Effects of bortezomib on bone disease in multiple myeloma

Abstract

Bone disease is a well-recognized feature of multiple myeloma (MM), with skeletal-related complications (including pathologic fractures, spinal cord compression, hypercalcemia, and pain) affecting greater than 80% of patients. For many patients, a pathologic fracture or severe bone pain resulting from bone destruction within the marrow cavity is the sentinel event heralding a diagnosis of MM [1]. Importantly, bone disease in MM differs from that of other malignancies. Usually, osteoclast-mediated bone resorption is closely followed both spatially and temporally by osteoblastmediated bone formation. In MM, however, this coupling does not occur even following years of complete remission. Rather bone disease in MM results both from the increased proliferation and activation of osteoclasts and the suppression or absence of osteoblast activity [2]. This finding explains the clinical observation that bone scans (which detect bone formation) in approximately half of patients with MM are normal despite the presence of severe osteolytic disease. Intravenous bisphosphonates (pamidronate and zoledronic acid) bind to bone at sites of exposed hydroxyapatite to limit osteoclast-mediated bone resorption and are the current primary therapy for MM bone disease [3]. Despite bisphosphonate therapy, however, 50% of patients experience a skeletal-related event at the time of relapse [4]. Importantly, however, bisphosphonates target only mature osteoclasts and do not appear to affect osteoblast activity [5]. Further, bisphosphonate therapy is not without risk as MM patients have the highest incidence (among all groups of patients with malignancies receiving bisphosphonate therapy) of avascular osteonecrosis of the jaw [6]. Given these limitations, new strategies to improve skeletal-related outcomes are warranted. In this issue of the American Journal of Hematology, Pennisi et al. [7] demonstrate a skeletal anabolic effect for bortezomib, a first in class proteasome inhibitor with antineoplastic activity in both newly diagnosed and relapsed MM. The authors report that bortezomib (but not melphalan) treatment of immunodeficient mice engrafted with human myeloma cells increased bone formation in previously implanted nonfetal rabbit bone, as documented by both radiographic and bone histomorphometric analyses. The increase in bone formation was associated with an increase in osteoblast numbers along with a simultaneous reduction in the number of osteoclasts. Importantly, however, significant heterogeneity in both the skeletal anabolic and MM cell responses to proteasome inhibition was found. Thus, increases in bone mass were not uniformly found in all subjects treated with bortezomib, and changes in bone mass did not appreciably correlate with the changes in MM tumor cell burden (i.e., subjects with a significant anti-MM cell response to bortezomib may have had little or no bone anabolic response). Accordingly, though the data presented are interesting, it is important to place them in the context of our current understanding of MM bone disease. Over the past several years, the importance of the bone marrow microenvironment in the establishment and progression of MM and MM bone disease has received increasing attention. In this context, the report by Pennisi et al. follows a growing body of literature examining the role of bortezomib in modulating the activity of primary bone cells, namely osteoblasts and osteoclasts. Although the mechanism(s) by which proteasome inhibition leads to increased osteoblast activity has not been firmly established, previous work suggests that proteasome inhibition decreases proteolytic degradation of the Gli3 protein, leading to increased bone morphogenetic protein-2 (BMP-2) expression in osteoblastic cells [8]. Other studies have shown that proteasome inhibition can also increase the activity of the osteoblast transcription factor Runt-related transcription factor 2 (Runx2)/core binding factor a1 (Cbfa1) [9]. These studies are consistent with a recent report by Mukherjee et al. which demonstrated that in mice, bortezomib induces the Runx2/Cbfa1-dependent differentiation of mesenchymal stem/progenitor cells into osteoblasts, leading to new bone formation [10]. In addition, bortezomib also appears to decrease levels of the Wnt pathway inhibitor DKK-1 in patients with MM [11]. As demonstrated in preclinical mouse models of MM, increases in Wnt signaling (and hence osteogenic differentiation) lead to decreased MM bone disease and tumor growth within the bone marrow microenvironment [12,13]. In total, these results are consistent with the increases in bone formation markers demonstrated to occur in MM patients treated with bortezomib [14,15]. As further shown by Pennisi et al., bortezomib also appears to influence osteoclast activity. Although the molecular mechanism(s) by which bortezomib inhibition of proteasome activity leads to decreased osteoclast activity is somewhat unclear, inhibition of nuclear factor-kappa B (NFjB) signaling (perhaps through decreases in TNF Receptor Associated Factor 6 (TRAF6) levels [16] or stabilization of inhibitor of NF-jB (IjB) levels [17]) appears to be important. In turn, NF-jB inhibition serves to limit both the ability of osteoclast precursors to differentiate into mature osteoclasts and the capacity of mature osteoclasts to resorb bone [18]. Importantly, Terpos et al. have shown that bortezomib can reduce serum levels of receptor activator of NF-