Mechanisms of microvascular thrombosis in thrombotic thrombocytopenic purpura

Abstract

Recent studies have demonstrated that thrombotic thrombocytopenic purpura (TTP), a serious thrombotic disorder affecting the arterioles and capillaries of multiple organs, is caused by a profound deficiency in the von Willebrand factor cleaving metalloprotease, ADAMTS13. ADAMTS13, a 190-kD plasma protease originating primarily in hepatic stellate cells, prevents microvascular thrombosis by cleaving von Willebrand factor when the substrate is conformationally unfolded by high levels of shear stress in the circulation. Deficiency of ADAMTS13, due to genetic mutations or inhibitory autoantibodies, leads to accumulation of superactive forms of vWF, resulting in vWF-platelet aggregation and microvascular thrombosis. Analysis of ADAMTS13 has led to the recognition of subclinical TTP and atypical TTP presenting with thrombocytopenia or acute focal neurological deficits without concurrent microangiopathic hemolysis. Infusion of plasma replenishes the missing ADAMTS13 and ameliorates the complications of hereditary TTP. The patients are at risk of both acute and chronic renal failure if they receive inadequate plasma therapy. The more frequent, autoimmune type of TTP requires plasma exchange therapy and perhaps immunomodulatory measures. Current studies focus on the factors affecting the phenotypic severity of TTP and newer approaches to improving the therapies for the patients.

Thrombotic thrombocytopenic purpura (TTP) typically presents with thrombocytopenia and hemolytic anemia with fragmentation of the red cells. Other manifestations include mental changes, focal neurological deficits, seizures, hematuria, proteinuria, fever, abdominal pain with or without pancreatitis, and electrocardiographic abnormalities.¹ The conventional classification of microangiopathic disorders, based primarily on the clinical features of the patients, is notable for ambiguity and uncertainty. Recent discoveries of deficient ADAMTS13 (a disintegrin and metalloprotease with thrombospondin type 1 repeat, member 13) in TTP² and defective complement regulation in non-shiga toxin-associated hemolytic uremic syndrome³ have provided a new framework to classify thrombocytopenia and microangiopathic hemolysis on the basis of molecular defects and associated disease conditions (Table 1).

Table 1.

A molecular and etiologic classification of thrombotic microangiopathy

Renal failure: uncommon	Renal failure: common
Defective vWF proteolysis	Dysregulation of complement activation
Mutations of ADAMTS13	Mutations of CFH, IF, MCP, BF
Inhibitory antibodies of ADAMTS13	Autoantibodies of CFH
Tumor cell embolism	Shiga toxins (e.g. E. coli O157:H7)
Paroxysmal nocturnal hemoglobinuria	Neuraminidase (e.g. S. pneumoniae)
Infectious vasculitis (e.g. C. difficile, R. rickettsii, B. anthracis)	Drugs (e.g. calcineurin inhibitors, mitomycin, gemcitabine, quinine)
Idiopathic	Autoimmune disorders (e.g. SLE, scleroderma)
	BM/stem cell transplantation Miscellany (e.g. HELLP syndrome, surgery, pancreatitis)
	Idiopathic

BF, complement factor B; BM, bone marrow; CFH, complement factor H; HELLP, hemolysis with elevated liver enzymes and low platelet counts; IF, complement factor I; MCP, membrane cofactor protein (CD46); SLE, systemic lupus erythematosus; vWF, von Willebrand factor.

PATHOGENESIS OF THROMBOSIS IN TTP

In this review, TTP refers to the thrombotic disorder resulting from severe deficiency of ADAMTS13, due to genetic mutations or inhibitory autoantibodies. Very low levels of ADAMTS13 activity may occasionally be observed in patients with various pathological states, such as disseminated intravascular coagulopathy, liver disease, or sepsis. Nevertheless, it is unclear how accurately a low ADAMTS13 value measured in in vitro assays reflects decreased activity in patients with any of these conditions.

ADAMTS13, a metalloprotease of the M12B subfamily, cleaves vWF at the Tyr1605–Met1606 bond in the central A2 domain of the vWF polypeptide whenever this normally cryptic bond is rendered accessible by circulatory shear stress or chaotropic agents. This cleavage progressively converts the endothelial vWF polymer to smaller multimers that are conformationally less flexible and less adhesive. When ADAMTS13 is deficient, vWF multimers are conformationally unfolded by shear stress but are not cleaved, resulting in accumulation of hyperactive forms of vWF that cause platelet aggregation and microvascular thrombosis characteristic of TTP.¹

The evidence in support of this model of TTP is as follows: (1) exposure to arteriolar levels of shear stress in capillary tubes renders vWF susceptible to cleavage by ADAMTS13;⁴ (2) exposure of vWF to shear stress in various devices causes conformational unfolding that is detectable by either atomic force or fluorescent microscopy, increases its platelet-aggregating capacity, and causes direct vWF–platelet aggregation that is abolished by ADAMTS13.^5–9

The current model of ADAMTS13 as a regulator of vWF–platelet aggregation explains why the thrombi of TTP characteristically contain abundant vWF and platelets but little fibrin, and are limited to the high shear environments of arterioles and capillaries.^10,11 It also explains why the size of vWF increases early in the course of TTP, before the large multimers become progressively depleted along with worsening thrombocytopenia.¹² In animal models of ADAMTS13 deficiency, no thrombosis occurs if vWF is absent.¹³

In cell cultures or in ex vivo vascular perfusion studies, endothelial cells, when profoundly perturbed by agonists such as histamine, may become decorated with elongated strands of adherent vWF, providing the matrix for platelet adhesion.^14,15 This process, occurring under very low levels of shear stress, similar to those encountered in the venules, may not account for the thrombosis of TTP, which affects arterioles and capillaries but not venules. Furthermore, serial investigation of patients with relapsing TTP has shown that endothelial perturbation follows rather than precedes the onset of thrombocytopenia.¹⁶

Recent studies have suggested that binding of vWF to platelets or factor VIII may promote its cleavage by ADAMTS13 under shear stress.^17,18 Nevertheless, there is no evidence that vWF proteolysis by ADAMTS13 is impaired in patients with thrombocytopenia or factor VIII deficiency.

Clinical experience indicates that platelet thrombosis does not occur when ADAMTS13 is greater than 10% of the normal. However, there is no fixed threshold level of ADAMTS13 below which microvascular thrombosis invariably occurs, as the level is likely to be affected by multiple factors such as circulatory shear stress profile, platelet receptor and reactivity levels, vWF, and perhaps other as yet unknown factors.

In animal studies, inactivation of the ADAMTS13 gene causes microvascular thrombosis in some but not in other strains of mice.¹⁵ It is intriguing to speculate that the epistatic genes affecting the response to ADAMTS13 deficiency in mice might also contribute to the heterogeneity of phenotypic severity observed in patients with severe ADAMTS13 deficiency.

HEREDITARY TTP

In hereditary TTP, severe deficiency of ADAMTS13 results from homozygous or double heterozygous mutations of the ADAMTS13 gene. The mutations, mostly non-recurrent, cause severe deficiency of plasma ADAMTS13 activity levels by decreasing its biosynthesis, intracellular trafficking and secretion, and/or proteolytic activity.

A ‘two-hit’ hypothesis has been postulated by some researchers to explain the variable phenotypic severity associated with ADAMTS13 deficiency. Nevertheless, a review of hereditary TTP cases described in the literature and in our own series shows that ADAMTS13 was severely deficient in 19 of 59 siblings of the index cases. Of these 19 siblings, 17 had clinical disease; the other 2 cases had older sisters that were symptomatic only during pregnancy. Thus, current evidence indicates that, with few exceptions, severe ADAMTS13 deficiency is sufficient to confer TTP.

AUTOIMMUNE TTP

In autoimmune TTP, autoantibodies inhibit the proteolytic activity of ADAMTS13. Structural-functional analysis shows that the central spacer domain of ADAMTS13 is an essential component for the ADAMTS13 epitope recognized by TTP autoantibodies.^19,20 ADAMTS13 variants truncated upstream of the spacer domain are proteolytically active but are not suppressible by the autoantibodies of TTP. Such variants might have therapeutic advantages for autoimmune TTP.¹⁹

The incidence of TTP, estimated to be 1.74 cases per 10⁶ person-years in the Oklahoma registry,²¹ is likely to vary in different communities. The risk factors of autoimmune TTP include female gender, middle age (30–50 years), HIV infection, and perhaps African/Hispanic ethnicity. With the exception of ticlopidine,²² no etiological agents have been definitively associated with autoimmune TTP.

The levels of ADAMTS13 inhibitors tend to fluctuate, often causing one or more episodes of relapse before remitting to low or undetectable levels. Occasionally, the antibodies may increase not only in molar concentration but also in inhibitory potency, suggesting activation of the immune system and somatic hypermutation of the B cells.²³

Pregnancy is not a risk factor for autoimmune TTP. Nevertheless, pregnancy decreases the level of ADAMTS13 by ~30%, and more if it is complicated by pre-eclampsia/eclampsia or HELLP (hemolysis with elevated liver enzymes and low platelet counts) syndrome.^24,25 Thus, pregnancy may precipitate acute exacerbation of TTP in a woman with low baseline ADAMTS13 levels owing to genetic mutations or autoantibodies of ADAMTS13.

THERAPEUTIC IMPLICATION

The molecular mechanism of TTP provides a basis to assess the efficacy of plasmapheresis and other therapeutic modalities. For hereditary TTP, a small amount of plasma, for example, ≤ 15 ml per kg body weight every 2–3 weeks, is sufficient to prevent acute exacerbation. Nevertheless, the regimen of plasma replacement should be tailored not only to prevent acute exacerbation but also to minimize chronic injury to vital organs. Some factor VIII concentrates contain ADAMTS13 activity and may be a potential alternative to fresh frozen plasma for patients with allergy to whole plasma.²⁶

For autoimmune TTP, plasma exchange is necessary to overcome the inhibitors. Plasma-exchange therapy is effective for TTP because most patients have low inhibitor levels.²⁷ The efficacy of corticosteroids remains controversial. Cryoprecipitate-depleted plasma and fresh frozen plasma contain similar levels of ADAMTS13 activity,²⁶ and are similarly effective for TTP.^28,29 Immunotherapy with rituximab may benefit patients with persistent autoimmune TTP,^30,31 but its role for patients with acute TTP is not clear.

Future investigation should define the role of immunomodulation in the management of autoimmune TTP, and develop non-suppressible variants of ADAMTS13 or antibody blockers that may be used to bypass or suppress the autoimmune inhibitors.

RENAL ABNORMALITIES IN TTP

Conflicting data exist for the frequency and severity of renal failure in TTP. An earlier review of TTP suggested that renal dysfunction was common, with creatinine > 176.8 µmol/l in 45% of the cases, and that dialysis therapy was necessary in 12% of the patients.³² Nevertheless, the definition of TTP varied, and it is likely that not all of the patients in the reviewed series had TTP as defined currently. In a recent series of TTP with demonstrated ADAMTS13 deficiency, serious renal failure is infrequent (Table 2). Current evidence suggests that advanced renal failure, hypertension, fluid overload, and need of dialysis therapy are uncommon in autoimmune TTP.

Table 2.

Renal failure in autoimmune and hereditary TTP

Autoimmune TTP	No. of cases	Cases with Cr > 265.2 mmol/l	Cases requiring dialysis
Literature33–36	115	3 (2.6%, range 0–11%)	4 (3.5%, range 0–10%)
Author’s series
Non-referral	38	0 (0%)	0 (0%)
Referral	300	2 (0.75%)a	0 (0%)
Total	453	5 (1.1%)	4 (0.9%)
Hereditary TTP	No. of cases	Acute renal failureb	Chronic renal failure
Literature37–42	68	7	8 (dialysis: 6)
Author’s series43	25	5	1 (dialysis: 0)
Total	93	12 (13%)	9 (9.7%)

TTP, thrombotic thrombocytopenic purpura.

^aOne case developed graft dysfunction and TTP immediately after renal allograft transplantation;⁴⁴ the other case was found to have crescenteric nephropathy because of anti-glomerular basement membrane antibody 11 months after his TTP, which was complicated with acute renal failure.⁴⁵

^bRequiring dialysis therapy.

In contrast, acute or chronic renal failure is not infrequent with hereditary TTP (Table 2). ADAMTS13 is synthesized primarily in the hepatic stellate cells^46,47 but also in glomerular podocytes and endothelial cells.^48–50 Thus, patients with hereditary TTP may have higher risk of renal failure because they lack the protection provided by locally secreted ADAMTS13. Mutation in complement factor H has been described in a patient with ADAMTS13 mutations and renal failure.³⁷

Plasma therapy reverses acute renal failure caused by genetic ADAMTS13 deficiency, and is also effective for preventing the development of chronic renal failure. The high prevalence rate of renal dysfunction among the hereditary TTP patients reflects the current practice, favoring no or minimal maintenance therapy. A strategy of periodic monitoring of blood counts, urinalysis, and renal function, with maintenance plasma therapy individually tailored to ameliorate thrombocytopenia, hematuria, proteinuria, or deteriorating renal function, may provide the best strategy of preventing acute and chronic renal failures for patients with hereditary TTP.