Lung Transplantation in Rheumatoid Arthritis Related Interstitial Lung Disease: Risk Factors and Mortality in a Universal Health Care System

ABSTRACT

Background

Despite therapy, some patients with rheumatoid arthritis related interstitial lung disease (RA‐ILD) require lung transplantation for progressive disease. We aimed to identify factors associated with lung transplantation and post‐transplantation survival in RA‐ILD.

Methods

We performed a retrospective observational cohort study using data from Ontario, Canada between 2003 and 2022. RA‐ILD patients were identified from the Ontario RA Database. We used a cause‐specific hazards model, considering the competing risk of death, to identify factors associated with lung transplantation. A Cox proportional hazards model was used to evaluate post‐transplant survival.

Results

Among 6470 RA‐ILD patients, 77 (1.2%) underwent lung transplantation, equating to 2.40 transplants per 1000 person years. The median time to transplant was 3.7 years. Younger age [HR 0.93 per increasing year (95%CI 0.91–0.95), p < 0.0001], male sex [HR 2.40 (95%CI 1.523.78), p = 0.0002], and comorbid diabetes [HR 1.70 (95%CI 1.06–2.73), p = 0.003], congestive heart failure [HR 1.87 (95%CI 1.15–3.04), p = 0.01] and COPD [HR 2.14 (95%CI 1.25–3.68), p = 0.006] were associated with transplantation. The median survival following transplantation was 4.3 years.

Conclusion

Lung transplantation is an uncommon, but important therapy for RA‐ILD. Select comorbidities are associated with lung transplantation. Post‐transplant survival among RA‐ILD patients in Ontario was comparable to that reported elsewhere.

Abbreviations

ADG

Aggregated diagnosis groups

CAD

Coronary artery disease

CIHI

Canadian institute for health information

CIHI‐DAD

Canadian institute for health information discharge abstract database

COPD

Chronic obstructive pulmonary disease

EAM

Extra‐articular manifestations

ICD

International classification of diseases

ILD

Interstitial lung disease

IPF

Idiopathic pulmonary fibrosis

MLTC

Ministry of long‐term care

OHIP

Ontario health insurance plan

ORAD

Ontario rheumatoid arthritis database

PH

Proportional hazards

RA

Rheumatoid arthritis

RA‐ILD

Rheumatoid arthritis related interstitial lung disease

SARD

Systemic autoimmune rheumatic disease

1. Introduction

Rheumatoid arthritis (RA) is a chronic inflammatory disease characterized by joint pain and destruction. Systemic complications can also occur in RA, of which interstitial lung disease (ILD) is one of the most concerning. RA‐associated ILD (RA‐ILD) is a significant cause of morbidity and mortality, with some progressing to experience severe restriction and respiratory failure. Despite advancements in the management of RA, treatment options for advanced lung disease remain limited.

Lung transplantation offers some hope for patients with severe RA‐ILD, having been reported to improve survival and quality of life. Although rates of lung transplant listing and performance for ILD attributed to systemic autoimmune rheumatic disease (SARD) have increased in recent years, this indication has been reported to comprise less than 1% of performed lung transplants [1, 2]. There are limited data on post‐lung transplantation outcomes in RA‐ILD, with most data being limited to small studies, often focusing on SARD‐ILD as a whole, including only a subset of RA‐ILD patients. Singer et al. reported a 1‐year post transplant survival of 67% among 10 RA‐ILD patients, similar to patients with idiopathic pulmonary fibrosis (IPF) [3]. A study of 31 SARD‐ILD patients, 4 with RA‐ILD, reported similar 1‐ and 5‐year post‐transplant survival to IPF [4]. More recently, post‐lung transplant survival among 42 patients with RA‐related lung disease from a single American center reported a median survival of 5.3 years, similar to other lung transplant populations [5]. The paucity of data highlights the need for larger studies focused on risk factors for lung transplantation and post‐transplant outcomes in RA‐ILD. Using health services datasets from Ontario, Canada, we aimed to identify patient factors associated with lung transplantation and post‐transplant survival in RA‐ILD.

2. Methods

2.1. Study Design and Setting

Ontario is Canada's largest province, providing universal health care to over 15 million people as of 2024 [6]. We conducted an observational study using a population‐based cohort from Ontario, Canada between 2003 and 2022.

2.2. Data Sources

We used deidentified provincial health services data from ICES. ICES is an independent, non‐profit research institute funded by an annual grant from the Ontario Ministry of Health (MOH) and the Ministry of Long‐Term Care (MLTC). As a prescribed entity under Ontario's privacy legislation, ICES is authorized to collect and use health care data for the purposes of health system analysis, evaluation and decision support. Secure access to these data is governed by policies and procedures that are approved by the Information and Privacy Commissioner of Ontario. These datasets were linked using unique encoded identifiers and analyzed at ICES. RA cases were identified using the validated Ontario Rheumatoid Arthritis Database (ORAD) [7]. The ORAD was generated using data from the Canadian Institute for Health Information Discharge Abstract Database (CIHI‐DAD) and Ontario Health Insurance Plan (OHIP) Physician Services Claims database. The Ontario Registered Persons Database (RPDB) (provided by CIHI‐DAD, and MOH) and OHIP databases were used to ascertain death date and comorbidity data. The RPDB and the Immigration, Refugees and Citizenship Canada Permanent Resident database were used to derive patient demographics and immigration status. The use of the data in this project is authorized under section 45 of Ontario's Personal Health Information Protection Act (PHIPA) and does not require review by a Research Ethics Board.

2.3. Inclusion Criteria

Adults over 18 years of age with incident RA‐ILD residing in Ontario between 2003 and 2022 were included. Individuals were excluded if they were not eligible for health insurance, which is provided to all permanent residents of Ontario. Those lacking demographic information were excluded.

2.4. RA‐ILD Case Definition

ORAD cases are required to meet the following definition: (1) one hospitalization for RA, indicated by an International Classification of Diseases, Ninth/Tenth Revision (ICD9/10), Clinical Modification code specific for RA (714, M05, M06), or (2) ≥ 3 OHIP physician claims (714) for RA in a two‐year period, with ≥1 claim submitted by an internist, rheumatologist, or orthopaedic surgeon.

We defined RA‐ILD as ≥ 2 physician outpatient or inpatient OHIP claims (515) or inpatient ICD9/10 codes (ICD9: 515, 516.3, 714.8; ICD10: J84.1, J84.2, J84.9, M05.1) for ILD in patients within the ORAD. ILD claims were at least two weeks apart, with one following the RA diagnosis. We excluded patients with other ILDs diagnosed within 5 years of RA, including: sarcoidosis, hypersensitivity pneumonitis, pneumoconiosis, radiation toxicity, pulmonary hemosiderosis, pulmonary alveolar microlithiasis, lymphangioleiomyomatosis, langerhans cell histiocytosis, scleroderma lung disease, and paediatric ILD. These patients were identified using OHIP/ICD9/ICD10 codes (Table S1). The RA‐ILD diagnosis date was set as the first claim for ILD. Incident RA‐ILD cases were identified using a look‐back window of 5‐years prior to the RA diagnosis to distinguish between prevalent cases and excluded patients without 5‐years of available OHIP data prior to RA.

2.5. Outcome Measures

Lung transplantation date was identified using OHIP claims (M155). We did not distinguish between single and double lung transplantation. Two physician claims for ILD were required prior to transplant performance. We recorded all‐cause mortality and the date of death using data from the RPDB.

2.6. Variable Selection

To investigate factors associated with lung transplantation and post‐transplant survival we assessed demographic and comorbidity variables a priori. Age at RA‐ILD diagnosis and lung transplant, sex, income quintile, rurality, and immigration status were recorded. Pulmonary [chronic obstructive pulmonary disease (COPD), asthma], cardiovascular [hypertension, coronary artery disease (CAD), congestive heart failure (CHF), stroke], diabetes, malignancy, extra‐articular manifestations (EAM), and musculoskeletal (osteoarthritis, osteoporosis) comorbidities were documented within five years of RA‐ILD diagnosis. EAM were identified using OHIP and ICD codes are listed in Table S2. The burden of comorbidity was evaluated using the ACG System Aggregated Diagnosis Groups (ADGs), corresponding to the Johns Hopkins ACG System Version 10 [8]. We analyzed ADG comorbidity severity by the number of conditions [low (0–5), intermediate (6–9) and high groupings (10+)].

2.7. Statistical Analysis

Descriptive statistics were used to compare demographics and comorbidities associated with lung transplantation. Chi‐squared (or Fisher's exact test) and Wilcoxon rank sum tests were used to compare categorical and continuous variables between patients who did and did not undergo transplantation.

The rate of lung transplant for RA‐ILD was estimated per 1000 person years. The time used to calculate rates began at RA‐ILD diagnosis and ended at the time of transplant, date of last contact, death, or cohort completion (April, 1 2023). We calculated the incidence rate ratio (IRR) and generated a Poisson regression model to evaluate for significant changes in the rate of lung transplant in the first versus second half of our cohort.

We fit a multivariable cause‐specific hazards model for time to RA‐ILD lung transplantation accounting for the competing risk of death. The included variables incorporated patient age, sex, median neighbourhood income, location of residence (rural vs. urban), immigration status, and comorbidities (COPD, CHF, and diabetes). The number of included comorbidities was limited due to low event rates; we included select comorbidities that were more common in those undergoing transplant in univariate comparisons and were felt to be clinically relevant. We evaluated for multicollinearity between variables (variance inflation factor >5) and confirmed the absence of PH assumption violations using the Komolgorov‐type supremum test. Because subjects needed two RA‐ILD physician claims to enter our cohort, all individuals experienced a period where there was no possibility of death (immortal time). To mitigate the influence of immortal time, the time to transplantation began when patients fulfilled our disease definition (second physician claim). The unadjusted time to transplant was calculated from when patients were initially diagnosed with RA‐ILD (first physician claim).

A multivariable Cox PH model was used to identify factors associated with post‐lung transplant mortality. Our survival analysis began at the date of lung transplant. Given the number of deaths post‐transplant was small, we restricted our model to patient demographics (age at transplantation, sex), income [high (top two quintiles) vs. low] and rurality to reduce the degrees of freedom. There were not enough events to include comorbidities.

Patients with missing demographic data were excluded from regression analyses (<1% of our cohort). Statistical analyses were performed using SAS (Cary, North Carolina, USA).

3. Results

3.1. Patient Characteristics and Performed Transplants

6470 incident cases of RA‐ILD were identified in Ontario (Figure 1), of which 77 (1.2%) received a lung transplant. Transplant recipients were younger at the time of RA‐ILD diagnosis [58.0 vs. 70.0 (median years), p < 0.0001], more likely to be male (53.3%, p = 0.002), and less comorbid according to ADGs (p < 0.001) (Table 1). Most comorbidities were similar or less common in transplant recipients, except for COPD (72.7% vs. 59.8%, p = 0.02).

FIGURE 1.

Flow diagram showing the exclusion criteria that led to the identification of incident rheumatoid arthritis related interstitial lung disease cohort in Ontario between 2003 and 2022.

TABLE 1.

RA‐ILD patient demographics and comorbidities stratified by lung transplantation.

	RA‐ILD lung transplant recipients (n = 77)	RA‐ILD patients not receiving a lung transplant (n = 6393)	p‐value
Female, n (%)	36 (46.7)	4078 (63.8)	0.002
Age at ILD diagnosis (years), median (IQR)	58.0 (10)	70.0 (16.0)	<0.001
Income quintile * 1 (lowest) 2 3 4 5 (highest)	10 (13.0) 14 (18.2) 22 (28.6) 14 (18.2) 17 (22.1)	1313 (20.6) 1365 (21.4) 1245 (19.5) 1235 (19.4) 1212 (19.0)	0.19
Rural dwelling	12 (15.6)	790 (12.4)	0.39
ADG comorbidity Low Medium High	18 (23.4) 34 (44.2) 25 (32.5)	706 (11.0) 2307 (36.1) 3380 (52.9)	<0.001
COPD	56 (72.7)	3821 (59.8)	0.02
Asthma	21 (27.3)	1899 (29.7)	0.64
CHF	33 (42.9)	2262 (35.4)	0.17
Malignancy	15 (19.5)	1605 (25.1)	0.26
Diabetes	33 (42.9)	2104 (32.9)	0.07
Hypertension	51 (66.2)	4697 (73.5)	0.15
Osteoarthritis	19 (24.7)	2540 (39.7)	0.007
Extra‐articular manifestations	21 (27.3)	1947 (30.5)	0.55

Abbreviations: ADG, aggregated diagnosis groups; CHF, congestive heart failure; COPD, chronic obstructive pulmonary disease; IQR, interquartile range, RA‐ILD, rheumatoid arthritis related interstitial lung disease.

^*Represents variables where there is a small amount of missing data. Several comorbidities of interest are not shown in the table due to small cell size.

The rate of lung transplantation was estimated as 2.40 transplants per 1000 person years. When comparing patients diagnosed with RA‐ILD in the first and second half of our cohort, the transplant rate increased from 1.89 to 2.93 per 1000 person years. The IRR for lung transplantation between time periods was 1.55 (95%CI 0.99–2.45, p = 0.06).

3.2. Time to Lung Transplantation in RA‐ILD

In those undergoing transplantation, the unadjusted median time to lung transplantation was 3.7 years (IQR 3.6). After adjusting for comorbidities, younger age [HR 0.93 per increasing year (95%CI 0.91–0.95), p < 0.0001] and male sex [HR 2.40 (95%CI 1.52–3.78), p = 0.0002] were associated with an increased hazard of lung transplantation (Figure 2). RA‐ILD patients with COPD [HR 2.14 (95%CI 1.25–3.68), p = 0.006], CHF [HR 1.87 (95%CI 1.15–3.04), p = 0.01], and diabetes [HR 1.70 (95%CI 1.06–2.73), p = 0.003] experienced an increased hazard of lung transplantation.

FIGURE 2.

Hazard ratios forest plot for the time to lung transplantation in RA‐ILD. Age at ILD diagnosis represents increasing age in years. Comorbidities were modeled based on their presence as compared to absence. CHF = congestive heart failure, COPD = chronic obstructive pulmonary disease.

3.3. Mortality Following Lung Transplantation in RA‐ILD

Forty‐one (53.2%) RA‐ILD patients that received a lung transplant died during follow‐up. The median unadjusted survival following lung transplant was 4.3 (IQR 10.7) years. Age at lung‐transplantation, sex, income, and rural dwelling were not significantly associated with post‐transplant survival (Table 2). Due to the low number of deaths, we were unable to include comorbidities in our model.

TABLE 2.

Summary of hazard ratios for survival post‐lung transplantation for patients with rheumatoid arthritis related interstitial lung disease.

	HR (95% CI)	p‐value
Male	1.33 (0.69–2.59)	0.39
Age at lung transplant	0.99 (0.95–1.03)	0.59
Income High vs. Low	1.03 (0.52–2.04)	0.93
Rural dwelling	0.80 (0.36–1.76)	0.57

4. Discussion

Using data from a North American population with universal health coverage, we found the frequency of lung transplantation among RA‐ILD patients to be low, approximating 1% of patients. Patients receiving a lung transplant were generally younger and less comorbid. However, those with COPD, diabetes and CHF had an increased hazard for lung transplantation, suggesting that in select cases, these comorbidities may contribute to an increased urgency for transplant. The median survival post‐lung transplantation was 4.3 years. Patient demographics were not clearly associated with post‐transplant survival, though our analysis may have been underpowered.

The proportion of lung transplants performed for ILD has increased substantially over time, where ∼40% of performed lung transplants in 2017 were for idiopathic interstitial pneumonias (IIP) or non‐IIP ILD [2]. However, lung transplantation is an uncommon therapy for SARD‐ILD, accounting for only 0.9% of all lung transplants globally between 1995 and 2018 [2]. We observed that 1.2% of patients with RA‐ILD underwent lung transplantation, representing a rate of 2.40 transplants per 1000 person years. Although lung transplant remains rare, these results suggest it is a viable option for a select subgroup of RA‐ILD patients with advanced disease.

We found male sex was associated with an increased hazard for transplantation in RA‐ILD after multivariable adjustment. This might suggest that men with RA‐ILD experience increased disease severity. We previously reported annual rates of RA‐ILD related death to be generally higher in men as compared to women, supporting the presence of sex differences in disease severity [11]. The increased frequency of smoking, and usual interstitial pneumonia pattern seen in men with RA‐ILD are potential factors that could also contribute to poorer prognosis [12, 13, 14]. Alternatively, unrecognized biases in transplant referrals and listing between the sexes may be contributing. Irrespective of disease severity, men with scleroderma ILD were reported to undergo lung transplantation sooner and more frequently than women, suggesting potential biases may exist [15].

We found lung transplantation recipients with RA‐ILD had a lower burden of comorbidity; transplant candidates are required to have a certain degree of fitness and significant comorbidity represents a major barrier to transplant. However, patients with COPD, CHF and diabetes experienced an increased hazard for lung transplantation. This might suggest certain comorbidities may signify more aggressive disease, necessitating earlier intervention. Patients with COPD and RA‐ILD may represent a unique subgroup with enhanced disease severity; combined fibrosis and emphysema (CPFE) is associated with pulmonary hypertension and severe hypoxemia and may warrant expedited transplantation [9]. The increased hazard for those with CHF in our cohort could signify right sided heart failure and pulmonary hypertension, though this is challenging to confirm outside of a detailed chart review. An additional hypothesis is that the increased frequency of diabetes among transplant recipients is a sequela of more intensive RA treatment, given chronic corticosteroid use is common in this population. However, these results need to be interpreted cautiously, as the association between the above comorbidities and lung transplantation could alternatively represent coding misclassification, referral biases, or an alternative disease phenotype where COPD may be the predominant pulmonary process.

Following lung transplant, we report a median survival of 4.3 years among RA‐ILD patients, indicating a modest post‐transplant prognosis. This is shorter than the previously reported median survival of 6.7 years among all lung transplant recipients according to the International Society for Heart and Lung Transplantation [2, 10]. However, given the heterogeneity in transplant recipients and small cohort size, we need to interpret these results cautiously and consider populations that may be more similar as opposed to the transplant population as a whole. Compared to IPF, where the median post‐transplant survival has been reported as approximately 5.2 years, our findings align more closely [16]. The shared characteristics of older age at ILD onset, male predominance and often UIP pattern fibrosis may account for a more similar prognosis [17]. A history of chronic immunosuppressive therapy and chronic inflammatory disease effects in RA‐ILD may in part explain the worse survival in our RA‐ILD cohort. Razmjou et al. recently published on post lung transplant survival in 42 patients with RA‐related lung disease (83% of which had ILD) at a single centre in California, USA, reporting a median survival of 5.3 years [5]. The reasons for an attenuated survival in our cohort are unclear, and given the small cohort size, our results may be best interpreted as broadly comparable. Differences in candidate selection and practices within Ontario could be contributing; accepting higher risk recipients could lead to enhanced post‐transplant mortality risk.

We failed to identify any significant impact of patient demographics on post‐transplant survival. This was unexpected, as past research has reported that older age at lung transplant, female sex, and socioeconomic disadvantage are associated with worse post‐transplant survival [18, 19, 20]. Alternatively, it could be due to low statistical power. We were unable to evaluate the impact of comorbidities on post‐transplant outcomes due to the small number of events.

There are important limitations of this research that require acknowledgement. First, the use of physician claim data to identify RA‐ILD cases may predispose to case misclassification. Although we used a validated cohort for RA detection, our disease definition for ILD using OHIP and/or ICD codes was not formally validated and may be prone to misclassification. To try and reduce this bias, we applied a disease definition requiring repeated physician claims for ILD which could utilize both inpatient and outpatient codes. Second, because our disease definition required repeated physician claims, a period of immortal time was introduced. To reduce immortal time bias, our time to event analysis started when patients fulfilled RA‐ILD criteria. Third, it's possible that some patients underwent lung transplantation due to the effects of COPD as opposed to RA‐ILD. It is challenging to determine the exact influence of specific diagnoses on transplant need. However, we believe that repeated ILD claims would suggest significant RA‐ILD, and that this diagnosis is likely to contribute to disease severity as compared to COPD alone. Lastly, there are several unmeasured factors in the pre‐ and post‐transplant period that could impact transplant performance and mortality; data regarding lung function, smoking, serology, radiographic pattern of fibrosis and immunosuppressive treatments were not available or included in our analysis. We were unable to comment on the presence of RA‐overlap syndrome and how this may have impacted disease severity. Double‐lung versus single‐lung transplantation has been reported to improve survival in IPF, however, we did not distinguish between the type of transplant performed, listing urgency or performance of repeat transplantation [21].

5. Conclusion

Lung‐transplantation was performed in 1.2% of patients with incident RA‐ILD over a near 20‐year period. Patient demographics and certain comorbidities were associated with an increased hazard for lung transplantation. Post‐transplant survival in RA‐ILD patients was 4.3 years, which is largely comparable to that reported elsewhere. Concurrent COPD is a common comorbidity and influences the time to lung transplantation among RA‐ILD patients. We believe these results provide meaningful information for physicians managing RA‐ILD in the pre‐ and post‐transplant periods.

Author Contributions

Lee M. Fidler: conceptualization, methodology, formal analysis, writing – original draft, funding acquisition. Omri A. Arbiv: methodology, formal analysis, writing – review and editing. Joseph S. Munn: methodology, formal analysis writing – review and editing. Jolene H. Fisher: writing – review and editing, funding acquisition. Shane Shapera: writing – review and editing, funding acquisition. Peter C. Austin: methodology, formal analysis, writing – review and editing. Andrea S. Gershon: conceptualization, methodology, formal analysis, writing – original draft, writing – review and editing data, supervision, funding acquisition.

Funding

Previous grants provided by the University of Toronto Pettit Block Grant and the Canadian Pulmonary Fibrosis Foundation contributed to this research.

Conflicts of Interest

LF has received grant funding from the University of Toronto and the Canadian Pulmonary Fibrosis Foundation and has received honoraria from Boehringer Ingelheim. OAA has no conflicts of interest to declare. JSM has no conflicts of interest to declare. JHF has received grant funding from the University of Toronto and the Canadian Pulmonary Fibrosis Foundation and has received honoraria from Boehringer Ingelheim and AstraZeneca. SS has consulted and participated in advisory boards (Hoffman La‐Roche, Boehringer Ingelheim, AstraZeneca), accepted honoraria ((Hoffman La‐Roche, Boehringer Ingelheim, AstraZeneca), and participated in clinical trials for Boehringer Ingelheim and United Therapeutics. PCA has no conflicts of interest to declare. ASG has no conflicts of interest to declare.

Supporting information

Supporting File 1: ctr70644‐supp‐0001‐SuppMat.docx.

Supporting File 2: ctr70644‐supp‐0002‐SuppMat.docx.

Data Availability Statement

The dataset from this study is held securely in coded form at ICES. While legal data sharing agreements between ICES and data providers (e.g., healthcare organizations and government) prohibit ICES from making the dataset publicly available, access may be granted to those who meet pre‐specified criteria for confidential access, available at www.ices.on.ca/DAS (email: das@ices.on.ca). The full dataset creation plan and underlying analytic code are available from the authors upon request, understanding that the computer programs may rely upon coding templates or macros that are unique to ICES and are therefore either inaccessible or may require modification.