The findings of this study indicate that in both RA and RA-ILD, mortality showed a significant decline in the last 20 years and the presence of ILD decreased survival in patients with RA. While RA is more common in women than in men, previous studies have suggested that the prevalence of RA-ILD was greater among men than among women or was similar between genders. In the present study, although the ASMR of females was greater than that of males, the ASMR ratio of RA-ILD to RA was higher in men than in women, trending upwards in men and downwards in women. This observation suggests that the proportion of male patients with RA dying from ILD was higher than the proportion of female patients, and the proportion of male patients increased between 1999–2003 and 2014–2018. Hence, male patients with RA-ILD have an increased risk of death, and the risk has been increasing over the course of 20 years. These findings are consistent with those of previous cohort studies (11, 12).
Compared to other connective tissue diseases, patients with RA-ILD may be older, and exhibit a higher proportion of usual interstitial pneumonia (UIP). Owing to immunosenescence and inflammaging, the aged lung undergoes functional and structural changes that facilitate the occurrence of pulmonary fibrosis. Hence, ILD is more likely to occur in older patients with RA. In the present study, we found that the elderly accounted for the majority of patients with RA-ILD deaths, whereas RA-ILD patient mortality in the 55–64 years group showed significant reductions. Furthermore, the proportion of ILD cases increased in patients aged 55–74 years whose death certificates mentioned RA and decreased in patients aged > 74 years. This suggests that RA-ILD survival was extended. However, the reasons accounting for this increase in life expectancy are not well characterised.
To answer this question, we conducted an analysis of UCD in RA-ILD, which indicated that arthropathies and ILD are ranked as the top two most frequent UCDs in patients with RA and ILD mentioned in death certificates. This was the case during the entire 20-year follow-up period. Surprisingly, although there was a decrease in the death rate among patients with RA-ILD mentioned in death certificates, little significant changes occurred in the mortality rate when ILD was the UCD. Therefore, the decline in RA-ILD mortality may be more due to an improvement in living standards, control of other complications, and active treatment of primary diseases than due to treatment of ILD. One possible explanation is that there is a critical unmet need with respect to the management of RA-ILD patients. There is also evidence that treatment with methotrexate in patients with RA increased the prevalence of interstitial lung disease by approximately 0.3–11% based on numerous case reports and case series (13, 14). Other evidence suggests that treating with abatacept versus TNF inhibitors (TNF-Is) might be associated with the occurrence of ILD in RA patients, which can lead to more severe pulmonary symptoms and even death (15, 16). Further prospective cohort studies are warranted to better illustrate the association or causation between TNF-Is and ILD. A longitudinal multicenter study of 68 RA-ILD patients was conducted from 2007 until 2018 in Madrid. The results showed that patients receiving rituximab were less likely to develop functional respiratory impairment than were patients treated with other therapies (17). Tozumab showed a good safety profile in patients with RA-ILD and a potential effect on the stabilisation of pulmonary manifestations. Nevertheless, large sample-size randomised controlled trials and prospective studies are needed to validate these findings. In addition, there are a lack of sensitive and specific disease predictors. Studies showed that possible predictors of mortality include lung carbon monoxide diffusion function, high IgM rheumatoid factor levels (18), and UIP patterns (19). However, these indicators are not specific or are difficult to use for the early identification of a disease. A recent study indicated that the MUC5B promoter variant was the strongest genetic risk factor. The MUC5B promoter variant was strongly correlated with the occurrence of RA-ILD and with UIP on imaging (20). Nevertheless, the clinical application of this test appears to be far off. A prospective cohort study (21) showed that there was an association between active articular RA and development of RA-ILD, and decreasing systemic inflammation might modify the natural history of RA-ILD. Therefore, more attention should be paid to active prevention, early diagnosis, and effective management of this condition.
Ischaemic heart disease is a very common UCD in patients with RA or RA-ILD. A previous population-based cohort study indicated that ischaemic heart disease and congestive heart failure were more likely to occur in the RA-ILD group than patients with RA but not ILD, with the difference being more significant for congestive heart failure (8.5% in the RA-ILD group and 4.4% among those with RA but not ILD). A cross-sectional cohort study of 2013 patients with RA from 21 hospitals in China showed that treatment with hydroxychloroquine (HCQ) was a protective factor against cardiovascular disease, whereas ILD, hypertension, and hyperlipidemia were independent risk factors. In our study, the UCD ranking of ischaemic heart disease in RA-ILD patients decreased from third to fifth, probably owing to the widespread use of HCQ and the effective management of ischaemic heart diseases. In this study, malignant neoplasms ranked fourth on the UCD list in RA-ILD patients during 2014–2018, and compared with 1999–2003, they showed an increasing trend. In fact, certain meta-analyses have suggested that RA patients, compared with the general population, tend to have lymphomas and lung malignancies (22). RA patients with ILD and a UIP pattern were often subjected to higher mortality rates than those with other patterns and were also more likely to die of lung cancer (19). Systemic connective tissue disorders also showed an upward trend in the UCD ranking of RA-ILD, possibly because of the widespread use of various autoantibody detection techniques. Some studies indicated that infectious diseases were increasingly frequent causes of death associated with RA, and that the most common cause of death was pneumonia based on chronic ILD (19, 23). Our study reached similar conclusions. However, surprisingly, influenza and pneumonia dropped out of the top 10 most frequent UCDs during the study period. In contrast, the incidence of other bacterial diseases was elevated. This could be related to more attention being paid to lung infections based on ILD.