Evaluation of interstitial lung disease in mixed connective tissue disease (MCTD)

Abstract

Objective. Interstitial lung disease (ILD) may be a characteristic, often serious, manifestation of mixed connective tissue disease (MCTD). In this retrospective study, the frequency and clinical picture of ILD were determined in patients with MCTD using two diagnostic tests: high-resolution computed tomography (HRCT) and inhaled aerosol clearance times of ^99mTc-labelled diethylene-triamine pentaacetate (^99mTc-DTPA). In addition, pulmonary function, effects of therapy and a variety of immunoserological markers were also assessed.

Methods. One hundred and forty-four consecutive patients with MCTD were selected from the clinic, irrespective of the presence or absence of ILD. All patients underwent a detailed clinical assessment, chest HRCT scanning, chest radiography, inhaled aerosol of ^99mTc-DTPA clearance times, and all pulmonary function tests. Patients who had active ILD received corticosteroid (CS) or CS in combination with cyclophosphamide (CPH). All investigations were repeated after 6 months of immunosuppressive therapy.

Results. Ninety-six out of 144 MCTD patients (66.6%) had active ILD, 75 of this group (78.1%) showed ground glass opacity, 21 patients (21.8%) ground glass opacity with mild fibrosis with HRCT. Forty-five patients with active ILD received 2 mg/kg/day CS for 6–8 weeks alone and 51 patients CS in combination with CPH (2 mg/kg/day). Six months later, after therapy, 67 out of 96 MCTD patients with ILD (69.8%) showed a negative HRCT pattern, ground glass opacity with mild fibrosis developed in 15 patients (15.6%), and fibrosis was detected in 13 patients (13.5%). Only one patient showed subpleural honeycombing. ^99mTc-DTPA was rapid in all 96 MCTD patients with active ILD (28.7 ± 8.2 min, normal value >40 min). After therapy the ^99mTc-DTPA was normalized, 79 out of 96 patients (82.3%). Carbon monoxide diffusion capacity (DLCO) was reduced in 33 out of 96 MCTD patients with active ILD (34.3%), while there were no significant differences in the pulmonary function tests between the active versus inactive stage of ILD or versus patients without ILD. The sera of 96 MCTD patients with active ILD contained a high level of immune complexes (ICs), and the total haemolytic complement levels (CH50/ml U) decreased. After 6 months of therapy, the IC levels decreased and CH50/ml levels normalized (MCTD patients before and after active ILD: IC optical density = 355 ± 227 vs 206 ± 92, P<0.001; CH50/ml, 38.0 ± 12.6 U vs 64.3 ± 13.0 U, P<0.001).

Conclusions. HRCT is the gold standard for diagnosis of ILD. However, we used another method, ^99mTc-DTPA, in order to compare this technique with HRCT. This latter technique has not been studied previously in MCTD. The elevated levels of IC and increased complement consumption indicated that IC-mediated alveolocapillary membrane damage and tissue injury might play a role in the pathogenesis of ILD in MCTD.

Mixed connective tissue disease (MCTD) is a systemic autoimmune disease [1–3]. The most common clinical features of MCTD include Raynaud's phenomenon, polyarthritis, sclerodactyly, oesophageal hypomotility, renal disease and inflammatory myositis. The most relevant immunolaboratory marker of MCTD is autoantibody to nuclear ribonucleoprotein (U1RNP) [4–7].

Pulmonary impairment is not usually clinically evident early in the course of the disease. Indeed, the respiratory complications of MCTD were not reported in the original publication describing MCTD [1]. In a prospective study published in 1976, 80% of MCTD patients had evidence of pulmonary disease, but 69% were asymptomatic [8].

Recently, it has become evident that interstitial lung disease (ILD), such as fibrosing alveolitis, as well as pulmonary hypertension (PH) are rather serious pulmonary complications of MCTD [9–11]. In a recent report, ILD was detected in 48% of MCTD patients using high-resolution computed tomography (HRCT) [12]. These authors suggest that the prognosis of ILD may be better in MCTD than in systemic sclerosis (SSc). However, the pathomechanism and the therapy of ILD have not been widely studied in MCTD.

Recently, ^99mTc-labelled diethylene-triamine pentaacetate (^99mTc-DTPA) aerosol inhalation scintigraphy was described as a sensitive and non-invasive test commonly used to assess pulmonary epithelial membrane permeability. This technique was introduced into the present study in order to evaluate its possible applicability in MCTD-associated ILD [13].

In this study, ILD was investigated in patients with MCTD using HRCT and ^99mTc-DTPA scintigraphy. Results of HRCT, ^99mTC-DTPA scintigraphy and lung function tests were correlated with other clinical symptoms, as well as with immunolaboratory markers of MCTD. MCTD patients with ILD received either corticosteroid (CS) monotherapy or the combination of CS and cyclophosphamide (CPH). Therefore, all of the above tests were performed at baseline and repeated after 6 months of therapy.

Patients and methods

Patients and screening methods

The study population consisted of 144 consecutive patients with MCTD who were admitted to and underwent follow-ups at the special clinics of our institution.

The diagnosis of MCTD was established according to the criteria described by Alarcon-Segovia and Villarreal, which include Raynaud's phenomenon, swelling of the hands with a spindle-like swelling of the fingers, sclerodactyly, polyarthritis, myositis and the presence of anti-U1RNP autoantibodies in the patient's sera [14]. All patients fulfilled the diagnostic criteria for MCTD. Demographic data and the clinical manifestations of these patients are shown in Table 1. The mean age of 144 patients with MCTD was 49.1 ± 8.6 yr (range 29 to 73 yr). The mean follow-up of the patients was 13.4 ± 5.1 yr (range 4 to 25 yr).

Medical history was taken with special reference to previous cardiopulmonary disease, coughs, dyspnoea, sputum, chest pain and risk factors for pulmonary disease, such as smoking. Smokers and patients with diagnosed extrinsic asthma were excluded.

All patients underwent echocardiography, electrocardiography (ECG), chest radiography, HRCT scanning, ^99mTc-DTPA clearance, full pulmonary function tests, and immunolaboratory investigations were performed over a 1-week time period. Twenty patients had pulmonary hypertension (PH) without ILD.

The study received ethical approval from the Institutional Review Board of the University of Debrecen Medical Center. In addition, the subjects’ consent was obtained according to the Declaration of Helsinki.

Immunolaboratory investigations

Immunolaboratory tests included the assessment of immunoglobulins, immune complexes, rheumatoid factor, autoantibodies and antinuclear antibodies in the sera of patients. Autoantibodies to U1RNP were determined by enzyme-linked immunsorbent assay (ELISA) (Pharmacia & Upjohn, Freiburg, Germany). Antinuclear antibodies were detected on Hep-2 cells by an indirect immunofluorescence method. Total haemolytic complement activity (CH50/ml) was measured in the haemolytic assay. The serum level of C3 was assayed by nephelometry (Beckman) [15]. Serum immune complexes were assayed by polyethyleneglycol (PEG) precipitation.

Pulmonary HRCT

Pulmonary HRCT was performed as previously described [16] using a Siemens Somatom DR CT scanner. Scanning time was 7 s. The high-resolution protocol was performed with the window levels set for optimal imaging of the lungs. The HRCT images were obtained at 120 kV and 420 mA s. A high spatial resolution algorithm (bone algorithm) was used. A thin collimation of 1–2 mm was applied. Routinely, patients were examined in the supine position. All scans were obtained at full inspiration. No contrast material was used. Three scans were performed sequentially in all cases: one 1–2 cm above the upper level of diaphragm, one at the level of the tracheal carina and one at the level of the aortic arch. The results of HRCT were assessed independently by two radiologists. HRCT was performed at baseline and after 6 months’ treatment for ILD. All of the first and the second scans were blindly read by the same investigator (K.D.). Because of the close association between HRCT findings and lung histology, we used the grading system described by Wells et al. with minor modifications by Dévényi and Czirják [16, 17].

^99mTc-DTPA clearance

Patients inhaled 3 ml of aerosol containing ⁹⁹Tc-DTPA (Mallinckrodt, The Netherlands). The aerosol was produced using a nebulizer (Acorn). The aerosol contained submicrometre particles (mass median aerodynamic diameter 1.4 ± 0.2 mm). Patients inhaled the aerosol for 3 min in a supine position. One-minute frames were acquired in the posterior projection for 30 min, using a gamma camera (Siemens Orbiter, Siemens, Germany). Regions of interest (ROI) were defined for each lung using a 10% edge cut-out. Background subtraction was employed. Clearance rates were found to be mono-exponential; rates were calculated over 30 min in either lung separately and both lungs together. As the difference between these two measures was minimal for analysis, the clearance rate for both lungs together was used. Clearance rate was defined as when radioactivity reached half of its initial radioactivity (t_1/2) and was expressed in minutes. Based on our previous work in our laboratory, an abnormal clearance was <45 min which is >2 s.d. below the mean. ⁹⁹Tc-DTPA clearance was evaluated at baseline and after 6 months of therapy.

Lung function tests

Lung function tests expressed as percentages of the predicted values based on age, sex, height and weight, were completed within 1 week before or after HRCT scanning and ^99mTc-DTPA clearance. Forced expiratory volume in 1 s (FEV1), total lung capacity (TLC) and diffusion capacity for carbon monoxide (DLCO) were analysed. DLCO was corrected for haemoglobin and alveolar volume.

Data were expressed as percentages of the predicted values. Lung function was regarded as abnormal when TLC and FEV1 were <80% of the predicted value and DLCO when they were <75% of the predicted values. Quality control procedures and reference values were applied according to guidelines of the European Community for Coal and Steel [18]. Lung function tests were performed at baseline and after 6 months of immunosuppressive therapy.

Regimen of therapy

At the time of the first HRCT evaluation, 56/144 patients received non-steroidal anti-inflammatory drugs (NSAID), 39/144 patients CS (10–20 mg methylprednisolone), 31/144 patients CS plus sulphasalazine and 18/144 patients CS plus antimalarials. After the diagnosis of ILD, all 96 patients received 2 mg/kg/day CS (methylprednisolone), and 51/96 patients not responding to initial CS therapy administered for 4 to 6 weeks received CS in combination with CPH (2 mg/kg/day p.o. or 15 mg/kg five times i.v.).

Results

Epidemiology and clinical features of ILD

The clinical features of the patients, both with and without ILD, are described for comparison in Tables 2 and 3. There was no difference in the age and the duration of the disease between the MCTD patients with versus without ILD.

Dyspnoea, tachycardia and bibasilar crackles were significantly associated with the existence of active ILD (P<0.001). Bibasilar crackles were found in 48 out of 96 patients (50%) with active ILD, while after treatment crackles were present in only seven cases. Among the patients without ILD, crackles were observed in only two patients. There was no difference in the results of pulmonary function tests between MCTD patients with versus without ILD. In addition, no occupational or other risk factors for the development of ILD, other than MCTD, were present in any patients (data not shown).

Radiological investigations

Chest X-rays showed abnormalities consisting of small irregular opacities predominantly in the bases and middle regions in 87/96 cases (90.6%). However, the radiological picture could not differentiate between pulmonary infiltrates and vascular abnormalities. Therefore, lung HRCT was performed in all cases. Ground glass opacities with hyperattenuated regions were the most common abnormalities on HRCT (78.2% of patients), and ground glass opacity with mild fibrosis, with thickened non-septal or septal lines in 21.8% of the patients (Table 4). After 6 months of therapy, 67/75 (89.3%) abnormal HRCT cases became negative. In addition, ground glass opacity with mild fibrosis developed in 15/96 patients (15.6%), mild fibrosis in 13/96 patients (13.5%), and the HRCT showed subpleural honeycombing in only one case (1.0%).

^99mTc-DTPA scintigraphy and clearance

^99mTc-DTPA scintigraphy was performed in all patients. The clearance time was abnormally rapid in all 96 MCTD patients complicated with ILD (mean t_1/2 28.7 ± 8.2 min; range 14.4–43.2 min) (Table 5). Six months later, after immunosuppressive treatment, the clearance time normalized in 79/96 patients (82.3%; mean t_1/2 57.0 ± 10.3 min; range 32–78 min), while in MCTD patients without ILD, ^99mTc-DTPA clearance stayed within the normal range, with some age-related fluctuations (<40 min).

Lung function tests

The mean values of the FEV1 and TLC did not differ between the acute stage of ILD and after treatment, but the DLCO values were significantly lower in the active pulmonary stage (DLCO in the active ILD and after treatment: 82.5 ± 25.3 vs 94.3 ± 23.3; P<0.001) (Table 6).

Six months later, DLCO was impaired in 5/96 patients (5.2%), while restrictive ventilation damage (decreased TLC and FEV1) developed in 31/96 patients (32.3%). Lung function tests were normal in 60/96 cases (62.5%) (data not shown).

Immunolaboratory investigations

There were high levels of anti-U1RNP autoantibodies, immune complexes, complement C3 factor and CH50 in the sera of patients with active ILD (Table 6). We could not detect anti-Jo1 autoantibodies in the investigated sera. After immunosuppressive therapy, serum anti-U1RNP autoantibody levels, immune complex concentrations, total complement activity (CH50/ml), C3 factor and serum CRP significantly decreased in comparison with baseline values (anti-U1RNP 28.5 ± 35.4 vs 10.5 ± 17.0 U/ml; P<0.01; immune complexes optical density (OD) = 355 ± 227 vs OD = 236 ± 133, P<0.001; CH50/ml 38.0 ± 12.6 U vs 64.3 ± 13.0 U, P<0.001; C3 0.92 ± 0.24 g/l vs 1.15 ± 0.21 g/l, P<0.01; CRP 25.7 ± 17.7 mg/l vs 6.3 ± 3.5 mg/l, P<0.001) (Table 7).

Discussion

Pulmonary manifestations were not described in the original publication on MCTD by Sharp et al. [1]. However, long-term follow-up studies demonstrated that the lung is a common target organ in 25 to 85% of MCTD patients [19–21]. The two most relevant pulmonary complications in MCTD are PH and ILD. PH is considered to be the most serious, often fatal, complication of MCTD [22–26].

ILD is the most common pulmonary manifestation of MCTD. The reported frequency of ILD in MCTD is between 21 and 50% [27, 28]. Sullivan et al. described chest X-ray abnormalities in 10/34 MCTD patients [29]. Small irregular opacities were found predominantly in the lower and midlung fields during a 1 to 18 yr follow-up period.

HRCT is a sensitive method for detecting the small alveolar structures and can allow a confident evaluation of diffuse infiltrative lung diseases [30–33]. Pulmonary involvement in MCTD may be detected using HRCT. Kozuka et al. found that the predominant abnormalities include areas of ground glass attenuation with subpleural micronodules [34]. Saito et al. [12] compared the pulmonary manifestations of MCTD, SSc, systemic lupus erythematosus (SLE) and polymyositis using HRCT. Ground glass opacities were found in 4/35 MCTD patients, which was significantly lower than found in other connective tissue diseases. In that study, the predominant HRCT pattern of MCTD was the interlobular septal thickening (82.9%).

In the present retrospective study, ILD was detected by HRCT in 66.6% of our MCTD patients. Clinical signs of ILD included dyspnoea at rest in 75%, and dry cough in 12.5% of the patients. Two patients had dry cough and haemoptysis, raising the differential diagnostic possibility of pulmonary embolism. Most MCTD patients (78.8%) had evidence of ground glass shadowing on HRCT, which was associated with active alveolitis. This is usually considered as a feature of early and potentially reversible disease [33].

⁹⁹Tc-DTPA scintigraphy is useful for assessing lung epithelial permeability and is a sensitive marker of pulmonary inflammation [35]. The clearance of ^99mTc-DTPA is increased in adult respiratory distress syndrome [13]. Hill et al. [36] monitored ILD in one patient with dermatomyositis, and found that HRCT and ^99mTc-DTPA clearance were useful adjuncts in the initial assessment and the effect of therapy in ILD associated with dermatomyositis. Kula et al. [37] found increased ^99mTc-DTPA clearance in patients infected with hepatitis C virus. In our present study, ^99mTc-DTPA clearance was assessed in all 96 MCTD patients with ILD. Clearance was abnormally fast in all investigated MCTD patients at the active time of ILD, and after therapy the clearance time normalized in 75/96 patients (78.1%). Therefore, although HRCT is still the gold standard for the diagnosis of ILD, ^99mTc-DTPA scintigraphy may also give some information on lung function.

A reduction of FEV1 and total lung capacity was found in 34 and 41% of MCTD patients, respectively. In addition, decreased DLCO was detected in MCTD [21, 29]. In our MCTD population, reduced DLCO was the most sensitive test for predicting the presence of fibrosing alveolitis on HRCT, but the overall correlation of pulmonary function with radiographic appearance was poor. After 6 months of CS or/and CPH therapy, 89.3% of the positive HRCT cases became negative, and mild fibrosis developed in one-third of the patients.

There have been no large cohort studies on the outcome of immunosuppressive treatment on MCTD-associated ILD. It has been suggested that the acute inflammatory process is more likely to respond to CS and cytotoxic drugs, while pulmonary fibrosis is unlikely to do so [38]. In another prospective study, 27/34 MCTD patients with ILD required CS treatment and 7/34 were treated with CPH due to severe pulmonary inflammation [39]. Takeda et al. [40] reported a 43-yr-old woman with MCTD who rapidly developed interstitial infiltrates and responded well to pulse CS therapy (methylprednisolone 1 mg/day for 3 days). A 55-yr-old MCTD patient with rapidly declining pulmonary function and interstitial fibrosis failed to respond to CS, but improved on azathioprine therapy [41]. Treatment of MCTD-associated ILD with CS and CPH is based on the immunopathogenesis of ILD and also on the concept that in ILD inflammation usually leads to fibrosis rather than healing and resolution [42–45].

In this study, 46.9% of MCTD patients responded well to 2 mg/kg/day CS. However, 53.1% of the patients did not respond properly to 4–6 weeks of CS monotherapy; therefore, these patients received combination of CS and CPH. CPH was highly effective in the group of our patients not responding to CS therapy. Eventually, mild pulmonary fibrosis developed in only 30.2% of patients with ILD. Our results suggest that CS is effective in suppressing ILD in about half of MCTD patients. However, if CS seems to be ineffective the combination of CS and CPH may be used to suppress inflammation.

Histopathological abnormalities of ILD in MCTD are similar to those seen in idiopathic pulmonary fibrosis (IPF), including alveolar septal infiltration by lymphocytes, plasma cells, as well as the deposition of type III collagen [42, 43]. The development of ILD in MCTD involves alveolar injury from an unknown trigger, a persistent or repeated immune–inflammatory phase, and dysregulated tissue repair [44]. Studies in experimental animal models demonstrated a role for C5aR and FcgRIII in the initiation of IgG immune complex-mediated inflammation in mice. Induction of immune complex-mediated alveolitis resulted in strong accumulation of neutrophils in the lungs, and in increased production of pro-inflammatory cytokines [45]. In this study, we were unable to perform bronchoalveolar lavage (BAL) in most patients in order to analyse the composition of bronchial and alveolar cells and the different cytokines in BAL. However, the sera of MCTD patients with ILD contained abundant immune complexes, and there was obvious complement consumption indicated by decreased CH50 and serum C3 levels.

In summary, this is the first study where HRCT, ^99mTc-DTPA scintigraphy, lung function and immunolaboratory tests have been comprehensively used to analyse MCTD-associated ILD. Lung HRCT is a useful method for diagnosing ILD. In addition, ^99mTc-DTPA scintigraphy may give additional information on lung function. CS monotherapy or a combination of CS and CPH may be effective in preventing the progression of ILD and the development of fibrosis in the early inflammatory stage of ILD associated with MCTD.

The authors have declared no conflicts of interest.

References

Author notes

Divisions of Clinical Immunology and 1Rheumatology, 3rd Department of Internal Medicine, 2Department of Radiology and 3Department of Nuclear Medicine, University of Debrecen, Medical and Health Science Center, Debrecen, H-4004, Hungary.