American Journal of Respiratory and Critical Care Medicine

The aim of the present study was to assess the prevalence and characteristics of airways involvement in rheumatoid arthritis (RA) patients in the absence of interstitial lung disease. We prospectively evaluated, with high-resolution computed tomography (HRCT) and pulmonary function tests (PFTs), 50 patients with RA (nine males and 41 females; mean age: 57.8 yr), including 39 nonsmokers and 11 smokers (mean cigarette consumption: 15.3 pack-yr) without radiographic evidence of RA-related lung changes. PFTs demonstrated airway obstruction (i.e., reduced FEV1/VC) in nine patients (18%) and small airways disease (SAD) (i.e., decreased FEF25–75, defined as exceeding the predicted value by 1.64 residual SD [RSD] or more, and/or an increased phase III slope > 2 SD by single breath nitrogen washout) in four patients (8%). HRCT demonstrated bronchial and/or lung abnormalities in 35 cases (70%), consisting of air trapping (n = 16; 32%), cylindral bronchiectasis (n = 15; 30%), mild heterogeneity in lung attenuation (n = 10; 20%), and/or centrilobular areas of high attenuation (n = 3; 6%). Airway obstruction and SAD were correlated with the presence of bronchiectasis and bronchial- wall thickening (p = 0.003), and with bronchial infection (p = 0.01), but were unrelated to rheumatologic data. FEF25–75 was reduced and the slope of phase III was increased in patients with airway changes on HRCT scans, whereas no PFT abnormalities were found in 13 of 15 patients with normal HRCT scans. HRCT depicted features of SAD in 20 of the 33 patients with normal PFTs. HRCT findings were unrelated to rheumatologic data. A high prevalence of airway abnormalities as assessed with HRCT and/or PFTs was observed in our RA population. HRCT appears to be more sensitive than PFTs for detecting small airways disease.

Pulmonary involvement is a frequent extraarticular manifestation in rheumatoid arthritis (RA), but the prevalence and nature of rheumatoid lung disease have not yet been definitely established. Interstitial lung disease and subclinical alveolitis have been found in up to 40% of RA patients (1, 2), and the prevalence of airways disease has been reported to have a similar percentage in some series (3, 4). A possible association between RA and bronchiectasis was suggested by Walker (5), and has recently been reassessed with high-resolution computed tomography (HRCT), enabling the identification of bronchiectasis in up to 35% of cases (6, 7). Small airways involvement is more controversial, with some series showing airway dysfunction in 38 to 65% of patients (3, 4, 8, 9), whereas reduction in midexpiratory flow rates in RA has been related to cigarette smoking or lung restriction (10). At present, neither large nor small airways disease has been convincingly shown to constitute a definite manifestation of RA disease. The purpose of the present study was therefore to evaluate the clinical, functional, and HRCT characteristics of airways involvement in RA patients without interstitial lung disease.

In 1994 and 1995, we selected 69 consecutive patients with RA from a hospital rheumatology department to be included in a prospective study aimed at evaluating airway changes in RA. Only rheumatoid subjects, with RA as defined by the American Rheumatism Association criteria, were included (11). The systematic evaluation of lung function in RA patients was approved by the hospital ethics committee, and written informed consent was obtained from all patients before their inclusion in the study. Of the initial population, 19 patients were excluded owing to: (1) the concurrent presence of lung changes precluding an accurate evaluation of airways, including a previous history of chronic obstructive lung disease (n = 7, including two asthmatic subjects), pleuropulmonary sequelae of tuberculosis (n = 3), a previous history of thoracic radiation for breast carcinoma (n = 3), pleural effusion (n = 2), or lung fibrosis secondary to RA (n = 2); or (2) the inability to undergo both pulmonary function tests (PFTs) and HRCT (n = 2) at the same session.

The study group consisted of 50 outpatients (nine males and 41 females), ranging in age from 34 to 73 yr (mean ± SEM: 57.9 ± 1.5 yr). Forty patients were lifelong nonsmokers; six others were current smokers and four others were ex-smokers with a mean (± SEM) cigarette consumption of 15.9 (± 2.8) pack-yr. At the time of pulmonary evaluation, the mean (± SEM) duration of articular disease was 14.4 (± 1.3) yr (range: 1 to 48 yr). None of the patients included in the study had a previous history of occupational dust or fume exposure.

Clinical Evaluation

A standardized pulmonary evaluation was systematically conducted, and included a determination of the patient's smoking history and detailed determination of pulmonary symptoms, and a clinical examination. Clinical evaluation of articular disease was based on the Steinbrocker functional capacity (range: I to IV, where IV represents maximal disability), the Ritchie articular index (range: 0 to 78, where 78 represents maximal pressure-induced pain in all joints), and the health assessment questionnaire (HAQ) in its validated French version (12). The eight domains of the HAQ measure arthritis-specific function on a scale of 0 to 3, where 0 represents no limitation. The overall HAQ score is calculated as the average of these scores, and ranges from 0 to 3.

Ongoing and prior drug treatments and any adverse reactions to the drugs were noted. Patients were also questioned about symptoms of dry mouth and dry eyes. The diagnosis of secondary Sjögren's syndrome was defined by the European criteria (13).

Immunologic Assessment

Titers of rheumatoid factor (RF) were determined by laser nephelometry. A titer 40 IU was considered to indicate seropositivity. Antinuclear antibodies (ANAs) were detected by indirect immunofluorescence with the HEP2 cell line.

Pulmonary Function Tests

Spirometry was performed with a pneumotachometer in a body plethysmograph (Model 1085; Medgraphics, St. Paul, MN), and included measurement of FEV1, FVC, slow vital capacity (SVC), FRC, residual volume (RV), and TLC. Maximal midexpiratory flow (FEF25–75) was obtained from the best flow–volume curve. Diffusing capacity for carbon monoxide (Dl CO) and single-breath nitrogen washout (SBN2) measurements were made with a Medgraphics PFDX. Dl CO was measured by the single-breath technique and expressed in ml/min/mm Hg. The SBN2 test was performed as previously described (14). A minimum of three tracings were obtained for each subject. Tracings in which VC was not within 10% of the largest SVC or in which the expiratory flow exceeded 0.75 L/s were excluded. The slope of the alveolar plateau or phase III (ΔN2/L) was chosen as the most relevant parameter of the nitrogen washout test.

Results were expressed as percentages of predicted values (15), and FEV1 reversibility was defined as an increase in FEV1 as a percentage of the initial value. We defined as abnormal any spirometry measurement that was more than 1.64 residual standard deviations (RSD) below or above the predicted value. Dl CO/Va was considered significantly decreased when it was below 80% of the predicted value, since the predicted values we used do not include an RSD for this parameter (15). The slope of phase III was considered abnormal if it was more than 2 SD above the predicted values for nonsmoking men and women (14).

Airway obstruction was defined by a decrease in the FEV1/VC ratio of more than 11.8% from predicted values in men and 10.7% in women (15). Small-airways disease (SAD) was defined by a decreased FEF25–75 and/or an increased ΔN2/L as defined earlier.

High-resolution CT Scan

A total of 50 HRCT examinations were performed with a Somatom Plus S CT unit (Siemens, Erlangen, Germany). Scans were obtained with 1-mm-thick sections at 10-mm intervals, extending from the lung apices to below the costophrenic angles. A 350-mm field of view and a 512 × 512 reconstruction matrix were used. Examinations were performed at 137 kV and 255 mA. Images were reconstructed with a high-spatial-frequency algorithm for parenchymal analysis and a standard algorithm for mediastinal evaluation. CT scans were obtained at the suspended end-inspiratory volume, with imaging times of 1 s. Patients were scanned in the supine position, with additional scans obtained in the prone position to demonstrate the reversibility of dependent areas of attenuation whenever necessary. Expiratory scans were systematically obtained at 30-mm intervals to detect areas of air trapping. In six patients, evaluation of subtle rounded opacities within the lung parenchyma was optimized by means of a focal spiral CT acquisition over a 1-cm region of interest (ROI) using a sliding-thin-slab, maximum-intensity projection (i.e., the STS–MIP technique) (16).

In cases of articular limitation at the level of the shoulders, CT was performed with the patient's arms positioned alongside the body without any adverse effects on image quality. All images were obtained at window levels appropriate for lung parenchyma (window width: 1,600 HU; window level: −600 HU) and mediastinum (window width: 350 HU; window level: 50 HU). CT scans were interpreted in random order by two observers who were blinded to the patient's RA disease activity, pulmonary symptoms, and PFT results; decisions were made by consensus.

Three categories of CT abnormalities were evaluated on inspiratory scans: (1) CT signs suggestive of airways disease (17), including bronchial-wall thickening; bronchiectasis, including central bronchiectasis (characterized by lack of progressive tapering of lobar and segmental bronchi on successive scans) and peripheral bronchiectasis (recognized by the finding of abnormal visualization of subsegmental and smaller airways in peripheral locations); centrilobular nodular and branching areas of high attenuation, suggestive of bronchiolar involvement; mosaic perfusion, recognized by the presence of areas of decreased attenuation on inspiratory lung scans; air trapping, as assessed on scans obtained at end-expiration, and recorded as present upon the identification of areas of abnormally low attenuation larger than three adjacent secondary pulmonary lobules; (2) CT features suggestive of rheumatoid lung lesions (7), including rounded areas of attenuation ranging from 3 mm to 30 mm in diameter, suggestive of rheumatoid nodules; rounded areas of attenuation less than 3 mm in diameter, and immediately adjacent to the chest wall, which were recorded as subpleural micronodules; and (3) additional CT abnormalities, including emphysema, characterized by areas of decreased attenuation, disruption of the vascular pattern, and absence of a well-defined wall (i.e., nonbullous emphysema); bullae, defined as regions of emphysema with a well-defined wall 1 to 2 mm thick; areas of linear attenuation, which included septal lines (identified as fine linear opacities or as a polygonal pattern of multiple polygonal lines) and nonseptal lines; increased lung attenuation, which included areas of ground-glass attenuation (defined as areas of minimally to markedly increased attenuation in which the bronchi and vessels remained visible); and air-space consolidation (defined as areas of increased attenuation with obscuration of adjacent bronchial walls and vessels). Each of the aforementioned CT signs was coded separately as being present or absent.

To determine the distribution of parenchymal abnormalities, each lung was divided into three zones: upper (superior to the level of the carina), middle (between the level of the carina and the level of the inferior pulmonary veins), and lower (inferior to the level of the inferior pulmonary veins). Each sign of lung involvement seen at CT was coded separately as being present or absent in the six areas, thus defining the vertical extent of lung changes. Transverse distribution of lung changes, including anteroposterior and central versus peripheral distribution, was also subjectively assessed on a visual basis in each of the six areas previously defined, with an average distribution determined for the overall lung. The distribution of parenchymal lesions in the right and left lungs was also noted.

Statistical Analysis

When applicable, data are reported as mean ± SEM, and are expressed as absolute values or as percent of predicted normal values. The Mann–Whitney U test was used to analyze differences between groups. Comparison of qualitative data between groups was tested with Pearson's chi-square probability, with Yates's correction when appropriate. Correlations between variables were assessed with Spearman's rank order test. A value of p < 0.05 was considered significant.

Characteristics of RA Disease

The mean duration of RA was 14.4 ± 1.3 yr. Twenty patients had subcutaneous nodules and 13 had a secondary Sjögren's syndrome. The mean HAQ score was 1.66 ± 0.1. The mean erythrocyte sedimentation rate (ESR) was 46 ± 4 mm/h. Fourteen patients were negative for RF and 12 were ANA positive. Thirty-three patients were taking oral steroids and 38 were taking disease-modifying drugs at the time of assessment. Drug history was, however, difficult to analyze, since patients had taken a mean of 5 ± 2.8 second-line drugs for RA at the time of assessment. Characteristics of RA disease and its treatment at the time of the study are summarized in Table 1.

Table 1. CLINICAL AND BIOLOGIC CHARACTERISTICS OF ARTICULAR DISEASE

Steinbrocker functional class, n
I4
II18
III28
Ritchie articular index7.9 ± 0.9
(0–27)
HAQ index1.66 ± 0.1
(0–2.87)
ESR, mm/h30 ± 4
CRP, mg/L35.4 ± 5
Rheumatoid factor, IU/L266 ± 109
Antinuclear antibodies, n12/50 (24%)
Corticosteroids, n33/50 (66%)
Disease modifying drugs, n
Hydroxychloroquine4
Methotrexate15
Tiopronine4
Salazopyrine9
Aurothiomalate5
Cyclosporine2
Azathioprine1

Definition of abbreviations: CRP = C-reactive protein; ESR = erythrocyte sedimentation rate; HAQ = Health Assessment Questionnaire.

Pulmonary Symptoms

Thirty-five patients (70%), of whom 31 were lifelong nonsmokers (p = 0.054), complained of the following respiratory symptoms: dyspnea (n = 31, including two smokers; 20%), and chronic cough (n = 15, including two smokers; 30%), which was associated with chronic purulent sputum in nine nonsmokers and two smokers (22%). Bronchial suppuration preceded articular manifestations in four of these 11 patients (age range: 1 to 37 yr), whereas it appeared after joint involvement in seven patients (age range: 5 to 20 yr). Repeated episodes of acute bronchitis (⩾ 2/yr) were found in six patients (12%), including one smoker and five nonsmokers. No patient complained of wheezing. Pulmonary examination revealed crackles in 10 patients and wheezing in two nonsmokers.

Pulmonary Function Tests

Our study subjects had mean values of lung volume, air flow, and Tl CO that were within the normal range (Table 2). There was no significant difference between current or ex-smokers and nonsmokers in spirometric results, Tl CO, or SBN2 results (Table 2).

Table 2. RESULTS OF PULMONARY FUNCTION TESTING

Overall Group (n = 50)Nonsmokers (n = 40)Current or Ex-smokers (n = 10)
SVC, %pred99.5 ± 2.5100.3 ± 2.896.8 ± 5.9
(66–144)(66–144)(73–131)
FVC, %pred103.8 ± 2.8104.5 ± 3.2101.4 ± 5.7
(60.5–155)(61–155)(63–139)
FEV1, %pred98.7 ± 3.399.9 ± 3.894.1 ± 6.3
(59–139)(39–169)(63–139)
FEV1/SVC, %pred98.8 ± 2100 ± 2.493.9 ± 3.9
(46–128)(46–128)(70–108)
FEV1/FVC, %pred94.9 ± 1.495.2 ± 1.693.4 ± 3.1
(64–109)(64–109)(75–102)
TLC, %pred103.6 ± 2.3105.3 ± 2.496.8 ± 6.2
(72–141)(72–141)(72–133)
FRC, %pred106.2 ± 3.9107.7 ± 4.299.9 ± 9.5
(54–182)(54–182)(55–147)
RV, %pred110 ± 4.9113.5 ± 5.795.9 ± 8.1
(64–211)(64–211)(71–137)
FEF25–75, %pred77.6 ± 579.7 ± 5.869.1 ± 8.8
(15–186)(15–186)(34–128)
Tl CO, %pred89.5 ± 2.990 ± 3.487 ± 5
(44–136)(44–136)(60–105)
Tl CO/Va, %pred105.2 ± 3.3103.5 ± 398.1 ± 6.9
(58–145)(66–145)(58–126)
N2/L, %pred137 ± 14145 ± 18107.5 ± 9
(28–479)(28–479)(67–158)

Note: Values are mean ± SEM (range). All comparisons between current or ex-smokers and nonsmokers were not statistically significant.

Airway obstruction was demonstrated in nine subjects (18%) through a reduced FEV1/VC ratio, and SAD was detected in four patients (8%) through a decreased FEF25–75 (n = 2), an increased phase III slope (n = 1), and a combination of both in the remaining case.

The proportion of patients with an obstructive pattern was similar among smokers and nonsmokers (p = 0.74). A mild restrictive defect was found in four subjects (8%), with a TLC between 72% and 78% predicted. The number and proportion of patients with clinically relevant abnormal results in each individual test is shown in Table 3.

Table 3. PATIENTS WITH ABNORMAL PULMONARY FUNCTION TEST RESULTS

Decreased ValuesIncreased Values
SVC3NS
FVC4NS
FEV1 5NS
FEV1/VC9NS
FEF25–75 12NS
TLC42
RV212
RV/TLC18
Tl CO 10NS
Tl CO/Va 5NS
N2/LNS8

Note: Each number corresponds to the number of patients with a PFT value more than 1.64 standard deviation (2 SD for N2/L) below (“decreased”) or above (“increased”) the predicted estimate for a normal population; NS refers to abnormal results without clinical significance.

HRCT Findings

HRCT scans were normal in 15 patients (30%), whereas abnormal CT findings were made in 35 patients (70%) (Table 4). CT features suggestive of airway disease (Figures 1 and 2) were observed with the following frequency: (1) air trapping was identified in 16 patients (32%); it was diffusely distributed in nine cases and observed exclusively in the lower lung zones in seven cases; the extent of air trapping was always smaller than a subsegment; (2) cylindral bronchiectasis was identified in 15 patients (30%) at the level of the upper (n = 10), mid- (n = 11), and/or lower (n = 15) lobes, involving proximal (n = 13) and/or peripheral (n = 15) bronchi; the bronchiectasis was associated with abnormal bronchial-wall thickening in five patients; in every patient, cylindral bronchiectases were bilateral; (3) mild heterogeneity in lung attenuation was seen in 10 patients (20%), with a peripheral (n = 10) and bilateral (n = 9) distribution at the level of the upper (n = 2), mid- (n = 2), and/or lower (n = 8) lung zones; (4) centrilobular nodular and branching areas of high attenuation (n = 3; 6%) were observed with a bilateral distribution (n = 3) in the upper (n = 2), mid- (n = 2), and/or lower (n = 1) lung zones. Among the 29 patients with CT evidence of airway disease, 11 patients exhibited two signs and five had a combination of three signs. There was no relationship between the presence of these CT signs and smoking history (p = 0.78). CT features suggestive of rheumatoid lung lesions were observed as follows: (1) subpleural micronodules were seen in 14 patients (28%), with a predominant or exclusive distribution in the upper lung zones (n = 13) and a unilateral (n = 7) or bilateral (n = 7) distribution; (2) sharply demarcated, rounded opacities were observed in three patients (6%), with a peripheral distribution in the mid- (n = 2) and lower (n = 1) lung zones, and were unilateral in two cases. Three additional CT abnormalities were identified: (1) nonseptal linear opacities, observed in 15 patients (30%) at the level of the peripheral (n = 10) and/or central (n = 5) portions of the lung parenchyma, with a unilateral (n = 4) or bilateral (n = 11) distribution; these opacities were identified in the upper (n = 1), mid- (n = 5), and/or lower (n = 10) lung zones; (2) focal areas of consolidation were found in five patients (10%) at the level of the mid- (n = 3) and/or lower (n = 3) lung zones, with a random distribution in all cases; (3) nonbullous emphysema, involving less than 25% of the lung surface, was identified in two patients (4%).

Table 4. HIGH-RESOLUTION COMPUTED TOMOGRAPHIC FINDINGS

HRCT Abnormalitiesn (%)
HRCT features of airway disease
Air trapping16 (32)
Cylindral bronchiectasis15 (30)
Mild heterogeneity in lung attenuation10 (20)
Bronchial-wall thickening5 (10)
Centrilobular areas of high attenuation3 (6)
HRCT features of rheumatoid nodules
Subpleural micronodules14 (28)
Sharply demarcated, rounded opacities3 (6)
Miscellaneous CT findings
Nonseptal linear opacities15 (30)
Consolidation5 (10)
Nonbullous emphysema2 (4)

Relationship Between PFT and CT Findings

Among the 15 patients with normal HRCT scans, PFTs were normal in 13, whereas a mild obstructive pattern was found in two patients (FEV1/VC 70% and 79% predicted, respectively). In the group of patients with normal PFTs (n = 33), HRCT scans were normal in 13 patients (39%) and showed various combinations of abnormalities in the remaining 20 patients (Table 5). Among these 20 patients, 15 patients showed at least one CT feature of airway disease, (i.e., air trapping, bronchiectasis, mild heterogeneity in lung attenuation, bronchial- wall thickening). In the subgroup of 13 patients with PFT evidence of airways disease, the most frequent CT findings consisted of bronchiectasis (n = 8, of which six showed a diffuse distribution), air trapping (n = 5), and centrilobular nodular and branching areas of high attenuation (n = 3). Patients with HRCT evidence of airways involvement (i.e., bronchiectasis, air trapping, and/or centrilobular nodular and branching areas of high attenuation) (n = 24) were characterized by a significantly reduced FEF25–75 and Dl CO, whereas ΔN2/L was markedly higher in this subgroup (Table 6). Functional results did not significantly differ between patients with and without air trapping.

Table 5. RELATIONSHIPS BETWEEN INDIVIDUAL HIGH-RESOLUTION COMPUTED TOMOGRAPHIC FINDINGS AND PULMONARY FUNCTION TESTS

HRCT FindingsPulmonary Function Tests
Normal (n = 33)Obstructive (n = 13)Restrictive (n = 4)
Normal HRCT scan1320
HRCT features of airway disease
Air trapping852
Mild heterogeneity in lung attenuation631
Bronchiectasis681
Centrilobular areas of high attenuation030
Bronchial-wall thickening230
RA lung nodules
Subpleural micronodules662
Sharply demarcated, rounded opacities210
Miscellaneous CT abnormalities
Nonseptal linear opacities752
Consolidation230
Nonbullous emphysema110

Definition of abbreviations: HRCT = high-resolution computed tomography; RA = rheumatoid arthritis.

Table 6. RESULTS OF PULMONARY FUNCTION TESTS ACCORDING TO THE PRESENCE OR ABSENCE OF HRCT FEATURES OF AIRWAY INVOLVEMENT

No bronchial Involvement on HRCT (n = 26)Bronchial Involvement on HRCT*(n = 24)p Value
SVC102.6 ± 3.5 96.1 ± 3.5NS
FVC106.1 ± 3.4101.4 ± 4.4NS
FEV1, %pred102.8 ± 3.8 94.3 ± 5.4NS
FEV1/FVC, %pred   97 ± 1.7   92 ± 2.30.07
FEV1/SVC, %pred100.4 ± 2.4   97 ± 3.5NS
TLC, %pred104.9 ± 3.2102.2 ± 3.3NS
FRC, %pred106 ± 5106.2 ± 6.2NS
RV, %pred108.5 ± 6.4111.6 ± 7.6NS
RV/TLC, %pred100.8 ± 4.4107.4 ± 4.7NS
FEF25–75, %pred 84.3 ± 5.6 70.2 ± 8.30.04
Tl CO, %pred96.7 ± 3 81.9 ± 4.60.01
Tl CO/Va, %pred105.4 ± 3.6 98.7 ± 4.2NS
N2/L, %pred  98.8 ± 10.4   175 ± 24.50.007

* Note: Bronchial involvement refers to the presence of bronchiectasis, and/or air trapping, and/or centrilobular nodular and branching areas of high attenuation. NS = nonsignificant.

Relationship Between Clinical Data and PFT or HRCT Findings

Evidence of airway obstruction or SAD was correlated with the presence of repeated bronchial infections (p = 0.01), but was poorly related to chronic cough and sputum (p = 0.08 and p = 0.06, respectively). Patients with functional signs of airways disease were significantly older than those without airways manifestations (p = 0.009), whereas other clinical and laboratory features of RA in patients with and without airways involvement on PFT were comparable. Two patients among 15 with HRCT identification of diffuse bronchiectasis did not complain of chronic sputum and/or repeated episodes of acute bronchitis. Conversely, only two patients with these symptoms had a normal HRCT scan.

No correlation was found between RA data and PFT variables, including tests of small airways function.

The presence of secondary Sjögren's syndrome was unrelated to bronchial symptoms or to PFT or HRCT findings. RA disease-activity scores were similar in patients with and without HRCT evidence of airways involvement (HAQ: 1.8 ± 0.13 versus 1.5 ± 0.16; p = 0.15).

Because of the possible influence of antirheumatic drugs on the lung, we analyzed the relationship between abnormal PFT or HRCT scan results and past or present treatment with gold, penicillamine, tiopronin, and/or methotrexate. There were no significant differences in pulmonary function or HRCT data in patients given different treatments. However, the complexity of previous drug histories precluded a meaningful evaluation of the relationship between each single drug and pulmonary involvement.

Since an obstructive pattern was found in 18% of patients in the study, and HRCT features of airway disease were found in 34% of cases, the study suggests a high prevalence of airways involvement by RA in the absence of interstitial lung disease. With regard to functional impairment, our findings are in agreement with the 16% frequency of airway obstruction reported in a population of 81 unselected RA patients (versus 0% in matched controls) (18). However, it is noticeable that airway obstruction was observed in 24% and 41% of RA patients, respectively, in two surveys (3, 4). On the basis of a significantly decreased FEF25–75 value and/or a markedly increased ΔN2/L, we identified SAD in 8% of our population. FEF25–75 is a well-recognized test of small airways involvement, and single-breath nitrogen washout has been shown to be a highly sensitive test for SAD in smokers, preceding spirometric deterioration (19). The slope of phase III is significantly correlated with inflammation score for the small airways in smokers (20). The criteria used in the present study to assess SAD may account for the apparent discrepancies with previous reports. As defined by FEF25–75 values below 80% predicted, SAD was observed in 65% of patients by Collins (3), but this threshold is likely to produce overestimation of the prevalence of SAD. Defining abnormal values as below 95% of the predicted one-tailed confidence limits, Geddes found a significant decrease in FEF25–75 in 38% of RA patients (4). In contrast to Geddes and Mountz (4, 9), we found no significant difference in the proportions of airflow obstruction and SAD among smokers and nonsmokers, suggesting a minor role for tobacco smoke in such manifestations.

On HRCT scans, we observed a low frequency of direct signs of small airways disease (i.e., centrilobular nodular and branching areas of high attenuation [6%]), whereas indirect signs were identified with a frequency of from 20% to 32%. The identification of centrilobular abnormalities has previously been reported as a CT feature of follicular and constrictive bronchiolitis (17), which are considered the histologic hallmarks of bronchiolar disease in RA (21, 22).

Hayakawa reported nodular and centrilobular branching structures in nine of 15 patients with RA and biopsy-proven bronchiolar disease (22). The frequency of centrilobular abnormalities observed in the present study is similar to the 6% (six of 77 patients) previously reported (7). Owing to the ability of a recent CT technique (i.e., STS–MIP) to help detect mild forms of micronodular infiltration, one would await the results of further investigations to evaluate the accuracy of CT in detecting direct signs of bronchiolar disease in RA (16). Among the indirect signs of SAD in our study, the most frequently encountered CT finding was air trapping, observed in 16 patients (32%). The second most frequent CT abnormality consisted of cylindral bronchiectasis, identified in 15 patients (30%) in the absence of interstitial lung disease. The frequency of this finding is comparable to the frequency of cylindral bronchiectasis previously reported in RA, which varied from 20% to 25% of the study group (6, 7, 23, 24). In accord with the findings of Remy-Jardin and colleagues (16) and Lena and coworkers (25), bronchiectasis was observed in our study with a predominant or exclusive involvement of the lower half of the bronchial tree (right middle lobe/lingula and lower lobes), a distribution similar to that observed for each of the CT features of SAD evaluated in the present study. It is notable that 85% of our patients with HRCT evidence of bronchiectasis complained of chronic bronchial suppuration. The temporal relationship between bronchiectasis and RA is still debated. Prior evidence of bronchial infection and/or chronic sputum has led to the hypothesis that chronic bacterial infection may play an etiologic role in triggering immune reactions leading to articular involvement (5, 26-28). A parallel evolution of pulmonary exacerbations and flares of articular disease in some patients (29) is an additional argument for an immunologic relationship between RA and respiratory infection. By contrast, it has been suggested that several factors related to RA are likely to increase the incidence of respiratory infection (i.e., secondary Sjögren's syndrome, corticosteroids, and disease-modifying drugs), thus accounting for the delayed occurrence of bronchiectasis in RA (30, 31). Another possibility is that RA and bronchiectasis share a common genetic predisposing factor. This hypothesis is supported by the recent demonstration of a higher frequency of DQB1*0601, DQB*0201, and DQA*0501 human leukocyte antigen (HLA) genotypes in patients with RA and bronchiectasis (24). The long-term clinical significance of these airways changes is an important issue.

Swinson and colleagues recently demonstrated a poor 5-yr survival in patients with coexisting bronchiectasis and RA as compared with patients matched for age, sex, and disease duration who had RA alone (32). The death rate was 5.0 times greater in the former group, the leading cause of death being pulmonary infection. In the absence of acute respiratory symptoms, the presence of mild heterogeneity in lung attenuation on HRCT scans in 10 patients (20%) was interpreted as a mild form of mosaic pattern, a frequent indirect sign of SAD on HRCT scans. Until now, this CT feature has been evaluated only by Hayakawa and associates, who observed patchy low-attenuation areas in five of 15 patients with documented obliterative bronchiolitis (33%) (22).

Our study design allowed us to correlate PFT results and HRCT findings. Among the 15 patients with normal CT scans, PFTs depicted a moderate obstructive pattern in two cases. This apparent discordance can be explained by the well-known limitations of CT in depicting mild changes in bronchial-wall thickness and/or diameter. However, it is noticeable that HRCT depicted features of SAD in 15 of the 33 patients with normal PFTs, or 30% of the study population. In this subgroup of patients, the most frequent abnormalities consisted of air trapping, mild heterogeneity in lung attenuation, and bronchiectasis. These results suggest that CT is a more sensitive method for detecting small-airways abnormalities than are PFTs. Among the 13 patients with an obstructive pattern, CT features of SAD were observed in 10 patients, and consisted mainly of bronchiectasis and air trapping. It is noticeable that all patients with centrilobular areas of high attenuation, suggestive of RA bronchiolitis, presented with an obstructive pattern. However, and in contrast to Hansell and colleagues' recently reported findings (33), we failed to demonstrate any correlation between SAD and the presence of air trapping.

The design of the present study comprised a few limitations. Recruitement of patients from a university hospital rheumatology department could introduce some bias through selection of patients with somewhat more severe articular involvement than that in the overall RA population. However, all patients included in the present study were ambulatory at the time of pulmonary assessment, and disease activity was mild to moderate in a significant proportion of patients. Moreover, no relationship could be demonstrated between RA disease parameters and PFT or HRCT findings.

The HRCTs were not read blindly with respect to the diagnosis of RA, but the two readers had no information about the clinical or functional pulmonary status of the patient. Second, an independent reading, rather than a reading by consensus, could have been indicated, since several abnormalities were subtle and thus prone to over- or underestimation. However, the majority of patients with CT features of airways disease exhibited more than one sign of such involvement.

Risk factors for airways involvement in RA are still poorly defined. In the present study, we failed to demonstrate any significant relationship between airways involvement on PFTs and/or HRCT scans and clinical or immunologic features of RA. These findings contrast with those of Vergnenègre and associates (18), who recently reported a significant relationship between FEF25–75 and duration of articular disease, and between FEV1/FVC and the Ritchie index. However, the level of these correlations was low, and these findings might have been underestimated in our smaller study group. In agreement with the finding in previous studies (5, 28, 32), we observed no correlation of the presence of bronchiectasis with the severity of RA disease. With regard to the pathophysiologic substratum of bronchiectasis in RA, and in agreement with the findings of Vergnenègre and coworkers (18), we failed to observe any relationship between the frequency of secondary Sjögren's syndrome and the presence of bronchiectasis, which had previously been pointed out by several authors (8, 34, 35).

Conclusion

In conclusion, the present data support and extend the concept of a high incidence of airways involvement in RA patients. Evidence of airways involvement on PFTs and/or HRCT scans was correlated with respiratory symptoms. However, smoking history, Sjögren's syndrome, and RA disease activity did not appear as significant predictors of airways involvement. Although correlations between PFTs and HRCT findings were highly significant, HRCT appeared to be a more sensitive means of detecting SAD than were PFTs. In the clinical setting of RA with chronic sputum production and/or repeated episodes of acute bronchitis, these noninvasive techniques appear to be complementary for a comprehensive assessment of airways changes. The precise characterization of airways involvement allowed by these two techniques may help in evaluating the temporal and immunologic relationships between airways disease and RA, as well as the long-term significance of airways disease as a manifestation of RA.

1. Frank S. I., Weg J. G., Harkleroas L. E., Fitch R. F.Pulmonary dysfunction in rheumatoid arthritis. Chest6319732734
2. Perez T., Farre J. M., Gosset P., Wallaert B., Duquesnoy B., Voisin C., Delcambre B., Tonnel A. B.Subclinical alveolar inflammation in rheumatoid arthritis: superoxide anion, neutrophil chemotactic activity and fibronectin generation by alveolar macrophages. Eur. Respir. J.21989713
3. Collins R. L., Turner R. A., Johnson A. M., Whitley N. O., McLean R. L.Obstructive pulmonary disease in rheumatoid arthritis. Arthritis Rheum191976623628
4. Geddes D. M., Webley M., Emerson P. A.Airways obstruction in rheumatoid arthritis. Ann. Rheum. Dis.381979222225
5. Walker W. C.Pulmonary infection and rheumatoid arthritis. Q. J. Med.1421967239251
6. Donagh J., Greaves M., Wright A. R.High-resolution computed tomography of the lungs in patients with rheumatoid arthritis and interstitial lung disease. Br. J. Rheumatol331994118122
7. Remy-Jardin M., Remy J., Cortet B., Mauri F., Delcambre B.Lung changes in rheumatoid arthritis: CT findings. Radiology1931994375382
8. Radoux V., Menard H. H., Begin R.Airway disease in rheumatoid arthritis patients: one element of a general exocrine dysfunction. Arthritis Rheum301987249252
9. Mountz J. D., Turner R. A., Collins R. I., Gallup K. R., Semble E. I.Rheumatoid arthritis and small airways function: effects of disease activity, smoking, and alpha-1-antitrypsin deficiency. Arthritis Rheum271984728736
10. Banks J., Banks C., Umachandran V., Smith A. P., Jessop J. D., Pritchard M. H.An epidemiological and clinical investigation of pulmonary function and respiratory symptoms in patients with rheumatoid arthritis. Q. J. Med.851992795806
11. Arnett F. C., Edworthy S. M., Bloch D. A., Shane D. J., Fries J. F., Cooper N. S., Healy L. A., Kaplan S. R., Liang M. H., Luthra H. S., Medsger T. A., Mitchell D. M., Neustadt D. H., Pinals R. S., Schaller J. G., Sharp J. T., Wilder R. L., Hunder G. G.The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum311988315324
12. Guillemin F., Briançon S., Pourel J.Mesure de la capacité fonctionnelle dans la polyarthrite rhumatoide: adaptation Française du Health Assessement Questionnaire. Rev. Rhum. Mal. Ostéoartic581991459465
13. Vitali C., Bombardieri S., Moutsopoulos H. M.Preliminary criteria for the classification of Sjögren's syndrome: results of a prospective concerted action supported by the European Community. Arthritis Rheum.361993340347
14. Buist A. S., Ross B. B.Quantitative analysis of the alveolar plateau in the diagnosis of early airway obstruction. Am. Rev. Respir. Dis.108197310781087
15. European Respiratory Society. 1993. Report of the working party: standardized lung function testing. Eur. Respir. J. 6(Suppl. 16):5–52.
16. Remy-Jardin M., Remy J., Artaud D., Deschildre F., Duhamel A.Diffuse infiltrative lung disease: clinical value of sliding-thin-slab maximum intensity projection CT scans in the detection of mild micronodular patterns. Radiology2001996333339
17. Muller N. L., Miller R. R.Diseases of the bronchioles: CT and histopathologic findings. Radiology1961995312
18. Vergnenègre A., Pugnère N., Antonini M. T., Arnaud M., Melloni B., Treves R., Bonnaud F.Airway obstruction and rheumatoid arthritis. Eur. Respir. J.10199710721078
19. Cosio M., Ghezzo H., Hogg J. C.The relations between structural changes in small airways and pulmonary function tests. N. Engl. J. Med.298197812771281
20. Berend N., Wright J. L., Thurlbeck W. M., Marlin G. E., Woolcock A. J.Small airways disease: reproducibility of measurements and correlation with lung function. Chest791981263268
21. Colby, T. V. 1994. Bronchiolar pathology. In G. R. Epler, editor. Diseases of the Bronchioles. Raven Press, New York. 77–100.
22. Hayakawa H., Sato A., Imokawa S., Toyoshima M., Chida K., Iwata M.Bronchiolar disease in rheumatoid arthritis. Am. J. Respir. Crit. Care. Med.154199615311536
23. Hassan W. U., Keaney N. P., Holland C. D., Kelly C. A.High resolution computed tomography of the lung in lifelong non-smoking patients with rheumatoid arthritis. Ann. Rheum. Dis.541995308310
24. Hillarby M. C., Mahon M. J., Grennan D. M.HLA association in subjects with rheumatoid arthritis and bronchiectasis but not with other pulmonary complications of rheumatoid disease. Br. J. Rheumatol321993794797
25. Léna H., Desrues B., Le Coz A., Belleguic C., Quinquenel M. L., Kernec J., Chales G., Delaval P.Dilatation des bronches et polyarthrite rhumatoı̈de: une association méconnue. Rev. Mal. Respir.1419973743
26. Walker W. C., Wright V.Pulmonary lesions and rheumatoid arthritis. Medicine471968501520
27. Solanski T., Neville E.Bronchiectasis and rheumatoid disease: is there an association? Br. J. Rheumatol311992691693
28. Mc Mahon, M. J., D. R. Swinson, R. W. Wolstenholme, C. Chattopadhyay, P. Smith, P. Johns, and N. H. CrosbyBronchiectasis and rheumatoid arthritis: a clinical study. Ann. Rheum. Dis.521993776779
29. Bach F.Penicillin for the rheumatoid type of arthritis. Ann. Rheum. Dis.719482528
30. Bamji A., Cooke N., Rheumatoidarthritis and chronic bronchial suppurationScand. J. Rheumatol.1419851521
31. Schadick N. A., Fanta C. H., Weinblatt M. E., O'Donnel W., Coblyn J. S.Bronchiectasis: a late feature of severe rheumatoid arthritis. Medicine731994161170
32. Swinson D. R., Symmons D., Suresh U., Jones M., Booth J.Decreased survival in patients with co-existent rheumatoid arthritis and bronchiectasis. Br. J. Rheumatol361997689691
33. Hansell D. M., Rubens M. B., Padley S. P. G., Wells A. U.Obliterative bronchiolitis: individual CT signs of small airways disease and functional correlation. Radiology2031997721726
34. Andonopoulos A. P., Constantopoulos S. H., Drosos A. A., Moutsopoulos H. M.Pulmonary function of non smoking patients with rheumatoid arthritis in the presence and absence of secondary Sjögren's syndrome: a controlled study. Respiration531988251258
35. Begin R., Masse S., Cantin A.Airway disease in a subset of nonsmoking rheumatoid patients: characterization of the disease and evidence for an autoimmune pathogenesis. Am. J. Med.721982743
Correspondence and requests for reprints should be addressed to Thierry Perez, M.D., Hôpital Albert Calmette- Boulevard Jules Leclerc, 59037 Lille Cedex, France. E-mail:

Related

No related items
American Journal of Respiratory and Critical Care Medicine
157
5

Click to see any corrections or updates and to confirm this is the authentic version of record