Predictive Value of Serial High-Resolution Computed Tomography Analyses and Concurrent Lung Function Tests in Systemic Sclerosis

ARTHRITIS & RHEUMATOLOGY Vol. 67, No. 8, August 2015, pp 2205 2212 DOI 10.1002/art.39166 VC 2015, American College of Rheumatology redictive Value of Serial High-Resolution Computed Tomography Analyses and Concurrent Lung Function Tests in Systemic Sclerosis Anna-Maria Hoffmann-Vold, 1 Trond M. Aaløkken, 2 May Brit Lund, 2 Torhild Garen, 2 Øyvind Midtvedt, 2 Cathrine Brunborg, 2 Jan Tore Gran, 1 and Øyvind Molberg 1 Objective. Systemic sclerosis (SSc) carries a high risk of progressive interstitial lung disease (ILD), but tools for stratifying individual risk are scarce. The purpose of this study was to assess detailed data from serial lung fibrosis measurements and paired pulmonary function tests (FTs) as outcome prediction tools in a prospective cohort of SSc patients. Methods. aired FTs and high-resolution computed tomography (HRCT) scans were obtained at baseline and at followup in 305 SSc patients who met the American College of Rheumatology/European League Against Rheumatism 2013 classification criteria. The extent of fibrosis was scored on 10 sections from every HRCT scan and expressed as the percentage of the total lung volume. Results. Baseline HRCT analyses revealed 3 SSc subgroups: those with >20% lung fibrosis (n 5 40), those with 1 20% fibrosis (n 5 157), and those with no fibrosis (n 5 108). At followup HRCT (mean of 3.1 years later), all 108 group 3 patients were still free of fibrosis. In group 2 patients, 146 continued to have 1 20% fibrosis (group 2a), whereas 11 (marked by short disease duration of 1.3 years) had experienced progression to >20% fibrosis (group 2b). The annual fibrosis progression rate differed across the 4 groups: 0.9% in group 1, 0.7% in group 2a, 5.9% in group 2b, and 0% in group 3. The annual fibrosis progression rate correlated with the total Supported by grants from the Norwegian Women s ublic Health Association and the Norwegian Rheumatology Foundation. 1 Anna-Maria Hoffmann-Vold, MD, hd, Jan Tore Gran, MD, Øyvind Molberg, MD: Oslo University Hospital, Rikshospitalet, and University of Oslo, Oslo, Norway; 2 Trond M. Aaløkken, MD, hd, May Brit Lund, MD, hd, Torhild Garen, MSc, Øyvind Midtvedt, MD, Cathrine Brunborg, MSc: Oslo University Hospital, Rikshospitalet, Oslo, Norway. Address correspondence to Anna-Maria Hoffmann-Vold, MD, hd, Oslo University Hospital, Rikshospitalet, ostboks 4950, Nydalen, Oslo 0424, Norway. E-mail: anna.hoffmannvold@gmail.com. Submitted for publication October 19, 2014; accepted in revised form April 16, 2015. decline in the forced vital capacity (FVC) (7.1%, 5.7%, 8.7%, and 2.9% in groups 1, 2a, 2b, and 3, respectively), but not the diffusing capacity for carbon monoxide (DLCO) (8.4%, 7.7%, 7.7%, and 8.6%, respectively). Multivariate analyses identified anticentromere antibodies (odds ratio [OR] 4.7) and baseline DLCO (OR 1.04) as predictors of no fibrosis at followup and baseline fibrosis (OR 1.3) and FVC (OR 0.96) as predictors of >20% fibrosis at followup. Conclusion. These prospective cohort data suggest that HRCT performed at baseline predicts the development of fibrosis, the rate of progression of fibrosis, and the decline in pulmonary function in SSc. Systemic sclerosis (SSc) is a major challenge in clinical practice. The disease is heterogeneous, its outcome is unpredictable, and no disease-modifying therapies are available. Mortality rates are increased in SSc and are mainly driven by 2 cardiopulmonary disease components: pulmonary hypertension (H), which causes right-sided heart failure, and interstitial lung disease (ILD), which progresses to extensive lung fibrosis (1 5). SSc-associated ILD is frequent in both diffuse cutaneous SSc (dcssc) and limited cutaneous SSc (lcssc) (2). SSc-related ILD typically has an insidious onset, with subtle clinical symptoms (6). This may explain why SScrelated ILD is often diagnosed at an advanced stage, when extensive lung fibrosis is already present. Better, more-targeted strategies for the identification and risk stratification of SSc-related ILD early in the disease course are therefore warranted (7). ulmonary function tests (FTs) are key instruments for the assessment of SSc-related ILD in clinical practice, and a recently published expert opinion report highlighted forced vital capacity (FVC) as a core outcome measure for clinical trials (7,8). FTs are easily 2205

2206 HOFFMANN-VOLD ET AL accessible but have limitations. Abnormal findings on FTs are not specific for ILD, and wide ranges of normal values make baseline data difficult to evaluate (unless premorbid data are available). Benchmark FTs obtained at baseline are nonetheless critical for predicting outcome in SSc-related ILD. In fact, serial FT data suggest that a decline in the FVC to 50 70% predicts end-stage ILD and death within 5 years (9 11). A decline in the diffusing lung capacity for carbon monoxide (DLCO) also predicts all-cause mortality in SSc, but the risk is not restricted to ILD-related deaths (9,12). SSc patients with abnormal FT findings and/or with ILD symptoms are routinely evaluated with highresolution computed tomography (HRCT). This modality detects ILD with higher sensitivity than ordinary radiography (13), differentiates abnormalities in the airways from those in the parenchyma, and allows the quantitative assessment of the changes observed (13 16). In recent cross-sectional studies, a simple HRCT staging system was used for baseline evaluation of the total extent of ILD (i.e., ground-glass attenuation and reticular pattern changes consistent with fibrosis). Those studies showed that the baseline presence of extensive disease, defined as either ILD involving.20% of the total lung volume or ILD involving 10 20% of the total lung volume and an FVC of,70%, were associated with increased mortality rates (17,18). Consistent with this, data from the placebo arm of the Scleroderma Lung Study showed that.25% fibrosis on baseline HRCT predicted a decline in the FVC (19). Taken together, these studies demonstrated the predictive value of baseline HRCT analyses in patients with established SSc-related ILD. Serial HRCT analyses were not included in the Scleroderma Lung Study (19) or the Goh et al study (17), but Moore et al (18) assessed followup HRCT data in a subset of patients who had ILD at baseline HRCT (93 of 172) and found that an increasing extent of ILD seen on HRCT correlated with a decreasing FVC. Overall, we believe that there is still a need for data that allow the early stratification of SSc patients into those with a high risk of ILD (with a low threshold for aggressive therapy) and, equally importantly, those with a low risk of ILD. The issue of low risk was indirectly addressed in a retrospective analysis of serial HRCT data, in which 40 of 90 patients had normal findings on HRCT at baseline. At followup, 34 of the 40 patients (85%) continued to have normal findings on HRCT, indicating a good long-term prognosis (20). In the present study, we examined data from serial HRCTs and corresponding FTs performed in a large prospective cohort of SSc patients covering the whole spectrum of disease severity. Based on previous data, we focused the HRCT analyses on reticular pattern changes consistent with fibrosis (17 19) and addressed the following 5 research questions. What is the frequency of ILD in SSc patients over time? What are the annual fibrosis progression rates in SSc patients who develop ILD? What is the predictive value of baseline HRCT findings for fibrosis progression and lung function? Is there an association between fibrosis progression and death? What are the main risk factors for the development and progression of fibrosis? ATIENTS AND METHODS SSc patient cohort and inclusion criteria. Beginning in 2008, consecutive SSc patients from Oslo University Hospital were included in a prospective study (the Norwegian Systemic Connective Tissue Disease and Vasculitis Registry [NOSVAR]) with annual recording of data, including FT and lung HRCT findings (21). Inclusion criteria for the current study were fulfillment of the American College of Rheumatology (ACR)/European League Against Rheumatism (EULAR) 2013 classification criteria for SSc (22,23) plus data available on paired FTs and HRCTs performed both at baseline and at followup (at least 6 months apart). Clinical data and outcome measurements. atients were categorized as having lcssc or dcssc (24). Clinical parameters included smoking history, presence of H, SScrelated antibodies, and vital status. H was defined according to the European Society of Cardiology criteria (i.e., mean pulmonary artery pressure of $25 mm Hg at rest, as determined by right-sided heart catheterization) (25). All patients with clinical symptoms or systolic pulmonary artery pressure of.40 mm Hg on echocardiography underwent right-sided heart catheterization. Disease onset was defined as the first non Raynaud s phenomenon (non-r) symptom, and the time from disease onset until study end (January 2014) or until the time of death was defined as the total disease duration. The observation period was defined as the interval between the baseline and followup HRCTs. rimary outcomes were no lung fibrosis and.20% fibrosis at followup. Secondary outcomes were an FVC of,70% at followup and a decline in the FVC of.10% from baseline to followup. Survival was evaluated at study end but was excluded from outcome analyses, since the cohort had a survival bias. HRCT analyses and FTs. Low-dose thin-section CT images were obtained with the patient in the supine position, during breath-holding and inspiration. Images were reconstructed at a 1.25-mm section thickness at 10-mm intervals with the use of a high-spatial-frequency (bone) algorithm. Images were reviewed on a icture Archiving and Communication System (ACS) screen independently and in random order by an experienced chest radiologist (TMA) and a rheumatologist (A- MH-V) who was trained in CT interpretation. The observers were blinded with regard to the patient s data. In cases of differing interpretations, final conclusions were reached by consensus. Abnormal HRCT findings of reticular pattern changes (fine intralobular fibrosis without evident cysts, microcystic, and macrocystic) and honeycombing were defined as fibrosis. Superimposed ground-glass opacities were defined as being equivalent to fibrosis (26). Observers evaluated the presence and extent of

SERIAL LUNG ANALYSES IN SSc 2207 Table 1. Comparison of clinical characteristics between the SSc subgroups, stratified by extent of fibrosis on HRCT* Total cohort (n 5 305) Group 1 (n 5 40) Group 2a (n 5 146) Group 2b (n 5 11) Group 3 (n 5 108) Age at SSc onset, mean 6 SD years 48 6 15.0 46 6 15.7 49 6 15.2 47 6 12.0 47 6 14.8 Extent of fibrosis on HRCT Baseline.20 1 20 1 20 0 Followup.20 1 20.20 0 Time from SSc onset to baseline HRCT, mean 6 SD years 4.2 6 4.5 5.4 6 6.4 4.3 6 4.9 1.3 6 1.3 4.1 6 4.6 Time between HRCT and FT, mean 6 SD years 0.01 6 1.5 0.2 6 1.6 0.03 6 1.0 0.5 6 1.6 0.01 6 1.8 Time between baseline and followup HRCT, mean 6 SD years 3.1 6 2.6 3.4 6 2.7 3.3 6 2.7 3.6 6 3.2 2.8 6 2.3 Total observation period, mean 6 SD years 10.8 6 7.8 11.0 6 8.2 11.0 6 7.3 6.5 6 2.7 11.1 6 8.4 Male, no. (%) 63 (21) 14 (35) 33 (23) 3 (27) 13 (12) Ever smoked, no. (%) 124 (41) 13 (33) 62 (43) 4 (36) 45 (42) Deceased, no. (%) 67 (22) 12 (30) 32 (22) 4 (36) 19 (18) SSc subtype, no. (%) Diffuse cutaneous 87 (29) 22 (55) 45 (31) 6 (55) 17 (16) Limited cutaneous 218 (71) 18 (45) 100 (69) 5 (45) 91 (84) ulmonary hypertension, no. (%) 44 (14) 11 (28) 19 (13) 4 (36) 10 (9) Anti topoisomerase I antibodies, no. (%) 52 (17) 19 (48) 25 (17) 2 (18) 6 (6) Anticentromere antibodies, no. (%) 139 (46) 1 (5) 59 (40) 2 (18) 76 (70) * Systemic sclerosis (SSc) patients were grouped according to the extent of fibrosis seen on baseline high-resolution computed tomography (HRCT) scans. FT 5 pulmonary function testing. Time from SSc onset (first non Raynaud s phenomenon symptom) to end of study period or death. 5 0.014 versus group 1., 0.001 versus group 1. 5 0.006 versus group 1. fibrosis independently for each thin-section image. Area measurements were done precisely by drawing a freehand region of interest on the ACS screen to score the overall extent of fibrosis and relate this to the total lung volume. Fibrosis was expressed as the percentage of total lung volume and was stratified by extent (0%, 1 20%, and.20%) (17). rogression of fibrosis was expressed as the annual fibrosis progression rate, which was defined as the difference in extent of fibrosis between the baseline and followup HRCTs divided by the actual followup period (in years). Interrater variability between the readings by the radiologist and the rheumatologist was assessed in a subset of the patients, and the analyses showed interrater agreement on all fibrosis-free lung CT scans as well as on the grouping of patients into 1 20% and.20% fibrosis subsets. FTs were performed within 1 month of the corresponding HRCT. All tests were carried out according to American Thoracic Society/European Respiratory Society guidelines (27), using an automated Vmax V6200 system (SensorMedics). Recorded variables were FVC, forced expiratory volume in 1 second (FEV 1 ), FEV 1 /FVC, and DLCO, all of which are expressed as the percentage of predicted values, with reference values (27). Statistical analysis. Analyses were performed with SSS version 21 and Stata version 13 software. Odds ratios (ORs) with 95% confidence intervals (95% CIs) were used for analyses of potential risk factors and outcomes. Independent risk factors with a significance level of 20% on univariate analyses were included in the multivariate logistic regression analysis. A manual backward stepwise elimination procedure was performed to identify independent risk factors for the primary and secondary outcomes. Multivariate analyses were preceded by estimation of the correlation between risk factors. earson s chi-square test and independent sample t-test were used as appropriate. Items with significant effects on survival were entered into the Cox proportional hazards model. RESULTS Demographic and clinical variables in the study cohort. The study cohort included all of the 305 consecutive patients in the prospective Oslo University Hospital SSc cohort who had at least 2 pairs of lung HRCTs and corresponding FTs available and met the ACR/ EULAR 2013 SSc classification criteria. The mean 6 SD age at onset of the first non-r symptom in these 305 patients was 48 6 15.0 years. The mean 6 SD time from the first non-r symptom to the baseline HRCT was 4.2 6 4.5 years, and the mean time between baseline HRCT and baseline FT was 0.01 years (Table 1). The mean 6 SD observation time between baseline and followup HRCT was 3.1 6 2.6 years, and the mean 6 SD total disease duration, from the first non-r symptom to study end or death was 10.8 6 7.8 years. Right-sided heart catheterization was performed in 78 patients in whom the procedure was indicated, and 44 of these patients were diagnosed as having H (Table 1). A total of 67 of the 305 patients (22%) died during the time between the followup HRCT and the end of the study (Table 1). Longitudinal development of lung fibrosis. The mean 6 SD extent of lung fibrosis was 6.8 6 13.0% at baseline and 8.4 6 14.8% at followup, giving a total fibrosis progression rate of 1.6 6 5.5% and an annual fibrosis progression rate of 0.5 6 2.2%. At baseline, 197 of the 305 SSc patients (65%) had lung fibrosis (157 with 1 20% fibrosis and 40 with.20% fibrosis), while

2208 HOFFMANN-VOLD ET AL Figure 1. Course of lung fibrosis in patients with systemic sclerosis, stratified by the extent of fibrosis (0%, 1 20%, and.20%) on the baseline high-resolution computed tomography scans. Each symbol represents a different patient; lines connect the individual data points. Broken line indicates the cutoff for,20% and.20% fibrosis. 108 patients had no signs of fibrosis. Strikingly, none of the 108 SSc patients without lung fibrosis at baseline (group 3) had developed fibrosis on the followup HRCT obtained a mean of 3.1 years later (Table 1). The 108 patients with no fibrosis (group 3) were predominantly female (88%), had lcssc (84%), and were anticentromere antibody (ACA) positive (70%) (Table 1). In contrast, the 40 patients with.20% fibrosis at baseline (group 1) mostly had dcssc (55%), were anti topoisomerase I (anti topo I) positive (48%), and had a higher frequency of H (28%) (Table 1). Smoking habits did not differ across the groups (Table 1). Of the 157 patients with 1 20% fibrosis at baseline (groups 2a and b), 11 experienced progression to.20% fibrosis at followup (group 2b) (Table 1). These 11 patients were characterized by significantly shorter mean disease duration at baseline (1.3 years) than all of the other groups (Table 1). The mean annual rate of fibrosis progression was highest (5.9%) in the 11 patients who experienced progression from a level of 1 20% at baseline to a level of.20% at followup (Figure 1 and Table 2). Only 6 patients had an annual fibrosis progression rate.10% (Table 2). Lung fibrosis and disease duration. The mean 6 SD FVC at baseline in the total SSc cohort was 92.7 6 20.9%. Group 1 patients (.20% fibrosis at baseline) had the lowest FVC at 69.9% (Table 2). At followup, the FVC in the total cohort had declined by 4.8 6 12.7% (mean 6 SD). A decline of.10% in the FVC was observed in 85 of the 305 patients (28%), with the highest frequency (46%) in the 11 patients with 1 20% fibrosis at baseline and.20% fibrosis at followup (group 2b) (Table 2). The mean 6 SD unadjusted DLCO at baseline was 66.5 6 20.8%, and the mean 6 SD decline in DLCO from baseline to followup was 8.2 6 13.8%. atients with.20% fibrosis at followup had lower mean DLCO values at baseline than Table 2. Comparison of data from HRCT and pulmonary function testing between the SSc subgroups, stratified by extent of fibrosis on HRCT* Group 1 (n 5 40) Group 2a (n 5 146) Group 2b (n 5 11) Group 3 (n 5 108) Extent of fibrosis on HRCT Baseline.20 1 20 1 20 0 Followup.20 1 20.20 0 Baseline lung fibrosis, mean 6 SD % 34.7 6 15.6 4.0 6 5.3 10.1 6 8.8 0 6 0 Annual rate of fibrosis progression Mean 6 SD % 0.9 6 4.1 0.7 6 2.3 5.9 6 4.8 0 6 0.10%, no. (%) 2 (5) 1 (1) 3 (27) 0 (0) Forced vital capacity Baseline, mean 6 SD % 69.9 6 18.5 94.9 6 18.6 76.7 6 19.2 100.6 6 17.6 Total decline, mean 6 SD % 7.1 6 10.7 5.1 6 13.5 8.7 6 8.6 2.9 6 12.3.10% decline, no. (%) 15 (38) 44 (30) 5 (46) 20 (19),70% at followup, no. (%) 29 (73) 23 (16) 6 (55) 5 (5) Diffusing capacity for carbon monoxide Baseline, mean 6 SD % 47.5 6 17.1 66.1 6 17.7 48.3 6 23.1 76.1 6 19.6 Decline, mean 6 SD % 8.4 6 12.7 7.7 6 14.5 7.7 6 15.2 8.6 6 13.2.15% decline, no. (%) 13 (33) 41 (28) 3 (27) 31 (29) * Systemic sclerosis (SSc) patients were grouped according to the extent of fibrosis seen on high-resolution computed tomography (HRCT) scans at baseline., 0.001 versus group 1. 5 0.024 versus group 1. 5 0.008 versus group 1.

SERIAL LUNG ANALYSES IN SSc 2209 Table 3. rediction model with univariate analyses for primary and secondary outcomes at the time of followup HRCT in the SSc patients* rimary outcome Secondary outcome Variable No fibrosis, OR (95% CI) Fibrosis.20%, OR (95% CI) FVC,70%, OR (95% CI) FVC decline.10%, OR (95% CI) ulmonary hypertension 0.5 (0.23 1.03) 0.061 3.7 (1.81 7.50) 0.000 2.9 (1.49 5.74) 0.002 2.0 (1.03 3.89) 0.040 Male sex 0.4 (0.20 0.78) 0.007 2.0 (1.03 3.94) 0.041 2.5 (1.37 4.62) 0.003 1.0 (0.56 1.93) 0.889 Diffuse cutaneous SSc 0.4 (0.19 0.63),0.001 2.8 (1.45 5.23) 0.001 3.9 (2.20 6.91) 0.000 1.9 (1.16 3.38) 0.013 Anti topo I antibody 0.2 (0.80 0.47),0.001 5.0 (2.57 9.86) 0.000 2.7 (1.41 5.07) 0.003 1.3 (0.69 2.51) 0.395 Anticentromere antibody 5.1 (3.03 8.41),0.001 0.1 (0.03 0.21) 0.000 0.1 (0.06 0.26) 0.000 0.5 (0.26 0.76) 0.003 FVC at baseline 1.0 (1.02 1.04),0.001 0.9 (0.90 0.94) 0.000 0.01 (0.00 0.03) 0.000 1.2 (0.60 2.37) 0.613 DLCO at baseline 1.0 (1.03 1.05),0.001 0.9 (0.91 0.95) 0.000 0.2 (0.10 0.33) 0.000 0.9 (0.54 1.54) 0.746 Fibrosis at baseline NA NA 1.3 (1.21 1.39) 0.000 1.1 (1.09 1.15) 0.000 1.9 (0.95 3.77) 0.070 * HRCT 5 high-resolution computed tomography; SSc 5 systemic sclerosis; OR 5 odds ratio; 95% CI 5 95% confidence interval; FVC 5 forced vital capacity; anti topo I 5 anti topoisomerase I; DLCO 5 diffusing capacity for carbon monoxide; NA 5 not applicable. did the other groups, but there was no significant between-group differences in the mean decrease in DLCO or in the frequency of patients with a decline in the DLCO of.15% (Table 2). Fibrosis progression rate in relation to disease duration. Disease duration at the time of baseline HRCT was,3 years in 58% of the patients (177 of 305), 4 10 years in 23%, and.10 years in 19% of the patients (data available upon request from the corresponding author). The percentage of fibrosis at baseline did not differ between these 3 groups ( 5 0.216), but the group with,3 years of disease duration had a significantly higher annual rate of fibrosis progression (0.7%) than the groups with 4 10 years (0.3%) and.10 years (0.02%) of disease duration ( 5 0.049). Subgroup analyses of the 177 patients with,3 years of disease duration showed that the fibrosis progression rate was highest in the patient subgroup that experienced progression from 1 20% fibrosis at baseline to.20% fibrosis at followup (group 2b; n 5 11) (data available upon request from the corresponding author). These 11 patients predominantly had dcssc (64%) and had lower baseline FVC and DLCO values than did the patients who did not experience progression beyond the 1 20% fibrosis noted at followup (data available upon request from the corresponding author). Outcome analyses. Univariate analyses showed significant associations between the absence of lung fibrosis at followup and the presence of ACAs and between the presence of.20% fibrosis at followup and male sex, dcssc, anti topo I antibodies, and H (Table 3). Both of the primary outcomes were also associated with the baseline FVC and DLCO values. Univariate analyses of the secondary outcomes (FVC,70% and decline in the FVC of.10% from baseline to followup) were also performed, and the results are shown in Table 3. In the multivariate analyses, predictors of no lung fibrosis at followup were the presence of ACAs (OR 4.7) and the values for the DLCO (OR 1.0) at baseline, while.20% fibrosis was predicted by the baseline values for fibrosis (OR 1.2) and FVC (OR 0.96) (Table 4). FVC,70% was predicted by baseline fibrosis (OR 1.0), FVC Table 4. rediction model with multivariate analyses for lung outcomes in the SSc patients* rimary outcome Secondary outcome Variable No fibrosis, OR (95% CI) Fibrosis.20%, OR (95% CI) FVC,70%, OR (95% CI) FVC decline.10%, OR (95% CI) ulmonary hypertension NS NS NS 2.2 (1.12 4.35) 0.023 Male sex NS NS NS NS Diffuse cutaneous SSc NS NS NS NS Anti topo I antibody NS NS NS NS Anticentromere antibody 4.7 (2.72 7.98),0.001 NS 0.3 (0.08 0.77) 0.017 0.4 (0.25 0.73) 0.002 FVC at baseline NS 1.0 (0.93 0.99),0.001 0.9 (0.82 0.90),0.001 NS DLCO at baseline 1.04 (1.02 1.05),0.001 NS NS NS Fibrosis at baseline NS 1.3 (1.18 1.36) 0.045 1.0 (1.00 1.08) 0.049 NS * SSc 5 systemic sclerosis; OR 5 odds ratio; 95% CI 5 95% confidence interval; FVC 5 forced vital capacity; NS 5 not significant; anti topo I 5 anti topoisomerase I; DLCO 5 diffusing capacity for carbon monoxide. 5 0.211 for goodness of fit; area under the receiver operating characteristic curve 0.78. 5 0.511 for goodness of fit; area under the receiver operating characteristic curve 0.96. 5 0.434 for goodness of fit; area under the receiver operating characteristic curve 0.96. 5 0.322 for goodness of fit; area under the receiver operating characteristic curve 0.62.

2210 HOFFMANN-VOLD ET AL Table 5. Multivariate analyses of mortality risk factors in SSc patients who died during the period between the followup HRCT and the end of the study* atient characteristic HR (95% CI) ulmonary hypertension, no. (%) 2.0 (1.16 3.52) 0.013 Age at disease onset, years 1.1 (1.06 1.11),0.001 Baseline FVC, % 0.9 (0.97 0.99) 0.017 FVC decline, % 0.9 (0.95 0.99) 0.002 Baseline DLCO, % 0.9 (0.97 0.99) 0.019 * Hazard ratios (HRs) and 95% confidence intervals (95% CIs) were estimated by Cox regression analysis using a manual backward elimination procedure. The test of proportional-hazard assumption for the model showed 5 0.590, with all covariates showing. 0.05. SSc 5 systemic sclerosis; HRCT 5 high-resolution computed tomography; FVC 5 forced vital capacity; DLCO 5 diffusing capacity for carbon monoxide. (OR 0.9), and ACAs (OR 0.3). A decline in the FVC of.10% was also predicted by the presence of ACAs (OR 0.4) and by the development of H (OR 2.2) (Table 4). Survival. During the period from the followup HRCT to the end of the study, 67 of the 305 patients (22%) died. Univariate analyses showed that the patients who died were older at disease onset than the survivors (age 53 years versus 46 years; hazard ratio [HR] 1.1 [95% CI 1.05 1.11],, 0.001), had a longer time from disease onset to baseline HRCT (5.2 years versus 4.0 years; HR 0.9 [95% CI 0.88 0.98], 5 0.011), were more likely to have dcssc (42% versus 25%; HR 2.0 [95% CI 1.24 3.28], 5 0.005), and had a higher frequency of H (37% versus 8%; HR 3.8 [95% CI 2.28 6.18],, 0.001). Baseline values for the FVC % predicted (83.9% versus 95.3%; HR 0.9 [95% CI 0.97 0.99], 5 0.002) and the DLCO % predicted (56.4% versus 69.3%; HR 0.9 [95% CI 0.96 0.99],, 0.001) were significantly associated with survival, and a decline in the FVC was higher among the patients who died (10.7% versus 3.1%; HR 0.9 [95% CI 0.94 0.98],, 0.001). The frequency of patients with.20% fibrosis was numerically higher in the patients who died than in those who survived (24% versus 16%), but the difference was not statistically significant. Multivariate analyses demonstrated that H (HR 2.0), age at disease onset (HR 1.1), baseline FVC (HR 0.9), baseline DLCO (HR 0.9), and decline in the FVC (HR 0.9) were significantly associated with survival (Table 5). DISCUSSION ILD is a major clinical problem in SSc, but data on key issues, such as the rate of fibrosis progression and the risk factors for progressive disease, are limited. In the present study, we performed detailed HRCT analyses of the annual rate of fibrosis progression in a prospective cohort of SSc patients with comprehensive followup data. The major findings were that a baseline HRCT showing no fibrosis was highly predictive of a fibrosis-free followup HRCT, and that.20% fibrosis was mostly already present at baseline and was associated with a high annual rate of fibrosis progression, declining FT values, and the development of H. The 108 patients who remained free of fibrosis from baseline to followup shared several disease characteristics with the 34 patients without progressive lung disease described retrospectively by Launay et al (20); the patients were predominantly women with lcssc and ACAs, and they had stable lung function, with an FVC of.70%. As expected from the study by Wells (13), the patients also had high DLCO values at baseline. Surprisingly, the DLCO in these 108 patients was not stable, but declined by the same degree as that in patients with lung fibrosis. This finding emphasizes that the mechanisms behind SSc-related changes in the DLCO probably are multifactorial and may involve pathologic changes of the vasculature. This is underscored by the observation that H was present in all subgroups of the SSc cohort, which supports the notion that in SSc patients, screening for H should be conducted independently of screening for fibrosis (28 30). Detailed knowledge about the time course of SSc-related ILD is critical for early detection and therapeutic targeting (7). Evidence suggests that a reduction in FT values may occur early in the course of SSc (9,11). Much less is known about the development of structural abnormalities noted on HRCT. It is possible to detect HRCT abnormalities early in SSc, but only a few studies have followed up such changes over time (13,18,20,31,32). The current study is, to our knowledge, the largest that systematically correlates serial HRCT findings with corresponding changes in FT values. Nearly 60% of the patients underwent HRCT for the first time within the first 3 years after disease onset (data available upon request from the corresponding author). Even though these 177 patients were probably selected for early HRCT because of clinical symptoms in the lungs, the frequency of patients with baseline fibrosis.20% was strikingly high. This suggests that severe fibrosis may develop very early in the course of SSc. Still, we observed that fibrosis progression occurred throughout the disease course and that it was most pronounced in patients with.20% fibrosis at baseline, irrespective of disease duration. Not unexpectedly, a high proportion of lung fibrosis and deterioration of lung function were predicted by male sex, dcssc subtype, and anti topo I antibodies. Since this observational study required baseline and followup HRCT scans, it selected against patients

SERIAL LUNG ANALYSES IN SSc 2211 who died early in the disease course. Despite this bias, it was interesting to note that all FVC parameters we assessed predicted death in the whole SSc cohort. Rather surprisingly, neither the presence of fibrosis at baseline nor the annual rate of fibrosis progression was a significant predictor of death. These findings were in apparent contrast to the associations between extensive disease and death reported by Goh et al (17) and Moore et al (18). We believe that this may be partly due to the survival bias in our cohort and partly due to the few failure events and the statistical power of our study. Consistent with other studies, we found that older age at disease onset, the presence of dcssc, and the presence of H were associated with death (1,5). The strengths of our study are as follows. First, and probably the most important, is that our study provided paired HRCT and lung function data on a largely unselected cohort of SSc patients. Second, the study included a long observation period and provides complete data sets on HRCT findings, FT results, and clinical characteristics. Third, because of the unique identification numbers as well as the official mortality statistics in Norway, no patients with serial assessments were lost to followup at the end of study. The most important limitations of our study are the interindividual variations in disease duration at study inclusion and the varying observation periods. These variations reflect the largely observational nature of the study. It was designed as a prospective study with inclusion of consecutive patients and systematic data registered at annual followup visits. Notably, the content of these annual visits was determined by the treating physician and not by a standardized protocol. Another potential limitation was that the survival bias with the current study design was unavoidable. Finally, since we assessed fibrosis and conducted FTs in a rather homogeneous Caucasian population, our results are not necessarily transferable to other populations. In conclusion, we report that the frequency of SSc patients with lung fibrosis was stable across the observation period. The annual fibrosis progression rate was closely associated with the extent of fibrosis at baseline. The baseline HRCT findings predicted the development of fibrosis and the deterioration in lung function. The extent of lung fibrosis and the fibrosis progression rate were not associated with death in our SSc cohort. The main risk factors for the development and progression of fibrosis were male sex, dcssc subtype, presence of anti topo I antibodies, an FVC of,70%, and the extent of fibrosis at baseline. Taken together, the results indicate that a baseline examination in patients with newly diagnosed SSc should include both lung HRCT and FTs. It appears that, at least in our cohort, these two modalities together efficiently segregate patients into subsets with a low risk versus a high risk of SSc-related ILD. SSc patients with a low risk of ILD should not undergo serial HRCT examinations, but they probably need serial FTs as an adjunct H detection tool. Our data do not provide unequivocal answers about the value of serial HRCT examinations in patients with fibrosis at the baseline HRCT. Hence, we believe that further studies are warranted to clarify this important issue. AUTHOR CONTRIBUTIONS All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Hoffmann-Vold had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study conception and design. Hoffmann-Vold, Lund, Garen, Gran, Molberg. Acquisition of data. Hoffmann-Vold, Aaløkken, Lund, Garen, Midtvedt. Analysis and interpretation of data. Hoffmann-Vold, Lund, Brunborg, Gran, Molberg. REFERENCES 1. Steen VD, Medsger TA. Changes in causes of death in systemic sclerosis, 1972-2002. Ann Rheum Dis 2007;66:940 4. 2. Tyndall AJ, Bannert B, Vonk M, Airo, Cozzi F, Carreira E, et al. Causes and risk factors for death in systemic sclerosis: a study from the EULAR Scleroderma Trials and Research (EUSTAR) database. Ann Rheum Dis 2010;69:1809 15. 3. Walker UA, Tyndall A, Czirjak L, Denton C, Farge-Bancel D, Kowal-Bielecka O, et al. Clinical risk assessment of organ manifestations in systemic sclerosis: a report from the EULAR Scleroderma Trials and Research group database. Ann Rheum Dis 2007;66:754 63. 4. Komocsi A, Vorobcsuk A, Faludi R, inter T, Lenkey Z, Kolto G, et al. The impact of cardiopulmonary manifestations on the mortality of SSc: a systematic review and meta-analysis of observational studies. Rheumatology (Oxford) 2012;51:1027 36. 5. Hoffmann-Vold AM, Molberg O, Midtvedt O, Garen T, Gran JT. Survival and causes of death in an unselected and complete cohort of Norwegian patients with systemic sclerosis. J Rheumatol 2013;40:1127 33. 6. Khanna D, Seibold JR, Wells A, Distler O, Allanore Y, Denton C, et al. Systemic sclerosis-associated interstitial lung disease: lessons from clinical trials, outcome measures, and future study design. Curr Rheumatol Rev 2010;6:138 44. 7. Saketkoo LA, Mittoo S, Huscher D, Khanna D, Dellaripa F, Distler O, et al. Connective tissue disease related interstitial lung diseases and idiopathic pulmonary fibrosis: provisional core sets of domains and instruments for use in clinical trials. Thorax 2014;69:428 36. 8. Behr J, Furst DE. ulmonary function tests. Rheumatology (Oxford) 2008;47 Suppl 5:v65 7. 9. Steen V. redictors of end stage lung disease in systemic sclerosis. Ann Rheum Dis 2003;62:97 9. 10. Morgan C, Knight C, Lunt M, Black CM, Silman AJ. redictors of end stage lung disease in a cohort of patients with scleroderma. Ann Rheum Dis 2003;62:146 50. 11. lastiras SC, Karadimitrakis S, Ziakas D, Vlachoyiannopoulos G, Moutsopoulos HM, Tzelepis GE. Scleroderma lung: initial forced vital capacity as predictor of pulmonary function decline. Arthritis Rheum 2006;55:598 602.

2212 HOFFMANN-VOLD ET AL 12. Steen VD, Medsger TA Jr. Severe organ involvement in systemic sclerosis with diffuse scleroderma. Arthritis Rheum 2000;43: 2437 44. 13. Wells AU. High-resolution computed tomography and scleroderma lung disease. Rheumatology (Oxford) 2008;47 Suppl 5: v59 61. 14. Winklehner A, Berger N, Maurer B, Distler O, Alkadhi H, Frauenfelder T. Screening for interstitial lung disease in systemic sclerosis: the diagnostic accuracy of HRCT image series with high increment and reduced number of slices. Ann Rheum Dis 2012;71:549 52. 15. Wells AU, Hansell DM, Rubens MB, King AD, Cramer D, Black CM, et al. Fibrosing alveolitis in systemic sclerosis: indices of lung function in relation to extent of disease on computed tomography. Arthritis Rheum 1997;40:1229 36. 16. Wells AU, Margaritopoulos GA, Antoniou KM, Denton C. Interstitial lung disease in systemic sclerosis. Semin Respir Crit Care Med 2014;35:213 21. 17. Goh NS, Desai SR, Veeraraghavan S, Hansell DM, Copley SJ, Maher TM, et al. Interstitial lung disease in systemic sclerosis: a simple staging system. Am J Respir Crit Care Med 2008;177: 1248 54. 18. Moore OA, Goh N, Corte T, Rouse H, Hennessy O, Thakkar V, et al. Extent of disease on high-resolution computed tomography lung is a predictor of decline and mortality in systemic sclerosis-related interstitial lung disease. Rheumatology (Oxford) 2013;52:155 60. 19. Khanna D, Tseng CH, Farmani N, Steen V, Furst DE, Clements J, et al. Clinical course of lung physiology in patients with scleroderma and interstitial lung disease: analysis of the Scleroderma Lung Study lacebo Group. Arthritis Rheum 2011;63: 3078 85. 20. Launay D, Remy-Jardin M, Michon-asturel U, Mastora I, Hachulla E, Lambert M, et al. High resolution computed tomography in fibrosing alveolitis associated with systemic sclerosis. J Rheumatol 2006;33:1789 801. 21. Hoffmann-Vold AM, Midtvedt O, Molberg O, Garen T, Gran JT. revalence of systemic sclerosis in south-east Norway. Rheumatology (Oxford) 2012;51:1600 5. 22. Van den Hoogen F, Khanna D, Fransen J, Johnson SR, Baron M, Tyndall A, et al. 2013 classification criteria for systemic sclerosis: an American College of Rheumatology/European League Against Rheumatism collaborative initiative. Arthritis Rheum 2013;65:2737 47. 23. Hoffmann-Vold AM, Gunnarsson R, Garen T, Midtvedt O, Molberg O. erformance of the 2013 American College of Rheumatology/European League Against Rheumatism classification criteria for systemic sclerosis (SSc) in large, well-defined cohorts of SSc and mixed connective tissue disease. J Rheumatol 2015;42:60 3. 24. Hachulla E, Launay D. Diagnosis and classification of systemic sclerosis. Clin Rev Allergy Immunol 2011;40:78 83. 25. Galie N, Hoeper MM, Humbert M, Torbicki A, Vachiery JL, Barbera JA, et al. Guidelines for the diagnosis and treatment of pulmonary hypertension: the Task Force for the Diagnosis and Treatment of ulmonary 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) [published erratum appears in Eur Heart J 2011;32:926]. Eur Heart J 2009;30:2493 537. 26. Hansell DM, Bankier AA, MacMahon H, McLoud TC, Muller NL, Remy J. Fleischner Society: glossary of terms for thoracic imaging. Radiology 2008;246:697 722. 27. Quanjer H, Tammeling GJ, Cotes JE, edersen OF, eslin R, Yernault JC. Lung volumes and forced ventilatory flows: report Working arty Standardization of Lung Function Tests, European Community for Steel and Coal: official statement of the European Respiratory Society. Eur Respir J Suppl 1993; 16:5 40. 28. Schwaiger J, Khanna D, Gerry Coghlan J. Screening patients with scleroderma for pulmonary arterial hypertension and implications for other at-risk populations. Eur Respir Rev 2013;22: 515 25. 29. Khanna D, Tan M, Furst DE, Hill NS, McLaughlin VV, Silver RM, et al. on behalf of the QuERI-SSc Investigators. Recognition of pulmonary hypertension in the rheumatology community: lessons from a quality enhancement research initiative. Clin Exp Rheumatol 2014;32 Suppl 86:21 7. 30. Hsu VM, Chung L, Hummers LK, Wigley F, Simms R, BolsterM,etal.Developmentofpulmonaryhypertensionina high-risk population with systemic sclerosis in the ulmonary Hypertension Assessment and Recognition of Outcomes in Scleroderma (HAROS) cohort study. Semin Arthritis Rheum 2014;44:55 62. 31. Goldin J, Elashoff R, Kim HJ, Yan X, Lynch D, Strollo D, et al. Treatment of scleroderma-interstitial lung disease with cyclophosphamide is associated with less progressive fibrosis on serial thoracic high-resolution CT scan than placebo: findings from the Scleroderma Lung Study. Chest 2009;136:1333 40. 32. Kim HJ, Brown MS, Elashoff R, Li G, Gjertson DW, Lynch DA, et al. Quantitative texture-based assessment of one-year changes in fibrotic reticular patterns on HRCT in scleroderma lung disease treated with oral cyclophosphamide. Eur Radiol 2011;21:2455 65.