Predictive value of alkaline phosphatase for response and time to
progression in bortezomib-treated multiple myeloma patients
Maurizio Zangari,1*Dixie Esseltine,2Federica Cavallo,1Rachel Neuwirth,2Francesca Elice,3
Michael J. Burns,1Shmuel Yaccoby,1Paul Richardson,5Pieter Sonneveld,4and Guido Tricot1
1The Myeloma Institute for Research and Therapy, The University of Arkansas for Medical Sciences, Little Rock, Arkansas
2Millennium Pharmaceutical Company, Cambridge, Massachusetts
3Department of Hematology, San Bortolo Hospital, Vicenza, Italy
4Department of Hematology, University Hospital Rotterdam, Rotterdam, The Netherlands
5Dana-Farber Cancer Institute, Boston, Massachusetts
Myeloma bone disease is characterized by osteolytic destruction associated with suppressed osteoblastic
activity. Using data from the APEX (Richardson et al., N Engl J Med 2005;352:2487–2498) study, we have
assessed the relationship of changes in alkaline phosphatase (ALP) levels during bortezomib therapy with
response and time to progression on this therapy. The percentage of ALP increments in responders (com-
plete and partial response) and nonresponders was analyzed at different thresholds and time points. For all
bortezomib-treated patients enrolled in the trial (N 5 333), at least a 25% increase in ALP from the baseline
at 6 week was the most powerful predictor of treatment response (P < 0.0001) and time to progression (206
vs. 169 days) relative to patients with less than a 25% increase in ALP (P 5 0.01). Markers of osteoblastic
activation may predict quality and duration of response in multiple myeloma. In addition, our data suggest
that bone anabolism could inhibit myeloma growth. Am. J. Hematol. 82:831–833, 2007.
C 2007 Wiley-Liss,
Multiple myeloma (MM) is typically associated with osteo-
lytic bone disease. Changes in bone remodeling precede
the progression of MM [1,2]. In more advanced stages of
the disease, the number of osteoclasts increases while that
of osteoblasts decreases, resulting in an uncoupling of the
processes of osteoclastic resorption and osteoblastic bone
formation [3,4]. Experimental studies suggest that bone dis-
ease drives the progression of MM. Cocultures of osteo-
blast-derived mesenchymal stem cells affect bone mineral
density and MM growth in myelomatous human bones of
SCID-hu mice, suggesting that osteoblasts may control
tumor growth . We previously reported the association
between osteoblast activation and bortezomib response
. Bortezomib-responsive patients (those with complete
response, CR + partial response, PR) enrolled in the APEX
trial showed a statistically significant elevation of alkaline
phosphatase (ALP) compared to the other individuals. In
contrast, ALP levels remained stable throughout the dura-
tion of the treatment in the dexamethasone-responsive
patients (CR + PR) enrolled in the control arm of the trial.
Other groups have also described that specific bone
markers such as bone-specific ALP and osteocalcin are
increasing during bortezomib therapy . We now report
the correlation between ALP changes, not only with
response but also with time to progression on bortezomib
Materials and Methods
The APEX clinical trial enrolled patients with relapsed or progressive
MM and randomly assigned them to bortezomib (1.3 mg/m2on 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/day 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). Patients had been treated with frontline therapy or
up to three prior therapies. The details of the trial have been published
. Responses were assessed according to the European Group for
Blood and Marrow Transplant criteria . Complete plus partial res-
ponders were considered responders. Percent changes in ALP levels
from baseline were analyzed at different time points in both responsive
and nonresponsive patients in the bortezomib arm. Institutional Review
Board-approved informed consent was obtained for all patients.
A Cochran–Mantel–Haenszel x2test was used to com-
pare response rates with adjustment for stratification factors
(number of previous treatments, time to progressive dis-
ease after the last treatment, and b2-microglobulin values).
A logistic regression model was used to assess the rela-
tionship between response rate (CR/PR) and percent
changes in ALP levels from baseline. Time to tumor pro-
gression was estimated with the Kaplan–Meier product limit
method. Differences in time to tumor progression between
ALP percent change categories were compared with the
use of the stratified log-rank test.
Results and Discussion
The characteristics of the overall study population have
been described . A total of 333 patients were included in
this analysis with a median age of 60 years, of which 59%
were male. No differences in baseline prognostic factors
were observed between responders and nonresponders. In
the APEX trial, significant elevation in serum ALP in Borte-
zomib-responsive patients (those with complete and partial
*Correspondence to: Maurizio Zangari, MD, 4301 W Markham Street, Slot
776, Little Rock, AR 72205. E-mail: firstname.lastname@example.org
Received for publication 22 January 2007; Revised 13 March 2007; Accepted
16 March 2007
Am. J. Hematol. 82:831–833, 2007.
Published online 1 June 2007 in Wiley InterScience (www.interscience.
C 2007 Wiley-Liss, Inc.
American Journal of Hematology
response) has been previously reported . A statistically
significant difference in response was observed even with a
minimum of 10% increment of ALP from baseline at Week
6 from enrollment (Table I). A 25% increase in ALP (N ¼
105) at Week 6 was the strongest indicator associated with
quality of response (P < 0.0001) and also with the time to
progression (206 vs. 169 days) relative to patients with less
than a 25% increase in ALP (N ¼ 228; P ¼ 0.01; Fig. 1).
The median ALP value recorded at the time of relapse was
not statistically different from the one recorded at the time
of enrollment (at baseline before bortezomib therapy start).
Even in the early stages of myeloma development, there is
evidence of the close interaction between the microenviron-
ment and disease activity. Histomorphometric studies sug-
gest that changes in bone remodeling precede disease pro-
gression to overt myeloma. Coupled increases in the num-
ber of osteoblasts and osteoclasts have been identified
even in patients with monoclonal gammopathy . This
report is the first clinical study showing that a bone-related
marker (elevation of ALP 25% from baseline) is strongly
associated not only with myeloma response to bortezomib,
but also, probably more importantly, with time to disease
progression. Positive clinical outcome in patients with
relapsed and refractory MM treated with bortezomib ap-
peared to be independent of many patient and disease fac-
tors historically considered to be adverse prognostic factors
in myeloma .
Myeloma research has been mostly focused on the my-
eloma-induced osteoclastic activity. Our experience sug-
gests that increased osteoblastic activity closely correlates
with disease control in bortezomib-treated patients, sug-
gesting that agents promoting osteoblast differentiation may
have potential clinical application in the treatment of this
1. Bataille R, Chappard D, Marcelli C, et al. Recruitment of new osteoblasts and
osteoclasts is the earliest critical event in the pathogenesis of human multiple
myeloma. J Clin Invest 1991;88:62–66.
2. Taube T, Beneton MN, McCloskey EV, Rogers S, Greaves M, Kanis JA.
Abnormal bone remodelling in patients with myelomatosis and normal
biochemical indices of bone resorption. Eur J Haematol 1992;49:192–
3. Barille-Nion S, Bataille R. New insights in myeloma-induced osteolysis. Leuk
4. Roodman GD. Mechanisms of bone metastasis. N Engl J Med 2004;350:
TABLE I. Probability of Response by % of ALP Elevation at Week 6
% of ALP
(CR + PR)
(N ¼ 128)
(N ¼ 153) ORP value
10 < % < 25
aValues in parentheses indicate percentages.
bValues in square brackets indicate 95% CI.
Based on APEX bortezomib patients who are evaluable for response
and have baseline and Week 6 values. Logistic regression model; OR:
odds of response relative to the ?10% change.
Kaplan-Meier product limit method. A 25% increase in ALP level by week 6 in bortezomib treated patients (total) was
associated with time to progression (206 vs. 169 days) relative to patients with less than a 25% increase in ALP (N 5
228; P 5 0.01). [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]
Time to progression by 25% increase in ALP level. Time to tumor progression was estimated with the
832 American Journal of Hematology DOI 10.1002/ajh
5. Yaccoby S, Wezeman MJ, Zangari M, et al. Inhibitory effects of osteoblasts
and increased bone formation on myeloma in novel culture systems and a
myelomatous mouse model. Haematologica 2006;91:192–199.
6. Zangari M, Esseltine D, Lee CK, et al. Response to bortezomib is associated
to osteoblastic activation in patients with multiple myeloma. Br J Haematol
7. Heider U, Kaiser M, Mu ¨ller C, et al. Treatment of bortezomib increases osteo-
blast function in patients with multiple myeloma. Blood 2005;106:3457a.
8. Richardson PG,SonneveldP, Schuster MW, et al. Bortezomib or high-dosedexa-
9. Blade ´ J, Samson D, Reece D, et al. Criteria for evaluating disease response and
progression in patients with multiple myeloma treated by high-dose therapy and
haemopoietic stem cell transplantation. Br J Haematol 1998;102: 1115–1123.
10. Politou M, Terpos E, Anagnostopoulos A, et al. Role of receptor activator of
nuclear factor-kappa B ligand (RANKL), osteoprotegerin and macrophage
protein 1-alpha (MIP-1a) in monoclonal gammopathy of undetermined signifi-
cance (MGUS). Br J Haematol 2004;126:686–689.
11. Richardson PG, Barlogie B, Berenson J, et al. Clinical factors predictive of
outcome with bortezomib in patients with relapsed, refractory multiple my-
eloma. Blood 2005;106:2977–2981.
American Journal of Hematology DOI 10.1002/ajh