Eur Respir J 1989, 2, 7-1 3

Subclinical alveolar inflammation in rheumatoid arthritis: superoxide anion, neutrophil chemotactic activity and fibronectin

generation by alveolar macrophages

Th. Perez*, J.M. Farre**, Ph. Gosset*, B. Wallaert*, B. Duquesnoy ... C. Voisin*, B. Delcambre**, A.B. Tonnel*.

Subclinical alveolar inflammation in rheumaJoid arthritis: superoxide anion, neutrophil chemotactic activity and fibronectin generation by alveolar macro­phages. Th. Perez, J.M. Farre, Ph. Cosset, B. Wallaert, B. Duquesnoy C. Voisin, B. Delcambre, A.B. Tonne/. ABSTRACT: Interstitial lung disease (D...D) can be detected by pulmonary function testing (PFI') in 30-40% of rheumatoid arthritis (RA) patients. We assessed by bronchoalveolar lavage (BAL) the patterns of alveolitis in 21 RA patients: group 1 comprised 12 patients without evidence of lLD, and group 29 patients with clinical lLD defined by abnormal pulmonary function tests and/or chest X-ray. Cellular characteristics of BAL were stud­ied In both groups. In addition, alveolar macrophages (AM) from patients in group 1 were Isolated, and three parameters of cellular activation were studied: superoxlde anion, flbronectin and neutrophil chemotactic activity generation. Total cell counts were not increased In group 1 but significantly Increased In group 2 compared to controls. In group 1, 5/12 patients had elevated lymphocyte percentage (>18%) suggesting subclinical lymphocyte alveolitis. In contrast, neutrophil alveolltis (>4%) was found In 7/9 patients in group 2, mean percentage 12.9±4.2, compared with 1.2±6.4% In con­trols and 1.9±0.5% In group 1. These changes were not correlated with disease duration nor rheumatoid factor titres. Marked elevation of lympho­cyte percentage was observed In patients with abnormal serum beta-2- mlcr oglobuUn. Alveolar macrophages from group 1 patients released Increased amounts of superoxlde anion (7260±2700 11s controls 850±120 URL/S.tOS cells), neutrophil chemotactic activity (21±4.8 vs controls 8.1±0.7 cells!HPI<'), and fibronectln (6.1±1.6 vs controls 1.3±0.2 ng·106 cells/hour). Whether or not lympbocyte alveolitis and/or AM dysfunction are patho­genic mechanisms of subsequent Interstitial lung disease In patients who are still free of symptoms remains to be determined. Eur Respir J., 1989, 2, 7- 13.

• Departement de Pneumologie, Hopilal A. Calmette et Institut Pasteur, Lille, France. .,. Clinique Rhumatologique, Centre Andre Verhaeghe, Hopital de la Charite, Iille, France.

Correspondence: Dr. Th. Perez, Departement de Pneumologie, Hopital A. Calmcue et Institut Pas­teur, Lille, France.

Supported in pan by Univcrsitc de Lille, INSERM (reseau de recherche clinique, partici· pation des cellules inflammatoires en pathologie respiratoire n° 850027), and by Fonds de Recher­che de I' Association Fran~aise de la Lutte contre la Rhumatisme.

Received: January 1988; Accepted after revision August 15, 1988

Keywords: Alveolar macrophage; bronchoalveo­lar lavage; fibronectin; lung immunology; neutro­phil chemotactic activity.

Pulmonary involvement is a major extra-articular manifestation of rheumatoid arthritis (RA). Interstitial lung disease (lLD) was reported in less than 5% of RA patients in studies using radiographic evaluation [1]. However, there is evidence that chronic infiltrative lung disease may be present in patients with normal chest roentgenograms [2]. More recently pulmonary function tests studies detected abnormalities in 30 to 40% of RA patients [3, 4). Moreover, combined use of diffusing capacity, X-ray and lung biopsy showed disturbances in up to 80% of patients [5). Thus subclinical lLD as a frequent feature of RA requires precise tools for diagnosis, pathophysiologic approach and prognostic evaluation.

fibrosis [6, 7). Bronchoalveolar lavage (BAL) has emerged as a useful investigative method for evaluation of alveolitis [8, 9]. In addition, a good correlation has been shown between BAL disturbances and histologi­cal findings for parenchyma! inflammation assessment [10]. In this context, a subclinical alveolitis was dem­onstrated by BAL in 48% of patients with various col­lagen vascular diseases [11] and recent studies described BAL differential cell count abnormalities in RA patients without evidence of lLD [12, 13].

Alveolitis is usually recognized as the first stage of chronic interstitial lung disorders, preceding irreversible

With this in mind, we initiated this study to evalu­ate by BAL the cellular characteristics of alveolitis in RA patients without evidence of ILD and compared our results with those obtained from healthy controls and from RA patients with ILD.

The occurrence of pulmonary fibrosis requires a

8 TH. PEREZ ET AL.

combination of parenchyma! injury, recruitment of in­flammatory cells, and fibroblast proliferation. In this re­gard, AM are considered to play a key role in different steps of alveolitis by producing a variety of inflammatory and cell recruiting mediators. Conse­quently, free radical generation, neutrophil chemotactic activity (NCA) and fibronectin production by AM from symptom free patients (group 1) were evaluated exclud­ing corticosteroid and/or immunosuppressive treated pa­tients with ll..D (group 2).

Subjects and Methods

Patients

Twenty-one patients with classical or definite RA [14) were prospectively studied. They comprised 3 males and 18 females with a mean age of 55 yrs (range 29- 71). Pulmonary symptoms, clinical signs, and medication history were determined in all patients. All patients were non smokers and none had a significant environ­mental exposure to toxic gases or dusts, or previous pulmonary disease unrelated to RA.

The control group (18 subjects) included hospital staff and healthy volunteers. All were non smokers.

Methods

Informed consent was obtained from all patients. Mean disease duration was 10 yrs (range 1-23). Screen­ing of secondary Sj6gren syndrome was systematically performed based on clinical criteria (xerophthalmia and/ or xerostomia, positive Schirmer test) and minor sali­vary gland biopsies. Rheumatoid factor was assessed by nephelometry (Hyland disc 120, Hyland Lab). Beta-2-microglobulin was radioimmunoassayed using a com­mercial test (Phadebas Pharmacia, Upsala, Sweden). R~ suits were considered abnormal if over 2.5 mg·l·1•

Definition of patient groups

Chest roentgenograms. Posteroanterior and lateral roent­genograms were evaluated by a radiologist unaware of clinical and BAL data. Pulmonary function tests (PFT). Forced vital capacity (FVC), forced expiratory volume in one second (FEV

1)

and total lung capacity (TLC) were determined using a Jaeger spirometer. Carbon monoxide diffusing capac­ity (DLco) was obtained by a single breath method and corrected for alveolar volume and haemoglobin. The predicted values for each subject based on sex, age and height were obtained from standard tables [15]. Data were expressed as a percentage of the predicted values. Pulmonary function tests were considered abnormal if at least two of the following parameters were reduced: FEV

1, TLC, FVC<80% predicted, and/or DLco<75%

predicted. According to clinical data, radiological and PFT find-

ings, patients with RA were divided in 2 groups: group 1 included 12 patients without evidence of interstitial lung disease; group 2 included 9 patients with clinical interstitial lung disease defined as abnormal X-ray consistent with an interstitial disorder and/or a restric­tive pattern of PFT. Clinical and biological data in both groups are shown in table 1.

Table 1. - Clinical biological and pulmonary function data in group 1 and group 2 patients. Results are expressed as mean±seM.

Group units

Number Age yrs Disease duration yrs Sjogren's syndrome Dyspnoea ESR Beta-2-micro­globulin FVC• TLC* TLco•

mm·h

1 2

12 9 53.9±4.7 60.9±3 11.7±2.1 10.5±2.6

4/12 2/9 2/12 6/9 49.iti 39±7

3.3±0.3(n=l0) 2.5±0.2 (n=6) 97±2.6 59.5±4.7

lll±72 69.8±5.1 107±6.8 66.4±85

ESR: erythrocyte sedimentation rate; PVC: forced vital capac­ity; TLC: total lung capacity; TLco: diffusing capacity for carbon monoxide; t; mean-B2-mi(.Toglobulin is not significantly different betwen the two groups and between each group and the controls. *; 1 versus 2: p<0.05. FVC, TLC and TLco are expressed as % of predicted value.

Bronchoalveolar lavage

Bronchoalveolar lavage was performed after premedi­cation with atropine under local anaesthesia with lignocaine using a wedged fibreoptic bronchoscope (Olympus model BF-B3; Olympus Corps of America, New Hyde Park, N.Y.). 250 ml of sterile saline solu­tion were instilled in five 50 ml aliquots with immedi­ate gentle vacuum aspiration after each aliquot, as previously described [11). Total number and differen­tial cell counts were determined and the cellular vari­ability of alveolar macrophages was assessed by trypan blue exclusion.

Alveolar macrophage isolation

The fluid was filtered through several layers of ster­ile surgical gauze and centrifuged at 400 g for 10 min at 4°C. After three washings the pellet was resuspended at a cell concentration of l.Sxl06 ml-1 in Hank's balanced salt solution (HBSS). 2 ml of the cell suspen­sion were cultured at 37°C in humidified air with 5% C0

2 in 35 mm diameter Petri dishes. The cell popula­

tion consisted of at least 95% alveolar macrophages after a 2 h adherence phase; basophil cells and mas­tocytes were absent from the culture as previously demonstrated [16).

SUBCLINICAL ALVEOLJTIS IN RHEUMATOID ARTHRITIS 9

Evaluation of alveolar macrophage function

To detennine whether alveolar macrophages from patients in group 1 without clinical interstitial lung disease were activated, three secretory products were evaluated: superoxide anion generation before and after stimulation by phorbol-myristate-acetate (PMA), release of neutrophil chemotactic activity for neutrophils (NCA) and of fibronectin.

The chemiluminescence of AM was investigated us­ing a lucigenin-dependent chemiluminescence method adapted from Wll.l.IAMS and Coi..E [17, 18]. Lucigenin (IO""M) (Sigma) was dissolved in HBSS buffered with 18 mM HEPES (N-2-hydroxyethylpiperazine-N-2-ethane-sulphonic-acid). PMA (Sigma) was dissolved in dimethylsulphoxide at a concentration of 0.5 mg-mJ·' . PMA (10~1) was dissolved just before use in 5 ml of HBSS-HEPES. Superoxide dismutase (SOD) (Sigma) was dissolved in HBSS-HEPES (120 ~g·ml·'). AM with HBSS were centrifuged at 800 g (10 min. 4°C) and resuspended in HBSS-HEPES to a concentration of 1x 1()6 viable AM per ml. The suspension was kept on ice in a siliconized glass container until use. Chemilu­minescence (CL) was measured at 37°C using a nucleo­timeter 107. AM suspension aliquots (500~ were added to vials containing lucigenin aliquots (900!1/) with or without PMA (100~ and/or SOD (100~. The total volume in each vial was brought to 1650 ~ by adding a 50~ aliquot from 3% gelatin solution and the appropriate amount of HBSS-HEPES. Intensity of luminescence was integrated for 60 sec after a 12 min incubation at 37°C. The results are expressed in rela­tive luminescent units (RLU) per 0.5x106 viable AM.

Alveolar macrophage derived chemotactic factor for neutrophils was measured as previously described [19). In brief, alveolar macrophages were cultured at 1.5x 106 cells per ml in serum-free RPMl-1640 at 37°C for three hours. The supemates were collected and assayed for neutrophil chemotactic activity (NCA) by counting the number of neutrophils that passed through a 3 ~ micropore filter (Nucleopore Corp., Pleasanton, catifor­nia) from the upper compartment of a 48-well micro­chemotaxis chamber (Neutro Probe, Cabin John, Md) in response to the macrophage supernate placed in the lower chamber. In order to distinguish chemokinesis from chemotaxis, experiments were performed in which supernatant samples were placed on both sides of the filter membrane. This procedure demonstrated no sig­nificant difference between the negative controls and the experimental wells and proved the existence of a chemotaxis mechanism. The number of neutrophils that had migrated through the filter was determined micro­scopically by the use of an oil immersion lens. Four fields were read per well. Experiments were conducted in groups of four. The results were expressed as the difference between the mean number of cells per field in the experimental well and the mean number of cells per field on the control well (migration towards medium). Positive control experiments consisted of migration towards FMLP 10·7M (Peninsula Laboratories Inc., San Carlos, California).

Alveolar macrophage fibronectin production was measured by competitive radioimmunoassay as previ­ously described [20]. AM were cultured in the same conditions and the fibronectin secreted in the medium was measured. The results were expressed as nanogram of fibronectin per 106 cells per hour.

Statistical analysis

Results are expressed as mean±sEM. Comparison be­tween groups used Mann-Whitney U-test since most data are non parametric. Correlations were analysed with the Speannan rank test using a 5% significance level.

Results

Bronchoalveolar lavage results

Total cell count: Total cell numbers were not signifi­cantly elevated in group 1 (12.7±2.0x10'·ml·1) compared with controls (10.2±1.1x 10'·ml·1). Group 2 had the high­est cellular concentration (20±4.7xlQ4 cell-m!·') signifi­cantly different from controls (p<0.05) but there was no statistical difference between group 1 and group 2. Cell Differential: Results are expressed as a percentage of (1) lymphocytes, (2) neutrophils and (3) eosinophils. ( 1) Lymphocytes. Interestingly I ymphocyte percentage was markedly elevated (> 18%) in 5 out of 12 patients without lLD (group 1) (range 22-43%). However, no significant difference could be seen between mean lym­phocyte percentage in group 1 {17.3±3.8%) and the control group (9±1.2%) (fig. 1). In contrast only 1 of the 9 group 2 patients exhibited an abnonnal lympho­cyte percentage (24%). (2) Neutrophils. Neutrophil count abnormalities were more prominent (fig. 1). In group 1. 3 out of 12 patients exhibited a slightly elevated percentage (range 5-6%). Mean proportion of neutrophils (2.2±0.6%) was not significantly different from controls (1.2±0.4%). An increased proportion of neutrophils was detected in group 2 {12.9±4.2%), significantly different from con­trols and from group 1 (p<O.Ol); 5 out of 9 patients had a marked elevation of neutrophil percentage {>10%). (3) Eosinophils. No difference appeared between patient groups and controls. Only one patient in group 2 had a slightly elevated percentage (3.5%).

Relationship between differential counts, clinical and biological data

Neutrophil and lymphocyte percentages did not differ significantly between patients with recent or long tenn disease duration. Surprisingly, secondary sicca syndrome was not associated with higher lym­phocyte counts. Similarly, scrum rheumatoid factor titres were unrelated to cell differential abnormalities. But

10 TH. PEREZ ET AL.

60

.. 50

"' >. <.> 0 40 .r::. Cl.

• E

;?:>

0 30 t • .. a

~ 20 : c:

~ A "' Q. 10

l I l w 0

Controls

• I • • ••

I I

I . • I I

Group 1 Group 2

50

45 • 40

.!! :E Cl.

35

e :; 30

"' c: 25 0 • ~ 20 ~

15 • c:

~ 10 : "' Q.

5 - t I - - I 0 Group 2 Controls Group 1

Fig. 1. - Lymphocyte and nemrophil perceruagcs from the two patients groups. A s•gniOcant elevauon of neuuophil proponion was demonstrated in group 2 compared with controls; group r and 2 per­centages were significantJy different. • indicates ()<0.01.

40

30

20

10

0 -1-..l..ao""" Superoxlde NCA

Fig. 2. - Spontu.neous supcroxide anion relative luminescent units <RLU), neu trophil chcmotac~ic activity (NCA) (cells/HPF) and t'ibroncctin (ng·IO' cells/hour) generation by alveolar macrophages (AM) from subgroups or:__l)31ients according tO BAL differential counts. !2'21 Normal BAL; t7.tl Abnormal BAL.

interestingly, patients with abnormal serum beta-2-microglobulin had significantly more lymphocytes in BAL (mean 21±3.7%) than did patients with nor­mal serum beta-2-microglobulin (7.8±1.4%, p<0.05). There was no correlation between clinical data and neutrophil percentage.

Table 2. - Alveolar macrophage functions in controls and group 1 patients

Secretory products units Control Group 1

Superoxide anion

- Spontaneous RLU 850±120 7260:!:2700** - PMA induced RLU 16320±1440 28700±6220

Neutrophil chemotactic activity

Fibronectin

cells/HPF 8.1±0.7

ng·106cells h-1 1.3±0.2

21±4.8*

6.1±1.6*

*Group 1 versus controls p<0.05; **Group 1 versus controls p<O.Ol. PMA: phorbol-myrisate-acetate, RLU: relative lumi­nescent writs.

Alveolar macrophage secretory products

Superoxide anion release: When studied at baseline, CL response of AM was significantly greater (7260±2700 RLU) compared with controls (850±120 RLU; p<O.Ol) (table 2). Mean SOD inhibition was 79% (range 57-96%).

In contrast, after stimulation with PMA there was only a mild increase in group 1 (28700±6220 RLU) compared with controls (16320±1440 RLU, p>0.05). SOD inhibited 83 to 97% of PMA induced CL. Neutrophil chemotactic activity: An increased NCA was detected in AM supematants from six group 1 patients (range 14-41 cells/HPF, mean 21±4.8) and a significant difference appeared with controls (8.1±0.7 cells/HPF, p<0.05) (table 2). No correlation was found between NCA and neutrophil proportion in BAL fluid, but low chemotaxis values ( < 10 cells/HPF) were associated with low polymorpho nuclear neutrophil (PMN) percentage (0-1%) in BAL fluid Fibronectin release: Compared to levels usually recov­ered from control alveolar macrophages 6 out of 9 patients released abnonnal amounts of fibronectin (>3 ng-106 cells/hour, range 6-13) (table 2) and mean pro­duction was significantly greater than in controls (6.1±1.6 vs 1.3±0.2 ng-106 cells/h-1).

Correlation between alveolar macrophage function. There were no correlations between spontaneous CL, NCA and fibronectin. A significant correlation was found between PMA induced CL and NCA (r=0.76 p<0.05).

Correlation between cell differentials and alveolar macrophage functions. There were no correlations between lymphocyte proportion and CL, NCA or fi­bronectin production. In contrast, a significant correla­tion was found between neutrophil percentage in BAL fluid and spontaneous CL (r=0.64 p<0.05). Group 1 was divided into a subgroup of patients who had nonnal BAL differential counts and a subgroup of those who had elevated neutrophil and/or Iymphocyte percentages (fig. 2).

Interestingly, AM from some patients with nonnal BAL differential counts released increased amounts of superoxide anion and/or NCA, and/or fibronectin.

SUBCLINICAL ALVEOLITIS IN RHEUMATOID ARTHRITIS 11

Conversely, nonnal AM secretory functions were only found in those patients with nonnal BAL differential counts.

Discussion

Our results demonstrate a high incidence of BAL abnonnalities in RA patients with and without clinical evidence of ll..D and allow defmition of two preferen­tial patterns of alveolitis according to the group studied.

An increased number and proportion of neutrophils in BAL suggesting a neutrophil alveolitis was promi­nent in 7 out of 9 patients with cJinical ll..D. Indeed neutrophil alveolitis with increased total cell concentra­tion is a classical feature of alveolitis in idiopathic pulmonary fibrosis (IPF) (6] and ll..D associated with systemic scJerosis [21).

In patients without clinjcal ll..D, abnonnalities in BAL content were less prominent, but 5 out of 12 patients had a pattern of lymphocyte alveolitis. Among them, only one had an associated secondary sicca syndrome. These results parallel histopathologic findings of lung biopsies. Beside rheumatoid nodules, bronchiolitis obl­iterans and usual interstitial pneumonia, frequent pri­mary patt~rns include lymphoid hyperplasia and mononuclear interstitial infiltrates [22).

Our data are in agreement with a recent study of GARCIA et al (12). Elevated percentages of lymphocytes were detected in 5 of the 15 symptom free patients with nonnal X-ray and PFr. In contrast, TISCHLER (23] demonstrated a lymphocyte alveolitis only in patients with mild X-ray interstitial changes.

Mechanisms responsible for lymphocyte alveolitis in RA remain unclear. Although lymphocyte alveolitis was reported to be an usual feature of primary Sjogren syndrome [11) abnonnal lymphocyte counts were not significantly associated with secondary sicca syndrome. Interestingly, lymphocyte alveolitis was related in our study to high serum levels of beta-2-microglobulin. An elevation of plasma, urinary and synovial beta-2-microglobulin is well known in RA and seems to be correlated with inflammatory parameters, circulating immune complex levels and severity of joint disease [24-26). Although non specific beta-2-microglobulin appears as a good index of lymphocyte total mass and of cell membrane turnover. Thus lymphocyte prolifera­tion in alveolar spaces might be related to the general­ized hyperreactivity of lymphoid tissue associated with RA [27] paralleling synovial disorders.

In the second part of our study, three parameters of alveolar macrophage activation were examined in patients without clinical ll..D.

Among a wide range of secretory activities [28, 29] hydrogen peroxide and superoxide anion production by AM are triggered by cell membrane stimulation. Lucigenin dependcm chemiluminescence correlates closely with oxygen radical release Ll7) and inhibition by SOD suggests a close relationship with superoxide anion generation. Our results demonstrate a prominent increase of spontaneous CL indicating oxidant species

production by BAL cells. A significant correlation appeared with PMN percentage suggesting a possible participation of PMN. Nevertheless, experimental con­tamination of AM suspension with similar amounts of PMN produced only a slight increase of lucigenin­dependent CL response, in contrast with luminol dependent method [17]. In addition lymphocytes do not produce detectable amounts of oxygen radical. Thus free radical generation in BAL fluid from our patients probably reflects mainly AM production. PMA in­duced CL was not significantly different from controls. A previous activation occurring in vivo might explain such a low response to PMA.

Abnonnal free radical production by AM is well established in IPF and both pulmonary and extrathora­cic sarcoidosis [18, 30]. Lymphokines enhance in vitro generation of oxygen reactive intermediates by mon­ocytes and monocyte-derived macrophages [31] but no correlation was found in our study between lymphocyte percentages in BAL fluid and AM CL. Oxygen radical production is a potent mechanism of respiratory mem­brane damage [32, 33] but the relative contribution of AM and recruited PMN to lung lesions is not yet detennined in ll..D. Experimental studies on lung ex­plants showed a significant injury by IPF-AM which was reduced by antioxidants [34]. In addition a marked synergistic oxidant mediated cytotoxicity by IPF cells and alveolar epithelial lining fluid has been recently demonstrated by CANTIN et al. [35]. Whatever its relevance in lung damage, oxidant generation reflects an AM activation which likely implicates other mecha­nisms and mediators.

An increased neutrophil chemotactic activity was detected in AM supernatants from six patients from group 1. Similarly, a significant neutrophil chemotactic activity was found by GARCIA et al. (13] in BAL fluid from RA patients without ILD but who demonstrated a subclinical lymphocyte alveolitis. Paradoxically there was no correlation in our study between NCA and PMN proportion in BAL. Such a discrepancy has re­cently been described in unaffected family mem­bers of autosomal idiopathic fibrosis [36}. In contrast, AM from IPF patients were shown to spontaneously release neutrophil chemotactic factor mainly when BAL neutrophils exceeded 10% [37]. NCA was not correlated with spontaneous superoxide anion and fi­bronectin generation, but an unexpected significant cor­relation appeared with PMA induced CL. Although mechanisms of secretion, specificity and physical prop­erties of NCA were not detennined in our study, several lines of evidence suggest that local immune complexes may be responsible for AM activation generating NCA [37, 38). The prognostic significance of elevated NCA in symptom free patients with mjni­mal or without neutrophil alveolitis remains unclear. Long term follow-up of group 1 patients is necessary to confinn the hypothesis that NCA secretion by AM leads to a secondary and progressive neutrophil alveo­litis and established interstitial disease.

Elevated fibronectin secretion was a third parameter of AM activation detected in 6 out of 9 patients. This

12 TH. PEREZ ET AL.

large glycoprotein produced by both fibroblasts and AM mediates cell matrix interactions such as cell adhesion and directed movement of parenchymal cells [39]. It is also a well recognized competence factor for fibroblast growth [ 40] and an elevated local production by AM is a classical feature of IPF [41]. Whether or not chronic fibronectin elevation could take part in subse­quent interstitial structural derangement remains to be determined, since optimal fibroblast proliferation usually requires both a competence factor and a progression factor.

Although unrelated to each other, superoxide anion generation, NCA and fibronectin secretions are indica­tive of an obvious though subclinical AM activation in our patients. Furthermore activation was detected in some patients despite normal BAL differential count~. suggesting that AM dysfunction might preceed recruit­ment of inflammatory cells such as neutrophils and lymphocytes. Thus study of AM secretory products appears as a useful and sensitive tool to investigate the early stage of subclinical alveolitis. Nevertheless the most relevant problem relates to the long term signifi­cance of this subclinical alveolitis in RA patients. In agreement with other studies, we found a high incidence of BAL cell abnormalities in symptom free patients. Moreover, increased AM generation of toxic and cell recruiting mediators was a frequent finding. One could hypothesize that in vivo simultaneous inhibitory mecha­nisms prevent lung damage in these patients. Such mechanisms include protective antioxidant interventions, namely antioxidant enzymes and oxygen radical scav­engers. Similarly, a low molecular weight factor which inhibits PMN chemotaxis and superoxide release was recently shown to be produced by AM [42]. Assessing the pathogenic significance of these abnormalities requires further follow-up studies of clinical and PFT data in these patients.

References

1. Walker WC, Wright V. -Diffuse interstitial pulmonary fibrosis and rheumatoid arthritis. Ann Rheum Dis, 1969, 28, 252-258. 2. Epler GR, M cLoud TC, Gaensler EA, Mikus JP, Carrington CB. - Normal chest roentgenograms in chronic diffuse infiltrative lung disease. N Eng J Med, 1978, 298, 934-939. 3. Frank ST, Weg JG, Harkleroad LE, Fitch RF. - Pulmo­nary dysfunction in rheumatoid disease. Chest, 1973, 63, 27-34. 4. Popper MS, Bogdonoff ML, Hughes RL. - Interstitial rheumatoid lung disease: a reassessment and review of the literature. Chest, 1972, 62, 243-249. 5. Cervantes Perez P, Toro-Perez AA, Rodriguez-Juardo P. - Pulmonary involvement in rheumatoid arthritis. lAMA, 1980, 243, 1715-1719. 6. Crystal RG, Bitterman PB, Rennard SI. Hance AJ. Keogh BA. - Interstitial lung diseases of unknown causes: dis­orders characterized by chronic inflammation of the lower respiratory tract (first of two parts). N Engi 1 Med, 1984, 310, 154-166.

7. Keogh BA, Crystal RG. - Alveolitis: the key to the in­terstitial lung disorders. Thorax, 1982, 37, 1-10. 8. Daniele RP, Elias JA, Epstein PE, Rossman MD. -Bronchoalveolar lavage: Role in the pathogenesis, diagnosis and management of interstitial lung disease. Ann Intern Med, 1985, 102, 93-108. 9. Rossi GA. - Bronchoalveolar lavage in the investigation of disorders of the lower respiratory tract. Eur J Respir Dis, 1986, 69, 293-315. 10. Haslam P, Turton C, Heard B, Lukoszek A, Collins N, Salsbury AJ, Turner-Warwick M. - Bronchoalveolar lavage in pulmonary fibrosis: comparison of cells obtained with lung biopsy and clinical features. Thorax, 1980, 35, 9- 18. 11. Wallaert B, Hatron PY, Grosbois JM, Tonne] AB, Devul­der B, Voisin C. - Subclinical involvement in collagen vas­cular diseases assessed by bronchoalveolar lavage. Relation­ship between alveolitis and subsequent changes in lung func­tion. Am Rev Respir Dis, 1986, 133, 57~580. 12. Garcia JGN, Parhami N, Killam D, Garcia PL, Keogh BA. - Bronchoalveolar lavage fluid evaluation in rheumatoid arthritis. Am Rev Respir Dis, 1986, 133, 450-454. 13. Garcia JGN, James HL, Zinkgrak S, Perlman MB, Keogh B. - Lower respiratory tract abnormalities in rheumatoid in­terstitial disease: potential role of neutrophils in lung injury. Am Rev Respir Dis, 1987, 136, 811-817. 14. Ropes MW, Bennett GA, Cobb S, Jacox R, Jessar RA. - Revision of diagnostic criteria for rheumatoid arthritis. Bull Rheum Dis, 1958, 9, 175-176. 15. Quanjer PH. - Standardized lung function testing report working party "Standardization" of lung function tests. Euro­pean Community for Coal and Steel, Luxembourg, 1981. 16. Joseph M, Tonne! AB, Torpier G, Capron A, Arnoux B, Benveniste J. - Involvement of immunoglobulin E in the se­cretory processes of alveolar macrophages from asthmatic patients. 1 Clin Invest, 1983, 71, 221-228. 17. Williams AJ, Cole PJ. -Investigation of alveolar macro­phage function using lucigenin-dependent chemiluminescence. Thorax, 1981, 36, 866-869. 18. Aerts C, Wallaert B, Grosbois JM, Voisin C. - Super­oxide anion release by alveolar macrophages in pulmonary sarcoidosis. Ann NY Acad Sci, 1986, 465, 193-200. 19. Gosset PH, Tonne! AB, Joseph M, Prin L, Mallart A, Charon J, Capron A. - Secretion of a chemotactic factor for neutrophils and eosinophils by alveolar macrophages from asthmatic patients. J Allergy Clin Immunol, 1984, 74, 827-834. 20. Ouaissi MA, Neyrinck JL, Capron A. - Development of a competitive radioimmunoassay for human plasma fi­bronectin. lnt Archs Allergy Appi Immunol, 1986, 81, 75-80. 21. Silver RM, Metcalf JF, Stanley JH, Leroy EC. - Inter­stitial lung disease in scleroderma. Analysis by bronchoalveo­lar lavage. Arthritis Rheum, 1984, 27, 125~1262. 22. Yousem SA, Colby TV, Carrington CB.- Lung biopsy in rheumatoid arthritis. Am Rev Respir Dis, 1985, 131, 770-777. 23. Tishler M, Grief J, Fiereman E, Y aron M, Topilsky M. - Bronchoalveolar lavage. A sensitive tool for early diagno­sis of pulmonary involvement in rheumatoid arthritis. J Rheu­matol, 1986, 13, 547-550. 24. Duquesnoy B, Asfour M, Santoro F, Vandemeule­Broucke B, Hochart JP, Delcambre B. - Beta-2-microglobulin, anti beta-2-microglobulin activity and circulating immune complexes in rheumatoid arthritis: serum and synovial study. Rev Rhum, 1980, 47,481-487. 25. Latt D, Weiss JB, Jayson MIV. - Beta-2-microglobulin levels in serum and urine of rheumatoid arthritis patients on gold therapy. Am Rheum Dis. 1981, 40, 57-60.

SUBCLINICAL ALVEOLITIS IN RHEUMATOID ARTHRITIS 13

26. Manicourt D, Brauman H, Orloff S. - Plasma and urinary levels of Beta-2-microglobulin in rheuamtoid arthritis. Ann Rheum Dis, 1978, 37, 328-332. 27. Janossy G, Duke C, Poulter LW, Panayi G, Bofill M, Goldsein G. - Rheumatoid arthritis: a disease of T-lympho­cyte/macrophage immunoregulation. Lancet, 1981, ii, 839-842. 28. Du Bois RM. - The alveolar macrophage. Thorax. 1985, 40, 321-327. 29. Herscowitz AB. - In defense of the lung: paradoxical role of the pulmonary alveolar macrophage. Ann Allergy, 1985, 55, 634--648. 30. Wallaert B, Ramon P, Foumier E, Prin L, Tonne! AB, Voisin C. - Activated alveolar macrophage and lymphocyte alveolitis in extrathoracic sarcoidosis without radiological mediastina pulmonary involvement. Ann NY Acad Sci, 1986, 465, 201-210. 31. Nakawara A, Desantis NM, Nogueira N, Nathan CF. -Lymphokines enhance the capacity of human monocytes to secrete reactive oxygen intermediates. J Clin Invest, 1982, 70, 1042-1048. 32. Brigham KL. - Role of free radicals in lung injury. Chest, 1986, 89, 859-863. 33. Martin WJ, Gadek JE, Hunninghake GW, Crystal RG. - Oxidant injury of lung parenchyma! cells. J Clin Invest. 1981, 68, 1277-1288. 34. Martin WJ, Davis WB. Gadek JE, Keogh BA, Crystal RG. - Alveolar macrophages from patients with idiopathic pulmonary fibrosis contribute to lung cell injury. Am Rev Respir Dis, 1982, 125, 91 (abstract). 35. Cantin AM, North SL, Fells GA, Hubbard RC, Crystal RG. - Oxidant mediated epithelial cell injury in idiopathic pulmonary fibrosis. J Clin Invest, 1987, 79. 1665-1673. 36. Bitterman PB, Rennard SI, Keogh BA, Wewers MD, Adelberg S, Crystal RG. - Familial idiopathic fibrosis. Evi­dence of lung inflammation in unaffected family members. N Engl J Med, 1986, 314, 1343- 1347. 37. Hunninghake GW, Gadek JE, Lawley TJ. - Mechanisms of neutrophil accumulation in the lungs of patients with idio­pathic pulmonary fibrosis. J Clin Invest, 1981, 68, 259- 269. 38. Hunninghake GW, Gadek JE, Pales HM, Crystal RG. -Human alveolar macrophage derived chemotactic factor for neutrophils: stimuli and partial characterization. J C/in Invest, 1980, 66, 473-483. 39. Ruoslahti E, Enguall E, Hayman EG. - Fibronectin: current concepts of its strucrure and functions. Collagen Rei Res. 1981, 1, 95- 128. 40. Bitterman PB, Rennard SI. Adelberg S, Crystal RG. -Role of fibronectin as a growth factor for fibroblasts. J Cell Bioi, 1983, 97, 1925-1932.

41. Rennard SI, Hunninghake GW, Bitterman BP, Crystal RG. - Production of fibronectin by the human alveolar macro­phage. Mechanism for the recruitment of fibroblasts to the sites of tissue injury in interstitial lung diseases. Proc Nat Acad Sci, 1981, 78, 7147- 7151. 42. Sibille Y, Merrill WW, Care SB, Cooper JAD, Dutton TA, Reynolds HY. - Further characterization of a factor re­leased by human alveolar macrophages which inhibit leu­cocyte ftmction. Am Rev Respir Dis, 1987, 135, A338 (ab­stract).

Inflammation alveolaire in Fraclinique au cours de la po/­yarthrite rehumatoiae: production par Jes macrophages alveolaires d'anions superoxyde, d'une activite chimiotactique pour les polynucteaires neontrophiles et de Fibronectine. Th. Perez, J M . Farre, Ph. Cosset, B. Wallaert, B. Duquesnoy, C. Voisin, B. Delcambre, A.B. Tonne/. RESUME: Les pneumopathies interstitieUes sont frequentes et volontiers infracliniques au cours de la polyarthrite rhumatolde puisqu'une alteration de transfert du CO est d&:eh~e chez 30 a 40% des patients. Le but de cette crude est de preciser grace au lavage broncho-alveolaire (LBA) !'incidence et les car­acteristiques de l'alveolite chez 21 patients non fumeurs ay­ant ou non une atteinte interstitielle dece1able. Le groupc 1 (n=l2) est constitue de patients asymptomatiques dont !'ex­ploration fonctionnelle respiratoire et la radiographic de thorax sont normales. Le groupe 2 (n=9) rassemble les patients ayant un syndrome interstitiel radiologique et/ou un syndrome restrictif avec trouble du transfert du CO. Une alveolite lymphocytaire (>18%) est decelee chez 5/12 patients du groupe 1, alors que 7/9 des patients du groupc 2 presen­tent une alveolite a polynuclCaires neutrophiles (>4%). Le pourcentage de lymphocytes est significativement plus eleve chez Jes patients ayant Un taUX serique de bcta-2-microglob­uline anormal. Trois activites secretoires des macrophages alveolaires (MA) en culture sont etudiees chez les patients du groupe 1. La production spontanee d'anions superoxyde est significativement elevee par rapport aux MA temoins (7260±2700 versus 850±120 unites lumineuses relatives/5.1 OS cellules). La secretion de fibronectine est significativement augmentee comparativement aux temoins (6.1±1.6 versus 1.3±0.2 ng·106 cellules/heure ), ainsi que I' activite chimiotac­tique du sumageant de culrure pour les polynucleaires neutrophiles (PNN) (21±4.8 versus 8.1±0.7 PNN/champ microscopique). La signification pronostique de l'alveolite lymphocytaire et de !'activation macrophagique decelees ehez des patients sans atteinte respiratoire decelable reste a deter­miner. Eur Respir J., 1989. 2, 7- 13.