Although primarily developed for treatment of B-cell lymphoproliferative disorders, the use of rituximab has been increasingly reported in many other, mainly immune-mediated, conditions. Autoimmune-mediated hematological disorders encompass a broad spectrum of disorders, characterized by an immune attack on self-antigens. Of these disorders, immune-mediated thrombocytopenia, autoimmune hemolytic anemia (AIHA), and thrombotic thrombocytopenic purpura (TTP) have been studied most extensively, although others may also cause serious clinical problems. Despite major progress during the last decades, the pathogenesis of immune-mediated disorders remains incompletely understood. The role of T-cells in autoimmune cytopenias has been recognized for some time, but it appears that B cells also play a major role not only by being responsible for the production of autoantibodies but also by production of cytokines and by acting as antigen-presenting cells (APCs) supporting the activation of autoreactive cells. This major contribution of B cells makes them an attractive target in the treatment of immune-mediated disorders and provides a rationale for the use of B-cell depleting therapy in these disorders. Although rituximab has been used off-label in most immune-mediated disorders, the drug has been licensed worldwide for the treatment of rheumatoid arthritis. Here, we describe the currently available evidence on the use of rituximab in adult immune-mediated hematological disorders, in addition to potential side effects, which might hamper the initial enthusiasm on its use in these conditions.
Mechanism of Action of Rituximab
Rituximab is a genetically engineered chimeric mouse/human monoclonal antibody against CD20, which is expressed on the surface of premature and mature B-lymphocytes, but not on pro-B cells or plasma cells . The mechanisms of action of rituximab in lymphoproliferative disorders have been extensively reported and include antibody-dependent cell-mediated cytotoxicity (ADCC), complement-dependent cytotoxicity (CDC), direct apoptotic activity, disturbed T-cell activation, and possible vaccinal effects .
Although the existence of autoimmune human diseases has been recognized for several decades, the pathogenic mechanism of autoimmune pathology remains poorly understood. Generalized immune system dysfunction seems to play a major role, leading to failure of self-tolerance. The exact mechanism of this specific loss of immunologic tolerance to self-antigens remains unknown. Although the role of T cells has been established for some time, B cells also play a major role in immune-mediated disorders. Recent findings have highlighted the major role of B cells by producing and secreting autoantibodies, producing inflammatory cytokines, and acting as APCs, supporting activation of T cells . As a consequence, depletion of B cells by rituximab was introduced as a new treatment option in immune-mediated disorders. However, as this depletion occurs in all patients treated with rituximab irrespectively of their response—though patients with immune thrombocytopenic purpura (ITP) not responding seems to have higher B-cell levels when compared with responders—other mechanisms of action may play an important role as well .
The production of autoantibodies by autoreactive B cells is regulated by cellular mechanisms, involving mainly responses to signaling from CD4 positive T lymphocytes [T-helper (Th) cells]. Autoimmune disorders are characterized by imbalances between Th1 and Th2 subsets, in favor of Th1 cells. Recently published data from patients with ITP have shown that this pattern is skewed toward an increase in Th2 cells during remission of the disorder, suggesting that active disease may be predominantly caused by Th1 activation [5, 6]. Stasi et al.  showed that reversal of this Th1/Th2 imbalance was also seen in patients responding to rituximab, whereas the Th1/Th2 ratio remained unchanged in nonresponders.
In addition to the important role of Th cells, CD4+ CD25+Foxp3+ regulatory T cells (Tregs) have recently emerged as natural immune master regulators, playing a major role in the maintenance of peripheral tolerance . Deficiency in generation and/or defective functions of Tregs may contribute to loss of immunologic self-tolerance in autoimmune disorders by failure to suppress autoreactive T cells and B cells, which results in continued autoantibody production. Stasi et al.  reported their findings in rituximab-treated patients with ITP, particularly responders, who showed restored numbers of Tregs as well as restored regulatory functions when compared with pretreatment findings. Finally, rituximab also induces downregulation of CD40 and CD80 on B cells, leading to further disturbed T-cell activation [10, 11]. For a review on the effects of rituximab on the different immunological cells, we refer to a recently published article by Cooper and Arnold .
Although in most patients a response occurs 3–8 weeks after the first infusion of rituximab, an interesting and frequently reported observation is the very rapid response following rituximab administration in a small proportion of patients, even if low doses of rituximab are given [13–15]. It has been suggested that such prompt responses could result from macrophage Fc-receptor blockade by rituximab-opsonized B cells, referred to as immune complex decoy hypothesis . The same investigators were able to show that low-dose rituximab can promote rapid clearance and destruction of chronic lymphocytic leukemia (CLL) cells by the mononuclear phagocytic system in liver and spleen . However, as only a minority of patients experience a response in the first week following first rituximab administration, the significance of this mechanism may be limited.
Taken together it is clear that rituximab interferes with both B cell and T cell participation in antibody production (Fig. 1).
Despite the high number of responses to rituximab in immune-mediated disorders, a substantial proportion of patients fail to respond. The mechanisms of this failure are largely unknown and may be heterogeneous. Possible explanations include insufficient B-cell depletion, especially in lymphoid organs outside the peripheral blood, the presence of FcγR polymorphisms, and evolution to B-cell-independent T-cell expansion.
Experience with Rituximab in Immune-Mediated Hematological Disorders
Here, we present a narrative review of published material on the use of rituximab in immune-mediated hematological disorders. PubMed was searched for articles in English using the keywords “rituximab” and each immune-mediated disorder from 2000 to November 1, 2010. Specific limitations in including manuscripts with regard to number of patients are given in the text. Each coauthor contributed for a pre-established part of the article with the whole manuscript being reviewed by all coauthors. The relevance of each article was weighted by each author for his own part of the article with priorities given to randomized trials and studies published as full papers.
Immune thrombocytopenic purpura
ITP is characterized by the presence of IgG or, less frequently, IgM autoantibodies against platelet glycoproteins, leading to platelet destruction in the splenic mononuclear phagocytic system. However, antibodies are not detectable in up to 50% of the patients [18, 19]. Besides immune platelet destruction, it has been shown that the rate of thrombopoiesis is inadequate to compensate for the increased platelet destruction in ITP, making thrombopoietic growth factor therapy an attractive treatment . ITP can be divided into primary ITP, indicating the absence of any obvious initiating and/or underlying cause, and secondary ITP, including all forms of immune-mediated thrombocytopenias except primary ITP .
First-line treatment for ITP comprises oral corticosteroids, mainly prednisone or dexamethasone. Although initially durable responses were seen in less than 30% of adult patients, the use of high-dose dexamethasone has led to sustained responses in about 70% [22, 23]. In cases of failure or relapse, splenectomy is traditionally considered as second-line therapy, leading to durable responses in about two-thirds of the patients. In refractory or relapsing patients, immunosuppressive therapy has shown variable responses [24, 25]. Recently, the second-generation thrombopoietin receptor agonists romiplostim and eltrombopag have shown impressive activity in both splenectomized and nonsplenectomized patients with ITP [26, 27].
When compared with many other immune-mediated disorders, the use of rituximab in ITP has been reported extensively. However, most available evidence is based on retrospective case reports and case series, with only few uncontrolled prospective trials, and until recently, only one controlled prospective trial . Another major limitation of these trials, as recently reviewed by Ruggeri et al. , is the impressive heterogeneity in criteria used for reporting responses and clinical outcomes. Despite these drawbacks, rituximab appears to have efficacy in the treatment of ITP with overall response rate (ORR) and complete response rate (CRR) exceeding 60% and 40%, respectively. Table I provides an overview of all publications regarding the use of rituximab in adult ITP. Papers in abstract form, not written in English language, and publications describing less than five patients are not included in this table [4, 14, 15, 29–50].
AIHA is characterized by the formation of autoantibodies against own red blood cell (RBC) antigens, leading to hemolysis; AIHA can be classified according to the characteristic temperature reactivity of the RBC autoantibody in warm-antibody AIHA (reacting at 37°C) and cold-antibody AIHA (reacting optimally at lower temperatures). Both types of AIHA can be developed idiopathically (“primary” AIHA) as well as in association with a wide range of underlying conditions (“secondary” AIHA), including lymphoproliferative disorders, systemic autoimmune diseases, infections, medication, and immunodeficiency syndromes. The degree of hemolysis largely depends on characteristics of the bound antibody. In warm-antibody AIHA, IgG antibodies opsonize RBCs and are predominantly cleared in the spleen being recognized by mononuclear phagocytic cell Fc receptors (extravascular hemolysis). As complement activation by IgG requires the presence of a dimer of IgG molecules on the RBC surface, IgG antibodies are poor activators of the classical complement pathway. As macrophages also have receptors for complement factors C3b, this can potentiate the extravascular hemolysis. In cold-antibody-mediated AIHA, however, most antibodies are of the IgM type, being pentamers and fixing complement in the bloodstream, leading to activation of the complement cascade with pore formation and cytolysis of the RBCs (intravascular hemolysis). Because of the complement fixation, part of IgM-associated hemolysis is extravascular, the Kuppfer cells in the liver being the principal effector cells. Occasionally, patients may have a combination of warm and cold autoantibodies and are classified as mixed-type AIHA [51, 52]. Standard management of warm-antibody AIHA consists of corticosteroid therapy, leading to a high initial response rate, but only to 20–35% permanent remissions. In case of failure or relapse, splenectomy can be performed leading to CRRs of about 50% in warm-antibody-type AIHA. If these conventional therapies fail or are not tolerated, immunosuppressive regimens can be used . In primary chronic cold agglutinin disease (CAD), steroids and splenectomy are most often not effective, and immunosuppressive drugs are usually used upfront if therapy is needed .
Experience on the use of rituximab in AIHA has been mostly limited to smaller series. The ORR in the largest series of the literature and especially in warm-type AIHA was beyond 60%. Especially in primary CAD, use of rituximab seems very promising, taking into consideration that no real effective therapy is available for this subtype of AIHA. However, CRs are rare [53–56]. To improve response rates and duration, Berentsen et al. recently performed a prospective trial combining rituximab and fludarabine, a purine analog used in several lymphoproliferative disorders, in patients with primary CAD. ORR was 76% with 21% of the patients achieving a CR. Interestingly, response duration was estimated to be more than 66 months and responses were seen as well in patients not responding to rituximab monotherapy before. In contrast to the favorable toxicity profile of rituximab, combination therapy was associated with a higher frequency (41%) of Grades III–IV hematologic toxicity .
The combined use of rituximab and chemotherapy might be an attractive treatment option in AIHA associated with lymphoproliferative disorders, especially CLL, as these more aggressive rituximab-containing combinations may be superior to standard immune-mediated disorder treatment because of their efficacy in treating the underlying disorder. Gupta et al.  investigated the use of immunochemotherapy in CLL-based AIHA. Retreatment with RCD (rituximab, cyclophosphamide, and dexamethasone) in five of eight patients was effective in achieving a new response on relapse of AIHA. More recently, Kaufman et al.  reported on the treatment of 20 patients with CLL-associated AIHA using the same RCD-regimen. In this series ORR was 100% with a median duration of response of 22 months. In another recently published paper, 17 patients with CLL-associated AIHA were treated with R-CVP (rituximab, cyclophosphamide, vincristine, prednisone), leading to similar responses .
Table II provides an overview of publications regarding the use of rituximab in adult AIHA. Papers in abstract form, not written in English language, and publications describing less than five patients are not included in this table [33, 35, 48, 54–69].
Evans' syndrome (ES) is an uncommon disorder, defined by the simultaneous or sequential development of ITP and AIHA, without a known underlying condition (primary or idiopathic ES) . However, analogous to ITP and AIHA, recent evidence reveals the existence of “secondary” ES, associated with other disorders such as systemic lupus erythematosus, lymphoproliferative disorders, and primary immunodeficiencies . The management of ES is primarily based on treatment recommendations in AIHA and ITP. The use of rituximab in ES is, again, described in case reports and small series, including less than 20 patients, with success rates exceeding 80%. However, Michel et al. recently published a large retrospective series of 68 adult patients with diagnosis of ES. Eleven patients, of whom five had undergone prior splenectomy, had been treated with rituximab, leading to an initial response rate of 82% and long-term response rates in 64% of the patients . No other reports describing five or more patients have been published to date.
Thrombotic microangiopathy (TMA) is a life-threatening disorder, characterized by microangiopathic hemolytic anemia, thrombocytopenia, and formation of hyaline microthrombi in different organs [72, 73]. Several different syndromes can be categorized as TMA according to their similar histological and biochemical properties (Table III) [74, 75].
Table III. Classification of Thrombotic Microangiopathy
• Autoimmune disorders (antiphospholipid syndrome, systemic l upus erythematosus, etc.)
Acquired ADAMTS13-deficient TTP is a typical autoimmune disorder, which is caused by the formation of autoantibodies against ADAMTS13. This enzyme is responsible for cleavage of ultralarge von Willebrand factor (ULvWF) multimers. Because of these antibodies, ULvWF multimers are not, or insufficiently, cleaved in patients with TTP, leading to profound platelet consumption, fragmentation of RBCs, and occlusion of small blood vessels in various organs [72, 76].
Daily therapeutic plasma exchange (TPE) with fresh frozen plasma is the only treatment that has shown superior survival in a randomized fashion . However, treatment with rituximab is a promising approach in immune-mediated TMA, both as an adjunctive treatment to TPE as in case of relapse . Currently, the experience with rituximab is limited to a few smaller case series and case reports on relapsing and refractory cases of TTP. Taken together, rituximab therapy seems to provide most benefit in patients with severe ADAMTS13 deficiency due to ADAMTS13 autoantibodies, which seems to be associated with a significantly higher relapse rate when compared with patients with nonseverely deficient ADAMTS13 activity levels . Currently, as shown in Table IV, more than 130 patients have been reported, with severe ADAMTS13 deficiency being documented in almost 85% of these patients. If mentioned, response was in most cases defined as complete hematological and clinical remission [80–102].
In most of these reports, rituximab was used at the moment of relapse in adjunction with TPE, reaching response rates as high as 80–100%. Jasti et al. reported on 12 TPE-refractory patients treated with rituximab in adjunction to TPE, leading to a complete clinical response in 83% of the patients. Unfortunately, data on ADAMTS13 were only available in half of the patients . More detailed information on ADAMTS13 was given by Ling et al. who treated 13 patients with ADAMTS13 deficiency (12 with severe deficiency) with rituximab, leading to durable CRRs in 92% of the patients. Besides clinical remissions, most responses were also associated with a decrease in ADAMTS13 antibodies and an increase in plasmatic ADAMTS13 activity . The largest series till now has been published by Scully et al., who described 25 patients with relapsing or refractory TTP, treated with conventional rituximab doses in addition to TPE. All patients showed a complete remission of their disorder, which occurred within 11 days following rituximab administration. This study confirmed the inverse relationship between successful rituximab therapy and decreased antibody levels and increased ADAMTS13 activity . In a recent publication of the regional United Kingdom TTP Registry, the number of patients having received adjuvant rituximab therapy in the period 2004–2006 was significantly increased in comparison with other adjuvant therapies. Of the 48 patients treated with rituximab, only one patient relapsed . Besides the use of rituximab as adjunctive therapy, there are a few reports on the successful use of rituximab as a monotherapy in patients with relapsing TTP [84, 86]. In an effort to better understand the therapeutic benefit of rituximab in relapsed/refractory TTP, the Canadian Apheresis Group recently initiated a Phase II trial incorporating anti-CD20 therapy together with standard TPE .
Rituximab may also be used prophylactically in patients with relapsing TTP. In a multicentric open-label prospective trial, Fakhouri et al. treated five patients with severe relapsing TTP and persistent ADAMTS13 autoantibodies during a period of clinical remission. In all patients, the antibodies disappeared with appearance of significant ADAMTS13 activity following rituximab administration . In another small case series, Bresin et al. recently reported on five patients with TTP treated pre-emptively with rituximab based on persistent unmeasarable ADAMTS13 activity and high autoantibody titers. In three of these patients, durable responses were noticed, whereas the other two patients eventually relapsed despite initial improvement of ADAMTS13 activity and disappearance of autoantibodies .
These promising results have raised the question whether rituximab should be used upfront together with TPE as standard treatment for TTP. In 2006, George et al. started up a study on this specific subject, which will help to identify the place of rituximab in first-line therapy of TTP . At the last American Society of Hematology Meeting (2009), the initial results of a French prospective multicentric open-label Phase II study using first-line rituximab in patients with idiopathic TTP experiencing a nonoptimal result to TPE were presented. When compared with a historical TPE- ± vincristine-treated population, treatment with rituximab prevented relapses during the first year following initial treatment, but not long-term relapses . As rituximab may be removed by subsequent plasmapheresis, several centers advocate to postpone the next TPE for at least 24 hr following rituximab administration. However, Froissart et al.  were able to show a profound and sustained peripheral blood B-cell depletion despite associated daily TPE.
Pure red cell aplasia
Pure red cell aplasia (PRCA) is an immune-mediated disorder characterized by an isolated depletion of erythroid precursors from the bone marrow, leading to severe normocytic anemia and reticulocytopenia. PRCA can be either primary or secondary to parvovirus B19 infection, hematological malignancies, immune disorders (especially systemic lupus erythematosus), thymoma, solid tumors, treatment with recombinant human erythropoietin or other drugs, pregnancy, and ABO-incompatible hematopoietic stem cell transplantation. Initial therapy, if not self-limiting, includes classical immunosuppressive medication including steroids, cyclosporine A, cyclophosphamide, and intravenous immunoglobulins . Evidence on the use of rituximab in PRCA is limited to several case reports, showing both successes and failures. By reviewing the literature, 35 cases were identified. Rituximab was mostly given in 4 weekly doses of 375 mg/m2, although both lower and higher doses have been described. Dungarwalla et al. reported on three cases with severe therapy-resistant PCRA. In none of these patients were responses observed . The authors concluded that the role of T cells in the pathogenesis may be more important than in other immune-mediated disorders, explaining the poor response to rituximab when compared with the use of alemtuzumab in PRCA [109, 110].
Of note, parvovirus B19-induced PRCA has also been described as a complication of rituximab therapy, especially when combined with chemotherapy [111, 112].
Autoimmune neutropenia (AIN) encompasses a heterogeneous group of disorders, characterized by autoantibodies against neutrophils, leading to their destruction. As in most other immune-mediated hematological disorders, the disease can be classified as either “primary” or “secondary” to many underlying disorders, including systemic disorders, malignancies, infections, and drug exposure . Initial treatment consists of standard immunosuppressive therapy or granulocyte colony-stimulating factor . Only a few case reports with regard to the use of rituximab in AIN have been described [14, 108, 115]. Dunjarwalla et al. reported on six cases with severe AIN treated with rituximab in their pilot trial. Only one patient showed a CR with response duration of more than 20 months, whereas all other patients failed to respond to anti-CD20 therapy . Interestingly, as will be discussed later, rituximab is one of the causes of drug-induced (late onset) neutropenia, potentially leading to severe and life-threatening infectious complications.
Graft-versus-host disease (GvHD) is probably the most challenging and troublesome consequence of allogeneic hematopoietic stem cell transplantation (allo-HSCT) and is associated with significant morbidity and mortality . Systemic corticosteroid therapy with or without additional immunosuppressive agents remains the first-line therapy for GvHD in both its acute and chronic forms. Long-term corticosteroid use is, however, limited by an increased risk of infection, contributing to GvHD-related mortality. Additionally, there is no agreement in regards to the best option of treatment for those patients who have failed to respond to first-line therapy .
Several case reports and case series as well as a number of prospective studies have evaluated the use of rituximab in the treatment of chronic GvHD (Table V), especially in those patients refractory to corticosteroid therapy. ORRs between 43 and 83% were observed in those studies containing more than five patients [118, 119, 122, 124, 126–128]. CRRs were universally low. Organ responses were variable amongst the different studies; however, better responses were achieved in those patients with skin, musculoskeletal, and hematological manifestations [117–128]. Ocular manifestations were consistently poorly responsive to rituximab therapy [121, 122, 124, 127]. A recent meta-analysis of three prospective studies and four retrospective studies (containing more than five patients) reported ORR of 60%, 36%, 29%, 31%, and 30% for skin, oral, hepatic, gastrointestinal, and lung GvHD manifestations, respectively . Data concerning time-to-response are quite variable ranging from 7 to 138 days with responses to skin and hematological GvHD occurring earlier [117, 118, 120, 123–125]. Kamble et al.  also reported complete and early responses in three patients with acute GvHD treated with rituximab.
Table V. Responses with Rituximab in GvHD
Number of patients (n)
Treated with CS (%)
Treated with other ISAs(%)
Abbreviations: ref, reference; CS, corticosteroids; ISAs, immune-suppressive agents; ORR, overall response rate; TTR, time to response; NR, not reported; ITP, immune thrombocytopenia; PRCA, pure red cell aplasia; AIHA, autoimmune hemolytic anemia; MSK, musculoskeletal; TTP, thrombotic thrombocytopenic purpura.
The dose of rituximab used in almost all the studies was 375 mg/m2 given in weekly infusions for four consecutive weeks, although there was great heterogeneity in the number of cycles that patients actually received. Interestingly, von Bonin et al.  reported similar ORR (69%) in 13 patients using rituximab at 50 mg/m2 given weekly for four doses. Teshima et al.  reported a median dose reduction of corticosteroid use of 67%, with none of the patients treated requiring any additional immunosuppressive agents within 1 year of rituximab therapy. In one study, 68% of patients achieved corticosteroid dose reductions of 50% or more . Other studies reported rates of 82–86% [124, 126]. Rituximab-related toxicities were largely infusion reactions as well as some infectious complications (especially pneumonia and viral reactivations), with no deaths directly attributed to rituximab treatment in the larger studies .
There is increasing evidence supporting the role of B cells in the pathogenesis of GvHD, although the full extent of these mechanisms has not been fully elucidated. Early animal studies had shown a lower rate of acute GvHD in B-cell-depleted animals . These were supported in humans by a recent retrospective analysis showing a reduced incidence of acute GvHD in patients who have received rituximab within 6 months of allo-HSCT . Interestingly, there was no difference in the risk of chronic GvHD. Miklos et al.  had previously demonstrated that alloantibodies against at least one H-Y antigen developed in 52% male recipients of female donor hematopoietic stem cells. They also demonstrated that the cumulative incidence of chronic GvHD was significantly higher in the presence of at least one of these alloantibodies (89% vs. 31%). Later, the same group demonstrated durable responses in four patients with chronic GvHD and H-Y antibodies who were treated with rituximab .
Despite the limited number of prospective studies available and the relatively small patient numbers in those studies, rituximab has shown promising results in the treatment of steroid-refractory chronic GvHD. A Phase II trial by the Stanford University and the National Institutes of Health is currently recruiting and hoping to determine the efficacy of two, 4-week courses of rituximab at standard dose as first-line treatment in patients with chronic GvHD (clinicaltrials.gov, NCT00350545). A Canadian group has also recently started recruiting for a randomized trial comparing rituximab in combination with corticosteroids versus standard GvHD therapy (i.e., corticosteroids ± calcineurin inhibitors) as first-line treatment for patients with chronic GvHD (clinicaltrials.gov, NCT01066598).
Acquired hemophilia is a rare but serious bleeding disorder caused by the development of autoantibodies (inhibitors) against circulating clotting factors. By far, the commonest of these autoantibodies is against factor VIII (acquired hemophilia A). Certainly all the literature reviewed here concerning the use of anti-CD20 in the treatment of acquired hemophilia discusses the management of patients with factor VIII inhibitors. A large 2-year prospective study by the UK Haemophilia Centre Doctors' Organisation (UKHCDO) reported an incidence for acquired hemophilia of 1.5 cases per million per year with a median age of 78 years . Almost 10% of patients studied sustained a fatal bleed with an overall mortality rate of 41%.
Although up to one-third of patients may achieve spontaneous remissions, this outcome is unpredictable . Successful treatment is, therefore, dependent on the prompt diagnosis and inhibitor eradication by early initiation of immunosuppressive therapy [135, 136]. Corticosteroids alone or in combination with cyclophosphamide have become the accepted standard of care, the combination showing higher response rates and better inhibitor suppression, but without any significant difference in overall survival or disease-free survival [133, 135–137].
The evidence for the use of rituximab in patients with acquired hemophilia (Table VI) is promising with many patients achieving early and CRs [138–152]. With its good safety profile, it appears particularly attractive in managing patients with postpartum acquired hemophilia [145, 150–152]. Rituximab was given at standard dose (375 mg/m2) of weekly infusions in all the studies reviewed. There was considerable variation to the length of treatment, ranging from 1 to 8 weeks. Stasi et al.  reported a partial and transient response rate to rituximab in their two patients with inhibitor titers above 100 BU/mL. However, by adding pulse intravenous cyclophosphamide to rituximab, they were able to achieve a complete and sustained response in these patients. Aggarwal et al.  proposed an algorithm of treatment using inhibitor titer levels. They suggested the addition of rituximab to standard therapy in those patients with inhibitor titer above 5 BU/mL and in particular to those patients with inhibitor titers above 30 BU/mL. A recent case report shows successful treatment of a patient with acquired hemophilia A and inhibitor titers of 79 BU/mL with a single dose of 100 mg rituximab .
Table VI. Responses with Rituximab in Acquired Hemophilia
Inhibitor titer fell to <1 BU, but factor VIII levels were not reported.
Safety of Rituximab in Autoimmune Disorders
Although rituximab is generally considered well tolerated and safe, severe and life-threatening events have been described. Arnold et al. reported a systematic review of published reports describing the use of rituximab in adult patients with ITP. Among the 29 reports (306 patients) that described toxicities, nine deaths (2.9%) occurred. However, in most cases an association with rituximab could not be confirmed .
The vast majority of adverse events in patients with lymphoma treated with rituximab are infusion related, including fever, chills, rigor, and hypotension, and are in most cases seen during the first infusion . These symptoms are mediated by the release of several inflammatory cytokines, including tumor necrosis factor-alpha, interleukin (IL)-8, and interferon-gamma . Most of the complications can be ameliorated by slowing or temporary interruption of the infusion and with premedication with paracetamol, antihistamines, and steroids. Although not always mentioned in the different reports, the incidence of infusion-related side effects seems to be comparable in immune-mediated disorders. In the systematic review by Arnold et al., 21.6% of the patients showed mild to moderate adverse events, of which 83.3% were infusion related . Although rare, life-threatening symptoms such as bronchospasm, angioedema, hypoxia, and shock have been described.
An important concern with the use of rituximab has been the risk of developing infection. Indeed, the use of rituximab has been associated with several severe or fatal infections, including cytomegalovirus reactivation, Pneumocystis jirovecii pneumonia and parvovirus B19 infection . In a recent meta-analysis, Aksoy et al.  reported a higher incidence of infection and neutropenia in patients with lymphoma during rituximab maintenance therapy.
Since the introduction of rituximab, hepatitis B reactivation and hepatitis B-related complications have been increasingly reported . Pei et al. reported on 115 patients with B-cell lymphoma treated with rituximab-containing therapy. Eight of 10 HBsAg-positive patients without lamivudine prophylaxis, and four of 95 HBsAg-negative patients developed HBV-related hepatitis, with three patients dying from fulminant hepatitis/hepatic failure . In another trial, Yeo et al.  found that among HBsAg-negative/anti-HBc-positive diffuse large B-cell lymphoma patients treated with R-CHOP, 25% developed HBV reactivation. Recently, it was shown that entacavir, another oral nucleoside analog, was more effective than lamivudine in the prevention of hepatitis B reactivation in patients with lymphoma treated with chemotherapy and that immunochemotherapy could be successfully reinitiated after reduction of the viral load by entecavir [162, 163]. Taking into account the fact that HBV reactivation is also a well-known complication of cytotoxic therapy, the real impact of rituximab is not entirely clear. Besides, the incidence of this complication in rituximab-treated patients with immune-mediated disorders is currently unknown. However, in our opinion, hepatitis B screening should always be performed prior to starting rituximab therapy.
Another emerging infectious problem in patients treated with rituximab and other monoclonal antibodies is progressive multifocal leukoencephalopathy (PML). PML is a usual fatal infection caused by reactivation of the JC polyoma virus. After the publication of several case reports on PML following immunomodulatory therapy, an attempt was made to collect as many cases as possible following rituximab therapy. In this way, the Research on Adverse Drug Events and Reports Project reported on 57 cases, mostly heavily pretreated patients with lymphoma. However, in addition, five patients with immune-mediated disorders were described, of which one patient with ITP received only corticosteroids and rituximab without any other immunosuppressive therapy .
However, the role of rituximab in these infectious complications is not always clear, as most of these patients received multiple prior immunosuppressive or cytotoxic treatments. The low risk for severe infections following treatment with rituximab in patients with rheumatoid arthritis was recently established in a French registry study. In this analysis, five severe infections per 100 patient years were observed, of which half occurred within 3 months following rituximab administration. In multivariate analysis, chronic heart and lung disease, extra-articular involvement, and low IgG before start of treatment were associated with increased infection risk. The authors suggested that immunoglobulin levels should be checked before start of treatment and may be taken into account in the decision to start rituximab .
Although rare, cytopenias are a well-known complication of rituximab therapy [166, 167]. Especially late-onset neutropenia (LON) has emerged as an important side effect of this specific immunotherapy. The time period between the administration of rituximab and the occurrence of LON can vary widely, although most cases have been reported between 2 and 12 months following rituximab therapy. Several hypotheses for this phenomenon have been suggested, as reviewed by Ram et al. [157, 168]. Whether late-onset cytopenia is a contraindication for future rituximab administration has not been clarified yet, and further studies, both biological and clinical, are needed to unravel pathogenesis and potential consequences for further treatment.
Rituximab has been associated with several other side effects including intestinal perforation and obstruction, although the exact contribution of the monoclonal antibody remains unknown . However, the most severe and fearful complication seems to be delayed interstitial pneumonitis. Although very rare in most studies or series, Liu et al. reported on 107 patients with lymphoma treated with rituximab-containing immunochemotherapy. Nine patients developed interstitial pneumonitis, which was in most cases reversible after prompt initiation of treatment with corticosteroids . Recently, Bitzan et al. performed a systematic review of pediatric cases with rituximab-associated lung injury. Interestingly, of the 31 reported patients, the underlying disorder was immune mediated in three of the cases (one focal segmental glomerulosclerosis, one GvHD, and one ITP) .
Adverse events reported in randomized studies
A causal relationship between adverse events occurring in patients treated with rituximab and the use of the antibody is confounded by the concomitant use of other immunosuppressive drugs and by symptoms resulting from the underlying disorder, with the exception of infusion-related complications, most of which can be ascribed with confidence to the use of the antibody. Theoretically, randomized studies comparing patients given or not given rituximab could shed some light on the issue of rituximab-induced toxicity, although these studies are generally not calibrated to identify rare adverse events and although the assessment of adverse events in randomized studies is not always optimal, as toxicity evaluation often is a minor objective in such trials. We could identify 23 randomized studies suitable for an analysis of toxicity: five studies comparing rituximab to placebo, and 18 assessing rituximab given in addition to chemotherapy or immunotherapy. Only randomized studies published as full texts, including at least 100 patients, and reporting on toxicity with quantitative data allowing for a comparison between patients given rituximab and control patients were taken into account. Overall, 8,130 patients were included in these studies [23, 171–192]. Six studies, totaling 3,456 patients, were devoted to aggressive malignant lymphoma. No major toxicity resulting from the addition of rituximab to chemotherapy could be evidenced, with the exception of the only study on HIV patients that reported an increased incidence of deaths resulting from severe infections in the rituximab arm . Only one study mentioned an increased incidence of neutropenia after rituximab. Of note, a 5-year follow-up study of late adverse events could show no excess of secondary tumors or of deaths unrelated to the lymphoma in the rituximab arm. Nine studies, totaling 2,427 patients reported on patients with follicular or mantle cell lymphoma. In five trials, neutropenia was more common in patients given with rituximab, while the incidence of infections was increased in only one study. Eight publications, totaling 2,187 patients dealt with nonmalignant conditions, including three devoted to rheumatoid arthritis, two to multiple sclerosis, one to kidney transplantation, one to systemic lupus erythematosus, and one to ITP. No significant excess of toxicity was evidenced in the studies on rheumatoid arthritis and on kidney transplantation, while more serious adverse events, more treatment interruption, and more infections were reported in one study on multiple sclerosis. In the only randomized study on ITP, more adverse events but no more serious adverse events were mentioned in patients given with rituximab , while an excess of neutropenia and of serum sickness was found in the study on patients with systemic lupus erythematosus .
Overall, these randomized studies suggest that the use of rituximab is comparatively safe, even in the long term, with the notable exception of HIV patients. It cannot be excluded that some very uncommon severe complications, like interstitial pneumonitis or PML, that were related to the use of rituximab in nonrandomized studies may have escaped large randomized trials. However, the current evidence suggests that their relationship with the use of rituximab remains unproven and their occurrence is uncommon. Other complications, like hepatitis B reactivation, may have passed unnoticed because more and more clinicians would use preventive measures in patients at risk. Finally, as previously discussed, the severity of some infusion-related reactions should not be minimized as, besides the hazards resulting from the reaction itself, they can lead to discontinuing therapy in conditions clearly benefiting from therapy with anti-CD20 antibodies.
Rituximab versus splenectomy
Currently, rituximab is increasingly being used as second-line splenectomy-avoiding therapy in immune-mediated disorders. Although the use of minimal invasive surgical techniques and the widespread use of vaccination presplenectomy has reduced complications and mortality rates, safety issues remain a matter of concern making many patients and physicians reluctant to undergo or recommend splenectomy [193, 194]. Whether rituximab effectively can be considered an alternative to splenectomy in chronic ITP patients was recently investigated by Godeau et al in a multicentric prospective phase II study including 60 patients. After administration of rituximab 1 and 2 year responses, defined as a platelet count ≥ 50 × 109/L with at least a doubling of the initial value, were observed in 40% and 33% of the patients respectively. The authors concluded that rituximab was an apparently safe and effective splenectomy-avoiding option in some adult patients with chronic ITP . Although in most studies prior splenectomy was not associated with a better response to rituximab, Penalver et al. reported in their retrospective Spanish multicentric study a higher early response rate in nonsplenectomized patients, but no difference in maintained response, in patients with ITP. The authors suggested that an intact spleen might be necessary to achieve an early response . The possible important role of the spleen has recently been confirmed in another retrospective analysis showing that previous splenectomy was associated with a poor response to rituximab . The issue of second-line therapy in ITP has recently been complicated by the advent of the new thrombopoietin receptor agonists romiplostim and eltrombopag. Clearly randomized trials are needed to define the exact place of these different treatment modalities .
Optimal dosing schedule
The optimal rituximab schedule in immune-mediated hematological disorders has not been established. In most reports, dosage was 375 mg/m2 once weekly for four consecutive weeks, just as in early non-Hodgkin's lymphoma studies [196, 197]. However, as the B-cell mass in immune-mediated disorders is expected to be less high, recently, lower doses have been successfully used in the treatment of ITP [14, 15, 44]. Provan et al described 11 patients with immune-mediated cytopenias treated with rituximab 100 mg weekly for 4 consecutive weeks , whereas Zaja et al reported on 48 adult ITP patients treated with the same regimen . In both trials levels of B-cell depletion, duration of B cell depletion and response rates appeared similar to those of patients treated with conventional dosages, although some concern exists with regard to duration of response. One possible hypothesis, as proposed by Zaja, may be the fact that low dose rituximab may cause insufficient extravascular B-cell depletion, especially in the spleen . Recently, Barcellini et al.  showed that patients with warm-type AIHA might benefit as well when low-dose rituximab is associated with standard corticosteroid therapy. On the other hand, Fairweather et al.  described their experience with one or two doses of rituximab 375 mg/m2 in immune-mediated hematological disorders, also showing that responses with an abbreviated course of rituximab seems to be comparable with standard dosages.
As lower doses seems to be associated with similar efficacy and substantially lower cost, further studies on this topic are required. Whether this approach will also lead to a reduction in noninfusion-related complications is currently not clear, although limited data point to a more favorable side effect profile when compared with the standard dose .
Instead of lowering rituximab doses, Hasan et al. performed a prospective randomized pilot trial in patients previously treated with rituximab. Sixteen patients were included, of which seven had responded to prior rituximab therapy, and were randomized to receive rituximab combined with R-CVP or double-dose rituximab (4 × 750 mg/m2 during four consecutive weeks). No responses were seen in patients who were previously refractory to rituximab, whereas patients with prior response showed no increased response rate when compared with retreatment with conventional rituximab dose .
Duration of response and retreatment with rituximab
Response duration in patients with immune-mediated cytopenias treated with rituximab has shown considerable variation ranging from 1 month to more than 8 years, with most patients showing responses exceeding 1 year. However, long-term follow-up is limited and, although rituximab can induce durable responses in a proportion of patients, most of the patients eventually will relapse (Tables I and II). In this regard, Patel et al. reported on their long-term follow-up in patients with ITP treated with rituximab. According to their findings, about 16% of the patients treated with rituximab will show a sustained response lasting longer than 2.5 years with little further relapse up to 5 years following rituximab administration. However, these data have only been published in abstract form and need confirmation in a full paper with complete data .
Retreatment with rituximab has been described as an attractive way to prolong duration of responses. In ITP, response rates after a second course of rituximab approached 50%, which is comparable with those after the first treatment cycle [46, 48, 200]. In the large series by Cooper et al., retreatment was successful in approximately 75% of the patients initially responding to a first course of rituximab . Hasan et al. reported on retreatment with rituximab in 20 patients with ITP who had achieved a response to prior rituximab. Response rates were 75% with 50% of the patients showing a CR, with similar response duration when comparing first and second rituximab course . Similar to ITP, retreatment with rituximab in AIHA was associated with an ORR of more than 80% . In TMA, limited evidence has also shown that retreatment with rituximab can reinduce CRs in patients previously responsive to anti-CD20 therapy [78, 80, 88, 89, 101].
Whether maintenance therapy with rituximab may improve duration of response has not been investigated yet, it might be a promising approach in patients with CR following one course of rituximab.
Pretreatment factors predictive for response
As shown in Tables I and II, in most studies performed until now, no clinical or biochemical parameter was able to predict outcome following rituximab treatment. In AIHA, identification of these characteristics have been poorly investigated [48, 54, 66, 69]. As more rituximab-treated patients with ITP have been described, more data are available in this population. Pretreatment factors associated with better ORRs in ITP include younger age [15, 30, 42, 45], female gender [30, 42], shorter duration of the disease [4, 15, 38, 45], achievement of CR [4, 38], and fewer previous treatments [38, 42]. However, such data were not confirmed in other studies. So, currently no clearly identifiable patient subpopulation has been shown to be predictive for response to treatment with anti-CD20 therapy.
In TMA, as mentioned previously, rituximab therapy appears to provide most benefit in patients with severe autoantibody-induced ADAMTS13 deficiency, although successful cases have been reported in which no ADAMTS13 deficiency was present before administration of rituximab [77, 78, 101, 102].
New anti-CD20 monoclonal antibodies
Although the use of rituximab-containing regimens has clearly improved response rates and outcome in B-cell malignancies, about 50% of the patients eventually will become rituximab resistant. This has led to the development of other (humanized) monoclonal antibodies targeting CD20. Ofatumumab is a completely human anti-CD20 antibody, binding to a different epitope when compared with rituximab. This different binding largely explains its more potent characteristics based on superior CDC . Veltuzumab is another humanized monoclonal antibody using the same framework as the humanized anti-CD22 antibody epratuzumab. Although the three main mechanisms of action seems similar to rituximab, veltuzumab has a slower off rate, probably explaining its superior in vivo activity . Another promising second-generation anti-CD20 antibody is ocrelizumab. This humanized antibody also binds to a different epitope when compared with rituximab, leading to improved ADCC and CDC . Apart from these so-called second generation antibodies, the use of third-generation antibodies have been recently introduced. In addition to humanized IgG, these antibodies are characterized by engineered Fc region leading to increased apoptosis and ADCC . GA101 is an example of this new generation of antibody .
Most of these agents are in Phases II and III clinical development for treatment of malignant lymphoproliferative disorders . Besides, some of them, in particular ofatumumab, veltuzumab, and ocrelizumab, have shown promising results in nonhematological immune-mediated disorders. However, experience in immune-mediated hematological disorders currently is very limited (Table VII). Saleh et al. reported on 25 evaluable patients treated in a Phase I/II trial with low-dose intravenous or subcutaneous veltuzumab. ORRs were 62%, including CR in 29%. Responses occurred in all dose levels (ranging from two administrations of 80 mg to two doses of 320 mg, which has been estimated as fixed dose for the ongoing Phase II part) and with both intravenous and subcutaneous administrations, with an acceptable toxicity profile .
Table VII. New Promising Monoclonal Anti-CD20 Antibodies in Immune-Mediated Hematological Disorders
Rheumatoid arthritis and relapsed remitting multiple sclerosis
• Enhanced CDC
• Similar mechanism to rituximab, but superior in vivo activity
Rheumatoid arthritis, relapsed remitting multiple sclerosis, and systemic lupus erythematosus
• Improved ADCC and CDC
• Improved ADCC
• Enhanced ADCC
• Improved ADCC and apoptosis
Immune-mediated hematological disorders encompass a broad group of hematological conditions characterized by the loss of self-tolerance to a variety of antigens, leading to anemia, neutropenia, and thrombocytopenia, either separately or combined. With the exception of TTP, in which urgent TPE is the treatment of choice, immunosuppressive therapy, mostly corticosteroids, are considered the cornerstone of treatment. However, relapses are common necessitating new and safe therapeutic options. Monoclonal antibodies, in particular anti-CD20 antibodies, may be an important alternative for the conventional therapies of these disorders. Indeed, in most immune-mediated cytopenias, rituximab appears to be a promising agent. However, the exact response rates are unknown as patients not responding are unlikely to be reported and only few prospective trials have been performed till now. In addition, most patients are treated with the standard “lymphoma” regimen comprising 4 weekly administrations of 375 mg/m2. Recent studies, especially in ITP, showed similar response rates and B-cell depletion with lower doses of rituximab. Clearly, further studies are needed to determine the best timing as well as the optimal schedule of rituximab administration in immune-mediated hematological disorders. Initially considered a safe therapy, caution is needed based on recent reports describing, sometimes severe, rituximab-related complications. Although most side effects have been reported in heavily treated patients with lymphoma, patients with immune-mediated disorders experiencing serious complications have also been described. By treating benign, although sometimes potentially life-threatening, immune-mediated disorders, efficacy and toxicity of rituximab treatment should be balanced carefully.