Progression and mortality of interstitial lung disease in mixed connective tissue disease: a long-term observational nationwide cohort study

Progression and mortality of interstitial lung disease in mixed connective tissue disease: a... Abstract Objectives To assess the prevalence, extent, progression, functional impact and mortality of interstitial lung disease (ILD) in a nationwide unselected MCTD cohort. Methods The study cohort included patients with high-resolution CT lung scans available at baseline (n = 135) and at follow-up (n = 119). The extent of disease was expressed as percentage of total lung volume (TLV). Results ILD was present in 41% of MCTD patients at follow-up. Median (interquartile) extent (% of TLV) was 5 (8) at baseline and 7 (17) at follow-up, mean length 6.4 years later. The lung disease progressed in 19% of patients across the observation period. Predictors of ILD progression were elevated anti-RNP titre [hazard ratio (HR) 1.5, 95% CI: 1.1, 2.0; P = 0.008], presence of anti-ro52 antibodies (HR = 3.5, 95% CI: 1.2, 10.2; P = 0.023), absence of arthritis (HR = 0.2, 95% CI: 0.1, 0.6; P = 0.004) and male gender (HR = 4.0, 95% CI: 1.4, 11.5; P = 0.011) after age and baseline disease adjustments. The risk of death increased by 2.9 (95% CI: 1.1, 7.9; P = 0.038) in patients where disease involved ⩾5% of TLV. Conclusion Lung disease extent and progression in MCTD are modest. Yet, the extension continues several years after MCTD diagnosis causing lung function decline and increasing the risk of mortality. The study identified male gender, elevated anti-RNP titre, presence of anti-ro52 antibodies and absence of arthritis as the strongest predictors of ILD progression. mixed connective tissue disease, interstitial lung disease, pulmonary fibrosis, mortality, anti-ribonucleoprotein antibodies Rheumatology key messages Interstitial lung disease is typically modest with slow progression and stable disease equally present. Progression of lung disease was evident in 1/5 of patients resulting in lung function deterioration. Risk factors for lung disease were male gender, high anti-RNP titres, anti-ro52 antibodies and absence of arthritis. Introduction MCTD is a chronic, immune-mediated disorder defined by the combined presence of serum anti-RNP antibodies and selected clinical features of SSc, SLE, RA and PM/DM. The entity was first described by Sharp et al. in 1972 [1, 2]. Since then it has been debated whether MCTD is a distinct disease, an overlap syndrome or a transient phase before developing into another CTD [3, 4]. Genetic studies from Europe [5–7] and Asia [8] support the notion that MCTD is a distinct disease. Interstitial lung disease (ILD) is a major cause of morbidity and mortality in the CTDs, particularly in SSc [9, 10] and PM/DM [11]. In MCTD prevalence figures from different cohorts have ranged from 47 to 78% [12, 13]. Previously, our research group investigated the prevalence and incidence of MCTD between the years of 2005 and 2008. One hundred and forty-seven patients were included in the nationwide MCTD cohort [14]. Analyses of lung high-resolution CT at study inclusion found that reticular pattern was generally present in the basal zones and the most common abnormality, resembling the pattern seen in SSc [15]. Studies from other countries have reported different CT findings. In a study on 41 MCTD patients from Japan, interlobular septal thickening was the predominant pattern [16], while a Hungarian study reported that ground glass attenuation was highly frequent [17]. In a recent report from Brazil the most common patterns were ground glass and reticular changes [18]. Histological descriptions of MCTD-related lung pathologies are scarce. Hajas et al. [12] reviewed lung biopsies from 16 MCTD patients where the ILD diagnosis could not be resolved by imaging alone. All these 16 patients had findings compatible with non-specific interstitial pneumonitis, where 11 had cellular interstitial pneumonitis with little or no fibrosis and the remaining 5 had significant fibrosis. Data on the course and outcome of ILD in MCTD are limited. Bodolay et al. [17] reported mild fibrosis in the majority of MCTD patients, but did not present how this was evaluated. Kawano-Dourado et al. [18] recently presented serial CT analyses in 37 MCTD patients followed at a tertiary centre, and showed that the median percentage ILD score in the lower lobes increased from 7.5 to 11.3% over 10 years, indicating restricted disease progression. There are no larger longitudinal studies on the overall progression of ILD in unselected MCTD cohorts, and outcome risk factors are needed. Predictive factors for progressive ILD have been studied in SSc [19] but remain unknown in MCTD. ILD is known to predict mortality in SSc [19] and PM/DM [11, 20–22] while the effect of ILD on mortality has not been established in MCTD cohorts. Here, we present a long-term observational study of the unselected, nationwide Norwegian MCTD cohort, with serial CT and pulmonary function test analyses. The aim of this study was to examine the prevalence, extent, progression and functional effect of ILD in MCTD. We also wanted to identify predictive factors of ILD progression in MCTD, aiming to aid the clinician in identifying MCTD patients at risk of deteriorating lung function. Finally, we wanted to estimate the effect of ILD on mortality in MCTD. Methods Patient population The Norwegian nationwide MCTD cohort recruited patients from Departments of Rheumatology in Norway during the years of 2005–08 [15] with the following inclusion criteria: age above 18 years; fulfillment of at least one of the three criteria sets of MCTD, the modified Sharp’s criteria [1], the criteria of Alarcón-Segovia and/or Kasukawa [23]; and exclusion of another CTD. In the present study we included all patients from the nationwide MCTD cohort where baseline CTs were attainable for assessment (n = 135), encompassing all 126 patients included in the evaluation of ILD prevalence and severity from 2012 [15] as well as nine additional patients from the nationwide MCTD cohort. Clinical data Clinical parameters included symptoms and signs included in the three classification criteria as well as smoking history, medications and vital status. The clinical parameters were recorded at two time points: baseline and follow-up. Baseline was defined as study entry. Myositis was confirmed by muscle biopsy and/or electromyogram and CK elevation. Smokers had smoked >1 cigarette a day for >1 year. Precapillary pulmonary hypertension was defined according to the updated European Society of Cardiology criteria by mean pulmonary artery pressure ⩾25 mmHg and pulmonary wedge pressure ⩽15 mmHg by right-sided heart catheterization at rest [24]. Vital status at the end of the study was obtained from the National Population Register of Norway. Acquisition and analysis of CT images CT lung examination was included in the study protocol at baseline and at follow-up. High-resolution CT images were obtained with the patient in the supine position, during breath-holding and inspiration. The images were reconstructed at 1–1.25 mm section thickness with 10-mm intervals and at 2.5-mm section thickness reformatted in the axial, coronal and sagittal planes. Images were reviewed on a picture archiving and communication system screen in consensus and in random order by two specialist chest radiologists. The observers evaluated the presence and extent of ILD. These findings included reticular pattern, ground glass attenuation, traction bronchiectasis, interlobular septal thickening and airspace consolidations according to the CT criteria of ILD recommended by The Nomenclature Committee of the Fleischner Society [25]. Although the term honeycombing is common in radiological studies, the definition is not standardized [26]. Reticular pattern (i.e. the coarseness of fibrosis) was therefore graded as follows: grade 1, fine intralobular reticular pattern without evident cysts; grade 2, predominantly micro-cystic reticular pattern involving air spaces ⩽4 mm in diameter; and grade 3, a predominantly macro-cystic reticular pattern with air spaces >4 mm in diameter. If ground glass attenuations were superimposed on a reticular pattern, the abnormality was recorded as being reticular. When a grade 1 pattern was superimposed on a grade 2 pattern, the abnormality was recorded as grade 2. If different types of reticular patterns were present in different lung zones of an individual patient, both pattern types were recorded. Abnormal CT findings of ground glass attenuation and reticular patterns grades 1–3 were defined as ILD. The distribution of disease was reviewed in four lung zones: zone 1, above the aortic arch; zone 2, between the aortic arch and at the level of the carina; zone 3, between the level of the carina and the level of the inferior pulmonary veins; and zone 4, below 1 cm above the top of the right hemidiaphragm. The extent of involvement was evaluated independently for each zone and each zone was assigned a percentage of the lung parenchyma that showed evidence of ILD. The total extent of disease in each patient was added in the four lung zones and expressed as a percentage of total lung volume (TLV). ILD progression was defined as a disease extension as a percentage of TLV and stable ILD was defined as no disease extension as a percentage of TLV between baseline and follow-up. Minimal time between the baseline and follow-up CT was 12 months. Follow-up CT scans were available in 119 of 135 patients. Clinical information about the 16 patients where follow-up lung CT examination was not available can be found in supplementary Table S1, available at Rheumatology Online. The CT scans were subjectively assessed for possible oesophageal motility disorders. Substantial distension was recorded when columns of air and distension larger than doubling the oesophageal wall thickness was identified. Pulmonary function test analyses The pulmonary function test analyses were performed within 8 weeks of the corresponding CT and performed according to the American Thoracic Society/European Respiratory Society guidelines [27], using an automated Vmax V6200 system (SensorMedics, VIASYS Respiratory Care Inc, Yorba Linda, CA, USA). Recorded variables were the percentage of predicted value of full vital capacity (FVC% pred) and the percentage of predicted value of diffusing capacity of the lung for carbon monoxide (DLCO% pred). Autoantibody analyses Serum levels of anti-RNP autoantibody at baseline and follow-up in addition to anti-ro52 antibodies at baseline were determined by line immunoassay (ANA Profile 5 Euroline Blot test kit, Euroimmun, Lübeck, Germany). Strips were scanned to obtain optical density values according to the kit instructions. Anti-ro52 antibodies optical density values of ⩾10 were recorded as positive. Statistical analyses The data were presented as means (s.d.) or median and interquartile range (IQR) depending on the level of resemblance to the normal distribution. Comparisons of groups were performed appropriately by chi-square test, Fisher’s exact test, or one-way ANOVA with post hoc comparisons by Tukey’s test or the Kruskal–Wallis test and Mann–Whitney U test. Cox regression analyses were used to find predictive factors of lung disease progression. Using a manual backward elimination procedure, variables at a significance level of P < 0.25 in the univariable analyses were considered a candidate for the multivariable model. The prediction was quantified by the hazard ratio (HR) with its 95% CI. Patients were followed from the date of inclusion in the cohort until death or end of follow-up on 30th of April 2016. Kaplan–Meier survival curves were used to examine differences in survival between patients with and without disease compromising ⩾5% and ⩾10% of TLV. The log-rank test was used to assess differences. The proportional hazard assumptions were tested by plotting the logarithm of the integrated hazards (log–log survival plots). To estimate the total effect of ILD on mortality with minimal bias, we performed multivariable cox regression analyses adjusting for variables selected from the directed acyclic graph presented in supplementary Fig. S1, available at Rheumatology Online [28]. Data extraction and analyses were conducted using SPSS Statistics version 22 (IBM Corp., Armonk, NY, USA) and STATA version 22 (StataCorp, College Station, TX, USA). Ethical approval The study was approved by the Norwegian Regional Committee for Medical and Health Research Ethics and conducted in accordance with the guidelines of the Helsinki declaration. All patients had given informed written consent to participate in the study. Results Prevalence, extent and distribution Serial lung CT images available for evaluation were retrieved in 119 of 135 patients. There were no significant differences found in terms of gender, age at diagnosis, disease duration at baseline or classification criteria when comparing these 119 patients against the 16 patients with missing follow-up. Mean (s.d.) age at diagnosis in the 135 patients with baseline CT data was 35 (16) years and mean (s.d.) disease duration was 9 (8) years. The Alarcón-Segovia criteria were met by 90%, 97% fulfilled the Sharp criteria and 86% fulfilled the Kasukawa criteria. The mean (s.d.) time between baseline and follow-up CT was 6.4 (2.0) years. CT findings compatible with ILD were present at baseline in 40% and at follow-up in 41% of the patients (Table 1). There were only two new cases of ILD during the study observation. Median (IQR) extent of lung disease increased from 5 (8)% of TLV at baseline to 7 (17)% of TLV at follow-up (Table 1). The extension was generally slow, apparent by the number of patients with disease involving ⩾5, 10 and 20% of TLV at follow-up compared with baseline (Table 1). Distribution analyses demonstrated lung disease to be present largely in the lower lung lobes, but about half of the patients with lung disease did have findings of ILD in the upper part of the lungs (i.e. in zone 1: see ‘Methods’ section for details) (Table 1). Fine reticular pattern (grade 1) was the most common abnormality; however, macro-cystic reticular pattern (grade 3) covered a larger area of the lungs in the patients where this pattern was evident (Table 1). Table 1 Detailed description of lung CT findings at baseline and follow-up in MCTD patients (n = 119) Lung CT findings  Baseline  Follow-up  Age at CT examination, mean (s.d.), years  44.3 (15)  51.1 (14)  Disease duration at time of CT, mean (s.d.), years  9.5 (8)  16.5 (8)  Prevalence of interstitial lung disease, n (%)  47 (40)  49 (41)  Extent          Median (IQR) % of TLV  5 (2–10)  7 (2–19)      ≥5% of TLV, n (%)  24 (20)  30 (25)      ≥10% of TLV, n (%)  13 (11)  20 (17)      ≥20% of TLV, n (%)  7 (6)  12 (10)  Distribution          Zone 1, n (%)  25 (21)  28 (24)      Median (IQR) % of TLV  5 (2–10)  8 (3–10)      Zone 2, n (%)  27 (23)  32 (27)      Median (IQR) % of TLV  5 (3–15)  10 (5–19)      Zone 3, n (%)  40 (34)  42 (35)      Median (IQR) % of TLV  10 (3–15)  10 (5–23)      Zone 4, n (%)  46 (39)  47 (39)      Median (IQR) % of TLV  10 (5–21)  15 (5–45)  Patterns          Ground glass attenuation, n (%)  3 (3)  4 (3)      Median (IQR) % of TLV  13 (11–16)  11 (9–48)      Grade 1, n (%)  31 (26)  33 (28)      Median (IQR) % of TLV  3 (2–4)  3 (1–5)      Grade 2, n (%)  19 (16)  22 (18)      Median (IQR) % of TLV  5 (3–10)  5 (3–11)      Grade 3, n (%)  9 (8)  13 (11)      Median (IQR) % of TLV  14 (8–27)  13 (5–39)  Other CT changes          Traction bronchiectasis, n (% of patients with ILD)  14 (30)  20 (41)      Emphysema, n (% of patients with ILD)  11 (15)  11 (16)      Dilated oesophagus, n (% of patients with ILD)  43 (57)  52 (65)  Lung CT findings  Baseline  Follow-up  Age at CT examination, mean (s.d.), years  44.3 (15)  51.1 (14)  Disease duration at time of CT, mean (s.d.), years  9.5 (8)  16.5 (8)  Prevalence of interstitial lung disease, n (%)  47 (40)  49 (41)  Extent          Median (IQR) % of TLV  5 (2–10)  7 (2–19)      ≥5% of TLV, n (%)  24 (20)  30 (25)      ≥10% of TLV, n (%)  13 (11)  20 (17)      ≥20% of TLV, n (%)  7 (6)  12 (10)  Distribution          Zone 1, n (%)  25 (21)  28 (24)      Median (IQR) % of TLV  5 (2–10)  8 (3–10)      Zone 2, n (%)  27 (23)  32 (27)      Median (IQR) % of TLV  5 (3–15)  10 (5–19)      Zone 3, n (%)  40 (34)  42 (35)      Median (IQR) % of TLV  10 (3–15)  10 (5–23)      Zone 4, n (%)  46 (39)  47 (39)      Median (IQR) % of TLV  10 (5–21)  15 (5–45)  Patterns          Ground glass attenuation, n (%)  3 (3)  4 (3)      Median (IQR) % of TLV  13 (11–16)  11 (9–48)      Grade 1, n (%)  31 (26)  33 (28)      Median (IQR) % of TLV  3 (2–4)  3 (1–5)      Grade 2, n (%)  19 (16)  22 (18)      Median (IQR) % of TLV  5 (3–10)  5 (3–11)      Grade 3, n (%)  9 (8)  13 (11)      Median (IQR) % of TLV  14 (8–27)  13 (5–39)  Other CT changes          Traction bronchiectasis, n (% of patients with ILD)  14 (30)  20 (41)      Emphysema, n (% of patients with ILD)  11 (15)  11 (16)      Dilated oesophagus, n (% of patients with ILD)  43 (57)  52 (65)  TLV: total lung volume. Progression Patients were stratified in three groups: group 1, no ILD (n = 70); group 2, stable ILD (n = 26); and group 3, ILD progression (n = 23) (Table 2). The median (IQR) annual disease progression in group 3 was 1.1 (2)% of TLV. Compared with group 1 fewer patients had arthritis and anti-RNP titres were higher at baseline and at follow-up in group 3 (Table 2). Frequency of cases meeting the Alarcón-Segovia criteria was lowest in group 3. Group 2 (stable ILD) differed from the two other groups by a higher occurrence of patients with dilated oesophagus. Table 2 Characteristics of MCTD patients stratified by lung CT results Characteristics  Group 1  Group 2  Group 3  P-value    No ILD  Stable ILD  ILD progression      n = 70  n = 26  n = 23    Demographics              Time baseline to follow-up CT, mean (s.d.), years  6.5 (2)  6.2 (3)  6.0 (2)  NS      Disease duration at follow-up, mean (s.d.), years  16.7 (8)  17.0 (9)  16.3 (10)  NS      Male gender, n (%)  16 (23)  5 (13)  9 (39)  NS      Age at diagnosis, mean (s.d.), years  31.4 (15)  38.1 (18)  38.4 (18)  NS      Age at baseline, mean (s.d.), years  41.6 (14)  48.6 (13)  47.7 (17)  NS      Age at follow-up, mean (s.d.), years  49.0 (14)  55.7 (14)  54.0 (16)  NS      Smokers, n (%)  37 (54)  14 (54)  10 (44)  NS      Alarcon criteria, n (%)  66 (94)  23 (89)  17 (74)  0.025a      Kasukawa criteria, n (%)  58 (83)  26 (100)  22 (96)  0.030a      Precapilary pulmonary hypertension  1 (1)  2 (8)  2 (9)  NA  CT features              Extent of ILD at baseline, median (IQR), % of TLV  NA  3.5 (2–8)  8.8 (3–22)  NS      New cases, n (incidence proportion %)  NA  NA  2 (3)  —      Extent of ILD at follow-up, median (IQR), % of TLV  NA  3.5 (2–8)  17.5 (6–43)  <0.001b      Annual progression rate, median (IQR), %  NA  NA  1.1 (0.5–2.5)  —  SLE-like features cumulative at baseline              Arthritis, n (%)  63 (90)  20 (77)  12 (52)  <0.001a      Pericarditis, n (%)  7 (10)  6 (23)  2 (9)  NS  SSc-like features at baseline              Sclerodactily, n (%)  19 (27)  13 (50)  8 (35)  NS      Dilated oesophagus on CT, n (%)  15 (22)  20 (77)  8 (35)  <0.001a  PM-like features cumulative at baseline              Myositis, n (%)  13 (19)  4 (15)  5 (22)  NS  Laboratory features              Anti-RNP at baseline, median (IQR), ×10−3 U/l  65 (16–159)  123 (49–240)  240 (59–240)  0.004c      Anti-RNP at follow-up, median (IQR), ×10−3 U/l  30 (6–92)  150 (29–240)  194 (12–240)  0.005c      ESR at baseline, mean (s.d.), mm  18 (14)  22 (18)  29 (18)  0.012d      Hb at baseline, mean (s.d.), g/dl  13.3 (1.1)  13.0 (1.2)  13.3 (1.6)  NS      Anti-ro-52 positivity at baseline, n (%)  12 (19)  8 (32)  9 (39)  NS  Characteristics  Group 1  Group 2  Group 3  P-value    No ILD  Stable ILD  ILD progression      n = 70  n = 26  n = 23    Demographics              Time baseline to follow-up CT, mean (s.d.), years  6.5 (2)  6.2 (3)  6.0 (2)  NS      Disease duration at follow-up, mean (s.d.), years  16.7 (8)  17.0 (9)  16.3 (10)  NS      Male gender, n (%)  16 (23)  5 (13)  9 (39)  NS      Age at diagnosis, mean (s.d.), years  31.4 (15)  38.1 (18)  38.4 (18)  NS      Age at baseline, mean (s.d.), years  41.6 (14)  48.6 (13)  47.7 (17)  NS      Age at follow-up, mean (s.d.), years  49.0 (14)  55.7 (14)  54.0 (16)  NS      Smokers, n (%)  37 (54)  14 (54)  10 (44)  NS      Alarcon criteria, n (%)  66 (94)  23 (89)  17 (74)  0.025a      Kasukawa criteria, n (%)  58 (83)  26 (100)  22 (96)  0.030a      Precapilary pulmonary hypertension  1 (1)  2 (8)  2 (9)  NA  CT features              Extent of ILD at baseline, median (IQR), % of TLV  NA  3.5 (2–8)  8.8 (3–22)  NS      New cases, n (incidence proportion %)  NA  NA  2 (3)  —      Extent of ILD at follow-up, median (IQR), % of TLV  NA  3.5 (2–8)  17.5 (6–43)  <0.001b      Annual progression rate, median (IQR), %  NA  NA  1.1 (0.5–2.5)  —  SLE-like features cumulative at baseline              Arthritis, n (%)  63 (90)  20 (77)  12 (52)  <0.001a      Pericarditis, n (%)  7 (10)  6 (23)  2 (9)  NS  SSc-like features at baseline              Sclerodactily, n (%)  19 (27)  13 (50)  8 (35)  NS      Dilated oesophagus on CT, n (%)  15 (22)  20 (77)  8 (35)  <0.001a  PM-like features cumulative at baseline              Myositis, n (%)  13 (19)  4 (15)  5 (22)  NS  Laboratory features              Anti-RNP at baseline, median (IQR), ×10−3 U/l  65 (16–159)  123 (49–240)  240 (59–240)  0.004c      Anti-RNP at follow-up, median (IQR), ×10−3 U/l  30 (6–92)  150 (29–240)  194 (12–240)  0.005c      ESR at baseline, mean (s.d.), mm  18 (14)  22 (18)  29 (18)  0.012d      Hb at baseline, mean (s.d.), g/dl  13.3 (1.1)  13.0 (1.2)  13.3 (1.6)  NS      Anti-ro-52 positivity at baseline, n (%)  12 (19)  8 (32)  9 (39)  NS  aχ2 test. bGroup 2 vs group 3, Kruskal–Wallis and Mann–Whitney U test. cGroup 1 vs group 3, Kruskal–Wallis and Mann–Whitney U test. dGroup 3 vs group 1, one-way ANOVA and post hoc Tukey’s test. ILD: interstitial lung disease; NA: not applicable; NS: not significant. Functional impact The extent of ILD at baseline CTs was greater in group 3 than group 2, although not significantly (Table 2). Figure 1 illustrates the effect of ILD on mean lung function test results. FVC% pred differed significantly in group 3 compared with group 2 (P = 0.002) and group 1 (P < 0.001) at baseline and at follow-up (both P < 0.001). Group 3 differed significantly in lower mean values of DLCO% pred compared with group 1 (P < 0.001) but not group 2 at baseline, while at follow-up DLCO% pred was significantly lower in group 3 compared with both group 1 (P < 0.001) and group 2 (P = 0.001). FVC% pred in group 2 did not differ significantly from group 1 at baseline or at follow-up. DLCO% pred in group 2 differed significantly from group 1 at baseline (P = 0.003) but not at follow-up. Fig. 1 View largeDownload slide Pulmonary function test results stratified by lung CT results Fig. 1 View largeDownload slide Pulmonary function test results stratified by lung CT results Treatment Analyses of immune-modulating treatments did not indicate clear differences between groups 1–3, except for higher accumulated frequencies of AZA usage in patients with stable ILD (group 2) (supplementary Table S2, available at Rheumatology Online). Predictors of progression Multiple Cox regression analyses revealed that the strongest predictive factors at baseline for ILD progression at follow-up were male gender, elevated anti-RNP titre, the presence of anti-ro-52 antibodies and absence of arthritis when adjusting for baseline lung disease and age (Table 3). A rise of 50 U/ml in anti-RNP titre resulted in 50% increased risk of ILD progression (95% CI: 1.1, 2.0; P = 0.008), while a history of arthritis reduced the risk by 80% (95% CI: 0.1, 0.6; P = 0.004). The risk of ILD progression was 4-fold in male patients (95% Cl: 1.4, 11.5; P = 0.011) and >3-fold in patients with anti-ro52 antibodies (95% Cl: 1.2, 10.2; P = 0.023). According to Harrell’s C index, patient outcomes were accurately predicted by this model 91% of the time. Table 3 Predictors of interstitial lung disease progression in MCTD analysed by cox regression   Univariable  Multivariable  Clinical predictors  HR (95% CI)  P-value  HR (95% CI)  P-value  Alarcon criteria  0.26 (0.10, 0.67)  0.005      FVC % of predicted  0.95 (0.93, 0.98)  <0.001      DLCO % of predicted  0.94 (0.92, 0.97)  <0.001      ESR (per 10 mm increase)  1.5 (1.2, 1.9)  <0.001      % ILD of TLV  1.1 (1.1, 1.1)  <0.001  1.1 (1.0, 1.1)  0.002  Arthritis  0.15 (0.06, 0.36)  <0.001  0.22 (0.08, 0.61)  0.004  Anti-RNP value (per 50 U/l increase)  1.4 (1.1, 1.8)  0.009  1.5 (1.1, 2.0)  0.008  Age at baseline  1.0 (1.0, 1.1)  0.065  1.1 (1.0, 1.1)  0.028  Male gender  2.0 (0.9, 4.6)  0.112  4.0 (1.4, 11.5)  0.011  Anti-ro-52 positivity  3.2 (1.3, 7.7)  0.01  3.5 (1.2, 10.2)  0.023    Univariable  Multivariable  Clinical predictors  HR (95% CI)  P-value  HR (95% CI)  P-value  Alarcon criteria  0.26 (0.10, 0.67)  0.005      FVC % of predicted  0.95 (0.93, 0.98)  <0.001      DLCO % of predicted  0.94 (0.92, 0.97)  <0.001      ESR (per 10 mm increase)  1.5 (1.2, 1.9)  <0.001      % ILD of TLV  1.1 (1.1, 1.1)  <0.001  1.1 (1.0, 1.1)  0.002  Arthritis  0.15 (0.06, 0.36)  <0.001  0.22 (0.08, 0.61)  0.004  Anti-RNP value (per 50 U/l increase)  1.4 (1.1, 1.8)  0.009  1.5 (1.1, 2.0)  0.008  Age at baseline  1.0 (1.0, 1.1)  0.065  1.1 (1.0, 1.1)  0.028  Male gender  2.0 (0.9, 4.6)  0.112  4.0 (1.4, 11.5)  0.011  Anti-ro-52 positivity  3.2 (1.3, 7.7)  0.01  3.5 (1.2, 10.2)  0.023  DLCO: diffusing capacity of the lung for carbon monoxide; FVC: forced vital capacity; HR: hazard ratio; ILD: interstitial lung disease; TLV: total lung volume. Effect on all-cause mortality The effect of ILD on mortality was evaluated by dividing the patient cohort in two subsets: disease extent <5% (n = 108), and ⩾5% of TLV (n = 27) and <10% (n = 122) and ⩾10% (n = 13) of TLV. The 5-year cumulative survival rates for <5% and ⩾5% disease of TLV were 94% (95% CI: 88, 98%) and 82% (95% CI: 61, 92%), respectively. The 10-year cumulative survival rates were 87% (95% CI: 79, 92%) and 70% (95% CI: 49, 84%; log rank P = 0.015) (Fig. 2). The 5-year cumulative survival rates for <10% and ⩾10% disease of TLV were 93% (95% CI: 85, 95%) and 77% (95% CI: 44, 92%), respectively, and the 10-year cumulative survival rates were 86% (95% CI: 78, 91%) and 60% (95% CI: 29, 81%), respectively (log rank P = 0.01). Cox regression univariable analyses revealed that having disease extending ⩾5% of TLV was associated with 3-fold increased risk of death (HR: 2.9, 95% CI: 1.2, 7.0; P = 0.020), while having disease involving 10% of TLV increased the risk by almost fourfold (HR: 3.5, 95% CI: 1.3, 9.5; P = 0.015) (Table 4). A disease extension of 1% of TLV was associated with increased risk of death (HR: 1.1, 95% CI: 1.0, 1.1; P < 0.001) (Table 4). The total effect of % ILD extent on all-cause mortality (see supplementary Fig. S1, available at Rheumatology Online) was similar after adjustments to smoking, gender, disease duration and age (Table 4). Table 4 Effect of interstitial lung disease on all-cause mortality in the MCTD cohort Interstitial lung disease  Univariable    Multivariablea      HR (95% CI)  P-value  HR (95% CI)  P-value  Model 1: ≥5% of TLV  2.8 (1.6, 2.8)  0.020  2.9 (1.1, 7.9)  0.038  Model 2: ≥10% of TLV  3.5 (1.3, 9.5)  0.015  3.4 (1.1, 10.2)  0.028  Model 3: per 1% of TLV increase  1.1 (1.0, 1.1)  <0.001  1.1 (1.0, 1.1)  <0.001  Interstitial lung disease  Univariable    Multivariablea      HR (95% CI)  P-value  HR (95% CI)  P-value  Model 1: ≥5% of TLV  2.8 (1.6, 2.8)  0.020  2.9 (1.1, 7.9)  0.038  Model 2: ≥10% of TLV  3.5 (1.3, 9.5)  0.015  3.4 (1.1, 10.2)  0.028  Model 3: per 1% of TLV increase  1.1 (1.0, 1.1)  <0.001  1.1 (1.0, 1.1)  <0.001  Univariable and multivariable cox regression analyses. a Vriables included in the multivariable models: age at baseline, smoking, disease duration at baseline and gender. TLV: total lung volume. Fig. 2 View largeDownload slide Kaplan–Meier curves for interstitial lung disease involving ≥5% and <5% of total lung capacity Fig. 2 View largeDownload slide Kaplan–Meier curves for interstitial lung disease involving ≥5% and <5% of total lung capacity Discussion ILD is frequently reported in MCTD cohorts and studies of the long-term clinical impact of these changes are necessary to assist rheumatologists in the management of MCTD. Here, we found that 19% of patients in our nationwide, unselected MCTD cohort had ILD progression on CT, where the mean (s.d.) disease duration was 17 (8) years. The annual progression is modest, but results in deteriorating lung function in this relatively young patient population. Based on the presenting, high resolution, long-term observational study in addition to comparable work on smaller patient populations [17, 18] and cross-sectional studies [2], it appears reasonable to draw the following conclusions: interstitial lung changes occur frequently in MCTD; ILD extent in MCTD is typically modest; approximately half of the patients with established lung disease have progression; the progression is generally slow, but appears to continue several years after diagnosis causing deteriorating lung function; a subset of patients, in our study 10%, had developed extensive ILD (>20% of TLV); and, finally, ILD is associated with increased risk of mortality, and should therefore be regarded as a major complication in MCTD. Interestingly we found that increasing anti-RNP antibody titres at baseline was a strong predictor of ILD progression. Greidinger et al. [29] found anti-70K antibodies to be strongly correlated with lung disease in 70K immunized mice. Notably, in a recent study on juvenile MCTD our colleagues found increased anti-RNP antibody titres in patients with active disease [30] corresponding to previous findings of Burdt et al. [31]. The presence of anti-ro-52 antibodies was recently found to be associated with ILD in our MCTD cohort [32]. We here present that the presence of anti-ro-52 antibodies is additionally a predictor of ILD progression in MCTD. Furthermore, we found arthritis to be protective of lung disease progression, possibly indicating that MCTD patients with more SLE-like features are less likely to develop severe lung disease. Previous studies of MCTD have found oesophageal dysmotility to be related to ILD [18, 33]. We did not examine dysmotility directly but observed that patients with distended oesophagus on CT were significantly more frequent in patients with ILD (data not shown). To our knowledge, this is the first MCTD study to show that the extent of ILD has an effect on all-cause mortality after adjustments to possible confounders in multivariable analyses. The strengths of this study were that the MCTD patients were derived from a large unselected nationwide cohort and that CTs were systematically analysed at baseline and at follow-up. The present study is the largest dataset of MCTD patients with systematically analysed CTs at two time points. The study has a longitudinal design with consistent data at baseline and at follow-up allowing us to assess a number of relevant parameters in univariable and multivariable analyses. The large number of attained baseline lung CT examinations is a strength of this study, yet a limitation of this study is that a follow-up CT was not accessible in all 135 patients. However there were no significant differences found between the patient groups with and without follow-up lung CT. A possible limitation of this study is the relatively low number of outcomes in the multivariable analyses. However, studies have shown that one can adjust for more variables per number of outcomes than previously assumed [34] and the associations found in the presenting study are strong and significant in both univariable and multivariable models. Due to limited number of events; and lack of data on cause of death, we did not have the possibility to assess whether any of the deaths were directly or indirectly related to ILD. In conclusion ILD in MCTD is common; it generally involves a modest proportion of TLV with basal distribution. Stable lung disease and slow disease progression is equally present. The risk factors of ILD progression are male gender, high anti-RNP antibody titre, presence of anti-ro-52 antibodies and no prior arthritis. ILD is associated with increased risk of mortality in our long-term observational study. Our findings can hopefully aid rheumatologists in identifying MCTD patients at risk of deteriorating lung disease and reduced survival. Acknowledgements The authors want to acknowledge Siri Opsahl Hetlevik, Vibke Lilleby and the Norwegian MCTD Study Group: Åse Stavland Lexberg, Department of Rheumatology, Vestre Viken; Alvilde Sofie Strand Dhainaut, Department of Rheumatology, St Olavs Hospital University Hospital; Liv-Turid Bertelsen, Department of Rheumatology, Haukeland University Hospital; Karen Irgens, Department of Rheumatology, Ålesund Hospital; Andrea Becker-Merok, Department of Rheumatology, University Hospital of North-Norway; Jan Leidulf Nordeide, Department of Rheumatology, Førde Central Hospital; Helle Bitter, Department of Rheumatology, Sørlandet Hospital; Sonja Pedersen, Department of Rheumatology, Nordland Hospital; Anne Prøven, Department of Rheumatology, Martina Hansens Hospital; and Sigrid Svalastoga, Department of Rheumatology, Østfold Hospital for enrolling patients and collecting data for the study. Funding: This work was supported by the state-funded University of Oslo, Norway. Disclosure statement: The authors have declared no conflicts of interest. Supplementary data Supplementary data are available at Rheumatology Online. References 1 Sharp GC, Irvin WS, Tan EM, Gould RG, Holman HR. Mixed connective tissue disease–an apparently distinct rheumatic disease syndrome associated with a specific antibody to an extractable nuclear antigen (ENA). Am J Med  1972; 52: 148– 59. Google Scholar CrossRef Search ADS PubMed  2 Tani C, Carli L, Vagnani S et al.   The diagnosis and classification of mixed connective tissue disease. J Autoimmun  2014; 48–49: 46– 9. Google Scholar CrossRef Search ADS PubMed  3 Cappelli S, Bellando Randone S, Martinovic D et al.   "To be or not to be," ten years after: evidence for mixed connective tissue disease as a distinct entity. Semin Arthritis Rheum  2012; 41: 589– 98. Google Scholar CrossRef Search ADS PubMed  4 Ciang NC, Pereira N, Isenberg DA. Mixed connective tissue disease—enigma variations? Rheumatology  2017; 56: 326– 33. Google Scholar PubMed  5 Flam ST, Gunnarsson R, Garen T et al.   The HLA profiles of mixed connective tissue disease differ distinctly from the profiles of clinically related connective tissue diseases. Rheumatology  2015; 54: 528– 35. Google Scholar CrossRef Search ADS PubMed  6 Black C, Isenberg DA. Mixed connective tissue disease—goodbye to all that. Br J Rheumatol  1992; 31: 695– 700. Google Scholar CrossRef Search ADS PubMed  7 Paradowska-Gorycka A, Jurkowska M, Czuszynska Z et al.   IL-10, IL-12B and IL-17 gene polymorphisms in patients with mixed connective tissue disease. Modern Rheumatol  2015; 25: 487– 9. Google Scholar CrossRef Search ADS   8 Dong RP, Kimura A, Hashimoto H et al.   Difference in HLA-linked genetic background between mixed connective tissue disease and systemic lupus erythematosus. Tissue Antigens  1993; 41: 20– 5. Google Scholar CrossRef Search ADS PubMed  9 Goh NS, Desai SR, Veeraraghavan S et al.   Interstitial lung disease in systemic sclerosis: a simple staging system. Am J Respir Crit Care Med  2008; 177: 1248– 54. Google Scholar CrossRef Search ADS PubMed  10 Komocsi A, Vorobcsuk A, Faludi R et al.   The impact of cardiopulmonary manifestations on the mortality of SSc: a systematic review and meta-analysis of observational studies. Rheumatology  2012; 51: 1027– 36. Google Scholar CrossRef Search ADS PubMed  11 Johnson C, Pinal-Fernandez I, Parikh R et al.   Assessment of mortality in autoimmune myositis with and without associated interstitial lung disease. Lung  2016; 194: 733– 7. Google Scholar CrossRef Search ADS PubMed  12 Hajas A, Szodoray P, Nakken B et al.   Clinical course, prognosis, and causes of death in mixed connective tissue disease. J Rheumatol  2013; 40: 1134– 42. Google Scholar CrossRef Search ADS PubMed  13 Fagundes MN, Caleiro MT, Navarro-Rodriguez T et al.   Esophageal involvement and interstitial lung disease in mixed connective tissue disease. Respir Med  2009; 103: 854– 60. Google Scholar CrossRef Search ADS PubMed  14 Gunnarsson R, Molberg O, Gilboe IM, Gran JT, Group PS. The prevalence and incidence of mixed connective tissue disease: a national multicentre survey of Norwegian patients. Ann Rheum Dis  2011; 70: 1047– 51. Google Scholar CrossRef Search ADS PubMed  15 Gunnarsson R, Aalokken TM, Molberg O et al.   Prevalence and severity of interstitial lung disease in mixed connective tissue disease: a nationwide, cross-sectional study. Ann Rheum Dis  2012; 71: 1966– 72. Google Scholar CrossRef Search ADS PubMed  16 Kozuka T, Johkoh T, Honda O et al.   Pulmonary involvement in mixed connective tissue disease: high-resolution CT findings in 41 patients. J Thorac Imaging  2001; 16: 94– 8. Google Scholar CrossRef Search ADS PubMed  17 Bodolay E, Szekanecz Z, Devenyi K et al.   Evaluation of interstitial lung disease in mixed connective tissue disease (MCTD). Rheumatology  2005; 44: 656– 61. Google Scholar CrossRef Search ADS PubMed  18 Kawano-Dourado L, Baldi BG, Kay FU et al.   Pulmonary involvement in long-term mixed connective tissue disease: functional trends and image findings after 10 years. Clin Exp Rheumatol  2015; 33: 234– 40. Google Scholar PubMed  19 Winstone TA, Assayag D, Wilcox PG et al.   Predictors of mortality and progression in scleroderma-associated interstitial lung disease: a systematic review. Chest  2014; 146: 422– 36. Google Scholar CrossRef Search ADS PubMed  20 Marie I. Morbidity and mortality in adult polymyositis and dermatomyositis. Curr Rheumatol Rep  2012; 14: 275– 85. Google Scholar CrossRef Search ADS PubMed  21 Yu KH, Wu YJ, Kuo CF et al.   Survival analysis of patients with dermatomyositis and polymyositis: analysis of 192 Chinese cases. Clin Rheumatol  2011; 30: 1595– 601. Google Scholar CrossRef Search ADS PubMed  22 Yamasaki Y, Yamada H, Ohkubo M et al.   Longterm survival and associated risk factors in patients with adult-onset idiopathic inflammatory myopathies and amyopathic dermatomyositis: experience in a single institute in Japan. J Rheumatol  2011; 38: 1636– 43. Google Scholar CrossRef Search ADS PubMed  23 Kasukawa RS, Gordon C (eds). Mixed Connective Tissue Disease and Anti-Nuclear Antibodies. Proceedings of the International Symposium on Mixed Connective Tissue Disease and Anti-nuclear Antibodies. Tokyo, 29–30 August 1986 . Amsterdam: Excerpta Medica, 1987. 24 Galie N, Hoeper MM, Humbert M et al.   Guidelines for the diagnosis and treatment of pulmonary hypertension: the Task Force for the Diagnosis and Treatment of Pulmonary Hypertension of the European Society of Cardiology (ESC) and the European Respiratory Society (ERS), endorsed by the International Society of Heart and Lung Transplantation (ISHLT). Eur Heart J  2009; 30: 2493– 537. Google Scholar CrossRef Search ADS PubMed  25 Hansell DM, Bankier AA, MacMahon H et al.   Fleischner Society: glossary of terms for thoracic imaging. Radiology  2008; 246: 697– 722. Google Scholar CrossRef Search ADS PubMed  26 Galvin JR, Frazier AA, Franks TJ. Collaborative radiologic and histopathologic assessment of fibrotic lung disease. Radiology  2010; 255: 692– 706. Google Scholar CrossRef Search ADS PubMed  27 Miller MR, Hankinson J, Brusasco V et al.   Standardisation of spirometry. Eur Respir J  2005; 26: 319– 38. Google Scholar CrossRef Search ADS PubMed  28 Textor J, Hardt J, Knuppel S. DAGitty: a graphical tool for analyzing causal diagrams. Epidemiology  2011; 22: 745. Google Scholar CrossRef Search ADS PubMed  29 Greidinger EL, Zang Y, Jaimes K et al.   A murine model of mixed connective tissue disease induced with U1 small nuclear RNP autoantigen. Arthritis Rheum  2006; 54: 661– 9. Google Scholar CrossRef Search ADS PubMed  30 Hetlevik SO, Flato B, Rygg M et al.   Long-term outcome in juvenile-onset mixed connective tissue disease: a nationwide Norwegian study. Ann Rheum Dis  2016; 76: 159– 65. Google Scholar CrossRef Search ADS PubMed  31 Burdt MA, Hoffman RW, Deutscher SL et al.   Long-term outcome in mixed connective tissue disease: longitudinal clinical and serologic findings. Arthritis Rheum  1999; 42: 899– 909. Google Scholar CrossRef Search ADS PubMed  32 Gunnarsson R, El-Hage F, Aalokken TM et al.   Associations between anti-Ro52 antibodies and lung fibrosis in mixed connective tissue disease. Rheumatology  2016; 55: 103– 8. Google Scholar CrossRef Search ADS PubMed  33 Szodoray P, Hajas A, Kardos L et al.   Distinct phenotypes in mixed connective tissue disease: subgroups and survival. Lupus  2012; 21: 1412– 22. Google Scholar CrossRef Search ADS PubMed  34 Vittinghoff E, McCulloch CE. Relaxing the rule of ten events per variable in logistic and Cox regression. Am J Epidemiol  2007; 165: 710– 8. Google Scholar CrossRef Search ADS PubMed  © The Author 2017. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Rheumatology Oxford University Press

Progression and mortality of interstitial lung disease in mixed connective tissue disease: a long-term observational nationwide cohort study

Loading next page...
 
/lp/ou_press/progression-and-mortality-of-interstitial-lung-disease-in-mixed-wKsp8LTrMH
Publisher
Oxford University Press
Copyright
© The Author 2017. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: journals.permissions@oup.com
ISSN
1462-0324
eISSN
1462-0332
D.O.I.
10.1093/rheumatology/kex077
Publisher site
See Article on Publisher Site

Abstract

Abstract Objectives To assess the prevalence, extent, progression, functional impact and mortality of interstitial lung disease (ILD) in a nationwide unselected MCTD cohort. Methods The study cohort included patients with high-resolution CT lung scans available at baseline (n = 135) and at follow-up (n = 119). The extent of disease was expressed as percentage of total lung volume (TLV). Results ILD was present in 41% of MCTD patients at follow-up. Median (interquartile) extent (% of TLV) was 5 (8) at baseline and 7 (17) at follow-up, mean length 6.4 years later. The lung disease progressed in 19% of patients across the observation period. Predictors of ILD progression were elevated anti-RNP titre [hazard ratio (HR) 1.5, 95% CI: 1.1, 2.0; P = 0.008], presence of anti-ro52 antibodies (HR = 3.5, 95% CI: 1.2, 10.2; P = 0.023), absence of arthritis (HR = 0.2, 95% CI: 0.1, 0.6; P = 0.004) and male gender (HR = 4.0, 95% CI: 1.4, 11.5; P = 0.011) after age and baseline disease adjustments. The risk of death increased by 2.9 (95% CI: 1.1, 7.9; P = 0.038) in patients where disease involved ⩾5% of TLV. Conclusion Lung disease extent and progression in MCTD are modest. Yet, the extension continues several years after MCTD diagnosis causing lung function decline and increasing the risk of mortality. The study identified male gender, elevated anti-RNP titre, presence of anti-ro52 antibodies and absence of arthritis as the strongest predictors of ILD progression. mixed connective tissue disease, interstitial lung disease, pulmonary fibrosis, mortality, anti-ribonucleoprotein antibodies Rheumatology key messages Interstitial lung disease is typically modest with slow progression and stable disease equally present. Progression of lung disease was evident in 1/5 of patients resulting in lung function deterioration. Risk factors for lung disease were male gender, high anti-RNP titres, anti-ro52 antibodies and absence of arthritis. Introduction MCTD is a chronic, immune-mediated disorder defined by the combined presence of serum anti-RNP antibodies and selected clinical features of SSc, SLE, RA and PM/DM. The entity was first described by Sharp et al. in 1972 [1, 2]. Since then it has been debated whether MCTD is a distinct disease, an overlap syndrome or a transient phase before developing into another CTD [3, 4]. Genetic studies from Europe [5–7] and Asia [8] support the notion that MCTD is a distinct disease. Interstitial lung disease (ILD) is a major cause of morbidity and mortality in the CTDs, particularly in SSc [9, 10] and PM/DM [11]. In MCTD prevalence figures from different cohorts have ranged from 47 to 78% [12, 13]. Previously, our research group investigated the prevalence and incidence of MCTD between the years of 2005 and 2008. One hundred and forty-seven patients were included in the nationwide MCTD cohort [14]. Analyses of lung high-resolution CT at study inclusion found that reticular pattern was generally present in the basal zones and the most common abnormality, resembling the pattern seen in SSc [15]. Studies from other countries have reported different CT findings. In a study on 41 MCTD patients from Japan, interlobular septal thickening was the predominant pattern [16], while a Hungarian study reported that ground glass attenuation was highly frequent [17]. In a recent report from Brazil the most common patterns were ground glass and reticular changes [18]. Histological descriptions of MCTD-related lung pathologies are scarce. Hajas et al. [12] reviewed lung biopsies from 16 MCTD patients where the ILD diagnosis could not be resolved by imaging alone. All these 16 patients had findings compatible with non-specific interstitial pneumonitis, where 11 had cellular interstitial pneumonitis with little or no fibrosis and the remaining 5 had significant fibrosis. Data on the course and outcome of ILD in MCTD are limited. Bodolay et al. [17] reported mild fibrosis in the majority of MCTD patients, but did not present how this was evaluated. Kawano-Dourado et al. [18] recently presented serial CT analyses in 37 MCTD patients followed at a tertiary centre, and showed that the median percentage ILD score in the lower lobes increased from 7.5 to 11.3% over 10 years, indicating restricted disease progression. There are no larger longitudinal studies on the overall progression of ILD in unselected MCTD cohorts, and outcome risk factors are needed. Predictive factors for progressive ILD have been studied in SSc [19] but remain unknown in MCTD. ILD is known to predict mortality in SSc [19] and PM/DM [11, 20–22] while the effect of ILD on mortality has not been established in MCTD cohorts. Here, we present a long-term observational study of the unselected, nationwide Norwegian MCTD cohort, with serial CT and pulmonary function test analyses. The aim of this study was to examine the prevalence, extent, progression and functional effect of ILD in MCTD. We also wanted to identify predictive factors of ILD progression in MCTD, aiming to aid the clinician in identifying MCTD patients at risk of deteriorating lung function. Finally, we wanted to estimate the effect of ILD on mortality in MCTD. Methods Patient population The Norwegian nationwide MCTD cohort recruited patients from Departments of Rheumatology in Norway during the years of 2005–08 [15] with the following inclusion criteria: age above 18 years; fulfillment of at least one of the three criteria sets of MCTD, the modified Sharp’s criteria [1], the criteria of Alarcón-Segovia and/or Kasukawa [23]; and exclusion of another CTD. In the present study we included all patients from the nationwide MCTD cohort where baseline CTs were attainable for assessment (n = 135), encompassing all 126 patients included in the evaluation of ILD prevalence and severity from 2012 [15] as well as nine additional patients from the nationwide MCTD cohort. Clinical data Clinical parameters included symptoms and signs included in the three classification criteria as well as smoking history, medications and vital status. The clinical parameters were recorded at two time points: baseline and follow-up. Baseline was defined as study entry. Myositis was confirmed by muscle biopsy and/or electromyogram and CK elevation. Smokers had smoked >1 cigarette a day for >1 year. Precapillary pulmonary hypertension was defined according to the updated European Society of Cardiology criteria by mean pulmonary artery pressure ⩾25 mmHg and pulmonary wedge pressure ⩽15 mmHg by right-sided heart catheterization at rest [24]. Vital status at the end of the study was obtained from the National Population Register of Norway. Acquisition and analysis of CT images CT lung examination was included in the study protocol at baseline and at follow-up. High-resolution CT images were obtained with the patient in the supine position, during breath-holding and inspiration. The images were reconstructed at 1–1.25 mm section thickness with 10-mm intervals and at 2.5-mm section thickness reformatted in the axial, coronal and sagittal planes. Images were reviewed on a picture archiving and communication system screen in consensus and in random order by two specialist chest radiologists. The observers evaluated the presence and extent of ILD. These findings included reticular pattern, ground glass attenuation, traction bronchiectasis, interlobular septal thickening and airspace consolidations according to the CT criteria of ILD recommended by The Nomenclature Committee of the Fleischner Society [25]. Although the term honeycombing is common in radiological studies, the definition is not standardized [26]. Reticular pattern (i.e. the coarseness of fibrosis) was therefore graded as follows: grade 1, fine intralobular reticular pattern without evident cysts; grade 2, predominantly micro-cystic reticular pattern involving air spaces ⩽4 mm in diameter; and grade 3, a predominantly macro-cystic reticular pattern with air spaces >4 mm in diameter. If ground glass attenuations were superimposed on a reticular pattern, the abnormality was recorded as being reticular. When a grade 1 pattern was superimposed on a grade 2 pattern, the abnormality was recorded as grade 2. If different types of reticular patterns were present in different lung zones of an individual patient, both pattern types were recorded. Abnormal CT findings of ground glass attenuation and reticular patterns grades 1–3 were defined as ILD. The distribution of disease was reviewed in four lung zones: zone 1, above the aortic arch; zone 2, between the aortic arch and at the level of the carina; zone 3, between the level of the carina and the level of the inferior pulmonary veins; and zone 4, below 1 cm above the top of the right hemidiaphragm. The extent of involvement was evaluated independently for each zone and each zone was assigned a percentage of the lung parenchyma that showed evidence of ILD. The total extent of disease in each patient was added in the four lung zones and expressed as a percentage of total lung volume (TLV). ILD progression was defined as a disease extension as a percentage of TLV and stable ILD was defined as no disease extension as a percentage of TLV between baseline and follow-up. Minimal time between the baseline and follow-up CT was 12 months. Follow-up CT scans were available in 119 of 135 patients. Clinical information about the 16 patients where follow-up lung CT examination was not available can be found in supplementary Table S1, available at Rheumatology Online. The CT scans were subjectively assessed for possible oesophageal motility disorders. Substantial distension was recorded when columns of air and distension larger than doubling the oesophageal wall thickness was identified. Pulmonary function test analyses The pulmonary function test analyses were performed within 8 weeks of the corresponding CT and performed according to the American Thoracic Society/European Respiratory Society guidelines [27], using an automated Vmax V6200 system (SensorMedics, VIASYS Respiratory Care Inc, Yorba Linda, CA, USA). Recorded variables were the percentage of predicted value of full vital capacity (FVC% pred) and the percentage of predicted value of diffusing capacity of the lung for carbon monoxide (DLCO% pred). Autoantibody analyses Serum levels of anti-RNP autoantibody at baseline and follow-up in addition to anti-ro52 antibodies at baseline were determined by line immunoassay (ANA Profile 5 Euroline Blot test kit, Euroimmun, Lübeck, Germany). Strips were scanned to obtain optical density values according to the kit instructions. Anti-ro52 antibodies optical density values of ⩾10 were recorded as positive. Statistical analyses The data were presented as means (s.d.) or median and interquartile range (IQR) depending on the level of resemblance to the normal distribution. Comparisons of groups were performed appropriately by chi-square test, Fisher’s exact test, or one-way ANOVA with post hoc comparisons by Tukey’s test or the Kruskal–Wallis test and Mann–Whitney U test. Cox regression analyses were used to find predictive factors of lung disease progression. Using a manual backward elimination procedure, variables at a significance level of P < 0.25 in the univariable analyses were considered a candidate for the multivariable model. The prediction was quantified by the hazard ratio (HR) with its 95% CI. Patients were followed from the date of inclusion in the cohort until death or end of follow-up on 30th of April 2016. Kaplan–Meier survival curves were used to examine differences in survival between patients with and without disease compromising ⩾5% and ⩾10% of TLV. The log-rank test was used to assess differences. The proportional hazard assumptions were tested by plotting the logarithm of the integrated hazards (log–log survival plots). To estimate the total effect of ILD on mortality with minimal bias, we performed multivariable cox regression analyses adjusting for variables selected from the directed acyclic graph presented in supplementary Fig. S1, available at Rheumatology Online [28]. Data extraction and analyses were conducted using SPSS Statistics version 22 (IBM Corp., Armonk, NY, USA) and STATA version 22 (StataCorp, College Station, TX, USA). Ethical approval The study was approved by the Norwegian Regional Committee for Medical and Health Research Ethics and conducted in accordance with the guidelines of the Helsinki declaration. All patients had given informed written consent to participate in the study. Results Prevalence, extent and distribution Serial lung CT images available for evaluation were retrieved in 119 of 135 patients. There were no significant differences found in terms of gender, age at diagnosis, disease duration at baseline or classification criteria when comparing these 119 patients against the 16 patients with missing follow-up. Mean (s.d.) age at diagnosis in the 135 patients with baseline CT data was 35 (16) years and mean (s.d.) disease duration was 9 (8) years. The Alarcón-Segovia criteria were met by 90%, 97% fulfilled the Sharp criteria and 86% fulfilled the Kasukawa criteria. The mean (s.d.) time between baseline and follow-up CT was 6.4 (2.0) years. CT findings compatible with ILD were present at baseline in 40% and at follow-up in 41% of the patients (Table 1). There were only two new cases of ILD during the study observation. Median (IQR) extent of lung disease increased from 5 (8)% of TLV at baseline to 7 (17)% of TLV at follow-up (Table 1). The extension was generally slow, apparent by the number of patients with disease involving ⩾5, 10 and 20% of TLV at follow-up compared with baseline (Table 1). Distribution analyses demonstrated lung disease to be present largely in the lower lung lobes, but about half of the patients with lung disease did have findings of ILD in the upper part of the lungs (i.e. in zone 1: see ‘Methods’ section for details) (Table 1). Fine reticular pattern (grade 1) was the most common abnormality; however, macro-cystic reticular pattern (grade 3) covered a larger area of the lungs in the patients where this pattern was evident (Table 1). Table 1 Detailed description of lung CT findings at baseline and follow-up in MCTD patients (n = 119) Lung CT findings  Baseline  Follow-up  Age at CT examination, mean (s.d.), years  44.3 (15)  51.1 (14)  Disease duration at time of CT, mean (s.d.), years  9.5 (8)  16.5 (8)  Prevalence of interstitial lung disease, n (%)  47 (40)  49 (41)  Extent          Median (IQR) % of TLV  5 (2–10)  7 (2–19)      ≥5% of TLV, n (%)  24 (20)  30 (25)      ≥10% of TLV, n (%)  13 (11)  20 (17)      ≥20% of TLV, n (%)  7 (6)  12 (10)  Distribution          Zone 1, n (%)  25 (21)  28 (24)      Median (IQR) % of TLV  5 (2–10)  8 (3–10)      Zone 2, n (%)  27 (23)  32 (27)      Median (IQR) % of TLV  5 (3–15)  10 (5–19)      Zone 3, n (%)  40 (34)  42 (35)      Median (IQR) % of TLV  10 (3–15)  10 (5–23)      Zone 4, n (%)  46 (39)  47 (39)      Median (IQR) % of TLV  10 (5–21)  15 (5–45)  Patterns          Ground glass attenuation, n (%)  3 (3)  4 (3)      Median (IQR) % of TLV  13 (11–16)  11 (9–48)      Grade 1, n (%)  31 (26)  33 (28)      Median (IQR) % of TLV  3 (2–4)  3 (1–5)      Grade 2, n (%)  19 (16)  22 (18)      Median (IQR) % of TLV  5 (3–10)  5 (3–11)      Grade 3, n (%)  9 (8)  13 (11)      Median (IQR) % of TLV  14 (8–27)  13 (5–39)  Other CT changes          Traction bronchiectasis, n (% of patients with ILD)  14 (30)  20 (41)      Emphysema, n (% of patients with ILD)  11 (15)  11 (16)      Dilated oesophagus, n (% of patients with ILD)  43 (57)  52 (65)  Lung CT findings  Baseline  Follow-up  Age at CT examination, mean (s.d.), years  44.3 (15)  51.1 (14)  Disease duration at time of CT, mean (s.d.), years  9.5 (8)  16.5 (8)  Prevalence of interstitial lung disease, n (%)  47 (40)  49 (41)  Extent          Median (IQR) % of TLV  5 (2–10)  7 (2–19)      ≥5% of TLV, n (%)  24 (20)  30 (25)      ≥10% of TLV, n (%)  13 (11)  20 (17)      ≥20% of TLV, n (%)  7 (6)  12 (10)  Distribution          Zone 1, n (%)  25 (21)  28 (24)      Median (IQR) % of TLV  5 (2–10)  8 (3–10)      Zone 2, n (%)  27 (23)  32 (27)      Median (IQR) % of TLV  5 (3–15)  10 (5–19)      Zone 3, n (%)  40 (34)  42 (35)      Median (IQR) % of TLV  10 (3–15)  10 (5–23)      Zone 4, n (%)  46 (39)  47 (39)      Median (IQR) % of TLV  10 (5–21)  15 (5–45)  Patterns          Ground glass attenuation, n (%)  3 (3)  4 (3)      Median (IQR) % of TLV  13 (11–16)  11 (9–48)      Grade 1, n (%)  31 (26)  33 (28)      Median (IQR) % of TLV  3 (2–4)  3 (1–5)      Grade 2, n (%)  19 (16)  22 (18)      Median (IQR) % of TLV  5 (3–10)  5 (3–11)      Grade 3, n (%)  9 (8)  13 (11)      Median (IQR) % of TLV  14 (8–27)  13 (5–39)  Other CT changes          Traction bronchiectasis, n (% of patients with ILD)  14 (30)  20 (41)      Emphysema, n (% of patients with ILD)  11 (15)  11 (16)      Dilated oesophagus, n (% of patients with ILD)  43 (57)  52 (65)  TLV: total lung volume. Progression Patients were stratified in three groups: group 1, no ILD (n = 70); group 2, stable ILD (n = 26); and group 3, ILD progression (n = 23) (Table 2). The median (IQR) annual disease progression in group 3 was 1.1 (2)% of TLV. Compared with group 1 fewer patients had arthritis and anti-RNP titres were higher at baseline and at follow-up in group 3 (Table 2). Frequency of cases meeting the Alarcón-Segovia criteria was lowest in group 3. Group 2 (stable ILD) differed from the two other groups by a higher occurrence of patients with dilated oesophagus. Table 2 Characteristics of MCTD patients stratified by lung CT results Characteristics  Group 1  Group 2  Group 3  P-value    No ILD  Stable ILD  ILD progression      n = 70  n = 26  n = 23    Demographics              Time baseline to follow-up CT, mean (s.d.), years  6.5 (2)  6.2 (3)  6.0 (2)  NS      Disease duration at follow-up, mean (s.d.), years  16.7 (8)  17.0 (9)  16.3 (10)  NS      Male gender, n (%)  16 (23)  5 (13)  9 (39)  NS      Age at diagnosis, mean (s.d.), years  31.4 (15)  38.1 (18)  38.4 (18)  NS      Age at baseline, mean (s.d.), years  41.6 (14)  48.6 (13)  47.7 (17)  NS      Age at follow-up, mean (s.d.), years  49.0 (14)  55.7 (14)  54.0 (16)  NS      Smokers, n (%)  37 (54)  14 (54)  10 (44)  NS      Alarcon criteria, n (%)  66 (94)  23 (89)  17 (74)  0.025a      Kasukawa criteria, n (%)  58 (83)  26 (100)  22 (96)  0.030a      Precapilary pulmonary hypertension  1 (1)  2 (8)  2 (9)  NA  CT features              Extent of ILD at baseline, median (IQR), % of TLV  NA  3.5 (2–8)  8.8 (3–22)  NS      New cases, n (incidence proportion %)  NA  NA  2 (3)  —      Extent of ILD at follow-up, median (IQR), % of TLV  NA  3.5 (2–8)  17.5 (6–43)  <0.001b      Annual progression rate, median (IQR), %  NA  NA  1.1 (0.5–2.5)  —  SLE-like features cumulative at baseline              Arthritis, n (%)  63 (90)  20 (77)  12 (52)  <0.001a      Pericarditis, n (%)  7 (10)  6 (23)  2 (9)  NS  SSc-like features at baseline              Sclerodactily, n (%)  19 (27)  13 (50)  8 (35)  NS      Dilated oesophagus on CT, n (%)  15 (22)  20 (77)  8 (35)  <0.001a  PM-like features cumulative at baseline              Myositis, n (%)  13 (19)  4 (15)  5 (22)  NS  Laboratory features              Anti-RNP at baseline, median (IQR), ×10−3 U/l  65 (16–159)  123 (49–240)  240 (59–240)  0.004c      Anti-RNP at follow-up, median (IQR), ×10−3 U/l  30 (6–92)  150 (29–240)  194 (12–240)  0.005c      ESR at baseline, mean (s.d.), mm  18 (14)  22 (18)  29 (18)  0.012d      Hb at baseline, mean (s.d.), g/dl  13.3 (1.1)  13.0 (1.2)  13.3 (1.6)  NS      Anti-ro-52 positivity at baseline, n (%)  12 (19)  8 (32)  9 (39)  NS  Characteristics  Group 1  Group 2  Group 3  P-value    No ILD  Stable ILD  ILD progression      n = 70  n = 26  n = 23    Demographics              Time baseline to follow-up CT, mean (s.d.), years  6.5 (2)  6.2 (3)  6.0 (2)  NS      Disease duration at follow-up, mean (s.d.), years  16.7 (8)  17.0 (9)  16.3 (10)  NS      Male gender, n (%)  16 (23)  5 (13)  9 (39)  NS      Age at diagnosis, mean (s.d.), years  31.4 (15)  38.1 (18)  38.4 (18)  NS      Age at baseline, mean (s.d.), years  41.6 (14)  48.6 (13)  47.7 (17)  NS      Age at follow-up, mean (s.d.), years  49.0 (14)  55.7 (14)  54.0 (16)  NS      Smokers, n (%)  37 (54)  14 (54)  10 (44)  NS      Alarcon criteria, n (%)  66 (94)  23 (89)  17 (74)  0.025a      Kasukawa criteria, n (%)  58 (83)  26 (100)  22 (96)  0.030a      Precapilary pulmonary hypertension  1 (1)  2 (8)  2 (9)  NA  CT features              Extent of ILD at baseline, median (IQR), % of TLV  NA  3.5 (2–8)  8.8 (3–22)  NS      New cases, n (incidence proportion %)  NA  NA  2 (3)  —      Extent of ILD at follow-up, median (IQR), % of TLV  NA  3.5 (2–8)  17.5 (6–43)  <0.001b      Annual progression rate, median (IQR), %  NA  NA  1.1 (0.5–2.5)  —  SLE-like features cumulative at baseline              Arthritis, n (%)  63 (90)  20 (77)  12 (52)  <0.001a      Pericarditis, n (%)  7 (10)  6 (23)  2 (9)  NS  SSc-like features at baseline              Sclerodactily, n (%)  19 (27)  13 (50)  8 (35)  NS      Dilated oesophagus on CT, n (%)  15 (22)  20 (77)  8 (35)  <0.001a  PM-like features cumulative at baseline              Myositis, n (%)  13 (19)  4 (15)  5 (22)  NS  Laboratory features              Anti-RNP at baseline, median (IQR), ×10−3 U/l  65 (16–159)  123 (49–240)  240 (59–240)  0.004c      Anti-RNP at follow-up, median (IQR), ×10−3 U/l  30 (6–92)  150 (29–240)  194 (12–240)  0.005c      ESR at baseline, mean (s.d.), mm  18 (14)  22 (18)  29 (18)  0.012d      Hb at baseline, mean (s.d.), g/dl  13.3 (1.1)  13.0 (1.2)  13.3 (1.6)  NS      Anti-ro-52 positivity at baseline, n (%)  12 (19)  8 (32)  9 (39)  NS  aχ2 test. bGroup 2 vs group 3, Kruskal–Wallis and Mann–Whitney U test. cGroup 1 vs group 3, Kruskal–Wallis and Mann–Whitney U test. dGroup 3 vs group 1, one-way ANOVA and post hoc Tukey’s test. ILD: interstitial lung disease; NA: not applicable; NS: not significant. Functional impact The extent of ILD at baseline CTs was greater in group 3 than group 2, although not significantly (Table 2). Figure 1 illustrates the effect of ILD on mean lung function test results. FVC% pred differed significantly in group 3 compared with group 2 (P = 0.002) and group 1 (P < 0.001) at baseline and at follow-up (both P < 0.001). Group 3 differed significantly in lower mean values of DLCO% pred compared with group 1 (P < 0.001) but not group 2 at baseline, while at follow-up DLCO% pred was significantly lower in group 3 compared with both group 1 (P < 0.001) and group 2 (P = 0.001). FVC% pred in group 2 did not differ significantly from group 1 at baseline or at follow-up. DLCO% pred in group 2 differed significantly from group 1 at baseline (P = 0.003) but not at follow-up. Fig. 1 View largeDownload slide Pulmonary function test results stratified by lung CT results Fig. 1 View largeDownload slide Pulmonary function test results stratified by lung CT results Treatment Analyses of immune-modulating treatments did not indicate clear differences between groups 1–3, except for higher accumulated frequencies of AZA usage in patients with stable ILD (group 2) (supplementary Table S2, available at Rheumatology Online). Predictors of progression Multiple Cox regression analyses revealed that the strongest predictive factors at baseline for ILD progression at follow-up were male gender, elevated anti-RNP titre, the presence of anti-ro-52 antibodies and absence of arthritis when adjusting for baseline lung disease and age (Table 3). A rise of 50 U/ml in anti-RNP titre resulted in 50% increased risk of ILD progression (95% CI: 1.1, 2.0; P = 0.008), while a history of arthritis reduced the risk by 80% (95% CI: 0.1, 0.6; P = 0.004). The risk of ILD progression was 4-fold in male patients (95% Cl: 1.4, 11.5; P = 0.011) and >3-fold in patients with anti-ro52 antibodies (95% Cl: 1.2, 10.2; P = 0.023). According to Harrell’s C index, patient outcomes were accurately predicted by this model 91% of the time. Table 3 Predictors of interstitial lung disease progression in MCTD analysed by cox regression   Univariable  Multivariable  Clinical predictors  HR (95% CI)  P-value  HR (95% CI)  P-value  Alarcon criteria  0.26 (0.10, 0.67)  0.005      FVC % of predicted  0.95 (0.93, 0.98)  <0.001      DLCO % of predicted  0.94 (0.92, 0.97)  <0.001      ESR (per 10 mm increase)  1.5 (1.2, 1.9)  <0.001      % ILD of TLV  1.1 (1.1, 1.1)  <0.001  1.1 (1.0, 1.1)  0.002  Arthritis  0.15 (0.06, 0.36)  <0.001  0.22 (0.08, 0.61)  0.004  Anti-RNP value (per 50 U/l increase)  1.4 (1.1, 1.8)  0.009  1.5 (1.1, 2.0)  0.008  Age at baseline  1.0 (1.0, 1.1)  0.065  1.1 (1.0, 1.1)  0.028  Male gender  2.0 (0.9, 4.6)  0.112  4.0 (1.4, 11.5)  0.011  Anti-ro-52 positivity  3.2 (1.3, 7.7)  0.01  3.5 (1.2, 10.2)  0.023    Univariable  Multivariable  Clinical predictors  HR (95% CI)  P-value  HR (95% CI)  P-value  Alarcon criteria  0.26 (0.10, 0.67)  0.005      FVC % of predicted  0.95 (0.93, 0.98)  <0.001      DLCO % of predicted  0.94 (0.92, 0.97)  <0.001      ESR (per 10 mm increase)  1.5 (1.2, 1.9)  <0.001      % ILD of TLV  1.1 (1.1, 1.1)  <0.001  1.1 (1.0, 1.1)  0.002  Arthritis  0.15 (0.06, 0.36)  <0.001  0.22 (0.08, 0.61)  0.004  Anti-RNP value (per 50 U/l increase)  1.4 (1.1, 1.8)  0.009  1.5 (1.1, 2.0)  0.008  Age at baseline  1.0 (1.0, 1.1)  0.065  1.1 (1.0, 1.1)  0.028  Male gender  2.0 (0.9, 4.6)  0.112  4.0 (1.4, 11.5)  0.011  Anti-ro-52 positivity  3.2 (1.3, 7.7)  0.01  3.5 (1.2, 10.2)  0.023  DLCO: diffusing capacity of the lung for carbon monoxide; FVC: forced vital capacity; HR: hazard ratio; ILD: interstitial lung disease; TLV: total lung volume. Effect on all-cause mortality The effect of ILD on mortality was evaluated by dividing the patient cohort in two subsets: disease extent <5% (n = 108), and ⩾5% of TLV (n = 27) and <10% (n = 122) and ⩾10% (n = 13) of TLV. The 5-year cumulative survival rates for <5% and ⩾5% disease of TLV were 94% (95% CI: 88, 98%) and 82% (95% CI: 61, 92%), respectively. The 10-year cumulative survival rates were 87% (95% CI: 79, 92%) and 70% (95% CI: 49, 84%; log rank P = 0.015) (Fig. 2). The 5-year cumulative survival rates for <10% and ⩾10% disease of TLV were 93% (95% CI: 85, 95%) and 77% (95% CI: 44, 92%), respectively, and the 10-year cumulative survival rates were 86% (95% CI: 78, 91%) and 60% (95% CI: 29, 81%), respectively (log rank P = 0.01). Cox regression univariable analyses revealed that having disease extending ⩾5% of TLV was associated with 3-fold increased risk of death (HR: 2.9, 95% CI: 1.2, 7.0; P = 0.020), while having disease involving 10% of TLV increased the risk by almost fourfold (HR: 3.5, 95% CI: 1.3, 9.5; P = 0.015) (Table 4). A disease extension of 1% of TLV was associated with increased risk of death (HR: 1.1, 95% CI: 1.0, 1.1; P < 0.001) (Table 4). The total effect of % ILD extent on all-cause mortality (see supplementary Fig. S1, available at Rheumatology Online) was similar after adjustments to smoking, gender, disease duration and age (Table 4). Table 4 Effect of interstitial lung disease on all-cause mortality in the MCTD cohort Interstitial lung disease  Univariable    Multivariablea      HR (95% CI)  P-value  HR (95% CI)  P-value  Model 1: ≥5% of TLV  2.8 (1.6, 2.8)  0.020  2.9 (1.1, 7.9)  0.038  Model 2: ≥10% of TLV  3.5 (1.3, 9.5)  0.015  3.4 (1.1, 10.2)  0.028  Model 3: per 1% of TLV increase  1.1 (1.0, 1.1)  <0.001  1.1 (1.0, 1.1)  <0.001  Interstitial lung disease  Univariable    Multivariablea      HR (95% CI)  P-value  HR (95% CI)  P-value  Model 1: ≥5% of TLV  2.8 (1.6, 2.8)  0.020  2.9 (1.1, 7.9)  0.038  Model 2: ≥10% of TLV  3.5 (1.3, 9.5)  0.015  3.4 (1.1, 10.2)  0.028  Model 3: per 1% of TLV increase  1.1 (1.0, 1.1)  <0.001  1.1 (1.0, 1.1)  <0.001  Univariable and multivariable cox regression analyses. a Vriables included in the multivariable models: age at baseline, smoking, disease duration at baseline and gender. TLV: total lung volume. Fig. 2 View largeDownload slide Kaplan–Meier curves for interstitial lung disease involving ≥5% and <5% of total lung capacity Fig. 2 View largeDownload slide Kaplan–Meier curves for interstitial lung disease involving ≥5% and <5% of total lung capacity Discussion ILD is frequently reported in MCTD cohorts and studies of the long-term clinical impact of these changes are necessary to assist rheumatologists in the management of MCTD. Here, we found that 19% of patients in our nationwide, unselected MCTD cohort had ILD progression on CT, where the mean (s.d.) disease duration was 17 (8) years. The annual progression is modest, but results in deteriorating lung function in this relatively young patient population. Based on the presenting, high resolution, long-term observational study in addition to comparable work on smaller patient populations [17, 18] and cross-sectional studies [2], it appears reasonable to draw the following conclusions: interstitial lung changes occur frequently in MCTD; ILD extent in MCTD is typically modest; approximately half of the patients with established lung disease have progression; the progression is generally slow, but appears to continue several years after diagnosis causing deteriorating lung function; a subset of patients, in our study 10%, had developed extensive ILD (>20% of TLV); and, finally, ILD is associated with increased risk of mortality, and should therefore be regarded as a major complication in MCTD. Interestingly we found that increasing anti-RNP antibody titres at baseline was a strong predictor of ILD progression. Greidinger et al. [29] found anti-70K antibodies to be strongly correlated with lung disease in 70K immunized mice. Notably, in a recent study on juvenile MCTD our colleagues found increased anti-RNP antibody titres in patients with active disease [30] corresponding to previous findings of Burdt et al. [31]. The presence of anti-ro-52 antibodies was recently found to be associated with ILD in our MCTD cohort [32]. We here present that the presence of anti-ro-52 antibodies is additionally a predictor of ILD progression in MCTD. Furthermore, we found arthritis to be protective of lung disease progression, possibly indicating that MCTD patients with more SLE-like features are less likely to develop severe lung disease. Previous studies of MCTD have found oesophageal dysmotility to be related to ILD [18, 33]. We did not examine dysmotility directly but observed that patients with distended oesophagus on CT were significantly more frequent in patients with ILD (data not shown). To our knowledge, this is the first MCTD study to show that the extent of ILD has an effect on all-cause mortality after adjustments to possible confounders in multivariable analyses. The strengths of this study were that the MCTD patients were derived from a large unselected nationwide cohort and that CTs were systematically analysed at baseline and at follow-up. The present study is the largest dataset of MCTD patients with systematically analysed CTs at two time points. The study has a longitudinal design with consistent data at baseline and at follow-up allowing us to assess a number of relevant parameters in univariable and multivariable analyses. The large number of attained baseline lung CT examinations is a strength of this study, yet a limitation of this study is that a follow-up CT was not accessible in all 135 patients. However there were no significant differences found between the patient groups with and without follow-up lung CT. A possible limitation of this study is the relatively low number of outcomes in the multivariable analyses. However, studies have shown that one can adjust for more variables per number of outcomes than previously assumed [34] and the associations found in the presenting study are strong and significant in both univariable and multivariable models. Due to limited number of events; and lack of data on cause of death, we did not have the possibility to assess whether any of the deaths were directly or indirectly related to ILD. In conclusion ILD in MCTD is common; it generally involves a modest proportion of TLV with basal distribution. Stable lung disease and slow disease progression is equally present. The risk factors of ILD progression are male gender, high anti-RNP antibody titre, presence of anti-ro-52 antibodies and no prior arthritis. ILD is associated with increased risk of mortality in our long-term observational study. Our findings can hopefully aid rheumatologists in identifying MCTD patients at risk of deteriorating lung disease and reduced survival. Acknowledgements The authors want to acknowledge Siri Opsahl Hetlevik, Vibke Lilleby and the Norwegian MCTD Study Group: Åse Stavland Lexberg, Department of Rheumatology, Vestre Viken; Alvilde Sofie Strand Dhainaut, Department of Rheumatology, St Olavs Hospital University Hospital; Liv-Turid Bertelsen, Department of Rheumatology, Haukeland University Hospital; Karen Irgens, Department of Rheumatology, Ålesund Hospital; Andrea Becker-Merok, Department of Rheumatology, University Hospital of North-Norway; Jan Leidulf Nordeide, Department of Rheumatology, Førde Central Hospital; Helle Bitter, Department of Rheumatology, Sørlandet Hospital; Sonja Pedersen, Department of Rheumatology, Nordland Hospital; Anne Prøven, Department of Rheumatology, Martina Hansens Hospital; and Sigrid Svalastoga, Department of Rheumatology, Østfold Hospital for enrolling patients and collecting data for the study. Funding: This work was supported by the state-funded University of Oslo, Norway. Disclosure statement: The authors have declared no conflicts of interest. Supplementary data Supplementary data are available at Rheumatology Online. References 1 Sharp GC, Irvin WS, Tan EM, Gould RG, Holman HR. Mixed connective tissue disease–an apparently distinct rheumatic disease syndrome associated with a specific antibody to an extractable nuclear antigen (ENA). Am J Med  1972; 52: 148– 59. Google Scholar CrossRef Search ADS PubMed  2 Tani C, Carli L, Vagnani S et al.   The diagnosis and classification of mixed connective tissue disease. J Autoimmun  2014; 48–49: 46– 9. Google Scholar CrossRef Search ADS PubMed  3 Cappelli S, Bellando Randone S, Martinovic D et al.   "To be or not to be," ten years after: evidence for mixed connective tissue disease as a distinct entity. Semin Arthritis Rheum  2012; 41: 589– 98. Google Scholar CrossRef Search ADS PubMed  4 Ciang NC, Pereira N, Isenberg DA. Mixed connective tissue disease—enigma variations? Rheumatology  2017; 56: 326– 33. Google Scholar PubMed  5 Flam ST, Gunnarsson R, Garen T et al.   The HLA profiles of mixed connective tissue disease differ distinctly from the profiles of clinically related connective tissue diseases. Rheumatology  2015; 54: 528– 35. Google Scholar CrossRef Search ADS PubMed  6 Black C, Isenberg DA. Mixed connective tissue disease—goodbye to all that. Br J Rheumatol  1992; 31: 695– 700. Google Scholar CrossRef Search ADS PubMed  7 Paradowska-Gorycka A, Jurkowska M, Czuszynska Z et al.   IL-10, IL-12B and IL-17 gene polymorphisms in patients with mixed connective tissue disease. Modern Rheumatol  2015; 25: 487– 9. Google Scholar CrossRef Search ADS   8 Dong RP, Kimura A, Hashimoto H et al.   Difference in HLA-linked genetic background between mixed connective tissue disease and systemic lupus erythematosus. Tissue Antigens  1993; 41: 20– 5. Google Scholar CrossRef Search ADS PubMed  9 Goh NS, Desai SR, Veeraraghavan S et al.   Interstitial lung disease in systemic sclerosis: a simple staging system. Am J Respir Crit Care Med  2008; 177: 1248– 54. Google Scholar CrossRef Search ADS PubMed  10 Komocsi A, Vorobcsuk A, Faludi R et al.   The impact of cardiopulmonary manifestations on the mortality of SSc: a systematic review and meta-analysis of observational studies. Rheumatology  2012; 51: 1027– 36. Google Scholar CrossRef Search ADS PubMed  11 Johnson C, Pinal-Fernandez I, Parikh R et al.   Assessment of mortality in autoimmune myositis with and without associated interstitial lung disease. Lung  2016; 194: 733– 7. Google Scholar CrossRef Search ADS PubMed  12 Hajas A, Szodoray P, Nakken B et al.   Clinical course, prognosis, and causes of death in mixed connective tissue disease. J Rheumatol  2013; 40: 1134– 42. Google Scholar CrossRef Search ADS PubMed  13 Fagundes MN, Caleiro MT, Navarro-Rodriguez T et al.   Esophageal involvement and interstitial lung disease in mixed connective tissue disease. Respir Med  2009; 103: 854– 60. Google Scholar CrossRef Search ADS PubMed  14 Gunnarsson R, Molberg O, Gilboe IM, Gran JT, Group PS. The prevalence and incidence of mixed connective tissue disease: a national multicentre survey of Norwegian patients. Ann Rheum Dis  2011; 70: 1047– 51. Google Scholar CrossRef Search ADS PubMed  15 Gunnarsson R, Aalokken TM, Molberg O et al.   Prevalence and severity of interstitial lung disease in mixed connective tissue disease: a nationwide, cross-sectional study. Ann Rheum Dis  2012; 71: 1966– 72. Google Scholar CrossRef Search ADS PubMed  16 Kozuka T, Johkoh T, Honda O et al.   Pulmonary involvement in mixed connective tissue disease: high-resolution CT findings in 41 patients. J Thorac Imaging  2001; 16: 94– 8. Google Scholar CrossRef Search ADS PubMed  17 Bodolay E, Szekanecz Z, Devenyi K et al.   Evaluation of interstitial lung disease in mixed connective tissue disease (MCTD). Rheumatology  2005; 44: 656– 61. Google Scholar CrossRef Search ADS PubMed  18 Kawano-Dourado L, Baldi BG, Kay FU et al.   Pulmonary involvement in long-term mixed connective tissue disease: functional trends and image findings after 10 years. Clin Exp Rheumatol  2015; 33: 234– 40. Google Scholar PubMed  19 Winstone TA, Assayag D, Wilcox PG et al.   Predictors of mortality and progression in scleroderma-associated interstitial lung disease: a systematic review. Chest  2014; 146: 422– 36. Google Scholar CrossRef Search ADS PubMed  20 Marie I. Morbidity and mortality in adult polymyositis and dermatomyositis. Curr Rheumatol Rep  2012; 14: 275– 85. Google Scholar CrossRef Search ADS PubMed  21 Yu KH, Wu YJ, Kuo CF et al.   Survival analysis of patients with dermatomyositis and polymyositis: analysis of 192 Chinese cases. Clin Rheumatol  2011; 30: 1595– 601. Google Scholar CrossRef Search ADS PubMed  22 Yamasaki Y, Yamada H, Ohkubo M et al.   Longterm survival and associated risk factors in patients with adult-onset idiopathic inflammatory myopathies and amyopathic dermatomyositis: experience in a single institute in Japan. J Rheumatol  2011; 38: 1636– 43. Google Scholar CrossRef Search ADS PubMed  23 Kasukawa RS, Gordon C (eds). Mixed Connective Tissue Disease and Anti-Nuclear Antibodies. Proceedings of the International Symposium on Mixed Connective Tissue Disease and Anti-nuclear Antibodies. Tokyo, 29–30 August 1986 . Amsterdam: Excerpta Medica, 1987. 24 Galie N, Hoeper MM, Humbert M et al.   Guidelines for the diagnosis and treatment of pulmonary hypertension: the Task Force for the Diagnosis and Treatment of Pulmonary Hypertension of the European Society of Cardiology (ESC) and the European Respiratory Society (ERS), endorsed by the International Society of Heart and Lung Transplantation (ISHLT). Eur Heart J  2009; 30: 2493– 537. Google Scholar CrossRef Search ADS PubMed  25 Hansell DM, Bankier AA, MacMahon H et al.   Fleischner Society: glossary of terms for thoracic imaging. Radiology  2008; 246: 697– 722. Google Scholar CrossRef Search ADS PubMed  26 Galvin JR, Frazier AA, Franks TJ. Collaborative radiologic and histopathologic assessment of fibrotic lung disease. Radiology  2010; 255: 692– 706. Google Scholar CrossRef Search ADS PubMed  27 Miller MR, Hankinson J, Brusasco V et al.   Standardisation of spirometry. Eur Respir J  2005; 26: 319– 38. Google Scholar CrossRef Search ADS PubMed  28 Textor J, Hardt J, Knuppel S. DAGitty: a graphical tool for analyzing causal diagrams. Epidemiology  2011; 22: 745. Google Scholar CrossRef Search ADS PubMed  29 Greidinger EL, Zang Y, Jaimes K et al.   A murine model of mixed connective tissue disease induced with U1 small nuclear RNP autoantigen. Arthritis Rheum  2006; 54: 661– 9. Google Scholar CrossRef Search ADS PubMed  30 Hetlevik SO, Flato B, Rygg M et al.   Long-term outcome in juvenile-onset mixed connective tissue disease: a nationwide Norwegian study. Ann Rheum Dis  2016; 76: 159– 65. Google Scholar CrossRef Search ADS PubMed  31 Burdt MA, Hoffman RW, Deutscher SL et al.   Long-term outcome in mixed connective tissue disease: longitudinal clinical and serologic findings. Arthritis Rheum  1999; 42: 899– 909. Google Scholar CrossRef Search ADS PubMed  32 Gunnarsson R, El-Hage F, Aalokken TM et al.   Associations between anti-Ro52 antibodies and lung fibrosis in mixed connective tissue disease. Rheumatology  2016; 55: 103– 8. Google Scholar CrossRef Search ADS PubMed  33 Szodoray P, Hajas A, Kardos L et al.   Distinct phenotypes in mixed connective tissue disease: subgroups and survival. Lupus  2012; 21: 1412– 22. Google Scholar CrossRef Search ADS PubMed  34 Vittinghoff E, McCulloch CE. Relaxing the rule of ten events per variable in logistic and Cox regression. Am J Epidemiol  2007; 165: 710– 8. Google Scholar CrossRef Search ADS PubMed  © The Author 2017. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: journals.permissions@oup.com

Journal

RheumatologyOxford University Press

Published: Feb 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Unlimited reading

Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.

Stay up to date

Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.

Organize your research

It’s easy to organize your research with our built-in tools.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve Freelancer

DeepDyve Pro

Price
FREE
$49/month

$360/year
Save searches from
Google Scholar,
PubMed
Create lists to
organize your research
Export lists, citations
Read DeepDyve articles
Abstract access only
Unlimited access to over
18 million full-text articles
Print
20 pages/month
PDF Discount
20% off