Microangiopathic Anemia

The term thrombotic microangiopathy (TMA) describes an etiologically very heterogeneous group of diseases that – in conjunction with endothelial damage – lead to sustained thrombotic clotting in vessels of the microcirculation. Secondarily, this leads to the consumption of platelets and mechanical hemolysis. The classic triad of direct antiglobulin test (DAT)-negative hemolytic anemia with detection of schistocytes in the blood smear, thrombocytopenia and ischemic end-organ damage defines a TMA. Depending on the vascular systems involved, renal failure, neurological symptoms, cardiac complications, respiratory failure, visual disturbances, pancreatitis, intestinal ischemia and – rarely – skin changes occur [1, 2]. Lethality is high, with 70–90% of patients dying if left untreated [3, 4]. The recognition of a TMA and its classification according to pathophysiological aspects within the first 48 h are prerequisites for successful treatment. Over the past 20 years, studies on the heterogenous pathophysiology of TMA have contributed significantly to standardized diagnostic procedures and treatment approaches guided by the underlying causes of disease. As a result, the terminology has been clarified. The term TMA includes diseases such as classical thrombotic thrombocytopenic purpura (TTP), both congenital (cTTP) and immune-mediated (iTTP), complement-mediated TMA (cmTMA) (former atypical hemolytic uremic syndrome [aHUS]), Shiga toxin-induced HUS (STEC-HUS), and Streptococcus pneumoniae-associated HUS (SP-HUS) as well as secondary TMA diseases. Most important is the rapid discrimination of STEC-HUS, SP-HUS (both mainly found in children) and TTP. A reliable distinction can be difficult at the onset of disease. The medical history (e.g., child with diarrhea: STEC-HUS), as well as the determination of the ADAMTS13 activity and the ADAMTS13 antibodies are decisive.

A case of fatal TMA was first described by Eli Moschcowitz in 1924 in a 16-year-old female patient. The young woman, who had previously been in excellent health, fell ill with weakness, paleness, and high fever. She also reported pain in her wrists and elbows. After being admitted to Beth Israel Hospital in New York, the patient’s condition continued to deteriorate. Already upon admission to the clinic, petechial bleeding on the forearms was described. The laboratory diagnostics available at that time were quite limited, but anemia with a hematocrit of 40% and leukocytosis are well documented. Platelet counts are not determined. The creatinine level at admission was 1.1 mg/dL. A test that is no longer in use today showed a “fragility” of the erythrocytes as a sign of hemolysis. In the following days, a complex clinical picture developed with pulmonary edema, bleeding tendency, hemiparesis and finally coma. On the sixth day of the hospital stay, the patient died. An autopsy was performed. A very detailed description of the histological diagnostics of the examined organs (kidneys, heart muscle, liver, and lungs) describes the detection of hyaline thrombi in the small and smallest vessels of all examined organs, especially in the heart but also in the kidneys. This case report was first published in 1924 and is considered the first description of TTP.

The leading finding of TMA is microangiopathic-hemolytic anemia with low platelets, low haptoglobin, reticulocytosis, and usually elevated lactate dehydrogenase level (LDH). The detection of schistocytes in the blood smear is an important and easy-to-perform laboratory test at the time of diagnosis and should always be carried out promptly – even at night – in case of suspected TMA. The morphological assessment requires experience and patience as the number of schistocytes varies greatly. In case of doubt, the smear should be examined again the following day. The detection of a leucoerythroblastic blood count with the appearance of red and white precursors may be indicative of bone marrow carcinosis. Since extensive bone marrow involvement is frequently associated with acquired coagulation disorders, routine testing, e.g., for hyperfibrinolysis, is recommended. In doubt, a bone marrow biopsy may be necessary.

In addition to a machine blood count, basic diagnostics include a blood smear for morphological assessment and hemolysis parameters (LDH, reticulocytes, haptoglobin, DAT). To assess organ dysfunction, it is useful to determine creatinine, troponin, NT-proBNP, and liver enzymes. It is important to differentiate the disease from other TMA-like clinical pictures as well as between the different forms of TMA. In particular, the early differentiation of TTP from other forms of TMA is essential. The key test is the determination of ADAMTS13 activity in the blood. It is imperative that the blood sample (usually citrate blood) for this test is taken before any transfusion of blood products (erythrocyte concentrates, platelet concentrates, fresh frozen plasma) as related ADAMTS13 replacement induces changes in the measured activity. As a consequence, ADAMTS13 activity may be measured too high, resulting in incorrect interpretation of the underlying pathophysiologic process.

The DAT is typically negative. A positive DAT does not completely rule out the diagnosis of TMA. The combination of DAT-positive autoimmune hemolysis and immune thrombocytopenia (Evans syndrome) is very rare; however, schistocytes are not elevated in this case. A complete coagulation status (prothrombin time, activated partial thromboplastin time, fibrinogen concentration, antithrombin activity, D-dimers) allows the differentiation between disseminated intravascular coagulation and hyperfibrinolysis, e.g., in disseminated cancer. Severely impaired renal function at diagnosis is a characteristic of HUS and is not typical for TTP. If HUS is suspected, microbiological diagnostics are necessary to detect a toxin-mediated disease (STEC-HUS, SP-HUS). Recommended laboratory tests are shown in Table 1. A comprehensive diagnostic workup in TMA patients is sketched in Figure 1 [5].

Even during the early workup of TMA, the differential diagnosis needs to be carefully evaluated since many other medical conditions may imitate TMA and need to be identified and treated differently. Table 2 provides an overview of diseases that may resemble TMA.

If ADAMTS13 testing is not available in a timely manner, the PLASMIC score can be determined as a surrogate [6]. The scoring system was published in 2017 to calculate the probability of severe ADAMTS13 deficiency in TMA patients in order to facilitate the early identification of TTP patients [6]. In the meantime, the parameters have been validated many times [7]. The following parameters are included in this score: platelet count, reticulocytes, haptoglobin, indirect bilirubin, creatinine, INR, MCV, and medical history (malignant disease, organ transplantation, hematopoietic stem cell transplantation). Table 3 specifies the PLASMIC score parameters. Deviations in the INR indicate coagulation disorders in general. Therefore, any measurable coagulation disorder should be taken into account when applying the PLASMIC score. The diagnosis of TTP with ADAMTS13 deficiency is likely in cases with a PLASMIC score above five [6]. Nevertheless, sensitivity and specificity of the PLASMIC score are limited and rapid ADAMTS13 testing is the preferable standard.

Due to the importance of the von Willebrand factor (vWF) cleavage protease ADAMTS13 and the detection of antibodies against ADAMTS13, TTP is now also referred to as ADAMTS13-dependent TMA or acquired/immune-mediated TTP (iTTP) [1]. In iTTP there is severe ADAMTS13 deficiency, leading to uncontrolled thrombus formation. The thrombi consist largely of platelets and vWF and little fibrin, in contrast, the thrombi of STEC-HUS, SP-HUS and cmTMA are enriched with fibrin and contain less platelets [1]. Thrombi develop in the microcirculation of all organs, mostly in the central nervous system, the heart and the kidneys. The mechanical stress of erythrocytes in partially occluded blood vessels as well as the formation of vWF fibrils and the formation of neutrophil extracellular traps is presumably responsible for the formation of schistocytes in vivo [8, 9].

Clinically, these patients often present with non-specific prodroma such as “flu-like” symptoms or diarrhea. The highly variable clinical symptoms range from general weakness to stroke or sudden cardiac death. What is particularly important is that clinical symptoms often fluctuate and very sudden, sometimes catastrophic, deteriorations can occur within a few hours. The pentad originally described as typical [3], including fever, neurological disorders, hemolysis, thrombocytopenia and renal dysfunction, only occurs when the disease is fully developed. However, presentation with only a single or few of these clinical characteristics is very common. First and foremost are neurological symptoms such as headaches, confusion, neurological deficits and seizures. Severe hemolysis and pronounced thrombopenia with clinical symptoms usually occur only after a certain period of the disease, when the consumption of platelets and mechanical hemolysis comprise a large vascular cross-section. Conversely, thrombus formation in a few vessels relevant for organ blood flow – for example, in the brain stem – can lead to severe neurological symptoms without predominant thrombocytopenia or hemolysis. In such cases, the diagnosis of TTP is usually not made and individual patients often suffer recurrent ischemic episodes and insults for years, e.g., in the context of cTTP, the so-called Upshaw-Schulman syndrome. Even severe ADAMTS13 deficiency is not necessarily linked to clinical apparent TTP, so a second trigger such as autoimmune disorder, pregnancy, infection (e.g., HIV, CMV, EBV), cancer, or application of certain drugs is required for initiating the prothrombotic process [10]. Several cases of iTTP may be associated with the aforementioned conditions since anti-ADAMTS13 antibodies can be induced under such circumstances [11]. Those cases are attributed to the term secondary iTTP. Conditions most frequently associated with secondary iTTP are HIV, systemic lupus erythematodes or the use of ticlopidin [12].

The quantitative determination of the activity of vWF-cleaving protease ADAMTS13 and the ADAMTS13 inhibitor are essential for the differential diagnosis of TTP. As previously stressed, blood sampling for ADAMTS13 diagnostics must be taken before transfusion or plasma therapy is initiated and a sample in citrate blood is mandatory for the diagnostic workup. EDTA blood is unsuitable for determining the required parameters since EDTA irreversibly inhibits protein activity [13, 14]. Follow-up studies of ADAMTS13 activity and inhibitor levels allow monitoring of the effectiveness of immunosuppressive therapy. Although there is an association between the risk of recurrence in iTTP and low ADAMTS13 activity cannot predict recurrence in the individual patient. The intra-individual variation of protease levels are very strong [15]. Thus, in addition to low ADAMTS13 activity, additional factors or triggers are necessary to cause a clinical manifestation of TTP as mentioned above.

In addition to the common, acquired form, there is the very rare congenital TTP [16], also called Upshaw-Shulman syndrome, with hereditary deficiency of the ADAMTS13. Only about 5% of all TTP cases can be traced back to genetic defects, the prevalence of cTTP is 0.2–0.5 per one million [17]. Typically, low ADAMTS13 activity in the absence of anti-ADAMTS13 antibodies is seen in cTTP. The first description was made by Upshaw and Schulmann, who in the 1960s and 1970s reported patients with recurrent episodes of TMA. Plasma infusion was described to stop the TMA episode [18]. The cTTP is caused by an autosomal recessive homozygous or compound heterozygous mutation of the ADAMTS13 gene with severely decreased ADAMTS13 plasma activity. Heterozygous carriers are asymptomatic. More than 200 mutations are currently described, distributed throughout the whole coding region of the ADAMTS13 gene. Depending on the underlying mutation, patients are symptomatic in early childhood or do not develop the disease until the 3rd–4th decade of life despite low levels ADAMTS13 for years [19]. Possible triggers include infections, pregnancy, surgery, but also alcohol abuse [2].

The introduction of plasma exchange (PEX) has reduced the lethality of TTP from >70 to 90% to less than 10% [4]. PEX is intended to increase ADAMTS13 activity and eliminate anti-ADAMTS13 antibodies; nevertheless, ADAMTS13 activity is often still low after PEX. To avoid irreversible organ damage, plasma exchange was considered to be necessary within the first 24 h after suspected diagnosis. But there is growing evidence that plasma-free treatment may be possible by using caplacizumab [20]. Caplacizumab, a bivalent humanized single-variable domain of heavy-chain nanobody, is an anti-vWF antibody and currently the only approved drug for the treatment of iTTP. By blocking the binding site of platelets to vWF, the A1 domain, thrombus formation is interrupted and further end-organ damage can be avoided [20]. On the other hand, caplacizumab increases the risk of bleeding, especially when antiplatelet agents and/or anticoagulants are used at the same time. In an emergency, the use of vWF concentrates may be necessary to correct hemostasis. Caplacizumab is approved for the treatment of iTTP in combination with plasma exchange at a dose of 10 mg per day for up to 30 days after completion of plasma exchange therapy. The daily treatment costs currently amount to around EUR 4,500. In iTTP, the combination with steroids is recommended, but it is usually not possible to control the antibody-producing B cells in the long term. The use of the B cell antibody rituximab shortens the time to response to plasma exchange and reduces the risk of recurrence [10]. Rituximab is not approved for the treatment of iTTP (off-label use), but widely recommended by almost all guidelines [10, 21, 22]. The dose of rituximab is the subject of current studies; currently, the administration of 375 mg/m² once a week for 4 weeks after plasma exchange session is common. The cost of one dose of rituximab is approximately EUR 3,500. A pre-emptive strategy of rituximab at ADAMTS13 relapse with known iTTP may prevent clinical recurrence. Rituximab is the drug of choice for the long-term control of immune-mediated TTP. Cases not responding to rituximab treatment have been successfully treated with other immunosuppressants such as bortezomib, daratumumab, vincristine, ciclosporin, azathioprine, cyclophosphamide, or MMF [1, 12]. Clinical trials for the treatment of iTTP with recombinant ADAMTS13 (apadamtase alfa) are ongoing [23].

As with iTTP, plasma therapy may be also necessary for patients with cTTP. The intensity of treatment depends on the degree of clinical severity. Asymptomatic patients with no organ damage, no signs of hemolysis, and normal platelet counts may also be observed initially only [24]. In case of an acute episode with organ dysfunction, plasma infusions are necessary and may be repeated daily until complete resolution of TMA symptoms and may be tapered to intervals of about 14 days (on demand treatment). Patients with frequent, repeated episodes can be assigned to permanent, preventive treatment. In severe cases with ongoing trigger – e.g., pregnancy – plasma exchange may be necessary [12, 23, 25]. Recombinant ADAMTS13 is already approved by the Food and Drug Administration (FDA) and is meanwhile available in Europe for the treatment of ADAMTS13 deficiency in patients with cTTP [26].

TMA with predominant involvement of the kidneys is assigned to the term HUS. The most common form of HUS occurs usually after infection with Shiga toxin-producing Escherichia coli (STEC-HUS), also called entero-hemolytic E. coli (EHEC). In children, about 80–90% of all HUS cases are STEC-induced. Less than 5% of cases with HUS are described after infections with other pathogens such as S. pneumoniae (SP-HUS). Only about 5–10% of cases can be attributed to cmTMA, formerly called atypical HUS (aHUS). Especially in adult patients, many cases are attributed to be secondary TMA. To this group all forms of TMA with underlying, hitherto perhaps not diagnosed, diseases or causes are attributed. Examples for those secondary TMA are transplantation-associated taTMA (after stem cell or solid organ transplantation), drug-induced TMA, autoimmune diseases (e.g., catastrophic antiphospholipid), pregnancy or malignancy-associated TMA [27]. Those forms are not always clearly determined by a certain pathophysiology, but there is growing evidence that in some situations an overwhelming activation of the complement system may play an important role.

All forms of HUS have in common endothelial cell damage, which mainly affects the capillary region of the kidney, and less frequently in other organs. If there is a transient trigger (Shiga toxin, S. pneumoniae, medication, surgery), spontaneous remission usually occurs with supportive therapy. However, if there is a genetic defect or acquired ongoing dysregulation of the complement system or the coagulation system, TMA can persist even after controlling the trigger and may lead to persisting damage to affected organs – usually the kidneys. Examples include complement regulation defects due to mutations of factor H, C3, membrane cofactor protein (MCP, CD46), thrombomodulin, diacylglycerol kinase epsilon (DGKE), cobalamin metabolic defects, or autoantibodies against factor H [28]. In patients with the mutations mentioned above, they may suffer from TMA if persisting or long-lasting acquired triggers, such as allogeneic stem cell transplantation (PBSCT), malignancy, or pregnancy occur. Jodele and co-workers [29] investigated this in a collective of children after allogeneic PBSCT. In some of the patients with taTMA, genetic variations, especially of factor H, could be detected. Whether such predisposing mutations worsen the prognosis in patients with STEC-HUS or in TMA after exposure to calcineurin inhibitors remains unclear, but individual case reports already exist [30, 31]. Such “silent” genetic mutations may explain, why only a small part of patients infected with EHEC develop STEC-HUS. However, there are up to now no systematic studies on this.

In 1955, the term “hemolytic uremic syndrome” was first used by Conrad Gasser, a Swiss pediatrician [32]. In the 1970s, a link with Shigella infections was suspected [33], an association with E. coli O157:H7 (EHEC) more than 10 years later. Over the years, other toxin-producing E. coli strains have been identified. Starting from an intestinal infection, transmission of Shiga toxin through the intestinal wall leads to the hematogenous distribution of the toxin. Binding to endothelial cells of the kidney via CD77 (globotriaosylceramide, Gb3) induces their apoptosis with release of vWF. Thrombosis of the renal vessels, and in many cases also of other electrical areas, leads to the end-organ damage typical of TMA. The activation of the complement system in the typical HUS also leads to the destruction of endothelial cells. Renal failure requiring dialysis is common, and some patients suffer from permanent renal impairment [34].

HUS diseases after severe infections with other pathogens such as S. pneumoniae have been documented, especially in children [35] and are referred to as SP-HUS. Overall, however, these cases are much rarer than STEC-HUS [28].

A small proportion (5–10%) of patients with HUS have no typical history of infectious, bloody diarrhea and may be assigned to have cmTMA. There are no exact data on epidemiology of cmTMA for adults, but the proportion of hereditary cmTMA cases is likely to be lower than in children. Strictly speaking, the term cmTMA may only be applied after the exclusion of an underlying cause, disease or infection. To diagnose cmTMA, Shiga toxin-induced TMA (typical HUS, Shiga toxin detection in stool or blood, blood cultures), ADAMTS13-mediated TMA (TTP), secondary TMA such as drug-induced TMA, malignancy-associated TMA must be excluded (Fig. 1). Evidence of complement mutations, inborn coagulation abnormality, or DKGE (diacylglycerol kinase ɛ) mutation is not mandatory for the diagnosis of cmTMA [28], but according to the current knowledge activation of the alternate pathways of the complement system is the key finding in many patients with cmTMA. cmTMA is based on dysregulation of the alternative complement pathway (Fig. 2). Congenital defects in complement regulation are more common than acquired mutations: the most common are factor H mutations, and more than 80 mutations have now been identified. Most of the genetic changes are inherited heterozygous. Not all carriers of the mutation develop the disease; the age of the first manifestation varies greatly from person to person. However, the risk of cardiovascular events is greatly increased, presumably caused by chronic complement activation with the development of arteriosclerosis [36]. Very rare genetic causes affect the metabolic pathways for Cobalamin and DGKE. Both diseases are mainly diagnosed in childhood, usually in the first year of life [28]. Acquired forms of cmTMA are very rare and are based, among other things, on antibodies against factor H (6–10% of all cmTMA patients), which usually occurs in childhood or adolescence and is frequently associated with a complement factor H mutation (CFH protein 1 and protein 3) [27].

Clinically, patients with cmTMA present with recurrent episodes of TMA, and in the absence of specific therapy, permanent renal failure occurs in about 25–30% of cases [34]. The standard treatment is complement-inhibiting therapy with the terminal complement inhibitors with anti-C5 activity eculizumab or ravulizumab. In cases with anti-factor H antibodies, immunosuppression aimed at elimination of the antibodies in combination with complement inhibition is standard of care. Ravulizumab is approved for the treatment of cmTMA (aHUS) at loading dose of 2,700 mg and a maintenance dose of 3,300 mg every 8 weeks (body weight 80 kg). The annual treatment costs for ravulizumab are approximately EUR 324,000. The most important side effect of using complement inhibitors are infections with Neisseria meningitidis. Therefore, vaccination against the most important strains (serogroups A, B, C, W, Y) is necessary, if possible before starting treatment. Antibiotic prophylaxis is required until vaccination protection is complete.

In addition to the previously mentioned, well-defined entities, there are several clinical pictures that are far less clearly classified. These conditions are called secondary TMA or secondary HUS. It may be based on malignant disorder, stem cell or solid organ transplantation, drug (ab)use, pregnancy, autoimmune diseases or primary kidney diseases.

For the first time, a connection between the intake of quinine and its derivatives and the occurrence of hemolysis was documented in the 1960s. A dose-independent formation of antibodies against platelets and endothelial cells is assumed [37]. The prognosis of quinine-induced HUS is unfavorable; permanent renal dysfunction is documented in about 70% of patients [38]. There are also indications of such immune-mediated damage for gemcitabine. A second, dose-dependent pathomechanism is directed to endothelial cells with consecutive formation of TMA. This mechanism has been described for most of the drugs. Classic representatives are the calcineurin inhibitors cyclosporine A and tacrolimus [39]. The number of drugs associated with TMA has grown in the past years mostly due to the progress in treating patients with malignant diseases. Most frequently reported drugs causing TMA in cancer patients are listed in Table 4 [40].

If TMA occurs after allogeneic stem cell transplantation or solid organ transplantation, it is referred as taTMA. Complement activation is suspected, the causes are probably endothelial cell damage in the context of conditioning, drug side effects (calcineurin inhibitors), chronic graft versus host disease, and infections. Clinically, the disease is extremely heterogeneous. In addition to the hematological complications, multiple end-organ damage involving the CNS, kidneys, intestine, and heart occur in some patients. The prognosis of taTMA is unfavorable, and a standard treatment is not yet known. The use of complement inhibitors needs to be evaluated by controlled trials [41]. Recently the new complement inhibitor for the treatment of taTMA narsoplamib – an inhibitor of the lectin pathway – has shown promising data [42].

Thrombopenia associated with an LDH elevation in pregnancy raises the suspicion of a pregnancy-related disease, such as HELLP syndrome, acute gestational fatty liver disease, preeclampsia or in contrast, a pregnancy-triggered disease like HUS or TTP. A smear of peripheral blood is always required. If schistocytes are found, there is much to suggest a pregnancy-triggered TMA. If schistocytes are missing and other findings indicate a pregnancy-related disease, rapid caesarean delivery is necessary. The initiation of plasma exchange should not be delayed until the ADAMTS13 result is available as soon as a TMA is detected. The rapid determination of the ADAMTS13 activity usually allows a discrimination between TTP and HUS. Data from the registry analyses show that TTP is more likely during pregnancy and cmTMA is more likely after delivery. Both diseases can be congenital, and pregnancy is merely the trigger [43‒45].

If the constellation of a TMA occurs in the context of an advanced tumor disease, a drug-induced TMA (e.g., after administration of gemcitabine) should always be considered in the first place. In case of extensive metastasis to the bone marrow and/or flushing of tumor cells into the vascular system, a TMA-like picture can develop, often accompanied by hyperfibrinolysis. The prognosis is extremely unfavorable and bone marrow histology may help confirm the diagnosis.

In antiphospholipid syndrome, arterial and venous thrombosis occurs in young patients (often during pregnancy). Prolonged aPTT and the detection of antibodies against phospholipid-binding proteins are typical. The uncontrolled thrombus formation in the context of a so-called “catastrophic antiphospholipid syndrome” not only has characteristics of an microangiopathic-hemolytic anemia, but some data also suggest an uncontrolled activation of the classical complement pathway with potential efficacy of complement inhibitors [46].

During the last decade medical advance has brought substantial progress to patients with TMA. Rapid ADAMTS13 testing results in early identification and classification of TMA and enables the use of targeted drugs such as caplazicumab in iTTP, complement inhibition in cmTMA and recombinant ADAMTS13 in cTTP. There is still a large group of patients with TMA, where pathophysiology of the disease remains unclear and needs more scientific workup. Since there is no specific and reliable test to identify cmTMA patients, extensive diagnostic workup is mandatory to identify patients who may have benefit from complement inhibition. New complement inhibitors may enhance therapeutic efficiency in selected cases.

M.B. has received honoraria from Alexion, Sanofi, and Takeda. J.B. states no conflict of interest.

This study was not supported by any sponsor or funder.

M.B. drafted and reviewed the manuscript and approved its final version. J.B. reviewed and revised the manuscript.