Hemolytic Uremic Syndrome

Hemolytic uremic syndrome (HUS) encompasses a group of diseases arising from endothelial cell injury and leading to thrombocytopenia, microangiopathic anemia, and AKI. HUS includes several subtypes, reflecting distinct mechanisms of endothelial cell injury.¹ Over the past two decades, the two main breakthroughs in the field of HUS have derived from complement genetics and the availability of complement inhibitors.² These breakthroughs have completely changed the landscape of one peculiar form of HUS, atypical complement-mediated HUS. Their effect on other forms of HUS remains, to date, more limited.

The article by Werion et al.³ in this issue of CJASN is a timely reminder of many remaining unresolved questions in the field of HUS. The authors retrospectively analyzed the causes and outcomes of 336 consecutive adult patients with a first episode of thrombotic microangiopathy seen in two reference centers in Belgium. The main type of thrombotic microangiopathy was secondary thrombotic microangiopathy (associated with various conditions/diseases), accounting for 56% of cases. Secondary thrombotic microangiopathy represented a heterogeneous group of distinct triggers of endothelial cell injury, ranging from hematopoietic stem-cell transplantation (30%) and solid-organ transplantation (23%) to malignant hypertension (13%), drugs (11%), and malignancy (6%). By contrast, atypical complement-mediated HUS represented only 15% of cases.

The authors focused their analysis on secondary thrombotic microangiopathy. Their results highlight the high morbidity and mortality of this form of thrombotic microangiopathy, with a dramatically higher risk of death or kidney failure compared with atypical HUS, even after the exclusion of the sickest patients (those with hematopoietic stem-cell transplantation and cancer). Nevertheless, the inclusion of hematopoietic stem-cell transplantation and patients with cancer probably explains the high mortality rate (44%) not seen in two previous series.⁴ The authors also report a low prevalence (3%) of pathogenic or likely pathogenic complement gene variants, similar to the one reported in healthy individuals¹ (as opposed to 49% prevalence of such variants in patients with atypical HUS). Interestingly, no pathogenic complement gene variant was detected in a substantial cohort of patients undergoing hematopoietic stem-cell transplantation—a finding that challenges the assumption that constitutional complement dysregulation plays a predominant role in this type of thrombotic microangiopathy.⁵

The results by Werion et al. also illustrate the persistent uncertainties regarding the optimal treatment of secondary thrombotic microangiopathy, in sharp contrast to atypical HUS for which C5 blockade has become the gold-standard treatment, at least in settings where it is available. One third of patients with secondary thrombotic microangiopathy underwent plasma exchanges, and a small minority (4%) received eculizumab, when there is no definite proof of the efficacy of the former or the latter in this type of thrombotic microangiopathy.³

The study by Werion et al. is in keeping with previously published data regarding secondary thrombotic microangiopathy/HUS.^4,6,7 The design of the study and its results illustrate the difficulties in defining secondary thrombotic microangiopathy/HUS. For instance, Werion et al. identified secondary thrombotic microangiopathy based (1) on the exclusion of ADAMTS13 deficiency and of an infection with shigatoxin-producing enterobacteriae and (2) on the presence of “any other disease or condition associated with [thrombotic microangiopathy] TMA.” They included pregnancy and malignant hypertension-related thrombotic microangiopathy as well as postkidney transplantation de novo HUS within secondary thrombotic microangiopathy. However, several studies have previously indicated that pregnancy-associated HUS is a form of complement-mediated atypical HUS.⁸ Similarly, it is acknowledged that complement-mediated atypical HUS may present with malignant hypertension in more than half of the cases.^9,10 Furthermore, de novo HUS after kidney transplantation may simply be a complement-mediated HUS undiagnosed before transplantation.

Twenty years ago, HUS was mainly a therapeutic issue. With the advent of complement genetics and complement inhibitors, it has become a diagnostic and semantic issue. Indeed, there is no universally accepted definition of secondary thrombotic microangiopathy/HUS, and its spectrum varies from one country to another and even from one institution to another within the same country.

We fully concur with Werion et al. that the current terminology of thrombotic microangiopathy, in general, and HUS, in particular (typical, atypical, and secondary), is far from being optimal and even potentially misleading. For instance, does a complement-mediated HUS triggered by an infection qualify as a secondary HUS? Similarly, should a HUS linked to a constitutional metabolic defect within the endothelial cell, such as in cobalamin C-deficiency, be included in the entity of secondary thrombotic microangiopathy/HUS?

We also agree with the authors that a terminology of thrombotic microangiopathy and HUS on the basis of the underlying mechanism is needed because the ultimate goal of any medical classification is to help clinicians choose a specific efficacious treatment. Nevertheless, such classification is not easy to implement.

First, the definition of HUS itself does not take into account the heterogeneous presentation of HUS, including cases with normal platelet count, and most particularly pathological features of kidney thrombotic microangiopathy. The diagnosis of HUS remains challenging in some situations, especially in the setting of hematopoietic stem-cell transplantation or pregnancy, where several conditions (infections, drugs, and hypertensive disorders of pregnancy) may mimic the classical triade of thrombotic microangiopathy.

Second, although significant breakthroughs have been achieved in our understanding of some forms of thrombotic microangiopathy, the exact mechanisms of endothelial cell injury remain elusive in several types of secondary thrombotic microangiopathy (malignancy-associated and some forms of drug-associated thrombotic microangiopathy/HUS). Third, distinct mechanisms of endothelial cell injury may overlap in a given form of thrombotic microangiopathy,⁷ as illustrated by hematopoietic stem-cell transplantation–associated thrombotic microangiopathy in the study by Werion et al. Finally, and most crucially, a positive diagnostic test for a predominant or partial contribution of complement activation to the pathogenesis of a given HUS is still lacking.

In the quest for an optimal classification of HUS, the underlying question is to determine in what type of thrombotic microangiopathy/HUS complement should be inhibited. To assess the implication of complement activation in the pathogenesis of a form of thrombotic microangiopathy, one may currently rely on the frequency of rare pathogenic complement variants, markers of complement activation, relapse rate, and response to complement blockade. When applied to secondary thrombotic microangiopathy/HUS, these criteria do not help draw any definite conclusion. The frequency of pathogenic complement gene variants is similar in patients with secondary HUS and healthy controls, as outlined once again in the study by Werion et al. However, this does not preclude transient complement activation. Systemic or urinary markers of complement activation have not been fully assessed in the available studies. The relapse rate of secondary HUS is not reported by Werion et al. but was lower than 1% in two previous studies.⁶ Finally, the benefit of C5 blockade on the outcome of secondary HUS is not proven. In some countries, C5 blockers are not even funded for secondary HUS, as in the study by Werion et al. In others, the effect of C5 blockade is hard to evaluate because of the heterogeneity of secondary HUS and the effect of the underlying conditions/disease.⁶ Not surprisingly, industry-sponsored trials failed, to date, to give any definite answer. We have recently supported the use of short-term C5 blockade in some forms of secondary thrombotic microangiopathy/HUS (drug and autoimmune diseases–associated forms) as a salvage therapy,¹¹ but this was not based on definite evidence.

How can we move forward? More studies are needed to understand the intimate mechanisms of secondary thrombotic microangiopathy/HUS, a first step toward specific targeted therapies. Most importantly, clinicians involved in the care of patients with thrombotic microangiopathy and HUS should refer to a common terminology. An optimal pathogenesis-driven classification of microangiopathy and HUS may be within reach in the few coming years. Despite all the major breakthroughs in the past two decades, HUS has still its share of unresolved questions and continuing debate.

Disclosures

F. Fakhouri reports consultancy for, honoraria from, and advisory roles for Alexion, Alnylam, Apellis, Novartis, Roche, and Vifor. V. Frémeaux-Bacchi served as consultant for Alexion Pharmaceuticals, Apellis, BioCryps, Novartis, Roche, Sobi, and UCB; reports research funding from Alexion Pharmaceuticals; reports honoraria from Alexion Pharmaceuticals, Apellis, BioCryps, Novartis, Roche, Sobi, and UCB; and reports advisory or leadership roles for Alexion Pharmaceuticals, Apellis, BioCryps, Novartis, and Sobi. The remaining author has nothing to disclose.

Funding

None.

Author Contributions

Conceptualization: Fadi Fakhouri, Véronique Frémeaux-Bacchi, Nora Schwotzer.

Writing – original draft: Fadi Fakhouri, Véronique Frémeaux-Bacchi, Nora Schwotzer.

Writing – review & editing: Fadi Fakhouri, Véronique Frémeaux-Bacchi, Nora Schwotzer.