Serum and urinary galectin-9 and C-X-C motif chemokine ligand 10

Abstract

Background: Systemic lupus erythematosus (SLE) is characterized by a type I interferon (IFN) signature, and traditional methods for its measurement like gene expression analysis are cumbersome for routine use. Thus, we aimed to study galectin-9 as a biomarker and compared it with a validated marker, C-X-C motifchemokine ligand 10(CXCL-10). Methods: Ninety-seven patients with SLE (26 years; 89 females) were included and stratified based on renal involvement and activity into - active (SLEDAI > 4) renal (35), active non-renal (32) and inactive renal subgroups (30) along with 20 healthy controls (HC, 25 years; 15 females). The median disease duration was 24 months (6-48), and SLEDAI 2K was 9 (2-15). Serum and urine galectin-9 and CXCL-10 levels were measured by ELISA. Urine levels were normalized with spot urine creatinine values. Follow-up serum and urine galectin-9 levels were measured for those in the active renal group at 6 months. Results: Patients with SLE had higher serum galectin-9 (5.6 vs 1.7 μg/mL, p = .0001) but not urine galectin-9 (0.52 vs 0.32 μg, p = .7) levels as compared to HC. Serum galectin-9 but not urine galectin-9 was higher in patients with active as compared to inactive lupus (12.9 - active renal, 16.7 - active non-renal vs 3.57 μg/mL, p = .04 and .005). Serum CXCL-10 (0.16 vs 0.05, p = .01) and urine CXCL-10 (0 vs 0, p = .01) were both significantly higher in the SLE group as compared with HC. Serum but not urine CXCL-10 was higher in the active as compared to inactive lupus (0.2 - active renal, 0.3 - active non-renal vs 0.08 μg/mL, p = .9 and .02). Serum galectin-9 showed a modest correlation with CXCL-10 0.4 (0.2-0.6), whereas none was found between their urine levels.Serum galectin-9 and CXCL-10 showed a moderate positive correlation with SLEDAI 2K. Serum galectin-9 showed a greater AUC than CXCL-10 (0.77 vs 0.67) in differentiating active from inactive SLE, and both tested together had the best AUC of 0.82. However, urinary levels had no association with SLEDAI 2K or renal SLEDAI. In a subset of patients with active renal disease, serum galectin-9 but not urine levels declined significantly after 6 months. Conclusion: Serum galectin-9 is a good marker of lupus activity; however, it does not differentiate between active renal and active non-renal disease. It performs slightly better than CXCL-10. Urinary galectin-9 does not reflect renal activity.