Prophylaxis for Pneumocystis pneumonia in patients with rheumatoid arthritis treated with biologics, based on risk factors found in a retrospective study

We retrospectively analyzed 702 traceable patients with RA who had started treatment with biologic agents between April 2005 and September 2009 in the First Department of Internal Medicine, University of Occupational and Environmental Health (Fukuoka, Japan). During the first step, we compared the clinical features and laboratory data between patients who developed PCP and the remaining patients who did not receive trimethoprim-sulfamethoxazole (TMP/SMX) prophylaxis. The following clinical parameters were evaluated to identify risk factors for the development of PCP: age, the duration of RA, the Disease Activity Score in 28 joints score and erythrocyte sedimentation rate (DAS28 ESR), coexisting pulmonary disease, diabetes mellitus (DM), ratio of patients who received glucocorticoid treatment, dose of glucocorticoid, dose of methotrexate (MTX), serum immunoglobulin G (IgG) and serum KL-6. Before starting biologic therapy, all patients underwent a bidirectional chest X-ray and/or chest computed tomography (CT) to identify coexisting pulmonary disease, including interstitial pneumonia, pleuritis, diffuse panbronchiolitis, bronchiectasia, old tuberculosis and inflammatory nodules. We excluded the patients who were proven soon after the initiation of biologic agents to have complications due to either malignancy or HIV, which influence the risk for PCP development.

TMP/SMX prophylaxis was begun when each physician considered a patient to be at relatively high risk for PCP based on factors including age, dose of glucocorticoid, serum IgG level, coexisting pulmonary disease and complications of DM.

We compared the characteristics of patients in the PCP group were compared with those in patients without PCP determined risk factors for PCP development. The sensitivity and specificity of the combination of risk factors were used to plan the primary prophylactic procedure, and that procedure was applied to patients with RA who initiated treatment with biologic agents starting in October 2009.

We analyzed 214 patients who initiated biologic therapy between October 2009 and September 2010 to estimate the incidence of PCP, the continuation rate and the adverse events associated with TMP/SMX prophylaxis during the observation period. The observation period was continued after TMP/SMX withdrawal.

One tablet containing 80 mg of trimethoprim and 400 mg of sulfamethoxazole was given every day, or two tablets were given three times weekly as prophylaxis against PCP, and the doses were tapered in patients with renal dysfunction. TMP/SMX was changed to inhaled aerosolized pentamidine in patients who experienced adverse events due to TMP/SMX treatment. They inhaled aerosolized pentamidine at a dose of 300 mg once monthly.

This study is compliant with the Declaration of Helsinki, and the study protocol was approved by the Ethics Committee of the University of Occupational and Environmental Health School of Medicine. Informed consent to participate in the study and have information published was obtained from each patient.

In the study to develop the prophylactic procedure, we evaluated 702 patients with RA who were being treated with biologic therapy (281 patients treated with IFX, 197 with ETN, 117 with adalimumab (ADA) and 107 with tocilizumab (TCZ)). The patient population comprised 574 females (81.7%) and 128 males (18.3%). The mean age of the patients when they received their first biologic agent was 58.0 years (range, 14 to 86 years) (Table 1). The mean observation period was 16.6 months (range, 2 weeks to 60 months) and 141 patients (20.1%) received TMP/SMX prophylaxis. A total of 265 patients (37.7%) had coexisting pulmonary disease, 303 patients (43.2%) received glucocorticoid therapy and the mean dose (±SD) of glucocorticoid (converted to the prednisolone (PSL) equivalent) was 4.46 mg ± 3.64 mg. We included not only active lung disease but also nonactive lung lesions, including inactive pulmonary fibrosis as coexisting pulmonary disease, which resulted in the high ratio of patients with coexisting lung disease. A total of 78.5% of patients were given MTX, and the average dose of MTX in patients treated with MTX was 8.96 mg/wk.

Nine (1.28%) of the seven hundred two patients developed PCP, and none of the patients in the group treated with TMP/SMX prophylaxis developed PCP. Four of nine patients with PCP were receiving the prophylactic treatment at the initiation of biologic therapy, although all of them stopped the prophylaxis (because of side effects of prophylaxis in two patients and for unknown reasons in two patients). The other five patients were not receiving the prophylaxis during the treatment with biologics. The mean interval between the first infusion of the biologic agent and the onset of PCP was 7.17 months (range, 2 weeks to 20 months). The mean age of patients with PCP at the start of biologic therapy was 69.5 years (range, 57 to 78 years), and the mean morbidity period of RA for the PCP patients was 76.3 months (Table 2). Coexisting pulmonary disease was detected in seven (77.8%) of the nine patients. Eight patients (88.9%) who developed PCP were receiving glucocorticoid therapy at a mean dose (±SD) of 8.83 mg ± 14.9 mg. Six (66.7%) of the nine patients were given MTX, and the average dose of MTX given to them was 9.83 mg/wk. The average in the nine patients was 6.56 mg/wk.

Because TMP/SMX is known to inhibit the development of PCP almost completely, we excluded from the analysis the 141 patients treated with the prophylaxis. We first analyzed the characteristics of these 141 patients who were thought to be at high risk for the development of PCP and received the prophylaxis. Compared with the other 561 patients, the mean age in their group was significantly older (68.9 years vs. 55.3 years); the serum level of IgG was significantly lower (1,496 mg/dl vs. 1,619 mg/dl); the dose of MTX was significantly lower (6.09 mg vs. 7.28 mg); and the ratios of patients treated with glucocorticoids, PCP complicated with coexisting pulmonary disease and PCP complicated with DM were significantly higher (53.9% vs. 40.8%, 72.3% vs. 29.1% and 21.3% vs. 10.5%, respectively).

Next, we analyzed the clinical and laboratory records of the 561 patients who did not receive TMP/SMX prophylaxis (Figure 1 and Table 3). The patients in the PCP group (n = 9) were significantly older, had a higher percentage of coexisting pulmonary disease, had a higher serum level of KL-6 and were treated with a higher PSL dose than the patients in the non-PCP group (n = 552). In the PCP group, the percentage of patients treated with glucocorticoids was significantly higher than that in the non-PCP group. There were no significant differences in the duration of morbidity , RA DAS28 score, percentage of patients with DM, dose of MTX or serum IgG level between the PCP and non-PCP groups. Because of the low incidence of PCP, we could not analyze the differences associated with each biologic therapy as a risk factor for PCP. The logistic regression analysis revealed that, among patients with PCP, the hazard ratio (HR) for patients at least 65 years of age was 4.37 (95% confidence interval (CI) = 1.04 to 18.2), the HR for those with coexisting pulmonary disease was 8.13 (95% CI = 1.63 to 40.0) and the HR for use of glucocorticoids was 11.4 (95% CI = 1.38 to 90.9) (Figure 2).

We calculated the sensitivity and specificity of these risk factors as tests to represent the incidence of PCP. The group of patients with one or more risk factors for PCP includes all patients with PCP, in addition to many patients without PCP (sensitivity = 100%, specificity = 34.1%). If prophylaxis were given only to patients with three risk factors, we would have missed four of nine patients with PCP (sensitivity = 55.6%, specificity = 94.4%). On the basis of these results, we assumed that patients with two or three risk factors could benefit from prophylaxis with TMP/SMX (sensitivity = 77.8%, specificity = 76.1%). Assuming that TMP/SMX inhibits the development of PCP completely, the number needed to treat to prevent one case of PCP was 19.9 in the analysis of 561 patients.

The inclusion criteria for PCP prophylaxis were applied for 214 patients with RA for whom treatment with biologic agents was initiated between October 2009 and September 2010. The mean age of the 214 patients (67 patients treated with IFX, 51 with ETN, 64 with ADA and 32 with TCZ) was calculated for 174 females (81.3%) and 40 males (18.7%). The mean age in this group was 58.9 years (range = 15 to 89 years). The mean observation period was 5.7 months. The mean (±SD) DAS28 (ESR) score was 5.45 ± 1.27, and the mean duration of RA was 92.9 ± 119 months. A total of 94 patients (43.9%) received prophylaxis against PCP. Of the 214 patients, 141 (65.8%) had coexisting pulmonary disease and 49 (22.9%) received glucocorticoid therapy, and the mean glucocorticoid dose (converted to the PSL equivalent) was 6.28 mg ± 6.46 mg. In this group of 214 patients, 87.9% were given MTX and the average dose of MTX was 9.51 mg/wk.

Among the primary prophylactic group (94 patients), 12 patients (12.8%) received prophylaxis with inhaled aerosolized pentamidine starting from the initiation period for the following reasons: four patients had renal dysfunction, four had cytopenia and four had other conditions. Among the patients who received prophylaxis with TMP/SMX (82 patients), adverse events occurred in 11 patients (13.4%). These side effects included liver dysfunction (five patients), cytopenia (two patients), skin eruption (one patient), renal dysfunction (one patient), hyponatremia (one patient) and an unspecified other adverse event (one patient). TMP/SMX prophylaxis was changed to inhaled aerosolized pentamidine in all 11 of these patients. There were no severe adverse events that required hospital admission. The observation period was continued after TMP/SMX withdrawal.

It is noteworthy that, during the observation period, no case of PCP was observed among the patients in the second study. Compared with the results of the first analysis of 702 patients, the incidence of PCP decreased from 1.28% (9 of 702 patients) and 0.93/100 patient-years to 0.00% and 0/100 patient years, which confirms the effectiveness of the prophylactic procedure.