Thrombotic Thrombocytopenic Purpura-Hemolytic Uremic Syndrome

A Newsletter for Physicians

James N. George, M.D.

The Oklahoma Blood Institute and University of Oklahoma

The following text was written for the Education Program Book for the annual meeting of The American Society of Hematology, December, 1998. It provides a concise summary of TTP-HUS and describes our experience and our approach to diagnosis and treatment.

Thrombotic thrombocytopenic purpura (TTP) and the hemolytic-uremic syndrome (HUS) today are different from the disorders as they were originally described. The presentation and diagnosis of TTP have changed because of the availability of effective treatment with plasma exchange during the past 20 years. TTP was initially defined in a review of 271 patients in 1966 by the pentad of clinical features including (1) thrombocytopenia, (2) microangiopathic hemolytic anemia, (3) neurologic symptoms and signs, (4) renal function abnormalities and (5) fever. In the 1966 series, the diagnosis of TTP was supported by pathologic demonstration of the characteristic hyalin thrombi in 93% of patients; 90% of patients died of their disease. Plasma exchange was demonstrated to be effective therapy in the 1970s. With effective therapy, there is now an urgency to make the diagnosis of TTP; urgency for diagnosis requires decreased stringency of diagnostic criteria; less stringent diagnostic criteria result in a broader spectrum of disorders. Currently only thrombocytopenia and microangiopathic hemolytic anemia, without another clinically apparent etiology, are sufficient to establish the diagnosis of TTP and initiate plasma exchange treatment. With the use of this diagnostic dyad rather than the classic pentad of clinical features, the apparent frequency of TTP has increased and the relative frequency of patients with overt neurologic symptoms, renal failure and fever has decreased.

HUS was first described in 1955 in a report on four infants and one 7 year old girl who had acute renal failure and died with renal cortical necrosis. These children would now be considered to have atypical HUS, as the current classic HUS only began in 1983 with the recognition of E. coli 0157:H7 and related serotypes as the etiology for a new disease, HUS following hemorrhagic colitis. This epidemic, post-infectious HUS is a distinct disorder, presenting with a prodrome of bloody diarrhea and occuring primarily in young children. These children typically recover without plasma exchange treatment.

Adults who have clinical features of TTP but in whom acute renal failure is the predominant abnormality have been described as "adult HUS" rather than TTP. However this distinction is never precise and plasma exchange treatment is always initiated upon suspicion of either TTP or HUS in adults, including patients who have a prodrome of bloody diarrhea. Therefore the distinction between TTP and HUS in adults is not relevant for initial management and these patients are best described as a single syndrome, TTP-HUS. However this is a heterogeneous group of disorders with different etiologies, clinical manifestations, and outcomes.

Epidemic HUS, which follows a prodrome of bloody diarrhea caused by shiga toxin-producing bacteria, predominantly E. coli 0157:H7, and occurs predominantly in young children, is distinct. This syndrome in young children has a low mortality and plasma exchange appears to provide no benefit. The remainder of TTP-HUS syndromes occur primarily in adults. In most adult patients, no etiology or association with another disorder is apparent. However in some patients there is a clear relationship to pregnancy, autoimmune disease, bone marrow transplantation, or an adverse reaction to a drug. Although these patients may have very different presentations and different clinical courses, all are assumed to have a disease that is potentially fatal without plasma exchange treatment. Although plasma exchange appears to be dramatically beneficial in most patients who have TTP-HUS without an apparent cause, it may not have the same benefit in patients who have TTP-HUS after bone marrow transplantation or mitomycin C treatment.

The decision for plasma exchange treatment dominates the initial evaluation, and treatment may be initiated in patients who are subsequently diagnosed with an alternative etiology for their disease. Therefore any analysis of patients with TTP-HUS must inevitably include patients in whom, upon follow-up, an alternative diagnosis becomes apparent. Examples include infection due to organisms such as cytomegalovirus, aspergillus, or beta-hemolytic streptococcus; disseminated carcinoma; and heparin-induced thrombocytopenia-thrombosis.

In some patients, such as those with systemic lupus erythematosus and/or the antiphospholipid antibody syndrome and all of the cardinal features of TTP-HUS, the issue of whether TTP-HUS is present as an additional disorder may be impossible to exclude. The practical decision in these patients is to determine if plasma exchange may be of value. Similarly, in women with severe pre-eclampsia, the distinction from TTP-HUS may be impossible and the decision to initiate plasma exchange may be empirical. In these women even the pathology demonstrated by renal biopsies cannot distinguish pre-eclampsia from TTP-HUS.

The discovery of drug-dependent antibodies with reactivity against platelets and other cells as an etiology for TTP-HUS has opened a new spectrum of this disorder. Quinine-induced TTP-HUS was initially described in 1991 and this has become a well-recognized and not rare etiology among our patients. The temporal association with quinine ingestion is dramatic; acute renal failure can be severe and, although complete recovery is the rule, permanent renal damage may occur. The frequency of quinine-induced TTP-HUS raises the concern that adverse reactions to other drugs can also cause this syndrome, but at this time no other case reports have documented other drug-specific anti-platelet antibodies as causing TTP-HUS.

The minimum criteria for an initial diagnosis are thrombocytopenia and microangiopathic hemolytic anemia without a clinically apparent etiology. Most patients have some neurologic symptoms or signs and some degree of renal failure. However even anemia and red cell fragmentation are not constant features. In patients who are followed closely after an initial episode of well documented TTP-HUS, recurrent episodes may be heralded only by asymptomatic thrombocytopenia. Suspicion of TTP-HUS in these patients would never occur without the history of a previous episode. Hemolysis occurs after the development of thrombocytopenia. The observation of fragmented red cells on the peripheral blood smear is important but not essential for the diagnosis. Quantitation of red cell fragments is inaccurate; the degree of hemolysis is best assessed by the increased serum LDH concentration, but the high LDH also reflects ischemic injury to multiple organs.

Plasma exchange is the cornerstone of treatment and the single most important modality. A randomized controlled clinical trial has demonstrated the efficacy of plasma exchange when compared to patients treated initially with only plasma infusion. Some centers always use high doses of glucocorticoid; other centers always use antiplatelet agents such as aspirin. All centers appear to have the same clinical outcomes, suggesting that the important therapeutic benefit is derived from plasma exchange. The efficacy of additional treatments remains uncertain. Plasma exchange is initiated on a daily basis with a single plasma volume exchange. The requirement for a central venous dialysis catheter is associated with substantial morbidity, both from insertion complications and from later infectious complications. The plasma replacement product may be either whole fresh frozen plasma or the supernatant fraction from cryoprecipitate preparations, termed cryosupernatant plasma or cryoprecipitate-poor plasma. Although some have suggested that cryosupernatant plasma has greater efficacy, this has not been demonstrated in a controlled study and many centers use these products interchangeably with no apparent therapeutic difference. Some patients may be initially unresponsive to several days of once daily plasma exchange, or may exacerbate after an initial response while on once daily plasma exchange. These patients typically respond to more intensive treatment with twice daily plasma exchanges.

The duration of plasma exchange required is empirical. An arbitrary parameter for stopping daily plasma exchange is the achievement of a normal platelet count; at this time the LDH level may or may not have returned to normal. Whether it is appropriate to discontinue plasma exchange at this time or rather to "taper" plasma exchange treatments by gradually increasing the interval between exchanges is unknown. However it is a common experience for the TTP-HUS to exacerbate promptly upon diminishing plasma exchange therapy, manifested by recurrent thrombocytopenia and increasing LDH values; this requires the resumption of daily plasma exchange. Our practice is to "taper" plasma exchange for several more treatments after the platelet count is normal. This appears to prevent prompt exacerbations. These exacerbations are distinct from relapses, which are arbitrarily defined as occurring after one month with no treatment and no signs or symptoms of TTP-HUS. The duration of plasma exchange required to achieve a durable remission is extremely variable, ranging from several days to many months of treatment. Whether the duration of plasma exchange affects the long term outcome is unknown. Renal function often requires many months for complete recovery, and whether plasma exchange treatment influences the recovery of renal function is unknown. Whether any initial treatment variable influences the risk for subsequent relapse is also unknown.

Although the mortality of TTP-HUS has improved dramatically since the era prior to plasma exchange, when 90% of patients died, current case series still report 15-30% mortality rates. These deaths are difficult to evaluate because some patients died immediately, before plasma exchange could be initiated, and others died from associated conditions. Older age clearly increases the risk for death. Among patients who initially respond following several days of plasma exchange, mortality is low.

The ability to achieve sustained clinical remissions in patients with TTP-HUS has opened the question of whether these patients are cured, or whether they are at risk for recurrent episodes. An important observation by a Canadian group is that relapses are common: In their series, with followup for 3-10 years on 63 of 72 survivors from an initial cohort of 102 patients, 17 patients (27%) have had 1-4 relapses. The projected 10 year risk of relapse was estimated to be 36%. Relapses among our patients appear to be less common. The mortality with relapsed TTP-HUS is very low because (1) there is no delay in diagnosis, and (2) a patient who has responded well to plasma exchange once predictably responds again. Other than overt relapses, it is unclear whether some patients who have recovered from TTP-HUS may have subtle sequelae. Some patients report the persistence of symptoms such as headache, memory difficulty, and fatigue; others report problems with blood pressure. Whether these are genuine sequelae of TTP-HUS, such as the higher blood pressures, proteinuria and decreased creatinine clearances observed in children who have recovered from epidemic HUS, is unknown.