Association of functional (GA)n microsatellite polymorphism in the FLI1 gene with susceptibility to human systemic sclerosis.

OBJECTIVE: Susceptibility genes that can account for characteristic features of SSc such as fibrosis, vasculopathy and autoimmunity remain to be determined. In mice, deficiency of Friend leukaemia integration 1 transcription factor (Fli1) causes SSc-like disease with these features. The human FLI1 gene contains (GA)n microsatellite, which has been shown to be associated with expression level. Because microsatellite polymorphisms are difficult to capture by genome-wide association studies, we directly genotyped FLI1 (GA)n microsatellite and examined its association with SSc. METHODS: Genomic DNA from 639 Japanese SSc patients and 851 healthy controls was genotyped for (GA)n microsatellite using the fragment assay. The cut-off repeat number for susceptibility to SSc was determined by receiver operating characteristics (ROC) analysis. Association with susceptibility and clinical characteristics was examined using logistic regression analysis. FLI1 mRNA levels were determined using quantitative RT-PCR. RESULTS: Based on the ROC analysis, (GA)n alleles with ≥22 repeats were collectively defined as L alleles and alleles with ≤21 repeats as S alleles. (GA)n L alleles were significantly associated with susceptibility to SSc (P = 5.0e-04, odds ratio 1.34, additive model). Significant association was observed both in diffuse cutaneous and limited cutaneous SSc. Among the SSc, (GA)n L alleles were significantly enriched in the patients with a modified Rodnan total skin thickness score ≥10 compared with those with a score <10. FLI1 mRNA levels were significantly decreased in healthy controls carrying (GA)n L alleles as compared with non-carriers. CONCLUSION: Extended repeat alleles of FLI1 (GA)n microsatellite may be associated with lower FLI1 mRNA levels and susceptibility to human SSc.

Utility of the neutrophil-to-lymphocyte ratio for predicting bacterial infection in patients with rheumatoid arthritis receiving Tocilizumab.

This study aimed to describe the utility of the neutrophil-to-lymphocyte ratio (NLR) for predicting bacterial infections in patients with rheumatoid arthritis (RA) treated with Tocilizumab (TCZ). We extracted RA patients treated with TCZ in whom an infection developed between April 2008 and March 2018 from our hospital database. We divided these patients into the bacterial infection and non-bacterial infection groups and compared their background, C-reactive protein (CRP) values, white blood cell count (WBC), the NLR at the time of infection diagnosis, and the ratio of the NLR at the time of infection diagnosis (post-NLR) to the NLR at baseline (pre-NLR). Of the 196 patients who received TCZ, 21 experienced a bacterial infection and 20 had a non-bacterial infection. The median CRP level, WBC count, post-NLR, and post-NLR/pre-NLR ratio in the bacterial infection group were significantly higher than in the non-bacterial infection group. In receiver operating characteristics (ROC) curve analysis for predicting bacterial infection, the area under the curve (AUC) for CRP, WBC, NLR, and the post-NLR/pre-NLR ratio were 0.787, 0.857, 0.887, and 0.975, respectively. The cut-off value of 2.25 for the post-NLR/pre-NLR ratio showed the greatest sensitivity (90.5%) and specificity (100%). The post-NLR/pre-NLR ratio may be a useful surrogate marker for predicting bacterial infections in patients with RA treated with TCZ.

Human leukocyte antigen in Japanese patients with idiopathic inflammatory myopathy.

Objective: The human leukocyte antigen (HLA) is the strongest genetic risk factor for idiopathic inflammatory myopathy (IIM), and different HLA alleles have been reported to be associated with IIM susceptibility among different ethnic groups. In this study, we have investigated HLA alleles associated with IIM in Japanese patients.Methods: Genotyping of HLA-DRB1 and DPB1 were performed in 252 Japanese IIM patients (166 dermatomyositis [DM] and 86 polymyositis [PM] patients) and the association was analyzed with comparison to controls (n = 1026 for DRB1 and n = 413 for DPB1).Results: DRB1*08:03 was associated with IIM (p = 1.60 × 10-5, pc = .0005, odds ratio [OR] 2.11, 95% confidence interval [CI] 1.52-2.92) and DM (p = .0004, pc = .0128, OR 2.06, 95%CI 1.40-3.02). DPB1*05:01 was also associated with IIM (p = .0001, pc = .0021, OR 1.96, 95%CI 1.38-2.77) and DM (p = .0005, pc = .0075, OR 2.05, 95%CI 1.37-3.08). DRB1*09:01 (p = .0012, pc = .0368, OR 0.35, 95% CI 0.18-0.69) and DPB1*04:01(p = .0004, pc = .0057, OR 0.05, 95% CI 0.00-0.85) were protectively associated with PM. Two locus analyses suggested that DRB1*09:01 and DPB1*04:01 were independently associated with PM.Conclusion: Protective associations of HLA were detected in Japanese PM patients.

Characteristics and risk factors of an emergency department visit in patients with systemic lupus erythematosus.

This study aimed to examine the characteristics of patients with systemic lupus erythematosus (SLE) visiting the emergency department (ED) and the risk factors of an ED visit by these patients. This 4-year retrospective study was performed at a tertiary care center in Japan. We included all 205 patients with SLE who were treated in our outpatient clinic between April 1, 2008 and March 31, 2012 and divided them into two groups: those who visited the ED (the ED-user group) and those who did not (the ED-non-user group). We statistically compared the patient backgrounds and characteristics of the groups and identified the risk factors of an ED visit. Of all the patients, 118 visited the ED during study period and 87 did not. In total, 269 events were identified in the ED-user group. Of these, 91 (33.8%) were cases of infection, 32 (11.9%) were orthopedic problems, 32 (11.9%) were cases of gastrointestinal disease, 31 (11.5%) were cases of neurological disease, and 25 (9.3%) were cardiovascular events. Twenty-four events (8.9%) were due to SLE flares, of which ten (41.7%) were cases of neuropsychiatric lupus (NPSLE). The glucocorticoid dosage and the presence of a psychiatric illness, NPSLE, and lupus nephritis were higher among the ED-user group. Multivariate logistic regression analysis demonstrated high glucocorticoid dosage to be a risk factor of an ED visit. Among SLE patients, infections were the principal reason for visiting the ED. The most common reasons for an ED visit were common diseases rather than flares.

Effect of sulfasalazine use on the presence of Pneumocystis organisms in the lung among patients with rheumatoid arthritis: A test-negative design case-control study with PCR tests.

OBJECTIVE: To evaluate the effect of sulfasalazine (SSZ) on the presence of Pneumocystis jirovecii (P. jirovecii) in the lungs of rheumatoid arthritis (RA) patients. METHODS: We retrospectively studied episodes of suspected P. jirovecii pneumonia (PJP) which were examined for P. jirovecii with polymerase chain reaction (PCR). We employed a test negative design case-control study; the cases were episodes of suspected PJP that were positive for PCR, and the controls were episodes of suspected PJP that were negative for PCR. The odds ratio for the positive PCR result associated with SSZ use was estimated by Firth's logistic regression. RESULTS: Between 2003 and 2017, 36 cases and 83 controls were identified. While none of the cases received SSZ before the episode, 18 of the controls received the drug. In the primary analysis involving all the episodes, SSZ use was negatively associated with PCR positivity (adjusted odds ratio, 0.087; confidence interval, <0.001-0.789). The sensitivity analysis, excluding those who received PJP prophylaxis, showed the same association as the primary analysis (adjusted odds ratio 0.085, 95% CI <0.001-0.790). CONCLUSION: This study demonstrated that SSZ use is associated with the absence of P. jirovecii in the lung, suggesting the preventive efficacy of the drug against PJP.

Comparison of the clinical characteristics and severity of community-acquired pneumonia between patients with rheumatoid arthritis treated with tocilizumab and those treated with TNF inhibitor.

Objective: To examine the clinical characteristics and severity of community-acquired pneumonia (CAP) between patients with rheumatoid arthritis (RA) treated with tocilizumab (TCZ) and those treated with TNF inhibitors. Methods: We extracted RA patients treated with biological DMARDs who developed CAP between 2003 and 2015 from our hospital database. We compared the patient backgrounds, duration from the onset of symptoms to diagnosis, and the severity of CAP between patients who developed CAP after treatment with TCZ or tumor necrosis factor (TNF) inhibitor. Results: Of 98 patients who received TCZ, seven developed CAP (IL-6 inhibitor group). Of 560 patients who received TNF inhibitors, 27 developed CAP (TNF inhibitor group). Between the two groups, there was no difference in the duration from the onset of symptoms to diagnosis (7 [4-21], 7 days [1-15]). The IL-6 inhibitor group had a lower body temperature (36.5 °C [36.4-36.8], 37.8 °C [35.9-40.5]) and CRP level (0.09 mg/dL [0.02-2.5], 6.76 mg/dL [0.63-15.2]) at diagnosis than the TNF inhibitor group. The CURB-65 score did not differ significantly between groups. Conclusion: There were no delays in the diagnosis of CAP or any difference in the severity of CAP between patients with RA treated with TCZ and those treated with TNF-inhibitors.

Association of human leukocyte antigen alleles with chronic lung diseases in rheumatoid arthritis.

OBJECTIVES: Chronic lung diseases including interstitial lung disease and airway disease (AD) occur in RA patients. Interstitial lung disease and AD in RA are extra-articular manifestations that influence the prognosis quoad vitam of RA. Studies on associations of HLA alleles with RA have been carried out, and shared epitopes of several alleles are reported to be associated with RA susceptibility. Few association studies in RA subpopulations with chronic lung diseases have been conducted. The aim of the study was to identify HLA alleles predisposing to RA phenotypes including the presence of AD. METHODS: Associations of HLA-DRB1 and DQB1 alleles with chronic lung diseases in RA were analysed. RESULTS: A positive association was found between the DR4 serological group and resistance to usual interstitial pneumonia [P = 0.0250, odds ratio (OR) 0.62, 95% CI: 0.41, 0.93]. The DR2 serological group was associated with susceptibility to usual interstitial pneumonia (P = 0.0036, OR = 1.86, 95% CI: 1.23, 2.81). An association was found for shared epitopes alleles with bronchiolitic AD (P = 0.0040, OR = 2.06, 95% CI: 1.24, 3.41). DQB1*03:01 was associated with bronchiectatic AD (P = 0.0021, corrected P-value (Pc) = 0.0315, OR = 1.99, 95% CI: 1.30, 3.06), as well as with emphysema (P = 0.0007, Pc = 0.0104, OR = 2.43, 95% CI: 1.49, 3.95). In combined analysis, a predisposing association of DQB1*03:01 (P = 1.94 ×10(-5), Pc = 0.0003, OR = 2.16, 95% CI: 1.53, 3.06) and a negative association of DQB1*03:02 (P = 0.0008, Pc = 0.0117, OR = 0.33, 95% CI: 0.17, 0.67) with bronchiectatic AD or emphysema were observed in RA. CONCLUSION: The present study identified an association of HLA-DQB1*03:01 with predisposition to, and DQB1*03:02 with resistance to, bronchiectatic AD or emphysema in RA.

Severe Thrombocytopenia Induced by First Infliximab Administration for Rheumatoid Arthritis.

Thrombocytopenia due to antitumor necrosis factorα agents is very rare. A 68-year-old woman with rheumatoid arthritis on methotorexate received infliximab (IFX). Three days after the first IFX infusion, she developed gingival bleeding, petechia, and gross hematuria. Her platelet count fell to 2000/µL. We administered a platelet transfusion and intravenous methylprednisolone. Three days after admission, her platelet count was 7000/µL and her bleeding persisted. After double filtration plasmapheresis, her bleeding stopped and her platelet count recovered over 2 weeks. Thrombocytopenia is a rare but severe complication of IFX. Double filtration plasmapheresis may be useful for removing IFX or possible antibodies against platelets when IFX remaining in the patient's blood interferes with improvement of the patient's condition.

MMP-3 can distinguish isolated PMR from PMR with GCA: A retrospective study regarding PMR and GCA in Japan.

OBJECTIVE: We sought to identify clinical features at diagnosis that can distinguish isolated polymyalgia rheumatica (PMR) without giant cell arteritis (GCA) from PMR with GCA, and clinical features at diagnosis of isolated PMR that can predict subsequent relapse and corticosteroid discontinuation. METHODS: A retrospective study of 115 patients with isolated PMR and 29 patients with GCA was performed. A comparison between isolated PMR patients, GCA patients (with or without PMR), and PMR with GCA patients was performed. Predictors of relapse and corticosteroid discontinuation were identified using a logistic regression in the patients with isolated PMR. RESULTS: Matrix metalloproteinase-3 (MMP-3) level was significantly different among the patient groups. MMP-3: 230.5 ± 201.5 ng/mL in isolated PMR, 80.5 ± 47.5 ng/mL in GCA (p < 0.01), and 96.8 ± 54.8 ng/mL in PMR with GCA (p = 0.03). In the patients with isolated PMR, female gender (odds ratio [OR], 2.73; 95% confidence interval [CI], 1.16-6.41; p < 0.05) and creatinine (Cr) < 50 µmol/L (OR, 2.48; 95% CI, 1.02-5.99; p < 0.05) were significant prognostic factors that predicted relapse. CONCLUSION: A low level of MMP-3 is an excellent positive predictor for PMR with GCA. Among patients with isolated PMR, female gender and Cr < 50 µmol/L were significant prognostic factors that predicted relapse.

Transgastric long tube placement following percutaneous endoscopic gastrostomy for severe chronic intestinal pseudo-obstruction related to systemic sclerosis.

Medical management of systemic sclerosis (SSc)-associated chronic intestinal pseudo- obstruction (CIPO) has often proved inadequate. Percutaneous endoscopic colostomy (PEC) has been proposed as a method of treatment, but it is associated with a relatively high incidence of serious complications. We report herein a very severe case of SSc-associated CIPO in which complications were successfully controlled by long tube placement via a gastrostomy. Transgastric long tube placement may offer a relatively safe alternative to PEC in treating severe SSc-associated CIPO.

Validation of RAPID3 using a Japanese version of Multidimensional Health Assessment Questionnaire with Japanese rheumatoid arthritis patients: characteristics of RAPID3 compared to DAS28 and CDAI.

OBJECTIVES: To validate Routine Assessment of Patient Index Data 3 (RAPID3) using a Japanese version of Multidimensional Health Assessment Questionnaire (MDHAQ) with Japanese rheumatoid arthritis (RA) patients and to describe the characteristics of RAPID3 by comparison with Disease Activity Score 28 (DAS28) and Clinical Disease Activity Index (CDAI). METHODS: The original MDHAQ was translated into Japanese with minor cultural modifications and was translated back in English. Test-retest reliability was evaluated in 50 Japanese RA patients and further validation was performed in 350 Japanese RA patients recruited by seven rheumatologists. RAPID3, CDAI, and DAS28 were assessed on two consecutive visits. RESULTS: The test-retest reliability and the internal reliability of RAPID3 were excellent. Spearman's correlation coefficients between RAPID3 score versus CDAI score and DAS28 score were 0.761and 0.555. However, the agreement measured by kappa (weighted) for RAPID3 category versus CDAI category and for RAPID3 category versus DA28 category were 0.225 (0.382) and 0.187 (0.336). The sensitivity and specificity of "RAPID3 ≤ 3 and swollen joint ≤ 1" for predicting Boolean remission were 90.0% and 93.4%, respectively. CONCLUSIONS: RAPID3 obtained by Japanese MDHAQ was validated with Japanese RA patients and the remission criteria were found to have excellent clinical utility in usual care.

Late-onset scleroderma renal crisis induced by tacrolimus and prednisolone: a case report.

Scleroderma renal crisis (SRC) is a rare complication of systemic sclerosis (SSc) but can be severe enough to require temporary or permanent renal replacement therapy. Moderate to high dose corticosteroid use is recognized as a major risk factor for SRC. Furthermore, there have been reports of thrombotic microangiopathy precipitated by cyclosporine in patients with SSc. In this article, we report a patient with SRC induced by tacrolimus and corticosteroids. The aim of this work is to call attention to the risk of tacrolimus use in patients with SSc.

Effect of Recent Antirheumatic Drug on Features of Rheumatoid Arthritis-Associated Lymphoproliferative Disorders.

OBJECTIVE: In this study, we examine how advancements in novel antirheumatic drugs affect the clinicopathologic features of lymphoproliferative disorder (LPD) in patients with rheumatoid arthritis (RA). METHODS: In this multicenter study across 53 hospitals in Japan, we characterized patients with RA who developed LPDs and visited the hospitals between January 1999 and March 2021. The statistical tools used included Fisher's exact test, the Mann-Whitney U-test, the log-rank test, logistic regression analysis, and Cox proportional hazards models. RESULTS: Overall, 752 patients with RA-associated LPD (RA-LPD) and 770 with sporadic LPD were included in the study. We observed significant differences in the clinicopathologic features between patients with RA-LPD and those with sporadic LPD. Histopathological analysis revealed a high frequency of LPD-associated immunosuppressive conditions. Furthermore, patients with RA-LPD were evaluated based on the antirheumatic drugs administered. The methotrexate (MTX) plus tacrolimus and MTX plus tumor necrosis factor inhibitor (TNFi) groups had different affected site frequencies and histologic subtypes than the MTX-only group. Moreover, MTX and TNFi may synergistically affect susceptibility to Epstein-Barr virus infection. In case of antirheumatic drugs administered after LPD onset, tocilizumab (TCZ)-only therapy was associated with lower frequency of regrowth after spontaneous regression than other regimens. CONCLUSION: Antirheumatic drugs administered before LPD onset may influence the clinicopathologic features of RA-LPD, with patterns changing over time. Furthermore, TCZ-only regimens are recommended after LPD onset.

Clinical characteristics and risk factors for Pneumocystis jirovecii pneumonia in patients with rheumatoid arthritis receiving adalimumab: a retrospective review and case-control study of 17 patients.

OBJECTIVES: To investigate the clinical characteristics and risk factors of Pneumocystis jirovecii pneumonia (PCP) in rheumatoid arthritis (RA) patients treated with adalimumab. METHODS: We conducted a multicenter, retrospective, case-control study to compare RA patients treated with adalimumab with and without PCP. Data from 17 RA patients who were diagnosed with PCP and from 89 RA patients who did not develop PCP during adalimumab treatment were collected. RESULTS: For the PCP patients, the median age was 68 years old, with a median RA disease duration of eight years. The median length of time from the first adalimumab injection to the development of PCP was 12 weeks. At the onset of PCP, the median dosages of prednisolone and methotrexate were 5.0 mg/day and 8.0 mg/week, respectively. The patients with PCP were significantly older (p < 0.05) and had more structural changes (p < 0.05) than the patients without PCP. Computed tomography of the chest revealed ground-glass opacity without interlobular septal boundaries in the majority of the patients with PCP. Three PCP patients died. CONCLUSIONS: PCP may occur early in the course of adalimumab therapy in patients with RA. Careful monitoring, early diagnosis, and proper management are mandatory to secure a good prognosis for these patients.

Response to hydroxychloroquine in Japanese patients with systemic lupus erythematosus using the cutaneous lupus erythematosus disease area and severity index (CLASI).

We evaluated the cutaneous lupus erythematosus disease area and severity index (CLASI) in Japanese patients with systemic lupus erythematosus (SLE) in order to design a clinical trial of hydroxychloroquine (HCQ) in Japan. Our prospective cohort study consisted of seven SLE patients with active skin disease who started HCQ at Tokyo Metropolitan Tama Medical Center. The therapeutic responses were assessed at 4 months. Patients were categorized as responders (improved) or non-responders (unchanged or worsened) using the criteria of a 4-point or 20% decrease in the CLASI activity score. We also assessed joint pain determined by patient visual analog scale (VAS), malaise (VAS), patient global assessment of SLE (VAS), and constitutional and musculoskeletal symptoms according to the British Isles Lupus Assessment Group (BILAG) disease activity index. Six patients (86%) were categorized as responders. The median (range) CLASI activity score of all patients at assessment had changed from 8.0 (2-22) to 4 (2-10). All five patients with joint pain and all five patients with malaise showed improvement in patient VAS but the BILAG findings failed to capture these improvements. In conclusion, the cutaneous aspects of SLE can be measured by the CLASI. The CLASI activity score may be a reasonable primary endpoint when performing a clinical trial of HCQ.

Bucillamine-induced gigantomastia with galactorrhea and hyperprolactinaemia.

Gigantomastia is characterised by excessive breast growth and can occur as a rare, drug-induced adverse event. D-penicillamine is the most frequent cause of drug-induced gigantomastia. Only one case of gigantomastia due to bucillamine, an analogue of D-penicillamine, has been reported so far. We herein report a case of bucillamine-induced gigantomastia presenting with acute enlargement of the bilateral breasts and accessory breast tissue in the axillae 7 months after the start of bucillamine therapy. Awareness about this rare adverse event is important since bucillamine is still widely used in Japan and Korea.

A Proposal to Standardize Low Disease Activity Criteria in Rheumatoid Arthritis Based on the Outcome Measures in Rheumatology Minimal Disease Activity Definition.

OBJECTIVE: We aimed to standardize the definition of low disease activity in rheumatoid arthritis (RA) using the Outcome Measures in Rheumatology (OMERACT) group's proposed definition of minimal disease activity. METHOD: Based on a nationwide RA database, we proposed new Boolean low disease activity criteria using OMERACT's core set definition of minimal disease activity that requires the fulfillment of at least five of the following seven core set measures: a pain score of 2 or less, a swollen joint count (SJC28) of 1 or fewer, a tender joint count of 1 or fewer, a Health Assessment Questionnaire score of 0.5 or less, a Physician's Global Assessment score of 1.5 or less, a Patient's Global Assessment score of 2 or less, and an erythrocyte sedimentation rate (ESR) of 20 mm/h or less. Using receiver operating characteristic analysis, we determined the cutoffs for the Simplified Disease Activity Index (SDAI), Clinical Disease Activity Index (CDAI), Routine Assessment of Patient Index Data 3 (RAPID3), and the Disease Activity Score in 28 joints (DAS28)-ESR. RESULTS: Of 8298 patients, 56.2% met the proposed Boolean low disease activity criteria. We determined an SDAI score of 5.5 or less and a CDAI score of 5 or less to be the new cutoffs, and we chose a DAS28 of 2.85 or less (the original cutoff for DAS-based minimal disease activity) and a RAPID3 score of 6 or less (the original cutoff for RAPID3-based low disease activity) with or without a swollen joint count (SJC) (SJC of 2 or fewer) as the cutoffs for DAS28 and RAPID3. The agreement between the new cutoffs for DAS28 of 2.85 or less vs. CDAI score of 5 or less, CDAI score of 5 or less vs. RAPID3 score of 6 or less (with SJC of 2 or fewer), and DAS28 of 2.85 or less vs. RAPID3 score of 6 or less (with SJC of 2 or fewer), was 0.619, 0.612 (0.702), and 0.474 (0.531), respectively. CONCLUSION: OMERACT's minimal disease activity definition may be used to standardize the criteria for low disease activity.

Diffuse chronic leptomeningitis with seropositive rheumatoid arthritis: report of a case successfully treated as rheumatoid leptomeningitis.

A case of biopsy-confirmed chronic leptomeningitis complicating rheumatoid arthritis in a 53-year old female is reported. Her symptoms included weight loss, severe depression, and pyrexia. Magnetic resonance imaging was useful in diagnosis. Intravenous methylprednisolone was prescribed (1 g/day for 3 days), followed by prednisolone (initial dose of 30 mg daily), and this treatment was effective. Her IgG-index, serum levels of soluble interleukin-2 receptor and ferritin, and cerebrospinal level of interleukin-6 paralleled her clinical course.

Prophylactic effect of sulfasalazine against Pneumocystis pneumonia in patients with rheumatoid arthritis: A nested case-control study.

OBJECTIVES: To evaluate the prophylactic effect of sulfasalazine against Pneumocystis jirovecii pneumonia (PJP) among rheumatoid arthritis (RA) patients. METHODS: We used a nationwide Japanese multicenter RA database to extract data from 2005 and 2014. To identify PJP cases, we selected patients hospitalized for PJP and verified their diagnosis. Two control groups, one unmatched and the other matched for age, sex, glucocorticoid, methotrexate, and tacrolimus dosage, and the use (and type, if used) of biological disease-modifying antirheumatic drug were selected by incidence-density sampling. The odds ratios for PJP associated with sulfasalazine use and other clinical factors were estimated by exact and standard conditional logistic regression. RESULTS: From 18,668 participants, 60 cases, 356 unmatched controls, and 337 matched controls were selected. None of the cases received sulfasalazine before PJP onset. A comparison of the cases with the unmatched controls showed that sulfasalazine use carried a decreased risk of PJP (adjusted odds ratio 0.18, 95% confidence interval 0.00-0.92). A comparison of the cases and matched controls also showed that sulfasalazine use had a decreased risk of PJP (0.08, 0.00-0.36). In an analysis of the cases and unmatched controls who did not receive sulfasalazine, an increased risk of PJP was associated with lung disease (3.88, 1.89-7.95) and the use of glucocorticoid (5.71, 2.68-12.19), methotrexate (5.25, 2.01-13.74), and tumor necrosis factor inhibitors (2.32, 1.10-4.93). CONCLUSIONS: The results of this nested case-control study demonstrated the preventive effect of sulfasalazine against PJP. The results await confirmation by future prospective studies.