This retrospective nested case–control study compared JAKi with ABT to clarify the safety and efficacy of JAKi in patients with RA-ILD. The results showed that there were no differences in the drug persistency rate and pulmonary complication rates between JAKi and ABT. The reason for drug discontinuation did not differ significantly between the two groups, and no exacerbation of the existing ILD was reported in either group. JAKi was as effective as ABT for treating arthritis, which enabled us to taper the PSL dosage. To the best of our knowledge, this is the first study to show that JAKi can be a safe and effective option for patients with RA-ILD by comparing it with ABT.
RA treatments have progressed considerably; in fact, remission or low disease activity of RA can be achieved in many patients using a variety of drugs, but no standard treatments have been established for RA-ILD yet. It is one of the leading causes of death and an essential prognostic factor [1, 2]. Choosing an anti-rheumatic drug is often difficult to for several reasons. First, some anti-rheumatic drugs can cause drug-induced pneumonia or exacerbate existing ILD. MTX and LEF are known to induce drug-induced pneumonia [4, 5]. As for bDMARDs, ILD exacerbation is reportedly caused by TNFi [24, 25] and anti-IL-6 drugs [26, 27]. Additionally, there is also fear for an increased risk of infections due to bDMARDs [6, 7].
There is no established evidence for the safety of anti-rheumatic drugs for patients with RA-ILD; thus, treatments are often given empirically. There is growing evidence about the safety of ABT and RTX for patients with RA-ILD. The safety of abatacept for patients with RA-ILD has been shown in some case reports and retrospective studies [28–30]. A multicenter, retrospective observational study in Spain showed the efficacy of ABT for arthritis without exacerbating the existing ILD in 263 patients with RA-ILD [28]. Likewise, the safety of RTX is also reported by some studies [31, 32]. Currently in Japan, ABT is commonly used because RTX is not allowed for RA patients. Presently, there is a lack of solid evidence; thus, the question ‘which anti-rheumatic drug can be used safely for patients with RA-ILD’ remains unanswered.
JAKi inhibits the JAK–STAT pathways and broadly blocks the pathways on inflammation [10]. In the lung specimen from patients with RA-ILD, numerous CD4 positive T cells are present; thus, it can be speculated that JAKi can suppress the activation of alveolar inflammatory cells by inhibiting multiple cytokines via the JAK–STAT pathway [13, 33]. Moreover, there is growing evidence that the JAK–STAT pathway is activated under a number of profibrotic cytokines and engages in fibrotic processes. Especially, the JAK2–STAT3 pathways are predominant in ILD, and phosphorylation of the pathway leads to cellular fibrotic processes, including epithelial/fibroblast to mesenchymal transition [14]. Another study has shown that JAK2–STAT5A & B pathway is involved in fibrosis in RA-ILD [15]. A basic research study showed that tofacitinib successfully suppressed the progression of ILD in SKG mice, facilitated the expansion of myeloid-derived suppressor cells in the lung and ameliorated ILD by suppressing the proliferation of T cells and differentiation of Th17 cells [34]. From these findings, we speculate that JAKi can inhibit the progression of RA-ILD by blocking these processes.
In the clinical field, some case reports suggest the efficacy of JAKi in ILD. In MDA5 antibody-positive dermatomyositis, several case reports show the efficacy of tofacitinib [35–37]. As for RA-ILD, only a few case reports or retrospective studies showed the efficacy of JAKi. JAKi as treatment for arthritis without exacerbating the existing ILD was effective in four patients [38]. Moreover, our previous observational study showed is efficacy for arthritis without causing exacerbation of ILD or increasing the risk for infection in relatively mild patients with RA-ILD [16]. Comparing JAKi with RTX, no significant difference in hospitalization and mortality rates was reported [18]. Additionally, a retrospective study in Italy recently showed that treatment with JAKi or ABT was related to stability or improvement of RA-ILD in many of the patients [17]. As mentioned, there is growing evidence of JAKi use for patients with RA-ILD, but there is no solid evidence yet, especially in terms of its safety. The present study showed no difference in the drug persistency and pulmonary complication rates between ABT and JAKi. Our result provides information about the safety of JAKi in patients with RA-ILD and indicates that JAKi can be a safe treatment option in these patients, which can be helpful in clinical practice.
Our research also showed the efficacy of JAKi for arthritis without exacerbating ILD. There is quite solid evidence showing the efficacy of JAKi for RA without extraarticular lesions [11, 12]. However, its effectiveness for cases of arthritis with pulmonary complications has rarely been reported. The treatment of the arthritis is essential because the high disease activity of RA can be a risk factor for ILD exacerbation [3]. Therefore, our result provides clinically important evidence. Especially, JAKi can be a good treatment option for patients who cannot use ABT for some reasons. For example, ABT is reported to be less effective in ACPA-negative patients [39, 40]; thus, JAKi can be a preferable option for such patients. JAKi. can be also used for patients for whom ABT was not effective or ABT cannot be used because of adverse events.
Another clinically important result is that JAKi helped in tapering the PSL dose
in patients with RA-ILD. In addition to the numerous side effects of PSL, such as diabetes or osteoporosis, PSL usage increases the risk of developing progressive fibrosing interstitial lung disease which is associated with increased mortality in patients with ILD [41]. Additionally, PSL can increase the risk of infections, which are also a risk factor for ILD exacerbation [42]. For these reasons, tapering of the PSL dose not only can improve the patients with RA-ILD quality of life by getting rid of the troublesome side effects, but also can improve their prognosis.
This study included 26 and 45 patients for the JAKi and ABT groups, respectively, which is greater than the number of cases analyzed previous studies. Moreover, by selecting patients with propensity score matching, there is little influence of confounding factors. This result cannot be applied to all RA-ILD patients because our study included a higher number of patients with an NSIP pattern (> 80%) than those previously reported [22, 23]. We should also be monitor patients for ILD exacerbation as a case of ILD exacerbation possibly caused by JAKi has been reported [43]. In addition, we should pay close attention to cases of pulmonary infections due to the increased risk of infectious disease by JAKi [11], especially in patients with bronchiectasis or a UIP pattern.
This study has several limitations. First, this retrospective study was conducted at a single center, and the number of the patients is limited. Second, we did not evaluate the lung function by respiratory function test or CT to investigate the efficacy of the treatments for ILD. Third, our sample included more patients with NSIP patterns, who carry quite a low risk for exacerbations. Therefore, information on high-risk cases was not sufficient. Finally, because of the limited number of patients included, we did not conduct a subgroup analysis; thus, the question ‘who among the patients will benefit more from the JAKi treatment’, or ‘which JAKi type is the most effective’ has remained unanswered. It can be speculated that JAKi are effective especially for those with NISP patterns whose main pathophysiology is the invasion of inflammatory cells. As for differences among the JAKi types, it is difficult to determine which JAKi is the best for patients with RA-ILD. Regarding efficacy, pan-JAK inhibitors, rather than JAK1 selective inhibitors, may work more effectively, especially for cases with a UIP pattern, considering a predominant role of JAK2 in fibrosis of ILDs. However, in terms of the risks for infection, the use of JAK1 selective inhibitors seem to be safer. To clarify these questions, further studies, especially multicenter, prospective cohort studies, are necessary in the future.