Reproductive Effect by Rheumatoid Arthritis and Related Autoantibodies

RFs, nonspecific antibodies targeting the Fc portions of immunoglobulins, include different isotypes, such as immunoglobulin M (IgM), IgA, IgG, and IgE. IgM-RF and IgA-RF are the most common RFs in clinical RA [5, 14], and they have also been reported to be predictors of poor prognosis and are strongly associated with extra-articular symptoms and long-term cardiovascular events [15, 16].

Despite their name, RFs are not specific for RA. RFs can also be detected in individuals with systemic lupus erythematosus (SLE), Sjögren’s syndrome (SS), and even in healthy populations [14]. Furthermore, RFs interact only with the immunoglobulins that bind to their corresponding antigens. Specifically, the CH2-CH3 domain within the Fc region, the RF-binding epitope, is exposed only when the immunoglobulin conformation changes [17, 18].

In 1948, Rose et al. discovered that RFs could be detected via sensitized sheep red blood cells, the so-called Waaler–Rose test [19, 20]. Then, several test methods, such as nephelometry and latex fixation, were gradually developed. Unfortunately, these methods are unable to distinguish the various immunoglobulin isotypes and mainly recognize the IgM isotype owing to its polyvalency [21]. Currently, common laboratory tests include enzyme-linked immunosorbent assays (ELISAs), chemiluminescence immunoassays (CLIAs), and addressable laser bead immunoassays (ALBIAs), among which ELISA was reported to be more objective and sensitive while identifying particular isotypes [21].

ACPAs are a group of autoantibodies targeting citrullinated proteins or peptides. Before the concept of “ACPAs” appeared, they were also called anti-cyclic citrullinated peptide (CCP) antibodies because CCP was a main antigen in testing. Citrullination is an enzymatic reaction that converts arginine to citrulline, mediated by peptidylarginine deiminase (PAD) [17], which results in the loss of the protein’s basic charge(s) and the generation of new antigenic targeted epitopes [22].

Resembling the broad isotype of RFs, IgA-, IgM-, and particularly IgG-ACPA can also be isolated from the serum of patients with RA, with a high specificity (94.9–97.8%) for diagnosis [23, 24]. To date, a variety of tests to detect ACPAs and anti-CCP assays have been developed to the third generation. In comparison with the initial anti-CCP1 assay relying on the cyclic derivative derived from citrullinated filaggrin [25], second- and third-generation anti-CCP assays not only retain the high specificity of anti-CCP1 tests but also have a higher sensitivity [25, 26]. A detailed description of the three generations of CCP assays is shown in Supplementary Table 1.

As another well-studied AMPA in addition to ACPAs, anti-carbamylated protein (anti-CarP) antibodies have been found in up to 46% of patients with RA and 16–20% of patients who are ACPA-negative [27]. Similar to citrullination, carbamylation is the conversion of lysine to homocitrulline under the influence of cyanate, which is at a lower level under physiological conditions. Structurally, homocitrulline is similar to citrulline except for an additional methyl group (CH₂) on the side chain, which may explain the cross-reactivity of ACPAs for citrullinated and carbamylated proteins or peptides.

Similarly, anti-CarP antibodies also show a broad range of isotypes and IgG subclasses, including IgM, IgG1–4, and IgA. With its increased application, the accuracy of screening individuals at a maximum risk of RA has been greatly improved in background populations, especially for patients with pre-RA. Compared with patients who are single and double positive, the specificity for RA screening could increase to 98–100% when patients are triple positive for ACPAs, RFs, and anti-Carp antibodies (Table 1) [24, 28‐31]. Moreover, beyond their diagnostic value, anti-CarP antibodies have also been considered a good clinical response predictor in a subgroup of patients with RA receiving abatacept treatment [32].

Decades of studies have already shown the link between RA and fertility (Table 2) [33‐41]. As mentioned above, the female/male incidence ratio is 4–5 among individuals younger than 50 years [4]. Considering the great impact of RA on women at reproductive age, multiple studies have evaluated whether RA could impair women’s fertility, and this impairment was commonly quantified on the basis of fertility (the production of biological children) and fecundity (the time to achieve pregnancy, TTP) [6, 39, 41].

Several large cohort studies have demonstrated that patients with RA show a trend toward infertility. Brouwer et al. observed that RF and ACPA positivity were significantly higher in women who did not conceive compared to pregnant women (p = 0.01, p = 0.05). The authors identified subfertility, defined as a TTP of greater than 12 months, and nonpregnant status in 67% of patients in a high disease activity group (28-joint Disease Activity Score > 5.1), in contrast to 30% and 37% of patients in low disease activity and remission groups, respectively [38]. Additionally, a strong association between a history of infertility and disease duration was observed among women with RA in Nigeria. Assays for RFs and anti-CCP were also performed, with positive rates of 80% and 66%, respectively. Unfortunately, their differences in history of infertility did not reach significance, which was attributed to limitations of small sample size (only 50) [41]. Considering that the majority of RA-affected women are not diagnosed until late childbearing age [42, 43], even past the age of reproduction, the role of RFs and/or ACPAs in RA-related infertility is difficult to extrapolate, as the age-related fertility decline may interfere with the analysis of the effect of RA on fertility among diagnosed patients.

PL has always been a concerning issue in the field of reproduction and is defined as a natural PL within 20 weeks [7]. In terms of etiology, studies have investigated multiple causes, including chromosomal abnormalities of couples and fetuses, infections, endocrine disorders, advanced maternal age, structural uterine abnormalities, etc. [7]. Among these presumed etiologies, immunological and rheumatological factors remain a focus of debate.

Prior studies have noted that elevated levels of serological markers such as RFs and ACPAs in RA seem to be strongly associated with adverse pregnancy outcomes (APO), such as premature delivery and low-birth-weight infants [44]. As early as 1994, Geva et al. found that autoimmune disorders caused by circulating RFs may be responsible for biochemical pregnancy after implantation [45]. Although several studies suggested that the miscarriage rate of patients with RA was not increased significantly compared with those of the reference or background populations [46‐48], a retrospective study based on data from the medical birth registry of Norway showed that individuals with RA had a slightly higher likelihood of PL. The frequency of early PL (< 12 weeks) in patients with RA was 46%, higher than in reference women (34%, p < 0.001) [49]. Moreover, a multicenter study found that PL (< 20 weeks) accounts for approximately 44% of APO in patients with RA [50]. In light of these results, a study stratified research participants by maternal age and indicated that those under the age of 35 and diagnosed with RA within the first 5 years of pregnancy were more likely to also have PL, with an adjusted odds ratio of 1.25 [51]. Then, to further explore the association of serum markers of RA with the miscarriage rate, a study was conducted within the cohort of the Pregnancy-induced Amelioration of RA (PARA), and the data of 162 patients with RA who tried to become pregnant or were in the first trimester of pregnancy were collected. The study showed that ACPA positivity was more common in female patients with RA who had a miscarriage than in those without, although the difference was not statistically significant (p = 0.058) [52]. Additionally, a cohort study of 299,629 Chinese pregnant women performed by the University of Oxford and the Chinese Centre for Disease Control and Prevention found a certain association between RA and PL, either spontaneous or induced. This risk was significantly positively correlated with the number of PL [53].

PCOS, characterized by chronic anovulation and hyperandrogenism, is associated with a multitude of conditions, including metabolic abnormalities such as diabetes mellitus, obesity, and cardiovascular disorders [54]. In addition, irregular menstrual cycles, metabolic conditions, and infertility are common in individuals with RA. Whether women with PCOS are at an increased risk for developing RA is still being debated.

The association of PCOS and RA was first elaborated in 2003 when Merlino et al. reported that there was an elevated risk of RA for elderly women with a history of PCOS, either in age-adjusted or multivariable-adjusted models [55]. Then, at the 2017 ACR/ARHP annual meeting, Edens reported that the prevalence of PCOS in individuals with rheumatic diseases (3.9%) was similar to that in the general population (3–10%) [56]. Recently, a single-center epidemiological study showed that in the field of rheumatology, some systemic autoimmune diseases, especially rheumatic diseases, were more common among the PCOS population [9]. The prevalence of RA reached 2.25% (17/737) in the patients with PCOS, although the difference was not significant (p = 0.0747) when compared to the control population (1.26%, 19/1489). However, in contrast to the literature data on the female population, the prevalence of RA in patients with PCOS was numerically higher (2.25% vs. 1.40%, p < 0.0001).

Notably, previous studies reported that dehydroepiandrosterone sulfate (DHEAS) may reduce susceptibility to RA [57, 58]. Higher levels of testosterone in young men were found to play a protective role in the incidence of RA [58]. In this way, the high level of androgen present in PCOS would have a protective effect against RA, but this contradicts the trend of a high prevalence of RA in PCOS backgrounds. Although researchers argue that the protective role of hyperandrogenism in the context of PCOS may be offset by estrogen’s effects on the immune system [59], the role of sex hormones in RA still needs to be verified in larger multicenter cohort studies.

Endometriosis, as a progressive and heterogeneous chronic inflammatory disease, is estimated to affect 190 million reproductive-age women worldwide [60]. Studies have demonstrated that women with endometriosis are more likely to have coexisting conditions, such as SLE, RA, SS, and inflammatory bowel disease (IBD) [10].

Concerning RA in patients with endometriosis, several epidemiological studies worldwide have found a higher incidence of RA among patients with endometriosis compared to the general female population (Table 3) [61‐68]. In 1986, Lamb and Nichols reported certain links between RA and endometriosis [61]. Recently, a study, which the authors claimed to be the largest population-based cohort in Asia, found that women with endometriosis had a higher risk for RA [66]. The incidence of RA in that study also matched the findings in cross-sectional cohort studies by both Sinaii and Huang [62, 64]. With regard to the prevalence of endometriosis in patients with RA, both Merlino and Shafrir showed that patients with RA were more likely to develop endometriosis [55, 68].

The gradual decline in ovarian reserve with age is an inevitable physiological phenomenon. However, women younger than 40 years of age can experience extensively impaired ovarian function, termed POI. Although prior studies increasingly observed that approximately 4–30% of POI cases coexisted with autoimmune diseases [69], these studies mainly focused on the effects of anti-ovarian antibodies and thyroid antibodies. Even the European Society of Human Reproduction and Embryology (ESHRE) guideline group also recommended screening only for thyroid and adrenocortical antibodies [70]. Information on the relationship between POI and RA is lacking.

Serum anti-Müllerian hormone (AMH) is often used as the best proxy of ovarian function. Patients with RA were reported to possess a lower AMH level, which was verified again in a recent systematic review and meta-analysis [71‐74]. Brouwer et al. followed 128 patients with RA and found that 39% of patients had AMH levels lower than the 10th percentile of controls [74]. In a PARA cohort study of 209 women, AMH levels were lower in 36 women and were inversely associated with ACPA positivity (p = 0.009) [71]. Similarly, a recent interdisciplinary prospective study by Grossmann and colleagues reported that approximately 38.5% patients with premature ovarian failure (POF) had coexisting symptoms of joint pain, and a patient who was positive for RFs and anti-CCP antibodies was clinically diagnosed with RA 5.5 years later [11]. Hoek et al. also observed that 41% of patients with POF tested positive for RFs [75]. Therefore, it is reasonable to speculate that there is a certain relationship between RA and ovarian insufficiency. Additionally, researchers suggested that RA itself may accelerate the depletion of the follicle pool, resulting in a lower ovulation rate [76]. This relationship was revalidated in a large prospective cohort in France that demonstrated a potential correlation between early menopause (≤ 45 years) and RA [29].

The coordination between immune and endocrine cytokines throughout the menstrual cycle is of great importance to female reproductive functions. Patients with RA have been reported to have inborn Th1/Th2 balance differences characterized by predominantly Th1 activity [77]. Additionally, a comparison of the cytokine profile between women with recurrent pregnancy loss (RPL) (low Th2 and high Th1 responsiveness) and healthy controls showed that the probability of normal fecundity increased more than tenfold in those with prominent Th2 responsiveness [78]. For patients with PL, low serum levels of mannose-binding lectin (MBL), which is a significant constituent of human innate immunity, are also a newly discovered risk factor [79]. MBL deficiency is thought not only to be involved in the pathogenesis of RA but also to be associated with the occurrence of PL [80, 81]. A study describing the link between the outcome of pregnancy in RA and circulating MBL levels reported that MBL deficiency was a risk factor for miscarriage in patients with RA (p = 0.007) [82].

Regarding the impaired fertility status of women with RA, some researchers have postulated that RA-induced disturbances of immune cells and cytokines could alter the so-called endometrial receptivity of female patients [43]. A diminished number and function of regulatory T cells have been observed in both RA and infertility [83, 84]. Demir also found that individuals with infertility had a higher concentration of interleukin-6 (IL-6), which is a primary mediator in the pathogenesis of RA [85]. Coincidentally, analogous cytokine disorders were also observed in RA and PCOS [86, 87], which made some scholars believe that PCOS and RA populations may possess similar cytokine disturbances (Fig. 3). Patients with PCOS are in a state of chronic low-grade inflammation, which is characterized by unregulated and continuous synthesis and secretion of cytokines, including tumor necrosis factor alpha (TNFα), interleukin-1 (IL-1), and IL-6 [88]. Taking the example of IL-6, multiple studies have suggested that patients with PCOS have high IL-6 levels [86, 89]. Additionally, the development of IL-6 blocking agents, such as biological antirheumatic drugs, has greatly improved the treatment of RA. Similar effects were also reported in PCOS by targeting IL-6 molecules [89].

Peritoneal fluid biopsy from patients with endometriosis revealed that aberrant numbers and functions of almost all types of immune cells led to enhanced humoral immunity and high levels of autoantibodies [60, 66]. These immune dysfunction conditions may contribute to maintaining the survival and invasion of ectopic endometrial cells while producing a range of cytokines and then causing systemic chronic inflammation, eventually leading to the occurrence of RA.

Similar to other complex diseases, the etiology of some reproductive-related diseases also involves genetic factors, such as infertility and endometriosis. Achour et al. found that copy number variations of glutathione S-transferase M1 (GSTM1) were a risk factor for anti-CCP antibody positivity in Tunisian patients with RA [90]. Then, a GSTM1 polymorphism was reported to be strongly associated with ovarian endometriosis and endometriosis-related infertility [91]. As endometriosis genomic studies have gradually increased, it has been postulated that endometriosis and RA share a similar genetic background. Studies have suggested that the pathogeneses of both endometriosis and RA involve chemokine CC motif ligand 21 (CCL21) and HLA-DRB1 [92]. Furthermore, there is another controversial gene, protein tyrosine phosphatase nonreceptor 22 (PTPN22). PTPN22 is well established in RA. For endometriosis, the correlation with PTPN22 variants is dependent on racial background, as the correlation is found in Italians and Brazilians [93, 94] but not in Polish individuals [95].

A significant proportion of patients with RA of childbearing age have expressed that they have a high degree of psychosocial stress, primarily including anxieties about relapse due to drug withdrawal, raising a baby with limited maternal function, side effects of medications on infants, and the inheritance of RA [96]. A trend toward infertility became more pronounced when multiple concerns were present simultaneously [37]. Moreover, Camacho et al. suggested that a history of APO, especially RPL, may psychologically prompt the later development of RA [97]. For endometriosis, others have suggested that stress due to symptoms such as chronic pelvic pain may play a crucial role in the onset of RA [98, 99].

Psychosocial stress often manifests in the form of sexual dysfunction. It was reported that 27–67% of women with RA suffered from sexual problems, manifested as reduced desire and pleasure, pain during intercourse, and erectile dysfunction [100]. The use of antirheumatic treatment has not been found to improve sexual problems, which is attributed to the persistence of fatigue symptoms, although pain is alleviated. A study performed to address the frequency of sexual intercourse in patients with RA suggested that 47.8% of patients showed less sexual activity after the diagnosis of RA. However, the overemphasis on the reduced sexual activity in couples with RA is controversial. A recent comparison of patients with RA in Toulouse trying to conceive showed that the frequency of sexual intercourse did not vary between women who successfully conceived and those who failed [101].

Some published data have shown that unexplained infertility was strongly related to the use of disease modifying anti-rheumatic drugs (DMARDs), such as nonsteroidal anti-inflammatory drugs (NSAIDs), methotrexate (MTX), and prednisone, during the periovulatory period [38, 39, 102, 103]. However, this is not the case with all antirheumatic drugs. A study showed that there was no statistically significant difference between the perinatal use of sulfasalazine (SSZ) and low fertility [38]. Periconceptional and perinatal use of TNF inhibitors appeared to be safe, with no maternal or fetal adverse effects [104, 105]. Researchers have evaluated the safety of adalimumab (a TNF inhibitor) during pregnancy in patients with RA [106]. A prospective multicenter study in Romania reported that planned pregnancies and low-dose corticosteroids (≤ 10 mg/day) reduced the incidence of APO in RA pregnancies (p < 0.001, p = 0.016, respectively) [50]. New expert opinion stated that pregnancy-compatible conventional synthetic DMARDs (csDMARDs) and TNF inhibitors are more favorable protective factors for APO in RA pregnancies compared to high-dose prednisone or long-term NSAIDs [107]. It is necessary to assess relevant risk factors, perform precise individual risk stratification, and strive to control the disease with csDMARDs during pregnancy planning, which is exhaustively recommended by the European League Against Rheumatism (EULAR) and the American College of Rheumatology (ACR) [107‐109].

When exploring the impact of RA on reproductive-related diseases, paternal RA is another consideration. Perez-Garcia et al. observed that the occurrence of miscarriage was strongly associated with paternal RA, and they found that couples in which the man had been diagnosed with RA also had a significantly increased risk of miscarriage [110]. Currently, data on the fertility problems of male patients with RA mainly focus on antirheumatic drugs. In male model rats treated with SSZ, a suppression of epididymal acrosomal membrane protein synthesis was reported, resulting in decreased sperm motility and high rates of morphologically abnormal sperm. However, these side effects were reversible in animal models after withdrawal [111, 112]. MTX was also found to induce chromosomal abnormalities in sperm in animal studies; however, animal data was not entirely applicable to humans. The systematic review conducted by Mouyis et al. indicated that paternal exposure to MTX or other DMARDs had minor negative impacts on pregnancy outcomes, beyond the known adverse effects of SSZ and cyclophosphamide on sperm [113].