Rituximab Treatment for Thrombotic Thrombocytopenic Purpura Associated with Human Immunodeficiency Virus Failing Extensive Treatment with Plasma Exchange: A Report of Two Cases

Introduction

T hrombotic microangiopathies (TMA) comprise a spectrum of diseases including thrombotic thrombocytopenia purpura (TTP) and the hemolytic uremic syndrome (HUS) as well as pregnancy-, drug- and transplant-related microangiopathic hemolytic anemias (MAHAs). These clinico-pathologic entities have in common thrombocytopenia and MAHA as evidenced by a marked elevation in the lactate dehydrogenase (LDH) and schistocytosis. An elevation in the LDH to greater than 600 U/L and two or more schistocytes per high-power field are frequently seen. Neurologic abnormalities and renal failure are frequently associated with TTP and HUS, respectively, but the presence of these findings is quite variable and not required for diagnosis. ¹ A deficiency in the von Willebrand factor (vWF) cleaving protein, ADAMTS-13, has been implicated in the development of TTP, ² but its role is not completely understood. Deficiency of ADAMTS-13 may be secondary to an inherited genetic abnormality or an acquired inhibitor. ¹ The development of TTP involves deficiencies in complement and ADAMTS-13 resulting in vWF- and platelet-rich microthrombi.

The Oklahoma series of 362 consecutive TTP patients revealed HIV infection in 6 individuals. ³ One had antibody-mediated ADAMTS-13 deficiency; 5 patients were found to have other causes of TMA indicating that other mechanisms must be involved. Levels of ADAMTS-13 are not routinely determined as assays for this enzyme are not standardized, and thus its routine use is not recommended. ¹ Ongoing HIV infection likely contributes a high rate of TMA-like syndromes by causing endothelial injury via cytokine-induced apoptosis and/or depletion of complement and ADAMTS-13. ^{4 –6}

HIV is endemic in Baltimore where more than 3% of the population is infected. Forty-two percent of these infections are thought to have been acquired thru injection drug use (IDU). ⁷ TTP occurs in HIV-positive individuals but has decreased in the highly active antiretroviral therapy (HAART) era, presumably due to fewer patients with long-lasting CD4 T-lymphocyte depletion. ⁸ Unfortunately, the utilization of HAART therapy varies by region and is often influenced by socioeconomic factors. The incidence of severe ADAMTS-13 deficiency was greater in women and those of African decent in an analysis of the Oklahoma TTP-HUS registry. ⁹ Analysis of the same registry also found that less than 2% of TTP patients were infected with HIV. ³ In the last 9 years, 100 patients have been referred to our apheresis service for treatment of TMA of all causes. Of these patients 20 were HIV infected, the majority were African American women.

In the literature, the extent and nature of the relationship of HIV to TTP is controversial. The largest series come from South Africa where 15% of the population and 70% of hospital patients are HIV infected. In a comparison of TTP patients with and without HIV infection, Novitzky and colleagues ¹⁰ in Capetown found that patients with HIV associated TMA (n = 21) had a 95% response rate to plasma infusion and prednisone treatment. In contrast, only 56% of HIV-negative patients (n = 23) had a complete response to this treatment, but often did respond to subsequent plasma exchange. This pattern of response suggests a depleted plasma factor and an immune mechanism as well as disease heterogeneity.

Prior to the institution of plasma therapy, TTP followed a rapidly fatal course in the majority of cases. As noted, Novitzky and colleagues ¹⁰ established that HIV-associated TTP responds to plasma infusion. In their experience, plasma exchange is useful for patients not responding to plasma infusion. Plasma exchange is likely more effective in these refractory cases because it not only replaces plasma components that may be deficient, but also removes plasma proteins that may contribute to the microangiopathic process, including autoantibodies to ADAMTS-13. TTP in the United States is almost universally treated with plasma exchange. When plasma exchange is judged to have failed, rituximab is increasingly used as second-line therapy, as this treatment modality adds the effect of depleting the active plasma cell population responsible for autoantibodies against ADAMTS-13. The role of rituximab in HIV-infected patients with suspected TTP has not been explored. Its use in HIV-infected patients with lymphoma has been associated with an increased risk of infectious death when used in combination with chemotherapy, ¹¹ but rituximab monotherapy has been used safely in patients with HIV-associated Castleman disease. ¹² We report the successful use of rituximab in two HIV-positive African American women with plasma exchange-responsive TTP whose symptoms returned when treatment was stopped.

Methods

We reviewed an existing Institutional Review Board-approved database of all cases of suspected TTP treated with plasma exchange between 2000 and 2009 for HIV-positive patients treated with rituximab. TTP was diagnosed in each patient based on the presence of thrombocytopenia, schistocytes, and an elevated LDH activity. Other diseases that can present in a similar fashion, such as infection with disseminated intravascular coagulation or malignant hypertension were excluded. Patients were treated initially with plasma exchange and glucocorticoids until the platelet count was maintained above 150,000/mm³ for 2 consecutive days. Treatment failure was established when repeated treatment with plasma exchange and glucocorticoid was not successful in establishing durable resolution of TTP, defined as at least 4 weeks without recurrence of symptoms and/or clinical findings consistent with this diagnosis. One hundred patients were identified, 20 were HIV infected at the time of diagnosis, and 4 were treated with rituximab after obtaining informed consent. One patient presented with multiorgan failure and died on the day of rituximab treatment and was excluded from our review. A second case of presumed TTP failing plasma exchange was treated effectively with a single dose of rituximab. This patient had an atypical presentation, with only a modest increase in LDH activity, and subsequently developed Hodgkin's lymphoma. This patient was also excluded from our review. The medical records of the remaining 2 cases were examined in detail. Two patients presented a similar syndrome of repeated responsiveness to plasma exchange without inducing remission.

Cases

For details regarding HIV disease characteristics and manifestations of TTP at diagnosis, please refer to Table 1 and 2.

Case 1 was a 39-year-old woman presenting in 2008 with new TTP who received daily plasma exchange for 6 courses of rapidly responding but promptly relapsing platelet counts (41 total plasma exchanges over 3 months) before she received rituximab (Fig. 1). She was begun on HAART 1 month into this course. Following her first dose of rituximab thrombocytopenia improved and she has remained well on HIV treatment for 21 months.

OPEN IN VIEWER

FIG. 1.

Graphs showing the platelet counts in our two patients with HIV associated thrombotic thrombocytopenic purpura (TTP) as they responded to repeated cycles of plasma exchange but promptly failed when plasma exchange was stopped until each was treated with rituximab (dark arrows). The initiation of highly active antiretroviral therapy (striped arrow) was not associated with stabilization of the platelet count.

Case 2 was a 50-year-old woman presenting in January 2009 who underwent four courses of daily plasma exchange with a good response in her platelet count but prompt relapse with discontinuation. She had been on HAART intermittently for greater than seven years. ADAMTS-13 levels were less than 5% prior to treatment with rituximab. Following her second dose of rituximab, she has remained well for 9 months with out TTP relapse and continued HIV treatment.

Discussion

We describe two patients with HIV-associated TTP who consistently responded to plasma exchange but whose symptoms returned when it was stopped. Patient 1 was diagnosed with HIV at the time of TTP presentation. Patient 2 was stable on intermittent antiretroviral therapy for over 6 years prior to developing TTP. Immune reconstitution has been described as a cause of TTP in HIV-infected patients undergoing HAART ¹³ but did not appear to be a factor in our patients.

Following multiple courses of plasma exchange with prompt but not durable responses, rituximab resulted in long-lasting remissions for many months in both patients (Fig. 1). There was no increase in infectious complications following rituximab treatment. Both patients were treated with HAART. Although the institution of HAART therapy could be playing a role in the patients' recoveries, the time course is highly suggestive of a key role for rituximab.

The pathophysiology of TTP in HIV-infected patients remains poorly understood. Many patients have no apparent underlying conditions or precipitating events. From others, a history of an antecedent infectious or inflammatory condition is obtained. TTP has been reported after pneumonia, ^14,15 cardiovascular surgery, ¹⁶ pancreatitis, ¹⁷ and patients with HIV infection. ^2,4,5 Inflammation in the form of systemic infection may cause direct endothelial damage, and may provoke a misdirected immune response resulting in autoantibodies to ADAMTS-13. ¹⁸ HIV disease is associated with varying degrees of immune system suppression and dysregulation, as well as frequent infectious complications. HIV-infected patients are likely to be at an increased risk for the development of TTP via direct endothelial damage, an immune mediated mechanism or a combination of both. It is likely that HIV infected patients with a TMA have multiple etiologic factors resulting in the characteristic triad of an elevated LDH, thrombocytopenia, and microangiopathic hemolytic anemia.

Benjamin and colleagues ⁴ suggested that AIDS-related diseases may mimic TTP and that a thorough investigation should be undertaken prior to starting treatment with plasma exchange. Their study also suggested that ADAMTS-13 levels and the presence of ADAMTS-13 inhibitors were not reliable tests for the diagnosis of TTP in AIDS patients. However, because of the high mortality rate of untreated TTP and the profound benefit of plasma exchange, it is common practice to initiate treatment with plasma exchange promptly on the basis of an elevated LDH, evidence of a MAHA, and thrombocytopenia. AIDS patients with TTP-like illness appear to benefit in the short-term from plasma exchange despite the known complications. ¹⁹

The treatment approach to HIV-infected patients with a TMA is complicated by the urgency of the clinical situation, uncertainty regarding a definitive diagnosis and concerns regarding infection in these immunosuppressed individuals. The observed course of our two patients, in agreement with other reports in the literature, indicates that HIV-associated TMA may respond rapidly to plasma exchange, but the response can be short lived. Treatment of these patients with rituximab was effective, not associated with infectious complications, and induced long-lasting remissions.