Thrombotic thrombocytopenic purpura and hemolytic uremic syndrome: will recent insight into pathogenesis translate into better treatment?

In 1925, Moschocowitz described a 16-year-old girl who developed thrombocytopenia, hemolytic anemia, renal impairment, fever and fluctuating neurologic abnormalities. The patient died after a rapid fulminating course of what we now recognize was thrombotic thrombocytopenic purpura (TTP).1 The pathology was as startling to Moschocowitz as it was mysterious: hyaline thrombi were present throughout the small capillary vessels in this young person. Some 30 years later, Gasser and coworkers described the occurrence of thrombocytopenia, hemolytic anemia, and renal failure in a syndrome that was to become known as hemolytic uremic syndrome (HUS).2 HUS, like TTP, is also characterized by microvascular thrombi, but in the case of HUS, they tend to occur within the renal microcirculation.3 Although both syndromes have been recognized for many years, there has been a dissociation between our understanding of the pathophysiology of each disorder and the development of effective treatments. Particularly for TTP, successful treatment has followed serendipitous and often empiric observations, which have been independent of our understanding of the basic pathophysiology. In this editorial, I will comment on some of the recent advances in our understanding of these disorders and speculate on whether they will translate into therapeutic advances. Both TTP and HUS are characterized by destructive thrombocytopenia and schistocytic hemolytic anemia, also known as microangiopathic hemolytic anemia. Hence, some physicians (the “lumpers”) have suggested that they represent overlap syndromes. But thrombo-cytopenia and schistocytic hemolytic anemia are also observed in other quite different disorders, including disseminated intravascular coagulation, a variety of vasculitic disorders, and, to a variable extent, malignant hypertension, and certain prosthetic heart valves. Consequently, I will comment on both the differences and the overlapping characteristics. TTP is classically defined by the pentad of thrombocytopenia, schistocytic hemolytic anemia, renal impairment, fever, and neurologic impairment, which is often fluctuating. However, it is uncommon for all components of the pentad to occur in the same patient. In the largest prospective trial examining TTP,4 all patients had thrombocytopenia and RBC fragmentation. Seventy-one percent of patients had neurologic abnormalities (typically waxing and waning); 59 percent had renal impairment of some type (typically an elevated blood urea nitrogen level); and 24 percent had fever. Untreated, TTP is often a catastrophic disorder in a previously well person, and retrospective studies have suggested that most untreated TTP patients (about 80%) will die.5, 6 The majority of TTP patients have a sudden onset of illness that may be precipitated by an immunologic stimulus such as an infection, a vaccination, or a pregnancy, among other triggers. There are several distinct subsets of TTP that have dramatically different triggers, but a similar natural history, to classical TTP. The best examples are drug-induced TTP associated with ticlopidine,7, 8 clopidogrel,9 and, rarely, quinidine.10 Approximately 20 to 35 percent of patients with TTP will have a relapse of their disorder. Relapse can occur many years after a complete recovery, suggesting that the underlying predisposition to TTP continues in these patients. Splenectomy can induce a long-term remission or cure in these patients.11 Two dramatically different TTP-related syndromes, which have unusual presentations and are often refractory to therapy, occur in patients who have undergone BMT,12 or in patients with cancer.13 HUS usually differs from TTP in presentation, natural history and the response of patients to treatment, leading other physicians (the “splitters”) to conclude that HUS and TTP are quite different disorders. HUS is more common in children, whereas TTP is more common in adults. HUS is divided into two major types with several subcategories.14 Most episodes (80-90%) of HUS are diarrhea-associated and are typically triggered by an infection.14 These diarrhea-associated cases of HUS are also known as classic HUS, or epidemic HUS. The large number of cases have provided a general, but not invariably consistent, picture. Typically, patients are affected at the extremes of life (the very young, less than 2-3 years of age; and the very old, greater than 80 years of age). Infection by the most commonly implicated agent (Shiga toxin-producing Escherichia coli O157:H7)15-17 often is related to transmission from the gut of otherwise healthy cattle.18 This can occur through inadequately cooked meat (“hamburger disease”), contaminated water from the run-off of pastures that infects the water supply; apples (with contaminated apple cider); and, less commonly, vegetables. The outcome of infection by E. coli O157:H7 can be conceptualized as a pyramid. An uncertain proportion of exposed individuals will have no symptoms and some will become carriers. For others, symptoms occur after an incubation phase of 3 to 5 days. Intra-abdominal cramps occur in most children and these are followed by watery diarrhea in the majority of those affected. A smaller group, 20 to 40 percent, will have bloody diarrhea. Approximately 5 to 15 percent of these infected children will progress to a HUS-like syndrome.14, 18-20 A small subset of patients with HUS, the tip of the pyramid, will have a disorder that is essentially identical to adult TTP. The outcome of HUS in these patients is frequently the same as the outcome for adult TTP, with progressive deterioration and death due to thrombotic events. It is intriguing to note that individual susceptibility determines both the likelihood that patients exposed to E. coli will develop HUS or TTP and their overall survival. For example, approximately 10 percent of affected children with these infections progress to HUS, and most recover.21, 22 However, approximately 25 percent of elderly patients in a nursing home who were reported to have a similar infection developed a HUS and/or TTP syndrome, and many of these patients died.23 The second type of HUS is much less common (comprising approximately 10% of HUS cases), and is a sporadic disease that is not preceded by diarrhea. This is sometimes known as atypical HUS, diarrhea-negative HUS, or endemic HUS.14 Lying conceptually between classical TTP and classical HUS is an exceptionally rare disorder that has been the focus of much attention in recent years. Familial TTP/HUS, sometimes called the Upshaw-Schulman syndrome24-26 or congenital TTP/HUS, is a recurring, chronic and relapsing TTP/HUS-like disorder, which often occurs in families. These patients most frequently present as babies or young adults with thrombocytopenia, schistocytic hemolytic anemia, and other episodes of renal impairment that can resemble classic sporadic HUS.27, 28 Some patients have neurologic impairment that can resemble classical TTP. The treatment for this disorder is plasma infusion given either intravenously or by exchange, and laboratory investigation of these patients has been the subject of recent studies. Recently, there has been excitement about some advances in our understanding of the pathogenesis of TTP/HUS. Perhaps this rational exuberance reflects the hopes of scientists to identify a common thread that would tie TTP and HUS together. That thread is vWF. vWF is a massive, multimeric, circulating protein that plays a pivotal role in binding platelets to the vessel wall (adhesion). The larger the multimer of vWF, the more reactive it is with platelets. Hence, one can anticipate that very large multimers of vWF could trigger platelet clotting and potentially cause a platelet-mediated thrombotic disorder such as TTP or HUS. Very large multimers of vWF are produced by endothelial cells, but are cleaved into smaller and less platelet-reactive multimers by a vWF-cleaving protease. The first evidence implicating vWF in TTP was produced by Moake and coworkers in 1970.29 These investigators described unusually large plasma multimers of vWF in patients with the chronic and relapsing familial form of TTP. The largest multimers were present during convalescent intervals and absent during relapses, leading Moake and colleagues to postulate that the large vWF multimers participate in the platelet-thrombotic events that characterize the syndrome. Subsequent studies by Moake et al.30 demonstrated that these large multimers probably originated from endothelial cells rather than platelets. The observations of my group extended the work of Moake and coworkers: we identified a platelet-aggregating factor in a patient with familial relapsing TTP/HUS.31 This patient behaved similarly to the patients described by Moake et al. because the largest multimers of vWF were inversely associated with disease activity. My group has continued to investigate platelet-aggregating and platelet-lytic factors in TTP;32-34 however, most other investigators have studied vWF. Over the past several years, the focus of activity has been on identifying the vWF-cleaving protease from normal plasma.35, 36 The vWF-cleaving protease was functionally characterized and was shown to be dependent on metallic ions.35, 36 However, the question still remained unanswered: was the protease activity that cleaves vWF causally related to familial TTP? Was it related to infectious or endemic HUS, and, perhaps most importantly, was it related to sporadic episodes of adult TTP?37 At least part of the answer is now in. In a landmark study, Levy and 18 investigators representing 13 research centers used genome-wide linkage analysis in four separate families with familial (chronic and relapsing) TTP/HUS.38 These investigators measured the plasma level of vWF-cleaving protease in a functional assay (see subsequent discussion) and related it to a gene locus on chromosome 9q34. Affected patients who also had abnormal levels of protease (as measured in the functional assay) consistently had gene defects that could be mapped to ADAMTS-13. ADAMTS-13 is a member of the ADAM (a disintegrin and metalloproteinase) family. This novel gene had, in addition, a thrombospondin-like domain, hence its ADAMTS-13 designation. Levy and coworkers provided compelling evidence that a variety of mutations within this gene were associated, presumably causally, to the familial form of TTP/HUS. However, today the key question remains whether this remarkable observation can be extrapolated more broadly to the much more common types of either adult TTP or childhood and adult HUS. At the time of writing this editorial, the issue remains unresolved. Evidence implicating this or other related enzyme defects in the role of vWF-protease in TTP comes from two groups. Both Furlan et al.39, 40 and Tsai and Lian41 have provided evidence of deficienct vWF protease activity in patients not only with the familial form of TTP but, more importantly, with the acute and sporadic adult forms of TTP. Some patients appeared to have evidence of antibodies that inactivated the protease. However, a recent study that my group performed draws this hypothesis into question. In a blinded, coded fashion, we measured vWF protease activity in a large group of TTP patients, as well as in thrombocytopenic and nonthrombocytopenic controls.42 We observed unmeasurable levels of vWF-cleaving protease activity in two unrelated patients with familial TTP, similar to the observations of Levy and coworkers.38 However, we found sometimes normal, but sometimes reduced, levels of vWF-cleaving protease in sporadic TTP patients, and these levels were not related to disease activity. More concerning was the observation of deficient vWF protease activity in many thrombocytopenic controls who did not have TTP. These observations have been confirmed by other investigators.43 Hence, whether or not vWF proteinase impairment is a contributor to the much more commonly encountered forms of TTP and HUS, or whether it is an abnormality primarily seen in patients with the familial, relapsing TTP/HUS remains uncertain. The observation of apparent mechanical fragmentation of RBCs in patients with a variety of schistocytic hemolytic anemias,44 including TTP, HUS, eclampsia, and malignant hypertension, among others, suggested the possibility that these anemias could have a similar pathophysiology. Other investigators noted that patients with disseminated intravascular coagulation also had fragmented RBCs.45, 46 This led to the scientifically based use of heparin as a treatment, first for disseminated intravascular coagulopathy, and then subsequently for TTP and HUS. However, heparin proved uniformly ineffective. Because of the exceptionally high mortality rate in untreated patients with TTP, many different empirical and intuitive treatments were tried. In 1959, Rubenstein and coworkers described remission in a TTP patient after fresh whole blood exchange.47 This observation was quickly confirmed, and in a series of reports48, 49 it was shown that plasma exchange or infusion could result in significantly improved survival for most TTP patients. However, to this day it has not been shown whether plasma infusion replaces a deficient or depleted factor, or whether plasma exchange removes a “toxic” factor. Although Rock and coworkers (including this author) found that a significant survival advantage is conferred by plasma exchange compared with plasma infusion,4 the study itself did not resolve the “toxin removal vs. deficient factor replacement” theory, because those patients treated with plasma exchange received more plasma than those treated by infusion. Subsequently, these investigators have produced some data (largely retrospective) showing that there is an added advantage with the use of cryosupernatant (which is depleted of vWF, among other factors) as the replacement fluid, an observation indirectly supporting a role for vWF in TTP.50 Other advances in the management of patients with TTP have followed empiric observations. For example, there is some evidence that corticosteroids are useful in patients with TTP, an observation consistent with the clinical observations implicating an immune pathogenesis.51 Similarly, our group has demonstrated a benefit of splenectomy as a maneuver to prevent relapses of TTP patients in a mode that may parallel the efficacy of splenectomy as a treatment for immune thrombocytopenia.11 The treatment of both endemic and epidemic HUS continues to be largely supportive, although it is my opinion that those patients who have severe TTP-like disease benefit from plasma exchange therapy. There have been small studies of plasma exchange in patients with TTP; however, the relatively (in comparison with TTP) low mortality of HUS makes it methodologically difficult to demonstrate efficacy. Advances in the understanding of underlying toxins that trigger epidemic HUS, such as E. coli O157:H7, have led to maneuvers to reduce infections or potentially neutralize toxins. To date, the results have been disappointing, and a recent study52 demonstrated that antibiotic treatment unexpectedly appeared to increase the overall risk of HUS development following E. coli O157:H7 infections.52 The authors speculated that this was related to the release of Shiga toxins from E. coli after antibiotic treatment. Today, we have a better understanding of the natural history and treatment options for several different types of HUS and TTP. Recent advances in our understanding of the role of vWF-protease activity in a subtype of TTP and HUS have produced exuberant anticipation of dramatic advances in therapy. The next several years will define whether this was rational or irrational exuberance.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.