pregnancy outcomes in these women may improve with early (possibly prenatal) thyroid hormone replacement, similar approaches are presently being studied among women with RPL [8, 9]. Evaluation of endocrine disorders should include measurement of the thyroid-stimulating hormone (TSH) level. Other testings that might be indicated based on the patient’s presentation include insulin resistance testing, ovarian reserve testing, serum prolactin in the presence of irregular menses, antithyroid antibody testing, and, very rarely, luteal phase endometrial biopsies. Therapy with insulin-sensitizing agents for the treatment of RPL that occurs in the presence of PCOS has recently gained popularity.

The role of infectious diseases in RPL is not clarified yet, but proposed an incidence of 0.5–5% [6, 10]. There are some candidate infectious diseases such as Listeria monocytogenes, Toxoplasma gondii, rubella, herpes simplex virus (HSV), measles, cytomegalovirus, and coxsackie viruses. Infectious diseases may cause pregnancy loss by the following mechanisms such as direct infection of the uterus, fetus, or placenta, placental insufficiency, chronic endometritis/endocervicitis, amnionitis, or intrauterine miscellaneous infections. Infections of mycoplasma, ureaplasma, Chlamydia trachomatis, L. monocytogenes, and HIV are speculated to play a role in RPL. Chronic infection is the most pertinent risk for RPL secondary to acute stage in an immunocompromised patient. Evaluation for chronic infections may be warranted for those patients. Overall, prevention of infectious diseases is not necessary, but favorable for the patient of RPL to relieve their anxiety.

This problem happens mainly in Caucasian people. Inherited and combined inherited/acquired thrombophilias are common with more than 15% of the white population carrying an inherited thrombophilic mutation. The factor V Leiden mutation is the most common. This is the mutation in the promoter region of the prothrombin gene and mutations in the gene encoding methylenetetrahydrofolate reductase (MTHFR). These common mutations are associated with mild thrombotic risks, and it remains controversial whether homozygous MTHFR mutations are associated with vascular disease at all. The potential association between RPL and heritable thrombophilias is based on the theory that impaired placental development and function. This mechanism causes venous and/or arterial thrombosis and thereafter induces miscarriage. Pregnancy losses by this type of thrombophilia take place at greater than 10 weeks of gestation rather than prior to 10 weeks of gestation, because maternal blood begins to flow within intervillous spaces of the placenta at approximately 10 weeks of gestation. Of course, the transfer of nutrition from the maternal blood to the fetal tissues depends on uterine blood flow regardless of gestational age. Therefore, thrombotic events occurring at any gestational age play a role for thrombophilias in pregnancy losses [13]. The heritable thrombophilias associated with RPL include hyperhomocysteinemia resulting from MTHFR mutations, activated protein C resistance associated with factor V Leiden mutations, protein C and protein S deficiencies, prothrombin promoter mutations, and antithrombin mutations. Acquired thrombophilias associated with RPL include hyperhomocysteinemia and activated protein C resistance. Definite causative links between these heritable and acquired conditions have yet to be solidified. However, testings for factor V Leiden mutation, protein S levels, prothrombin promoter mutations, homocysteine levels, and global activated protein C resistance are appropriate targets for the selection of treatments. Once diagnosis is determined, appropriate therapy for heritable or acquired thrombophilias should be initiated. Specific for individual disorder should be performed as follows: supplementation of folic acid for those patients with hyperhomocysteinemia and prophylactic anticoagulation in cases of isolated defects with no personal or family history of thrombotic complications. Therapeutic anticoagulation should be performed in cases of combined thrombophilia defects.

Although exact causes of RPL have not been elucidated and still unexplained or idiopathic, some of the cause can be explained by various factors such as described above. Quintessential possibility of these causes for idiopathic RPL is that these couples are producing more aneuploid embryos, leading to higher miscarriage occurrence. The role of chromosomal abnormalities in miscarriage has been widely reported, with 50–70% of first-trimester miscarriages attributed to aneuploidy. Furthermore, it has been demonstrated that analyses of fetal chromosomes miscarried could explain 80% of unexplained RPL in older women [25]. A higher rate of aneuploidy in RPL patients has been confirmed by many authors [10, 26–34]. Preimplantation genetic screening (PGS) has been proposed as a method for reducing miscarriage by selecting euploid embryos for transfer, because of the prevalence of aneuploidy in first-trimester losses and the increased prevalence of aneuploidy in the RPL population. The current standard of care for patients with unexplained RPL espoused by the American Society for Reproductive Medicine is expectant management [35]. However, the emotional trauma that can accompany clinical miscarriages and a perceived urgency to conceive felt by many RPL patients lead them toward alternative treatment options, including assisted reproductive technology, and specifically to in vitro fertilization (IVF) and PGS. Therefore, PGS for the indication of idiopathic RPL is that euploidy embryos could be selected for embryo transfer, leading to a decreased pregnancy loss rate in idiopathic RPL patients. All studies using PGS for this indication have evaluated that the miscarriage rate after this procedure has shown a decrease [36–39]. Again, it has been widely well accepted that aneuploidy is the most common genetic abnormality in embryos and also the most common cause of miscarriage [40, 41]. No matter how good treatment the patient is offered, if the embryo implanted was aneuploidy, it never works.

A term, aneuploidy, has been used to describe a loss or gain of genetic material of a chromosome(s) since the first human with aneuploidy. Since then, aneuploidy has been demonstrated to be a very common cause, accounting for no less than 15–20% of all clinical pregnancies. The majority of aneuploid embryos will never result in a clinical pregnancies and live birth, making aneuploidy the leading cause of miscarriage, but some are compatible with live birth, making aneuploidy the leading cause of congenital malformations and mental retardation. Aneuploidy has been identified as a significant factor contributing to IVF cycle failures, specifically implantation failure and/or spontaneous miscarriage in the field of assisted reproductive technology (ART) [42]. However, recent advances in reproductive medicine and molecular cytogenetics have completely changed the treatment protocol designed for infertile couples suffering from recurrent aneuploid losses. Genetic testings such as chorionic villus sampling, amniocentesis, and NIPT (noninvasive prenatal test) from maternal blood have been available prenatally. When these techniques are applied, if unfavorable results are revealed, a subsequent termination of living fetuses would still be necessary. Fluorescence in situ hybridization (FISH) was the main methodology of PGD or PGS over the past two decades, and aneuploidy detected by FISH technology with reference to infertility was reported in the beginning [43–45]. In spite of the confirmation of the high rate of aneuploidy in both repeated IVF failures and miscarriages, improvement of IVF outcomes with PGS by FISH was not demonstrated successfully [46–49]. These earlier studies were typically performed with the use of FISH evaluation of cleavage-stage embryos and typically tested only 7–12 chromosomes. In one meta-analysis [50], four observational studies [41, 51–53] were evaluated in which fertile patients with RPL underwent day 3 cleavage-stage biopsy of 1–2 cells and were compared with natural conception RPL patients. All four studies performed FISH screening 3–9 chromosomes. The spontaneous abortion rate (SABR) ranged from 0 to 10% (mean 9%) in RPL patients with PGS compared with 14–52% (mean 28%) with natural conception (P = 0.0013). Thereafter, array CGH technology appeared by analyzing all 24 chromosomes, as opposed to FISH, allowing more accurate results when detecting for aneuploidy. There are several methods of comprehensive chromosome screening (CCS), including single nucleotide polymorphism (SNP) array, CGH, and quantitative polymerase chain reaction (PCR) [50]. Comparison of FISH with SNP array showed up to a 60% false-positive rate with FISH. When FISH was compared to CCS, it was found that mosaicism was three times more common in FISH [54]. Therefore, the European Society of Human Reproduction (ESHRE) recently recommended that this technique should be replaced by comprehensive methods of screening [55]. In conclusion, PGS by CCS should be applied to RPL patients in modern ART era.

It is difficult to find the ideal control group for RPL studies to determine if PGS is beneficial to reduce miscarriage. It is the question if the RPL couple should be compared with other couples undergoing PGS, with or without infertility, or only those with a history of RPL when PGS is applied and found that PGS using FISH significantly reduced miscarriage rates, from 36% expected rate to 13%. Patients that were offered PGS but rejected it had a 44% miscarriage rate, which is also another way to compare RPL patients using PGS with an appropriate control. This beneficial effect of PGS for RPL was observed in both fertile and infertile RPL patients undergoing IVF [40]. However, these studies used FISH, evaluated a limited number of chromosomes, and used day 3 embryo biopsy, which very recent evidence suggests it can negatively affect the implantation potential of the biopsied embryo, whereas blastocyst biopsy does not seem to be detrimental [56]. The clinical effectiveness of IVF and PGS compared with expectant management, which is the current standard of care in the treatment of RPL patients, has not been investigated with longitudinal prospective studies or randomized clinical trials. Furthermore, IVF–PGS is an expensive treatment option, and the cost-effectiveness of IVF–PGS compared with expectant management needs to be investigated. However, recent following report demonstrated the beneficial effect of PGS on RPL patients. It concluded that patients with RPL initiating PGS have a significantly higher LBR compared to expected management with no significant difference in miscarriage rate. Miscarriage rate would likely be lower if all IVF patients intending PGS completed the cycle as intended since aneuploidy is a common cause of first-trimester miscarriage. Of course, further studies are needed to investigate the cost-effectiveness of this treatment strategy for fertile RPL patients.