Infections associated with monoclonal antibody and fusion protein therapy in humans

Abstract

Monoclonal antibodies (mAbs), especially those that interact with immune or hematologic leukocyte membrane targets, have changed the outcome of numerous diseases. However, mAbs can block or reduce immune cells and cytokines, and can lead to increased risk of infection. Some of these risks are predictable and can be explained by their mechanisms of action. Others have been observed only after the mAbs were licensed and used extensively in patients. In this review, we focus on infectious complications that occur upon treatment with mAbs or Fc-containing fusion proteins targeting leukocyte membrane proteins, including CD52, CD20, tumor necrosis factor, VLA4, CD11a and CTLA4. We report their known infectious risks and the recommendations for their use. Although most of these drugs are clinically safe when the indications are respected, we emphasize the need for regular updating of pharmacovigilance data.

Introduction

Monoclonal antibodies (mAbs) have substantially changed the outcome of severe diseases such as rheumatoid arthritis and lymphoma in recent years. These molecules are now frequently used, and some of them have several indications for use in various disorders. The notable feature of mAbs is the specific interaction with an antigen, most of the time an immune or hematologic target. The consequences can be blockade or reduction of effector cell function, depletion of B or T lymphocytes, or inhibition of key intercellular or cytokine interactions; all these mechanisms influence the risk of infection. In some cases, a high risk of infection is expected when the immune target is important for an infectious response, e.g., tumor necrosis factor (TNF). In other cases, the risk of infection was realized after mAbs were licensed and extensively used in patients. Pharmacovigilance is therefore essential for the management of these new drugs. Nevertheless, there have been few reports or recommendations for the uses of therapeutic mAbs.

We review here the known infectious risks and the recommendations for use of the following mAbs and Fc fusion proteins that have been particularly associated with infectious complications in humans in terms of frequency and severity: anti-CD52 alemtuzumab; anti-CD20 rituximab; the TNF-targeting agents infliximab, adalimumab and etanercept; anti VLA4 natalizumab; anti-CD11a efalizumab; and the CTLA4-Ig fusion proteins abatacept and belatacept.

Anti-CD52 Monoclonal Antibody: Alemtuzumab

Alemtuzumab is a humanized mAb (IgG1kappa) distributed with the trade names of CAMPATH^® in the US and MABCAMPATH^® in Europe. This antibody is specific to CD52, which is a 21–28 kDa glycoprotein expressed mainly on normal or pathologic B and T peripheral blood lymphocytes. The antigen is also expressed on monocytes, thymocytes, natural killer (NK) cells and macrophages, but not on erythrocytes or platelets. Alemtuzumab targets normal or pathologic mononuclear cells to destroy them, without affecting stem or progenitor cells. This pathway explains the use of alemtuzumab in chronic lymphoid disease and Hodgkin lymphoma, and also in transplantation and graft versus host disease and multiple sclerosis. The drug can also increase regulatory cells in the immune reconstitution phase, induce regulatory T-cell differentiation and inhibit of T-cell transmigration. (1) Different doses are required for different indications, e.g., for hematologic diseases, doses are much higher to obtain effective malignant cell depletion; for transplantation, alemtuzumab was tested as induction therapy to reduce the use of steroid and other conventional immunosuppressive drugs.

Its action appears to be related to antibody-dependent cell-mediated cytotoxicity (ADCC), complement cytotoxicity² and apoptosis induction,³ which leads to neutropenia and reduction in CD4⁺ and CD8⁺ T cells, as well as B and NK cells. The cell depletion develops early in treatment, persists for up to a year after therapy is discontinued, and explains opportunistic and non-opportunistic infections. Infectious risk is directly linked to different doses of alemtuzumab administered for different indications.

Several authors reported opportunistic infections and also septicemia and pulmonary infections in refractory chronic lymphocytic leukemia patients treated with alemtuzumab.^4,5 Such infections led to recommendations regarding pneumocystosis and herpes infections, resulting in reduction in the rates of opportunistic infections as reported in the Keating et al. multicenter study.⁶

Martin et al. published a retrospective study in 2006 involving 27 refractory chronic lymphocytic leukemia patients, with nine patients treated with alemtuzumab combined with prophylactic treatment against pneumocystosis and herpes virus.⁷ Fifteen patients (56%) developed opportunistic infections and 22 patients (82%) developed non-opportunistic infections; ten patients died, seven from infections. However, although cytomegalovirus (CMV) viremia significantly increased, survival was greater in the alemtuzumab-treated group.⁷

In 2007, Peleg et al. reported a retrospective study involving 547 transplant patients who received alemtuzumab.⁸ Fifty-six patients (10%) developed opportunistic infections, mainly due to CMV, BK virus and Candida; 12 patients died, seven from infections. Patients who received alemtuzumab for induction therapy were significantly less likely to develop opportunistic infections compared with patients who received alemtuzumab for rejection therapy (4.5% vs. 21%; p < 0.001).

In 2010, Reddy et al. compared alemtuzumab induction to rabbit antithymocyte globulin induction in simultaneous kidney and pancreas transplantation.⁹ There was no difference in the rates of CMV infection or BK nephropathy between the two groups.

Because symptomatic CMV infections are probably the most frequently occurring infections with alemtuzumab therapy (estimated at 10–25%), different management guidelines, especially in chronic lymphocytic leukemia, recommend monitoring of CMV reactivation by weekly PCR analysis.^10,11 When CMV reactivation becomes symptomatic or viremia increases, alemtuzumab therapy should be discontinued and anti-CMV therapy started.

Alemtuzumab induces profound T- and B-cell depletion and is used in refractory LLC and organ transplant induction treatment. The risk of opportunistic infection is high, but seems to depend on the dose that is used regarding the different indications. Its use should be evaluated for each patient according to the risk/benefit ratio (Table 1).

Table 1.

Synthetic wee of target molecule, infections, recommendations and references

Target Molecule	International non-proprietary name	Trade name	Structure	Functions	Indication	Type of immune suppression	Overall safety signals	Recommendations	References
CD25	Alemtuzumab	Campath or Mabcampath	Humanized IgG1	ADCC, complement cytotoxicity, apoptosis induction, normal or pathologic mononuclear cells	Approved indication: Chronic lymphoid disease Transplantation and GVH disease Clinical trials: Autoimmune diseases	Persistent neutropenia and decreased T, CD4/8+, B and NK cells	Opportunistic infectious disease	Recommendations for pneumocystosis, herpes and CMV infections	6,8
CD20	Rituximab	Mabthera or Rituxan	Chimeric IgG1	ADCC, complement cytotoxicity, apoptosis induction on B cells	B cell lymphoma, auto-immune disease, rheumatoid arthritis, transplantation	Profound B-cell lymphopenia; some cases of hypogamma-globulinemia	Viral infections	Discussion regarding lamivudine for HBV reactivation, prophylactic treatment for pneumocystis, registry for JC virus, immunoglobulin supplementation if agammaglobulinemia	12, 24, 25
TNF	Infliximab; Adalimumab; Etanercept	Remicade; Humira; Enbrel	Chimeric IgG1; Human IgG1; Human IgG1 Fc fused with extracellular portion of TNF α receptor	Neutralization of soluble TNFα. induction of T cells and macrophages, apoptosis by ADCC or CDC	Inflammatory bowel disease, rheumatoid arthritis, chronic juvenile arthritis	T cells and macrophages lymphopenia	Intracellular infection+++	Screening for tuberculosis Screening for non-tuberculosis mycobacteria infections? HBV prophylaxis if chronic hepatitis	34
VLA-4	Natalizumab	Tysabri	Humanized IgG4	Blocks the binding of VLA-4 on VCAM-1, reduces migration of activated leukocytes through endothelium	Neurological disease, e.g., multiple sclerosis Crohn disease	Reduces tissue inflammation of the intestines and blood-brain barrier	Infection with JC virus leading to progressive multifocal leucoencephalopathy	Forbidden in cases of immune deficiency (HIV+), leucopenia or with other immunosuppressive drugs	42
CD11a	Efalizumab	Raptiva	Humanized IgG1	Blocks the binding of CD11a on ICAM-1, reduces migration of activated leukocytes through endothelium	Psoriasis	Reduces tissue inflammation	Infection with JC virus leading to progressive multifocal leucoencephalopathy	Withdrawn from the market in 2009	42
CTLA-4-Ig	Abatacept; Belatacept	Orencia; Amevive	Fc of IgG1 fused to the extracellular domain of CTLA-4 (abatacept and belatacept differ by 2 amino acids)	Inhibits T cell costimulation	Rheumatoid arthritis, polyarticular juvenile arthritis, graft survival	Blocks T-cell activation	Bacterial or viral infection without opportunistic or tuberculosis infection		50

Abbreviations: ADCC, antibody-dependent cell-mediated cytotoxicity; CD, cluster of differentiation; CDC, complement-mediated cytotoxicity; CTLA, cytotoxic T-lymphocyte antigen; Fc, crystallizable fragment; GVH, graft vs. host; HBC, hepatitis B virus; HIV, human immunodeficiency virus; ICAM, intercellular adhesion molecule; Ig, immunoglobulin; NK, natural killer; TNF, tumor necrosis factor; VLA, very late antigen.

Anti-CD20 Monoclonal Antibody: Rituximab

Rituximab (RITUXAN^®, MABTHERA^®) is a chimeric IgG1 that targets CD20, an antigen expressed on both normal and abnormal B cells. Rituximab thus destroys healthy and normal CD20-expressing B without any effect on progenitor stem cells, T cells, myeloid cells or plasma cells. Rituximab is mainly used in oncology for B-cell lymphoma. However, rituximab is increasingly used in B-cell dysfunction, such as auto-immune diseases and rheumatoid arthritis, and also in organ transplantation. As with alemtuzumab, doses and frequency of rituximab administration vary according to the indications, which could explain the differences in infectious complications.

As with alemtuzumab, the mechanisms of action of rituximab depend on ADCC, complement cytotoxicity and apoptosis. Rituximab induces profound B-cell lymphopenia without hypogammaglobulinemia or T-cell lymphopenia. However, some cases of hypogammaglobulinemia have been reported after prolonged treatment due to the loss of plasma cells after recurrent doses of rituximab.

The risks of infection with rituximab are relatively low, except for HIV-infected patients and those receiving other immunosuppressive agents. In 2007, Schult et al. published a meta-analysis on six randomized studies involving B lymphoma or Hodgkin disease patients treated with CHOP (cyclophosphamid, doxorubicine, vincristine, prednisone) with or without rituximab. No significant difference was observed in five studies,¹² and a significantly increased rate of infection was reported only in one study in which all the patients were HIV-positive.¹³ Boué et al. confirmed the low risk for HIV patients treated with rituximab if the number of CD4⁺ T cells is over 50/µl.¹⁴

Apart from the meta-analysis, some reported cases involved related CMV, herpes, parvovirus, BK, JC or enterovirus infections.^15–17 In 2009, Carson et al. reported 57 cases of JC virus infections in HIV-negative patients treated with rituximab.¹⁸ However, rituximab was used with other immune suppressive treatments, making any conclusion impossible. Reactivation of hepatitis B virus (HBV) infection has also been reported with rituximab, with increased risk of mortality.^19–23 In a few cases, bacterial infections with hypogammaglobulinemia were observed, leading to immunoglobulin supplementation. Similarly, Kamar et al. recently reported an increased infection rate after rituximab therapy in a retrospective study involving kidney transplant recipients;²⁴ however, the infections reported may also have been due to the combination with other immunosuppressive agents.

Brinkman et al. reported ten studies on the use of rituximab in rheumatoid arthritis. Infections and serious infections were reported in 10–65% and 0–5.4% of patients, respectively, with incidence rates of 0.8–1.55% and 0.038–0.08 events per year.²⁵ The published data showed neither increased nor serious infection compared to control groups (placebo or other DMARDs). Two open-label extension studies showed a higher level of serious infection after four or five courses, but in a small number of patients.^26,27

Although the risk of infection seems relatively low with rituximab, some authors have suggested prophylactic treatment with lamivudine because of reported HBV reactivation. Prophylactic treatment of pneumocystosis can also be discussed in cases of corticosteroid-associated treatment or T lymphopenia. In addition, the JC virus infections reported must be seen in relation to the large number of patients already treated. Nevertheless, a patient register has been opened to follow this risk towards natalizumab history (Table 1).

Anti-TNF: Infliximab, Adalimumab, Etanercept

Two different types of molecules that neutralize TNF are now marketed. Infliximab and adalimumab (distributed as REMICADE^® and HUMIRA^®, respectively), are mAbs; etanercept (ENBREL^®) is a fusion protein constructed with the human IgG1 Fc fragment and an extracellular portion of the TNFα receptor. TNF produced by macrophages plays a major role in the defense against intracellular organisms, the maintenance and formation of granulomata, and in leukocyte trafficking and immune complex clearance.²⁸ Infliximab, adalimumab and etanercept neutralize soluble TNFα, but the mAbs can induce death of activated T cells and macrophages.²⁹ There seems to be a relationship between FcγRIII receptor polymorphism and infliximab biological responses via an ADCC mechanism. The three drugs have several indications in human inflammatory diseases, such as inflammatory bowel disease, rheumatoid arthritis and chronic juvenile arthritis. Their different modes of action explain why etanercept cannot be used for Crohn disease and the infectious risk varies with the drug used. As for rituximab, the use of anti-TNF therapy has markedly increased in the last five years.

Preliminary studies with anti-TNF agents in polyarthritis disease reported tuberculosis infection that was confirmed by post-marketing studies conducted between 1998 and 2001, with 24.4 infections per 100,000 treated patients compared to 6.4 infections in the control group.³⁰ The risk of tuberculosis led to recommendations such as clinical examination, chest radiography, and tuberculin test or INF-Elispot test before the use of anti-TNF treatment. The systematic detection of tuberculosis before beginning treatment has considerably decreased the risk to 0.02–0.16 cases per 100 treated patients per year.^31,32

The British Society for Rheumatology Biologics register has reported several infectious agents (typical and atypical mycobacteria, listeria, nocardia, coccidiosis, streptococci) with infections occurring in 2–6 patients per 100 patients treated with anti-TNF therapy per year, and a 4-fold increased risk of tuberculosis.³³ A few cases of fungal infections have also been described with anti-TNF treatment. The updated results of 2011 emphasize the increased risk of serious infection in patients with RA, especially in the first six months of treatment.³⁴ The highest serious infection rate is reported for infliximab (46/1,000), followed by adalimumab (43/1,000) and etanercept (38/1,000); however, these differences disappear with adjusted analysis. These data suggest that there is in fact a small but significant overall risk of serious infection with anti-TNF therapy, mostly in the first six months of treatment.

Moreover, a review of the US Food and Drug Administration (FDA) Medwatch database has underlined an increase in non-tuberculosis mycobacteria (NTM) infections, with most cases reported since 2005, especially with infliximab-treated patients.³⁵ It is difficult to say whether the risk is higher with infliximab itself or it is due to the characteristics of patients or the more frequent association with methotrexate. However, it raises the question of the need for screening for NTM infections before initiation of anti-TNF therapy.

Finally, because of increased risk of viral infection, particularly HBV reactivation, with combined infliximab/methotrexate treatment, prophylactic treatments have been recommended with anti-TNF therapy.^36,37

All patients should be screened for tuberculosis before receiving any anti-TNF treatment, and should be informed about the risk of infection (Table 1).

Anti-integrin: Anti-VLA-4 (Natalizumab) and Anti-CD11a (Efalizumab)

Natalizumab (TYSABRI^®) is a humanized IgG4 that targets the α4 chain of the α4β2 integrin (VLA-4). The heterodimeric antigen is expressed on leucocytes and belongs to adhesion molecules involved in leukocyte migration. Natalizumab blocks the binding of VLA-4 to vascular cell adhesion molecule-1 expressed on endothelial cells; it reduces migration of activated leukocytes through the endothelium, and therefore reduces tissue inflammation. Natalizumab is indicated in neurological diseases, particularly in aggressive multiple sclerosis and in Crohn disease.

The most severe infectious complication that occurs with natalizumab treatment is a JC virus infection, responsible for progressive multifocal leucoencephalopathy (PML), a fatal form of encephalopathy. The first cases were described in 2005 in multiple sclerosis and in Crohn disease patients, resulting in temporarily suspension of the license.^38,39 Other cases of PML were described in 2006 (1 in the US, 6 in Europe), but natalizumab was relicensed for relapsing multiple sclerosis in 2008 because of its efficacy. Moreover, it seems that the risk of PML increases proportionally with the duration of exposure (especially after about two years of treatment), which means that drug monitoring is going to be a major challenge for the use of this drug.⁴⁰ However, its prescription is forbidden in cases of immune deficiency (HIV⁺ patients) and leucopenia and in combination with immunosuppressive drugs.

Efalizumab (RAPTIVA^®) is a humanized mAb that targets the CD11a molecule of heterodimer LFA1 (CD11a/CD18). As with VLA-4, this integrin is involved in transendothelial leukocyte migration. Efalizumab blocks CD11a (LFA1) binding to intracellular adhesion molecule-1 (also known as CD54) expressed on endothelial cells. Efalizumab was prescribed for psoriasis, but the product was withdrawn in 2009 because of reports of PML in 4 patients who had never received immunosuppressive treatment.⁴¹

Anti-integrin mAbs are associated with a high risk of PML, justifying evaluation of the risk/benefit ratio of these molecules, with specific monitoring and ongoing research on the pathogenesis of this severe disease (Table 1).⁴²

CTLA4-Ig: Abatacept, Belatacept

Abatacept (ORENCIA^®) is a fusion protein composed of an immunoglobulin Fc fused to the extracellular domain of CTLA4. It selectively modulates the CD80/CD86 CD28 co-stimulatory signal required for T-cell activation.⁴³ It is licensed for the treatment of rheumatoid arthritis as a first line of treatment. A new indication is for polyarticular juvenile arthritis. A second-generation fusion protein, belatacept, is currently being tested in Phase 2 clinical trials to provide extended graft survival with limited toxicity.

The safety of abatacept has been reported in several randomized trials.^44–50 In the first pilot study of abatacept compared with placebo, 90 patients received abatacept; the only infectious event in the double-blind period was a case of septic arthritis 40 days after a steroid infusion.⁴⁴ In 2005, Kremer et al.⁴⁵ compared 100 patients treated with placebo or CTLA4-Ig at 2 or 10 mg/kg. The serious adverse events were comparable, with an average of 12 patients in each group having upper respiratory tract infections and one case of cellulitis of the foot in the treated group. There was no opportunistic infection in the treated group. Genovese et al. reported serious infection in 2.3% of 258 abatacept-treated patients, comparable with the placebo group. There were no unusual or opportunistic infections, and the infection rate was similar in the abatacept and the placebo groups (37.6 vs. 32.3%, p = 0.3), with cases of nasopharyngitis, sinusitis, upper respiratory tract infection and bronchitis that were mild or minor.⁴⁶ The incidence and types of adverse events were consistent with the double-blind period over 18 months of long-term extension treatment for 218 patients.⁴⁷ In 2006, 385 methotrexate-resistant patients were treated with abatacept or placebo in a randomized, double-blind controlled trial. There was a higher incidence of serious infections (2.5 vs. 0.9%; difference, 1.6 percentage points).⁴⁸ A similar increase in infection rate was reported by Weinblatt et al. in 959 patients (2.9 vs. 1.9%) with serious infections without opportunistic infection or tuberculosis.⁴⁹ To date, only one study has combined data from double-blind and open-label phases of all abatacept clinical studies.⁵⁰ It included 4,134 patients (8,392 patient-years). The control group comprised patients treated for RA with disease-modifying antirheumatic drugs, age and sex adjusted (137,000 patients). There were 3.05 vs. 2.15 observed incidence rates per 100 patients respectively in the treated and the placebo groups, with no significant difference for infections requiring hospitalization. Three cases of tuberculosis, one of aspergillosis, one of blastomycosis and one of systemic Candida infection were reported in the abatacept group, similar to the DMARD group. This study suggests that in the long term the risk with abatacept in open-label studies is comparable with the risk for patients receiving non-biological DMARD therapy. The FDA is in the process of evaluating a biologic license application for belatacept: little information regarding risks of infection is available to date, but since belatacept does not selectively target alloreactive T cells, it must be used with other immunosuppressive drugs, which increases the risk of infection.⁵¹

Conclusion

The therapeutic use of mAbs and fusion proteins has revolutionized the treatment of several severe inflammatory and hematological diseases. However, blockade of key immunologic pathways in both innate and adaptive responses can directly or indirectly induce immune deficiency. Risk of infection increases according to the disease, the immune target, the dose and the other drugs administered concurrently. Pharmacogenetics will probably play a major role in the future, as shown with the importance of the FcγRIII receptor in the use of infliximab. The wide experience accumulated has improved the understanding of the mechanisms involved in these infections and has subsequently improved clinical safety, supported by the respect for indications, weighing of the risk/benefit ratios, identification of patients at risk, detection of latent infections before commencing treatment and careful clinical and biological monitoring. This approach requires regular updating of pharmocovigilance data on mAbs, specific informed patient consent and clear information for medical practitioners.

Nevertheless, there are cases where the relationship between the cell expression pattern of the target antigen and the immunosuppression pattern are not always clear (e.g., agammaglobulinemia and anti-CD20, which do not target plasma cells). More fundamental research is needed to improve understanding of the paradoxical immunosuppression pattern, especially the impact of mAbs on untargeted cells, e.g., neutropenia observed with rituximab.⁵²

We should also be asking ourselves how these new drugs may be made safer. Drug monitoring could be considered, but is still not usual except for certain drugs (infliximab). The hypothesis that rituximab could lead to regulatory B-cell destruction could be an argument to use intermittent B-cell depletion, with regular “resting” periods for the immune system.⁵³ Nevertheless, it has not yet been demonstrated whether the infectious risk is related to the dose or to the mechanism of action: it is quite probable that there is more than one cause that could be associated. Moreover, we should consider the targets that have to be selected for the development of new mAbs. Nevertheless, the use of anti-CD20 and anti-TNFα is associated with infectious risk when these drugs mainly induce depletion. We therefore have to be careful with the second generation of anti-CD20 (ocrelizumab, ofatumumab, veltuzumab), which are more effective in inducing apoptosis than rituximab and could probably lead to a higher risk of infection.⁵⁴ Finally, we have to bear in mind that the best model for mAb studies is the human body. Registers, fundamental research (based on clinical observations) and genetic studies are necessary to improve the safety of these new drugs and to develop biomarkers that could help us to determine when a patient is going to enter a biological period of infectious risk or which patient is genetically determined to be at high risk of infection.