Alterations in Platelet Levels of Brain-Derived Neurotrophic Factor Are Associated with Clinical Response to Bortezomib in Patients with Multiple Myeloma

Introduction: Brain-derived neurotrophic factor (BDNF) was previously shown to support the proliferation of human multiple myeloma (MM) cell lines and promote their resistance to bortezomib (Pearse et al, 2005). As platelets appear to be the main transportation system of BDNF in the human circulation (Fujimura et al, 2002), we hypothesized that early alterations in platelet derived BDNF levels may be associated with clinical response to bortezomib in patients with MM.

Methods: The total BDNF levels as well as the levels of its precursor peptide (proBDNF) were measured by ELISA (DY248 and DY3175 Dou set ELISA kits, respectively, R&D Systems, Minneapolis, MN) in serum, platelet-poor plasma (PPP), and platelet lysate that were isolated from MM patients at diagnosis and along their treatment course with bortezomib-based protocol. Relative expression of precursor and mature forms of BDNF in the platelet lysates were analyzed by Western blot gel electrophoresis. BDNF levels at diagnosis were compared between patients who achieved complete response (CR) to patients with less then partial response (LTPR) to bortezomib treatment. The differential effect of bortezomib treatment between the two groups is presented by the calculated delta between BDNF levels at diagnosis and after 2 treatment cycles.

Results: Twelve MM patients were analyzed. Five patients achieved CR to bortezomib and 7 patients showed LTPR to bortezomib. The overall BDNF levels were 6-22 ng/ml (mean±SD 13.08±5.9), 0-10.65 ng/ml (mean±SD 2.49±2.67) and 2.27-23.04 ng/mg total protein (mean±SD 8.77±5.68) for serum, ppp and platelet lysates respectively. proBDNF was not detected in patients samples either by ELISA or by Western blot gel electrophoresis, suggesting that total BDNF is comprised mainly of the mature form of BDNF. Significantly higher BDNF levels were found in the serum and platelet lysates, obtained at diagnosis from patients who later achieved CR compared to patients with LTPR (serum: 19.94±1.84 ng/ml vs. 8.46±6.78ng/ml , p<0.05; platelet lysate: 13.79±6.31 ng/ml vs. 4.25±1.41 ng/ml p<0.01). No difference was found in the levels of BDNF in the ppp (2.18±1.6 ng/ml vs 2.72±2.03 ng/ml p=n.s.). Analysis of the delta BDNF levels revealed a decrease in BDNF levels in the serum, plasma, and platelet lysates of patients with CR and an increase of BDNF levels in patients with LTPR (serum: -7.05±2.53 ng/ml vs. +2.66±8.09 ng/ml in LTPR, p<0.05; Delta BDNF in ppp: -1.29±1.85 ng/ml vs. +1.29±2.47 ng/ml p<0.05; platelet lysate: -7.83.±4.57 ng/ml vs. +3.60±5.21 ng/ml, p<0.01).

Discussion and conclusions: Our preliminary data of higher platelet BDNF levels at diagnosis together with therapy related reduction of BDNF levels in patients achieving CR may indicate that myeloma activity in responding patients is highly dependent on BDNF. Validating these preliminary observations in a large number of patients may help to establish a new clinical tool which facilitates the ability to identify bortezomib resistance at earlier step of therapy.