Serum levels of syndecan-1 in B-cell chronic lymphocytic leukemia: correlation with the extent of angiogenesis and disease-progression risk in early disease.

Syndecan-1 is a transmembrane proteoglycan generally not expressed in mature B-cell neoplasias like chronic lymphocytic leukemia (CLL). Moreover, information dealing with the evaluation of soluble syndecan-1 in CLL are lacking. We measured syndecan-1 concentrations in serum drawn at the time of diagnosis from 67 B-cell CLL patients (Binet stage A, 46; stage B, 7; stage C, 14). For this purpose a syndecan-1 enzyme-linked immunosorbent assay (ELISA, Diaclone, Besancon, France) was used. Detectable levels of syndecan-1 were found in all patients, although serum concentrations were significantly lower in CLL patients in comparison to age- and sex-matched controls (P=0.02; Mann-Whitney test). No correlation was found with Binet clinical stages (P=0.796), beta 2-microglobulin (P=0.923), hemoglobin level (P=0.605), platelet count (P=0.992) and lymphocyte doubling time (P=0.709). Only an association with absolute peripheral blood lymphocytosis (PBL) (P=0.01) and LDH (P=0.05) could be detected. Serum levels of syndecan-1 did not parallel those of several angiogenic cytokines such as vascular endothelial growth factor ( VEGF) (P=0.963), basic fibroblastic growth factor (FGF-2) (P=0.216), angiogenin (P=0.478), metalloproteinase-9 (MMP-9) (P=0.125) as well as bone marrow (BM) microvessel density (P=0.110). The same applied with adhesion molecules such as CD54 (P=0.233), CD108 (P=0.799), CD44 (P=0.816) and CD31 (P=0.508). Interestingly, the inverse correlation (r=-0.4967; P=0.03) between serum concentrations of syndecan-1 and plasma levels of stromal derived growth factor-1 (SDF-1) is in keeping with the different function, respectively, pro- and antiapoptotic, of these molecules. In 46 Binet stage A patients, serum levels of syndecan-1 were further evaluated as a dichotomous variable with respect to progression-free survival (PFS), an end-point surrogate for overall survival in early B-cell CLL. The best separation of curves was seen with a cut-off point at the median value of syndecan-1 (i.e. 36.5 pg/ml). Median PFS was not reached in the patient group with low syndecan-1, compared to a median of 34 months observed in the remaining patients (P=0.018; HR=0.208; 95% CI=0.115-0.816). At the multivariate analysis performed including variables significant in the univariate analysis [i.e. PBL (P=0.03) and syndecan-1 (P=0.01)], only syndecan-1 retained a trend of significance (P=0.08). Despite the pro- angiogenic activity of syndecan-1 which mediates FGF-2 binding and activity, no correlation with either angiogenic cytokines or the extent of BM angiogenesis was found in CLL. The inverse correlation with plasma levels of SDF-1 suggests an involvement in the processes leading to apoptosis. Finally, our results highlight the involvement of syndecan-1 in the mechanisms of disease-progression of early CLL.