Pneumocystis jiroveci pneumonia following rituximab treatment in Wegener's granulomatosis




Wegener's granulomatosis (WG) is a devastating small-vessel vasculitis in children. Standard treatment consists of immunosuppressive medications with cyclophosphamide potentially associated with significant infectious side effects, including Pneumocystis jiroveci pneumonia (PCP). Recently, rituximab, a monoclonal antibody against B cells, has successfully been used in refractory disease.


We describe the first pediatric patient with refractory WG with sinus and lung disease who developed PCP 6 months after treatment with rituximab, while being treated with methotrexate and prednisone. This 9-year-old child had no CD20+ B cells at time of infection, with normal lymphocyte and CD4 counts.


This study provides a review of the published literature, including current protocols, which suggest chemoprophylaxis only in WG patients receiving T cell–targeted immunosuppression such as cyclophosphamide. However, clinical and laboratory evidence points toward a possible role of B cells in the defense against PCP.


Routine PCP chemoprophylaxis should be strongly considered in patients with WG treated with rituximab.


Wegener's granulomatosis (WG) is a relapsing, antineutrophil cytoplasmic antibody (ANCA)–positive systemic vasculitis affecting adults and children. WG is associated with significant morbidity not only from disease activity, but also from the mandatory, aggressive immunosuppressive therapy (1). Pneumocystis jiroveci pneumonia (PCP) is a common and life-threatening complication of immunosuppressive therapy in ANCA vasculitis, in particular of cyclophosphamide treatment (2). PCP can effectively be prevented by institution of chemoprophylaxis, usually with cotrimoxazole. Rituximab, a monoclonal antibody against CD20, a B cell surface molecule, is increasingly used in ANCA vasculitis. Efficacy has been demonstrated for its use in refractory WG (3). PCP has been strongly linked to low T cell counts, especially in patients infected with the human immunodeficiency virus. Therefore, B cell depletion induced by rituximab treatment is not commonly considered a PCP risk factor. We report, to our knowledge, the first case of a pediatric patient with severe refractory WG who developed PCP subsequent to treatment with rituximab, and we review the published evidence for chemoprophylaxis.

Case report

A 9-year-old girl presented with an 8-day history of fever, congestion, productive cough, cracked lips, lymphadenitis, and conjunctivitis. She was diagnosed as having Kawasaki disease and was treated with 2 courses of intravenous immunoglobulin (2 gm/kg) and acetylsalicylic acid (100 mg/kg/day) with no significant improvement. On reexamination, she reported increasing cough with hemoptysis, rhinorrhea, and epistaxis. Physical examination revealed an erythematous nasal cavity with blood-tinged mucus, cervical lymphadenopathy and hepatosplenomegaly, and decreased air entry to her left lung on auscultation. Further investigations yielded a urinalysis with mild proteinuria and granular casts. A computed tomography (CT) scan of the chest revealed multiple nodules, ground-glass appearance, and a critical left main stem bronchus stenosis. A CT scan of the sinuses showed opacification of the left frontal, ethmoidal, and maxillary sinuses. Laboratory testing demonstrated significantly elevated inflammatory parameters with an erythrocyte sedimentation rate (ESR) of 54 mm/hour, a C-reactive protein (CRP) level of 214 mg/dl, a white blood count (WBC) of 13.4 × 109/liter, and a platelet count of 688 × 109/liter. She had a positive cytoplasmic ANCA pattern, with a high titer on anti–proteinase 3 enzyme-linked immunosorbent assay (>100 units/ml). A biopsy of nasal granulomata was performed at diagnosis and the results showed acute and chronic inflammation with mainly lymphocytic and neutrophil infiltrate with interspersed multinucleate giant cells and focal necrosis. The patient was diagnosed as having WG.

She was initially treated with prednisone (2 mg/kg/day) and 7 dosages of intravenous monthly cyclophosphamide (750 mg/m2). During this treatment she received PCP prophylaxis with cotrimoxazole (4 mg/kg of trimethoprim, 3 days per week), which was discontinued after 8 months. Her clinical features, including the respiratory symptoms, and inflammatory markers normalized. No bronchoscopic intervention was required for the stenosis, which had improved on subsequent chest CT studies. Postinduction, she was switched to maintenance therapy with enteric-coated mycophenolic acid (28 mg/kg/day) and low-dosage prednisone (0.15 mg/kg/day), and she remained in clinical remission. Six months into this therapy, and 14 months after diagnosis, she presented with increasing clinical signs of sinusitis, hearing loss, and mild symptoms of inflammation of the upper airway. A repeat chest CT scan showed multiple large cavitating lesions in both lungs (Figure 1A). She was treated with rituximab (4 weekly dosages at 375 mg/m2), which led to significant clinical and radiographic improvement (Figure 1B). Mycophenolic acid was discontinued and prednisone was tapered off during the following 4 months. Laboratory testing revealed complete depletion of CD19+ and CD20+ B cells.

Figure 1.

Computed tomography chest imaging of the patient A, prior to treatment with rituximab, showing multiple cavitating nodules in both lungs, and B, 6 months after treatment with rituximab, showing almost complete resolution of the lesions.

Five months after rituximab therapy, the patient presented with nasal crusting, discharge, and bleeding, as well as continuous tearing from both eyes. Her infectious evaluation remained negative, but nasal endoscopy showed perforation of the nasal septum. Since her blood work at that time showed increasing CD20+ B cells and elevated inflammatory parameters, she received retreatment with rituximab (2 doses of 375 mg/m2) 6 months after the initial course. She also received a therapeutic dosage of cotrimoxazole (4 mg/kg of trimethoprim daily) shortly before the start of rituximab retreatment, which was discontinued after 2 months. A repeat CT scan of her sinuses revealed progressive destruction of the nasal septum and the medial walls of both maxillary sinuses (Figure 2A). A visible saddle nose deformity was noted (Figure 2B). Sinus lavage and multiple biopsy samples returned negative results for microbial growth, including Pneumocystis and fungus. Treatment with methotrexate (MTX; 13 mg/m2/week by mouth) and prednisone (0.5 mg/kg/day) was added, with significant improvement.

Figure 2.

A, Computed tomography imaging of the patient. Note the perforation of the nasal septum and significant erosions of the medial wall of the maxillary sinuses. B, Photograph of the patient showing the saddle nose deformity.

Six months after rituximab retreatment and 27 months after the diagnosis of WG, on a routine visit for pulmonary function tests she was found to have a moderate cough, mostly at night, and temperatures of up to 38.5°C with chills and loss of energy and appetite. At that time, her treatment consisted of MTX (13 mg/m2/week by mouth) and prednisone at a dosage of 0.3 mg/kg/day. Pulmonary function tests showed a significant reduction in the forced expiratory volume in 1 second from 60% previously to 43%, as well as in the total lung capacity from 98% previously to 66% of the predicted value for age. A chest radiograph demonstrated increased bilateral opacification with a consolidation in the lingula (Figure 3A). Laboratory investigations revealed a slightly elevated WBC of 10.7 × 109/mm3 (neutrophils 8.04 × 109/liter and total lymphocyte count of 1.66 × 109/liter), a high CRP level of 60.2 mg/dl, and a mildly raised ESR of 31 mm/hour. Flow cytometry showed no CD19+ or CD20+ B cells with a normal CD4 count of 1,580/μl. IgG was slightly decreased at 6.0 gm/dl.

Figure 3.

A, Chest radiograph of the patient on admission showing bilateral basilar increased density. B, Computed tomography imaging of the patient 3 days later showing multiple diffuse ground-glass opacities.

She was admitted and treated with intravenous piperacillin tazobactam and gentamicin for a presumed bacterial pneumonia. MTX was held during the admission and restarted afterward. She improved with this treatment and was discharged 2 days later on oral cefuroxime axetil. Cultures of blood and sputum were negative. Two days later she was readmitted with high fevers and worsening respiratory distress. Despite resuming parenteral antibiotic treatment, she continued to require oxygen at a rate of 2 liters/minute. A chest CT scan revealed multiple ground-glass opacities bilaterally (Figure 3B). Bronchoalveolar lavage was performed, which demonstrated normal bronchial architecture, but yielded a positive result for P jiroveci by calcofluor staining in both the right middle lobe and the left lingula. She was started on intravenous cotrimoxazole and was switched to oral cotrimoxazole (both 20 mg/kg of trimethoprim daily) after 1 week. She subsequently recovered well, and after completing a 3-week course of oral cotrimoxazole, she was continued on prophylactic dosages of cotrimoxazole (4.5 mg/kg of trimethoprim daily) and prednisone.


To our knowledge, we report the first case of a child with WG who developed PCP within 6 months of treatment with rituximab. At the time of infection, she was receiving concurrent therapy with MTX and prednisone. She was B cell depleted as detected by flow cytometry. Total lymphocyte counts, immunoglobulin levels, and CD4 cell counts were within normal limits. She had previously received a 6-month course of intravenous cyclophosphamide. Cotrimoxazole was given while on cyclophosphamide and for a short term around the second course of rituximab. Repeated invasive evaluations for infections, including PCP or fungi, had previously yielded negative results. On admission, the patient was severely ill, as illustrated by the associated hypoxia. After confirming the diagnosis of PCP and institution of appropriate treatment, the patient recovered completely.

A total of 34 patients with PCP following rituximab have been reported in the literature to date: 26 (76%) of these were adult lymphoma patients receiving a chemotherapy protocol containing high-dose cyclophosphamide. Seven of the 8 remaining patients were treated with significant immunosuppression for autoimmune hemolytic anemia (AIHA), complications of renal transplantation, or refractory pemphigus. Among these, the only pediatric patient reported was a 14-year-old boy with AIHA, who was treated with rituximab in addition to cyclophosphamide and vincristine, among other medications (4). A 53-year-old man with rheumatoid arthritis and pulmonary emphysema treated with MTX and low-dose prednisolone, similar to the patient described in this report, developed fatal PCP within a month of receiving 2 doses of rituximab (5).

WG itself seems to confer an increased risk for PCP. In a large survey of hospitalized adult patients, PCP occurred at a rate of 89 cases in 10,000 hospitalizations of patients with WG per year. This compares with only 27 cases/10,000 in patients with inflammatory myopathy, and 12 and 2 cases in patients with systemic lupus erythematosus and rheumatoid arthritis, respectively (2). Treatment with cyclophosphamide has been linked to an increased incidence of PCP in patients with connective tissue disorders (6). Increased incidences were also demonstrated for patients treated with MTX and high-dose prednisone, predominantly in association with lymphopenia (7). In the patient described here, a role of MTX and prednisone in the development of PCP cannot be excluded, despite normal lymphocyte counts.

Chemoprophylaxis with cotrimoxazole has been shown to be highly effective in preventing the occurrence of PCP in adults and children. However, it is potentially associated with side effects, including bone marrow suppression and Stevens-Johnson syndrome, which preclude its unrestricted use (8). Current common usage and recent recommendations from the European League Against Rheumatism suggest institution of chemoprophylaxis in patients with WG when receiving cyclophosphamide or similar second-line immunosuppressive agents (9). This correlates well with a recent risk–benefit analysis, which cites a risk of occurrence of PCP in children and adults of >3.5% being sufficient to institute chemoprophylaxis (10). Meta-analysis of an unselected WG patient cohort containing both pediatric and adult cases found the risk to be close to this threshold (8).

Rituximab may directly interfere with cellular immunity. Interactions between B and T cells are complex, and B cells seem to play an integral part in the defense mechanisms against Pneumocystis. It has long been known that B cell–deficient mice are susceptible to murine P jiroveci infections with a high mortality (11). Mice with specific B cell defects in CD40 or the major histocompatibility complex, impairing B and T cell signaling, are unable to clear P jiroveci infections (12, 13). In addition, there are several case reports of PCP in patients with X-linked agammaglobulinemia (14). At the same time, it has been shown that CD20 is expressed on up to 6% of T cells, and that these T cells are also depleted by rituximab (15).

Although there are only a few reports of patients with PCP following rituximab treatment so far, there is strong laboratory evidence to suggest an important role of B cells in the defense against this pathogen. Only a small number of pediatric or adult patients with WG have been treated with rituximab to date; therefore, even a single case of PCP represents a significant concern. In the absence of evidence to the contrary, patients treated with rituximab should currently be considered at an increased risk for PCP. In patients with WG, PCP chemoprophylaxis should be instituted either continuously or, at the very least, while B cell numbers are depressed.


All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be submitted for publication. Dr. Benseler had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Hugle, Solomon, Harvey, James, Wadhwa, Amin, Bell-Peter, Benseler.

Acquisition of data. Hugle, Bell-Peter, Benseler.

Analysis and interpretation of data. Hugle, Solomon, Harvey, James, Wadhwa, Amin, Bell-Peter, Benseler.