Evidence that atherosis of the spiral artery represents atherosclerotic lesions similar to those of native and transplant-induced atherosclerosis: implications for understanding the pathophysiology of obstetrical syndromes and long-term cardiovascular risk

Atherotic lesions demonstrated: (1) a phenotypic pattern of macrophage and endothelial cell activation (ICAM-1 and HLA-DR cellular expression) and inflammation (inflammatory markers NF κ B and CRP), (2) activation of extravillous trophoblast (HLA-DR positive), (3) lipid deposition (Oil Red O staining) in macrophages, (4) expression of the scavenger receptors CD36 in macrophages, and (5) fibrin deposition. These features were observed in atherotic lesions regardless of obstetrical complications. The histopathologic and immunopathologic features of atherosis were similar to those observed in native coronary atherosclerosis and transplant-induced atherosclerosis. Therefore, the evidence suggests that the pathogenesis of atherosis and atherosclerosis are similar. This concept has important implications for the healthcare of women who have undergone pregnancy complications with atherotic lesions in the placenta. Moreover, the findings suggest that strategies developed to prevent or treat atherosclerosis may have a role in the prevention or treatment of vascular pathology in the placental bed in pregnant women. The lesions now recognized as acute atherosis were first hinted at by Williams in 1915 in 2 cases of premature separation of a normally implanted placenta. Williams reported degenerative changes in the intima of small decidual arteries and thickening beneath the endothelium, findings now interpreted as indicative of early vascular injury. 108 In 1945, Arthur T. Hertig provided a more complete description of the lesion in placentas of patients with preeclampsia, identifying foam cells, inflammation, and fibrinoid necrosis in spiral arteries. 109 This was expanded in 1950 by Sexton, Hertig, and Reed et al 110 who reported similar lesions in a case series of 476 patients with placental abruption. 110 The term “acute atherosis” was introduced by Zeek and Assali in 1950 to reflect the lesion’s morphologic resemblance to atherosclerosis, particularly the presence of lipid-laden macrophages and intimal damage. Importantly, they also linked these vascular changes to the development of placental infarctions, emphasizing their pathophysiologic significance. 111 The suffix “-osis”—rather than “-sclerosis”—was chosen to distinguish this acute, pregnancy-specific, and self-limited process from the chronic, fibrotic, and proliferative nature of typical atherosclerosis. Subsequent studies identified atherosis in other obstetrical syndromes such as preterm labor with intact membranes, preterm PROM, fetal death, and antiphospholipid antibodies, among others. 47 , 49 , 50 , 78 , 107 , 112 Atherosclerosis is characterized by endothelial activation and dysfunction, often triggered by factors such as hyperlipidemia, hypertension, or oxidative stress. 113 , 114 Endothelial dysfunction leads to increased vascular permeability, allowing for the infiltration of low-density lipoproteins (LDLs) into the intima, where they undergo oxidative modification. Oxidized LDL attracts circulating monocytes, which migrate into the intima and differentiate into macrophages. 115 These macrophages engulf oxidized lipids and become foam cells, forming the earliest visible lesion—known as the fatty streak. As the lesion progresses, there is ongoing chronic inflammation, characterized by the accumulation of macrophages, T lymphocytes, and production of cytokines. 116 T cells recognize antigens such as oxidized LDL-derived peptides, heat shock proteins, and modified phospholipids, presented by antigen-presenting cells in the vascular wall. These antigens trigger adaptive immune responses that contribute to plaque inflammation, progression, and instability (some microbial antigens and autoantigen complexes may also play a role via molecular mimicry or bystander activation). Smooth muscle cells migrate from the media into the intima, proliferate, and contribute to extracellular matrix production, leading to intimal thickening and the development of a fibrous cap. 117 Beneath this cap lies a necrotic core, composed of dead foam cells, lipid debris, and cholesterol clefts, often accompanied by calcification in advanced stages. 118 Neovascularization may occur within the plaque, increasing the risk of intraplaque hemorrhage. 119 As inflammation and matrix degradation persist, the fibrous cap may rupture or erode, exposing thrombogenic material to the bloodstream and potentially resulting in thrombus formation and acute vascular events such as myocardial infarction and stroke. 120 These histologic features reflect the dynamic and complex nature of atherosclerosis as a chronic inflammatory and lipid-driven disease. Atherosclerosis can arise in 2 distinct clinical contexts: native atherosclerosis and transplant-induced (or allograft) atherosclerosis. Native atherosclerosis develops gradually over the decades in response to traditional cardiovascular risk factors and reflects the chronic interplay between lipids, immune activation, and vascular injury. 120 In contrast, transplant-induced atherosclerosis is characterized by a diffuse, concentric intimal hyperplasia affecting both large and small arteries. 121 , 122 Unlike native plaques, which are often focal and eccentric, transplant atherosclerosis is driven primarily by immune-mediated injury, including T cell–mediated and antibody-mediated rejection, and is influenced by alloimmune responses to donor antigens. 123 Lesions in kidney, heart, and liver transplants, 73 – 77 , 79 , 116 , 124 – 148 and bronchiolitis obliterans of lung transplants, 149 – 154 represent examples of transplant-induced atherosclerosis. Atherosis of the spiral arteries, which we propose be named atherosclerosis of the spiral arteries, is a unique vascular lesion because it emerges specifically in the context of pregnancy, a state in which the placenta functions as a semi-allograft, expressing paternal antigens that are foreign to the maternal immune system. 46 Unlike native atherosclerosis or transplant vasculopathy, which develop in systemic vessels or those of the transplanted organ (graft), atherosis of the spiral arteries occurs at the maternal–fetal interface, particularly in the decidual segment spiral arteries. 79 The presence of trophoblast cells, which invade and remodel these vessels, introduces a graft-like component that may provoke maternal immune activation, especially in genetically predisposed individuals. This raises the possibility that atherosis of the spiral arteries reflects a chronic, subclinical vascular or immunologic disorder that preexists pregnancy and is exacerbated by the immunological challenge of trophoblast invasion. 155 The convergence of maternal cardiovascular risk factors with alloimmune recognition of the trophoblast appears to drive the development of this lesion, highlighting atherosclerosis of the spiral arteries as a distinct form of immune-mediated vascular pathology. 45 , 46 Our findings indicate that atherosis is characterized by endothelial and macrophage activation, as evidenced by the expression of ICAM-1 and HLA-DR, and the nuclear translocation of NF- κ B, a transcription factor that orchestrates inflammatory gene expression. Both endothelial cells and CD68-positive macrophages expressed CRP, particularly in its monomeric isoform, which is known to be proinflammatory. 102 Monomeric CRP facilitates the release of IL-1 ß , IL-6, and TNF- α and soluble IL-6 receptor by macrophages and foam cells. 156 – 168 It also enhances NF κ B activation in foam cells, propagating the local inflammatory response through cytokine production, leukocyte recruitment, and cellular activation. 162 , 169 – 175 Intravascular inflammation is likely to contribute to impair spiral artery remodeling. 79 , 176 Activated endothelial cells may impair or fail to support trophoblast invasion, resulting in incomplete spiral artery transformation. 177 An alternative hypothesis is that ICAM-1 upregulation may follow failure of vascular transformation. 79 Inflammatory injury may further support growth factor–mediated vascular proliferation in atherotic lesions. 178 Immune cell profiling in atherosis of the spiral arteries reveals infiltration of CD3 + CD8 − FoxP3 − T cells and CD56 + NK cells, with a relative paucity of regulatory FOXP3 + T cells, suggesting that nonregulatory CD4 + T cells may contribute to lesion formation. 179 , 180 Whether T-cell infiltration is a primary driver or a consequence of atherosclerotic changes remains unclear. 179 The inflammatory environment observed in atherosis of the spiral arteries also affects extravillous trophoblast. In normal pregnancy, trophoblasts do not express HLA-DR, but in atherotic lesions, HLA-DR expression is detected, likely in response to the local inflammatory milieu. This aberrant expression may be induced by monomeric CRP and cytokines and contribute to a phenotype resembling antigen-presenting cells. 181 , 182 The presence of HLA-DR in atherosis of the spiral arteries aligns with findings in both native and transplant-induced atherosclerosis, where upregulation of class II major histocompatibility complex molecules is a hallmark of vascular inflammation and immune activation. 183 – 187 The acquisition of HLA-DR expression by trophoblasts could impair their invasive capacity, further disrupt spiral artery remodeling, and perpetuate vascular pathology. Our findings with Oil Red O staining provide direct histological evidence of neutral lipid accumulation in atherotic lesions, confirming the presence of foam cells rich in triglycerides and cholesteryl esters. The findings of this study confirm our previous observation. 79 Oil Red O staining (is specific for lipids in frozen and paraffin sections) corroborates the immunohistochemical findings of CD36 and supports the lesion being renamed atherosclerosis of the spiral arteries as a placental analog of early-stage atherosclerosis. Foam cells in these lesions result from inflammatory and oxidative stress and are derived from both macrophages and vascular smooth muscle cells, consistent with observations in atherosclerosis. 188 , 189 Increased levels of oxidative stress markers such as 8-iso-prostaglandin F 2 α in the decidua of patients with preeclampsia further suggests the presence of lipid peroxidation in these lesions. 190 This molecule, likely generated through phospholipase A 2 activation, has been implicated in upregulating lectin-like oxidized low-density lipoprotein receptor-1, another receptor that promotes the uptake of oxidized LDL and the formation of foam cells. 191 Together, these findings suggest that atherosis of the spiral arteries shares fundamental pathophysiological features with native atherosclerosis, including lipid deposition, oxidative stress, and inflammatory cell infiltration, adapted to the unique immunologic context of pregnancy. We report here for the first time that lipid-laden CD68 + macrophages express CD36, a class B scavenger receptor involved in lipid uptake and foam cell formation. CD36 mediates the internalization of oxidized LDL and binds a variety of ligands including thrombospondin-1, oxidized phospholipids, and long-chain fatty acids. 192 CD36 expression in spiral artery lesions parallels findings in atherosclerotic plaques and supports its role in mediating inflammatory lipid accumulation. Lipid profiling studies of the decidua basalis of preeclamptic patients demonstrate elevated phospholipids, cholesterol, and lipid peroxides, highlighting the presence of an atherogenic environment. 193 – 195 Moreover, increased susceptibility of LDL to oxidation in pregnancies complicated by fetal growth restriction may promote foam cell generation via CD36. 196 , 197 Importantly, CD36 can interact with monomeric CRP, resulting in NF κ B activation and enhanced uptake of oxidized lipids, contributing further to the formation of foam cells. 173 , 175 , 181 , 198 , 199 These findings underscore the potential role of CD36 in integrating lipid metabolism and inflammatory signaling in the pathogenesis of atherosis of the spiral arteries. In this study, we identified fibrin deposition within atherotic lesions using a fibrin-specific antibody rather than one directed against fibrinogen, thereby providing more definitive evidence of localized thrombin activation and fibrin formation. This distinction is critical, as fibrinogen immunoreactivity can represent precursor protein or plasma leakage, whereas detection of cross-linked fibrin indicates activation of the coagulation cascade and clot stabilization. The presence of fibrin within the lesions underscores the role of endothelial injury and inflammation-induced coagulation in the pathophysiology of atherosis and suggests that this lesion is not merely inflammatory or lipid-laden but also prothrombotic. These findings gain further importance in the context of growing interest in anticoagulant therapies, such as low molecular weight heparin, for the prevention of obstetrical syndromes—particularly preeclampsia and fetal growth restriction. 200 – 205 The demonstration of true fibrin deposition within atherotic lesions supports the rationale for targeting maternal vascular thrombosis as part of a broader strategy to improve uteroplacental perfusion and pregnancy outcomes in high-risk populations. Atherosis of the spiral arteries is increasingly understood to result from immune dysregulation at the maternal–fetal interface, particularly involving interactions between invading trophoblasts and maternal immune cells inflammation. 124 , 206 – 209 In normal pregnancy, extravillous trophoblasts invade the maternal decidua and spiral arteries, where they are recognized by uterine natural killer cells and T cells through the detection of paternally derived human leukocyte antigen-C antigens expressed on the trophoblast surface cells. 79 , 210 – 214 These interactions are typically regulated to promote immune tolerance and allow for successful vascular remodeling. However, in certain individuals, maternal–fetal histocompatibility mismatch or altered immune recognition may disrupt this balance, triggering inappropriate maternal immune activation. Such responses can impair trophoblast invasion, induce local inflammation, and hinder the physiological transformation of spiral arteries, thereby predisposing to atherosis formation. 213 , 215 Intravascular inflammation that arises later in pregnancy, particularly in the context of preeclampsia, spontaneous preterm labor, preterm premature rupture of membranes, SGA fetuses, or maternal systemic infections, may aggravate preexisting or subclinical atherotic lesions. 216 – 220 This proinflammatory state can lead to activation of thrombin, 221 – 227 upregulation of the coagulation cascade, and decidual necrosis, ultimately predisposing to the separation of a normally implanted placenta—placental abruption. 216 , 217 , 219 , 220 This immune-based model of atherosis of the spiral arteries is supported by findings that link maternal T and NK cell activity with defective decidual vascular remodeling and has been proposed in our prior work. 124 , 215 , 228 The concept has since gained broader acceptance, as reviewed by Staff et al. 46 A maternal inflammatory response against fetal antigens—in essence, an alloimmune reaction—may lead to decidual inflammation, vascular injury, and eventual deposition of lipids and fibrin, hallmarks of the atherotic lesion. 229 – 236 This immunopathologic process situates atherosis within the broader spectrum of immune-mediated vascular disorders of pregnancy.

The clinical characteristics of cases with and without spiral artery atherosis are shown in Table 1 . We found that placentas with spiral artery atherosis and failure of physiologic transformation corresponded to mothers who delivered at a markedly lower gestational age ( Table 1 ). Of the 134 placentas without atherosis, 56.7% (76/134) had spiral arteries with physiologic transformation and 43.3% (58/134) had spiral arteries with failure of physiologic transformation (complete or partial) ( Table 1 ). Of the 134 placentas with atherosis, 93.3% (125/134) were from women with a complicated pregnancy and 6.7% (9/134) from uncomplicated pregnancies. Of the 134 placentas without spiral artery atherosis, 127 (94.8%) were from women with a complicated pregnancy and 7 (5.2%) from uncomplicated pregnancies ( Table 1 ). Atherosis was identified in the placentas of patients with preeclampsia (n=73) (early-onset preeclampsia [n=39] and late-onset preeclampsia [n=34]), an SGA neonate (n=31), preterm labor with intact membranes (n=8), preterm PROM (n=9), fetal death (n=4), and uncomplicated pregnancies (n=9) ( Table 1 ). Among placentas with preeclampsia, atherosis was identified mainly in the placentas of neonates below the fifth percentile in early-onset preeclampsia (n=36/39) ( Table 1 ). We identified a total of 1675 spiral arteries across the 1488 block samples evaluated (500 mm 3 each; median: 5 per case) from 268 placentas. Information across all arteries was summarized per case. This study then focused on the evaluation of the percentage of lesions (across all vessels) per case and the histopathologic and immunopathologic analyses of the cell composition of arteries with and without atherosis. The arteries were categorized according to the presence of physiologic transformation, failure of physiologic transformation (complete or partial), and spiral artery atherosis as shown in Table 2 . Quantitative evaluation of basal plate arteries showed a substantially higher number and percentage of vessels with complete and partial failure of physiologic transformation and atherosis in (1) placentas from pregnancies complicated by lower birthweight percentiles, (2) placentas from complicated pregnancies, and (3) placentas from pregnancies complicated by early-onset preeclampsia and lower birthweight percentiles ( Table 2 ). Arteries with atherosis had numerous Oil Red O-positive cells ( Figure 1 ) confirmed to be macrophages ( Figure 2 ), which were not observed in arteries with either physiologic transformation or failure of physiologic transformation without atherosis. Atherosis lesions were von Willebrand factor-positive and α -actin-positive ( Table 3 ) and had extensive areas of fibrin deposits within the wall embedding CD68-positive macrophages ( Figure 2 ). Evidence of activation and inflammation in atherosis was confirmed by the expression of ICAM-1 and HLA-DR in endothelial cells and macrophages ( Figure 3 ), and the demonstration of nuclear translocation of NF κ B and CRP in endothelial cells and within Oil Red O-positive/CD68-reactive foam cells ( Figure 4 ), respectively. Macrophages in atherosis showed expression of scavenger receptor CD36, which was not identified in spiral arteries with physiologic transformation or with failure of physiologic transformation ( Figure 5 ). The immunohistochemical characteristics of atherosis, as well as other conditions of spiral artery transformation in the placental basal plate, were the following: (1) physiologic transformation of the spiral arteries had complete muscular wall replacement by cytokeratin-positive trophoblasts that surrounded the von Willebrand factor-positive endothelial lining of the vessels, (2) spiral arteries with failure of physiologic transformation maintained a smooth-muscle alpha actin-positive muscular wall surrounding the von Willebrand factor-positive endothelial lining of the vessels and cytokeratin-positive trophoblastic cells (interstitial extravillous trophoblasts) surrounded but did not invade the arterial walls, and (3) spiral arteries with partial failure of physiologic transformation had a combination of smooth muscle cells and intra-arterial trophoblasts in the arterial walls surrounding the von Willebrand factor-positive endothelial lining of the vessels. Atherosis was present in spiral arteries with failure of physiologic transformation (either complete or partial) and had the presence of foam cells Oil Red O-positive in the arterial intima ( Figure 1 ) surrounding the von Willebrand factor-positive endothelial lining of the vessels. The immunohistochemical characteristics of spiral arteries with physiologic transformation, failure of physiologic transformation (complete or partial), and atherosis are shown in Table 3 . The quantitative analysis of cellular components in spiral arteries with atherosis and without atherosis stratified according to arteries with physiologic transformation, and failure of physiologic transformation is depicted in Table 4 . Smooth muscle cells were predominantly found in atherosis and failure of physiologic transformation, and intra-arterial trophoblast cells in physiologic transformation ( P =.0001). The percentage of endothelial cells was similar in arteries with atherosis and arteries without atherosis (with and without physiologic transformation) ( P =.12). Oil Red O-positive foam cells were found in atherosis but not in physiologic transformation or failure of physiologic transformation ( P =.0001). CD68-positive and CD36-positive macrophages were found predominantly in atherosis ( P =.0001). Fibrin was found in atherosis (100%) but not in arteries without atherosis ( P =.0001). Atherosis lesions had a considerably higher percentage of macrophages and endothelial cells positive for ICAM-1 (48.4% vs 13.8%), HLA-DR (48.0% vs 13.4%), NF κ B (48.9% vs 7.6%), and CRP (17.6% vs 0%) than spiral arteries with failure of physiologic transformation, but none of them were identified in arteries with physiologic transformation ( Table 4 ). The histopathologic and immunopathologic characteristics of vessels with physiologic transformation, failure of physiologic transformation, and atherosis are shown in Figure 6 . The histopathologic and immuno pathologic-defined cell phenotypes of atherosis and vessels without atherosis (with or without physiologic transformation) are shown in Supplemental Table 1 . The morphologic features of atherosis were consistent among different complications of pregnancy (preeclampsia, SGA, preterm labor, preterm PROM, or fetal death). Different forms of atherosclerosis such as native coronary atherosclerosis, transplant-induced atherosclerosis, and spiral artery atherosis consistently shared similar histopathologic and immunopathologic characteristics such as Oil Red O-positive and CD68-positive macrophages; vessel wall fibrin; ICAM-1-positive and HLA-DR-positive endothelial cells; and NF κ B-positive, CRP-positive, and CD36-positive cells, but these features were not present in unaffected arteries ( Figure 7 ; Supplemental Table 2 ).

The findings of this study have important implications for maternal cardiovascular health, obstetrics, and long-term preventive care in women with pregnancy complications. The demonstration that atherosis shares morphologic, immunologic, and molecular features with both native and transplant-induced atherosclerosis underscores its relevance beyond pregnancy and supports the concept that atherosis may be an early manifestation of systemic vascular susceptibility inflammation. 124 , 206 – 209 First, recognizing atherosis as a form of immune-mediated vascular injury supports the concept that women who exhibit this lesion during pregnancy may be at an increased long-term risk for cardiovascular disease. Atherosis involves chronic inflammation, lipid accumulation mediated by CD36, endothelial activation, and fibrin deposition—hallmarks of atherogenic pathology. 173 , 175 , 181 , 198 , 199 Therefore, the detection of atherosis in the placenta could serve as a sentinel event that identifies women in need of cardiovascular risk screening and early intervention postpartum. Second, strategies currently used for the early detection and prevention of atherosclerosis—such as assessment of lipid profiles, inflammatory markers (eg, CRP), and imaging of vascular health—may be repurposed for postpartum surveillance in women who experienced atherosis-related complications, such as preeclampsia, fetal growth restriction, spontaneous premature labor, or fetal death. These approaches could help stratify cardiovascular risk in a previously under-recognized population of young women. 200 , 201 Third, the mechanistic similarity between atherosis and atherosclerosis raises the possibility that therapeutic agents with vascular protective effects in nonpregnant populations may also have utility in pregnancy. For instance, statins—widely used for primary and secondary prevention of atherosclerotic disease—have shown promise in reducing the risk of preeclampsia and improving placental function by mitigating inflammation, improving endothelial function, and reducing oxidative stress. 191 , 192 Emerging experimental observations in animal models and results of clinical trials data support the safety and potential efficacy of pravastatin in women at high risk for preeclampsia, 237 – 252 validating the relevance of shared pathophysiologic pathways. 200 Moreover, the vascular inflammation, endothelial activation, and prothrombotic features of atherosis may help explain the efficacy of low-dose aspirin in preventing preeclampsia 253 , 254 and, in some cases, spontaneous preterm birth. 255 – 257 Aspirin improves uteroplacental perfusion by inhibiting platelet aggregation and thromboxane synthesis, while also exerting anti-inflammatory effects through cyclooxygenase inhibition 258 – 261 and NF κ B modulation—mechanisms relevant to the vascular pathology of atherosis of the spiral arteries. The evidence that reduces the frequency of preterm preeclampsia is now persuasive. 255 , 262 , 263 Several studies have shown that aspirin may reduce the rate of spontaneous preterm birth. 255 The ASPIRIN trial 256 found that 81 mg aspirin reduced spontaneous preterm birth and perinatal mortality in nulliparous women. A 2024 systematic review 264 confirmed that early aspirin use reduced delivery before 34 weeks (relative risk, 0.84; 95% confidence interval, 0.71–0.99), and a meta-analysis by Roberge et al 265 supported modest reductions in both spontaneous and very preterm birth when aspirin was initiated early. These findings suggest that aspirin’s protective effect may extend to a subset of women with underlying decidual atherotic lesions. Identifying histologic or molecular markers of this lesion could help tailor prophylactic strategies in the future. Fourth, the identification of fibrin deposition within atherotic lesions using a fibrin-specific antibody highlights the prothrombotic component of this lesion. This finding lends additional support to the therapeutic rationale for anti-coagulation (eg, low molecular weight heparin) in pregnancies at high risk for thrombosis-mediated placental insufficiency. 200 , 201 It also invites further exploration of the role of coagulation biomarkers and targeted therapies in improving maternal and fetal outcomes. Finally, these findings reinforce the importance of multidisciplinary care models that integrate obstetrics, cardiology, and primary care in the postpartum follow-up of women with adverse pregnancy outcome. Identification of placental vascular lesions should prompt clinicians to initiate conversations about lifestyle modification, blood pressure and lipid management, and tailored long-term follow-up. Importantly, these approaches represent not only a response to obstetrical syndromes but an opportunity for early prevention of chronic vascular disease in women. In conclusion, this study provides evidence that atherosis is not merely a pregnancy-specific lesion but a manifestation of a broader vascular and immunologic vulnerability. By bridging obstetric pathology with systemic vascular medicine, these findings offer a novel framework for prevention, risk stratification, and targeted therapy in maternal health.

The findings from this study provide a foundation for reclassifying atherosis as a placenta-specific manifestation of a broader family of chronic inflammatory vascular disorders, thereby reframing it as a valuable experimental model for studying immune-driven vascular injury in a temporally well-defined setting. The presence of CD36-expressing macrophages, oxidized lipid accumulation, monomeric CRP expression, NF κ B activation, and fibrin deposition in spiral arteries with atherosis 102 , 156 – 175 , 181 , 188 – 192 , 198 , 199 offers a unique opportunity to investigate the coordinated roles of lipid metabolism, innate immunity, and coagulation in a controlled, time-limited environment such as pregnancy. This model can provide mechanistic insights into how maternal immune responses to fetal antigens—particularly at the level of HLA-C recognition by uterine T and NK cells—initiate or amplify vascular inflammation, 79 , 210 – 215 a process relevant not only to obstetrics but also to transplant rejection and autoimmune vasculopathies. The detection of HLA-DR expression on extravillous trophoblasts in atherosis suggests that inflammatory cues can induce antigen-presenting features in nonimmune cells, 181 , 266 raising questions about the plasticity of trophoblast function and its regulation by local cytokines and damage-associated molecular patterns. Furthermore, the consistency of cellular and molecular features across diverse obstetrical syndromes 124 , 206 – 209 invites future research into shared molecular signatures and epigenetic modifiers that might predispose to atherosis irrespective of clinical phenotype. These findings also suggest that the decidua and placenta may serve as accessible tissues for understanding early-life vascular programming and for identifying molecular biomarkers of maternal vascular susceptibility. The proposed link between endometriosis and native atherosclerosis invites future evaluation of the role of systemic inflammation upon development of atherosclerosis 267 of the spiral arteries in great obstetrical syndromes. Finally, this work highlights the need for multiomics approaches—integrating transcriptomics, lipidomics, and immunophenotyping—to map the molecular architecture of spiral artery atherosclerotic lesions, identify therapeutic targets, and explore their relevance in other contexts of vascular inflammation.

A cross-sectional study was performed in 268 placentas (19–42 weeks of gestation), of which 208 were collected at the Hospital Italiano of Buenos Aires, Argentina, approved by its Institutional Review Board, and 60 were collected at Hutzel Women’s Hospital, the Detroit Medical Center, under protocols approved by the Wayne State University Institutional Review Board. The analyses were completed in 2022 at the California Medical Innovations Institute (CALMI2; San Diego, CA). Placentas with (n=134) and without (n=134) spiral artery atherosis from pregnancies with a gestational age range between 19 and 42 weeks and a fetal weight range between 220 and 3550 g were included. For each placenta with spiral artery atherosis, a placenta without spiral artery atherosis was chosen as a control. Placentas were obtained from patients with small for gestational age (SGA) of unexplained cause (n=64); preeclampsia (n=146); 77 early-onset preeclampsia and SGA; 15 late-onset preeclampsia and SGA; and 54 preeclampsia and normal birthweight (5 early-onset and 49 late-onset), preterm PROM (n=18; 14 infected confirmed by amniotic fluid culture and/or Gram stain and biochemical analysis), preterm labor (n=16; 8 infected confirmed by amniotic fluid culture and/or Gram stain and biochemical analysis), and unexplained fetal death (n=8). Uncomplicated pregnancies (n=16) were defined as those without medical/obstetrical complications and with birthweight adequate for gestational age (>10th percentile) at term (≥37 weeks of gestation); 14 babies were delivered vaginally and 2 by cesarean delivery (none of the mothers or newborns showed clinical signs and symptoms of infection). Preeclampsia was defined as new-onset hypertension after 20 weeks of gestation (systolic blood pressure ≥140 mmHg or diastolic blood pressure ≥90 mmHg on at least 2 occasions, 4 hours to 1 week apart), edema, and proteinuria (≥300 mg in a 24-hour urine collection or one dipstick measurement of ≥2+). Early-onset preeclampsia was defined as preeclampsia that developed before 34 weeks of gestation, whereas late-onset preeclampsia developed at or after 34 weeks of gestation. 80 – 82 Infants were considered to be SGA if they were <10th percentile for weight at their gestational age. Preterm birth was diagnosed before 37 weeks of pregnancy were completed. Idiopathic fetal deaths occurred between 23 and 35 weeks of pregnancy. Birthweight percentiles for cases recruited at Hutzel Women’s Hospital (Detroit, MI) were calculated using nationwide United States Natality data. 83 For cases recruited in Argentina, we used data from the Argentine population that references birthweight according to sex and gestational age. 84 All women provided written informed consent prior to the collection of placenta samples. Comparison of the findings in the spiral arteries were made with coronary arteries. Samples obtained from hearts with native coronary atherosclerosis (n=10), transplant-induced atherosclerosis (n=10), and hearts without atherosclerosis (n=10) were acquired as part of heart transplantation program at Methodist Research Institute, Clarian Health Partners, Indiana University, and Riley Hospitals (Indianapolis, IN). The specimens from patients without atherosclerosis were obtained from patients who had heart failure requiring transplantation. All patients provided informed consent for the examination of the specimens. The terms used in this study follow accepted nomenclature. 85 The placental bed is part of the decidua and adjoining myometrium that underlies the placenta and provides blood supply to the intervillous space. Extravillous trophoblasts are all trophoblasts (fetal-derived cells) located outside of the placental villi, and based on the location and course of the invasion, are categorized as vascular or interstitial. Interstitial extravillous trophoblasts are placental bed trophoblasts that have entered the decidual stroma and can differentiate into 2 subpopulations: those that invade the spiral arteries participating in arterial wall remodeling and those that do not. Extravillous trophoblasts invading spiral artery walls and replacing smooth muscle (interstitial) and endothelium (endovascular) are part of the intra-arterial trophoblasts. Physiologic transformation of the spiral arteries was defined as (1) the complete trophoblastic invasion of the spiral arteries, (2) partial failure of physiologic transformation of the spiral arteries as incomplete trophoblastic invasion of the arterial muscular wall (≤50% failure of trophoblast and fibrinoid replacement of the muscular media of the vessel wall), 48 , 78 , 86 and (3) failure of physiologic transformation of the spiral arteries as the complete absence of trophoblastic replacement of the arterial wall. Placental samples (10×10×5 mm) from the placental septa near the center of the basal plate (5.6 [standard deviation, 0.9] per placenta) of 268 placentas were included in the study. There is general agreement among placental biologists that the arterial and venous openings are irregularly distributed on the decidual basal plate but that most arterial openings are present near the decidual projections and the septa. 87 Samples from 60 placentas obtained immediately after delivery were embedded in Tissue-Tek optimum cutting temperature compound (Miles, Elkhart, IN) and snapfrozen in liquid nitrogen. The rest of the placental samples (n=208) were fixed in 10% formalin and embedded in paraffin. Seven serial sections (a total of 7 slides) from each basal plate sample were used to evaluate the single-antibody, double-antibody, and triple-antibody techniques performed. Sample slides having basal plate spiral arteries defined immunohistochemically were included in the study, and samples without any spiral artery were excluded. A triple-antibody technique 88 – 90 with antibodies to α -smooth muscle actin (1A4, Sigma-Aldrich, St Louis, MO) 88 , 89 , 91 to identify vascular smooth muscle cells, cytokeratin (A575, DAKO, Carpinteria, CA) 88 , 91 , 92 to identify intra-arterial trophoblasts, and cluster of differentiation (CD) 36 (185–1G2, ThermoFisher Scientific, Waltham, MA) to identify the class B scavenger receptor CD36 was used to detect the presence of CD36 in lesions of spiral artery atherosis. Double-antibody techniques using antibodies to α -smooth muscle actin (1A4, Sigma-Aldrich, St Louis, MO), von Willebrand factor (A0082, DAKO, Carpinteria, CA), 88 , 89 , 91 CD68 (EBM 11, DAKO, Carpinteria, CA), 93 – 95 fibrin (mouse monoclonal antibody that specifically recognizes fibrin but does not recognize fibrinogen; mouse antifebrin II ß chain [B ß 15–42] [NYBT2G1, Accurate, Westbury, NY and a-Fib Beta, American Diagnostica, Stamford, CT] reacts with fibrin following the cleavage of fibrinopeptides A and B), 90 , 96 – 99 intercellular adhesion molecule-1 (ICAM-1)/CD54 (LB2, Santa Cruz Biotechnology, Dallas, TX; G5, Santa Cruz Biotechnology, Dallas, TX), 79 , 91 , 92 , 96 – 99 human leukocyte antigen–DR (HLA-DR) (L243, BD Biosciences, San Jose, CA; TAL.1B5, DAKO, Carpinteria, CA; Santa Cruz Biotechnology, Dallas, TX), 91 – 95 , 100 , 101 nuclear factor kappa B (NF κ B) (p65 [F-6], Santa Cruz Biotechnology, Dallas, TX), and C-reactive protein (CRP) (CRP-8, Sigma-Aldrich, St Louis, MO) 102 counterstained with hematoxylin were performed in formalin-fixed, paraffin-embedded tissue samples. Primary and secondary antibodies were always titrated, and all experiments performed using immunofluorescence or light immunohistochemistry always included concomitant negative and positive controls. Immunohistochemical studies performed on samples obtained from the Perinatology Research Branch (Detroit, MI) and the Hospital Italiano (Buenos Aires, Argentina) showed similar reactivity. One of the authors (C.L.) with more than 35 years of experience in immunohistochemistry supervised the quality of the experiments and data obtained. Physiologic transformation of the spiral arteries was defined as cytokeratin-positive, α -actin-negative, and von Willebrand factor-positive (to identify endothelial cells that are also CD31/platelet endothelial cell adhesion molecule-1-reactive 103 ). Failure of physiologic transformation was characterized by cytokeratin-negative, α -actin-positive, and von Willebrand factor-positive spiral arteries. Atherosis was cytokeratin-negative, α -actin-positive, and von Willebrand factor-positive and showed the presence of CD68-positive macrophages in the intimal area of the spiral arteries. Serial sections (5 μ m each) from each placental block were used for light microscopy (hematoxylin and eosin) and for single-antibody, double-antibody, and triple-antibody immunohistochemistry. Immunohistochemistry clearly identified the spiral arteries in all slides from the different serial sections. Spiral arteries (1 or more per placenta) with different characteristics (physiologic transformation, partial or complete absence of physiologic transformation, and atherosis) were included in the analysis and compared as the frequency for each characteristic in each placenta, and the differences among groups were calculated. The deposition of neutral triglycerides and other lipids was examined using Oil Red O to confirm spiral artery atherosis. 104 – 107 Histopathologic and immunohistochemical studies were performed in frozen and in formalin-fixed paraffin-embedded tissues. Similar studies were performed in samples (10×10×5 mm each; n=30) obtained from hearts with native and transplant-induced atherosclerosis and from hearts without atherosclerosis. Individuals examining microscopic sections (C.A.L. and H.L.D.) were masked to the clinical diagnoses. Continuous measures were summarized using mean, standard deviation, median, minimum, and maximum and categorical measures were summarized using frequencies (proportions) across categories. Demographic variables were compared using Fisher’s exact (for discrete measures) and Kruskal-Wallis (for continuous measures) tests. These tests compared values for spiral artery atherosis groups (spiral artery atherosis vs no spiral artery atherosis). A comparison of continuous measures for groups (spiral artery atherosis vs no spiral artery atherosis) was evaluated using the Wilcoxon Rank Sum test. We also used analysis of variance to compare means from multiple groups since inference was unchanged from nonparametric approaches. Finally, a Bonferroni-adjusted level of significance of 0.05/3=0.01667 was used to establish statistical significance for 3 subsequent pairwise comparisons among the groups. Statistical analysis was performed by one of the authors (J.W.H.) using Stata version 18.0.

The primary strengths of this study are (1) the characterization of cells involved in spiral artery atherosis, (2) the quantification of cells participating in the lesions, and (3) the demonstration that spiral artery atherosis has phenotypic characteristics found in native and transplant-induced atherosclerosis. Although the present data show consistent findings in atherosis of the spiral arteries irrespective of the specific obstetrical syndrome, the 134 placentas with the lesion included in this cohort represent a substantial but still relatively small dataset, especially given the considerable complexity observed in placental diseases. Another limitation is the unbalanced patient sample distribution, with substantially more placentas with preeclampsia and SGA and fewer from fetal death and uncomplicated pregnancies. We emphasize that this study focuses on atherosis of the spiral arteries and that future work should examine the prevalence and clinical significance of vascular pathology of the placenta and long-term maternal and infant risk.

Atherosis of the spiral arteries presents evidence of cell activation and inflammation as well as the CD68-positive, lipid-loaded macrophages that express CD36 scavenger receptors. These findings are similar to those of native atherosclerosis. 268 Figure 8 summarizes the cellular and molecular mechanisms leading to atherosis of the spiral arteries. We propose that the term atherosis be replaced with atherosclerosis of the spiral arteries, which occurs in vessels with defective physiologic transformation. 269 – 271 This lesion represents an accelerated atherosclerotic process, which may lead to oxidative stress and hypoxia, 270 driven by the dramatic physiological and pathophysiological changes in pregnancy and its complications. This subclinical vascular pathologic process may explain the observation that women who experience complications of pregnancy are at an increased risk for developing maternal cardiovascular diseases and death later on in life. 271 The demonstration of a more prevalent proinflammatory macrophage phenotype in the decidua basalis with atherosis and an intravascular monocyte source for macrophages in the lesion 272 supports the hypothesis that both tissue-based and circulation-based cellular pathways are involved in the genesis of these lesions. 48 , 141 The improved understanding of the pathologic process leading to atherosis and its similarity to atherosclerosis opens avenues for treatment which may improve pregnancy outcome and long-term health of affected women. ■

A solid body of evidence indicates that mothers who develop preeclampsia, 1 – 6 preterm birth (spontaneous or medically indicated), 7 – 10 fetal growth restriction, 11 – 13 placental abruption, 14 – 17 fetal death, 18 – 21 and gestational diabetes mellitus 22 – 27 are at an increased risk for developing cardiovascular disease and for having a premature death compared to women with a normal pregnancy. Several hypotheses have been proposed to explain the link between the occurrence of these “great obstetrical syndromes” 19 , 28 – 33 and the subsequent risk for cardiovascular disease. One view is that the risk factors for pregnancy complications are similar to those of cardiovascular disease, for example, a high body mass index, hypertension, insulin resistance/hypertriglyceridemia, diabetes mellitus, and renal disease. 20 , 34 , 35 Other investigators have proposed that pregnancy and its complications can worsen preexisting and subclinical cardiovascular dysfunction in women. 20 – 22 , 36 , 37 One possibility is that a preexisting subclinical cardiovascular pathologic process leads to atherosclerosis in pregnancy and that the utero-placental vessels providing blood supply to the placenta are involved and its pathology mediates placental ischemia, 38 – 42 which, in turn, leads to the development of a wide range of obstetrical syndromes. 43 Indeed, the spiral arteries can be affected by a lesion termed atherosis, which is characterized by subendothelial lipid-filled foam cells, fibrinoid necrosis, and leukocyte infiltration. Originally described in preeclampsia, 44 – 48 the lesion has been observed in all major obstetrical syndromes, including preterm labor with intact membranes, 47 , 49 preterm prelabor rupture of the membranes (PROM), 50 fetal growth restriction, 47 , 50 , 51 abruptio placenta, 52 massive perivillous fibrin deposition (also known as maternal floor infarction), 53 – 55 fetal death, 50 gestational diabetes mellitus, 50 , 56 – 58 and antiphospholipid antibody syndrome, 59 – 62 among others. Atherosis of the spiral arteries is thought to decrease blood flow to the placenta, leading to several lesions referred to as maternal vascular lesions of malperfusion, 63 – 69 which have recently been shown to predict vascular impairments a decade after delivery. 70 , 71 We undertook the present study to characterize the morphologic characteristics of atherosis in the decidual segment of the spiral arteries in patients with different obstetrical syndromes and to determine if these lesions resemble those typically observed in native atherosclerosis (eg, coronary arteries of patients with myocardial infarctions) or transplant-induced atherosclerosis. The latter has been demonstrated in the arteries of transplanted organs (heart, lung, intestines, liver, etc.). 72 The strategy used in this study was to study atherotic lesions to determine whether there is evidence of lipid deposition, macrophage infiltration, and cellular activation (of endothelia and macrophages), and of inflammation. 72 – 77 A unique aspect of the biology of pregnancy is that trophoblasts invade the wall and lumen of the spiral arteries and replace the media of the arteries, a phenomenon called physiologic transformation of the spiral arteries. 48 , 78 , 79 Most atherotic lesions are observed in arteries that do not have physiologic transformation. 47 – 50 , 79 Therefore, our study needed to consider trophoblast invasion of the arteries and activation.