Interferon-inducible chemokines reflect severity and progression in sarcoidosis

Abstract

Background: Identification of serum proteins that track with disease course in sarcoidosis may have clinical and pathologic importance. We previously identified up-regulated transcripts for interferon-inducible chemokines CXCL9, and CXCL10, in blood of sarcoidosis patients compared to controls. The objective of this study was to determine whether proteins encoded by these transcripts were elevated in serum and identified patients with remitting vs. chronic progressive sarcoidosis longitudinally.

Methods: Serum levels of CXCL9, CXCL10, and proteins associated with inflammation and/or disease activity (sIL2R, ACE, ESR and CRP) were measured in a prospective cohort of sarcoidosis subjects and controls. Comparisons were made between groups and clinical course using pulmonary function measures and a severity score developed by Wasfi et al.

Results: In a cross-sectional analysis of 36 non-immunosuppressed sarcoidosis subjects, serum CXCL9, CXCL10, and sIL2R were significantly elevated compared to 46 controls (p < 0.0001). CXCL9 and CXCL10 were strongly inter-correlated (p = 0.0009). CXCL10 and CXCL9 were inversely correlated with FVC% predicted and DLCO% predicted, respectively. CXCL10 and CXCL9 significantly correlated with sarcoidosis severity score. sIL2R, ESR, CRP, and ACE serum levels did not correlate with pulmonary function measures or severity score. In the longitudinal analysis of 26 subjects, changes in serum CXCL10 level over time corresponded with progression versus remission of disease.

Conclusions: Interferon-γ-inducible chemokines, CXCL9 and CXCL10, are elevated in sarcoidosis and inter-correlated with each other. Chemokine levels correlated with measures of disease severity. Serial measurements of CXCL10 corresponded to clinical course.

Figure 1

Comparison of serum CXCL9 (N = 30), CXCL10 (N = 36), and sIL-2R (N = 36) protein levels in non-immunosuppressed sarcoidosis subjects. Data are presented as median and interquartile range. *p < 0.0001.

Figure 2

Serum levels of CXCL10 and CXCL9 in non-immunosuppressed subjects with sarcoidosis (N = 30).

Figure 3

Relationship between interferon-inducible chemokines and pulmonary function testing. (A) Serum CXCL10 is inversely correlated with FVC% predicted and (N = 36) (B) serum CXCL9 is inversely correlated with DLCO% (N = 23) predicted in non-immunosuppressed subjects with sarcoidosis.

Figure 4

Relationship between interferon-inducible chemokines and clinical severity. (A) Serum CXCL10 (N = 36) and (B) CXCL9 (N = 30) positively correlated with a sarcoidosis severity score previously published by Wasfi et al.

Figure 5

Changes in serum CXCL10 levels correspond to sarcoidosis disease course. CXCL10 remains persistently elevated in patients with chronic sarcoidosis (N = 13). CXCL10 declines over time in patients with remitting sarcoidosis (N = 13). Longitudinal serum CXCL10 levels in 14 healthy controls are presented for comparison.

Figure 6

The effect of systemic immunosuppression (IS) on serum CXCL10. In remitting disease subjects who did not receive IS at baseline and followup (N = 13), CXCL10 declined spontaneously. In chronic disease, CXCL10 remained largely unchanged if immunosuppression was unchanged (N = 7), and increased if immunosuppression was decreased (N = 4). *denotes a statistically significant result using the non-parametric trend test.