The Role of NK‑Cell Dysregulation in Adverse Pregnancy Outcomes Associated with Rheumatic Immune Diseases: Mechanisms and Translational Advances

Abstract

Rheumatic autoimmune diseases induce sustained systemic inflammatory imbalance and substantially elevate the incidence of adverse pregnancy complications, such as recurrent spontaneous abortion, preeclampsia, preterm delivery, and fetal growth restriction. Decidual natural killer (dNK) cells constitute the major lymphocyte subset at the maternal-fetal interface and maintain the homeostasis of immune tolerance, trophoblast invasion and spiral artery remodeling; peripheral NK (pNK) cells differ greatly from dNK cells in phenotype, cytotoxicity and secretory function under the regulation of local uterine microenvironment. Emerging evidence indicates that dysregulated NK‑cell number, subset distribution, receptor signalling and cytokine release may perturb the decidual immune‑vascular microenvironment and contribute to inflammatory placental‑associated pathology, with considerably more robust mechanistic support in antiphospholipid syndrome than in systemic lupus erythematosus, primary Sjögren’s syndrome, rheumatoid arthritis, and systemic sclerosis. Solid preclinical and clinical evidence confirms that excessive NK cell activation participates in placental injury in antiphospholipid syndrome, whereas the direct inflammatory pathogenic cascades underlying gestational complications remain insufficiently clarified for other rheumatic disorders. This work elucidates the physiological heterogeneity of pNK and dNK cells, delineates six core inflammatory axes driving NK-mediated placental injury, characterizes disease-specific aberrant inflammatory phenotypes of NK cells, summarizes translational advances in NK-targeted anti-inflammatory strategies, and identifies existing research bottlenecks. We also put forward prospective research schemes based on single-cell multi-omics and large-scale clinical cohorts to reveal NK-centered inflammatory pathology, which offers theoretical support for immune inflammation regulation in childbearing-age women with rheumatic autoimmune diseases.

Introduction

Patients with rheumatic autoimmune diseases (such as systemic lupus erythematosus, rheumatoid arthritis, antiphospholipid syndrome, etc.) have significantly elevated risks of adverse pregnancy outcomes, including miscarriage, preterm birth, preeclampsia, and fetal growth restriction.^1,2 Mendelian randomization studies have further revealed causal associations between the genetic predispositions to various autoimmune diseases and the aforementioned pregnancy complications.³

A successful pregnancy depends on the precise immune homeostasis at the maternal-fetal interface—both defending against pathogens and tolerating the semi-allogeneic fetus. Rheumatic autoimmune diseases are fundamentally characterized by systemic immune dysregulation, with disordered T/B cell function, production of autoantibodies, and imbalance of inflammatory cytokines, which can directly interfere with the immune tolerance microenvironment required for pregnancy.⁴ In this interactive network, natural killer (NK) cells have emerged as a critical bridge linking maternal autoimmunity to pregnancy outcomes, owing to their central role at the maternal‑fetal interface (comprising 50–70% of decidual lymphocytes in early pregnancy) and their dual functions in immune tolerance and placental development.

Existing literature has primarily focused on single pathological conditions, such as uterine NK cells in recurrent miscarriage^5,6 or peripheral NK biomarkers,⁷ failing to systematically integrate the mechanisms by which NK cells disrupt maternal-fetal interface homeostasis within the broader context of rheumatic immune diseases through multiple pathways. In recent years, key mechanisms that have emerged (such as the C3b-VSIG4 axis,⁸ CXCR4⁺dNK subset,⁹ NFAT residency signals)¹⁰ and emerging technologies (single-cell omics, iNK cell therapy) still lack integrative analysis across diseases. This paper reviews the heterogeneity of NK cell subsets, receptor-ligand interactions, core signaling pathways, and clinical translation progress, aiming to construct a mechanistic framework of NK cell abnormalities driving adverse pregnancy outcomes in the context of rheumatic immune diseases, with the expectation of providing theoretical pathways for risk stratification and targeted interventions.

Given the prominent clinical relevance and existing research base, this review focuses on five major rheumatic autoimmune diseases closely associated with gestational complications: antiphospholipid syndrome (APS), systemic lupus erythematosus (SLE), primary Sjögren’s syndrome (pSS), rheumatoid arthritis (RA), and systemic sclerosis (SSc). These conditions were selected because published studies have reported associations with elevated risks of adverse‑pregnancy‑related complications, display observable NK‑cell immunological perturbations, and span a spectrum from relatively robust mechanistic evidence (APS) to substantial unaddressed knowledge gaps (RA and SSc), helping to illustrate heterogeneity and key bottlenecks within this research field.

A targeted literature search was performed in PubMed and Web of Science from January 2016 to July 2026. Search terms combined keywords including “natural killer cells”, “decidual NK cells”, “rheumatic autoimmune diseases”, “antiphospholipid syndrome”, “systemic lupus erythematosus”, “Sjögren’s syndrome”, “rheumatoid arthritis”, “systemic sclerosis”, “pregnancy”, “recurrent miscarriage”, “preeclampsia”, “fetal growth restriction”. Original research articles and English‑language narrative reviews relevant to NK‑cell dysregulation and adverse gestational outcomes in rheumatic disorders were selected for synthesis.

Overview of Subpopulation Classification and Physiological Functions of NK Cells

NK cells are categorized into two major subpopulations based on the expression density of CD56 and CD16: CD56dimCD16+ predominantly exhibit cytotoxicity, mediating target cell killing through the perforin/granzyme pathway; CD56brightCD16− primarily secrete cytokines (eg, IFN-γ, TNF-α), thereby exerting immunoregulatory functions.^11,12 In different anatomical locations, NK cells exhibit significant phenotypic and functional differentiation, particularly at the pregnancy-related maternal-fetal interface, where uterine/decidual NK (uNK/dNK) cells and peripheral blood NK (pNK) cells demonstrate markedly distinct characteristics (Table 1).

Table 1.

Comparison of Major Characteristics Between Peripheral Blood NK Cells and Decidual NK Cells

Characteristics	Peripheral Blood NK (pNK)	Uterine/Decidual NK (uNK/dNK)
Primary Phenotype	CD56dimCD16+	CD56brightCD16−
Cytotoxicity	Strong	Weak
Primary Functions	Immune surveillance (antiviral, antitumor)	Immune regulation, tissue support
Key Products	Perforin, granzymes	VEGF, PlGF, IFN-γ, IL-811,12
Proportion (Early Pregnancy Decidua)	Not directly localized in the decidua	Comprise 50–70% of lymphocytes in the decidua5
Functional Role	Reflect systemic immune status	Maintain maternal-fetal immune tolerance, promote trophoblast invasion and spiral artery remodeling13

The functional specialization of dNK cells originates from the directional shaping by the local microenvironment (cytotrophoblast cells, cytokines and ligands derived from decidual stromal cells) and a unique transcription factor expression profile (such as the high expression of GATA-3 and Eomes).¹⁴ Single-cell transcriptome studies further reveal that multiple functional subpopulations of dNK cells exist during early pregnancy, with their proportions dynamically evolving over gestational weeks, suggesting that different subpopulations assume distinct roles in embryo implantation, vascular remodeling, and pregnancy maintenance.¹⁵

pNK and dNK exhibit an irreconcilable biological gap in terms of developmental origin, tissue localization, molecular characteristics, and pregnancy-specific functions. Direct study of uNK/dNK is necessary to more accurately represent the immunological environment at the maternal-fetal interface.¹⁶ Extrapolating pNK data to dNK lacks scientific justification and may lead to clinical misjudgment and mechanistic misinterpretation.

The Abnormal Mediation of NK Cells in the Pathogenesis of Adverse Pregnancy Outcomes

The establishment and maintenance of a normal pregnancy depend on a finely tuned immune tolerance balance at the maternal-fetal interface, where NK cells, as the most abundant immune cell population in the decidua, play a central role in regulating trophoblast invasion, spiral artery remodeling, and local microenvironment homeostasis. Abnormalities in the number, phenotype, or function of NK cells can disrupt the immune-vascular-inflammatory regulatory network through multiple pathways, leading to adverse pregnancy outcomes (Figure 1 Created with BioGDP.com).¹⁷

Figure 1.

NK-cell dysregulation in the pathogenesis of adverse pregnancy outcomes. Schematic illustration of multiple partly‑independent, intersecting pathogenic pathways driven by NK‑cell perturbations. These pathways do not follow a fixed sequential chronological order but may act individually or in concert to disturb maternal‑fetal homeostasis and precipitate adverse pregnancy outcomes. ↑ indicates increased expression or frequency; ↓ indicates decreased expression or frequency; cross‑mark (×) represents functional inhibition or pathological blockade.

Abnormal Enhancement of Cytotoxic Function

Normal dNK cells exhibit relatively low cytotoxicity and primarily exert their effects through the secretion of regulatory factors.⁵ Abnormal functional activation leads to a significant increase in NK cell cytotoxicity (such as upregulation of perforin and granzyme expression), which directly damages trophoblastic cells and disrupts the balance of immune tolerance.¹³ Premature activation of NK cells and their shift to a cytotoxic phenotype in late pregnancy can lead to the collapse of maternal-fetal interface tolerance, inducing preterm labor.⁵

Dysregulation of Cytokine and Angiogenic Factor Secretion

Uterine natural killer (uNK) cells regulate endometrial receptivity, spiral artery remodeling, and placental development by secreting angiogenic factors (such as related growth factors) and immunomodulatory factors (such as IL-10, IFN-γ, etc.).⁶ Abnormalities in uNK cell function are characterized by dysregulated secretion of angiogenic factors or excessive release of pro-inflammatory factors, leading to failed spiral artery remodeling, inadequate placental perfusion, restricted trophoblast invasion, and subsequently resulting in placental insufficiency and pregnancy loss.^13,18

Imbalance of KIR-HLA Receptor-Ligand Interactions

The interaction between the killer cell immunoglobulin-like receptors (KIR) expressed by maternal uNK cells and the HLA-C molecules expressed by fetal trophoblast cells is crucial for the establishment of immune tolerance.^6,19 Abnormalities in NK cells can lead to alterations in the KIR repertoire (such as a reduction in inhibitory KIR and an abnormal increase in activating KIR), disrupting the signaling balance; additionally, abnormalities in the soluble HLA-G and KIR2DL4 receptor pathway can interfere with the transcriptional program of tolerance genes,^20,21 thereby increasing the risk of recurrent miscarriage or preeclampsia.^16,22

Organizational Residency Functions and Microenvironment Regulatory Disorders

The NFAT signaling pathway is involved in the tissue residency of uNK cells, and its abnormality can lead to a disruption in the distribution of resident subpopulations, thereby affecting local immune functions.¹⁰ dNK cells interact with TNFSF14 (ligand) and TNFRSF14 (receptor) to inhibit the senescence of decidual stromal cells (DSC); dysfunction of this pathway (such as NK cell exhaustion or receptor deficiency) can lead to excessive aging of the uterine microenvironment, impairing pregnancy maintenance.²³ Changes in the microenvironment can also remodel the phenotype of NK cells (eg, upregulation of activation receptors such as NKG2D), further exacerbating functional dysregulation.²⁴

Systemic Immune Dynamic Imbalance

The level and functional abnormalities of pNK cells may reflect the systemic immune status and indirectly influence the local microenvironment of the uterus. In women with unexplained recurrent pregnancy loss (uRPL), the quantification, phenotypic characteristics (eg, changes in the CD56⁺ subpopulation), and functional dynamics of pNK cells during the pre-pregnancy and early gestational periods are significantly associated with pregnancy outcomes.^7,25 Additionally, hyperglycemic environments induced by metabolic disorders such as gestational diabetes (GDM) may also lead to NK cell dysfunction, contributing to the occurrence of pregnancy complications.²⁶

Dysregulation of the Inflammatory Immune Control System

A normal pregnancy requires the dynamic balance of “implantation (inflammation) - pregnancy maintenance (anti-inflammation) - parturition (inflammation)”.⁵ Aberrant activation of NK cells can disrupt this rhythm, such as when pro-inflammatory responses are prematurely triggered in pathological conditions, leading to uncontrolled inflammation at the maternal-fetal interface and impairing placental function.⁵

Abnormal Characteristics of NK Cells in the Context of Rheumatic Immunological Diseases

As a systemic immunodysregulatory condition, rheumatic and immunological diseases exhibit widespread abnormalities in natural killer (NK) cells, including alterations in their quantity, subpopulation distribution, cytotoxic function, cytokine secretion, and receptor expression. This section focuses on the intrinsic abnormalities of NK cells observed in various rheumatic and immunological diseases under non-pregnant conditions, providing background knowledge for subsequent discussions on their association with adverse pregnancy outcomes.

Imbalance in the Proportions of NK Cell Subsets and Phenotypic Abnormalities

In various rheumatological and immunological diseases, the distribution of NK cell subsets exhibits significant imbalance. For example, in patients with RA, those in sustained remission exhibit an increased proportion of CD8+CD57+KIR2DL1+ NK cells, while active RA patients lack this phenotype in their synovial fluid, suggesting a close correlation between NK cell subset composition and disease activity.²⁷ Additionally, in patients with pSS, the frequency of a CD56hiCD16⁺NKG2Dhi NK‑cell subset is increased; this subset exhibits enhanced cytotoxic potential, distinct from the conventional CD16⁻CD56hi NK subset which is predominantly cytokine‑producing with limited cytotoxic capacity.²⁸ Similarly, in patients with systemic autoinflammatory diseases such as Still’s disease, the frequency of peripheral NK cells is generally reduced, and the predominant phenotype is one of functional exhaustion.²⁹ Similar NK cell exhaustion phenotypes have also been observed in other systemic autoinflammatory diseases, such as VEXAS syndrome.³⁰ These evidences suggest that different rheumatological and immunological diseases can induce specific NK cell subset remodeling.

Aberrant NK Cell Cytotoxic Function

The cytotoxicity of NK cells is generally impaired in rheumatic immune diseases. In patients with Still’s disease, although NK cells exhibit an overly activated state, they show diminished response to cytokine stimulation, with a significant decrease in their killing capacity, and this abnormality is reversible after clinical remission, suggesting it is driven by inflammatory factors.²⁹ In RA patients, the CD16⁺ NK cells enriched in the synovium are also limited in their degranulation capacity and target cell killing efficiency; however, specific NK subsets in the peripheral blood during the remission phase still maintain normal degranulation responses, indicating that functional abnormalities are dependent on the disease stage.²⁷ Mitochondrial quality control disorders may contribute to NK cell dysfunction in SLE.³¹

Cytokine Secretion by NK Cells and Dysregulation of Immune Modulation

Natural killer (NK) cells in rheumatic autoimmune diseases not only exhibit impaired cytotoxicity but also display a significantly disordered cytokine secretion profile. In rheumatoid arthritis (RA) synovium, NK cells can secrete large amounts of IFN-γ, which induces synovial fibroblasts to transform into pro-inflammatory phenotypes (HLA-DR⁺CD90⁺), thereby enhancing the activation of CD4⁺T cells and forming a positive feedback inflammatory loop, a process that can be blocked by JAK inhibitors.³² Conversely, in patients with Still’s disease, NK cells lose their responsiveness to cytokines such as IL-12, IL-15, and IL-18, leading to a reduction in the production of effector molecules such a IFN-γ, indicative of a state of functional exhaustion.²⁹ In patients with pSS, the interaction between NK cells and cDC2 may enhance inflammatory responses, but the specific changes in the cytokine profile need to be analyzed in conjunction with the local microenvironment.²⁸

Aberrant Expression of Receptors and Signaling Pathways

Alterations in the repertoire of NK cell surface receptors constitute a crucial molecular basis for functional abnormalities in rheumatic immune diseases. In antiphospholipid syndrome (APS) patients, even in the non-pregnant state, there is already an imbalance in the expression of inhibitory receptor TIGIT and activating receptor CD226 on peripheral blood NK cells.³³ Peripheral blood NK cells in Systemic Lupus Erythematosus (SLE) patients exhibit high expression levels of TIM-3 and PD-1, which are positively correlated with disease activity.³⁴ In Rheumatoid Arthritis (RA) patients, the upregulation of TIGIT expression on NK cells is associated with suppressed NK cell activation.³⁵ These aberrant receptor expressions provide potential targets for targeted interventions.

Disease Activity and Treatment Effects on NK Cells

Abnormalities in NK cells in rheumatic immune diseases often dynamically change with disease activity. The proportion of CD161 on CD56 NK cells in peripheral circulation is associated with the clinical features and disease activity of pSS.³⁶ In patients with RA in remission, there can be an expansion of protective NK subpopulations, while in active phases, there is a manifestation of functional impairment.²⁷ In adult-onset Still’s disease (AOSD), the exhausted phenotype of NK cells can be reversible after clinical remission.²⁹ In terms of treatment, JAK inhibitors can block the pro-inflammatory circuit induced by IFN-γ in RA,³² calcineurin inhibitors may affect tissue-resident signals of NK cells,¹⁰ and the dose-dependent effects of glucocorticoids on NK cell function also warrant attention. These observations suggest that the assessment of NK cell abnormalities must consider the disease status and the background of medication use.

In summary, in rheumatic and immune diseases, NK cells exhibit a “bidirectional dysregulation”: in some conditions, they are characterized by excessive proinflammation, while in others, functional exhaustion predominates. This complexity suggests that when considering combined adverse pregnancy, the functional status of NK cells must be assessed based on the specific disease type and activity level (Table 2).

Table 2.

Aberrant Types of NK Cells and Representative Diseases in Rheumatic Immune Disorders

Exception Types	Representative Diseases	Main Findings	References
Imbalance in Subpopulation Proportions	RA, pSS, AOSD	Remission phase RA: CD8+CD57+KIR2DL1+NK↑. pSS: CD56hiCD16+NKG2Dhi NK↑. AOSD: total NK↓, Exhaustion phenotype.	[27–29]
Abnormal Cytotoxicity	AOSD, RA, SLE	AOSD: Cytotoxic function reduced, Marker↓. RA synovium: Degranulation limitation (phase dependence). SLE: Defects in Cytotoxicity.	[27,29,31]
Dysregulation of Cytokine Secretion	RA, AOSD, pSS	RA: IFN-γ↑, Pro-inflammatory loop. AOSD: Unresponsive to stimuli, effector molecules↓. pSS: Interacts with cDC2 to promote inflammation.	[28,29,32]
Receptor expression abnormalities	APS, SLE, RA	APS: dysbalance in TIGIT/CD226. SLE: TIM-3+PD-1+↑. RA: TIGIT↑.	[33–35]

Notes: ↑ = increased, ↓ = decreased/reduced.

Abbreviations: RA, Rheumatoid arthritis; pSS, primary Sjögren’s syndrome; AOSD, Adult‑onset Still’s disease; SLE, systemic lupus erythematosus; APS, Antiphospholipid syndrome.

Evidence and Mechanistic Gaps in Rheumatic Diseases with Adverse Pregnancy Outcomes

APS

APS represents the most well-characterized rheumatic autoimmune disease associated with adverse pregnancy outcomes, with the most robust body of evidence supporting aberrant NK cell involvement. Clinical studies have found that there is an imbalance in the expression of the inhibitory receptor TIGIT and the activating receptor CD226 on peripheral and local NK cells in patients with primary APS, which may disrupt the functional homeostasis of NK cells and exacerbate the risk of placental inflammation and pregnancy loss.³³ Multi-omics studies on APS mouse models have shown that there are histological abnormalities such as disordered angiogenesis in APS placentas; at the same time, biological and transcriptomic characterization of uterine NK (uNK) cells has found that their molecular expression profiles have changed, suggesting that abnormal uNK cell function may directly participate in APS-related placental developmental disorders.³⁷ In addition, antiphospholipid antibodies (APLs) can abnormally activate NK cells through the ADCC pathway and mediate trophoblast damage.⁴

SLE and pSS

A reduction in the number of NK cells, defects in cytotoxicity, abnormalities in cytokine secretion, and mitochondrial dysfunction are commonly observed in the peripheral blood of SLE patients, indicating impaired innate immune regulatory capacity.^31,34,38,39 Although the risk of pregnancy complications, such as preeclampsia, disease activity, and recurrence, is significantly increased in SLE patients,^40,41 the current literature lacks direct mechanistic studies linking the aberrations of NK cells in SLE patients to adverse pregnancy outcomes. Inferences can only be indirectly drawn based on the consensus that “dysfunction of NK cells is generally involved in pregnancy complications”.¹³

In patients with pSS, a cytotoxic CD56hiCD16⁺NKG2Dhi NK-cell subset is enriched in peripheral blood, and killer cell immunoglobulin-like receptor (KIR) polymorphisms may further modulate NK-cell functional activity.^28,42 Clinical cohorts and meta-analyses have confirmed that patients with primary Sjögren’s syndrome (pSS) have a significantly higher risk of adverse pregnancy outcomes, such as miscarriage, preterm birth, and fetal growth restriction, compared to the general population.^43,44 However, current research has not yet established a direct causal link between the abnormal phenotypes/functions of NK cells in pSS patients and specific pregnancy complications, and the exploration of underlying mechanisms remains at an initial stage.

Other Rheumatic Diseases

Research confirms that NK cells are involved in the immunopathological process of RA: NK cell infiltration can be detected in peripheral blood, spleen, lymph nodes, and inflamed joints, with their proportions correlating with disease progression.⁴⁵ NK cells are present in synovial fluid and may participate in joint inflammation by secreting pro-inflammatory factors such as TNF-α and IFN-γ.⁴⁶ Subsets of NK cells exhibit quantitative and functional abnormalities in newly diagnosed, untreated RA patients;⁴⁷ variations in TIGIT expression levels may influence NK cell activity;³⁵ high PD-1 expression indicates an exhausted state of NK cells, which may be regulated by microRNAs.⁴⁸ Therapeutic interventions, such as JAK inhibitors and rituximab, can affect the quantity and phenotype of NK cells in RA patients, suggesting an association with disease activity and treatment response.^49,50 Pregnancy often leads to the spontaneous remission of RA symptoms, which has been hypothesized in the literature to be associated with systemic immune regulation during pregnancy. This may involve mechanisms such as “innate and adaptive immune alterations”, “IgG glycosylation”, and “immune regulation mediated by HLA differences.⁵¹ ” However, there is a significant gap in the current literature regarding the dynamics of NK cells during RA pregnancy, with a lack of specific numerical data, subpopulation distribution, functional testing, or clinical correlation data.

In systemic sclerosis (SSc), alterations in NK cell numbers or phenotypes have been associated with complications such as pulmonary arterial hypertension.^52,53 However, there is a lack of reported research on NK cells in the context of adverse pregnancy outcomes.

Current studies demonstrate a prominent uneven distribution of research evidence across autoimmune rheumatic diseases in pregnancy. Specifically, APS has accumulated relatively solid mechanistic and clinical evidence, whereas SLE and SS only present tentative clinical correlations with inadequate mechanistic elucidation. Similarly, pregnancy-related immunological research targeting RA and SSc remains extremely scarce. Such heterogeneous evidence distribution can be largely explained by inter-disease disparities in pathological mechanisms and clinical management paradigms. The detailed NK cell abnormalities, existing research evidence and mechanistic gaps of each disease are summarized in Table 3.

Table 3.

Current Evidence and Mechanistic Gaps of NK Cell Dysregulation in Rheumatic Diseases with Adverse Pregnancy Outcomes

Disease	Key NK Cell Abnormalities	Current Pregnancy-Related Evidence	Primary Gaps
APS	Imbalanced TIGIT/CD226 expression, aberrant uNK transcriptome, and APL-induced NK activation via ADCC4,33,37	Robust evidence confirms NK dysfunction contributes to placental disorders and pregnancy loss4,33,37	Lack of dynamic mechanistic and intervention studies in human pregnancy
SLE	Reduced NK count, impaired cytotoxicity, dysregulated secretion and mitochondrial dysfunction31,34,38,39	Higher gestational complication risk; NK involvement is only indirectly inferred13,40,41	Absence of evidence for causative mechanisms
pSS	Enriched cytotoxic CD16+CD56hi NK subset and KIR polymorphism-modulated NK activity28,42	Elevated adverse pregnancy risk; causal NK-pathology relationship unconfirmed43,44	Lack of maternal-fetal interface data
RA	TIGIT/PD-1-related NK exhaustion and JAK inhibitor modulation (joint inflammation only)35,45,46	No pregnancy-related NK research available	Complete gap
SSc	Altered NK number and phenotype (associated with pulmonary hypertension)52,53	No NK studies on SSc-related adverse pregnancy outcomes	Complete gap

Abbreviations: APS, Antiphospholipid syndrome; SLE, systemic lupus erythematosus; pSS, primary Sjögren’s syndrome; RA, Rheumatoid arthritis; SSc, Systemic sclerosis.

In terms of disease pathogenesis, pregnancy complications secondary to APS arise from a well-defined and relatively unified pathogenic cascade, wherein antiphospholipid antibodies trigger placental thromboinflammation. This streamlined pathological mechanism enables standardized and targeted mechanistic research design. In comparison, adverse pregnancy outcomes associated with RA and SSc are multifactorial and confounded by fluctuating systemic disease activity, widespread multi-organ involvement, and heterogeneous medication regimens. Additionally, clinical recommendations for RA and SSc pregnancy management advocate the avoidance of high-dose prednisone and nonsteroidal anti-inflammatory drugs (NSAIDs) throughout gestation, further increasing the complexity of research implementation and mechanistic validation.

From the perspective of clinical research orientation, the standardized, guideline-based management of APS pregnancy, with anticoagulation as the core therapeutic strategy, has propelled extensive clinical outcome investigations and in-depth mechanistic explorations. In contrast, the antenatal care of RA and SSc emphasizes individualized, multidisciplinary dynamic monitoring and treat-to-target strategies. Such personalized clinical frameworks prioritize iterative optimization of clinical practice over standardized mechanistic research, ultimately leading to deficient basic and translational evidence concerning their gestational immune regulatory mechanisms.

Future research should prioritize the following directions: dynamic profiling of disease-specific NK cell subsets (particularly uNK cells), receptor signatures (TIGIT/CD226, KIRs), and downstream signaling pathways within the placental microenvironment across gestational stages; prospective cohort studies to validate the predictive value of NK cell phenotypic and functional parameters for pregnancy outcomes; and translational exploration of NK cell-targeted intervention strategies, including receptor modulation and cell-based therapy, to improve gestational safety in rheumatic autoimmune diseases.

In summary, as a pivotal immune regulator at the maternal-fetal interface, NK cell abnormalities and their pathogenic roles in APS-related adverse pregnancy outcomes have been validated by substantial preclinical and clinical evidence. However, mechanistic insights and translational research regarding NK cell dysregulation in most rheumatic autoimmune diseases remain insufficient. Multidisciplinary collaborative investigations are urgently required to fill these research gaps and advance the clinical translation of NK cell-based gestational immune regulation strategies.

Advances in Clinical Translational Research

Exploration of Biomarkers

Dynamic monitoring of changes in peripheral blood NK (pNK) cells can assist in predicting pregnancy outcomes. Prospective studies have shown that a decrease in pNK levels in the early pregnancy is an independent predictor of live birth in patients with unexplained recurrent pregnancy loss (uRPL).⁷ The expression level of CD56⁺ in endometrium of uterine NK (uNK) cells increases in patients with recurrent miscarriage (RM) and recurrent implantation failure (RIF).¹⁶ A reduction in the CXCR4⁺CD56⁺ dNK subset is associated with Th1/Th2 imbalance.⁹ Furthermore, specific KIR/HLA-C combinations and abnormal degranulation activity of uNK cells are also linked to adverse pregnancy outcomes.⁵⁴

In patients with preeclampsia, dNK cells exhibit abnormal phenotypic transformation and functional hyperactivity: single-cell transcriptomic analysis reveals that NK cells differentiate into an ITGAE+CD160+ tissue-resident subset with enhanced cytotoxicity, in which cytotoxicity-related genes are significantly upregulated, and this subset forms a positive feedback loop with SPP1 signaling derived from macrophages, thereby amplifying local inflammatory responses.⁵⁵ This state of immune imbalance may drive an anti-angiogenic phenotype through multiple pathways: on one hand, the hyperactivation of NK cells leads to an increased release of pro-inflammatory cytokines, such as TNF-α and IFN-γ, which can directly stimulate trophoblastic and endothelial cells of the placenta to upregulate sFlt-1 expression and inhibit PlGF synthesis;⁵⁶ on the other hand, impaired trophoblast invasion and insufficient placental perfusion mediated by NK cells induce hypoxic stress, which further exacerbates the imbalance of the sFlt-1/PlGF ratio through unknown mechanisms such as LAT1-NRF2.⁵⁷ Clinical evidence demonstrates that patients with preeclampsia exhibit a significant increase in the proportion of cytotoxic NK cells (CD56+CD16+perforin+) in peripheral blood, while the subset of NK cells with immunoregulatory functions (CD56+CD16-CD279+) is reduced. Furthermore, this phenotypic alteration is closely associated with KIR ligand loss.⁵⁸ It is noteworthy that an sFlt-1/PlGF ratio ≥ 38 has been demonstrated to predict the risk of developing severe preeclampsia within two weeks (adjusted hazard ratio 5.36, AUC = 0.88),⁵⁹ while abnormal NK cell function may initiate this angiogenic imbalance cascade early in the disease process. Therefore, NK cells are not only key effector cells in the immunopathology of preeclampsia but may also serve as an important bridge connecting the disruption of maternal-fetal immune tolerance with systemic endothelial dysfunction, thereby providing a theoretical basis for the development of combined diagnostic strategies targeting the immune-angiogenic axis. However, current standardized methods for uNK detection and reference ranges are lacking, which limits clinical application.

Targeted Intervention and Therapeutic Strategies for NK Cells

Available data show distinct NK-modulating profiles among antirheumatic agents. The NK‑cell‑modulating properties and pregnancy‑safety profiles of commonly used antirheumatic agents are summarized in Table 4. Notably, most available data regarding drug‑mediated NK‑cell modulation are derived from peripheral‑blood samples; evidence directly reflecting decidual NK‑cell responses remains scarce. A direct causal chain linking drug-induced NK abnormality to adverse pregnancy outcomes has not yet been established, but NK-suppressive effects should be considered in medication selection for women of childbearing age. Multidisciplinary consultation between rheumatologists and obstetricians is recommended to prioritize pregnancy-compatible agents and avoid potent NK inhibitors. Most low-risk drugs are also compatible with breastfeeding per current guidelines.⁶⁹

Table 4.

Summary of Antirheumatic Agents: NK-Cell-Modulating Effects and Pregnancy-Related Safety

Agent Class and Drug	Effects on NK-Cell Phenotype and Function	Pregnancy-Safety Notes	References
Methotrexate	Suppresses NK-cell function; reduces NK-cell degranulation and IFN-γ secretion	Contraindicated in pregnancy (teratogenic); discontinue ≥3 months before conception	[60–62]
Hydroxychloroquine	Limited data; may indirectly modulate NK cells via immunomodulation	Relatively safe in pregnancy; standard dose ≤400 mg/day without clear teratogenic risk	[63–65]
Adalimumab	Inhibits NK-cell degranulation (20.9% vs 31.3%, p=0.04); reduces IFN-γ secretion (17.4% vs 29.7%, p=0.007); impairs NK-cell killing activity	Permitted in pregnancy, especially first and second trimesters; discontinue in late pregnancy to reduce fetal exposure	[60,61,66,67]
Etanercept	Likely inhibits NK-cell function similar to other TNF inhibitors	Permitted in pregnancy; well-supported evidence	[61,66,67]
Rituximab	Indirectly affects NK cells via B-cell depletion; may enhance ADCC	Use with caution in pregnancy; only when benefit outweighs risk; may cause neonatal B-cell depletion	[61,67]
Tofacitinib	May affect NK-cell function via JAK1/3 pathway; limited data	Contraindicated in pregnancy; teratogenic in animal studies	[61,66,67]
Prednisone/Prednisolone	High dose may suppress NK-cell function; low dose has minor effects	Permitted in pregnancy; placental metabolism to inactive form; dose-dependent increased preterm birth risk	[61]
Non‑Steroidal Anti‑Inflammatory Drugs	May mildly affect NK-cell function; limited data	Short-term use in first/second trimester relatively safe; contraindicated in late pregnancy (premature ductus arteriosus closure)	[61,68]

Notes: NK-cell effects are summarized from in-vitro, animal and human peripheral-blood studies; direct evidence on human decidual NK cells under most drugs remains limited. Pregnancy safety statements follow EULAR recommendations for rheumatic and musculoskeletal diseases in reproduction.

In terms of long-term prospective intervention strategies, interventional strategies targeting dysfunctional NK cells are progressing from basic research to clinical investigation. In APS, NK cells are abnormally activated through antibody-dependent cellular cytotoxicity (ADCC), suggesting that blocking this pathway may represent a potential therapeutic target.⁴ Nanoparticles targeting CD2 for delivering interleukin-2 can simultaneously induce regulatory T cells and NK cells producing transforming growth factor-β, thereby restoring immune tolerance, which holds potential therapeutic value in recurrent pregnancy loss associated with autoimmune diseases.⁷⁰ In vitro-induced CD49a⁺ immature NK (iNK) cells exhibit angiogenic functions similar to dNK cells and can improve fetal growth in mouse models.⁷¹ Nevertheless, the above novel intervention approaches are only verified in animal model studies at present, with a lack of clinical data on human safety and fetal effects.

Challenges and Future Directions

Despite the widely recognized role of NK cells as a critical bridge between rheumatic autoimmune diseases and adverse pregnancy outcomes, the translation to clinical application remains fraught with numerous challenges. Firstly, the function of NK cells is highly dependent on the tissue microenvironment, and pNK cells are inadequate in reflecting the true state at the maternal-fetal interface, while uNK cells themselves exhibit spatiotemporal dynamic changes in functional subgroups. The integration of high-resolution single-cell multi-omics and spatial transcriptomics may potentially overcome the heterogeneity analysis bottleneck. Secondly, clinical studies are plagued by inconsistent detection methods, weak association between pNK and uNK, and conflicting conclusions, necessitating the establishment of internationally standardized detection consensus and the design of large-scale prospective multicenter cohorts. Integrating NK subgroups, KIR/HLA typing, and disease activity to construct a “maternal-fetal immune scoring” model is imperative. Lastly, targeted interventions face specificity and safety bottlenecks—systemic regulation of NK cells may impact host defense, and safety assessments during pregnancy are insufficient. Future efforts should prioritize evaluating the regulatory effects of existing anti-rheumatic drugs on NK cells to optimize treatment regimens, explore localized precise delivery (eg, decidual-targeted nanocarriers), or engineered NK cell-derived extracellular vesicles, and utilize humanized maternal-fetal interface models (organoid co-culture, placenta-on-a-chip) to accelerate translation. Through these efforts, it is anticipated that a risk stratification and individualized pregnancy management system based on NK cell status can ultimately be established.

Conclusions

Decidual NK cells are central to maintaining maternal-fetal immune tolerance, trophoblast invasion and spiral artery remodeling. Systemic immune dysregulation stemming from rheumatic autoimmune diseases can perturb the number, subset distribution, receptor signalling and secretory function of peripheral and decidual NK cells and may contribute to the development of adverse pregnancy outcomes. Distinct NK-cell-related disorder signatures exist across APS, SLE, pSS, RA and SSc, yet supporting research evidence is markedly uneven: only APS possesses solid mechanistic proof linking NK-cell dysfunction to gestational complications, whereas critical causal knowledge gaps persist for the remaining rheumatic conditions. Though peripheral NK biomarkers show preliminary predictive value for pregnancy prognosis, standardized detection criteria for decidual NK cells are lacking; conventional antirheumatic drugs exert disparate NK-modulating effects with clear pregnancy contraindications for potent NK suppressors, and novel NK-targeted therapies have only been validated in animal models without human clinical data. Key translational bottlenecks include the tissue heterogeneity between peripheral and decidual NK cells, inconsistent clinical testing methods and insufficient safety of systemic NK regulation. Given that approximately 2% of women of reproductive age worldwide suffer from rheumatic and immunological diseases, failure to implement precise immunological risk stratification may lead to avoidable adverse maternal-fetal events, increase the medical burden associated with obstetric complications, and result in suboptimal pregnancy consultation strategies for high-risk patients. So future work needs to apply single-cell spatial multi-omics to identify disease-specific decidual NK signatures, conduct large-scale multicenter prospective cohorts to build NK-based maternal-fetal immune risk scoring systems, clarify how medications alter NK function to worsen pregnancy risks, and develop decidua-targeted precise interventions, which will ultimately support individualized risk stratification and pregnancy management to safeguard maternal and fetal safety in rheumatic autoimmune patients of childbearing age.

Funding Statement

This research was funded by the Key Science and Technology Program of Jinhua City, grant number 2024-3-067.

Data Sharing Statement

Data sharing is not applicable to this article as no data were created or analyzed in this study.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

The authors declare that there are no conflicts of interest.