Miscarriage is not only a medical but also a social problem, which is multifactorial in nature and remains relevant despite the constant search for and implementation of new diagnostic and therapeutic strategies. Up to 50% of cases of recurrent miscarriage are considered idiopathic, because even the recommended complete clinical and laboratory examination does not reveal potential causes. In this regard, the study of additional risk factors that could “shed light” on the unexplored links in the pathogenesis of miscarriage continues. The evaluation of immunologic predictors of recurrent pregnancy loss, as well as the contribution of paternal factors to reproductive failure, is currently of greatest interest.

Initially, a literature search was conducted in PubMed, Scopus, eLIBRARY, and The Cochrane Database for the period from 1991 to 2023. The key words and terms used in the literature search were “recurrent pregnancy loss” and “recurrent miscarriage”. The resulting list of studies was manually analyzed to select relevant publications. This review included 69 references (domestic and foreign systematic reviews, meta-analyses, and original studies) that present the existing approaches to the problem of pregnancy failure, as well as the issues that remain unresolved to date.

There is no consistent definition of miscarriage. Currently, professional communities apply different interpretations of the term “miscarriage”, which ultimately leads to different scopes of diagnostic measures. Therefore, in order to understand what questions remain open in the topic of miscarriage, it is necessary to start with terms and definitions.

As defined by the European Society of Human Reproduction and Embryology (ESHRE), recurrent pregnancy loss is the termination of two or more consecutive pregnancies [1].

ESHRE guidelines emphasize the need for further epidemiological studies to form unambiguous definitions and to assess the extent to which the definition influences diagnosis, prognosis, and treatment.

The American Society for Reproductive Medicine (ASRM) also defines miscarriage as the loss of two or more consecutive clinical pregnancies [2].

There is a slightly different definition from the Royal College of Obstetricians and Gynecologists in the UK, which is “loss of three or more consecutive pregnancies occurring in 1% of couples trying to conceive” [3].

Russian Federal Clinical Guidelines “Recurrent miscarriage”, developed by the Russian Association of Obstetricians and Gynecologists and approved in 2021, define recurrent miscarriage as two or more clinical pregnancy losses that occurred up to 22 weeks gestation [4].

The interpretation of “miscarriage” in the case of consecutive losses of two pregnancies makes it possible to start the examination without waiting for the third episode. In this case, the division of recurrent miscarriage into primary and recurrent optimizes the search for the causes of miscarriage and the development of adequate tactics of management of the couple, which has a favorable effect not only on the prognosis but also on the psychological aspect of reproductive failures formed in the case of pregnancy loss.

The prevalence of recurrent miscarriage in the population ranges from 1 to 5%, with meta-analyses finding no significant differences based on geographic factors [5][6][7].

Despite its seemingly low incidence, pregnancy loss is a major problem. Every year, about 23 million pregnancies are lost worldwide. This means that 44 pregnancies are terminated every minute [8].

Numerous studies showed that the risk of recurrent miscarriage increased with the number of pregnancy losses in the history. For example, in women with a history of one miscarriage, the risk of recurrent miscarriage is comparable to the general population and is 13–17%. At the same time, after two spontaneous abortions, the risk of a third abortion increases more than two-fold to 36–38% [9]. The risk of recurrent loss in female patients was also found to correlate with increasing age and number of previous miscarriages [10]. For example, in women aged 25 to 29 years old with a history of three or more miscarriages, the risk of recurrent pregnancy loss is 40–42%, and in the 40–44 age group, it reaches 65%. The study on the relationship between the number and sequence of prior miscarriages and antiphospholipid syndrome (APS) [11] showed that clinical parameters such as the number of miscarriages and their sequence, as well as maternal age, did not differ significantly in couples with miscarriage in the presence of APS. Thus, testing for APS is warranted for all patients with a history of two or more consecutive or inconsecutive miscarriages.

Other risk factors for recurrent miscarriage are also well known. These include high-risk genetic thrombophilias, congenital uterine anomalies, as well as acquired myo- and endometrial diseases (polyps, synechiae, myomas, chronic endometritis, and decidualization disorders), clinical hypothyroidism and other endocrine causes, abnormal parental karyotype, obesity, stress, smoking, and excessive alcohol consumption [12]. However, even for many proven risk factors, new information systematically appears in publications. For example, it is known that in the combination of subclinical hypothyroidism and the presence of antibodies to thyroperoxidase, pregnant women should be prescribed L-thyroxine preparations [13], but this does not answer the question of whether it is necessary to prescribe these preparations to women with miscarriage and the presence of antibodies to thyroperoxidase with TTG values within the normal range. Thus, the study by van Dijk et al. showed that the administration of levothyroxine to women with recurrent pregnancy loss in the presence of AT to TPO in the euthyroid state did not lead to an increase in live birth rates [14].

Powerful advances in molecular technology are revealing new frontiers in the genetic causes of miscarriage [15]. Karyotyping of parents and cytogenetic studies of embryo tissues become routine diagnostic tools. At the same time, this study should be approached carefully enough to avoid overdiagnosis of those genetic causes that have minimal risks of repeated reproductive losses [16]. Genome sequencing-based studies are actively going to find genes potentially responsible for recurrent miscarriage, as well as the presence of common “candidate genes” in women with primary infertility and patients with repeated pregnancy loss [17]. The efficacy of using preimplantation genetic diagnosis in programs using assisted reproductive technologies versus wait-and-see tactics in couples with abnormal karyotypes is debatable [18].

The leading guidelines clearly regulate the order of examination of women with diagnosed miscarriage. The list of tests recommended for a couple with recurrent miscarriage by the clinical guidelines of the Russian Society of Obstetricians and Gynecologists is presented in the table below.

Despite the fact that the list of diagnostic measures is quite extensive, often, in clinical practice, even with a complete examination, it is not always possible to identify the cause of miscarriage and overcome it. Up to 50% of cases of recurrent pregnancy loss are considered idiopathic, which suggests that there are many unexplored factors that can lead to this problem.

Given the fact that the formation of the embryo and extraembryonic structures is possible only under conditions of adequate immune response of the maternal organism, it cannot be denied that the formation of a normally developing pregnancy, as well as the problem of pregnancy loss, largely depend on immune mechanisms. After all, the maternal immune system faces a difficult task: to form an immunologically favorable environment for the development of the embryo and placentation, while maintaining an adequate response to foreign microorganisms. In this context, it is very interesting to compare the tolerance of the maternal immune system to foreign (paternal) antigens expressed by the fetus with the tolerance to antigens expressed by commensal microbes [19].

To understand the immunologic basis of miscarriage, it is necessary to discuss the mechanisms of normal pregnancy formation.

During pregnancy, an immune microenvironment responsible for specific maternal tolerance to a semi-allogeneic (or fully allogeneic in the case of oocyte donation) fetus is formed [20]. The most important changes occur in such immune cells as macrophages, uterine natural killer cells, dendritic cells, and T cells. Maternal immune effectors such as T helper, T killers, T regulatory, and B cells involved in the regulation of adaptive immunity are activated by trophoblast antigens. Identification of active signaling through T cell receptors stimulates the differentiation of natural CD3+ and CD4+ T cells into specific T cell subsets such as Th1, Th2, Th9, Th17, Th22, and follicular Th cells (Tfh). Each of these subsets has a specific role in the development of pregnancy [21].

Th1 immunity, characterized by immune-inflammatory responses, becomes dominant in the preimplantation period, and “controlled” Th1 immunity has a positive effect on trophoblast invasion. Shortly after implantation, there is a normal shift in the immune response from Th1 to anti-inflammatory Th2. The predominant Th2 immunity, which suppresses Th1 immunity in the implantation zone, protects the embryo by counterbalancing Th1 immunity and coordinates placental and fetal development. Treg and Th9 cells regulate local inflammatory immune responses potentially harmful to the fetus. Th17 cells potentiate the development of an immune response against extracellular microorganisms during pregnancy. In this case, excessive Th17 immunity can cause uncontrolled neutrophil infiltration at the borderline between the embryo and the maternal body. Other subsets of Th cells, such as Tfh cells, also contribute to the course of pregnancy by generating favorable humoral immunity. Dysregulation of Th cell immunity during pregnancy can lead to obstetric complications such as recurrent miscarriage and pre-eclampsia [22].

A significant proportion of cases of pregnancy loss are thought to be associated with immune etiology [23]. Immunologic factors can be divided into two main groups: causes of endometrial origin and factors related to the elements of peripheral blood, but often, miscarriage has mixed etiology [24]. It is known that the endometrium plays an important role in the process of implantation and subsequent development of pregnancy, because the changes that are observed in it in some mammals and humans occur long before the appearance of the embryo in the uterine cavity [25][26].

Among the immunologic factors of endometrial origin, 3 major and 4 minor cell populations can be distinguished. The major ones include uterine NK cells, T cells, and macrophages, while the minor ones include monocytes, mast cells, dendritic cells, and B cells [27].

Uterine NK cells are mainly formed from peripheral blood NK cells [28]. They are short-lived highly differentiated cells that play one of the crucial roles in the transformation of spiral arteries. According to Tang et al, uterine natural killer cell count increases dramatically after ovulation, remains substantial in the early decidual membrane, but disappears before menstruation in the absence of pregnancy. Uterine NK cells are abundant around the spiral arteries, near the endometrial glands, and adjacent to the extravillous trophoblast in early pregnancy. However, there are relatively few of them in the stroma underlying the luminal epithelium [29].

Some studies reported an increase in uterine natural killer cells in the endometrium during the late secretory phase of the menstrual cycle in women with miscarriage [30][31][32].

Other studies suggest that uterine natural killer cell count does not differ between healthy women and patients with recurrent miscarriageы, but there is an alteration in the qualitative ratio of their subclasses, which causes a marked cytotoxic effect leading to pregnancy loss [33].

Whereas uterine NK cells are responsible for an adequate response against foreign antigens, regulatory T cells prevent autoimmune reactions against their own antigens. In addition, this cell type plays a major role in regulating the activity of T killers, namely, in limiting their activity, creating the most optimal microenvironment for the implanting embryo. Regulatory T cells are most necessary in the preimplantation period, as well as in the first weeks of pregnancy. Their role is being studied not only in pregnancy loss, but also in the formation of obstetric complications, in particular, gestational diabetes mellitus [34]. In addition, it is of scientific interest to determine the ratio of regulatory T cell subtypes (CD4+ and CD8+). The study by Wang et al. showed that an imbalance in the predominance of CD8+ T cells was more common in women with recurrent miscarriages.

Dendritic cells can have both positive and negative effects on the development of pregnancy due to the possibility of activation of T killer and regulatory T cells. In this regard, dendritic cells can be conditionally divided into two types: tolerogenic and inflammatory. Moreover, according to Mor et al., the presence of dendritic cells of both types is necessary during normal pregnancy [35].

The action of dendritic cells begins in the pre-implantation period. The seminal fluid has the ability to recruit dendritic cells into the decidual membrane. These cells act as antigen-presenting cells and present paternal antigens to regional lymph nodes, thereby triggering the expansion of the regulatory T cell population. By the time of implantation, the tolerance of the endometrial immune system to the half-foreign embryo increases [36]. Inadequate “work” of dendritic cells or aberrant activity can lead to a decrease in this tolerance and, as a consequence, to embryo rejection early in gestation. There is an association of dendritic cell dysfunction not only with recurrent pregnancy loss but also with preterm labor, development of preeclampsia, cardiomyopathy, and infectious complications of pregnancy. Prospects for future research in this direction may be the development of drugs for treatment and prevention (through vaccination) based on dendritic cells.

There are some described mechanisms of the pathogenesis of miscarriage of endometrial origin. [37]. One of them is the impaired transformation of stromal fibroblasts into a decidual phenotype [38]. On the one hand, it leads to impaired receptivity of embryonic signals even from embryos of adequate quality. On the other hand, it results in perverse sensitivity and increased chances of implantation of poor-quality embryos that will be rejected by the body later in gestation.

It should be noted that, despite the abundance of research and development, routine immunologic and immunogenetic studies in patients with pregnancy loss are currently not recommended due to the lack of developed and approved therapeutic strategies for this or that type of immunologic disorders.

Chronic endometritis, a local persistent inflammation of the endometrium, is a “mysterious” disease that often lacks clear diagnostic criteria and clinical manifestations. Chronic endometritis can be caused by changes in the qualitative and quantitative parameters of the endometrial microbiome or it can be abacterial, which greatly complicates diagnosis and the assessment of the contribution of this condition to the genesis of recurrent miscarriage and obstetric complications of pregnancy.

Currently, there is a controversial point in the clinical recommendations of the Ministry of Health of the Russian Federation. Thus, it is not recommended to routinely refer a patient with recurrent miscarriage for endometrial biopsy to clarify the cause of pregnancy loss. At the same time, the fact that the endometrium is an indicator of embryo quality is undeniable [39][40], so the loss of endometrial plasticity may be one of the causes of idiopathic miscarriage [41].

In accordance with the same domestic clinical recommendations, an immunohistochemical examination of the endometrium obtained by biopsy is necessary to clarify the causes of pregnancy loss. It aims to determine the expression of the plasma cell surface antigen CD-138. This study is recommended for patients with idiopathic recurrent miscarriage, repeated implantation failure after embryo transfer in assisted reproductive technology programs, and/or ultrasound evidence of chronic endometritis. An obstacle to the implementation of hysteroscopy with endometrial biopsy or pipel-biopsy into routine practice may be the lack of standardized criteria for the diagnosis of chronic endometritis [42]. Thus, the question of the method of identification of plasma cells (histologic examination with hematoxylin-eosin staining or immunohistochemical examination with staining for syndecan-1), as well as the method of counting endometrial plasmacytes (in the whole sample, in a certain number of fields of vision, according to calculation formulas) remains debatable.

Currently, the assessment of the endometrial microbiome is complicated and is not used in routine practice, and therefore, antibacterial therapy is either empirical or based on the results of the bacteriologic examination of material obtained from the cervical canal. At the same time, there are studies demonstrating the inconsistency of the microbial landscape in the cervical canal and uterine cavity [43], which stimulates the search for new ways to obtain material from the uterine cavity with minimal contamination of the vaginal microflora to assess the microbiome, in particular, the use of dual-lumen catheters originally designed for embryo transfer in assisted reproductive technology programs [44]. Endometrial sampling catheters with a special sheath that protects the sample as it passes through the cervix and vagina were also developed.

According to Moreno et al., the “healthiest” microbiota is the one represented predominantly by lactobacilli [45], and dysbiotic disorders of the vaginal microbiome are more common in women with recurrent pregnancy loss than in healthy women [46][47][48]. The influence of opportunistic microorganisms on the development of various diseases of the reproductive system, including pregnancy loss and repeated implantation failures, is being actively studied. For example, a prospective study by Shi et al. (2022) revealed a relationship between the prevalence of different ureaplasma species and the number of previous miscarriages, as well as the risk of preterm labor [49]. In contrast to this study, the work by Reschini et al. suggests a possible favorable effect of higher biodiversity on endometrial receptivity, and thus, the likelihood of pregnancy [44].

Interestingly, the role of endometrial polyps in the genesis of pregnancy loss remains controversial. This may be partly explained by the fact that polyp formation may be inflammatory in nature: in chronic endometritis, proinflammatory cytokines such as interleukin-1b, interferon-g, and tumor necrosis factor-α are overexpressed. The latter, in turn, enhances local estrogen biosynthesis, which ultimately leads to the formation of polyps and polyp-like endometrial hyperplasia [50][51].

It is of interest to determine the immunohistochemical characteristics of sex hormone receptor expression in the endometrium of patients with recurrent miscarriage in comparison with patients without this problem. The study by Tral et al. demonstrated the inconsistency of the histologic structure of the endometrium in women with miscarriage, desynchronosis of the receptor profile of the endometrium, and a decrease in immunologic markers (PIBF and SDF-1) associated with chronic endometritis, determining the violation of structural-functional and molecular mechanisms, which may be the cause of early pregnancy loss [52].

To identify autoimmune causes of recurrent miscarriage, the first step is to determine markers of APS. APS is a systemic autoimmune disease that belongs to acquired thrombophilias. It is characterized by recurrent vascular thrombosis (arterial or venous), obstetric pathology (more often fetal loss syndrome) and is associated with the formation of antiphospholipid antibodies – anticardiolipin antibodies and/or antibodies to β2-glycoprotein-1, and/or lupus anticoagulant [53].

Effective therapeutic regimens for obstetric APS were developed, but there are a number of unresolved problems in this area as well. First, seronegative APS, which significantly complicates the diagnosis and appropriate treatment. Second, the prescription of the recommended therapy for confirmed APS does not give a 100% result of overcoming miscarriage, which leads to active research into the efficacy and safety of the use of drugs from other groups. Thus, there is a stepwise strategy for the treatment of APS refractory to the “gold standard” therapy with low-molecular-weight heparins and low-dose aspirin, including the use of hydroxychloroquine, prednisolone, and plasmapheresis, as well as considering the use of tumor necrosis factor inhibitors and hydrophilic statins [54]. However, the use of all the above methods does not always lead to a favorable outcome of pregnancy, which dictates the need to search for other possible autoimmune causes of pregnancy loss.

Gluten-sensitive enteropathy (celiac disease) is a chronic genetically determined disease characterized by persistent gluten intolerance with the development of hyperregenerative potentially reversible atrophy of small intestinal mucosa and associated malabsorption syndrome. There is evidence that the incidence of idiopathic recurrent pregnancy loss is twice as high in women with celiac disease compared with healthy women [55].

Damage to the small intestinal mucosa in celiac disease occurs mainly because of the protein gliadin, which enters the body as part of gluten. Gliadin contains a large amount of the amino acid glutamine.

According to the recommendations of the German Society of Gynecology and Obstetrics (DGGG), the Austrian Society of Gynecology and Obstetrics (ÖGGG), and the Swiss Society of Gynecology and Obstetrics (SGGG), it is recommended to test for immunoglobulin class A to transglutaminase in women with a history of food sensitization (celiac disease) for the diagnosis of an autoimmune component of recurrent miscarriage, followed by biopsy of the proximal small intestine in case of a positive result [56].

In the presence of a genetic predisposition to celiac disease, the composition of the intestinal microbiome is of interest because HLA-DQ molecules, responsible for genetic predisposition and located in the major histocompatibility complex, are involved in immune activation, which leads to microbiota selection. In this regard, the qualitative and quantitative characteristics of the vaginal and endometrial microbiome in HLA-DQ2/DQ8-positive women (i.e., those with a genetic predisposition to celiac disease) may also differ from healthy women.

A study by Masucci et al. demonstrated a difference in the lactobacilli family in endometrial and vaginal samples of women with miscarriage and the presence/absence of genetic predisposition to celiac disease compared to controls. Thus, in women with recurrent miscarriage, Lactobacillus acidophilus was absent in both vaginal and endometrial samples. Lactobacillus iners was the most characteristic representative of the vaginal and endometrial microbiota of women with recurrent miscarriage (regardless of genetic predisposition to celiac disease, 26.4% of patients were HLA DQ2/DQ8 positive and 22.1% were HLA DQ2/DQ8 negative) [57].

Antinuclear antibodies are a group of antibodies against nuclear and cytoplasmic antigens, which are the main marker of autoimmune pathology, in particular systemic lupus erythematosus [58]. Four possible mechanisms of the influence of these antibodies on fetal egg rejection were established:

In a study by Chen et al, a meta-analysis showed that patients with recurrent miscarriage associated with both autoimmune diseases and “unexplained” causes of gestational loss had significantly higher levels of positive antinuclear antibodies than healthy women [59]. There are studies with controversial evidence that elevated antinuclear antibody levels are not specific for patients with recurrent miscarriage [60]. The ambiguity of the results requires further investigation of the relationship between the etiologic involvement of antinuclear antibody levels and recurrent pregnancy loss. The data on the results of immunotherapy for positive antinuclear antibodies in combination with pregnancy loss are also contradictory and require further research.

The role of the male factor in the development of recurrent miscarriage is now proven. Factors of male infertility are categorized into genetic, non-genetic, and other factors potentially leading to pregnancy loss. Genetic markers include abnormalities of the chromosome number and structure, Y-chromosome microdeletions, epigenetic factors, and gene polymorphisms. Non-genetic factors are represented primarily by the DNA fragmentation of spermatozoa, the age of the father, and abnormalities in spermogram parameters and lifestyle (obesity, smoking, and the presence of chronic diseases) [61]. While genetic factors directly influence the risks of pregnancy loss, the role of sperm DNA fragmentation remains controversial due to the lack of a standardized definition of “high” fragmentation [62]. There are studies aimed at evaluating the efficacy of ways to overcome pregnancy loss associated with increased sperm DNA fragmentation, primarily through different variants of ICSI (intracytoplasmic sperm injection). A study by Lepine et al. revealed that ICSI with hyaluronic acid showed a higher probability of clinical pregnancy in couples with recurrent miscarriage in assisted reproductive technology programs compared to the classical ICSI technique. However, the indicators of the desired outcome (live birth) do not have reliable differences in these groups [63].

A study of non-genetic factors showed relatively higher amounts of pro-inflammatory cytokines in the seminal plasma of men from couples suffering from recurrent miscarriages. These patients were significantly older than the control group and had more unfavorable lifestyle factors [64].

Another new direction in assessing the contribution of the male factor to recurrent miscarriage is quantitative proteomics, a method that identifies key proteins. Their expression level determines the qualitative characteristics of spermatozoa, as well as embryo development in early pregnancy. One of them is clusterin, a protein that has an anti-inflammatory effect by influencing the complement system and regulating the functioning of the vascular endothelial lining. Reduced clusterin expression in semen samples may lead to impaired compensatory metabolic reprogramming in early pregnancy and serve as a predictor of pregnancy loss [65].

Spermatozoa are significantly susceptible to oxidative stress due to weak antioxidant systems, insufficiently active DNA repair, and a large amount of polyunsaturated fatty acids in the plasma membrane, the main target of reactive oxygen species. Sperm leukocytes, while performing their primary function of defense against infectious agents, produce aggressive chemicals, initiate an inflammatory response, and, as a consequence, act as a source of oxidative stress [66].

A study by Al-Sheikh et al. that aimed to determine the concentration of key antioxidant/oxidant molecules revealed a significant increase in superoxide anion radicals in plasma and placental tissue samples in patients with miscarriage compared to healthy women [67]. There are also published works demonstrating the relationship between oxidative stress and APS, as well as a decrease in the concentration of serum markers of the latter during antioxidant therapy [68][69].

The multifactorial and polyetiological nature of recurrent miscarriage, the exact mechanisms of which are not fully understood, is one of the most complex problems in modern reproductive medicine. However, the strategy of searching for the causes of fertility disorders should be based on the principles of evidence-based medicine, maximizing compliance for the couple facing this problem.

Clinicians face a number of questions dealing with the issue of miscarriage on a daily basis. What investigations should be recognized as the “mandatory minimum” in idiopathic miscarriage? What should be the treatment for pregnancy loss combined with preterm pregnancies and full-term deliveries? Should the loss of two or more pregnancies terminated as a result of various etiologic factors be treated as recurrent miscarriage? What is the scope of the diagnostic search, terms and methods of rehabilitation, as well as the possibilities of fertility restoration in a couple with combined infertility?

The prospects for future research lie in obtaining answers to these and many other questions. The solutions and rational pre-conceptional preparation will contribute to overcoming recurrent miscarriage, successful conception, a favorable course of pregnancy, and the birth of a healthy child.