Abstract
Objective
There are few clinical predictors of progression of systemic sclerosis (SSc) related interstitial lung disease (ILD). Herein, we examine the predictive significance of key cytokines for long-term progression of ILD and survival in early SSc in two independent cohorts.
Methods
Plasma levels of 11 Th1/Th2 cytokines (IL-1β, IL-5, IL-6, IL-8, IL-10, IL-12, IL-13, tumor necrosis factor, CCL2, interferon-inducible T cell a chemoattractant, and interferon-g–inducible 10-kd protein) were measured in 266 SSc patients in the GENISOS cohort. Next, CCL-2, IL-10, and IL-6 were measured in 171 early SSc patients from the CSRG replication cohort. The primary outcome was a decline in forced vital capacity predicted (FVC%) over time. A joint analysis of longitudinal FVC% measurements and survival was performed.
Results
After adjustment for age, gender, and ethnicity, CCL-2 and IL-10 were significant predictors of ILD progression in the discovery cohort. Higher CCL-2 predicted faster decline in FVC% predicted (b=−0.57, p=0.032) while IL-10 predicted a slower decline (b=0.26, p=0.01). CCL-2 was also predictive of poorer survival (HR: 1.76, p=0.030). In the CSRG replication cohort, CCL-2 predicted faster decline in FVC% (b=−0.58, p=0.038) but neither IL10 nor IL-6 had predictive significance. CCL-2 also predicted poorer survival (HR: 3.89, p= 0.037).
Conclusion
Higher CCL-2 levels in the circulation were predictive of ILD progression and poorer survival in SSc, supporting its role as a biomarker and potential therapeutic target.
Keywords: Cytokines, CCL-2, IL-10, systemic sclerosis, ILD
Introduction
Systemic sclerosis (SSc) is an autoimmune, fibrotic disease with high disease related morbidity and mortality (1). Pulmonary involvement including interstitial lung disease (ILD) has become the leading cause of SSc-related death (2;3). The course of ILD is highly variable and predictive biomarkers could lead to more effective and individualized monitoring and treatment strategies. Previously, we investigated the predictive significance of a comprehensive list of demographic and autoantibody variables, including disease types for long-term ILD progression. None of the selected variables % at baseline predicted long-term decline in forced vital capacity (FVC) % (4). In a follow-up study, C-reactive protein (CRP) at baseline predicted long-term ILD progression in SSc (5). Another recent study indicated that serum IL-6 was predictive of ILD progression in patients with SSc (6). SSc patients have a distinct cytokine profile (7). However, the predictive significance of these cytokines for long-term progression of ILD has not been established in early SSc cohorts. The objective of the present study was to investigate the predictive significance of baseline Th1/Th2 cytokine/chemokine levels in the circulation for ILD progression using highly sensitive multiplex assays in two large, independent early SSc cohorts.
Patients and Methods
Study population
Genetics versus Environment in Scleroderma Outcome Study (GENISOS) is a collaborative effort from University of Texas Health Science Center (UTHSC) at Houston, UTHSC-San Antonio and University of Texas Medical Branch-Galveston. At the time of cytokine determination, 266 SSc patients with early SSc had been enrolled. All patients fulfilled the American College of Rheumatology and European League Against Rheumatism (ACR/EULAR) Classification Criteria for SSc (8) and had disease duration < 5 years at the time of enrollment (from the 1st non-Raynaud’s phenomenon symptom). Diffuse skin involvement was present in 156 patients (59%) and mean disease duration (determined from the first Non-Raynaud’s phenomenon symptom) at enrollment was 2.5 years (data available upon request from the corresponding author). To be enrolled in GENISOS, all subjects were ≥ 18 years old and were diagnosed by the local GENISOS rheumatologist investigators. 86 patients (32.3%) had died at the time of the death search query.
For the replication cohort, 171 eligible patients were identified in the Canadian Scleroderma Research Group (CSRG) registry. These patients had disease duration < 5 years, available serum samples, and at least two pulmonary function tests (PFTs). CSRG patients must have a diagnosis of SSc from a CSRG rheumatologist, be ≥ 18 years and be fluent in English or French. Over 98% meet the 2013 ACR/EULAR classification criteria for SSc (9). Diffuse skin involvement was present in 66 patients (39%) and mean disease duration was 2.2 years (data available upon request from the corresponding author). In the CSRG samples, 19 (10%) patients had died at the time of the death search query.
Cytokine determination
Plasma samples were examined from 266 SSc patients at the baseline visit in the GENISOS cohort, as well as 97 age-, gender-, and ethnicity-matched healthy controls. The blood samples were collected using EDTA collection tubes. After cells were removed by centrifugation for 10 min at 2600 rpm, plasma samples were stored at −80°C until analyzed. The plasma had not undergone more than 2 freeze-thaw cycles before the chemokine levels were determined. In the first step, eleven key Th1/Th2 cytokines and chemokines, namely Interleukin 1beta (IL-1β), Interleukin 5, 6, 8, 10, 12, 13 (IL-5, IL-6, IL-8, IL-10, IL-12, IL-13), tumor necrosis factor-alpha (TNF-α), C-C motif ligand-2 (CCL-2 also called monocyte chemoattractant protein-1 or MCP-1), Interferon-inducible T-cell alpha chemoattractant (ITAC, also called CXCL11), and Interferon gamma-induced protein-10 (IP-10) were measured in the GENISOS cohort.
In the replication step, IL-6, IL-10 and CCL-2 were measured in the baseline serum samples of 171 CSRG patients by the same methodology.
All cytokine quantifications were performed by Mesoscale highly sensitive multiplex sandwich immunoassay (Meso Scale Discovery, Rockville, MD) using the electro-chemiluminescence system accordingly to the protocol recommended by the manufacturer. Each sample was run in duplicate. All measurements had a coefficient of variation <0.2.
Severity of SSc-ILD
Pulmonary Function Tests (PFTs) were performed at the initial visit and annually thereafter. FVC% as a continuous variable was used as a surrogate for severity of SSc-ILD as this has been demonstrated to be a validated outcome measure for severity of ILD in randomized controlled studies of patients with SSc (10). Longitudinal FVC% predicted analysis was based on 1,016 individual FVC% measurements in the GENISOS cohort and 968 measurements in the CSRG subjects. The mean of follow up time in the GENISOS study was 4.36 years (up to 13.1 years) and 5.72 years (up to 9.71 years) in CSRG cases. The rate of change in longitudinally obtained FVC % predicted value served as a surrogate measurement of ILD progression.
Vital status determination
All deaths were recorded prospectively in the data base. To ensure complete capture of vital status in the GENISOS cohort, a National Death Index query was also performed.
Statistical Analysis
All cytokine levels were log-transformed. Cytokine level differences between patients and controls were investigated by t-test. Pearson’s correlations were calculated between cytokine levels and baseline FVC%. Linear regression analysis was also used to examine the univariable and multivariable associations between the cytokines and baseline FVC%.
A joint analysis of longitudinal measurements (sequentially obtained % predicted FVC) and survival data was also conducted in order to investigate the predictive significance of the baseline cytokine levels (as a continuous variable) for the long-term change in FVC%. This analysis allows inclusion of all FVC% measurements and adjusts for baseline differences in FVC% (random intercept). Furthermore, it accounts for the association between FVC% and survival, and reduces the bias resulting from the fact that patients with more rapid decline in FVC% have a higher mortality (4). The longitudinal component consisted of a linear model with random effects. The ultimate goal of this analysis was to examine whether baseline cytokine levels had predictive significance for faster (or slower) decline in FVC% over time. This was investigated by the interaction term between the cytokine levels and follow-up time (11). We focused on the interaction term between the cytokine and follow-up time rather than the main effect for cytokine as the main predictor variable because a time component has to be included to determine the rate of change in the outcome variable (i.e. FVC%). The main effect would provide information on the association with serially obtained FVC%s but would not identify cytokines that are associated with faster (or slower) rate of decline in FVC%. The survival component fitted a parametric Weibull model with the cytokines as predictors. The starting point of the survival analysis was time of enrollment. The overall Wald Chi-square test was highly significant (p<0.001) for the final models indicating a significant association between the predictors in the models and FVC. We also evaluated model fit with various marginal and subject-specific residual diagnostic plots. No deviations from the model assumptions were evident in any of the plots (12).
Results
Discovery cohort: Comparison of cytokine/chemokine levels in SSc patients vs. controls
First, we investigated 11 key Th1/Th2 cytokines/chemokines namely IL1-β, IL-5, IL-6, IL-8, IL-10, IL-12, IL-13, TNF-α, CCL-2, ITAC, IP-10 in plasma samples from 266 SSc patients and 97 age-, gender-, and ethnicity-matched unaffected controls. SSc patients had significantly higher IL-6, IL-8, IL-10, TNF-α, CCL-2, ITAC and IP-10 plasma levels compared to unaffected controls (Figure 1 and data available upon request from the corresponding author). Limiting the analysis to patients not treated with immunosuppressive agents at the baseline visit did not change the observed significant associations (data available upon request from the corresponding author).
Figure 1.
Comparison of plasma cytokine/chemokine levels in Genetics versus Environment in Scleroderma Outcome Study (GENISOS) patients versus controls. Plasma samples from 266 systemic sclerosis (SSc) patients and 97 age-, sex-, and ethnicity-matched unaffected controls were collected, and the levels of interleukin-1b (IL-1b), IL-5, IL-6, IL-8, IL-10, IL-12, IL-13, tumor necrosis factor (TNF), CCL2, interferon inducible T cell a chemoattractant (I-TAC), and interferon-g–inducible 10-kd protein (IP-10) were measured by multiplex platform assay. Data are shown as box plots. Each box represents the 25th to 75th percentiles. Lines inside the boxes represent the median.The depth of the box is the interquartile range (IQR). The line inside the box represents the median. The upper adjacent value is the largest observation that is less than or equal to the 75th percentile plus 1.5 times the IQR. The lower adjacent value is the smallest observation that is greater than or equal to the 25th percentile minus 1.5 times the IQR. Circles indicate outliers.
Discovery cohort: Correlation of cytokine levels with baseline FVC%
Next, we investigated the correlation of cytokines with the concomitantly obtained FVC% at the cross-sectional level. After adjustment for age, gender, and African American race, higher IL-1β, IL-6, IL-8, IL-10, IL-13, CCL-2, ITAC and IP-10 levels were associated with more severe restrictive lung disease at the baseline visit (i.e. significant negative correlation with the concomitantly obtained FVC%) although the observed associations were relatively weak (r > −0.3) (Table 1).
Table 1.
Cross-sectional correlation of cytokines with concomitantly obtained FVC% in the GENISOS cohort
| Cytokine/chemokine | Univarable Analysis | Multi-varable analysis* | |||
|---|---|---|---|---|---|
| r | p-value | b | 95% CI | p-value | |
| IL-1β | −0.17 | 0.012 | −2.90 | (−5.21; −0.59) | 0.014 |
| IL5 | −0.11 | 0.109 | −2.04 | (−4.26; 0.17) | 0.071 |
| IL6 | −0.16 | 0.020 | −3.23 | (−5.94; −0.52) | 0.020 |
| IL8 | −0.20 | 0.004 | −3.61 | (−5.80; −1.42) | 0.001 |
| IL10 | −0.18 | 0.008 | −2.78 | (4.78; −0.78) | 0.007 |
| IL12 | −0.12 | 0.079 | −1.64 | (−3.34; 0.06) | 0.059 |
| IL13 | −0.16 | 0.057 | −2.00 | (−3.94; −0.05) | 0.045 |
| TNF-α | −0.13 | 0.161 | −2.99 | (−6.98; 1.00) | 0.141 |
| CCL-2 | −0.18 | 0.008 | −7.36 | (−12.27; −2.46) | 0.003 |
| ITAC | −0.21 | 0.003 | −4.87 | (−7.91; −1.83) | 0.002 |
| IP10 | −0.08 | 0.272 | −1.66 | (−4.59; 1.26) | <0.001 |
Adjusted for age at enrollment, gender, and African American race.
Discovery cohort: Predictive significance of baseline cytokine levels for ILD progression
In the longitudinal analysis after adjustment for age, gender, and African American race, higher CCL-2 levels predicted faster decline in FVC% predicted (b=−0.57, 95% CI: −1.11 - −0.04, p=0.032) while higher IL-10 predicted a slower decline in FVC% (b=0.26, 95% CI: 0.06–0.46, p=0.01) and had a protective effect (Table 2). Similarly, higher IL-5and IL-12 cytokine levels (p=0.015 and p=0.018 respectively) were also predictive of slower decline in FVC% (Table 2). However, IL-10, IL-12 and IL-5 were highly correlated with each other (r>0.8) (data available upon request from the corresponding author). Due to problems arising from multi-co-linearity, they could not be investigated in the same multivariable model. From these three cytokines (IL-10, IL-12, and IL-5), IL-10 was pursued in the follow-up analyses because plasma IL-10 showed significant differential levels in the patient versus control comparison and had the strongest correlations with FVC% in the cross-sectional analysis (Figure 1 & Table 1).
Table 2.
Longitudinal analysis of predictive significance of baseline cytokine levels for rate of changes in FVC% in the GENISOS cohort*
| Cytokine/Chemokine | Main effect of cytokine | Time Effect | Interaction terms between baseline cytokine and follow up-time | |||
|---|---|---|---|---|---|---|
| b(95% CI) | p -value | b(95% CI) | p -value | b (95% CI) | p -value | |
| IL-1β | −2.12 (−4.01; −0.23) | 0.028 | −1.19(−1.67; −0.71) | <0.001 | 0.21 (−0.03; 0.45) | 0.085 |
| IL5 | −1.85 (−3.70; 0.00) | 0.050 | −1.18 (−1.58; −0.78) | <0.001 | 0.27 (0.05; 0.48) | 0.015 |
| IL6 | −3.00 (−4.76; −1.25) | 0.001 | −1.50 (−1.87; −1.13) | <0.001 | −0.12 (−0.36; 0.13) | 0.350 |
| IL8 | −2.46 (−3.85; −1.06) | 0.001 | −1.47 (−1.90; −1.05) | <0.0001 | 0.01 (−0.19; 0.21) | 0.928 |
| IL10 | −2.61 (−4.23; −0.98) | 0.002 | −1.42 (−1.76; =1.07) | <0.0001 | 0.26 (0.06; 0.46) | 0.010 |
| IL12 | −1.72 (−3.17; −0.28) | 0.019 | −1.40 (−1.75; −1.05) | <0.0001 | 0.21 (0.03; 0.38) | 0.018 |
| IL13 | −2.04 (−3.71; −0.36) | 0.017 | −1.46 (−1.82; −1.10) | <0.0001 | 0.13 (−0.09; 0.35) | 0.242 |
| TNF-α | −2.21 (−5.86; 1.44) | 0.235 | −1.73 (−2.29; −1.18) | <0.0001 | 0.41 (−0.12; 0.93) | 0.124 |
| CCL-2 | −5.30 (−9.49; −1.10) | 0.013 | 1.00 (−1.28; 3.28) | 0.391 | −0.57 (−1.11; −0.04) | 0.032 |
| ITAC | −4.33 (−6.99; −1.66) | 0.002 | −1.13 (−2.75; 0.48) | 0.168 | −0.06 (−0.39; 0.28) | 0.728 |
| IP10 | −0.91 (−3.44; 1.61) | 0.477 | −1.18 (−2.58; 0.22) | 0.099 | −0.06 (−0.37; 0.25) | 0.717 |
Adjusted for age at enrollment, gender, and African American race
The predictive significance of CCL-2 and IL-10 remained significant in a multivariable analysis that included both cytokines CCL-2 or IL-10 with disease type (diffuse versus limited cutaneous SSc), immunosuppressive treatment status at the baseline visit, anti-topoisomerase antibody (Topo I), C-reactive protein (Log transformed CRP also called LnCRP) in addition to age, gender, and African American race (Table 3) (b=−0.69, 95% CI: −1.24 - −0.15, p=0.013 for CCL-2 and b=0.28, 95% CI: 0.09 – 0.47, p=0.004 for IL-10). The other investigated cytokines did not show predictive significance for ILD progression. In the GENISOS cohort, IL-6 did not predict faster decline in FVC% predicted (b=−0.12, 95% CI: −0.36 - 0.13, p=0.350) (Table 2). Of note, IL-6, similar to CCL2 and IL-10 was associated with lower serially obtained FVCs (i.e. main effect in Table 2) which paralleled the findings at the cross sectional level (Table 1).
Table 3.
Extended multivariable analysis of predictive significance of CCL2 and IL-10 for rate of change in FVC% in GENISOS cohort *
| Cytokine | b | 95% CI | p-value |
|---|---|---|---|
| Time: CCL2 | −0.69 | (−1.24; −0.15) | 0.013 |
| Time: IL10 | 0.28 | (0.09; 0.47) | 0.004 |
| Time | 1.41 | (−0.98; 3.80) | 0.248 |
| CCL2 | −2.79 | (−7.07; 1.50) | 0.202 |
| IL10 | −2.19 | (−3.76; −0.63) | 0.006 |
| Diffuse | −5.24 | (−10.49; 0.01) | 0.051 |
| Immunosuppression | −8.44 | (−13.69; −3.20) | 0.002 |
| Topo I | −4.92 | (−11.79; 1.95) | 0.160 |
| LnCRP | −3.23 | (−4.96; −1.51) | <0.001 |
Adjusted for age at enrollment, gender, and African American race
In survival analysis, higher CCL-2 were also predictive of shorter survival (HR: 1.76, 95% CI: 1.06- 2.93, p=0.030) while IL-10 was not predictive of survival (HR: 0.99, 95% CI: 0.83- 1.18, p=0.896) after adjustment for age, gender, and African American race. Higher CCL-2 levels also showed a trend for predicting poorer survival (p=0.088) while IL-10 did not predict survival (p=0.763) in the extended model.
In a secondary analysis, we investigated whether CCL-2 or IL-10 predicted modified Rodnan Skin Score (mRSS) progression rate among patients with diffuse cutaneous involvement. Neither CCL-2 (b= −0.38, CI: −1.08 – 0.32, p=0.28) nor IL-10 (b= −0.16, CI: −0.46 - 0.14, p=0.29) was predictive of mRSS progression (data available upon request from the corresponding author).
Replication cohort
Next, we analyzed the association between CCL-2 and IL-10 and FVC% using serum samples from the CSRG replication cohort. We also investigated IL-6 levels in this cohort because IL-6 was predictive of FVC% progression in a previous study (6) and has been recently proposed as a treatment target in SSc (13). In agreement with results of the discovery cohort, both CCL-2 and IL-10 showed a significant negative correlation with the concomitantly obtained FVC% (r =−0.22, p=0.004 for CCL-2 and r =−0.18, p=0.024 for IL-10, respectively). IL-6 also showed a significant negative correlation (r = −0.19, p=0.016). The strength of correlations was also in the similar range.
In the simple longitudinal model, the predictive significance of the three cytokines CCL-2, IL-10 and IL-6 were investigated after adjustment for age and gender (data available upon request from the corresponding author). African American race was not included in the model because only three patients in CSRG were African American. Higher CCL-2 was a significant predictor of faster ILD progression (b=−0.58, 95% CI: −1.14- −0.03, p=0.038) (data available upon request from the corresponding author), while IL-6 and IL-10 levels were not predictive of differential rate in FVC% progression (p=0.068 and p=0.859, respectively). Of note, all three cytokines were associated with lower serially obtained FVC% levels which again paralleled the results of the cross-sectional analysis results.
In the extended model, CCL-2 showed predictive significance for faster decline in FVC% (b=−0.55, 95% CI: −1.08 - −0.01, p=0.046) after adjustment for disease subtype, immunosuppression, Topo I, LnCRP, age, and gender (Table 4). Higher CCL-2 levels were also predictive of poorer survival (HR: 3.89, 95% CI: 0.23–7.55, p=0.037) in the CSRG cohort.
Table 4.
Multivariable analysis of predictive significance of CCL2 level for rate of change in FVC% in the CSRG cohort
| Cytokine/Chemokine | b* | 95% CI* | P value* |
|---|---|---|---|
| Time: CCL2 | −0.55 | (−1.08; −0.01) | 0.046 |
| Time | 1.96 | (−1.21; 5.13) | 0.225 |
| CCL2 | −3.42 | (−7.21; 0.37) | 0.077 |
| Diffuse | −1.76 | (−8.03; 4.52) | 0.583 |
| Immunosuppression | −7.09 | (−13.56; −0.62) | 0.032 |
| Topo I | −10.42 | (−17.24; −3.59) | 0.003 |
| LnCRP | −0.18 | (−0.36; 0.01) | 0.064 |
Adjusted for age at enrollment, gender.
Discussion
Using a highly sensitive multiplex assay, we showed that SSc patients, regardless of treatment, had higher plasma levels of IL-6, IL-8, IL-10, TNF-α, CCL-2, ITAC and IP-10 compared to control subjects. Subsequently, we demonstrated for the first time that CCL-2 levels in the circulation were predictive of faster, long-term decline in FVC% in two independent, well characterized, early SSc cohorts. Furthermore, CCL-2 was predictive of poorer survival. IL-10 levels were predictive of slower decline in FVC% in the GENISOS cohort but this finding was not replicated in the CSRG cohort.
Previously, we reported in a large-scale skin global gene expression study that skin CCL-2 mRNA levels and plasma CCL-2 levels correlated with concomitantly obtained FVC%. Furthermore, immunofluorescence stain revealed that CCL-2 was overexpressed in the fibroblasts in affected skin and lung tissue from SSc patients compared to controls (14). In a previous study, CCL-2 was highly up-regulated in the Bronchoalveolar Lavage Fluid (BAL) samples from SSc-ILD patients compared with healthy controls or ILD negative SSc patients. Furthermore, patients with a high HRCT score (20 or greater) had higher CCL-2 levels when compared with patients with fewer fibrotic changes (15).
CCL-2 also stimulated pro-collagen α I mRNA expression in a dose-dependent manner on rat lung fibroblasts (16). In another study, CCL-2 (MCP-1) was highly expressed in dermal fibroblasts of SSc patients but not of unaffected controls. Furthermore, antibodies blocking CCL-2 decreased the chemotactic activity of fibroblasts. Although they noted an increase in the level of type I procollagen over time, there was no significant effect of recombinant CCL2 on procollagen a1(I) mRNA expression in SSc and healthy dermal fibroblasts (17). CCL-2 blocking antibody also significantly inhibited α-smooth muscle actin (α-SMA) expression in cultured SSc dermal fibroblasts (18). In a bleomycin induced dermal fibrosis mouse model, administration of anti-CCL-2 neutralizing antibody reduced bleomycin induced dermal fibrosis. Stimulation with CCL-2 up-regulated alpha 1 pro-collagen mRNA in normal human fibroblasts as determined by northern blot hybridization (19). CCL-2 might also contribute to polarization of macrophages to a pro-fibrotic subtype (20), which has been implicated in SSc pathogenesis (21). These studies indicated that CCL-2 might contribute to the fibrotic signaling pathways through its biological functions. Future studies are needed to elucidate the role of CCL-2 in this potentially important pro-fibrotic mechanism in SSc.
Reliable prediction models using molecular data are needed in SSc-ILD as its course is highly variable and clinical characteristics alone do not predict long-term rate of decline in FVC% (4). We have previously demonstrated baseline CRP levels predict long-term decline in FVC% (5). Importantly, the predictive significance of CCL-2 was independent of baseline CRP levels in the present study indicating that this cytokine can provide predictive information beyond CRP.
In the present study, IL-10 predicted a slower decline in FVC% and had a protective effect in the GENISOS cohort. Several mechanistic studies have indicated that IL-10 might have anti-fibrotic effect in the lung tissue. IL-10 gene delivery significantly suppressed hydroxyproline and TGF-β production in bleomycin induced lung fibrosis mouse model (22). Furthermore, IL-10 attenuated chronic LPS-induced airway inflammation and subepithelial fibrosis (23). Although the anti-fibrotic effect of IL-10 in the GENISOS cohort is supported by above animal model studies, this finding was not replicated in the CSRG cohort. Therefore, further studies are needed to investigate the potential role of circulating IL-10 as a protective biomarker for ILD progression.
The cytokine studies in the GENISOS cohort were performed in plasma samples while they were conducted in serum samples in the CSRG cohort. We focused on plasma samples in the GENISOS cohort, in order to avoid artificial fluctuation in cytokine levels resulting from coagulations. Plasma samples were not available in the CSRG biorepository, thus the confirmatory assays were conducted in serum samples. The fact that CCL2 showed predictive significance in both sample types underscores the robustness of this cytokine as a biomarker.
Another cytokine IL-6 has been investigated in previous studies and serum IL-6 was an independent predictor of FVC% and DLCO% decline in SSc (6). In the present study, IL-6 did not predict faster decline in FVC% neither in the GENISOS nor CSRG cohort, although it was associated with lower FVC% levels in both cohorts (ie. the main effect in the joint analysis).
Other potential candidate proteins for predicting course of SSc-ILD are pneumoproteins. These proteins are linked to lung parenchymal injury. Specifically, surfactant D (SPD), CCL-18, and Krebs von den Lungen-6 (KL-6) have been studied as potential biomarkers. In a previous study, SPD and CCL-18 were not predictive of long-term ILD progression in the GENISOS cohort (24). In a recently published study of 50 patients with untreated early SSc-ILD, higher KL-6 levels were predictive of development of end-stage lung disease (defined as FVC% < 50%, continuous oxygen supplementation, or death due to ILD). A threshold of 1,273 U/ml fir KL-6 had sensitivity of 87.5% and specificity of 100% for identifying patients who developed end-stage lung disease (25). Future independent studies are needed to confirm the predictive significance of KL-6 in general and the aforementioned threshold value in particular for SSc-ILD progression.
The present study has several strengths. We conducted the study in two independent, well characterized cohorts of early SSc patients. The mean disease duration in the GENISOS cohort and subgroup of CSRG cases was 2.5 and 2.2 years, respectively. The focus on early disease avoids problems arising from survival bias. Furthermore, we used analytic models that allows inclusion of all long-term, prospectively obtained FVC% measurements (1,016 and 968 PFTs in the GENISOS and CSRG cohorts, respectively) and accounts for the fact that patients with faster ILD progression have higher mortality (non-random missing data). Another strength of our study is that we utilized multiplex platform assays (Mesoscale Discovery) that are more sensitive than conventional ELISAs or most other multiplex platforms (e.g. Luminex), providing a higher dynamic range and more accurate measurement of low-abundance chemokines/cytokines (26).
Our study also has some weaknesses. The study was conducted in an observational cohort where treatment regimens are heterogeneous, although we tried to address this limitation by adjusting for treatment with immunosuppressive agents in our multivariable models. However, we cannot exclude residual confounding. Therefore, this study was not well-suited to develop predictive biomarkers for response to specific therapeutic agents. Furthermore, we did not restrict our analysis to patients who already had evidence of ILD on HRCT at the baseline visit because the HRCT information was not available in all patients. However, we used the rate of decline in FVC% as a surrogate for ILD progression, and patients who do not have clinically significant ILD, should not have an accelerated decline in FVC%. Another limitation is that we examined the overall mortality rather than SSc-specific mortality because reliable data on causes of death were not available in our samples.
The present study provides evidence that baseline CCL-2 levels in the circulation predict faster decline in lung function in SSc patients and can be used to identify patients at risk for faster ILD progression. It also provides further support for CCL-2 as a treatment target in SSc-ILD.
Acknowledgments
This study was supported by the following grants: Scleroderma Foundation New Investigator Award to Dr. Wu, the National Institutes of Health (NIH) National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS) Centers of Research Translation (CORT) P50AR054144 to Dr. Mayes, NIH grant K23AR061436 to Dr. Assassi, NIH/NIAMS Scleroderma Family Registry and DNA Repository grant N01-AR02251 to Dr. Mayes, NIH/NIAMS AR055258 to Dr. Mayes, NIH/NIAMS R01AR062056-01A1 to Dr. Agarwal, NIH National Center for Clinical and Translational Sciences grant 3UL1RR024148, DOD Congressionally Directed Medical Research Program W81XWH-13-1-0452, Proposal # PR120687 to Dr. Mayes.
Biography
Dr. Mayes: Dr. Mayes received royalties (less than $10,000 for each) from Oxford University Press for “The Scleroderma Book”; from British Medical Journal for “Monograph on Scleroderma”; from Henry Stewar Talks for “SSc and Environmental Factors”; and received consultation fee (less than $10,000) from Clearview Health Care for Scleroderma unmet needs; All other authors have no relevant financial disclosures.
Reference List
- 1.Elhai M, Meune C, Avouac J, Kahan A, Allanore Y. Trends in mortality in patients with systemic sclerosis over 40 years: a systematic review and meta-analysis of cohort studies. Rheumatology (Oxford) 2012;51(6):1017–26. doi: 10.1093/rheumatology/ker269. [DOI] [PubMed] [Google Scholar]
- 2.Steen VD, Medsger TA. Changes in causes of death in systemic sclerosis, 1972–2002. Ann Rheum Dis. 2007;66(7):940–4. doi: 10.1136/ard.2006.066068. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Tyndall AJ, Bannert B, Vonk M, Airo P, Cozzi F, Carreira PE, 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(10):1809–15. doi: 10.1136/ard.2009.114264. [DOI] [PubMed] [Google Scholar]
- 4.Assassi S, Sharif R, Lasky RE, McNearney TA, Estrada YMR, Draeger H, et al. Predictors of interstitial lung disease in early systemic sclerosis: a prospective longitudinal study of the GENISOS cohort. Arthritis Res Ther. 2010;12(5):R166. doi: 10.1186/ar3125. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Liu X, Mayes MD, Pedroza C, Draeger HT, Gonzalez EB, Harper BE, et al. Does C-reactive protein predict the long-term progression of interstitial lung disease and survival in patients with early systemic sclerosis? Arthritis Care Res (Hoboken) 2013;65(8):1375–80. doi: 10.1002/acr.21968. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.De LA, Sestini P, Pantelidis P, Hoyles R, Hansell DM, Goh NS, et al. Serum interleukin 6 is predictive of early functional decline and mortality in interstitial lung disease associated with systemic sclerosis. J Rheumatol. 2013;40(4):435–46. doi: 10.3899/jrheum.120725. [DOI] [PubMed] [Google Scholar]
- 7.Gourh P, Arnett FC, Assassi S, Tan FK, Huang M, Diekman L, et al. Plasma cytokine profiles in systemic sclerosis: associations with autoantibody subsets and clinical manifestations. Arthritis Res Ther. 2009;11(5):R147. doi: 10.1186/ar2821. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.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. Ann Rheum Dis. 2013;72(11):1747–55. doi: 10.1136/annrheumdis-2013-204424. [DOI] [PubMed] [Google Scholar]
- 9.Alhajeri H, Hudson M, Fritzler M, Pope J, Tatibouet S, Markland J, et al. 2013 American College of Rheumatology/European League against rheumatism classification criteria for systemic sclerosis outperform the 1980 criteria: data from the Canadian Scleroderma Research Group. Arthritis Care Res (Hoboken) 2015;67(4):582–7. doi: 10.1002/acr.22451. [DOI] [PubMed] [Google Scholar]
- 10.Furst D, Khanna D, Matucci-Cerinic M, Clements P, Steen V, Pope J, et al. Systemic sclerosis - continuing progress in developing clinical measures of response. J Rheumatol. 2007;34(5):1194–200. [PubMed] [Google Scholar]
- 11.Rizopoulos Dimitris. An R Package for the Joint Modelling of Longitudinal and Time-to-Event Data. Journal of Statistical Software. 2010;35(9):1–33. [Google Scholar]
- 12.Rizopoulos D, Verbeke G, Molenberghs G. Multiple-imputation-based residuals and diagnostic plots for joint models of longitudinal and survival outcomes. Biometrics. 2010;66(1):20–9. doi: 10.1111/j.1541-0420.2009.01273.x. [DOI] [PubMed] [Google Scholar]
- 13.Khanna D, Denton CP, Jahreis A, van Laar JM, Frech TM, Anderson ME, et al. Safety and efficacy of subcutaneous tocilizumab in adults with systemic sclerosis (faSScinate): a phase 2, randomised, controlled trial. Lancet. 2016;387(10038):2630–40. doi: 10.1016/S0140-6736(16)00232-4. [DOI] [PubMed] [Google Scholar]
- 14.Assassi S, Wu M, Tan FK, Chang J, Graham TA, Furst DE, et al. Skin gene expression correlates of severity of interstitial lung disease in systemic sclerosis. Arthritis Rheum. 2013;65(11):2917–27. doi: 10.1002/art.38101. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Schmidt K, Martinez-Gamboa L, Meier S, Witt C, Meisel C, Hanitsch LG, et al. Bronchoalveoloar lavage fluid cytokines and chemokines as markers and predictors for the outcome of interstitial lung disease in systemic sclerosis patients. Arthritis Res Ther. 2009;11(4):R111. doi: 10.1186/ar2766. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Gharaee-Kermani M, Denholm EM, Phan SH. Costimulation of fibroblast collagen and transforming growth factor beta1 gene expression by monocyte chemoattractant protein-1 via specific receptors. J Biol Chem. 1996;271(30):17779–84. doi: 10.1074/jbc.271.30.17779. [DOI] [PubMed] [Google Scholar]
- 17.Distler O, Pap T, Kowal-Bielecka O, Meyringer R, Guiducci S, Landthaler M, et al. Overexpression of monocyte chemoattractant protein 1 in systemic sclerosis: role of platelet-derived growth factor and effects on monocyte chemotaxis and collagen synthesis. Arthritis Rheum. 2001;44(11):2665–78. doi: 10.1002/1529-0131(200111)44:11<2665::aid-art446>3.0.co;2-s. [DOI] [PubMed] [Google Scholar]
- 18.Carulli MT, Ong VH, Ponticos M, Shiwen X, Abraham DJ, Black CM, et al. Chemokine receptor CCR2 expression by systemic sclerosis fibroblasts: evidence for autocrine regulation of myofibroblast differentiation. Arthritis Rheum. 2005;52(12):3772–82. doi: 10.1002/art.21396. [DOI] [PubMed] [Google Scholar]
- 19.Yamamoto T, Nishioka K. Role of monocyte chemoattractant protein-1 and its receptor, CCR-2, in the pathogenesis of bleomycin-induced scleroderma. J Invest Dermatol. 2003;121(3):510–6. doi: 10.1046/j.1523-1747.2003.12408.x. [DOI] [PubMed] [Google Scholar]
- 20.Trial J, Cieslik KA, Haudek SB, Duerrschmid C, Entman ML. Th1/M1 conversion to th2/m2 responses in models of inflammation lacking cell death stimulates maturation of monocyte precursors to fibroblasts. Front Immunol. 2013;4:287. doi: 10.3389/fimmu.2013.00287. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Higashi-Kuwata N, Jinnin M, Makino T, Fukushima S, Inoue Y, Muchemwa FC, et al. Characterization of monocyte/macrophage subsets in the skin and peripheral blood derived from patients with systemic sclerosis. Arthritis Res Ther. 2010;12(4):R128. doi: 10.1186/ar3066. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Nakagome K, Dohi M, Okunishi K, Tanaka R, Miyazaki J, Yamamoto K. In vivo IL-10 gene delivery attenuates bleomycin induced pulmonary fibrosis by inhibiting the production and activation of TGF-beta in the lung. Thorax. 2006;61(10):886–94. doi: 10.1136/thx.2005.056317. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Garantziotis S, Brass DM, Savov J, Hollingsworth JW, McElvania-TeKippe E, Berman K, et al. Leukocyte-derived IL-10 reduces subepithelial fibrosis associated with chronically inhaled endotoxin. Am J Respir Cell Mol Biol. 2006;35(6):662–7. doi: 10.1165/rcmb.2006-0055OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Elhaj M, Charles J, Pedroza C, Liu X, Zhou X, Estrada YMR, et al. Can serum surfactant protein D or CC-chemokine ligand 18 predict outcome of interstitial lung disease in patients with early systemic sclerosis? J Rheumatol. 2013;40(7):1114–20. doi: 10.3899/jrheum.120997. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Kuwana M, Shirai Y, Takeuchi T. Elevated Serum Krebs von den Lungen-6 in Early Disease Predicts Subsequent Deterioration of Pulmonary Function in Patients with Systemic Sclerosis and Interstitial Lung Disease. J Rheumatol. 2016;43(10):1825–31. doi: 10.3899/jrheum.160339. [DOI] [PubMed] [Google Scholar]
- 26.Chowdhury F, Williams A, Johnson P. Validation and comparison of two multiplex technologies, Luminex and Mesoscale Discovery, for human cytokine profiling. J Immunol Methods. 2009;340(1):55–64. doi: 10.1016/j.jim.2008.10.002. [DOI] [PubMed] [Google Scholar]