Σκοπός: H μελέτη πραγματοποιήθηκε με σκοπό τη διερεύνηση της αποτελεσματικότητας και της ασφάλειας της θεραπείας με anti-TNF-α παράγοντες, σε σύγκριση με μη-βιολογικούς παράγοντες (nbDMARDs), όσον αφορά στην πνευμονική νόσο, σε ασθενείς με ρευματοειδή αρθρίτιδα. Υλικά και μέθοδοι: 82 διαδοχικοί ασθενείς, 29 άνδρες, 53 γυναίκες, ηλικίας 42-79 ετών, που είχαν διαγνωστεί με ρευματοειδή αρθρίτιδα, κατάλληλοι για θεραπεία με anti-TNF-α παράγοντες χωρίστηκαν σε δύο ομάδες μελέτης: 42 με προϋπάρχουσα διάμεση πνευμονοπάθεια σχετιζόμενη με την ΡΑ (RA-ILD) και 40 χωρίς διάμεση πνευμονοπάθεια (RA-ILD αρνητικοί). Συγκροτήθηκαν δύο έτερες αντίστοιχες ομάδες ελέγχου οι οποίες αποτελούνταν: η μία από 44 ασθενείς με προυπάρχουσα διάμεση πνευμονοπάθεια και η έτερη από 44 ασθενείς χωρίς προϋπάρχουσα διάμεση πνευμονοπάθεια, αμφότερες υπό αγωγή με μη βιολογικούς παράγοντες (nb DMARDs). Όλοι οι ασθενείς υποβλήθηκαν σε HRCT θώρακος στις οποίες έγινε βαθμονόμηση (scoring) των ευρημάτων, λειτουργικές δοκιμασίες πνευμόνων και μελέτη βιοδεικτών στο περιφερικό αίμα κατά την έναρξη της θεραπείας και ένα έτος μετά. Αποτελέσματα: Καταγράφηκε στατιστικά σημαντική μείωση της έκτασης της παγίδευσης αέρα και της πάχυνσης του τοιχώματος των βρόγχων ένα έτος μετά την έναρξη της θεραπείας, σε αμφότερες τις ομάδες μελέτης (RA-ILD και RA-ΙLD αρνητικοί) (p<0,05 ). Το εύρημα αυτό συνοδεύτηκε από στατιστικά σημαντική βελτίωση των παραμέτρων σπιρομέτρησης MMEF75-25, RV και του λόγου RV/TLC και στις δύο ομάδες μελέτης (p<0,05). Στην ομάδα RA-ILD η βαθμονόμηση σχετιζόμενη με την έκταση της διάμεσης πνευμονοπάθειας παρέμεινε αμετάβλητη μετά τη θεραπεία με anti-TNF-α παράγοντες. Κανένας ασθενής από την ομάδα μελέτης RA-ILD αρνητικοί δεν ανάπτυξε νεοεμφανιζόμενη διάμεση πνευμονοπάθεια. Σε αμφότερες τις ομάδες ελέγχου, τόσο τα ευρήματα της HRCT θώρακος όσο και οι λειτουργικές δοκιμασίες πνευμόνων δεν διέφεραν σημαντικά ένα έτος μετά την έναρξη της θεραπείας. Η ανάλυση παλινδρόμησης των HRCT ευρημάτων σε ομάδα ασθενών που έλαβαν μεθοτρεξάτη (ΜΤΧ) ανέδειξε σε αμφότερες τις ομάδες μελέτης και ελέγχου μια αρνητική συσχέτιση μεταξύ της δόσης της ΜΤΧ και της βαθμονομημένης έκτασης της διάμεσης πνευμονοπάθειας ένα έτος μετά την έναρξη της θεραπείας, καθώς και μεταξύ της δόσης της ΜΤΧ και της έκτασης παγίδευσης αέρα, τόσο κατά την έναρξη, όσο και ένα έτος μετά τη μετά τη θεραπεία. Συμπεράσματα: Η anti-ΤΝF-α θεραπεία, σε αντίθεση με την μη-βιολογική θεραπεία (DMARDs), οδήγησε στη βελτίωση της νόσου των μικρών αεραγωγών. Δεν υπήρξε εμφάνιση διάμεσης πνευμονοπάθειας ούτε παρόξυνση προϋπάρχουσας διάμεσης πνευμονοπάθειας σε ασθενείς που έλαβαν θεραπεία με παράγοντες anti-TNF-α. Τα ανωτέρω ευρήματα υποστηρίζουν την αποτελεσματικότητα και το ασφαλές προφίλ της θεραπείας αυτής σε ασθενείς με ρευματοειδή αρθρίτιδα.
Rheumatoid arthritis (RA) is a chronic, autoimmune systemic inflammatory disease. RA represents the most frequent inflammatory disease, affecting approximately 0.5-2% of the worldwide adult population, aged 35-55 years. The interaction of genetic and environmental factors results in a cascade of immune reactions, which ultimately lead to the development of synovitis, joint damage, and structural bone damage. The disease is associated with progressive disability due to multi-articular involvement, systemic complications and premature death with subsequent socioeconomic costs. It is characterized by synovial inflammation and hyperplasia of the peripheral joints, cartilage and bone destruction, especially at the small joints of the wrists and hands, as well as at the small joints of the feet.
Autoantibody production such as rheumatoid factor (RF) and anti-citrullinated protein antibodies (anti-CCP) seem to play a crucial role in the pathogenesis of the disease. Systemic features include skeletal, cardiovascular, pulmonary involvement and psychological disorders. Extra-articular manifestations appear in approximately 40-50% of RA patients, with a different incidence from country to country. Although RA itself is more common in females, extra-articular manifestations of the disease appear to be more frequent in males, may occur at any age after onset and include skin, eye, heart, lung, renal, nervous and gastrointestinal system involvement. Extra-articular organ involvement in RA is more frequently seen in patients with severe active disease, is associated with increased mortality and is more likely in those who have RF and/or are HLA-DR4 positive. Nodules are the most common extra-articular feature, and are present in up to 30%; many of the other classic features occur in 1% or less in normal clinic settings.
Pulmonary involvement in Rheumatoid Arthritis
Thoracic involvement occurs in 1-40% of patients with RA. Even though not clinically recognizable at early stages, changes include pleural, parenchymal, airway and lung vasculature involvement. Pleural disease is a common manifestation but usually asymptomatic, detectable in only 10% of cases. Lung involvement has been identified in early symptomatic disease and occasionally prior to the onset of articular symptoms. Rheumatoid arthritis-associated interstitial lung disease (RA-ILD) is the most common pulmonary manifestation observed in 50% of chest CTs and in 5% of chest radiographs and is associated with > 20% morbidity in patients with prevalent disease. Airway involvement in RA patients has been less emphasized in the literature comparing to ILD, with a prevalence reaching 51% in symptomatic patients. Small airways disease has been reported in up to two-thirds of RA patients with a high prevalence, even among those without ILD. High resolution computed tomography (HRCT) scans appear to be more sensitive compared to pulmonary function tests (PFTs) for small airways disease detection. There is controversy regarding the effects of TNF-α targeted therapy on the lung. Some studies report stability and others new-onset or progression of RA-ILD following anti-TNF-α treatment. PURPOSE This prospective study aims to evaluate the effect of tumor necrosis factor-α (TNF-α) inhibitors on airways and lung parenchyma compared to non-biological disease-modifying anti-rheumatic drugs (DMARDs), with regard to efficacy and safety. MATERIALS AND METHODS This prospective study was conducted from January 2011 to August 2015 according to the principles of the Helsinki Declaration and approved by the Institutional Ethics Committee. All patients gave written informed consent. Inclusion criteria for patients in the study group consisted of diagnosed rheumatoid arthritis (RA) according to the revised classification criteria for RA of the American College of Rheumatology and the European League Against Rheumatism, refractory to conventional treatment with non biologic disease-modifying antirheumatic drugs ((nb)DMARDs) and therefore candidates for anti-TNF-α agent therapy. Exclusion criteria included history of asthma, primary pulmonary hypertension, left heart failure and exposure to silica. Eighty-two (82) patients, 29 males, 53 females, of mean age 60 (range: 42-79 years) were prospectively recruited. Pre-existing RA-ILD (RA-ILD group) was present in 42/82, diagnosed on the basis of clinical findings, pulmonary function tests (PFTs) and chest HRCT findings. The remaining 40/82 had no imaging or clinical findings of ILD (RA-non ILD group). Twenty (20/82) patients were active smokers, 12 patients in the RA-ILD positive and 8 in the RA-ILD negative group, with a mean cigarette consumption of 22 pack/years. A control group of eighty-eight (88) RA patients diagnosed with the same inclusion criteria and following the same exclusion criteria who underwent (nb) DMARDs treatment were divided into two control groups: 44 patients with pre-existing RA-ILD and 44 without RA-ILD. All patients underwent paired inspiratory and expiratory chest HRCT, PFTs and laboratory tests at baseline and following 1-year treatment. Assessment of clinical response of RA was performed using DAS 28. PFT's, laboratory tests and DAS 28 were performed on average within 2 days (mean, 1.71 ± 5.9) of obtaining chest HRCT scans. Treatment Eighty-two RA patients received anti-TNF-α treatment: 68 received infliximab (recommended dose of 3mg/kg as an intravenous regimen at 0, 2 and 6 weeks, followed by a maintenance regimen of 3mg/kg every 8 weeks thereafter), 10 patients received etanercept (50mg, weekly) and 4 patients adalimumab (40mg, weekly). Anti-TNF-α treatment in all above patients was combined with low-dose methotrexate (MTX) (7.5mg per os/week). Eighty-eight control group patients received (nb) DMARDs treatment, among them 68 MTX alone (mean dose of 15 mg per os/week), 20 hydroxychloroquinine (mean dose 400mg per os/day) combined with low-dose MTX (7.5mg per os/week). High-Resolution CT All patients underwent chest HRCT exam consisting of paired inspiratory and expiratory scans performed on a multislice CT scanner (Siemens Somatom Sensation 64, Erlangen, Germany) using identical technical parameters including submillimeter slice thickness. In order to achieve reproducibility, patients were carefully trained to deeply inhale and hold their breath during each inspiratory scan and to deeply inhale then to forcefully and rapidly exhale and do not breathe for 10 seconds in order to acquire end-expiratory phase images. Inspiratory scans were evaluated for ground glass opacities (GGOs), nodules, reticulation, honeycombing and airway involvement (including bronchial/bronchiolar wall thickening, bronchiectasis and bronchiolectasis Bronchial wall thickening was considered present when the ratio between bronchial wall thickness and diameter of bronchus was > 0.2. Bronchiectasis was considered present when the broncho-arterial ratio was >1 and bronchiolectasis when peripheral bronchi were visible within 1cm of the costal pleural surfaces. CT sections were scored at five levels: 1) the origin of the great vessels, 2) the carina, 3) the pulmonary venous confluence, 4) between levels 3 and 5, and 5) 1cm above the right hemidiaphragm. The following features were quantified at each level: a) The extent of each abnormality described in the previous paragraph was estimated to the nearest 5% for each level and the overall extent was computed as the mean of the five section scores, b) the extend of ILD was graded as follows: 0 : ground-glass opacification alone; 1: fine intralobular fibrosis; 2 : microcystic honeycombing (air spaces up to 4 mm in diameter); 3 : macrocystic honeycombing (air spaces greater than 4 mm in diameter). The five section scores were summed to give the total ILD extent score. In order to detect air trapping areas on expiratory scans, a side by side comparison of inspiratory and expiratory CT images of the same area was performed and lung attenuation difference was measured by small regions of interest (1-2cm). Hypodense areas on expiratory scans showing attenuation increase less than 80 HU and involving more than 25% of the lobe, were regarded as air trapping. Areas of emphysema were excluded. The extent of air trapping (AT) on expiratory scans corresponding to small airways disease was evaluated by visual assessment using a semiquantitative scoring system estimating the percentage of lung that appeared abnormal on each scan. A 5%-point scoring system, as the one proposed by Webb et al. and Stern et al. estimated air trapping, on expiratory scans, at three different lung fields for each lung, six lung fields in total for both lungs: upper lungs fields from the lung apices to just above the level of the carina, middle lung fields between the level of the carina and the pulmonary veins and inferior lung fields from the pulmonary veins level till the level of the costophrenic angles. At each level and for each lung, a 5-point scale was used to estimate the percentage of air trapping extent visible to each radiologist: 0 = no air trapping, 1=1-25% of cross sectional area of the affected lung, 2= 26-50%, 3=51-75% and 4=76-100%. HRCT images were independently read by two chest radiologists (E.D, E.M), blinded to clinical and laboratory data. In cases of discrepancy, images were also evaluated by a third chest radiologist (M.R) and final decision was then reached by consensus among the three. Lung function tests All patients underwent complete pulmonary function tests (PFTs), including spirometry, lung volume and diffusion capacity measurement, at baseline and after one-year treatment. Spirometry, lung volumes using the helium-dilution technique and DLCO (corrected for haemoglobin) using the single breath technique were performed using a computerized system (Jaeger 2.12; MasterLab, Würzburg, Germany). Predicted values were obtained from the standardized lung function testing of the European Coal and Steel Community, Luxembourg (1993). Observed values were expressed as percentage of the predicted value, were compared with individuals of similar sex, age and height and were considered as abnormal if they were < 80% of the predicted values adjusted for age, sex, and height. The composite physiologic index (CPI) which represents the extent of fibrosis on HRCT, adjusting for emphysema in patients with idiopathic pulmonary fibrosis (IPF) was calculated using the following formula: 91.0 – (0.65 x percent predicted DLCO) – (0.53 x percent predicted FVC) 1 (0.34 x percentage predicted FEV1) in the RA-ILD group of patients. Laboratory tests Patients were evaluated at baseline and after one-year treatment for presence of rheumatoid factor (RF) and anti-citrullinated protein antibodies (anti-CCP). Evaluation of RA disease activity included recording of erythrocytes sedimentation rate (ESR) and C-reactive protein (CRP). Statistics The Kolmogorov-Smirnov test was used to determine whether the data obtained follow a normal distribution pattern. Group comparisons were made by analysis of variance, Student t test, Wilcoxon rank-sum test, or chi-square testing as appropriate. Linear regression between clinical parameters and the obtained data were analysed with the Linear (Pearson) correlation test and regression analysis was performed for MTX use between groups. Probability values (p-values) <0.05 were considered statistically significant. Statistical calculations were performed using SPSS 11.5 software (SPSS, Chicago, IL, USA). RESULTS There were no statistically significant differences in demographic parameters between the two patient study groups and the two control groups, respectively (Table 1). All patients had moderately to severely active RA. Intraobserver and interobserver reproducibilities were good for the detection and extent of HRCT findings in all pre and post-treatment scans. Findings at baseline In the RA-ILD study group, a usual interstitial pneumonia (UIP) pattern was found in 22 patients (52.5%), non specific interstitial pneumonia (NSIP) in 12 (28.5%) and cryptogenic organizing pneumonia (COP) in 8 (19%) (Fig. 1). All patients that showed air trapping areas on expiratory HRCT demonstrated abnormally low MMEF75-25, and abnormally increased RV and RV/TLC values, corresponding to small airways disease. There was no patient with large airways obstruction. In the RA-ILD control group, a UIP pattern was detected in 13 (29%) patients, NSIP in 16 (35.5%) and a COP pattern in 1 patient (2.2%). In the RA-non ILD study group, among nodules found, there were two biopsy proven necrobiotic rheumatoid nodules. On expiratory scans air trapping was depicted in 19 (47.5%) patients. All these patients showed abnormally low MMEF75-25 values at PFTs. Changes following one-year of anti-TNF-α treatment During the first 6 months of anti-TNF-α treatment 5 patients showed signs and symptoms of lower respiratory tract infection, 3 due to Streptococcus pneumoniae, belonging to RA-ILD group, 1 patient due to Legionella pneumophila and 1 due to Listeria monocytogenes, both belonging to the RA-non ILD study group. Regarding control groups three cases of opportunistic lung infections where depicted, two due to Pneumocystis Jirovecii in the RA-ILD control group and one case due to Mycobacterium Avium intracellulare in the RA-non ILD control group. In both study and control groups no case of mycobacterium tuberculosis infection was recorded. All the above cases were successfully treated without further complications. In the RA-ILD study group, overall there was a statistically significant decrease of bronchial wall thickening and of air trapping extent. There was no significant difference in the extent of ILD pre- and post-treatment (Fig. 3). Patients with decreased air trapping extent showed significantly increased MMEF75-25 as well as decreased RV and RV/TLC values, versus normal range (Table 3, Fig. 2BD). Anti-CCP was found to be significantly decreased. In the RA-ILD control group, a slight increase in ILD extent score and also in air trapping extent score was recorded, both statistically insignificant (Table 2 Fig. 2 A). These patients also showed MMEF75-25 decrease and RV and RV/TLC increase, although statistically insignificant. In the RA-non ILD study group, there was a significant decrease of bronchial wall thickening and air trapping extent while there was no significant increase in the extent of inspiratory HRCT findings. There was no evidence of new-onset ILD. At post-treatment PFTs, patients with decreased air trapping extent showed also a significant decrease of RV and RV/TLC and increase of MMEF75-25 values versus normal range. In the RA-non ILD control group an insignificant increase of air trapping extent score was depicted on expiratory HRCT scans. Concerning PFTs, RV and RV/TLC slightly increased while MMEF75-25 values further decreased, all statistically insignificant. No patient of this group developed new-onset ILD. DAS 28 was found significantly decreased in both RA-ILD and RA-non ILD study groups. No significant correlation was detected between MMEF75-25 values and RV or RV/TLC, respectively, in all groups. Methotrexate regression analysis In order to establish a clear role of anti-TNFα agents in this study we performed a detailed regression analysis of the MTX implication in our results. In this view, we tested whether MTX use (dose in mg) altered any of the tested parameters (PFTs, inspiratory HRCT findings and expiratory HRCT findings) between study and control groups. No patient in the study groups and control groups established MTX-induced pneumonitis or exacerbation of previous ILD related to MTX use. Regression analysis demonstrated that no parameter, including HRCT and PFT findings, showed any correlation with MTX use in the RA-non ILD control group, including the subgroup under MTX alone and subgroup under hydroxychloroquinine combined with MTX, as well as between each of the above subgroups and the RA-non ILD study group. These regression analysis results remained unchanged both at baseline and following one-year treatment. MTX regression analysis revealed a significant negative correlation at one-year study between ILD extent score and MTX dose regarding RA-ILD control (both subgroups, under ΜΤΧ alone and hydroxychloroquinine with ΜΤΧ) and RA-ILD study group (p=0.0135, r= -0.3973, r2= 0.1578). Another negative correlation was also depicted in the RA-ILD study group and the control group with regard to air trapping extent and MTX dose, both at baseline (p=0.0139, r= -0,3908 , r2= 0.1527) and at one year follow-up study (p=0.0425, r= -0.3265, r2= 0.1066). Composite Physiologic Index (CPI) There were no significant alterations in CPI measurements in the RA-ILD study group before and after anti-TNF-α treatment. Regarding the RA-ILD control group, an increase in CPI was recorded, even though this difference was not statistically significant. CPI positively correlated with honeycombing and GGOs in both RA-ILD study and control groups, at baseline and after one-year treatment. Regarding PFTs, CPI negatively correlated with predicted FVC and TLC in both RA-ILD study and control groups, before and after treatment. DISCUSSION A statistically significant decrease in the extent of air trapping and bronchial wall thickening was observed after anti-TNF-α treatment in both RA-ILD and RA-non ILD study groups. All patients with decreased air trapping extent on expiratory HRCT showed improvement of MMEF75-25, RV and RV/TLC values versus normal range at post-treatment PFTs. The above suggest that TNF-α targeted therapy may play an important role in stabilizing and even improving small airways disease in RA patients, either with or without ILD. To the best of our knowledge, the present study, following a case report by Cortot et al. is the first prospective study in humans to prove through HRCT findings and PFTs, the beneficial effect of these agents on small airways disease. Our findings can be explained by knowledge of the mechanism of anti TNF-α treatment: TNF-α is a cytokine known to mediate and augment inflammatory reactions and to enhance fibroblast proliferation at the level of the bronchial and bronchiolar wall. A study in mice proved that targeted TNF-α overexpression in the lungs is related to chronic inflammatory infiltration of the interstitium by lymphocytes and macrophages, especially localized in areas adjacent to the pleura and bronchioles. TNFα inhibition has been suggested to reduce inflammation, epithelial loss, fibrosis, and bronchiolar obliteration early in the development of obliterative bronchiolitis. Thus, the effect of anti-TNF-α treatment in air trapping extent, bronchial wall thickening and PFTs in our study can be explained by its inhibiting effect on airway wall thickening. The increased thickness of the submucosa and proliferation of smooth muscle and connective tissue in RA patients, is related to bronchial wall thickening and air trapping on expiratory HRCT scans. This is in accordance with the significant association recorded in the present study between air trapping presence and extent with bronchial wall thickening in all patient groups, at baseline and after treatment. In the existing literature, the effect of TNF-α in fibrotic disease is controversial. Overexpression of this factor in the lungs has been linked to fibrosis, while inhibition of TNF-α signaling can possibly prevent interstitial lung disease. It has been suggested that inflammation leading to recurrent alveolitis may trigger pulmonary fibrosis. It is in this inflammatory stage that anti-TNF-α agents may have a beneficial effect. Moreover, previous studies suggest an association between seropositivity for antiCCP antibodies, and the presence of ILD in RA patients. In accordance, we detected a statistically significant higher serum level of anti-CCP antibodies in the RAILD study group compared to the RA-non ILD. Following treatment, serum levels of anti-CCP significantly decreased in the RA-ILD study group while such decrease was not registered in the RA-ILD control group. In the present study, most patients in the RA-ILD study group, according to PFT parameters and peripheral blood biomarkers, appeared in an inflammatory stage before anti-TNF-α treatment and this is probably the reason why anti-TNF-α appeared to stabilize inspiratory HRCT findings and reduce anti-CCP antibodies levels in this group. More trials are needed to investigate the effect of anti-TNF-α treatment in RA patients at a non-inflammatory stage. There are several reports suggesting that anti-TNF-α agents may trigger pulmonary fibrosis in RA patients. During this one-year follow-up no patient developed exacerbation or progression of pre-existing ILD while none of the RA-non ILD group patients established new onset ILD. All the above, in accordance with other studies, support the safety of anti-TNF-α agents in patients with RA-ILD with regard to the occurrence and extent of ILD. Methotrexate-induced lung toxicity has been widely described in literature. Pneumonitis related to MTX appears to occur less often than previously thought. In our study, no patient developed MTX-induced pneumonitis or any other type of lung toxicity related to MTX supporting the reported favourable safety of its use. Furthermore, no significant correlation was recorded between MTX use, PFTs and HRCT findings extent in the RA-non ILD study group and control groups. In contrast, in the RA-ILD study group and the control group, methotrexate regression analysis showed a significant negative correlation at one-year interval between ILD extent score and MTX dose further supporting the idea that low dose MTX treatment may be related to more extended HRCT findings of MTX-induced ILD comparing to higher dose treatment. Another negative correlation was also depicted in the RA-ILD study group and the control group with regard to air trapping extent and MTX dose, both at baseline and at one-year follow-up study. These negative correlations were not followed by similar negative correlations regarding PFT parameters. Therefore, even though the above correlations may indicate a possible relation between low dose MTX treatment and a more extended MTX induced pneumonitis as well as low dose MTX and more extended air trapping on expiratory HRCT, larger scale and longer term studies are needed to draw safer conclusions. CPI is an important diagnostic tool that strongly correlates with the CT extent of pulmonary fibrosis and is linked to mortality in histologically proven UIP and idiopathic pulmonary fibrosis (IPF). In our study, CPI levels did not differ significantly after treatment in all groups, thus supporting the stabilizing role of treatment with regard to lung fibrosis. It is not a surprise that CPI correlated positively with honeycombing and GGOs on inspiratory HRCT scans and negatively with predicted FVC and TLC, since all these parameters are linked to restrictive lung disease. The risk of opportunistic infections associated with anti-TNF-α treatment is widely known. During the first 6 months of anti-TNF-α treatment only 5 patients showed signs and symptoms related to opportunistic lower respiratory tract infection. Patients of both study groups were successfully treated without complications while no mycobacterium tuberculosis or atypical mycobacterial infection was recorded in both study groups. Limitations of this study include the relatively small number of patients, the wide age range and duration of disease as well as the relatively short term of follow-up. Furthermore, most patients in the RA-ILD and RA-non ILD study groups were treated with infliximab, while a small percentage received other anti-TNF-α agents which did not permit further statistic correlations to be performed separately for each agent. Even though a one-year follow-up period is adequate to detect any significant short or medium term effect regarding pulmonary manifestations, longer-term studies are needed in order to solidify the safety of these agents. In addition, other biomarkers such as KL-6 reported in patients with ILD, were not investigated. CONCLUSIONS 1. The results of the present study support the beneficial effect of anti TNF-α agents with regard to small airway disease by improving air trapping extent, bronchial wall thickening and MMEF75-25, RV and RV/TLC values. 2. In patients with RA-ILD anti-TNF-a treatment had a stabilizing effect with regard to ILD, probably more efficiently during the inflammatory stage of disease. 3. Anti-TNF-α agents were not associated with ILD development or progression and opportunistic infections were limited, thus supporting the safety profile of this treatment.