B-lymphocyte depletion: rituximab

Rituximab is a chimeric murine-human monoclonal IgG1 that targets CD20 on normal and malignant B lymphocytes, with rapid and durable depletion of these cells for 6 to 9 months. Serum immunoglobulin levels remain largely stable, although prolonged hypogammaglobulinemia has been described in some patients with non-Hodgkin lymphoma (nHL) receiving rituximab concurrently with autologous hematopoietic stem cell transplantation (HSCT). Rituximab does not significantly affect CD3, CD4, CD8 or natural killer (NK) T-cell populations, and in theory has minimal effects on cell-mediated immunity.

Rituximab was initially approved by the FDA in 1997 for the treatment of relapsed or refractory low-grade or follicular nHL, and has subsequently been approved for the treatment of several other CD20+ B-cell lymphomas, either alone or in combination with other chemotherapy, and treatment of moderate to severe rheumatoid arthritis (RA) in combination with methotrexate in patients with an inadequate response to tumor necrosis factor α (TNF-α) antagonists. According to the manufacturer, there have been more than a million patient exposures since its approval, giving rituximab the most extensive clinical use of any biologic therapy to date.

No appreciable increase in infectious complications was observed with rituximab therapy over placebo in several randomized controlled trials for the treatment of nHL or B-cell chronic lymphocytic leukemia (CLL), although 2 recent meta-analyses of randomized trials of rituximab maintenance therapy in lymphoma patients have reported a higher relative risk (2.90) of grade 3 to 4 infections with rituximab therapy.

In some studies of human immunodeficiency virus (HIV)-associated nHL, addition of rituximab to standard chemotherapy was associated with a higher incidence of serious infections, particularly in patients with profound CD4 lymphopenia. A pooled assessment of 3 phase II randomized trials in which patients with HIV-associated nHL and a median CD4 count of 161 cells/μL receiving rituximab with standard chemotherapy reported a 14% incidence of serious opportunistic infections (OIs) within 3 months of chemotherapy completion despite trimethoprim-sulfamethoxazole and fluconazole prophylaxis, with the development of infections typically associated with impaired cellular immunity, including CMV retinitis, tuberculosis (TB), Pneumocystis jirovecii pneumonia (PCP), and salmonellosis. Another phase III trial of patients with HIV-associated nHL treated with rituximab or placebo plus cyclophosphamide, hydroxydaunomycin, oncovin (vincristine), and prednisone (CHOP) reported a significant difference in infectious mortality, 14% with rituximab (R-CHOP) compared with 2% with placebo. Of patients treated with R-CHOP, the incidence of infectious mortality was far higher in patients with baseline CD4 counts less than 50 cells/μL (36%) than patients with CD4 counts 50 cells/μL or greater (6%). Several OIs developed within 6 months of rituximab therapy in patients treated with R-CHOP, including PCP, CMV, invasive candidiasis, and Mycobacterium avium intercellulare (MAI), whereas no OIs were observed in the CHOP arm. Using more stringent enrollment criteria (exclusion of patients with CD4 <100 cells/μL, prior OIs, or poor performance status), a phase II study of R-CHOP in HIV-associated patients with nHL reported a lower incidence of infection. The depletion of B lymphocytes in patients with severe deficits in cellular immunity seems to increase the risk of developing OIs.

Hepatitis B virus (HBV) reactivation has been consistently associated with rituximab treatment in postmarketing reports. Humoral immunity to HBV surface antigen is known to play an important role in the containment of HBV infection, and several reports describe a reverse seroconversion phenomenon, with loss of protective HBV surface antibody and sometimes fulminant reactivation of HBV infection in rituximab recipients, particularly in patients with chronic HBV and detectable surface antigen before treatment. A recent study of patients with resolved HBV infection (positive HBV core antibody and negative surface antigen) in an HBV-endemic area receiving R-CHOP or CHOP for nHL described a 23.8% reactivation rate at 6 months in patients receiving R-CHOP, including 1 case of progression to hepatic failure, and 0% reactivation in patients receiving CHOP alone. The investigators associated rituximab exposure, male gender, and a lack of HBV surface antibody with HBV reactivation in this setting. Although most HBV reactivation seems to occur within 6 months of starting rituximab-containing therapy, reactivations as late as a year following therapy have been reported. Assessment of HBV status before starting rituximab-containing chemotherapy is essential in patients from endemic areas or with risk factors for prior HBV infection. Lamivudine prophylaxis during rituximab-containing chemotherapy has been reported to prevent HBV reactivation, and some groups recommend prophylactic HBV antiviral therapy in HBV surface antigen-positive patients for at least 6 months after completing chemotherapy. Optimal management of patients with resolved HBV infection receiving rituximab chemotherapy is less well defined, but patients should at a minimum have HBV surface antigen, HBV viral load, and liver function test monitoring every few months to assess for HBV reactivation.

From the initial FDA approval of rituximab in 1997 to 2008, 76 cases of progressive multifocal leukoencephalopathy (PML) associated with rituximab use have been reported to the manufacturer’s global safety database. PML is traditionally associated with profound deficits in cellular immunity, and the role of rituximab-induced impairments in humoral immunity in the development of PML is far less clear. Most cases have been reported in patients with lymphoproliferative disorders, although cases have also been reported in patients receiving rituximab therapy for systemic lupus erythematosus, RA, and immune thrombocytopenic purpura. Most of these patients were previously or concurrently exposed to other immunosuppressive therapies. Patients received a median of 6 rituximab doses and were exposed to rituximab for a median of 16 months before their PML diagnosis. The case fatality rate was high (90%), and the clinical course of these patients was rapidly progressive, with a 2-month median time to death after PML diagnosis. In 9 of 14 cases with available data, CD4 counts were less than 500 cells/μL. Although the absolute overall incidence of PML cases in patients exposed to rituximab is low, rituximab has carried a black box warning for PML since 2007, and active postmarketing surveillance is ongoing.

A case-control study of patients with persistent, relapsing Babesia microti infection and severe morbidity and death despite repeated courses of antiparasitic therapy identified rituximab treatment as an important factor in the inability to clear this infection. Patients required prolonged courses of antiparasitic therapy for cure, for at least 2 weeks after clearance of parasites on blood smear.

A case series reported a higher rate of PCP infection in patients receiving rituximab in combination with CHOP-based chemotherapy for lymphoma (6%–13%) compared with patients receiving comparable CHOP-based regimens alone (4%). PCP has been linked to rituximab therapy in other case series and reports, although patients concurrently received steroids and other immunosuppressive therapy. Although PCP infection has been classically associated with CD4 lymphocyte deficits, B-cell deficient mice are exquisitely sensitive to PCP infection and are unable to generate a protective CD4 memory and effector T-cell response to PCP.

There have been several reports of TB and nontuberculous mycobacterial (NTM) infections in association with rituximab. A recent survey-based study of mycobacterial infections reported by members of the Emerging Infections Network identified 3 cases of TB and 5 cases of NTM associated with rituximab use. Severe MAI infection and disseminated M kansasii and M wolinskyi infections have been reported in patients receiving rituximab, although patients in these cases also received other concurrent immunosuppression. B lymphocytes seem to be important in the containment of TB in murine infection models, and B-lymphocyte knockout mice are unable to contain TB infections, with an increased pulmonary mycobacterial burden compared with mice with normal B-cell function. B lymphocytes are present in the periphery of tuberculous granulomas in active folliclelike centers associated with antigen-presenting cells and CD4 and CD8 T lymphocytes in human TB infection, and are believed to help orchestrate containment of infection.

Several other severe infections have been linked to rituximab use at a sporadic case report level, including persistent enteroviral meningoencephalitis, CMV disease, disseminated varicella-zoster virus (VZV), pure red cell aplasia from parvovirus B19 infection, West Nile virus meningoencephalitis, and nocardiosis, although all patients received rituximab in combination with other immunosuppressive therapies, and the causal role of rituximab itself is unclear.

Anti-TNF-α therapies: infliximab, adalimumab, etanercept, certolizumab pegol, golimumab

Infliximab, adalimumab, and golimumab are monoclonal antibodies directed against TNF-α, certolizumab pegol is a pegylated fragment antigen-binding (Fab) fragment of a humanized anti-TNF-α monoclonal antibody, and etanercept is a soluble receptor for TNF-α. All of these therapies abrogate TNF-α activity to varying degrees, and are effective treatment modalities for various inflammatory conditions. The monoclonal antibodies bind soluble and cell-surface TNF-α, with fixation of complement and lysis of T lymphocytes and neutrophils expressing surface TNF-α, whereas etanercept is able to bind only to soluble TNF-α and does not seem to have the same lytic effect on cells expressing membrane-bound TNF-α. TNF-α is essential for macrophage activation, phagosome activation, differentiation of monocytes into macrophages, recruitment of neutrophils and macrophages, granuloma formation, and maintenance of granuloma integrity, and therapy with TNF-α blockers is associated with a particularly increased risk of granulomatous and intracellular infections.

Infliximab was approved in 1998 for the treatment of RA, psoriatic arthritis and plaque psoriasis, ankylosing spondylitis, ulcerative colitis, and Crohn disease. Adalimumab was approved in 1999, and has the same treatment indications as infliximab, except for ulcerative colitis. Etanercept was approved in 2001 for the treatment of RA, psoriasis and psoriatic arthritis, and ankylosing spondylitis. In 2009, the FDA approved golimumab for RA, psoriatic arthritis, and ankylosing spondylitis, and certolizumab for RA and Crohn disease refractory to conventional therapy.

The overall incidence of serious infections associated with anti-TNF-α therapy has been estimated from comprehensive national registry data of RA patients from the United Kingdom and Germany at 5.2 to 6.2 per 100 patient-years in patients with infliximab, 6.3 per 100 patient-years with adalimumab, and 5.3 to 6.4 per 100 person-years with etanercept. The German biologics register study adjusted for differences in patient characteristics using propensity scores, and reported an adjusted relative risk for total serious adverse infectious events of 2.2 with etanercept and 3.0 with infliximab use. An observational cohort study of the Swedish biologics register of RA patients assessed the risk of hospitalization with infection with anti-TNF-α therapy, and reported an increased relative risk of 1.43 during the first year, 1.15 during the second year, and no difference in the risk of hospitalization in subsequent years compared with RA patients not receiving anti-TNF-α therapy. A meta-analysis of anti-TNF-α therapy trials in RA reported an odds ratio (OR) of 2.0 for serious infections and 3.3 for malignancy in patients receiving anti-TNF-α therapy, compared with placebo, with only 12 granulomatous infections (10 cases of TB, 1 of histoplasmosis, and 1 of coccidioidomycosis) in 126 serious infections.

An association between anti-TNF-α therapy and TB was noted a few years after the initial approval of infliximab. A query of the FDA MedWatch spontaneous reporting system in 2001 showed 70 TB cases developing a median of 12 weeks after initial infliximab exposure, with a high proportion of extrapulmonary dissemination (57%), and a frequent lack of granuloma formation in patients with biopsy samples. In the United States, the rate of granulomatous and intracellular infection in patients receiving anti-TNF-α therapy has been estimated from the FDA adverse event reporting system, which relies on spontaneous reporting of cases, unlike national registries. These infections were more common in patients receiving infliximab (129 events per 100,000 patients) than etanercept (60 events per 100,000 patients). The rate of TB was 54 per 100,000 patients in infliximab patients, with a rate ratio of 1.9 compared with etanercept patients, and the median time to TB was substantially shorter in infliximab patients (17 weeks) than etanercept patients (48 weeks). A case-control study of RA patients in an American pharmaceutical claims database also identified a higher rate ratio in patients treated with biologic therapy (1.5) compared with patients receiving nonbiologic RA therapy, and also reported an earlier median time to TB in patients receiving infliximab (17 weeks) than in patients receiving etanercept (79 weeks). A Monte Carlo simulation of time to reactivation of latent TB calculated a median monthly rate of TB reactivation of 20.8% in patients receiving infliximab, 12.1-fold higher than patients receiving etanercept. There was a clustering of infliximab-associated TB reactivation cases in the first year; the risk of progression of new TB infection to active disease was comparable in infliximab and etanercept patients, suggesting that much of the excess risk of TB with infliximab therapy over etanercept is a consequence of more efficient latent TB reactivation shortly after starting anti-TNF-α therapy, whereas both infliximab and etanercept fairly equally increase the risk of active incident disease. In TB patients receiving anti-TNF-α therapy, an immune reconstitution inflammatory syndrome-like reaction has been described after withdrawal of these agents, sometimes requiring the reinitiation of these agents to control an overly exuberant and deleterious host immune response.

NTM infections have also been associated with anti-TNF-α therapy. The rate of NTM in the United States has been estimated at 9 cases per 100,000 patients with infliximab and 6 cases per 100,000 patients with etanercept. A recent survey asking members of the Emerging Infections Network to identify all cases of TB and NTM in patients receiving anti-TNF-α therapy in their clinical practice during the prior 6 months found that reports of NTM (65%) exceeded reports of TB. Most cases were MAI infections, but cases of M chelonae , M abscessus , M marinum , M fortuitum , M haemophilum , M kansasii , and M scrofulaceum were also reported. Cases of lepromatous leprosy have been reported in patients from Louisiana, Texas, and the Brazilian Amazon receiving anti-TNF-α therapy for various indications, with high numbers of bacilli on biopsy. The progression of disease has been observed to be faster than usual in these patients. A type 1 reversal reaction was described in the 2 North American patients a month after discontinuing anti-TNF-α therapy and starting antibiotic therapy, with exacerbation of skin lesions, malaise, and greater organization of the inflammatory infiltrate on skin biopsy. Severe infections with other rare mycobacterial pathogens, such as M peregrinum , M aurum , M bovis , and M szulgai have been described on a case report level in patients receiving anti-TNF-α therapy.

Histoplasmosis has been associated with anti-TNF-α therapy, with a significantly higher rate in infliximab (19 cases per 100,000 patients) than etanercept recipients (3 cases per 100,000 patients) in the United States. Most cases have been reported within 6 months of initiation of anti-TNF-α therapy, and many of the reported cases have been associated with disseminated disease.

Coccidioidomycosis has been reported in association with anti-TNF-α therapy, also with significantly higher rates with infliximab (11 cases per 100,000 patients) than etanercept (1 case per 100,000 patients) exposure in the United States. An assessment of patients with inflammatory arthritis living in Coccidioides immitis –endemic areas reported a relative risk of approximately 5 for coccidioidomycosis with infliximab compared with other antirheumatic drugs.

The rate of cryptococcosis in the United States is estimated at 9 cases per 100,000 patients exposed to anti-TNF-α therapy, and there is no notable difference in patients treated with infliximab or etanercept.

TNF-α is essential for normal activation of macrophage phagosomes and clearance of intracellular pathogens, and several intracellular infections have been associated with anti-TNF-α therapy, including Listeria bacteremia and meningitis, which has been reported at higher rates in infliximab (9 cases per 100,000 patients) than etanercept (1 case per 100,000 patients) recipients, Legionella pneumonia, and salmonellosis. Several intracellular protozoal pathogens have been reported in association with anti-TNF-α therapy, including relapsing cutaneous and visceral leishmaniasis in endemic areas, overwhelming Plasmodium falciparum parasitemia, and a report of an eventually fatal progressive myositis caused by Brachiola algerae .

TNF-α enhances conidial phagocytosis by alveolar macrophages, augments the effectiveness of polymorphonuclear cells against Aspergillus hyphae , and contributes to the recruitment and activation of neutrophils and mononuclear cells in the lung; anti-TNF-α therapy has been associated with an increased risk of invasive fungal infections, with overall invasive aspergillosis (IA) and invasive candidasis rates of approximately 7 to 8 cases per 100,000 exposed patients. A review of invasive fungal infections associated with TNF-α inhibition identified 281 cases, most associated with infliximab therapy, and many associated with concurrent corticosteroid therapy. IA was associated with TNF-α therapy in 64 reports, invasive candidiasis in 64 cases, and zygomycosis in 4 cases. Other fungi that generally cause localized disease, such as Trichophyton rubrum and Sporothrix schenckii infection, have been associated with disseminated disease in recipients of anti-TNF-α therapy.

Other infections associated with anti-TNF-α therapy include PCP, nocardiosis (∼4 cases per 100,000 infliximab recipients), toxoplasmosis (∼2 cases per 100,000 infliximab recipients), bartonellosis, and brucellosis.

The relationship between TNF-α inhibition and reactivation of latent and chronic viral infections is less well defined. A recent assessment of herpes zoster in the large German biologics registry reported 86 episodes, with a crude incidence rate of 11.1 per 1000 patient-years with monoclonal antibodies (adalimumab, infliximab), 8.9 for etanercept, and 5.6 with conventional disease-modifying antirheumatic therapy. Adjusting for other factors, the hazard ratio of herpes zoster with monoclonal antibody therapy compared with conventional therapy was 1.8, and there was no discernible increase in risk with etanercept. An assessment of large patient databases from the United States and the United Kingdom identified an increased risk of zoster with conventional disease-modifying RA therapy and a higher risk (OR 1.5) in patients receiving biologic therapy. Reactivation of other herpesviruses has not clearly been associated with anti-TNF-α therapy, although 3 cases of herpes simplex virus encephalitis were recently reported in patients receiving infliximab or adalimumab. Several cases of severe HBV reactivation in patients with positive surface antigen at the start of anti-TNF-α therapy have been reported, although concurrent lamivudine treatment may decrease the risk of reactivation in these patients.

Allogeneic HSCT recipients with severe steroid-refractory acute graft-versus-host disease (GVHD) have a high risk of invasive fungal disease (IFD) and CMV reactivation, likely because of the loss of normal mucosal barrier integrity and heavy concurrent immune suppression; the addition of anti-TNF-α therapy further increases the risk of these infections. A retrospective evaluation of 21 patients receiving infliximab for steroid-refractory GVHD reported the development of bacterial infections in 81%, viral reactivations in 67% (predominantly CMV), and invasive fungal infections in 48% of patients. Patients receiving infliximab for severe steroid-refractory GVHD at the authors’ institution had a high incidence rate of IFD (6.8 cases per 1000 GVHD patient-days) compared with 0.53 cases per 1000 GVHD patient-days in patients not exposed to infliximab. The adjusted hazard ratio for IFD in patients exposed to infliximab was 13.6 compared with patients who were not exposed.

Clinical experience with certolizumab pegol and golimumab is limited. Some studies of certolizumab for Crohn disease and RA have reported an increased incidence of serious infections compared with placebo, including several cases of TB. Although many randomized trials of golimumab have shown no increase in the overall incidence of serious infections compared with placebo, a trial in RA patients also treated with methotrexate and a trial in patients with severe asthma refractory to high-dose inhaled steroids and β2 agonists reported a higher incidence of serious infections with golimumab therapy, with 1 case of TB and 1 case of Legionella pneumonia. As golimumab is similar in structure to infliximab, it is likely that a comparable pattern of OIs will emerge with further clinical use.

A consensus group statement on the use of biologic agents in patients with RA recommends measures to reduce infectious complications in patients receiving anti-TNF-α therapy, including screening for latent TB infection and assessment for latent or chronic HBV and hepatitis C virus infection before starting therapy.

Anti-interleukin 1 therapies: anakinra, rilonacept

The cytokine interleukin 1 (IL-1) is secreted by numerous cell types in response to inflammatory antigens, and has a wide range of biologic activity, including mediation of the febrile response to infection and inflammation, B-cell activation, induction of IL-2 with subsequent stimulation of T-cell maturation, and induction of IL-6, TNF-γ, and IL-8.

Anakinra, a recombinant IL-1 receptor antagonist, was approved for the treatment of moderate to severe RA in 2001. It competitively inhibits binding of IL-1 to IL-1 type I receptor, with a decrease in the response to inflammatory stimuli. The German biologics registry reported a rate of 3.2 serious infections per 100 patient-years in patients exposed to anakinra, although the total number of patients receiving anakinra was small. A meta-analysis of all randomized placebo-controlled trials evaluating anakinra in RA reported a 1.4% incidence of serious infections with anakinra, compared with 0.5% with placebo, and the OR of serious infection was 3.40 in patients treated with high-dose anakinra versus placebo, although this difference was not significant when results were adjusted for underlying comorbidities. Pneumonia and other bacterial infections accounted for most events; no OIs were reported. One case of TB was reported in a patient enrolled in an RA study who had underlying pneumoconiosis from mining, and 1 case of visceral leishmaniasis was reported in a child living in an endemic area of France receiving anakinra for systemic onset juvenile idiopathic arthritis 6 months after starting treatment.

Rilonacept is a dimeric fusion protein of the ligand-binding domain of the extracellular portion of the human IL-1 receptor and IL-1 receptor accessory protein linked to the Fc portion of human IgG1. It acts as a soluble decoy receptor and binds IL-1β, preventing its normal binding to IL-1 receptors. Rilonacept was approved in 2008 for the treatment of cryopyrin-associated periodic syndromes (CAPS), characterized by excessive IL-1β production. In a study of 47 patients with CAPS, rilonacept was associated with upper respiratory infections in 26% of patients, compared with 4% in patients receiving placebo. One case of Streptococcus pneumoniae meningitis developed during the open-label extension period, but was believed not to be related to the study drug. The rilonacept package insert reported a case of MAI olecranon bursitis in a patient receiving rilonacept for an unapproved indication and intra-articular glucocorticoid injections.

Anti-TNF-α therapies: infliximab, adalimumab, etanercept, certolizumab pegol, golimumab

Infliximab, adalimumab, and golimumab are monoclonal antibodies directed against TNF-α, certolizumab pegol is a pegylated fragment antigen-binding (Fab) fragment of a humanized anti-TNF-α monoclonal antibody, and etanercept is a soluble receptor for TNF-α. All of these therapies abrogate TNF-α activity to varying degrees, and are effective treatment modalities for various inflammatory conditions. The monoclonal antibodies bind soluble and cell-surface TNF-α, with fixation of complement and lysis of T lymphocytes and neutrophils expressing surface TNF-α, whereas etanercept is able to bind only to soluble TNF-α and does not seem to have the same lytic effect on cells expressing membrane-bound TNF-α. TNF-α is essential for macrophage activation, phagosome activation, differentiation of monocytes into macrophages, recruitment of neutrophils and macrophages, granuloma formation, and maintenance of granuloma integrity, and therapy with TNF-α blockers is associated with a particularly increased risk of granulomatous and intracellular infections.

Infliximab was approved in 1998 for the treatment of RA, psoriatic arthritis and plaque psoriasis, ankylosing spondylitis, ulcerative colitis, and Crohn disease. Adalimumab was approved in 1999, and has the same treatment indications as infliximab, except for ulcerative colitis. Etanercept was approved in 2001 for the treatment of RA, psoriasis and psoriatic arthritis, and ankylosing spondylitis. In 2009, the FDA approved golimumab for RA, psoriatic arthritis, and ankylosing spondylitis, and certolizumab for RA and Crohn disease refractory to conventional therapy.

The overall incidence of serious infections associated with anti-TNF-α therapy has been estimated from comprehensive national registry data of RA patients from the United Kingdom and Germany at 5.2 to 6.2 per 100 patient-years in patients with infliximab, 6.3 per 100 patient-years with adalimumab, and 5.3 to 6.4 per 100 person-years with etanercept. The German biologics register study adjusted for differences in patient characteristics using propensity scores, and reported an adjusted relative risk for total serious adverse infectious events of 2.2 with etanercept and 3.0 with infliximab use. An observational cohort study of the Swedish biologics register of RA patients assessed the risk of hospitalization with infection with anti-TNF-α therapy, and reported an increased relative risk of 1.43 during the first year, 1.15 during the second year, and no difference in the risk of hospitalization in subsequent years compared with RA patients not receiving anti-TNF-α therapy. A meta-analysis of anti-TNF-α therapy trials in RA reported an odds ratio (OR) of 2.0 for serious infections and 3.3 for malignancy in patients receiving anti-TNF-α therapy, compared with placebo, with only 12 granulomatous infections (10 cases of TB, 1 of histoplasmosis, and 1 of coccidioidomycosis) in 126 serious infections.

An association between anti-TNF-α therapy and TB was noted a few years after the initial approval of infliximab. A query of the FDA MedWatch spontaneous reporting system in 2001 showed 70 TB cases developing a median of 12 weeks after initial infliximab exposure, with a high proportion of extrapulmonary dissemination (57%), and a frequent lack of granuloma formation in patients with biopsy samples. In the United States, the rate of granulomatous and intracellular infection in patients receiving anti-TNF-α therapy has been estimated from the FDA adverse event reporting system, which relies on spontaneous reporting of cases, unlike national registries. These infections were more common in patients receiving infliximab (129 events per 100,000 patients) than etanercept (60 events per 100,000 patients). The rate of TB was 54 per 100,000 patients in infliximab patients, with a rate ratio of 1.9 compared with etanercept patients, and the median time to TB was substantially shorter in infliximab patients (17 weeks) than etanercept patients (48 weeks). A case-control study of RA patients in an American pharmaceutical claims database also identified a higher rate ratio in patients treated with biologic therapy (1.5) compared with patients receiving nonbiologic RA therapy, and also reported an earlier median time to TB in patients receiving infliximab (17 weeks) than in patients receiving etanercept (79 weeks). A Monte Carlo simulation of time to reactivation of latent TB calculated a median monthly rate of TB reactivation of 20.8% in patients receiving infliximab, 12.1-fold higher than patients receiving etanercept. There was a clustering of infliximab-associated TB reactivation cases in the first year; the risk of progression of new TB infection to active disease was comparable in infliximab and etanercept patients, suggesting that much of the excess risk of TB with infliximab therapy over etanercept is a consequence of more efficient latent TB reactivation shortly after starting anti-TNF-α therapy, whereas both infliximab and etanercept fairly equally increase the risk of active incident disease. In TB patients receiving anti-TNF-α therapy, an immune reconstitution inflammatory syndrome-like reaction has been described after withdrawal of these agents, sometimes requiring the reinitiation of these agents to control an overly exuberant and deleterious host immune response.

NTM infections have also been associated with anti-TNF-α therapy. The rate of NTM in the United States has been estimated at 9 cases per 100,000 patients with infliximab and 6 cases per 100,000 patients with etanercept. A recent survey asking members of the Emerging Infections Network to identify all cases of TB and NTM in patients receiving anti-TNF-α therapy in their clinical practice during the prior 6 months found that reports of NTM (65%) exceeded reports of TB. Most cases were MAI infections, but cases of M chelonae , M abscessus , M marinum , M fortuitum , M haemophilum , M kansasii , and M scrofulaceum were also reported. Cases of lepromatous leprosy have been reported in patients from Louisiana, Texas, and the Brazilian Amazon receiving anti-TNF-α therapy for various indications, with high numbers of bacilli on biopsy. The progression of disease has been observed to be faster than usual in these patients. A type 1 reversal reaction was described in the 2 North American patients a month after discontinuing anti-TNF-α therapy and starting antibiotic therapy, with exacerbation of skin lesions, malaise, and greater organization of the inflammatory infiltrate on skin biopsy. Severe infections with other rare mycobacterial pathogens, such as M peregrinum , M aurum , M bovis , and M szulgai have been described on a case report level in patients receiving anti-TNF-α therapy.

Histoplasmosis has been associated with anti-TNF-α therapy, with a significantly higher rate in infliximab (19 cases per 100,000 patients) than etanercept recipients (3 cases per 100,000 patients) in the United States. Most cases have been reported within 6 months of initiation of anti-TNF-α therapy, and many of the reported cases have been associated with disseminated disease.

Coccidioidomycosis has been reported in association with anti-TNF-α therapy, also with significantly higher rates with infliximab (11 cases per 100,000 patients) than etanercept (1 case per 100,000 patients) exposure in the United States. An assessment of patients with inflammatory arthritis living in Coccidioides immitis –endemic areas reported a relative risk of approximately 5 for coccidioidomycosis with infliximab compared with other antirheumatic drugs.

The rate of cryptococcosis in the United States is estimated at 9 cases per 100,000 patients exposed to anti-TNF-α therapy, and there is no notable difference in patients treated with infliximab or etanercept.

TNF-α is essential for normal activation of macrophage phagosomes and clearance of intracellular pathogens, and several intracellular infections have been associated with anti-TNF-α therapy, including Listeria bacteremia and meningitis, which has been reported at higher rates in infliximab (9 cases per 100,000 patients) than etanercept (1 case per 100,000 patients) recipients, Legionella pneumonia, and salmonellosis. Several intracellular protozoal pathogens have been reported in association with anti-TNF-α therapy, including relapsing cutaneous and visceral leishmaniasis in endemic areas, overwhelming Plasmodium falciparum parasitemia, and a report of an eventually fatal progressive myositis caused by Brachiola algerae .

TNF-α enhances conidial phagocytosis by alveolar macrophages, augments the effectiveness of polymorphonuclear cells against Aspergillus hyphae , and contributes to the recruitment and activation of neutrophils and mononuclear cells in the lung; anti-TNF-α therapy has been associated with an increased risk of invasive fungal infections, with overall invasive aspergillosis (IA) and invasive candidasis rates of approximately 7 to 8 cases per 100,000 exposed patients. A review of invasive fungal infections associated with TNF-α inhibition identified 281 cases, most associated with infliximab therapy, and many associated with concurrent corticosteroid therapy. IA was associated with TNF-α therapy in 64 reports, invasive candidiasis in 64 cases, and zygomycosis in 4 cases. Other fungi that generally cause localized disease, such as Trichophyton rubrum and Sporothrix schenckii infection, have been associated with disseminated disease in recipients of anti-TNF-α therapy.

Other infections associated with anti-TNF-α therapy include PCP, nocardiosis (∼4 cases per 100,000 infliximab recipients), toxoplasmosis (∼2 cases per 100,000 infliximab recipients), bartonellosis, and brucellosis.

The relationship between TNF-α inhibition and reactivation of latent and chronic viral infections is less well defined. A recent assessment of herpes zoster in the large German biologics registry reported 86 episodes, with a crude incidence rate of 11.1 per 1000 patient-years with monoclonal antibodies (adalimumab, infliximab), 8.9 for etanercept, and 5.6 with conventional disease-modifying antirheumatic therapy. Adjusting for other factors, the hazard ratio of herpes zoster with monoclonal antibody therapy compared with conventional therapy was 1.8, and there was no discernible increase in risk with etanercept. An assessment of large patient databases from the United States and the United Kingdom identified an increased risk of zoster with conventional disease-modifying RA therapy and a higher risk (OR 1.5) in patients receiving biologic therapy. Reactivation of other herpesviruses has not clearly been associated with anti-TNF-α therapy, although 3 cases of herpes simplex virus encephalitis were recently reported in patients receiving infliximab or adalimumab. Several cases of severe HBV reactivation in patients with positive surface antigen at the start of anti-TNF-α therapy have been reported, although concurrent lamivudine treatment may decrease the risk of reactivation in these patients.

Allogeneic HSCT recipients with severe steroid-refractory acute graft-versus-host disease (GVHD) have a high risk of invasive fungal disease (IFD) and CMV reactivation, likely because of the loss of normal mucosal barrier integrity and heavy concurrent immune suppression; the addition of anti-TNF-α therapy further increases the risk of these infections. A retrospective evaluation of 21 patients receiving infliximab for steroid-refractory GVHD reported the development of bacterial infections in 81%, viral reactivations in 67% (predominantly CMV), and invasive fungal infections in 48% of patients. Patients receiving infliximab for severe steroid-refractory GVHD at the authors’ institution had a high incidence rate of IFD (6.8 cases per 1000 GVHD patient-days) compared with 0.53 cases per 1000 GVHD patient-days in patients not exposed to infliximab. The adjusted hazard ratio for IFD in patients exposed to infliximab was 13.6 compared with patients who were not exposed.

Clinical experience with certolizumab pegol and golimumab is limited. Some studies of certolizumab for Crohn disease and RA have reported an increased incidence of serious infections compared with placebo, including several cases of TB. Although many randomized trials of golimumab have shown no increase in the overall incidence of serious infections compared with placebo, a trial in RA patients also treated with methotrexate and a trial in patients with severe asthma refractory to high-dose inhaled steroids and β2 agonists reported a higher incidence of serious infections with golimumab therapy, with 1 case of TB and 1 case of Legionella pneumonia. As golimumab is similar in structure to infliximab, it is likely that a comparable pattern of OIs will emerge with further clinical use.

A consensus group statement on the use of biologic agents in patients with RA recommends measures to reduce infectious complications in patients receiving anti-TNF-α therapy, including screening for latent TB infection and assessment for latent or chronic HBV and hepatitis C virus infection before starting therapy.

Anti-interleukin 1 therapies: anakinra, rilonacept

The cytokine interleukin 1 (IL-1) is secreted by numerous cell types in response to inflammatory antigens, and has a wide range of biologic activity, including mediation of the febrile response to infection and inflammation, B-cell activation, induction of IL-2 with subsequent stimulation of T-cell maturation, and induction of IL-6, TNF-γ, and IL-8.

Anakinra, a recombinant IL-1 receptor antagonist, was approved for the treatment of moderate to severe RA in 2001. It competitively inhibits binding of IL-1 to IL-1 type I receptor, with a decrease in the response to inflammatory stimuli. The German biologics registry reported a rate of 3.2 serious infections per 100 patient-years in patients exposed to anakinra, although the total number of patients receiving anakinra was small. A meta-analysis of all randomized placebo-controlled trials evaluating anakinra in RA reported a 1.4% incidence of serious infections with anakinra, compared with 0.5% with placebo, and the OR of serious infection was 3.40 in patients treated with high-dose anakinra versus placebo, although this difference was not significant when results were adjusted for underlying comorbidities. Pneumonia and other bacterial infections accounted for most events; no OIs were reported. One case of TB was reported in a patient enrolled in an RA study who had underlying pneumoconiosis from mining, and 1 case of visceral leishmaniasis was reported in a child living in an endemic area of France receiving anakinra for systemic onset juvenile idiopathic arthritis 6 months after starting treatment.

Rilonacept is a dimeric fusion protein of the ligand-binding domain of the extracellular portion of the human IL-1 receptor and IL-1 receptor accessory protein linked to the Fc portion of human IgG1. It acts as a soluble decoy receptor and binds IL-1β, preventing its normal binding to IL-1 receptors. Rilonacept was approved in 2008 for the treatment of cryopyrin-associated periodic syndromes (CAPS), characterized by excessive IL-1β production. In a study of 47 patients with CAPS, rilonacept was associated with upper respiratory infections in 26% of patients, compared with 4% in patients receiving placebo. One case of Streptococcus pneumoniae meningitis developed during the open-label extension period, but was believed not to be related to the study drug. The rilonacept package insert reported a case of MAI olecranon bursitis in a patient receiving rilonacept for an unapproved indication and intra-articular glucocorticoid injections.