Real-Time, Multidisciplinary Approach of Women With Pregnancy-Associated Thrombotic Microangiopathy Where did all the aHUS go?

Written by Cristina Popa MD

Reviewed by Sophia Ambruso MD

A woman is three days postpartum. Her platelet count is falling, creatinine is rising, and the peripheral blood smear shows schistocytes. The laboratory phenotype is unmistakably thrombotic microangiopathy (TMA); the underlying diagnosis is not.  HELLP syndrome, thrombotic thrombocytopenic purpura (TTP), and complement-mediated atypical hemolytic uremic syndrome (aHUS) can initially produce remarkably similar clinical and lab findings, yet their management can be very different. HELLP syndrome should begin to improve after delivery. TTP does not, and untreated disease can be rapidly fatal, while aHUS requires consideration of complement inhibition rather than plasma exchange as definitive therapy. 

Fakhouri et al captured the clinical importance of this distinction in 2020: differentiating these entities “is of paramount importance because TTP and aHUS do not resolve with delivery or transfusion of red blood cells”. Their international workgroup provided a practical diagnostic framework: urgent testing of ADAMTS13 activity (the plasma protease whose severe deficiency defines TTP), angiogenic biomarkers such as the soluble fms-like tyrosine kinase-1/ placental growth factor (SFlt-1/PIGF) ratio, and a stepwise approach to diagnosis and treatment. What an algorithm cannot provide is expert judgement when the phenotype is incomplete, several mechanisms coexist, and decisions must be made before definitive testing becomes available.Marques et al document where that gap opens. A 38-year-old woman at 40+5 weeks presented with hypertension, thrombocytopenia, and transaminases three times normal—textbook HELLP syndrome. She delivered by emergency cesarean. Two days later, worsening cytopenias and new neurologic symptoms forced reconsideration. ADAMTS13 activity returned at 6%. Genetic testing confirmed a homozygous congenital mutation. The diagnosis was ultimately correct, but the clinically important problem was the interval required to reach it. 

France attempted to address it. In October 2023, the National Reference Center for Thrombotic Microangiopathies established a standing multidisciplinary committee for severe or diagnostically uncertain pregnancy-associated TMA. The committee could be activated within 24 hours, seven days a week—ideally before clinical deterioration rather than in response to it. Dengo et al published in KI Reports in September 2026, described the structure and first two years of activity of the French national multidisciplinary consultation committee “TMA and pregnancy." The committee issued advisory opinions on diagnosis and treatment for women with suspected pregnancy-associated TMA, referred by treating physicians. Legal responsibility for care remained with the treating team. Cases were reviewed by videoconference by volunteer experts from eight specialties. Each opinion was issued as a dated written report within 24 hours. The study reports the committee’s operating procedures and activity indicators, the clinical and obstetric characteristics of the women referred, the diagnoses proposed, and the effect of the committee’s input on management.  

From diagnostic framework to clinical infrastructure

Fakhouri et al and Dengo et al address different stages of the same problem. 

Fakhouri et al defined how clinicians should reason through pregnancy-associated TMA: disease categories, diagnostic thresholds, and treatment pathways. Dengo et al examined what happened when those principles were operationalized through centralized, time-limited multidisciplinary review.

The unit of analysis therefore changes from diseases to decisions.

The committee could convene as many as 15 specialists across obstetrics, nephrology, hematology, intensive care, internal medicine, hepatology, immunology, and laboratory medicine by videoconference within 24 hours. Its purpose was to provide rapid, multidisciplinary interpretation of a clinical problem in which treatment frequently cannot wait for diagnostic certainty. 

According to the authors, no comparable real-time national structure dedicated specifically to obstetric TMA had previously been described.

Which patients reached the committee?

Between October 2023 and February 2026, 76 cases were reviewed; 75 ultimately received a diagnostic classification, while one case was excluded.

This was not a typical, unselected obstetric population. Referral to the committee already reflected diagnostic uncertainty and greater clinical severity. Preeclampsia was present in 47% of cases and postpartum hemorrhage (PPH) in 40%, with substantial overlap: 21% of women had both. Thirty-eight percent were primiparous. A key point is timing: 70% of cases were discussed postpartum, at a median of five days after delivery. This means the diagnostic challenge captured by the study was largely one of recognizing disease after delivery, rather than detecting it antenatally. 

But who was referred is only part of the story. The more interesting question raised by the paper is mechanistic: why can several seemingly distinct obstetric complications ultimately converge on a similar pattern of renal microvascular injury?

Endotheliosis: a third mechanism within the TMA differential

The prevailing clinical heuristic treats “pregnancy-associated TMA” as a single diagnostic entity encompassing preeclampsia/HELLP syndrome, atypical hemolytic uremic syndrome, and thrombotic thrombocytopenic purpura. Dengo et al challenge this taxonomy on mechanistic grounds. 

The authors propose that pre-eclampsia/HELLP-associated renal injury constitutes a distinct entity—severe glomerular endotheliosis—capable of producing genuine intrarenal thrombotic microangiopathy, with renal cortical necrosis as its most extreme phenotypic expression. This is presented as biologically separate from both TTP and classic complement-mediated aHUS, rather than as a milder variant of either. 

The underlying histopathology is well established. Farquhar  first described glomerular endotheliosis in the 1960s’; Vassalli et al. documented glomerular fibrin deposition in 1963; and Stevens et al later showed that mild glomerular endotheliosis occurs even in uncomplicated physiologic pregnancy.

The novel contribution is mechanistic: a proposed link between this angiogenic injury and complement dysregulation. Preeclampsia is characterized by excessive placental sequestration of vascular endothelial growth factor (VEGF) and placental growth factor (PIGF), both essential to endothelial integrity. Dengo et al propose that the resulting loss of VEGF signaling additionally suppresses endothelial synthesis of factor H, the principal regulatory inhibitor of the alternative complement pathway. In parallel, free heme released during HELLP-associated intravascular hemolysis is proposed to damage the endothelial glycocalyx and impair Factor H binding at the cell surface. 

By this model, two independent injurious pathways- antiangiogenic endothelial dysfunction and secondary complement amplification- converge on the same microvascular bed. Critically, the presence of complement activation in this setting does not, by itself, establish a diagnosis of primary complement-mediated aHUS: activation may be a downstream consequence of endothelial injury rather than its primary cause.  This distinction yields three mechanistically and therapeutically distinct categories, forming the basis of Dengo et al’s principal diagnostic classification:

• Endotheliosis-associated TMA: HELLP and/ or AKI attributable to preeclampsia, with renal cortical necrosis. Management is predominantly supportive, centered on surveillance rather than disease-specific TMA therapy. 

• aHUS: Classically a severe, postpartum-onset renal TMA following an otherwise unremarkable pregnancy, diagnosed by exclusion of alternative causes. The disease-specific treatment is terminal complement inhibition. 

• TTP: A hematologically dominant TMA caused by severe ADAMTS13 deficiency, either immune-mediated or hereditary. The cornerstone of acute management is therapeutic plasma exchange. 

These three entities share a common laboratory phenotype- thrombocytopenia, microangiopathic hemolysis, and variable organ dysfunction—but diverge sharply in underlying pathophysiology and, consequently, in indicated therapy.

The kidney as the principal target

The endotheliosis hypothesis generates a directly testable prediction: if pregnancy lowers the physiological threshold at which vascular injury precipitates clinically significant renal dysfunction, this effect should be detectable in the pattern and severity of AKI observed across the cohort. 

The data support this prediction. AKI occurred in 62% of the cohort, with a median KDIGO stage of 3- the most severe category on that scale. Among women with AKI, 45% (21 of 76 in the overall cohort) required renal replacement therapy. 

The table below compares three obstetric subgroups. Women with isolated preeclampsia (n=16, no hemorrhage) developed AKI in 62.5%. AKI was far more frequent with PPH: 93.8% with PPH plus preeclampsia (n=16) and 91.7% with PPH alone (n=12), i.e. 26/28 pooled versus 10/16 for isolated preeclampsia (p=0.017). The informative contrast is between the two PPH groups. Women with PPH plus preeclampsia lost less blood (median 850 mL vs 1500 mL; p=0.02) yet had virtually the same AKI rate. Preeclampsia combined with a smaller hemorrhage thus produced as much AKI as a larger hemorrhage alone, suggesting that preeclampsia lowers the blood loss needed to injure the kidney. Renal cortical necrosis did not follow this pattern: it occurred in 5/12 women with PPH alone, versus 1/16 with PPH plus preeclampsia and ⅙ with isolated preeclampsia (p=0.128).

The authors interpret this pattern as evidence that pregnancy—and particularly the superimposed endothelial injury of preeclampsia—lowers the hemorrhagic threshold required to precipitate severe AKI, such that the degree of renal injury observed is disproportionate to the measured volume of the blood loss. This interpretation is mechanistically coherent, but the underlying subgroup sizes are small, and the specific thresholds implied by these comparisons should be regarded as hypothesis-generating rather than as established physiological boundaries.

Tranexamic acid: contributor or marker of severe disease?

The paper also revisits a more specific hypothesis: whether tranexamic acid might contribute to renal injury. 

Tranexamic acid is routinely used for postpartum hemorrhage and is cleared by the kidney. Once AKI develops, clearance falls and exposure increases. Dengo et al propose a potential feedback loop: AKI prolongs tranexamic acid exposure, sustained exposure may promote intrarenal thrombosis, and this could further worsen kidney injury. 

This hypothesis builds on earlier work from the same French network. Frimat et al described 18 women with postpartum renal cortical necrosis: all had severe hemorrhage and required dialysis. Longer tranexamic acid exposure was associated with poorer renal recovery. But this remains an association, not evidence of causality. Importantly,the signal has been described with high cumulative obstetric exposure, not with the standard 2g dose used in the WOMAN trial or in trauma populations.

So, tranexamic acid should not be presented as an established cause of renal cortical necrosis. A more cautious interpretation is that prolonged or excessive exposure during evolving AKI could be a modifiable contributor.

TTP: uncommon, but a different disease mechanism

If endotheliosis-associated TMA accounts for the majority of renal disease in this cohort, TTP represents a smaller group governed by an entirely distinct pathophysiological process, a quantitative deficiency of a specific plasma protease, rather than an angiogenic or complement-mediated mechanism.

TTP was diagnosed in 13 of 76 women (17%): five had immune-mediated disease, confirmed by detectable anti-ADAMTS13 antibodies; two had severe ADAMTS13 deficiency without detectable antibodies and were classified as idiopathic; and six were considered suggestive of hereditary TTP, primarily on the basis of presentation during a first pregnancy in the absence of antibodies. 

The proportion attributed to hereditary disease- 6 of 13, or 46%- substantially exceeds prior estimates. The French national TTP registry reported hereditary disease in approximately one-quarter of pregnancy-associated TTP cases. A biologically plausible explanation exists: hereditary TTP characteristically unmasks during first pregnancy, and 38% of this cohort was primiparous. Nonetheless, a sample of thirteen cases cannot support a revised population-level prevalence estimate, and the authors appropriately refrain from making this claim. 

A terminological inconsistency also warrants comment. The manuscript text describes five immune-mediated, two idiopathic, and six hereditary-suspected cases (13 total); Table 1 instead groups these as “Idiopathic TTP” (n=7) and “suspicion of congenital TTP” (N=6); and Figure 2 labels the same subgroups “iTTP” (N=7) and “cTTP” (N=6). The aggregate totals are internally consistent across all three representations, but the category labels are not- a distinction that matters when citing a specific subgroup figure rather than the overall total.

The diagnosis that disappears: where was aHUS?

Among 76 referrals for severe pregnancy-associated TMA, no case was classified as confirmed spontaneous inaugural aHUS. 

First, aHUS lacks a definitive diagnostic test and remains a diagnosis of exclusion. Complement abnormalities were identified in only 46% of pregnancy-associated aHUS cases in Bruel et al, and 41% in Huerta et al. Negative genetic testing therefore does not exclude aHUS.

Second, complement genetics rarely returned in time to guide acute management. In women with preserved ASAMTS13, the committee often withheld eculizumab for 24-28 hours while assessing spontaneous improvement after delivery. 

This strategy may reduce unnecessary eculizumab exposure, but classification partly depends on the diagnostic threshold applied. The absence of confirmed aHUS could reflect a genuinely low incidence, avoidance of overdiagnosis, or classification of uncertain cases as endotheliosis when timely genetic evidence was unavailable.

The defensible conclusion is narrower: with the evidence available in real time, the committee did not reach sufficient certainty to diagnose spontaneous inaugural aHUS. This does not establish that aHUS cannot occur after preeclampsia, HELLP or PPH.

Did the committee change management?

Management changed in 20/76 cases (26%): treatment was escalated in five and de-escalated in 15. Clinically, the committee’s main measurable effect was to identify those in whom active treatment could be stopped.

Recombinant ADAMTS13 was given to five women: four with suspected hereditary TTP flares and one with refractory immune TTP, reportedly the first documented use in this setting. Eculizumab was newly initiated in three cases: two with renal cortical necrosis after PPH on mechanistic grounds rather than confirmed aHIS, and one with severe hypertension and critical illness. It was continued prophylactically in one woman with a known pathogenic CD46 variant. Therapeutic termination of pregnancy was recommended once, pre-viability, for maternal indications. 

Outcomes There was one maternal death: a woman with sickle cell disease (HbSC), already in multiorgan failure at referral, who died from a vaso-occlusive crisis rather than TMA management failure. There were ten intrauterine fetal deaths and one early neonatal death (Table 1). These outcomes emphasize that pregnancy-associated TMA occurs within complex multisystem disease; prognosis is not determined by the microangiopathy alone.

How does this study fit with the existing literature?

The work by Dengo et al did not replace the framework proposed by Fakhouri et al; it tests how well that framework works in clinical practice. 

The diagnostic approach is similar: urgent ADAMTS13 testing, angiogenic biomarkers, exclusion of alternative diagnoses, and close assessments of the postpartum trajectory. 

Fakhouri et al proposed sFlt-1/PIGF thresholds: >85 before 34 weeks or >110 thereafter supports preeclampsia/HELLP, while >38 argues against it. Dengo et al highlight a practical limitation: the ratio was rarely measured early enough to influence committee decisions. A useful biomarker has limited value if it is not available when diagnosis and treatment decisions are being made. 

Kumar et al provide a complementary bedside framework, distinguishing TTP, HELLP, complement-mediated TMA (CM-TMA/aHUS), and acute fatty liver of pregnancy (AFLP) by timing, renal involvement, neurologic manifestations, ADAMTS13 activity, and laboratory pattern. Their algorithm incorporates the PLASMIC score to estimate the probability of severe ADAMTS12 deficiency while awaiting results.

What holds up?

Dengo et al’s strongest contribution may be organizational rather than diagnostic. A disease-specific, 24/7 national consultation network provides real-time access to TMA expertise. Importantly, treating physicians retain responsibility, and deviations from committee recommendations must be documented. The management signal is also credible because it was driven mainly by de-escalation, rather than increased treatment. 

The network has broad geographic reach, including Reunion, Mayotte, Martinique, New Caledonia, and Belgium (supplementary Fig S1), suggesting that centralized expertise can support clinicians well beyond tertiary referral centers. 

What remains uncertain?

The central limitation is that this is an audit by expert clinical judgement, not a diagnostic validation against a gold standard. Placental histology was usually unavailable at committee review, while complement genetics were delayed or absent. "Endotheliosis-associated TMA” therefore remains a clinical classification rather than a histopathologically confirmed one. 

Selection bias is inherent: these 76 women were referred because their cases were sufficiently severe or diagnostically uncertain to warrant national review. The cohort cannot estimate incidence in the general obstetric population. Subgroups are also small. Table 4 compares renal severity across only 44 patients (16, 12, and 16 per group). The p= 0.017 difference in AKI frequency should therefore be considered exploratory, particularly without reported adjustment for the multiple comparisons across tables 3 and 4. 

The Figure 3 radar plot is descriptive, not inferential. It uses normalized medians from small groups (TTP N=13, RCN N=9, and endotheliosis N=33), without showing dispersion or overlap (Supplementary methods). It illustrates the threshold in the absence of timely generic testing. External validity is also uncertain: this is one French/Belgian network operating with specific referral pathways, expertise, and drug-access structures. 

Does zero aHUS fit with other cohorts?

Zero confirmed aHUS out of 76 referrals sounds reassuring, but is it? To check, we can compare against studies built specifically to find aHUS in pregnancy. 

These aren’t the same populations as Dengo et al- they enrolled women already suspected or confirmed to have aHUS, not every pregnancy-TMA referral that comes through the door. However, they prove the point that matters: aHUS in pregnancy is real and recurring. Seen against that backdrop, a flat zero over two years is a result that needs an explanation, not proof that postpartum aHUS doesn't happen.

Renal cortical necrosis, endotheliosis, complement, or both?

Postpartum renal cortical necrosis illustrates how the same lesion can be interpreted differently depending on the diagnostic framework.

These cohorts use different populations and diagnostic thresholds, so direct comparison is limited. Collectively, however, they argue against treating postpartum renal cortical necrosis as specific for either endotheliosis or aHUS. The renal lesion may be shared across different upstream mechanisms. 

The closest parallel: pregnancy-onset TTP

Beranger et al. from the French Reference Center Network classified pregnancy-onset TTP into immune, unclassified, and congenital disease. Most immune and unclassified cases showed an “open” ADAMTS13 conformation during the acute episode. Given the overlapping referral network and authorship, some of Dengo et al’s 13 TTP cases might also be represented in this cohort. The two datasets, therefore, should not automatically be considered independent. 

Why build a committee at all?

Pregnancy is an important trigger of secondary TMA: Bayer et al attributed approximately 35% of secondary TMA cases in their French cohort to pregnancy. The challenge is not recognizing thrombocytopenia, hemolysis, and organ injury. It is identifying the underlying mechanism quickly enough to select the appropriate treatment. This is where Dengo et al add something different: not a new biomarker or therapy, but an organizational intervention that brings specialist expertise closer to the point of diagnostic uncertainty. 

What should change at the bedside? 

The clinical implications are relatively focused:

• Do not treat pregnancy-associated TMA as a single diagnosis. It is a shared phenotype arising from different mechanisms that require different treatments. 

• Obtain ADAMTS13 urgently, ideally before plasma transfusion or exchange affects interpretation. Severe deficiency strongly supports TTP.

• Measure sFlt-PIGF early when preeclampsia versus another TMA is uncertain. Timing determines its clinical usefulness.

• Consider imaging for renal cortical necrosis in severe oliguric/anuric AKI after PHH; renal cortical necrosis can explain profound kidney injury without establishing aHUS.

• Use the postpartum trajectory diagnostically. Preeclampsia/HELLP should begin to improve after delivery; persistent or worsening thrombocytopenia, hemolysis, neurologic abnormalities, or renal failure over 48-72 hours should prompt reassessment for TTP, aHUS, or another process. 

Where expert review is available, its value may lie as much in avoiding unnecessary treatment as in identifying when treatment should begin. Dengo et al do not solve the diagnostic uncertainty surrounding pregnancy- associated TMA. Their contribution is more pragmatic: when several high-stakes diseases converge on the same clinical phenotype, rapid access to disease-specific expertise can itself function as an acute-care intervention.