Interstitial lung disease(ILD) is one of the most common and severe extra-articular manifestations of rheumatoid arthritis (RA)1–4. In more than three-quarters of patients, ILD occurs after a diagnosis of RA, usually within the first 5–10 years of disease progression1–7. In these cases, it is not uncommon for it to be diagnosed late, as it is often asymptomatic or paucisymptomatic in its early stages1–7. In fact, when the entire RA population was screened for ILD by chest CT scan, a not insignificant percentage of subclinical disease was detected (between 11.9% and 55.7%), which confirms that this complication is underdiagnosed8,9. Less frequently, ILD occurs at onset of RA or precedes joint manifestations by months or years. In the latter scenario, ILD-RA is often confused with idiopathic forms, despite their different prognosis and treatment1–3,10,11.

Although a significant improvement in its prognosis has been observed12, ILD is still the second leading cause of premature death in RA, after cardiovascular complications1–4,10. Despite its frequency and severity, there are currently no recommendations on screening for this complication, in either the initial assessment or in the follow-up of RA patients.

Based on the above, and to respond to this clinical need, the objective of the RA-EPIDSER project was to draft a multidisciplinary proposal for screening criteria that would enable the early identification of patients with ILD-RA.

A systematic review of the literature and consensus techniques (Delphi method with a minimum of two rounds) were used to determine the degree of consensus of experts based on their clinical experience and scientific evidence.

ILD is one of the most frequent and severe extra-articular manifestations of RA and is currently the second leading cause of premature death from RA. Median survival from diagnosis of ILD varies between 2.6 and 8.1 years13.

However, unlike in systemic sclerosis, in which ILD is also the first or second cause of death14, in RA there are no recommendations for screening for this complication, despite the fact that it is underdiagnosed8,9 and that it often goes unnoticed until advanced stages, when the extent of fibrotic changes limits therapeutic possibilities and survival (it should be noted that interstitial pneumonia is the most frequent histopathological pattern in RA).

To correct this situation and improve the prognosis and quality of life of patients with ILD-RA, the SER promoted the AR-EPIDSER study, in which a group of rheumatology and pulmonology experts was formed to jointly develop criteria for the early identification of patients with ILD-RA to serve as a benchmark for referral of these patients to rheumatology and pulmonology services. This multidisciplinary proposal aims to resolve some common clinical questions and facilitate decision-making in different clinical scenarios. In the absence of evidence to support the cost-effectiveness of universal screening, selective screening proposals were chosen based on the patients’ clinical risk.

Regarding the criteria for when to investigate for ILD in patients diagnosed with RA, the desired consensus was reached in most of the situations considered, and a screening algorithm useful in clinical practice was proposed (Fig. 1). As a limitation, in the situations referring to screening in asymptomatic patients with normal auscultation, consensus was not reached on some of the items included in the scoring system, specifically on the cut-off point and score given to age, on the score assigned to male sex, and on how cumulative smoking should be quantified. Consensus was not reached on how to screen patients with respiratory symptoms of more than 3 months' duration either. In these cases, the SC opted to include in the final version of the proposal the option with the highest degree of agreement, which for most of these criteria was 71.4%, i.e., very close to consensus. Sixty was finally selected as the cut-off point for age, which, in addition to being the most voted for, is also a prognostic factor associated with higher mortality13–35.

Fig. 1.

Proposed screening algorithm for interstitial lung disease (ILD) screening in patients diagnosed with rheumatoid arthritis (RA).

* See risk factors and their score in Table 2.

? If screening tests are negative, screening will be repeated once a year by spirometry + DLCO.

(0.28MB).

To prevent this proposal from rapidly becoming obsolete, the SC considered the option of incorporating the new biomarkers under investigation into the scoring system27–31, fundamentally anti-carbamylated protein (anti-CarP) antibodies. A correlation has been shown to exist between levels of all anti-CarP specificities (foetal bovine serum, fibrinogen, and chimeric fibrin/filaggrin homocitrullinated peptide) and the presence of ILD, after adjustment for other known risk factors27. Promising preliminary results have also been obtained with serum levels of extracellular matrix metalloproteinase 7 (MMP-7), interferon gamma-inducible protein-10 (IP-10) or CXCL10, interleukin 18, and 90 and 70-kDa heat shock protein (HSP90/70)27–31. Finally, it was decided not to include them as they are not yet available in clinical practice and have not yet been shown to have a higher predictive value for the development of this complication than ACPA. Genetic biomarkers, including mutations in the MUC5B gene32 and mutations in telomerase genes that cause accelerated telomere shortening33,34, are not yet available in clinical practice. Nevertheless, and as an indirect way of recognising its future usefulness and the importance of genetic predisposition in the development of ILD in a subgroup of patients, it was agreed to include a family history of ILD as a risk factor to consider in the algorithm.

An important message that the authors of both specialties want to highlight in this proposal is the usefulness and importance of respiratory auscultation in the rheumatology consultation for ILD screening. The presence of “Velcro-like” crackles correlates well with the presence of pulmonary fibrosis36–38. In addition, two studies have shown that conventional respiratory auscultation has acceptable diagnostic accuracy for the detection of ILD when compared with that obtained with a digital electronic stethoscope36,37.

Of the tests to screen for ILD, the scientific committee decided not to include lung ultrasound for the time being, considering that it is a technique that has not yet been sufficiently validated to diagnose this complication. A recent systematic review by the Outcome Measures in Rheumatoid Arthritis Clinical Trials (OMERACT) group confirms its good face and content validity, and its feasibility, but it does not yet have criterion validity, reliability or, above all, sensitivity to change39. Limitations still to be resolved include the lack of a clear definition of normality, the absence of agreed definitions of the lesions assessed, and the lack of standardisation of the technique, image acquisition, and the number of areas to explore (which in the studies varies from 10 to 72). There was also no consensus on the quantification of pneumopathy, on the scoring system, or on the optimal probe to use (3–3.5 to 5, or 7.5 MHz). Likewise, there is a lack of longitudinal follow-up studies and studies assessing their ability to diagnose ILD in early stages (subclinical ILD)39.

Another important clinical scenario that considers these recommendations and of which pulmonologists should be made aware is the pulmonary debut of RA. As already mentioned, in 10%–14% of cases of ILD-RA, pulmonary involvement precedes joint manifestations by months or years1–3,10,11. In these cases, early referral to the rheumatologist by the pulmonologist will make it possible not to delay diagnosis and to initiate remission-inducing treatment as soon as the clinical diagnosis of the disease is established, thus taking advantage of the so-called “window of therapeutic opportunity”, understood as a period of time very early in the onset of the disease in which, with appropriate intervention, we can modify the course of the disease and avoid joint damage and associated complications.

Therefore, the initial assessment of all patients with ILD should include a detailed clinical history and a thorough physical examination to investigate the presence of extrapulmonary manifestations, together with an analytical study that includes at least the erythrocyte sedimentation rate (ESR), C-reactive protein and testing for RF, ACPA, antinuclear antibodies, and myositis-specific autoantibodies, as recommended by the European and American societies of Pneumology40. We also include in our proposal the screening for RA in the follow-up of patients with ILD of undetermined aetiology, with the frequency described in the present study.

In this clinical scenario, the predictive value of ACPA appears to be superior to that of RF41–44. Positive RF is not pathognomonic of RA, having been described in 5%–10% of the healthy population, especially in elderly patients, although generally at low titres. Hence, in the proposal, it was decided to consider only RF titres ≥2 at ULN. The joint manifestations included, based on which to suspect early arthritis, are easily identifiable by a non-rheumatologist. Their validity has been contrasted in population-based studies, including one conducted in Spain45,46.

The number of experts participating in the Delphi is within the range accepted as valid and all of them completed their participation during the successive rounds. The threshold set at 75% to define consensus and the number of Delphi rounds are also within the parameters considered adequate47,48. Despite following a rigorous methodology, this proposal has several limitations. The main limitation, in addition to the aforementioned lack of consensus for some of the criteria, is that this proposal is only an expert consensus, the usefulness of which has not yet been tested in validation studies. Moreover, given that the panel was only a national one, it may not be possible to extrapolate these criteria to other health systems.

Current evidence flags up the need for a common strategy between rheumatologists and pulmonologists to improve the early diagnosis of patients with ILD-RA, given its prognostic and therapeutic implications. To respond to this need, for the first time a multidisciplinary group of rheumatologists and pulmonologists have developed a consensual proposal for selective screening criteria as an aid to clinical decision-making. Its usefulness in identifying these patients in clinical practice will need to be confirmed in future validation studies.

All authors have contributed substantially to: a) the conception and design of the study and analysis of the data; b) drafting the article or critical revision of the intellectual content; and c) the final approval of the submitted version.

This project was funded by Bristol Myers Squibb, who were not involved in the design of the study, the conduct of the study, or drafting of this article.

The authors have no conflict of interests to declare.