Abstract
Objectives
To investigate total and cause-specific mortality risk among rheumatoid arthritis-associated lung disease (RA-LD) compared with RA patients without lung disease (RA-no LD).
Methods
We conducted a retrospective cohort study using the Mass General Brigham Biobank (Boston, MA), comparing RA-LD cases to RA-no LD comparators matched by calendar date, age, sex and RA duration. RA-LD was verified by medical record review and chest imaging for clinically apparent RA-associated interstitial lung disease (RA-ILD) and/or RA-associated bronchiectasis (RA-BR). Outcomes included total and cause-specific mortality (respiratory, cancer, cardiovascular, infection). Total mortality was analysed using Cox regression, and cause-specific mortality was analyzed using Fine and Gray models estimating subdistribution hazard ratios (sdHR) and 95% confidence intervals (CI), adjusted for propensity scores.
Results
We analyzed 221 RA-LD cases (151 RA-ILD and 70 RA-BR) and 980 RA-no LD comparators. RA-LD patients had higher total mortality (34.0 vs 13.1 per 1000 person-years; HR 1.97, 95%CI 1.43–2.70). Infection-related mortality was more common in RA-LD patients (7% vs 3%, P = 0.002). RA-LD was associated with increased respiratory (13.1 vs 1.7 per 1000 person-years; sdHR 6.68, 95%CI 3.19–14.00) and lung infection-related mortality (6.3 vs 1.4 per 1000 person-years; sdHR 3.33, 95%CI 1.32–8.37), particularly in RA-ILD patients. RA-BR patients had higher cancer mortality (8.6 vs 2.3 per 1000 person-years; sdHR 2.89, 95%CI 1.11–7.47).
Conclusion
RA-LD, particularly RA-ILD, is associated with higher mortality, especially respiratory and infection-related mortality. RA-BR is associated with increased respiratory and cancer mortality. While larger studies are needed, these findings emphasize the need for improved treatments for RA-LD, effective infection prevention, and vigilant cancer surveillance.
Keywords: rheumatoid arthritis, interstitial lung disease, bronchiectasis, mortality
Rheumatology key messages.
Patients with rheumatoid arthritis-associated lung disease (RA-LD) had a 2-fold increased mortality risk.
Respiratory disease was the leading cause of death among RA-LD patients, with infection-related mortality, particularly from lung infections, viral infections and SARS-CoV-2 being prominent.
RA-associated bronchiectasis patients had a 3-fold higher risk of respiratory and cancer mortality.
Introduction
Rheumatoid arthritis-associated lung disease (RA-LD), which includes RA-associated interstitial lung disease (RA-ILD) and RA-associated bronchiectasis (RA-BR), is a notable extra-articular manifestation of rheumatoid arthritis (RA) [1, 2]. Previous studies [3, 4] have indicated patients with RA have about a 2-fold increased risk of total mortality compared with the general population, primarily due to cardiovascular diseases, cancer and respiratory causes. RA-ILD, in particular, is associated with about 3-fold higher mortality [5–9] than patients with RA without ILD. There have been fewer studies investigating RA-BR [10, 11] but this has also exhibited reduced survival compared with RA without BR or BR-only patients. However, few comparative studies [8, 9] investigated detailed analyses of specific causes of death, particularly infection-related mortality, and provided limited information on anatomic sites or pathogens of infection. Most studies did not account for competing risks from alternative causes of death, such as cardiovascular disease or cancer [7].
Understanding cause-specific mortality is essential for assessing the true burden of disease in RA-LD patients. Infections represent a leading cause of death among patients with RA, but the specific anatomic sites and pathogens involved in these deaths remain unclear. Due to chronic inflammation and immunosuppressive treatments, RA-LD patients are particularly vulnerable to infections [12, 13], which may complicate the clinical course and play a more prominent role in mortality compared with other causes [14].
Therefore, we aimed to investigate total and cause-specific mortality among patients with RA-LD by leveraging data from the Mass General Brigham (MGB) Biobank. In this cohort study, we compared patients with RA-LD cases to RA-no LD comparators and further investigated the distinct contributions of RA-ILD and RA-BR to mortality. We hypothesized that patients with RA-LD cases would have a higher mortality rate compared with RA-no LD comparators. Further, we hypothesized that both RA-ILD and RA-BR would be associated with higher total and cause-specific mortality, particularly for respiratory and infection-related mortality.
Methods
Study design and participants
This retrospective comparative cohort study utilized data from the MGB Biobank, a prospective research study consisting of banked biospecimens, surveys and linked electronic health record (EHR) data from over 150 000 participants in the Greater Boston area hospitals since 2010 [15]. The MGB institutional review board approved this sub-study (Protocol #2019P000264), ensuring compliance with the Declaration of Helsinki. All participants provided written informed consent.
Identification of RA
Patients with RA in the MGB Biobank were initially identified through a validated algorithm combing diagnosis codes, laboratory results and natural language processing of EHR data, which demonstrated a 95% positive predictive value and 97% specificity for RA [16]. Medical record review was then conducted to confirm that all participants met the 2010 American College of Rheumatology/European Alliance of Associations for Rheumatology (ACR/EULAR) classification criteria [17] and to obtain the date of clinical RA diagnosis.
RA-LD cases
RA-LD cases were identified as clinically apparent RA-ILD and/or RA-BR, following criteria established in previous studies [18, 19]. Potential RA-LD cases were screened using billing codes for ILD and/or BR, chest computed tomography (CT) imaging or lung biopsy/procedure. For these patients, detailed medical record review was conducted to confirm clinical evidence of RA-ILD or RA-BR. Additionally, all chest CT images underwent review by at least two expert chest radiologists. RA-LD cases had findings on chest CT and/or lung biopsy consistent with ILD or BR [20]. Clinical evidence of ILD or BR was established based on a physician’s clinical diagnosis, or characteristic symptoms including cough, sputum production and/or dyspnea, along with imaging evidence of ILD/BR on chest CT scans. These imaging findings were scored independently by two thoracic radiologists following a previously described scoring system [21, 22]. Discrepancies were resolved by consensus or a third radiologist. Further details of the scoring system are available in the previous study [19]. We performed medical record review to determine the date of clinical diagnosis for RA-ILD or RA-BR.
Because ILD can cause traction BR, patients with both ILD and BR were classified as RA-ILD. RA-BR cases were defined as isolated BR [19] without any clinical or imaging evidence of ILD in the EHR or CT scans.
The primary analysis combined all RA-ILD and RA-BR cases into a single RA-LD cohort. In secondary analyses, we investigated RA-ILD and isolated RA-BR separately. Based on imaging patterns [18], we further divided RA-ILD into those with the usual interstitial pneumonia (UIP) pattern vs other RA-ILD patterns, as well as fibrotic (UIP or fibrotic nonspecific interstitial pneumonia [NSIP]) vs non-fibrotic (cellular NSIP or all other patterns).
Matched RA-no LD comparators
The pool of possible RA-no LD comparators included the remaining patients that did not meet the criteria for RA-ILD or RA-BR, ensuring they had no clinical or imaging evidence of ILD or BR.
Each RA-LD case was matched with up to five RA-no LD comparators based on calendar date of birth (±5 years), sex and RA duration (±5 years) at the index date, which was defined as the later of RA or ILD/BR diagnosis for RA-LD cases. For comparators, we assigned them the same index date as the corresponding RA-LD case. All RA-LD cases and RA-no LD comparators were required to have at least one clinical encounter after the index date.
Mortality outcomes
The primary outcomes were total and cause-specific mortality, with causes categorized as respiratory, cancer, cardiovascular, infection, other or unknown. Date of death was obtained through linkage with the National Death Index, and missing dates were verified through medical record review. Causes of death were determined through medical record review conducted by at least two rheumatologists, who independently assessed the clinical documentation. Any discrepancies in cause classification were resolved by a third rheumatologist with expertise in the area. For infection-related deaths, the anatomic sites, pathogen types and specific pathogens were recorded. Of note, death certificates were not available on all patients, so we used the clinical documentation to ascertain the primary cause of death using clinical judgment from the medical records.
Covariates
Demographic data including age at index date, sex, race and Hispanic ethnicity were obtained from electronic query. Smoking status (never/past/current) and pack-years were collected from the MGB Biobank survey, with missing smoking data supplemented from medical record review. Comorbidities within 365 days prior to the index date, including diabetes, heart failure, myocardial infarction or stroke, cancer from solid tumor, leukaemia or lymphoma were assessed using the Charlson Comorbidity Index (CCI) [23, 24] based on ICD codes to quantify the overall comorbidity burden. We verified cancer presence at index date through medical record review, as codes can be inaccurate to determine presence of active cancer. Chronic obstructive pulmonary disease (COPD) and connective tissue disease components were excluded from the CCI calculation, as many RA-LD patients may have had COPD codes misclassified and all patients had RA.
We collected RA characteristics, including RA duration at index date, serostatus, anti-cyclic citrullinated peptide antibodies (ACPA) status, rheumatoid factor (RF) status, the most recent erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) within 365 days before index date, as well as the use of disease-modifying antirheumatic drugs (DMARDs) and glucocorticoids. The specific DMARDs were categorized by mechanism of action and classified as conventional synthetic DMARDs (csDMARDs), biologic DMARDs (bDMARDs) or targeted synthetic DMARDs (tsDMARDs).
Lung disease characteristics included the presence of ILD, BR, both ILD and BR, ILD before RA, and BR before RA, and patterns of ILD encompassing UIP, NSIP, organizing pneumonia (OP) and respiratory bronchiolitis (RB). We also categorized RA-ILD as UIP or other patterns, as well as fibrotic (UIP and fibrotic NSIP) or non-fibrotic patterns.
Furthermore, data on health care utilization, and previous serious infection within 365 days before index date were collected. Missing serostatus data were reviewed in medical records, and multiple imputation by chained equations [25] for this variable was used when no data were available. Otherwise, there were no missing data for covariates included in the propensity score.
All patients were followed until the occurrence of either the outcome or their last clinical encounter (see study design in Supplementary Fig. S1, available at Rheumatology online).
Statistical analysis
We compared the RA-LD cases and RA-no LD comparators at baseline using χ2 or Fisher’s exact tests for categorical variables and t-tests or Wilcoxon rank-sum tests for continuous variables, as appropriate. Means and standard deviations (SD) were reported for normally distributed data, while medians and interquartile ranges (IQR) were used for those with skewed distributions. Mortality rates (per 1000 person-years) were calculated from baseline, with the censoring event defined as the last clinical encounter.
Cox regression models were utilized to estimate the hazard ratios (HR) and 95% confidence intervals (CI) for total mortality. For each cause-specific mortality (respiratory, cancer, cardiovascular, infection), Fine and Gray models [26, 27] estimated subdistribution HRs (sdHR) with 95%CIs, accounting for competing risks for alternative causes of death. We evaluated both a two-level exposure model (comparing RA-LD to RA-no LD) and a three-level exposure model (comparing RA-ILD and RA-BR separately to RA-no LD). For total mortality, RA-ILD was further stratified by imaging patterns into fibrotic RA-ILD and other RA-ILD, as well as RA-ILD with a UIP or non-UIP patterns. Models were adjusted for propensity scores incorporating imbalanced baseline covariates which included smoking status, pack-years, CCI score, serostatus, glucocorticoid use, csDMARD use, bDMARD use, serious infection within 365 days before index date and healthcare utilization.
Baseline characteristics were compared between groups using standardized mean differences (SMDs). All analyses were conducted using R version 2024.04.2 + 764.
Results
Baseline characteristics
We analysed 221 RA-LD cases and 980 RA-no LD comparators. Baseline characteristics are shown in Table 1 and Supplementary Table S1, available at Rheumatology online. The matching factors of age (mean 63.4 years in RA-LD vs 61.6 years in RA-no LD), sex (75% vs 76% female) and RA duration (median 9.0 years vs 6.4 years) were similar between RA-LD cases and RA-no LD comparators.
Table 1.
Baseline characteristics of RA-LD cases and RA-no LD comparators
| Characteristics | RA-LD cases | RA-no LD comparators | SMD |
|---|---|---|---|
| (n = 221) | (n = 980) | ||
| Demographics | |||
| Age, years (mean and SD) | 63.4 (12.2) | 61.6 (12.4) | 0.14 |
| Female sex, n (%) | 165 (75) | 743 (76) | 0.03 |
| Race | 0.04 | ||
| Asian | 3 (1) | 13 (1) | |
| Black | 14 (6) | 54 (6) | |
| White | 191 (86) | 853 (87) | |
| Other | 8 (4) | 37 (4) | |
| Unknown | 5 (2) | 23 (2) | |
| Hispanic | 2 (0.9) | 8 (0.8) | 0.11 |
| Smoking | |||
| Smoking status, n (%) | 0.23 | ||
| Never | 80 (36) | 465 (47) | |
| Past | 126 (57) | 454 (46) | |
| Current | 15 (7) | 61 (6) | |
| Pack-years, median (IQR) | 5 (0, 22.5) | 0 (0, 12.1) | 0.28 |
| Comorbidities | |||
| Charlson Comorbidity Index (median and IQR) | 0 (0, 1) | 0 (0, 0) | 0.30 |
| Diabetes | 9 (4) | 49 (5) | 0.05 |
| Heart failure | 15 (7) | 31 (3) | 0.17 |
| Cardiovascular disease | 20 (9) | 44 (4) | 0.18 |
| Solid cancer | 7 (3) | 6 (0.6) | 0.19 |
| Leukaemia or lymphoma | 17 (8) | 36 (4) | 0.17 |
| RA characteristics | |||
| RA duration, years (median and IQR) | 9.0 (0.7, 18.9) | 6.4 (1.8, 15.0) | 0.18 |
| Seropositive | 180 (81) | 618 (63) | 0.42 |
| RA medications | |||
| Glucocorticoids, n (%) | 113 (51) | 355 (36) | 0.30 |
| Any DMARDs, n (%) | 136 (62) | 542 (55) | 0.13 |
| csDMARDs, n (%) | 108 (49) | 411 (42) | 0.14 |
| Methotrexate | 84 (38) | 305 (31) | 0.15 |
| Hydroxychloroquine | 27 (12) | 131 (13) | 0.03 |
| Sulfasalazine | 7 (3) | 38 (4) | 0.04 |
| bDMARDs, n (%) | 82 (37) | 283 (29) | 0.18 |
| TNFi | 59 (27) | 211 (22) | 0.12 |
| Abatacept | 5 (2) | 38 (4) | 0.09 |
| Rituximab | 10 (5) | 23 (2) | 0.12 |
| Tocilizumab | 7 (3) | 23 (2) | 0.05 |
| tsDMARDs, n (%) | 5 (2) | 23 (2) | 0.01 |
| Tofacitinib | 4(2) | 20 (2) | 0.02 |
| Serious infection within 365 days prior to index date, n (%) | 10 (5) | 18 (2) | 0.15 |
| Health care utilization | |||
| Any hospitalization within 365 days before index, n (%) | 47 (21) | 107 (11) | 0.28 |
| Number of outpatient visits within 365 days before index, median (IQR) | 11 (4, 23) | 6 (2, 15) | 0.38 |
bDMARDs: biological disease-modifying antirheumatic drugs; csDMARDs: conventional synthetic disease-modifying antirheumatic drugs; DMARDs: disease-modifying antirheumatic drugs; IQR: interquartile range; RA-LD: rheumatoid arthritis-associated lung disease; RA-no LD: rheumatoid arthritis without lung disease; SD: standard deviation; SMD: standardized mean differences; TNFi: tumor necrosis factor inhibitor; tsDMARDs: targeted synthetic disease-modifying antirheumatic drug.
The RA-LD cohort included 151 (68%) patients with RA-ILD and 70 (32%) patients with RA-BR (Supplementary Table S1, available at Rheumatology online). Among the RA-ILD cases, 97/151 (64%) were classified as fibrotic, 73/151 (48%) as UIP pattern and 30/151 (20%) as NSIP pattern.
Compared with RA-no LD comparators, RA-LD cases were more likely to be ever smokers (57% vs 46%, SMD = 0.23), had higher pack-years (median 5 vs 0, SMD = 0.28) and had higher CCI scores (SMD = 0.30). They were also more frequently seropositive (81% vs 63%, SMD = 0.42) and more likely to use glucocorticoids (51% vs 36%, SMD = 0.30). RA-LD cases had more healthcare utilization including more previous hospitalization (21% vs 11%, SMD = 0.28), and more frequent clinic visits (median 11 vs 6, SMD = 0.38) in the previous year.
Total mortality
During a total of 10 602.4 person-years (median 8.1 years per patient), there were 65 (29%) deaths among RA-LD cases and 114 (12%) deaths among RA-no LD comparators. Compared with RA-no LD comparators, RA-LD cases demonstrated a significantly higher mortality rate (34.0 vs 13.1 per 1000 person-years, HR 1.97, 95%CI 1.43–2.70, Table 1 and Fig. 1A).
Figure 1.
Cumulative mortality. (A) Cumulative mortality comparing rheumatoid arthritis-associated lung disease (RA-LD) cases and RA-no LD comparators. (B) Cumulative mortality comparing RA-associated interstitial lung disease (RA-ILD) cases, RA-associated bronchiectasis (RA-BR) cases, and RA-no LD comparators
After stratifying RA-LD cases into RA-BR and RA-ILD subgroups, RA-ILD patients (37.7 vs 13.1 per 1000 person-years, HR 2.27, 95%CI 1.61–3.19, Table 2 and Fig. 1B), particularly those with fibrotic pattern (46.3 vs 13.1 per 1000 person-years, HR 2.87, 95%CI 1.97–4.16, Table 3) or UIP pattern (53.3 vs 13.1 per 1000 person-years, HR 3.36, 95%CI 2.26–5.00, Table 3) had higher mortality rates comparing to RA-no LD comparators. However, the mortality risk associated with RA-BR cases was not statistically significant after propensity score adjustment (HR 1.36, 95%CI 0.79–2.35).
Table 2.
Hazard ratios for total mortality, comparing RA-LD cases and RA-no LD comparators
| Study group | Deaths, n | Person-years | Mortality rate per 1000 person-years | Unadjusteda HR (95%CI) | Multivariableb HR (95%CI) |
|---|---|---|---|---|---|
| 2-level exposure | |||||
| RA-no LD | 114 | 8691.8 | 13.1 | 1.0 (Ref) | 1.0 (Ref) |
| RA-LD | 65 | 1911.1 | 34.0 | 2.60 (1.92–3.53) | 1.97 (1.43–2.70) |
| 3-level exposure | |||||
| RA-no LD | 114 | 8691.8 | 13.1 | 1.0 (Ref) | 1.0 (Ref) |
| RA-BR | 15 | 584.4 | 25.7 | 1.87 (1.09–3.21) | 1.36 (0.79-2.35) |
| RA-ILD | 50 | 1326.7 | 37.7 | 2.94 (2.11–4.10) | 2.27 (1.61–3.19) |
Bold text represents statistically significant values.
RA-LD cases and RA-no LD comparators were matched by calendar date, age, sex and RA duration at index date of RA-LD diagnosis.
Adjusted for propensity score that included smoking status, pack-years, Charlson Comorbidity Index score, serostatus, glucocorticoids, csDMARDs, bDMARDs, previous serious infection and health care utilization.
bDMARDs: biological disease-modifying antirheumatic drugs; CI: confidence interval; csDMARDs: conventional synthetic disease-modifying antirheumatic drugs; HR: hazard ratio; RA: rheumatoid arthritis; RA-BR: rheumatoid arthritis-associated bronchiectasis; RA-ILD: rheumatoid arthritis-associated interstitial lung disease; RA-LD: rheumatoid arthritis-associated lung disease; RA-no LD: rheumatoid arthritis without lung disease.
Table 3.
Hazard ratios for total mortality, comparing RA-BR cases, subtypes of RA-ILD cases and RA-no LD comparatorsa
| Study group | Deaths, n | Person-years | Mortality rate per 1000 person-years | Multivariableb HR (95%CI) |
|---|---|---|---|---|
| Fibrotic ILD sub-analysis | ||||
| RA-no LD | 114 | 8691.8 | 13.1 | 1.0 (Ref) |
| RA-BR | 15 | 584.4 | 25.7 | 1.36 (0.79–2.35) |
| RA-fibrotic ILD | 39 | 842.1 | 46.3 | 2.87 (1.97–4.16) |
| Non-fibrotic RA-ILD | 11 | 484.6 | 22.7 | 1.30 (0.69–2.43) |
| UIP sub-analysis | ||||
| RA-no LD | 114 | 8691.8 | 13.1 | 1.0 (Ref) |
| RA-BR | 15 | 584.4 | 25.7 | 1.37 (0.79–2.36) |
| RA-ILD with UIP pattern | 33 | 619.6 | 53.3 | 3.36 (2.26–5.00) |
| RA-ILD with non-UIP patterns | 17 | 707.1 | 24.0 | 1.40 (0.83–2.34) |
Bold text represents statistically significant values.
RA-LD cases and RA-no LD comparators were matched by calendar date, age, sex and RA duration at index date of RA-LD diagnosis.
Adjusted for propensity score that included smoking status, pack-years, Charlson Comorbidity Index score, serostatus, glucocorticoids, csDMARDs, bDMARDs, previous serious infection and health care utilization.
bDMARDs: biological disease-modifying antirheumatic drugs; CI: confidence interval; csDMARDs: conventional synthetic disease-modifying antirheumatic drugs; HR: hazard ratio; RA: rheumatoid arthritis; RA-BR: rheumatoid arthritis-associated bronchiectasis; RA-ILD: rheumatoid arthritis-associated interstitial lung disease; RA-LD: rhematoid arthritis-associated lung disease; RA-no LD: rheumatoid arthritis without lung disease; UIP: usual interstitial pneumonia.
Cause-specific mortality
Respiratory disease was the leading cause of death among RA-LD patients (11% vs 2%, P < 0.001, Table 4). There were no significant differences in the proportions of other causes of death, including cancer (4% vs 2%, P = 0.09), cardiovascular (4% vs 2%, P = 0.24) or non-respiratory infection (2% vs 1%, P = 0.53).
Table 4.
Total and cause-specific mortality, comparing RA-LD cases and RA-no LD comparators
| Characteristics | RA-LD cases (n = 221) | RA-no LD comparators (n = 980) | P-value |
|---|---|---|---|
| Death, n (%) | 65 (29) | 114 (12) | <0.001 |
| Cause of death, n (%) | |||
| Respiratory disease | 25 (11) | 15 (2) | <0.001 |
| Infection | 16 (7) | 25 (3) | 0.002 |
| Cancer | 9 (4) | 20 (2) | 0.09 |
| Lung cancer | 2 (0.9) | 7 (0.7) | 0.67 |
| Cancer other than lung | 7 (3) | 13 (1) | 0.07 |
| Cardiovascular disease | 8 (4) | 22 (2) | 0.24 |
| Non-respiratory infection | 4 (2) | 13 (1) | 0.53 |
| Other | 8 (4) | 20 (2) | 0.21 |
| Unknown | 11 (5) | 24 (2) | 0.07 |
| Sepsis, n (%) | 4 (2) | 6 (0.6) | 0.09 |
| Anatomic site of infection, n (%) | |||
| Lung | 12 (5) | 12 (1) | <0.001 |
| Urinary tract | 2 (0.9) | 1 (0.1) | 0.09 |
| Gastrointestinal system | 1 (0.5) | 5 (0.5) | 1.00 |
| Musculoskeletal system | 1 (0.5) | 1 (0.1) | 0.33 |
| Skin and soft tissue | 0 0 | 3 (0.3) | 1.00 |
| Cardiovascular system | 0 0 | 2 (0.2) | 1.00 |
| Central nervous system | 0 0 | 1 (0.1) | 1.00 |
| Pathogen type, n (%) | |||
| Bacteria | 7 (3) | 12 (1) | 0.06 |
| Virus | 5 (2) | 6 (0.6) | 0.04 |
| Fungi | 1 (0.5) | 1 (0.1) | 0.33 |
| Unknown | 3 (1) | 6 (0.6) | 0.22 |
| Specific pathogen, n (%) | |||
| SARS-CoV-2 | 4 (2) | 4 (0.4) | 0.04 |
| Staphylococcus | 3 (1) | 4 (0.4) | 0.12 |
| Pseudomonas | 2 (0.9) | 1 (0.1) | 0.09 |
| Escherichia coli | 1 (0.5) | 2 (0.2) | 0.46 |
| Klebsiella pneumoniae | 1 (0.5) | 1 (0.1) | 0.33 |
| Pneumocystis | 1 (0.5) | 0 0 | 0.18 |
| Respiratory syncytial virus | 1 (0.5) | 0 0 | 0.18 |
| Adenovirus | 0 0 | 1 (0.1) | 1.00 |
| Candida albicans | 0 0 | 1 (0.1) | 1.00 |
| Enterobacter | 0 0 | 1 (0.1) | 1.00 |
| Enterococcus | 0 0 | 1 (0.1) | 1.00 |
| Proteus mirabilis | 0 0 | 1 (0.1) | 1.00 |
Bold text represents statistically significant values.
RA-LD: rheumatoid arthritis-associated lung disease; RA-no LD: rheumatoid arthritis without lung disease; SARS-CoV-2: severe acute respiratory syndrome coronavirus 2.
Infection-related deaths
Infection-related deaths were more common in RA-LD patients compared with RA-no LD comparators (7% vs 3%, P = 0.002, Table 4). RA-LD patients had a significantly higher proportion of deaths from lung infections (5% vs 1%, P < 0.001). However, infections at other anatomical sites showed no statistically significant differences. Viral infection-related mortality was more likely in RA-LD patients (2% vs 0.6%, P = 0.04). There were no significant differences in bacterial (3% vs 1%, P = 0.06) or fungal infection-related mortality (0.5% vs 0.1%, P = 0.33).
Among specific pathogens, deaths due to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) were higher in RA-LD patients (2% vs 0.4%, P = 0.04). However, no statistical differences were observed for other pathogens, including Staphylococcus (1% vs 0.4%, P = 0.12), Pseudomonas (0.9% vs 0.1%, P = 0.09), Escherichia coli (0.5% vs 0.2%, P = 0.46) and Klebsiella pneumoniae (0.5% vs 0.1%, P = 0.33).
Competing risk analysis for cause-specific mortality
Accounting for alternative causes of death, RA-LD was associated with higher respiratory mortality (13.1 vs 1.7 per 1000 person-years, sdHR 6.68, 95%CI 3.19–14.0) and lung infection-related mortality (6.3 vs 1.4 per 1000 person-years, sdHR 3.33, 95%CI 1.32–8.37, Table 5). No statistically significant differences were observed for cardiovascular or cancer mortality between RA-LD cases and RA-no LD comparators.
Table 5.
Subdistribution hazard ratios for cause-specific mortality, accounting for competing risk of other causes of mortality, comparing RA-LD cases and RA-no LD comparatorsa
| Study group | Deaths, n | Person-years | Mortality rate per 1000 person-years | Multivariableb sdHR (95%CI) |
|---|---|---|---|---|
| Outcome: Respiratory mortality | ||||
| 2-level exposure | ||||
| RA-no LD | 15 | 8691.8 | 1.7 | 1.0 (Ref) |
| RA-LD | 25 | 1911.1 | 13.1 | 6.68 (3.19–14.0) |
| 3-level exposure | ||||
| RA-no LD | 15 | 8691.8 | 1.7 | 1.0 (Ref) |
| RA-BR | 4 | 584.4 | 6.8 | 3.41 (1.05–11.0) |
| RA-ILD | 21 | 1326.7 | 15.8 | 8.06 (3.85–16.9) |
| Outcome: Lung infection mortality | ||||
| 2-level exposure | ||||
| RA-no LD | 12 | 8691.8 | 1.4 | 1.0 (Ref) |
| RA-LD | 12 | 1911.1 | 6.3 | 3.33 (1.32–8.37) |
| 3-level exposure | ||||
| RA-no LD | 12 | 8691.8 | 1.4 | 1.0 (Ref) |
| RA-BR | 2 | 584.4 | 3.4 | 1.80 (0.38–8.62) |
| RA-ILD | 10 | 1326.7 | 7.5 | 4.00 (1.56–10.3) |
| Outcome: Cardiovascular mortality | ||||
| 2-level exposure | ||||
| RA-no LD | 22 | 8691.8 | 2.5 | 1.0 (Ref) |
| RA-LD | 8 | 1911.1 | 4.2 | 1.08 (0.48–2.39) |
| 3-level exposure | ||||
| RA-no LD | 22 | 8691.8 | 2.5 | 1.0 (Ref) |
| RA-BR | 1 | 584.4 | 1.7 | 0.41 (0.06–2.84) |
| RA-ILD | 7 | 1326.7 | 5.3 | 1.40 (0.61–3.24) |
| Outcome: Cancer mortality | ||||
| 2-level exposure | ||||
| RA-no LD | 20 | 8691.8 | 2.3 | 1.0 (Ref) |
| RA-LD | 9 | 1911.1 | 4.7 | 1.48 (0.67–3.25) |
| 3-level exposure | ||||
| RA-no LD | 20 | 8691.3 | 2.3 | 1.0 (Ref) |
| RA-BR | 5 | 584.4 | 8.6 | 2.89 (1.11–7.47) |
| RA-ILD | 4 | 1326.7 | 3.0 | 0.90 (0.29–2.81) |
Bold text represents statistically significant values.
RA-LD cases and RA-no LD comparators were matched by calendar date, age, sex and RA duration at index date of RA-LD diagnosis.
Adjusted for propensity score that included smoking status, pack-years, Charlson Comorbidity Index score, serostatus, glucocorticoids, csDMARDs, bDMARDs, previous serious infection and health care utilization.
bDMARDs: biological disease-modifying antirheumatic drugs; CI: confidence interval; csDMARDs: conventional synthetic disease-modifying antirheumatic drugs; RA: rheumatoid arthritis; RA-BR: rheumatoid arthritis-associated bronchiectasis; RA-ILD: rheumatoid arthritis-associated interstitial lung disease; RA-LD: rheumatoid arthritis-associated lung disease; RA-no LD: rheumatoid arthritis without lung disease; sdHR: subdistribution hazard ratio.
In the 3-level exposure models, RA-ILD cases demonstrated more pronounced risks of respiratory mortality (15.8 vs 1.7 per 1000 person-years, sdHR 8.06, 95%CI 3.85–16.9) and lung infection-related mortality (7.5 vs 1.4 per 1000 person-years, sdHR 4.00, 95%CI 1.56–10.3) compared with RA-no LD comparators. No significant differences were observed for cardiovascular or cancer mortality in RA-ILD vs RA-no LD comparators. RA-BR cases also exhibited increased respiratory mortality (6.8 vs 1.7 per 1000 person-years, sdHR 3.41, 95%CI 1.05–11.0) and cancer mortality (8.6 vs 2.3 per 1000 person-years, sdHR 2.89, 95%CI 1.11–7.47).
Secondary analyses for total and cause-specific mortality
Both RA-ILD (33% vs 12%, P < 0.001) and RA-BR (21% vs 12%, P = 0.02) cases had higher proportions of deaths compared with RA-no LD comparators (Supplementary Table S2, available at Rheumatology online). RA-ILD cases had a significantly higher proportion of deaths from respiratory causes (14% vs 2%, P < 0.001) and infection-related causes (9% vs 3%, P < 0.001). Among infection-related deaths, bacterial pathogens were more frequently implicated in RA-ILD case vs RA-no LD comparators (4% vs 1%, P = 0.02), with Pseudomonas being significantly more common (1% vs 0.1%, P = 0.048). Other infection types or causes of death showed no statistical differences across groups. Compared with RA-no LD comparators, RA-BR cases showed elevated respiratory mortality (6% vs 2%, P = 0.03) and cancer mortality (7% vs 2%, P = 0.02), particularly from non-lung cancers (6% vs 1%, P = 0.02) including cancers of bladder, supraglottis and gallbladder. No significant difference in infection-related deaths was observed between RA-BR cases and RA-no LD comparators.
Discussion
In this retrospective comparative cohort study, we found that RA-LD patients had a 2-fold higher total mortality than RA-no LD comparators, with the risk being particularly elevated in RA-ILD patients. While respiratory disease was the leading cause of death, we identify infection-related mortality as another common cause of death among RA-LD, particularly RA-ILD with fibrotic and UIP patterns. Among infection-related deaths, lung infections, viral infections, particularly SARS-CoV-2, were more common in RA-LD patients. This is one of the first comparative studies to investigate RA-BR and mortality. While there was no increase in total or infection-related mortality, RA-BR patients had 3-fold higher risks of respiratory and cancer mortality. These findings emphasize the need for improved treatment strategies for RA-LD patients to reduce respiratory mortality and also emphasize the importance of mitigating infection risk and monitoring for cancer.
Our findings indicate that RA-LD, particularly RA-ILD, was associated with an increased risk of mortality, which aligned with prior studies [5, 6, 8] that reported a 2- to 10-fold higher mortality in RA-ILD vs RA-no ILD. This elevated mortality risk may be attributed to the progressive nature of pulmonary fibrosis, particularly in patients with the UIP pattern, a well-established poor prognostic factor in idiopathic pulmonary fibrosis and may similarly predict adverse outcomes in RA-ILD [28–30]. Our finding that patients with the UIP pattern exhibited the highest mortality rate further supports this and emphasizes the importance of early detection and targeted management strategies for this high-risk subgroup.
In contrast to RA-ILD, RA-BR did not show a significantly higher total mortality rate compared with RA-no LD comparators in our study. This may be attributed to the nature of bronchiectasis as a condition primarily involving airway dilation and recurrent infections [19, 31] rather than progressive, life-limiting lung damage. While BR can lead to chronic respiratory symptoms and increased susceptibility to infections, these complications may be effectively managed with antibiotics and airway clearance, reducing their impact on overall survival. However, studies by Puéchal [10] and Swinson [11] indicated that RA-BR was associated with worse survival compared with those with RA alone or BR alone. The discrepancy may be due to differences in study designs or patient populations. The Puéchal study [10] focused on diffuse BR, while Swinson’s case-control study [11] assessed 5-year survival. Because ILD can cause traction BR, we took care to classify RA-BR as being isolated without any evidence of ILD. Thus, previous studies showing worse survival in RA-BR may also be explained by some patients who had concurrent BR and ILD. Our study, utilizing the MGB Biobank data with a longer follow-up period, may provide a more accurate reflection of total mortality risk in RA-BR patients. However, it is also possible that we were underpowered to detect a modest association or we may have identified some mild RA-BR cases that diluted a true effect of more severe RA-BR on mortality.
Lung infection-related mortality, including those caused by viral infections and SARS-CoV-2 [32], were more common in RA-LD patients. Bacterial infection-related deaths, particularly those involving Pseudomonas infections, were significantly more frequent in RA-ILD patients, emphasizing the increased vulnerability of this subgroup to severe bacterial infections. This finding aligned with prior studies highlighting the role of structural lung changes in predisposing patients to bacterial colonization [33, 34] and recurrent infections [35]. The high prevalence of Pseudomonas infections in RA-ILD may stem from chronic inflammation and impaired mucociliary clearance [36, 37], which can foster an environment conducive to persistent bacterial growth. Additionally, the widespread use of immunosuppressive therapies in RA-ILD patients may further heighten this risk by weakening host defenses against bacterial pathogens [38, 39]. When analysed separately, RA-BR patients did not show significant differences in infection-related mortality, which may be due to relatively lower severity of BR or limited sample size. Further studies with larger sample size and more comprehensive data, including disease activity, are needed to clarify this association. These results emphasize the importance of proactive infection prevention strategies, including vaccinations, and close monitoring for bacterial pathogens, to mitigate infection-related mortality in this high-risk population.
Competing risk analyses in our study further demonstrated that RA-LD, particularly RA-ILD, was associated with significantly higher risks of respiratory and lung infection-related mortality, aligning with prior findings that have highlighted the poor pulmonary outcomes [40, 41] in RA-ILD patients. Factors including chronic lung inflammation, progressive fibrosis, structural lung damage and immunosuppressive therapies [42, 43] commonly used in RA management may increase the vulnerability to lung infections and acute exacerbations, which could in turn contribute to higher mortality from respiratory causes. In contrast, besides increased respiratory mortality, RA-BR patients also exhibited a notable increase in cancer mortality, particularly from solid tumors. Chronic airway inflammation and structural lung damage in bronchiectasis [44] may promote a pro-tumorigenic environment by sustaining oxidative stress [45], impairing mucociliary clearance, and promoting recurrent infections [46] that drive mutations and tumorigenesis. Additionally, the systemic inflammatory burden in RA-BR might facilitate cancer progression by creating conditions favorable for metastasis [47–49]. In contrast, RA-ILD is characterized primarily by fibrosis rather than recurrent infection, which may lack these tumor-promoting factors. Although RA-BR cases exhibited increased risks of respiratory and cancer mortality, total mortality was not significantly different compared with RA-no LD comparators. These findings should be interpreted with caution due to the small number of events in the RA-BR group. Future studies with larger sample size are needed to validate these associations.
This study has several strengths. The use of a validated algorithm with further confirmation by medical record review for RA identification and rigorous confirmation of RA-LD cases through medical record review and expert radiologist assessments enhance the accuracy of our findings. We were also able to investigate RA-ILD subtypes and RA-BR that were confirmed using these methods. We also had RA-no LD comparators matched by calendar date, age, sex and RA duration to enable the comparative study and minimize potential confounding. We adjusted for the propensity score as a covariate to address residual confounding; however, this approach does not allow evaluation of post-adjustment covariate balance. We chose this approach rather than matching or weighting to retain statistical power while adjusting for key covariates.
However, limitations should be acknowledged. First, there is the potential for residual confounding from unmeasured variables. For example, we did not have validated measures of RA disease activity or RA-LD severity available. Second, the study population, drawn from the MGB Biobank in the Greater Boston area, may limit generalizability to other settings or populations. Third, the reliance on EHR and billing codes may lead to misclassification for some covariates, though efforts were made to validate cases thoroughly. Importantly, both the exposure and outcome variables were all obtained from detailed medical record review, which likely reduced misclassification for key variables. While outcomes were captured within the MGB EHR, events that occurred outside the network may have been missed. However, this may not vary appreciably by RA-LD status. Fourth, we did not have death certificates available on all patients and relied on detailed medical record review to obtain cause of death. While we abstracted data using a structured format, we had to rely on available medical records to determine cause of death. There were some patients who died outside of our hospital system or had several potential underlying causes of death. Fifth, the small sample size of certain RA-LD subgroups, such as RA-BR, limits statistical power for subgroup analyses. Sixth, the slightly higher methotrexate use in RA-LD cases may reflect exposure prior to the diagnosis of lung disease. More active RA may also have prompted both MTX use and chest imaging, increasing the likelihood of detecting RA-LD and introducing potential detection bias. The overall low current use of MTX use in both groups likely reflects real-world heterogeneity in treatment decisions, comorbidity and tolerance. Finally, our matching strategy may have introduced selection bias by excluding people who later developed RA-LD from the pool of possible comparators. Future studies using incidence density sampling could address this issue.
In conclusion, we found a significantly higher mortality risk in patients with RA-LD, particularly RA-ILD with fibrotic or UIP patterns. RA-LD was associated with higher respiratory and lung infection-related mortality, while RA-BR was associated with higher respiratory and cancer mortality. Larger studies are needed to confirm the cancer mortality signal observed in RA-BR. These findings emphasize the need for improved treatments for RA-LD while mitigating infection risks and monitoring for cancer.
Supplementary material
Supplementary material is available at Rheumatology online.
Supplementary Material
Contributor Information
Qianru Zhang, Division of Rheumatology, Inflammation, and Immunity, Brigham and Women’s Hospital, Boston, MA, USA; Department of Rheumatology and Immunology, Beijing Tsinghua Changgung Hospital, School of Clinical Medicine, Tsinghua University, Beijing, China; Harvard Medical School, Boston, MA, USA.
Ying Qi, Division of Rheumatology, Inflammation, and Immunity, Brigham and Women’s Hospital, Boston, MA, USA.
Xiaosong Wang, Division of Rheumatology, Inflammation, and Immunity, Brigham and Women’s Hospital, Boston, MA, USA.
Gregory C McDermott, Division of Rheumatology, Inflammation, and Immunity, Brigham and Women’s Hospital, Boston, MA, USA; Harvard Medical School, Boston, MA, USA.
Sung Hae Chang, Division of Rheumatology, Inflammation, and Immunity, Brigham and Women’s Hospital, Boston, MA, USA; Division of Rheumatology, Department of Internal Medicine, Soonchunhyang University, Cheonan, South Korea.
Mark Chaballa, Bristol Myers Squibb, Pennington, NJ, USA.
Vadim Khaychuk, Bristol Myers Squibb, Pennington, NJ, USA.
Misti L Paudel, Division of Rheumatology, Inflammation, and Immunity, Brigham and Women’s Hospital, Boston, MA, USA; Harvard Medical School, Boston, MA, USA.
Jeffrey A Sparks, Division of Rheumatology, Inflammation, and Immunity, Brigham and Women’s Hospital, Boston, MA, USA; Harvard Medical School, Boston, MA, USA.
Data availability
The original contributions presented in the study are included in the article and its supplementary materials. Further inquiries can be directed to the corresponding author.
Funding
This work was supported by an investigator-initiated grant to J.A.S. from Bristol Myers Squibb. J.A.S. is supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases (grant numbers R01 AR080659, R01 AR077607, P30 AR070253 and P30 AR072577), National Heart, Lung, and Blood Institute (grant number R01 HL155522), the Arthritis Foundation, the R. Bruce and Joan M. Mickey Research Scholar Fund and the Llura Gund Award funded by the Gordon and Llura Gund Foundation. The funders had no role in the decision to publish or preparation of this manuscript.
Disclosure statement: J.A.S. has received research support from Boehringer Ingelheim, Johnson & Johnson and Sonoma Biotherapeutics unrelated to this work. He has performed consultancy for AbbVie, Amgen, Boehringer Ingelheim, Bristol Myers Squibb, Gilead, Inova Diagnostics, Johnson & Johnson, Merck, MustangBio, Optum, Pfizer, ReCor, Sana, Sobi and UCB unrelated to this work. The content is solely the responsibility of the authors and does not necessarily represent the official views of Harvard University, its affiliated academic health care centers, or the National Institutes of Health. J.A.S. is an Editorial Board member of Rheumatology. M.C. and V.K. are employees of Bristol Myers Squibb. All other authors report no conflicts of interest.
References
- 1. Turesson C, O'Fallon WM, Crowson CS, Gabriel SE, Matteson EL. Extra-articular disease manifestations in rheumatoid arthritis: incidence trends and risk factors over 46 years. Ann Rheum Dis 2003;62:722–7. 10.1136/ard.62.8.722 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Kadura S, Raghu G. Rheumatoid arthritis-interstitial lung disease: manifestations and current concepts in pathogenesis and management. Eur Respir Rev 2021;30:210011. 10.1183/16000617.0011-2021 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Kerola AM, Kazemi A, Rollefstad S et al. All-cause and cause-specific mortality in rheumatoid arthritis, psoriatic arthritis and axial spondyloarthritis: a nationwide registry study. Rheumatology (Oxford) 2022;61:4656–66. 10.1093/rheumatology/keac210 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Widdifield J, Paterson JM, Huang A, Bernatsky S. Causes of death in rheumatoid arthritis: how do they compare to the general population? Arthritis Care Res 2018;70:1748–55. 10.1002/acr.23548 [DOI] [PubMed] [Google Scholar]
- 5. Hyldgaard C, Hilberg O, Pedersen AB et al. A population-based cohort study of rheumatoid arthritis-associated interstitial lung disease: comorbidity and mortality. Ann Rheum Dis 2017;76:1700–6. 10.1136/annrheumdis-2017-211138 [DOI] [PubMed] [Google Scholar]
- 6. Juge PA, Wemeau L, Ottaviani S et al. Increased mortality in patients with RA-associated interstitial lung disease: data from a French administrative healthcare database. RMD Open 2023;9:e003491. 10.1136/rmdopen-2023-003491 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Sparks JA, Jin Y, Cho SK et al. Prevalence, incidence and cause-specific mortality of rheumatoid arthritis-associated interstitial lung disease among older rheumatoid arthritis patients. Rheumatology (Oxford) 2021;60:3689–98. 10.1093/rheumatology/keaa836 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Sugano E, Tanaka E, Inoue E et al. Impact of interstitial lung disease on clinical remission and unfavourable events of rheumatoid arthritis: results from the IORRA cohort. Rheumatology (Oxford) 2024;63:1022–9. 10.1093/rheumatology/kead317 [DOI] [PubMed] [Google Scholar]
- 9. Kim D, Cho SK, Choi CB et al. Impact of interstitial lung disease on mortality of patients with rheumatoid arthritis. Rheumatol Int 2017;37:1735–45. 10.1007/s00296-017-3781-7 [DOI] [PubMed] [Google Scholar]
- 10. Puéchal X, Génin E, Bienvenu T, Le Jeunne C, Dusser DJ. Poor survival in rheumatoid arthritis associated with bronchiectasis: a family-based cohort study. PLoS One 2014;9:e110066. 10.1371/journal.pone.0110066 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Swinson DR, Symmons D, Suresh U, Jones M, Booth J. Decreased survival in patients with co-existent rheumatoid arthritis and bronchiectasis. Rheumatology 1997;36:689–91. 10.1093/rheumatology/36.6.689 [DOI] [PubMed] [Google Scholar]
- 12. Singh JA, Cameron C, Noorbaloochi S et al. Risk of serious infection in biological treatment of patients with rheumatoid arthritis: a systematic review and meta-analysis. Lancet 2015;386:258–65. 10.1016/s0140-6736(14)61704-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Schoels M, Aletaha D, Smolen JS, Wong JB. Comparative effectiveness and safety of biological treatment options after tumour necrosis factor α inhibitor failure in rheumatoid arthritis: systematic review and indirect pairwise meta-analysis. Ann Rheum Dis 2012;71:1303–8. 10.1136/annrheumdis-2011-200490 [DOI] [PubMed] [Google Scholar]
- 14. Mena-Vázquez N, Redondo-Rodriguez R, Rojas-Gimenez M et al. Rate of severe and fatal infections in a cohort of patients with interstitial lung disease associated with rheumatoid arthritis: a multicenter prospective study. Front Immunol 2024;15:1341321. 10.3389/fimmu.2024.1341321 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Karlson EW, Boutin NT, Hoffnagle AG, Allen NL. Building the Partners HealthCare Biobank at Partners Personalized Medicine: informed Consent, Return of Research Results, Recruitment Lessons and Operational Considerations. J Pers Med 2016;6. 10.3390/jpm6010002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Liao KP, Cai T, Gainer V et al. Electronic medical records for discovery research in rheumatoid arthritis. Arthritis Care Res (Hoboken) 2010;62:1120–7. 10.1002/acr.20184 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Aletaha D, Neogi T, Silman AJ et al. 2010 Rheumatoid arthritis classification criteria: an American College of Rheumatology/European League Against Rheumatism collaborative initiative. Arthritis Rheum 2010;62:2569–81. 10.1002/art.27584 [DOI] [PubMed] [Google Scholar]
- 18. McDermott GC, Hayashi K, Juge PA et al. Impact of Sex, Serostatus, and Smoking on Risk for Rheumatoid Arthritis-Associated Interstitial Lung Disease Subtypes. Arthritis Care Res (Hoboken) 2025;77:185–94. 10.1002/acr.25432 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. McDermott G, Gill R, Gagne S et al. Demographic, Lifestyle, and Serologic Risk Factors for Rheumatoid Arthritis (RA)-associated Bronchiectasis: role of RA-related Autoantibodies. J Rheumatol 2022;49:672–9. 10.3899/jrheum.211242 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Bongartz T, Nannini C, Medina-Velasquez YF et al. Incidence and mortality of interstitial lung disease in rheumatoid arthritis: a population-based study. Arthritis Rheum 2010;62:1583–91. 10.1002/art.27405 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Washko GR, Lynch DA, Matsuoka S et al. Identification of Early Interstitial Lung Disease in Smokers from the COPDGene Study. Acad Radiol 2010;17:48–53. 10.1016/j.acra.2009.07.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Doyle TJ, Dellaripa PF, Batra K et al. Functional impact of a spectrum of interstitial lung abnormalities in rheumatoid arthritis. Chest 2014;146:41–50. 10.1378/chest.13-1394 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Charlson ME, Carrozzino D, Guidi J, Patierno C. Charlson Comorbidity Index: a Critical Review of Clinimetric Properties. Psychother Psychosom 2022;91:8–35. 10.1159/000521288 [DOI] [PubMed] [Google Scholar]
- 24. Aslam F, Khan NA. Tools for the assessment of comorbidity burden in rheumatoid arthritis. Front Med (Lausanne) 2018;5:39. 10.3389/fmed.2018.00039 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. White IR, Royston P, Wood AM. Multiple imputation using chained equations: issues and guidance for practice. Stat Med 2011;30:377–99. 10.1002/sim.4067 [DOI] [PubMed] [Google Scholar]
- 26. Nolan EK, Chen HY. A comparison of the Cox model to the Fine-Gray model for survival analyses of re-fracture rates. Arch Osteoporos 2020;15:86. 10.1007/s11657-020-00748-x [DOI] [PubMed] [Google Scholar]
- 27. Schreiner AD, Zhang J, Moran WP et al. FIB-4 and incident severe liver outcomes in patients with undiagnosed chronic liver disease: a Fine-Gray competing risk analysis. Liver Int 2023;43:170–9. 10.1111/liv.15295 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Kim EJ, Collard HR, King TE Jr. Rheumatoid arthritis-associated interstitial lung disease: the relevance of histopathologic and radiographic pattern. Chest 2009;136:1397–405. 10.1378/chest.09-0444 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Qiu M, Jiang J, Nian X et al. Factors associated with mortality in rheumatoid arthritis-associated interstitial lung disease: a systematic review and meta-analysis. Respir Res 2021;22:264. 10.1186/s12931-021-01856-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Solomon JJ, Chung JH, Cosgrove GP et al. Predictors of mortality in rheumatoid arthritis-associated interstitial lung disease. Eur Respir J 2016;47:588–96. 10.1183/13993003.00357-2015 [DOI] [PubMed] [Google Scholar]
- 31. Duarte AC, Porter J, Leandro MJ. Bronchiectasis in rheumatoid arthritis. A clinical appraisial. Joint Bone Spine 2020;87:419–24. 10.1016/j.jbspin.2019.12.006 [DOI] [PubMed] [Google Scholar]
- 32. Figueroa-Parra G, Gilbert EL, Valenzuela-Almada MO et al. Risk of severe COVID-19 outcomes associated with rheumatoid arthritis and phenotypic subgroups: a retrospective, comparative, multicentre cohort study. Lancet Rheumatol 2022;4:e765–74. 10.1016/S2665-9913(22)00227-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Sethi S, Maloney J, Grove L, Wrona C, Berenson CS. Airway inflammation and bronchial bacterial colonization in chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2006;173:991–8. 10.1164/rccm.200509-1525OC [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Estes RJ, Meduri GU. The pathogenesis of ventilator-associated pneumonia: i. Mechanisms of bacterial transcolonization and airway inoculation. Intensive Care Med 1995;21:365–83. 10.1007/BF01705418 [DOI] [PubMed] [Google Scholar]
- 35. Richardson H, Dicker AJ, Barclay H, Chalmers JD. The microbiome in bronchiectasis. Eur Respir Rev 2019;28:190048. 10.1183/16000617.0048-2019 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Chegini Z, Noei M, Hemmati J, Arabestani MR, Shariati A. The destruction of mucosal barriers, epithelial remodeling, and impaired mucociliary clearance: possible pathogenic mechanisms of Pseudomonas aeruginosa and Staphylococcus aureus in chronic rhinosinusitis. Cell Commun Signal 2023;21:306. 10.1186/s12964-023-01347-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Garcia-Clemente M, de la Rosa D, Máiz L et al. Impact of pseudomonas aeruginosa infection on patients with chronic inflammatory airway diseases. J Clin Med 2020;9:3800. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Jani M, Barton A, Hyrich K. Prediction of infection risk in rheumatoid arthritis patients treated with biologics: are we any closer to risk stratification? Curr Opin Rheumatol 2019;31:285–92. 10.1097/bor.0000000000000598 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Frisell T, Bower H, Morin M et al. ; ARTIS Study Group. Safety of biological and targeted synthetic disease-modifying antirheumatic drugs for rheumatoid arthritis as used in clinical practice: results from the ARTIS programme. Ann Rheum Dis 2023;82:601–10. 10.1136/ard-2022-223762 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Koduri G, Solomon JJ. Identification, monitoring, and management of rheumatoid arthritis-associated interstitial lung disease. Arthritis Rheumatol 2023;75:2067–77. 10.1002/art.42640 [DOI] [PubMed] [Google Scholar]
- 41. Chen RX, Zhao LD, Xiao XY et al. Distinctive clinical characteristics and outcome of ILD-onset rheumatoid arthritis and ACPA-positive ILD: a longitudinal cohort of 282 cases. Clin Rev Allergy Immunol 2021;60:46–54. 10.1007/s12016-020-08819-0 [DOI] [PubMed] [Google Scholar]
- 42. Liu N, Fan X, Shao Y et al. Resveratrol attenuates inflammation and fibrosis in rheumatoid arthritis-associated interstitial lung disease via the AKT/TMEM175 pathway. J Transl Med May 14 2024;22:457. 10.1186/s12967-024-05228-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. Yun H, Xie F, Delzell E et al. Comparative risk of hospitalized infection associated with biologic agents in rheumatoid arthritis patients enrolled in Medicare. Arthritis Rheumatol 2016;68:56–66. 10.1002/art.39399 [DOI] [PubMed] [Google Scholar]
- 44. Choi H, Park HY, Han K et al. Non-cystic fibrosis bronchiectasis increases the risk of lung cancer independent of smoking status. Ann Am Thorac Soc 2022;19:1551–60. 10.1513/AnnalsATS.202111-1257OC [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45. Matysiak-Budnik T, Mégraud F. Helicobacter pylori infection and gastric cancer. Eur J Cancer 2006;42:708–16. [DOI] [PubMed] [Google Scholar]
- 46. Ding S-Z, Goldberg JB, Hatakeyama M. Helicobacter pylori infection, oncogenic pathways and epigenetic mechanisms in gastric carcinogenesis. Future Oncol 2010;6:851–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47. Grivennikov SI, Greten FR, Karin M. Immunity, inflammation, and cancer. Cell 2010;140:883–99. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48. Karin M. Nuclear factor-κB in cancer development and progression. Nature 2006;441:431–6. [DOI] [PubMed] [Google Scholar]
- 49. Lewis CE, Pollard JW. Distinct role of macrophages in different tumor microenvironments. Cancer Res 2006;66:605–12. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The original contributions presented in the study are included in the article and its supplementary materials. Further inquiries can be directed to the corresponding author.