Abstract
Objective. Clara cell secretory protein (CC16) is a sensitive marker of bronchial epithelial cell damage. The CC16 serum level is elevated in patients with pulmonary fibrosis, but its predictive value on lung disease progression has not yet been studied. We aimed to assess the value of serum CC16 concentration in predicting lung disease deterioration in patients with systemic sclerosis (SSc).
Methods. We prospectively analyzed and followed 106 patients with SSc during a 4-year period for the risk of developing combined deleterious event, defined as a 10% decrease in total lung capacity or forced vital capacity from baseline, or death, according to serum CC16 at inclusion. Receiver-operating characteristic (ROC) curve analysis was performed for prediction of events during the first 2 years after inclusion. Cumulative risks of combined events were computed by Kaplan-Meier analysis.
Results. The best cutoff level of serum CC16 for prediction of a combined event was 33 ng/ml, with 76% sensitivity and 65% specificity (area under the ROC curve: 0.71, 95% CI 0.61–0.81, p < 0.0001). Progression of lung disease evaluated by a mean time-to-event differed between patients with high baseline serum CC16 (42.8 mos, 36.3–49.3) and those with low serum CC16 (56.3 mos, 50.9–61.7; log-rank test, p < 0.001). After adjustment for age, duration of disease, clinical and lung function measures, the risk of combined event occurrence in patients with high serum CC16 was significantly higher than in those with low CC16 (HR 2.9, 1.2–6.75, p < 0.05).
Conclusion. High baseline serum CC16 predicts lung disease worsening in patients with SSc.
- SYSTEMIC SCLEROSIS
- INTERSTITIAL LUNG DISEASE
- CLARA CELL SECRETORY PROTEIN
- LUNG FUNCTION
- PROGNOSIS
Systemic sclerosis (SSc) is a life-threatening disease characterized by immune system activation, microvascular injuries, and excessive production of collagen, resulting in skin and internal organ fibrosis1. Interstitial lung disease (ILD) is frequently encountered and is currently the leading cause of mortality in SSc2. The pathophysiology of SSc is complex and not fully understood3. Current therapeutic strategies, based on immunosuppressive agents, showed only modest efficacy4. However, SSc patients with severe or worsening ILD may benefit greatly from such treatments5,6. It is important to identify patients at risk for lung function deterioration, for whom early therapy may reduce lung damage. Various autoantibody subsets, including anti-topoisomerase I (anti-topo I) and centromere antibodies, can identify patients with SSc who will likely develop SSc-related ILD and pulmonary arterial hypertension, respectively7,8,9. There is, however, only a very small number of studies investigating longterm predictors of ILD progression in SSc10,11.
Among lung-specific secretory proteins, also named pneumoproteins, Clara cell 16 kDa (CC16), secreted by bronchiolar Clara cells or Club cells, is one of the most abundant pneumoproteins. Its physiological role is incompletely known, but it may participate in bronchial epithelium defense mechanisms against oxidative stress and inflammation12. In acute or chronic lung injuries, CC16 was secreted into bronchoalveolar lavage fluid and rapidly leaked into the bloodstream because of its small size and molecular weight. Thus, serum CC16 can be used as a biomarker of lung diseases reflecting the lung epithelial damage12. Serum CC16 concentration was increased in idiopathic pulmonary fibrosis13 and sarcoidosis14, but decreased in smokers15 and patients with chronic obstructive pulmonary disease13,16. In their retrospective study of 92 patients with SSc, Hasegawa, et al found higher serum CC16 levels in patients with pulmonary fibrosis than in those without pulmonary fibrosis17. The result was recently confirmed by a Polish report in a small group of patients with SSc having extended pulmonary fibrosis, defined as high-resolution computed tomography (HRCT) with an extent of more than 20%18. However, there was no report investigating the predictive value of CC16 on the occurrence of lung function decline in patients with SSc.
We hypothesized that an increase in serum CC16 levels is a potential noninvasive marker of alveolar-capillary barrier disruption, reflecting early damage to the lungs in SSc. We therefore undertook this prospective study to determine the value of serum CC16 as a predictive marker of subsequent lung function deterioration in patients with SSc.
MATERIALS AND METHODS
The present study was approved by the Ethics Committees of Ile-de-France (approval number 05654). All patients had been informed about the study and signed the consent form.
Patients
Enrollment was conducted from January 2005 and November 2009 in the Department of Internal Medicine, Saint-Antoine Hospital (Paris, France), and the followup of this time-to-event–driven study ended in May 2012. Patients were considered for inclusion if they were aged ≥ 18 years and had SSc as defined by the American College of Rheumatology criteria19. Exclusion criteria were the presence of recent upper airway tract infection, pneumonia, or other systemic infection in the 3 months prior to the study; asthma; chronic obstructive pulmonary disease; idiopathic pulmonary fibrosis or ILD associated with connective tissue diseases other than SSc and creatinine clearance inferior to 30 ml/min/1.73 m2, according to the Modification of Diet in Renal Disease Study equation.
Study design
At the time of inclusion, these data were recorded: demographic data, duration of disease (delay between the first symptom attributable for SSc and the inclusion date), subset of SSc, modified Rodnan skin score (mRSS), autoantibody status including anticentromere (ACA) and anti-topo I antibodies, and Health Assessment Questionnaire (HAQ). All patients underwent thoracic HRCT, pulmonary function test (PFT), and echocardiography. PFT were routinely performed according to the American Thoracic Society and the European Respiratory Society (ATS/ERS) recommendations20. ILD was considered present if pulmonary HRCT demonstrated compatible changes in reticular or airspace opacities according to ATS/ERS consensus classification21.
The endpoint was the occurrence of combined deleterious events, defined as a 10% decrease in total lung capacity (TLC) or forced vital capacity (FVC) from baseline, or death22. The rationale for using these criteria as final outcome was first because pulmonary fibrosis is actually the most frequent cause of mortality (19%) in SSc2. Second, the cutoff value of 10% decrease in TLC or FVC was generally accepted as strong evidence of disease progression in pulmonary fibrosis22,23.
Survival data were obtained from visits and telephone interviews. Patients were annually evaluated by PFT, chest HRCT, and echocardiography (or earlier in an additional visit if symptoms appeared). Patients who had not been seen within 6 months were called to confirm their vitality.
Measurement of serum CC16 concentrations
Serum CC16 was quantified using Human Uteroglobulin Quantikine ELISA kits (DUGB00, R&D Systems Europe). All serum samples were diluted to 1:4. All samples were assayed in duplicate. Intraassay coefficients of variation of < 10% were accepted.
Statistical analysis
First, we assessed the optimal threshold and the diagnostic performance of CC16 for identifying patients with SSc who were going to have a combined deleterious event within the 2 years after inclusion by using receiver-operating characteristic (ROC) curve analysis. We used the 1000 bootstraps technique for internal validation.
Second, patients with SSc were equally categorized into 2 groups: patients with a serum CC16 level higher than the optimal threshold, and other patients. Characteristics for the group of SSc patients with serum CC16 ≤ optimal cutoff versus those with greater levels of serum CC16 were compared using Mann-Whitney U tests for continuous variables and chi-square tests for categorical variables. Continuous and categorical values are presented as mean ± SD and number (%), respectively. A p value of < 0.05 was considered statistically significant.
To estimate the predictive value of serum CC16 on lung function outcomes during the whole followup period, cumulative risks were computed by Kaplan–Meier analysis. We first conducted unadjusted Cox analyses according to each baseline characteristic separately. A multivariable Cox model further adjusted CC16 as categorical variable for age, FVC, DLCO, and serum creatinine as continuous variables, and smoking status, the presence of ILD, and immunosuppressive therapy as categorical variables.
RESULTS
Population characteristics
The study included 106 patients. Their mean age was 55.4 ± 12.7 years with SSc duration of 13.2 ± 12.6 years at inclusion (ranging from 10 mos to 58 yrs), and 89% were women. One-third of patients had diffuse cutaneous SSc and half of patients had ILD. Eighteen of the 106 patients received immunosuppressive therapies (cyclophosphamide, n = 4; azathioprine, n = 1; mycophenolate mofetil, n = 6; methotrexate, n = 6; corticosteroids ≥ 10 mg daily, n = 10; tacrolimus, n = 2). During the mean followup of 46 ± 20 months, 5 patients died. The composite event (10% decrease in TLC or FVC from baseline, or death) occurred in 41 patients (39%).
Measurement of baseline serum CC16 was performed in 106 patients and the results were presented in a scatter plot (Figure 1) according to the limited and diffuse cutaneous subtypes of SSc. There was no significant difference in serum CC16 levels between the 2 groups (diffuse form: n = 37, 43 ± 28 ng/ml and limited: n = 69, 48 ± 39 ng/ml; p = 0.8).
ROC curve analysis
The area under the ROC curve of serum CC16 for prediction of occurrence of the combined events was 0.71 (95% CI 0.61–0.81, p < 0.0001). The best cutoff level was 33 ng/ml (95% CI 18.9–50.7) with sensitivity, specificity, positive predictive value, and negative predictive value of 0.76 (95% CI 0.6–0.88), 0.65 (95% CI 0.5–0.76), 0.57 (95% CI 0.43–0.7), 0.81 (95% CI 0.67–0.9), respectively.
Characteristics of SSc population according to the best cutoff of serum CC16
Patients in the group with serum CC16 > 33 ng/ml were older (59.4 ± 11.6 vs 51.4 ± 12.5 yrs; p < 0.01) and had higher serum creatinine level (87 ± 22 vs 76 ± 12 µmol/l; p = 0.01). No statistically significant difference was seen between the 2 populations concerning sex ratio, duration of SSc disease, diffuse cutaneous SSc, mRSS, HAQ score, presence of ILD, FVC, TLC, DLCO, systolic pulmonary arterial pressure, smoking status, presence of ACA and anti-topo I antibodies, and immunosuppressive therapy (Table 1). The combined event occurred in 30 patients (57%) in the group with serum CC16 > 33 ng/ml and in 11 patients (2%) in the group with serum CC16 ≤ 33 ng/ml.
Association between serum CC16 level at baseline and subsequent deterioration in lung function or death
Progression of lung disease differed between patients with high and low baseline serum CC16 (Figure 2), with a mean time to event of 42.8 months (95% CI 36.3–49.3) and 56.3 months (95% CI 50.9–61.7), respectively (log-rank test; p < 0.001).
In an unadjusted Cox model (Table 2), serum CC16 > 33 ng/ml had a predictive value for lung function deterioration, with HR of 3.14 (95% CI 1.57–6.27, p < 0.01). The other baseline characteristics associated with lung function worsening were mRSS (HR 1.04, 95% CI 1–1.08, p < 0.05), ILD (HR 2.11, 95% CI 1.49–3.89, p < 0.05), and serum creatinine (HR 1.01, 95% CI 1–1.03, p < 0.05).
In the Cox proportional hazards model adjusted for age, duration of SSc, mRSS, smoking status, FVC, DLCO, presence of ILD, immunosuppressive therapy, and serum creatinine (Table 3), patients with baseline serum CC16 > 33 ng/ml had a higher risk of combined event occurrence (HR 2.9, 95% CI 1.2–6.75; p < 0.05).
DISCUSSION
In this prospective cohort study, we show that elevated serum CC16 > 33 ng/ml is a predictive marker of subsequent lung function deterioration or death in patients with SSc. Reliable markers of ILD outcome are currently lacking in patients with SSc. Pulmonary function tests are used in daily practice, but Assassi, et al24 showed that baseline FVC is not a predictive marker of subsequent decline in percentage predicted FVC over time. Because of their specificity to the lungs, pneumoproteins are potential markers of ILD in patients with SSc. Studies of surfactant protein D and Krebs von den Lungen-6 suggested that their elevation in serum could be a marker of ILD25. Reports evaluating the predictive value of CC Chemokine-ligand 18 for detecting lung disease worsening in patients with SSc, however, yielded conflicting results 11,26.
CC16 is a small pneumoprotein and is a sensitive marker of damage to the epithelium of the lungs12. We, therefore, hypothesized that elevation of serum CC16 could identify damage to the lungs that leads to lung function worsening in patients with SSc. To date, only 1 retrospective study has shown that elevated serum CC16 was associated with pulmonary fibrosis in patients with SSc and that serum CC16 increased significantly in patients with active pulmonary fibrosis, as defined by HRCT worsening and PFT decline during the 1-year followup17. A more recent study of 28 patients with SSc showed that serum CC16 was correlated with the severity of SSc-related ILD as assessed by HRCT and PFT18. None of these reports considered the predictive value of this novel biomarker in the progression of SSc-related ILD. Our prospective study with longterm followup (around 4 yrs) is the first, to our knowledge, to show that high baseline serum CC16 predicts lung disease aggravation in patients with SSc. This predictive power is independent of baseline characteristics such as age, smoking status, disease duration and severity at baseline, including PFT measures (FVC, DLCO) and presence of parenchymal lesions revealed on chest HRCT.
Intravascular leakage through damaged lung epithelium can directly influence serum CC16 concentration, but other mechanisms might also account for changes in CC16 levels in patients with SSc12. Owing to its low molecular weight, CC16 is eliminated by glomerular filtration, and renal insufficiency can lead to elevation in serum CC1617. Patients with severe kidney disease were thus excluded from this study. Tobacco smoking has been shown to decrease production of CC16 in the respiratory tract, resulting in lower serum CC1615,16. Smoking status was available for all patients included in our study. We included tobacco smoking and serum creatinine in our multivariate analysis and showed that high serum CC16 predicts lung disease worsening independently of these 2 variables. Overall, 41 patients (39%) met the composite events, including 10 deaths (9.4%), during the followup period. Our hospital is one of the national reference centers for SSc, usually receiving patients with severe SSc, which can explain in part the high rate of combined events.
We were unable to avoid some limitations when conducting our study. First, our study population had relatively longstanding disease, and it would be interesting to investigate the usefulness of CC16 in patients with newly diagnosed SSc. Second, we used internal validation to show the robustness of CC16 threshold for prediction of lung disease worsening. External validation in an independent cohort is needed before serum CC16 is recommended for use in daily practice. Our study, therefore, encourages additional investigation of this biomarker. Finally, the limited value of specificity and positive predictive value of serum CC16 for predicting deterioration of pulmonary function or death (0.65 and 0.57, respectively) might be due to some confounding factors that were not controlled and/or adjusted in our study such as diurnal variation27, atopic status28, exercise29, and inhaled pollutants30. These factors if present might contribute to increase the serum CC16 even if patients did not worsen their SSc-related pulmonary fibrosis. These factors should be well controlled in future studies investigating CC16 as a biomarker of chronic lung diseases in general and SSc-related ILD in particular.
Our findings showed that baseline serum CC16 has a significant predictive value for worsening of lung disease in patients with SSc. Further studies are needed to show whether serum CC16 could be used in association with other pneumoproteins to obtain a better predictive value in SSc and other chronic pulmonary diseases31.
Footnotes
Supported by grants from the Legs Poix, University of Paris, and the Air Liquide Foundation.
- Accepted for publication September 1, 2017.
REFERENCES
- 1.
- 2.
- 3.
- 4.
- 5.
- 6.
- 7.
- 8.
- 9.
- 10.
- 11.
- 12.
- 13.
- 14.
- 15.
- 16.
- 17.
- 18.
- 19.
- 20.
- 21.
- 22.
- 23.
- 24.
- 25.
- 26.
- 27.
- 28.
- 29.
- 30.
- 31.