An open-label, randomized controlled trial of sulfamethoxazole–trimethoprim for Pneumocystis prophylaxis: results of 52-week follow-up

Abstract Objectives The aim was to investigate the long-term prophylactic efficacy, drug retention and safety of low-dose sulfamethoxazole–trimethoprim (SMX/TMP) prophylaxis against Pneumocystis pneumonia (PCP). Methods Adult patients with rheumatic diseases receiving prednisolone ≥0.6 mg/kg/day were randomized into the single-strength group (SS; SMX/TMP 400/80 mg daily), the half-strength group (HS; 200/40 mg daily) or the escalation group (ES; starting at 40/8 mg and increasing incrementally to 200/40 mg daily) and treated for 24 weeks, then observed for 52 weeks. The primary endpoint, the PCP non-incidence rate (non-IR) at week 24, has been reported previously. The secondary endpoints were the PCP non-IR at week 52, treatment discontinuation rate and adverse events. Results Fifty-eight, 59 and 55 patients in the SS, HS and ES, respectively, received SMX/TMP. PCP did not develop in any of the patients by week 52. The estimated PCP non-IR in patients receiving SMX/TMP 200/40 mg daily (HS and ES) was 96.8–100%. Throughout the 52-week observation period, the overall discontinuation rate was significantly lower in HS than in SS (22.7 vs 47.2%, P = 0.004). The discontinuation rates attributable to adverse events were significantly lower in HS (19.1%, P = 0.007) and ES (20.3%, P = 0.007) than in SS (41.8%). The IRs of adverse events requiring SMX/TMP dose reduction before week 52 differed among the three groups, with a significantly higher IR in SS than in HS or ES (P = 0.007). Conclusion SMX/TMP 200/40 mg had a high PCP prevention rate and was superior to SMX/TMP 400/80 mg in terms of drug retention and safety. Trial registration University Hospital Medical Information Network Clinical Trials Registry, UMIN000007727.


Introduction
Pneumocystis pneumonia (PCP) can have significant impacts on the clinical course of immunocompromized patients [1]. In particular, in non-HIV patients PCP can result in rapid deterioration and is more likely to have a poor prognosis [2] and should be prevented appropriately. Currently, sulfamethoxazole-trimethoprim (SMX/ TMP) is widely used as the first-line drug for PCP prophylaxis [1,3]. When used properly, the prevention rate is reportedly 89-100%. In a retrospective study of 1522 treatments of rheumatic diseases with high-dose CSs, the preventive efficacy of SMX/TMP was 93% [4]. In a study of RA patients using biologics, the PCP nonincidence rate (non-IR) was 100% in the prophylactic group and 98.4% in the non-prophylactic group [5]. However, many adverse events (AEs) related to this drug have been reported [3], which can necessitate switching to other drugs, discontinuation of prophylaxis and, eventually, development of PCP. Alternative drugs, such as pentamidine isethionate, atovaquone and dapsone, are considered to be less effective [6] and also produce various adverse drug reactions.
In PCP prophylaxis, striking a balance between benefit (i.e. prophylactic effect) and risk (i.e. AEs) is required. Guidelines for patients with HIV infection, haematological malignancies and solid organ transplantation indicate which patients are at risk for PCP and should receive prophylaxis [7][8][9][10]. Although PCP prophylaxis is indicated in some patients with rheumatic diseases [1,5,[11][12][13], no clear consensus on prophylactic regimens has been developed because of the lack of robust evidence, and no official guidelines for PCP prophylaxis exist for patients with rheumatic diseases. To minimize the risks inherent in PCP prophylaxis, reduction of the AEs of SMX/TMP using an appropriate regimen is desirable. Prasad et al. [14] reported that SMX/TMP 400 mg/80 mg three times a week in renal transplant patients reduced AEs without affecting prophylaxis. Takenaka et al. [15] showed that SMX/TMP 400 mg/80 mg daily with dose escalation was superior to SMX/TMP 400 mg/80 mg daily in terms of the continuation rate in 41 patients with rheumatic diseases who started treatment with CSs. Suyama et al. [16] retrospectively compared SMX/TMP 400 mg/80 mg daily with 400 mg/80 mg in a dose escalation regimen in 59 patients with SLE and found that the latter was safer. In our previous report, using an openlabel, randomized controlled trial, we compared the efficacy, drug retention and safety of three regimens (SMX/ TMP 400 mg/80 mg daily, 200 mg/40 mg daily or 200 mg/ 40 mg with dose escalation) in 183 patients with systemic rheumatic diseases who started at a dosage of 0.6 mg/kg/day or more of prednisolone or an equivalent dosage of CS [17]. At week 24, the 200 mg/40 mg daily regimen was superior in efficacy, drug retention and safety. Herein, we report the results of observations at week 52 to verify long-term prophylactic effects and safety of the regimens.

Patients
The inclusion and exclusion criteria of the present study were described previously [17]. In brief, the inclusion criteria were follows: age 20 years; admission to one of the participating institutions for treatment of new-onset or relapsed systemic rheumatic disease in the study period; written informed consent; an oral prednisolone starting dosage at 0.6 mg/kg/day or an equivalent dosage of CS regardless of concomitant immunosuppressive drugs; no previous use of SMX/TMP, pentamidine isethionate or dapsone; and serum creatinine values within the upper limit of the normal range according to the institutional standard. Patients were excluded if they received a biologic agent, had a history of PCP or were unable to start SMX/TMP within 10 days of starting prednisolone.

Study design
This study was an open-label, multicentre, randomized controlled trial, and the study design was described previously [17]. In brief, the patients were randomized into one of three arms at a 1 Endpoints PCP non-IRs at week 24 between SS and ES (the primary endpoint of this study) and the other comparisons between groups at week 24 have been reported previously [17]. The PCP non-IRs, treatment discontinuation rates and AEs at 52 weeks are reported here. PCP was diagnosed clinically by symptoms, laboratory tests and imaging by site investigators, with verification by the clinical event review committee, which included experts in pulmonary medicine, infectious diseases and rheumatology.

Statistical analyses
The calculated sample size was 58 patients per group, assuming the PCP non-IR of SS to be 93% and that of ES to be 98%. The non-inferiority limit was set at 5%, one-sided a at 0.05, and b at 0.20 [15,17]. For statistical analyses, treatment discontinuation rates were analysed using the Kaplan-Meier method and the log-rank test. Fisher's exact test with adjusted residuals was used to compare the incidences of AEs. As a post hoc analysis, the PCP non-IRs of each group and the combined group of HS and ES were estimated using the Clopper-Pearson exact confidence interval [18] and the rule of three [19], because PCP developed in none of the patients.

Randomization and follow-up
One hundred and eighty-three patients were randomized into one of three arms, with 58, 59 and 55 patients in SS, HS and ES, respectively, starting treatment with SMX/TMP. Twenty-nine, 43 and 34 patients in SS, HS and ES, respectively, continued the same regimen until week 52. Details are given in Supplementary Fig. S1, available at Rheumatology Advances in Practice online.
Reasons for discontinuation of the regimens were AEs, prescription errors or the investigators' discretion. All the patients were followed for 52 weeks, except those who died or were transferred to another hospital.
Baseline characteristics of the patients Supplementary

Efficacy and drug discontinuation rate
No PCP cases were reported up to week 52, and the PCP non-IR was estimated by using the Clopper-Pearson exact confidence interval in post hoc analysis [18]. The estimated non-IR at week 52 in SS, HS and ES was 93.8-100, 93.9-100 and 93.5-100%, respectively. Given that the patients in HS and ES received 200 mg/ 40 mg SMX/TMP daily for 52 and 47 weeks, respectively, we combined these two groups and estimated the PCP non-IR of the combined group to be 96.8-100% (n ¼ 114). Estimation using the rule of three showed almost the same results [19]. The cumulative discontinuation rates due to any reason are shown in Fig. 1a using the Kaplan-Meier curves. HS showed a significantly lower cumulative discontinuation rate than SS (47.2 vs 22.7%, P ¼ 0.004), which also tended to be lower in ES (31.6%) than in SS (P ¼ 0.059). Fig. 1b

Discussion
In the present study, PCP did not develop in any of the patients in the 52-week observation period. The accumulated overall discontinuation rate for 52 weeks was significantly lower in HS than in SS, suggesting that the half-strength regimen had the better benefit-risk balance. A number of reasons might account for the lack of development of PCP in the present study. The physicians in charge at the participating institutions were proficient in treating rheumatic diseases and were knowledgeable about PCP prophylaxis. Many patients continued PCP prophylaxis even after stopping the allocated treatment with SMX/TMP; the numbers of patients without prophylaxis were 6 in weeks 0-12, 10 in weeks 12-24, and 17 in weeks 24-52. The lack of development of PCP in this study is understandable in view of the fact that the incidence of PCP without prophylaxis is reportedly 2.3-8.97% [5] and that the median prednisolone dosage administered at week 24 in this study was <15 mg/day, at which discontinuation of prophylaxis may be considered safely [4]. Some differences were observed between HS and ES in the drug discontinuation rates. The cumulative discontinuation rates by AEs did not differ significantly between HS and ES; the observed AEs, such as thrombocytopenia, elevated liver function test and hyponatraemia, were mostly dose dependent, with the exception of rash. This breakdown of AEs is consistent with previous reports [15,16]. In contrast, HS had a lower drug discontinuation rate for all reasons than ES, mainly resulting from prescription errors in ES. In patients with allergy to multiple drugs, SMX/ TMP may be administered under the escalation regimen. Recently, Suyama et al. [16] reported that patients with SLE, especially those with positive anti-Ro/SSA antibodies, had a higher incidence of adverse drug reactions to SMX/TMP. If true, the escalation regimen might be a better treatment option for these patients.
This study has some limitations. First, there was a detection bias attributable to nonblinding. The investigators might have expected more AEs in SS. Second, the efficacy and safety of a follow-up period >52 weeks are unknown. However, the median time from the start of treatment with CS to the onset of PCP was reportedly 12 weeks [20], and a 52-week observation period appeared to be sufficient in this context. Third, patients with reduced renal function and patients with low body weight were excluded. In addition, >60% of the patients were female; the mean body weight of the patients was <60 kg in all three groups, and only five patients weighed >80 kg. Fourth, not a large number of patients were enrolled in this study. Fifth, we did not test an alternate-day regimen in this study. This could be a challenge for the future.

Conclusion
In patients with the characteristics included in the present study, SMX/TMP dosage reduction decreased the cumulative drug discontinuation rate and AEs requiring SMX/TMP dosage reduction. In patients with normal serum creatinine concentrations and similar body weights to those enrolled in the present study, SMX/TMP 200 mg/40 mg might provide a favourable benefit-risk balance in PCP prophylaxis.