Myeloma bone disease is characterised by osteolytic destruction not compensated by adequate new bone formation. It is know that the growth of myeloma in the bone marrow is associated with suppressed osteoblastic activity (Bataille et al, 1991) possibly through inhibition of the WNT signalling pathway (Tian et al, 2003). A 63-year-old woman with kappa light chain multiple myeloma, who relapsed after tandem autotransplant, received bortezomib 1 mg/m2 on days 1, 4, 8, and 11 on a 21-d cycle. Response was associated with a rapid increase in serum alkaline phosphatase (ALP) with otherwise normal liver function tests. This observation prompted analysis of three distinct protocols. In protocol UARK 2001-37 (Zangari et al, 2003), patients with resistant or relapsing multiple myeloma received either bortezomib 1 or 1·3 mg/m2, on days 1, 4, 8, and 11 for eight 3-week cycles. Thalidomide was added with the second cycle at increasing doses (50, 100, 150, and 200 mg/d) in cohorts of 10 patients. Dexamethasone (40 mg with each bortezomib dose) was added after two cycles for disease progression. In protocol M34100-025 (SUMMIT) (Richardson et al, 2003) patients with relapsed refractory myeloma received bortezomib 1·3 mg/m2 on days 1, 4, 8, and 11 every for eight 3-week cycles. If progression occurred after two cycles, or stable disease after four cycles, dexamethasone could be added (40 mg with each bortezomib dose). Protocol M34101-039 (APEX) (Richardson et al, 2004) patients with relapsed or progressive disease were randomised to bortezomib (1·3 mg/m2 on days 1, 4, 8, and 11 for up to eight 3-week cycles, followed by treatment on days 1, 8, 15 and 22 for up to three 5-week cycles) or to high dose dexamethasone (40 mg/d on days 1–4, 9–12 and 17–20 for up to four 5-week cycles, followed by dexamethasone on days 1–4 for up to five 4-week cycles). All patients in each analysis had baseline and follow-up normal liver function tests, and creatinine clearance >50 ml/min. Responses were assessed according to the European Group for Blood and Marrow Transplant criteria (Blade et al, 1998). Response was defined as the sum of complete response (CR) plus partial response (PR). The change from baseline in ALP was compared between responders and non-responders during treatment. In the APEX study, ALP levels were analysed by randomisation arm and by response category. To identify the source of ALP, bone-specific ALP (BALP) and intact parathyroid hormone (PTH) were measured in three patients treated with single agent bortezomib. Fisher's exact test was used to compare responses and the Wilcoxon Rank-Sum test was used to assess significance between groups and ALP level. Institutional Review Board-approved informed consent was obtained for all patients.
The prompt response to bortezomib observed in a 63-year-old woman with multiple myeloma was associated with a significant increase in alkaline phosphatase (ALP). After similar elevations were noted in patients responding to bortezomib, thalidomide, dexamethasone combination, ALP levels were analysed in two large bortezomib trials. A statistically significant elevation of ALP from baseline was observed in responding patients (complete and partial responders) within three cycles of therapy. The rise in ALP after bortezomib in three patients was explained by a parallel increase in bone-specific ALP and parathyroid hormone, suggesting that response to bortezomib in myeloma is closely associated with osteoblastic activation.
Twenty-four patients enrolled in the study were evaluable for analysis. Change in ALP from the baseline value after bortezomib ranged from −20% to 500%, with a median of 40%. In 12 patients (group A, responders) ALP increased ≥40% from baseline levels, during the first (∼3) cycles; the remaining 12 patients (group B, non-responders) did not show such an ALP increase. Paraprotein and/or bone marrow plasmacytosis reduction of 50% was observed in 15 patients (72·5%). An increase in ALP (≥40%) was associated with a significantly higher response rate (92% in group A versus 33% in group B, P = 0·009).
Millennium trial M34100-025 (SUMMIT)
Seventy-seven patients were evaluable. The responder group included 31 patients while 46 patients had <PR. The difference in median ALP between these two groups during and upon completion of three cycles of therapy was statistically significant (week 8 P = 0·0015, responder range 62–837) (Fig 1A).
Millennium trial M34101-039
A total of 422 patients were eligible for analysis. A total of 217 patients were randomised to bortezomib, 205 to dexamethasone. The increase of ALP levels between responder and non-responder was statistically significant within the bortezomib arm at week 3 (P = 0·014), week 6 (P = 0·002, responder range 31–272) and week 9 (P = 0·036) (Fig 1B). Analysing only responders in both arms, we observed significantly higher median ALP elevation in the bortezomib arm compared with the dexamethasone arm (P < 0·01, responder range 31–272) (Fig 1C). No difference in ALP elevation was observed in dexamethasone-responding versus dexamethasone-non-responding patients.
Increased ALP levels and increased PTH activity
Bone specific ALP and PTH were prospectively measured in three patients relapsing after transplantation and treated with single agent bortezomib. The profound responses observed were associated with elevations of ALP and parallel increase of BALP and PTH (Fig 2).
By protecting myeloma cells from apoptosis and inducing drug resistance, the bone marrow microenvironment is a key factor in myeloma survival (Hardin et al, 1994). Histologic studies of bone have revealed not only excessive bone resorption but also marked inhibition of bone formation in the vicinity of the myeloma cells (Masahiro et al, 2004). In rodents, the ubiquitin-proteasome pathway exerts exquisite control of osteoblast differentiation and bone formation. When administered systemically to mice, the proteasome inhibitors epoxomycin and proteasome inhibitor-1 increase bone volume and bone formation rates >70% after only 5 d of treatment (Garrett et al, 2003).
Our original observation of an increased ALP in patients responding to bortezomib (Zangari et al, 2003) has now been confirmed in two independent large data sets. These results indicate that the anti-myeloma activity of bortezomib is associated with osteoblast activation. Maximum ALP elevation occurred within the first three cycles of therapy and was typically correlated with excellent tumour reduction. An alternative explanation of our findings would be that ALP was released by apoptosing myeloma cells. This hypothesis, however, would not explain the simultaneous increase in PTH and the lack of increase in ALP in dexamethasone-responsive patients. Because most patients had been previously exposed to bisphosphonates, our experience confirms that osteoblastic activation is not abrogated by the use of bisphosphonates (Clark et al, 2000). This retrospective analysis will require confirmation by prospective studies. Activation of osteoblasts by PTH or other compounds could result in a similar anti-myeloma effect, although these agents do not have direct anti-myeloma activity.
The authors wish to thank all the investigators, clinical teams and the patients who contributed to the data collected for the UARK 2001-37, SUMMIT and APEX clinical trials.