Pregnancy and the postpartum period, usually spanning up to 3 months post-delivery, are well-established significant risk factors for TMA (4, 5). TMA can manifest in diverse conditions, including exclusive pregnancy-specific patterns referred to as “pregnancy-specific TMAs” like preeclampsia (PE) and HELLP syndrome, and patterns associated with specific diseases or conditions, collectively termed “pregnancy-associated TMAs,” the incidence of which rises within the pregnancy context (4, 5). Pregnancy-associated TMAs encompass TTP, CM-HUS, autoimmune disorders such as systemic lupus erythematosus and antiphospholipid syndrome (APS), notably the catastrophic APS (CAPS). Despite suggestions, it remains unconfirmed whether pregnant women with severe diffuse scleroderma have a higher incidence of renal crisis, a TMA-like disorder, compared to non-pregnant women (22, 23).

Moreover, both PE and HELLP syndrome may coincide or emerge as a result of pregnancy-associated TMAs, adding to the intricacy of diagnosis and management (24–27). Further complicating matters, TMA-like patterns may also be noted in case of obstetrical disseminated intravascular coagulation (DIC). This severe event, characterized by extensive thrombus formation (not limited to small vessels), is associated with various pregnancy complications, including acute peripartum hemorrhage, placental abruption, sepsis, PE/HELLP syndrome, retained stillbirth, amniotic fluid embolism, and acute fatty liver of pregnancy (28–30). Hence, any suspicion of TMA should prompt repeated analysis of platelet count, coagulation parameters (such as prothrombin time and fibrinogen and fibrin split products/D-dimer concentrations), while considering the pregnancy-associated intricacies of these metrics. This consideration has prompted certain researchers to advocate the use of pregnancy-specific diagnostic scoring systems for obstetrical DIC including fibrinogen concentrations, the PT difference and platelet count (31).

The biological criteria for TMA in pregnancy closely resemble those in the general population, except for the lower threshold of thrombocytopenia. Indeed, in the context of a healthy pregnancy, platelet count moderately decreases, gradually starting from the second trimester, reaching a nadir in the third trimester, while still remaining above 80 × 10⁹/L. This benign gestational thrombocytopenia has been related to hemodilution due to increased plasma volume, augmented platelet consumption in the utero-placental unit and potential heightened platelet clearance, and does not affect pregnancy outcome (32, 33). Diagnosis of pregnancy-related TMA is thus based on the following parameters: platelet count <100 × 10⁹/L, hemoglobin level < 10 g/dL, serum lactate dehydrogenase >1.5 times the upper limit of normal, undetectable serum haptoglobin, negative direct erythrocyte antiglobulin test, and presence of schistocytes on blood smear (5).

The detection of MAHA in a pregnant or newly delivered patient constitutes a diagnostic and therapeutic emergency. In this context, algorithms for the management of pregnancy-related TMA are regularly updated, considering the challenge of urgent differentiation (5, 21, 34). Table 1 provides a concise overview of pregnancy-specific and -associated TMAs, focusing on the distinguishing features of different forms of TMA.

The recommended initial workup for pregnant or postpartum women presenting with TMA is detailed in recent reviews (5, 21). This workup includes coagulation tests, quantification of ADAMTS-13 activity, investigation of autoimmune disorders (lupus anticoagulant, anticardiolipin antibodies, Beta-2-glycoprotein antibodies, antinuclear antibody) or metabolite deficiencies (vitamin B9, B12), exploration of complement levels (C3, C4, factor H, factor H autoantibody, factor I, factor B, and MCP/CD46 expression by flow cytometry), and, when available, measurement of sFLT1 and PlGF concentrations.

The most specific indicator is ADAMTS13 activity. An activity of less than 10%, definitively confirms TTP, while levels above 20% usually rule out TTP. Hence, it should be urgently assessed in all cases of pregnancy-MAHA. In contrast, there is currently a lack of specific biomarkers for distinguishing between CM-HUS, PE, or the HELLP syndrome.

TMA lesions are traditionally defined as being confined to microvessels, including arterioles, capillaries, small arteries, with a diameter cutoff of <200 μm (3). When assessing capillary lesions under light microscopy, endothelial swelling is the most prevalent, followed by subendothelial widening and capillary wall thickening. Among arteriolar lesions, fibrinoid necrosis, mucoid intimal thickening and luminal thrombi are reported (Figures 1A,B). Intraluminal thrombi usually comprise fibrins and platelets, although their composition can vary according to the pattern of TMA (Table 2) (28, 44, 56, 58, 99–102). While endothelial changes are consistently evident, thrombi may not always be visualized, especially in instances of early or mild disease. Downstream from the thrombi, necrotic lesions can be observed, resulting from the ischemic process. Consequently, acute tubular necrosis lesions are frequently reported in association with renal TMA lesions in cases of HUS (103), and liver enzymes are elevated in women with HELLP syndrome due to microangiopathy with sinusoidal obstruction, leading to hepatocyte necrosis (44, 100). Immunofluorescence typically reveals fibrinogen deposits localized to thrombi and also, in mesangium, in glomerular capillary walls and in afferent arterioles. Nonspecific entrapment of IgM and complement components is usually observed in glomeruli and affected vessels in around 50% of specimens, primarily in glomerular capillary walls, mesangium, and afferent arterioles. Staining for IgG and IgA consistently yields negative results (104, 105).

None of these lesions are specific, and the histological diagnosis of TMA remains complex, mainly due to the absence of established criteria. Furthermore, the potential overlap between acute (defined as occurring within 2 months from the initial presentation) and chronic features further complicates the histological diagnosis. Approaches to standardized analysis are beginning to be proposed, as evidenced recently in cases of TMA in renal transplant recipients, where they incorporate histological, clinical, and biological data. In this specific context, fibrin thrombi in arterioles and/or glomerular capillaries (100%), arteriolar intimal edema/mucoid changes (95%), and mesangiolysis (82%) demonstrated the highest level of consensus among the panelists (106).

The literature primarily provides renal pathology data for pregnancy-related TMAs, possibly due to the frequency of postpartum AKI and the feasibility of renal biopsies. Although there are some histological descriptions of organ involvement beyond the kidney (detailed in Table 2), most of these descriptions originate from post-mortem examinations (99–102), possibly introducing bias by representing the most severe clinical cases, which could overestimate the prevalence of diffuse TMA lesions.

The kidney serves as the principal target of TMA in conditions such as PE, HUS, and CAPS, suggesting a specific vulnerability of this organ to TMA lesion development. While various hypotheses and emerging data, notably concerning an elevated renal susceptibility to complement activation, have been proposed, our understanding of this renal vulnerability remains constrained (107, 108). As mentioned previously, renal TMA is less common in HELLP syndrome, with predominant tubular lesions reported in series associated with AKI (66). The presence of fibrin deposition in hepatic sinusoids with periportal hemorrhage and hepatocyte necrosis is well-documented in post-mortem examinations of individuals with PE associated with HELLP syndrome. In contrast, these lesions are observed in liver needle biopsies in only 25% of eclampsia patients and are exceptionally rare in PE patients (44). These findings support the hypothesis of a clinical-pathological correlation between the extent of morphological changes and the severity of presentation, as well as adverse outcomes.

There is no documented correlation between the pathological features observed in renal biopsies and the underlying etiology of TMA. In non-gravid contexts, it has been suggested that the presence of isolated intimal mucoid edema without glomerular fibrin thrombi reduces the likelihood of CM-TMA in the setting of TMA associated with hypertensive emergencies (109); however, this remains a subject of controversy (103, 110). Additionally, some data suggest that renal arteriolar involvement is more widespread in aHUS and TTP, while glomerular involvement appears to be more prevalent in STEC-HUS (105). Nonetheless, it is important to note that the coexistence of these lesions is relatively common and has not been definitively linked to a specific underlying pathophysiological mechanism.

In gravid contexts, PE results in notable glomerular changes, termed “endotheliosis”, characterized by glomerular enlargement and reduced capillary blood flow due to endothelial swelling. This substantial reduction in endothelial cell fenestrations has been also observed in others capillary networks, including, choroid plexus, and hepatic sinusoids (104). This phenomenon is closely associated with sFlt-1, which reduces circulation of vascular endothelial growth factor (VEGF), impacting placentation and endothelial cell function, especially in fenestrated endothelium (40, 111). However, this specific lesion is not unique to PE and has been observed in various other conditions, including placental abruption, parvovirus B19 infection, COVID-19, POEMS syndrome, and even uncomplicated pregnancies (112). The presence of arteriolar or arterial lesions should prompt consideration of underlying conditions, like hypertensive vasculopathy, or alternative diagnoses, such as CM-HUS (105). Post partum biopsies can provide information about healing process or progression. For example, in PE, the endothelial swelling, IgM and fibrin deposits typically resolve within 2 weeks of delivery. The subendothelial irregularities in the glomerular basement membrane may persist for months (105). Renal biopsies can also help in lupus patients due to potential clinical similarities between LN flares and PE (113, 114). In APL cases, the presence of acute TMA in glomerular and arteriolar lesions alongside chronic vascular changes aids in diagnosis (115).

In 2016, the Amsterdam Placental Workshop Group provided a consensus statement on placental lesion sampling and definitions (116). Presently, the accepted term is “maternal vascular malperfusion” (MVM), which encompasses a range of placental pathologic findings, including macroscopic features (fetal and placenta weights, infarcts, retroplacental hemorrhage) and microscopic findings (maternal decidual arteriopathy, distal villous hypoplasia, and accelerated villous maturation), Figures 2A–D. Maternal decidual arteriopathy comprises acute atherosis of decidual arteries, mural hypertrophy of arterioles within the placental membranes, abnormal persistence of mural smooth muscle within arteries at the placental basal plate, and the persistence of intramural endovascular trophoblast. The specific causes of MVM remain unclear, with potential factors including immunological and inflammatory elements, as well as pre-existing maternal susceptibility to microvascular dysfunction (117). MVM is described in PE and HELLP syndrome, and it also appears in instances of second-trimester spontaneous abortion, fetal demise, abruptio placentae, small for gestational age, preterm labor, preterm prelabor rupture of membranes, and maternal autoimmune diseases (117, 118).

Most of the placental histological data is centered around PE. Distinct lesions are predominant according the timing of PE during gestation, suggesting potential differences in underlying pathophysiological factors. Early-onset PE is characterized by more severe maternal MVM lesions, both macroscopic and microscopic (119). A positive correlation has been reported between the presence of decidual vasculopathy in vessels and adverse maternal and fetal outcomes (i.e., higher diastolic blood pressure, increased urinary protein levels, shorter gestational age, and lower birth weight) (112). In contrast, late-onset PE placentas show normal or excessive growth, milder microscopic features like accelerated villous maturation and hypertrophic arteriopathy (119), and significantly fewer cases of failed spiral artery remodeling, acute decidual arteriopathy, and infarcts compared to early-onset PE placentas (120). Mural hypertrophy of placental arterioles was linked to the development of stage 2 hypertension during follow-up (121), emphasizing the impact of preexisting cardiovascular risk factors (e.g., diabetes, chronic hypertension, obesity) in late-onset PE pregnancies. A limited number of studies have compared placental histopathological features in severe PE cases with and without HELLP syndrome (122–124). Notably, Weiner et al., in their analysis of 287 placentas, reported that vascular and villous lesions indicative of maternal malperfusion were independently linked to HELLP syndrome (124).

Information regarding other pregnancy-related TMAs is limited. In a cohort of 91 pregnancies involving 47 TTP-patients, Scully et al. examined placental histology from 15 deliveries. Interestingly, untreated congenital cases exhibited distal villous hypoplasia, widespread placental ischemia with infarcts of various ages, and acute atherosis, while treated congenital cases showed normal placental features (27). Similar results were reported in a recent study involving 28 placentas from congenital TTP cases (125), emphasizing also the direct role of ADAMTS-13 deficiency in placental findings. Moreover, some research groups have examined placental deposits of complement activation products, which have been identified in conditions such as PE, HELLP syndrome, lupus, and APS, without conclusive discriminatory features among these pathologies at present (126–128).

The data is still quite limited, and a more systematic analysis of placentas from patients being treated for TMAs could provide valuable insights, particularly when combined with the recently suggested use of electron microscopy to study placental pathologies during pregnancy (129).

In contrast to large vessels with their three tunics, capillaries, the primary site for TMA with arterioles, consist of a thin endothelium, basement membrane, subendothelial connective tissue, and a few pericytes, with diameters ranging from 4 to 10 μm and an endothelial wall of about 0.5 μm. They operate at low intravascular pressures (0–25 mmHg, up to 50 mmHg in kidney glomeruli) and slow flow rates (<1 mm/s), facilitating efficient plasma and tissue exchanges (131).

The phenotype of microvascular endothelial cells varies from one organ to another to fulfill specific functions and respond to the different needs of various tissues (132–134). As early as 2003, the spatial heterogeneity among endothelial cells was supported by distinct transcriptomic profiles (135) and differential expression of protein surface markers (136, 137). Most recently, the advent of single-cell RNA sequencing technologies has revolutionized our understanding of endothelial diversity, revealing significant heterogeneity within the confines of a single organ or a blood vessel type, in both animal models and humans (138–140). The endothelium is a dynamic tissue with plasticity properties, making it also subject to temporal heterogeneity (141). Endothelial cells react and respond to various biomechanical and biochemical stimuli that differ between organs, such as hormones, as discussed below, contributing to gender-driven endothelial heterogeneity. Additionally, site-specific epigenetic modifications play a significant role in generating endothelial heterogeneity (142, 143).

Significant sex-based differences in endothelial cells have been documented (144–146), emphasizing the heightened endothelial protection in women of childbearing age, which is associated with a lower cardiovascular risk compared to males within the corresponding age group or postmenopausal women. The studies have primarily focused on exploring the involvement and interactions of circulating sex steroid hormones, their receptors, and sex steroid-independent mechanisms. Notably, estrogen receptor E2-mediated vascular benefits encompass inhibition of lipoprotein oxidation, reduction of atherosclerotic lesions, modulation of blood coagulation, inhibition of collagen accumulation in vessels, and adjustment of vascular tone (144, 147). Studies conducted in animal models involving ovariectomy and hormone supplementation therapy have contributed to the understanding of sex hormone-dependent effects, particularly those of estrogen, on endothelial function, including changes in endothelium-dependent dilation, and the respective roles of nitric oxide (NO) and endothelium-derived hyperpolarizing factor (EDHF). These investigations have predominantly concentrated on medium and large arterial vessels (e.g., aorta, carotid, mesenteric arteries) (148–150). For instance, small arteries from omental tissue in premenopausal women demonstrate increased sensitivity to bradykinin (151); premenopausal women have shown increased levels of progenitor ECs in comparison to men, alongside enhanced NO production essential for vascular homeostasis (152); the profile of endothelial-derived factors in childbearing women is associated with a less pronounced pro-oxidant environment, characterized by reduced levels of vascular reactive oxygen species and other oxidative stress markers (153, 154). Additionally, levels of endothelin-1, which play a role in cardiovascular and renal pathology, are generally higher in men than in women (155); plasma levels of VEGF have been found to be elevated in males compared to females and have shown a positive association with the advancement of atherosclerosis (156).

Collectively, these studies underscore the heightened endothelial protection observed in women of childbearing age, elucidating their diminished susceptibility to cardiovascular events. They also illustrate the importance of incorporating both genders into experimental protocols, also including ex vivo and in vitro studies. Nonetheless, questions and controversies debates, especially concerning the strict favorable vascular impacts of estrogen and its E2 receptor (157).

Furthermore, as previously emphasized, research on vascular sex differences has predominantly focused on macrovessels, while data on the microcirculation are limited (158). Some studies have demonstrated sex differences in the permeability or inflammation-induced disruption of the blood–brain barrier (159, 160), and circulating levels of products derived from the glycocalyx (e.g., syndecan-1, heparan sulfate, and hyaluronan) (161, 162). More answers may potentially come from single-cell RNA-seq analyses. Already, a single-cell RNA-seq analysis of mouse endothelial cells has further confirmed the presence of sex-specific differences in endothelial cells across various mouse tissues. Paik et al. found that markers of tissue-specific endothelial cells, enriched pathways, and endothelial subpopulations differed between male and female mice, particularly in the brain, heart, and lung. The Lars2 gene, encoding mitochondrial leucyl-tRNA synthetase, was notably more enriched in male than in female endothelial cells (163). While this comprehensive approach has yet to be undertaken in pregnant individuals or TMA patients, specific endothelial attributes associated with pregnancy have been recognized and may contribute to the heightened thrombogenicity of the gravid endothelium.

Vascular adaptations during pregnancy, even in healthy pregnancies, are associated with significant modifications in the characteristics and function of endothelial cells. For instance, glomerular capillary endotheliosis, originally associated to PE, has been detected in patients with gestational hypertension without proteinuria (> 1+ in 8/8, and > 2+ in 4/8 cases) and, as cited above, in women with healthy pregnancies (>1+ in 5/12, and > 2+ in 1/5 cases) based on renal biopsies conducted between 27 and 39 weeks of gestation (112). These findings suggest a continuum of endothelial changes that range from normal pregnancy to the development of PE. Multiple vascular changes take place during pregnancy, potentially heightening the vulnerability to thrombosis, and our primary focus in this context is on the specific factors involved in the pathophysiology of TMAs (Figure 3).

Therefore, the “gravid endothelium” is inherently more prone to activation, contributing to the favorable progression of pregnancy. The transition to a pathological state with the development of TMA lesions requires additional factors, as described in the concept of a “multiple-hits disease” (18, 200). These additional factors can be either genetic and/or acquired, and act to heighten susceptibility and/or induce endothelial stress. Their cumulative impact (with varying significance depending on the specific factor) will ultimately lead to an exceeding of the regulatory capacities involved in endothelial homeostasis.

Genome-wide association studies have not been conducted in women experiencing pregnancy-related TMAs. Such investigations have primarily been carried out in the context of PE, often without a clear differentiation of patient phenotypes. While more than fifty candidate genes have been associated with PE, no universally accepted susceptibility genes have emerged (41, 201, 202), possibly due to the polygenic nature of PE. In a study of 244,564 related women, Cnattingius et al. determined that PE has a heritability of over 50%, with 35% linked to maternal genetics, 20% to fetal genetics, and 13% to a ‘couple’ effect (203). Mutations and/or variants have been reported in the context of PE or HELLP syndrome, related to genes involved in angiogenesis, blood pressure regulation, coagulation, or complement activation (40, 202, 204, 205), underscoring the close pathophysiological connections with TMA. For instance, mutations or variations in Factor H, C3, or CD46 have been identified in cohorts of women with PE (206, 207). These findings indicate that a multifaceted interplay of susceptibility factors is likely, potentially resulting in distinct patterns of preeclampsia. Supporting this concept, Schuster et al. documented groups of patients exhibiting common gene and protein networks linked to severe preeclampsia (208).

The extensive range of environmental factors capable of increasing endothelial susceptibility cannot all be comprehensively reviewed here. In an illustrative example, placental hypoxia is worth mentioning to illustrate the interplay of factors involved in TMA. The placenta is a physiological immunological niche, especially under the influence of physiological hypoxia that contributes to immunological regulation (209). However, placental hypoxia has others consequences as the generation of oxidative stress and trophoblastic secretion of sFLT-1, promoting antiangiogenic activities and endothelial injury (210), alteration in placental metabolism favoring interruption of the glycocalyx susceptible to contribute to localized complement activation (199, 211, 212). Thus, a pathological placental hypoxia, as observed in the case of PE, represents a risk situation for TMA.