To the Editor

Osteopontin (Opn) is a non-collagenous matrix protein responsible for the migration and attachment of osteoclasts to mineral matrix of bone surfaces [1]. In vitro studies suggest that Opn and vascular endothelial growth factor (VEGF) cooperatively enhance angiogenesis in MM [2, 3]. We compared the serum levels of Opn with those of established markers of angiogenic activity and bone destruction in 50 (23 males, median age 68 y.o.) untreated MM patients and 25 healthy age & sex-matched blood donors (median age 65 y.o.). In 25 patients, measurements were repeated after reaching plateau phase with chemotherapy. Ethics approval and informed consent from all subjects were obtained.

Osteopontin and VEGF were measured with a solid-phase sandwich enzyme-linked immunosorbent assay (Quantikine®, R&D Systems Inc. Minneapolis MN, USA). Bone marrow cellularity and percentage of infiltration by myeloma cells was estimated in all patients. Microvessel density (MVD) was assessed on bone marrow sections after staining endothelial cells with the anti-CD34 antibody (Immunotech, Marseille, France) [4]. The N-terminal propeptide of procollagen type I (Ntx) in urine was measured by a competitive inhibition ELISA (Ostex International, Seattle, WA, USA) [5]. Classification and regression tree (CART; Salford Systems, San Diego, CA, USA) analysis was used to identify the Opn levels best associating with disease parameters.

Opn among MM patients [median (range) 85 (5-232) ng/ml] was higher than controls [36 (2-190) ng/ml] but the difference did not reach statistical significance. However, Opn was significantly higher in 35 patients with Stage II or III compared with 15 Stage I patients (p < 0.001) (Figure 1). Similarly, there was a marked difference in Opn according to disease Grade (Figure 2), with increasing levels from Grades 0 to 3 (p = 0.006).

Figure 1
figure 1

Serum osteopontin levels (ng/ml) according to Durie-Salmon stage in multiple myeloma patients. Columns: 25-75% of values; bold line: median; whiskers: 95% confidence intervals. (* = p < 0.01 by Kruskal Wallis test).

Full size image
Figure 2
figure 2

Serum osteopontin levels (ng/ml) according to grade of bone disease (low grade = grades 0 and 1; total 23 patients, versus high grade = grades 2 and 3; total 27 patients). Columns: 25-75% of values; bold line: median; whiskers: 95% confidence intervals. (* = p < 0.01 by Kruskal Wallis test).

Full size image

CART analysis determined Opn >100 ng/ml as predictor of advanced MM disease. All 17 patients with Opn >100 ng/ml had advanced stage (II or III), whereas all 15 stage I patients had Opn <100 ng/ml (Odds Ratio 1.5, 95% Confidence Intervals 1.2-2.0; p = 0.036 Fisher's Exact test). Similarly, high grade (2 or 3) MM was associated with Opn >100 ng/ml (OR 1.9, 95% CI 1.2-2.9, p = 0.001) (Table 1).

Table 1 Association of serum osteopontin levels with: A) stage, and B) grade of Multiple Myeloma
Full size table

Opn correlated significantly with VEGF, MVD, Ntx and disease infiltration (Spearman's rho 0.47, 0.30, 0.35 and 0.45 respectively; p < 0.05). In 25 patients, Opn decreased from a median 78 ng/ml to 45 ng/ml post-chemotherapy (p < 0.05). MVD, Ntx, VEGF and infiltration by myeloma cells were also significantly reduced post-treatment (p < 0.001).

Our results add to the weight of data supporting a relationship between circulating Opn and bone marrow angiogenesis in myelomatous disease. We have defined a cut-off value (100 ng/ml) strongly associated with advanced MM disease and grade of bone destruction. This may prove useful in patient stratification. The applicability of our findings in different patient populations (i.e. genetic background, treatments used, comorbidities, etc) awaits validation.

Conflict of interest

The authors declare that they have no competing interests.