Article Text

Extended Report
Prevalence and severity of interstitial lung disease in mixed connective tissue disease: a nationwide, cross-sectional study
  1. Ragnar Gunnarsson1,2,
  2. Trond Mogens Aaløkken3,
  3. Øyvind Molberg1,2,
  4. May Brit Lund2,4,
  5. Georg Karl Mynarek3,
  6. Åse Stavland Lexberg5,
  7. Kari Time6,
  8. Alvilde Sofie Strand Dhainaut7,
  9. Liv-Turid Bertelsen8,
  10. Øyvind Palm9,
  11. Karen Irgens10,
  12. Andrea Becker-Merok11,
  13. Jan Leidulf Nordeide12,
  14. Villy Johnsen13,
  15. Sonja Pedersen14,
  16. Anne Prøven15,
  17. Lamya Samir Noori Garabet9,
  18. Jan Tore Gran1,2
  1. 1Rheumatology unit, Oslo University Hospital Rikshospitalet, Oslo, Norway
  2. 2Institute of Clinical Medicine, University of Oslo, Oslo, Norway
  3. 3Department of Radiology, Oslo University Hospital Rikshospitalet, Oslo, Norway
  4. 4Department of Respiratory Medicine, Oslo University Hospital Rikshospitalet, Oslo, Norway
  5. 5Department of Rheumatology, Buskerud Hospital, Drammen, Norway
  6. 6Haugesund Sanitetsforeningens Revmatismesykehus, Haugesund, Norway
  7. 7Department of Rheumatology, St Olavs Hospital, Trondheim University Hospital, Trondheim, Norway
  8. 8Department of Rheumatology, Haukeland University Hospital, Bergen, Norway
  9. 9Department of Rheumatology, Østfold Hospital, Moss, Norway
  10. 10Department of Rheumatology, Ålesund Hospital, Ålesund, Norway
  11. 11Department of Rheumatology, Institute of Clinical Medicine, University of Tromsø, Tromsø, Norway
  12. 12Department of Rheumatology, Førde Central Hospital, Førde, Norway
  13. 13Department of Rheumatology, Sørlandet Hospital, Kristiansand, Norway
  14. 14Department of Rheumatology, Nordland Hospital, Bodø, Norway
  15. 15Department of Rheumatology, Martina Hansens Hospital, Bærum, Norway
  1. Correspondence to Ragnar Gunnarsson, Rheumatology Unit, Oslo University Hospital Rikshospitalet, PO Box 4950 Nydalen, 0424 Oslo, Norway; ragunnar{at}gmail.com

Abstract

Background Mixed connective tissue disease (MCTD) is an immune-mediated, systemic disorder of unknown cause.

Objective To assess the prevalence, pattern and severity of interstitial lung disease (ILD) in a cross-sectional study of the nationwide, Norwegian MCTD cohort.

Methods 126 patients with MCTD were systematically examined for ILD by high-resolution CT (HRCT), pulmonary function tests (PFT), 6 min walk test (6MWT) and by the New York Heart Association (NYHA) functional classification of dyspnoea. The extent and type of HRCT lung abnormalities were scored according to the CT criteria of ILD recommended by the Fleischner Society.

Results All 126 patients were Caucasian, 75% women. At the time of the cross-sectional ILD study, the patients had a mean disease duration of 9.0 years. 52% of the patients had abnormal HRCT findings, most commonly reticular patterns consistent with lung fibrosis (35%). Lung fibrosis was quantified as minor in 7%, moderate in 9% and severe in 19% of the patients. Fibrosis was uniformly concentrated in the lower parts of the lungs and was not associated with smoking. Patients with severe lung fibrosis had lower PFT values, shorter 6MWT and a higher mean NYHA functional class. After a mean 4.2 years' follow-up, overall mortality was 7.9%. Mortality in patients with normal HRCT was 3.3%, as compared with 20.8% in patients with severe lung fibrosis (p<0.01).

Conclusions Severe lung fibrosis is common in MCTD, has an impact on pulmonary function and overall physical capacity and is associated with increased mortality.

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Introduction

Mixed connective tissue disease (MCTD) was first described as a separate immune-mediated connective tissue disorder by Sharp et al in 1972.1 Since then, the concept of MCTD has persisted, but it is still debated whether the disease is a distinct entity or represents an overlap between systemic sclerosis (SSc), systemic lupus erythematosus (SLE) and polymyositis or dermatomyositis (PM/DM).2,,5 MCTD is characterised by serum autoantibodies targeting the U1-ribonucleoprotein6,,8 and distinct clinical features which include Raynaud's phenomenon,1 ,9,,12 ‘puffy’ or swollen hands,10,,12 arthritis,10,,13 pleuritis,10,,12 pericarditis,10,,12 myositis,10 ,14,,16 oesophageal dysmotility,10,,12 pulmonary hypertension8 ,11 ,17 ,18 and interstitial lung disease (ILD).8 ,11 ,19,,24 Many of these clinical features are present in other diseases, but their combined presence and a positive antiribonucleoprotein (RNP), appear to be unique for MCTD.

Accumulating evidence suggests that pulmonary disease and, particularly ILD, is a major cause of morbidity and mortality in many systemic connective tissue diseases (CTDs), including SSc and PM/DM.8 ,11 ,25 There are no data on the epidemiology of pulmonary disease in unselected groups of patients with MCTD, but in studies from tertiary referral centres, abnormal pulmonary function tests (PFTs) and with mostly a restrictive pattern suggestive of ILD, were seen in up to 90% of patients with MCTD.8 ,11 ,17 ,26 Little is known about the impact of ILD in MCTD, but an early study reported that more than 20% of the patients developed hypoxaemia.8 Moreover, the data on lung imaging in MCTD are limited. To our knowledge, only four studies have assessed lung abnormalities in MCTD by high-resolution CT (HRCT).20 ,22 ,23 ,27 One of these studies, from our own hospital, focused only on juvenile-onset MCTD.23 In the three studies on adult MCTD, the reported frequencies of HRCT abnormalities compatible with lung fibrosis ranged from 0% to 20%, while the frequency of ground-glass attenuations varied between 12% and 100%.20 ,22 ,27

Recently, we established a nationwide MCTD cohort consisting of every identifiable, living adult with MCTD in Norway.12 Altogether, 147 patients fulfilling at least one of the three most commonly used MCTD criteria28,,30 were included, giving a prevalence of adult MCTD in Norway of 3.8 per 100 000.12 At inclusion, all the patients were assessed cross-sectionally by an interdisciplinary study protocol which included echocardiography, PFTs and HRCT of the lungs.

The HRCT abnormalities found in ILD are mostly classified by a system which is based on lung histopathology and was originally developed for primary lung diseases.31 Since this system is not ideal for CTD-associated ILD, an alternative method that classifies the HRCT abnormalities according to radiological characteristics was used.32 ,33 In this study, the aim was to assess the prevalence, pattern and functional impact of ILD in the nationwide Norwegian MCTD cohort by HRCT, PFTs and overall functional capacity tests.

Patients and methods

Patients

The establishment of the Norwegian nationwide MCTD cohort and the inclusion criteria have been described previously.12 Briefly, during the inclusion period, which lasted from March 2005 to December 2008, the aim was to identify and enrol every living, adult patient with MCTD in Norway. Patients were included from all the 16 Norwegian public hospitals that have a department of rheumatology. The inclusion criteria were (1) age ≥18 years; (2) a clinical diagnosis of MCTD verified by a rheumatologist; (3) a positive anti-RNP test; (4) fulfilment of at least one of the three criteria sets for MCTD (the modified Sharps criteria,30 the criteria of Alarcón-Segovia28 or those of Kasukawa)29 and (5) exclusion of other CTDs. All the patients who fulfilled the inclusion criteria gave informed consent to be enrolled in the MCTD cohort. At enrolment, all the patients were examined by a study protocol which included echocardiography; lung imaging by HRCT; PFTs; dyspnoea evaluation by the New York Heart Association (NYHA) functional classification34; 6 min walk test (6MWT)35 and blood tests. In addition, the patients underwent a systematic clinical assessment. The presence of pleuritis, pericarditis and arthritis was based on clinical symptoms and findings, whereas the presence myositis was defined by increased serum creatine kinase levels and proximal muscle weakness. The myositis was, in some cases, confirmed by muscle biopsy and/or electromyography.

Study inclusion

All the patients with MCTD who had available for review an electronic file containing complete lung HRCT scan data performed at enrolment were included in this study. Patients who did not have an accessible electronic HRCT file were excluded from the study.

Patient follow-up

The follow-up of the study population was carried out according to routine clinical care, at the local departments of rheumatology. Information about deaths during cohort follow-up, until 1 January 2011, was gathered from the Norwegian Statistical Institute (http://www.ssb.no).

Analyses of HRCT lung scans

Thin-section CT images were obtained in the supine position during breath-holding and deep inspiration. Supplementary expiratory and prone scans were obtained when indicated. Several different CT scanners at 10 hospitals were used. All the available HRCT images were reviewed in consensus and in random order, blinded to patient history and lung function, by two experienced chest radiologists (TMA and GKM).

The presence, extent and distribution of ground-glass attenuations, airspace consolidation, reticular patterns and interlobular septal thickening were evaluated according to the CT criteria of ILD recommended by the Nomenclature Committee of the Fleischner Society.32 If ground-glass attenuations were superimposed on a reticular pattern, the abnormality was recorded as being reticular. The presence of associated findings, including traction bronchiectases, nodules, pleural irregularities, cysts, emphysema and air trapping, was also assessed. Since ‘honeycombing’, is a poorly defined parameter,33 it was not assessed.

The reticular patterns (ie, the coarseness of the lung fibrosis) were graded as follows; 1, a fine intralobular pattern without evident cysts (figure 1A); 2, a pattern with predominant small cysts involving air spaces ≤4 mm in diameter (figure 1B) and 3, a pattern with larger cysts involving air spaces >4 mm (figure 1C). When a reticular pattern grade 1 was superimposed on a grade 2 pattern, the abnormality was recorded as a reticular pattern grade 2. If different types of reticular patterns were present in different lung zones of an individual patient, both pattern types were recorded.

Figure 1

(A–C) Overview of reticular pattern on high-resolution CT of the lungs.

The distribution of disease was reviewed in four lung zones: (A) above the aortic arch, (B) between the aortic arch and the level of the carina, (C) between the level of the carina and the level of the inferior pulmonary veins and (D) below the inferior pulmonary veins (figure 2).

Figure 2

Lung fibrosis distribution in the four lung zones.

The extent of ground-glass attenuations and reticular patterns in each zone was assigned an area score based on the percentage of the lung parenchyma involved; 0, no involvement; 1, 1–4% involved; 2, 5–14%; 3, 15–29%; 4, 30–49% and 5, >50% involved.

A total reticular pattern score was calculated by adding up the individual area scores (0–5) of the reticular pattern changes in each of the four lung zones. A total reticular score of 0 was defined as no fibrosis. A total reticular score of 1–2 was defined as minor fibrosis, 3–4 as moderate fibrosis and ≥5 as severe fibrosis.

Pulmonary function tests

The PFT of each patient was performed within 6 weeks before or after the HRCT and included dynamic spirometry and gas diffusion capacity. Recorded variables were forced vital capacity (FVC), forced expiratory volume in 1 s (FEV1), single-breath carbon monoxide transfer factor (TLCO) and TLCO divided by alveolar volume (TLCO/AV). The tests were performed according to published guidelines.36 ,37 The PFT values were expressed in absolute terms and as a percentage of the predicted value. Predicted values were derived from reference equations, separated by gender with age and height as predictor variables.38

Screening for precapillary pulmonary hypertension

Three of the 126 patients had pulmonary hypertension (PH) already at inclusion. Of the other 123, 121 were screened for PH by echocardiography at study inclusion. An estimated systolic arterial pressure >40 mm Hg was set as the predefined cut-off point for referral to right-sided heart catheterisation (RHC). Precapillary PH was defined by RHC according to the 2009 European Society of Cardiology and the European Respiratory Society Guidelines.39 All the patients with MCTD suspected to have PH at study inclusion or during follow-up were referred for RHC at the Department of Cardiology, Oslo University Hospital.

Ethical permission

The study was approved by the regional committee for research ethics in the Southern and Eastern Norway Regional Health Authority and The Norwegian Social Science Data Services.

Data analysis

Completed data forms were sent to the primary investigator at the Unit of Rheumatology, Oslo University Hospital Rikshospitalet, for central registration and analysis. Cardiff Teleform 10.1 scanning software (http://www.intelliscan.com/TeleForm1.htm) was used to import data to the Microsoft Office Access database software 2000/2007. Differences between groups were tested by the independent sample t test, the Mann–Whitney U test, Mantel–Haenszel χ2 test or Fisher's exact test as appropriate. All tests were two tailed, with a 95% significance level. Statistical analyses were performed by SPSS/PASW version 19 (http://www.spss.com/) and OpenEpi version 2.3.1 (http://www.openepi.com/OE2.3).

Results

Study population and patient characteristics

The study population included the 126 patients with MCTD for whom HRCT of the lung had been carried out when they were enrolled in the Norwegian MCTD cohort and had an electronic file with the results accessible for review. All the 126 patients included were Caucasian and 75% were women. At the time of enrolment the 126 patients had a mean age of 44.9 years (95% CI 40.0 to 53.9) and a mean disease duration of 9.0 years (95% CI 7.6 to 10.4). The 126 patients comprised 86% of the patients enrolled in the MCTD cohort. At least 11 of the 21 excluded patients, had undergone HRCT at the time of inclusion, but electronic files with the HRCT data were not accessible for review. No significant differences were noted between the excluded and included patients for mean age, sex ratio, mean disease duration, overall disease manifestations, smoking status and mortality (data not shown).

Overview of lung HRCT abnormalities

In 52% of the patients, one or more abnormalities on the HRCT were identified (table 1 and supplementary table SA). The most common abnormalities were reticular patterns, being present in 44 patients (35%).The reticular patterns were most often graded as type 1 (21%) and/or type 2 (16%), whereas type 3 changes were seen in 6% only (table 1). The reticular changes were mostly present in the basal lung zones (figure 2), while the other abnormalities had no clear distribution pattern (supplementary table SA). Isolated ground-glass attenuations were identified in two patients only (2%) (table 1).

Table 1

Overview of HRCT lung findings in the patients with mixed connective tissue disease. (Note that individual patients may have more than one single HRCT abnormality)

The degree of lung fibrosis in each patient was defined by the total reticular score. According to this scoring system, 9 (7.1%) of the patients with MCTD had minor fibrosis (reticular score 1–2), 11 (8.7%) had moderate fibrosis (reticular score 3–4) and 24 (19.0%) had severe fibrosis (reticular score ≥5). In mild and moderate cases, the lung fibrosis mostly affected the lower lung zones, while the severe fibrosis was more widespread (figure 2).

Association between HRCT abnormalities and patient characteristics

The 24 patients with severe lung fibrosis were older than the other patients, but had a shorter mean disease duration (table 2). There were equal percentages of smokers in those with or without HRCT changes (table 2). The presence of lung fibrosis was negatively associated with arthritis during the course of the disease (p<0.01), but not with arthritis at study inclusion (table 2). There was no association between lung fibrosis and pleuritis, pericarditis or myositis, (table 2).

Table 2

Clinical and epidemiological data of 126 patients with MCTD assessed by HRCT of the lungs

Five patients were identified with precapillary PH by RHC at inclusion or during follow-up. Three patients had precapillary PH confirmed by RHC already at inclusion and two developed PH during follow-up. One of these two patients developed progressive dyspnoea, whereas the other was asymptomatic, but echocardiography performed as part of her routine clinical follow-up indicated that the pulmonary artery systolic pressure was abnormal. Three of the patients diagnosed with PH had lung fibrosis at inclusion (two severe and one moderate) (table 2).

Seventy-four per cent of the patients fulfilled all three MCTD criteria sets, 24% fulfilled two and 2% only one of the criteria sets. In the group with severe lung fibrosis, 75% fulfilled the Alarcón-Segovia criteria,28 while 100% fulfilled the Sharp's30 and Kasukawa criteria29 (table 2).

Association between lung fibrosis, PFTs and physical capacity tests

Patients with severe fibrosis had lower mean FVC and FEV1 than the other patient groups, indicating that they had more severe restrictive lung disease (table 3). They also had lower mean TLCO, shorter mean 6MWT and higher mean NYHA functional class than the other groups (table 3).

Table 3

Pulmonary function tests, classification of dyspnoea and functional testing, in 105 patients with mixed connective tissue disease with normal HRCT and fibrotic HRCT changes

Association between HRCT findings and mortality

After a mean observation time of 4.2 years, altogether 10 of the 126 patients had died, giving an overall death rate of 7.9%. Among the 61 patients without HRCT abnormalities at inclusion there were two deaths (3.3%), while eight of the 65 patients (12.3%) with baseline HRCT abnormalities had died. The highest death rate (20.8%) was noted in patients with severe lung fibrosis. Lung fibrosis, independent of grade, was associated with increased mortality (p<0.05) (table 2).

Discussion

A number of small, single-centre studies8 ,11 ,19 ,20 ,22 ,26 ,27 have indicated that ILD is common in MCTD, but data on its overall prevalence and clinical impact are lacking. Here, we used a cross-sectional approach to identify the type, distribution, extent and impact of ILD in an unselected nationwide MCTD cohort. We report that HRCT identified lung abnormalities in more than half of the patients and that one in five had severe lung fibrosis. After a mean 4 years' observation time, the death rate in patients with severe lung fibrosis was increased.

As early as in 1975, Bennett et al26 described abnormal PFTs in nine of 10 patients with MCTD. In 1984, a prospective study showed that 21% of the patients with MCTD developed hypoxaemia.8 Although later studies have reinforced the view that pulmonary disease is a major clinical problem in MCTD, the available data on lung imaging by HRCT are limited.20 ,22 ,27 The three studies that exist differ in their design and are not easily comparable, as only one of them provided PFT data. In the first of these studies, which only included patients with an abnormal HRCT, 41/41 patients had ground-glass attenuations.22 The second study, also from Japan, reported HRCT abnormalities in 35/60 patients, seven with ground-glass attenuations and 18 with changes consistent with lung fibrosis.20 The third and largest study, from Hungary, assessed HRCTs taken before and after 6 months with prednisolone±cyclophosphamide in 144 patients with MCTD with a mean 13-year disease duration.27 Before treatment, 96/144 patients had ground-glass attenuations, but lung fibrosis was not seen. After 6 months, however, 20% of the patients had developed HRCT signs of lung fibrosis.27 The authors found no correlation between HRCT findings and PFTs, either before or after the treatment course.

There is no consensus on how lung fibrosis should be defined by HRCT.40,,43 Here, a system based on adding up the reticular changes in four individual lung zones to obtain a total reticular score corresponding to mild, moderate or severe fibrosis was employed. Interestingly, even though the cut-off values for the total reticular score were arbitrary, it appeared that the patient subset which had a score above five (defined as severe lung fibrosis), had lower mean PFT values, shorter mean 6MWT and a higher mean NYHA functional class than the other subsets. This patient subset also had the highest death rate. We believe that the strong associations found between the reticular score, the functional tests and all-cause mortality indicate that the scoring system applied is valid.

As the majority of the MCTD cohort had no lung histology available, the American Thoracic Society/European Respiratory Society criteria for idiopathic interstitial pneumonias,31 were not applied. We did notice, however, that the reticular patterns identified in MCTD resembled the HRCT pattern seen in patients with non-specific interstitial pneumonia (data not shown). Notably, in the radiological scoring system applied, the reticular changes were always defined as dominant, meaning that an area with ground-glass attenuations superimposed on reticular changes was recorded as an area with reticular changes. It is therefore possible that the low frequency of ground-glass attenuations in our study, compared with previous studies,20 ,22 may partly be explained by the methodology applied. Another, not mutually exclusive, possibility is differences in disease activity at the time of sampling. Since ground-glass attenuations are mainly thought to reflect acute inflammatory processes in the lungs, they are more likely to occur in patients admitted to hospital (owing to increasing disease activity, flares or infections) than in the current, cross-sectional included, patient cohort.

There are additional differences between our study and the three previous HRCT studies. First, our patients had mean disease duration of 9 years, which is shorter than the 13 years in the Hungarian study,27 but far longer than the 4.5 months22 and 4 years20 in the two studies from Japan. Second, at least 38% of our patients were former or current smokers, the Japanese studies reported smoking in only 7%22 and 20%,20 respectively, while the Hungarian study excluded smokers.27 Finally, the patients differed in age, sex ratio and probably also in ethnicity. Our patients were 4–7 years younger than in the other studies and the female to male ratio was lower (3:1 vs 12:127 and 40:122).

Our study has some weaknesses. First, HRCT scans from 21 of the 147 patients, in the Norwegian MCTD cohort were missing.12 Eleven of the 21 missing patients had undergone a baseline HRCT (which was described by a radiologist), but the electronic files could not be retrieved. Notably, there were no differences in disease characteristics, sex, age, smoking status and PFTs between the included and excluded patients, suggesting that the dataset obtained was representative for the whole cohort. Second, because of the multicentre nature of the study, the HRCTs were carried out with different scanners and the PFTs were performed at different respiratory laboratories. We do not, however, believe that this had any major impact on the results.

The patients with severe fibrosis had a shorter mean disease duration at study inclusion (6.4 years) than patients with minor or moderate fibrosis (13.2 years), indicating that the subset of patients who develop the most severe lung fibrosis also has a more rapidly progressive disease than the other subsets. The only detected clinical difference between patients with and without severe lung fibrosis was the frequency of clinical arthritis during the disease course. We have no good explanation for this negative association.

Here, the different MCTD criteria sets seemed to select for different patient subsets. The most striking example was seen in the patients with severe lung fibrosis. All these patients fulfilled both the Kasukawa29 and the Sharp criteria,30 but 25% of them did not fulfil the Alarcón-Segovia criteria.28 The reason is probably that pulmonary disease is not included in Alarcón-Segovia criteria set,28 but is an integral part of both the Kasukawa29 and the Sharp criteria.30

It is not known why pulmonary disease and particularly lung fibrosis, is so frequent in MCTD. It might be related to the presence of anti-RNP autoantibodies, but the fact that anti-RNP-positive patients with SLE rarely develop pulmonary disease, speaks against it. Another possibility is that the MCTD subset who develop lung fibrosis have an SSc-like or PM/DM-like phenotype. This possibility is not readily supported by our clinical data, but it does appear that the reticular pattern seen in many of the patients with MCTD, to some degree, resembles the pattern seen in SSc.44 ,45

In conclusion, this study shows that HRCT findings compatible with lung fibrosis are common in unselected patients with MCTD and that lung fibrosis has an impact on pulmonary function, overall physical capacity and even, mortality. Systematic follow-up studies of the nationwide, Norwegian MCTD cohort described will hopefully provide additional data on the long-term morbidity and mortality of pulmonary disease in MCTD.

Acknowledgments

The authors thank Mrs Torhild Garen at Rheumatology Unit at Oslo University Hospital Rikshospitalet for assistance in gathering data and Mrs Brynhildur Axelsdóttir librarian for assistance.

References

View Abstract

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Files in this Data Supplement:

    • Web Only Data - This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

Footnotes

  • Funding Grants from the Norwegian Rheumatism Association and the Scandinavian Rheumatology Research Foundation.

  • Patient consent Obtained.

  • Ethics approval The study was approved by the regional committee for research ethics (REK) in the Southern and Eastern Norway Regional Health Authority and the Norwegian Social Science Data Services (NSD).

  • Provenance and peer review Not commissioned; externally peer reviewed.