XIth International Myeloma Workshop,
Kos, 26/6/2007
The r
ole
role o
f
of b
iochemical
biochemical markers
o
f
of bone
metabolism in multiple myeloma.
Adverse events of bisphosphonates
administration.
Evangelos Terpos, MD
Department of Hematology & Medical Research,
251 General Air F
orce
Force Hospital Athens
,
Greece
,
&
Hon. Senior Lecturer, Faculty of Medicine
Imperial College London, London, UK
Myeloma bone disease:
bi
b ol
o ogy
Osteoblasts
OPG
bALP,
Myeloma cells
(-)
osteocalcin
RANKL
CD138
DKK1, sFRP-2
4
4 1 integrin
1
RANKL
RANKL
VCAM-1
VCAM-1
BMSCs
RANKL
IL-11, IL-1,
(-)
bFGF
OPG
IL-6
TNF, M-CSF
RANK
MIP-1, MIP-1,
SDF-1, IL-3,
HGF, OPN
Osteoclast
Activated osteoclasts
precursor
TRACP-5b
Collagen type-1
Terpos E & Dimopoulos MA.
degradation products:
Bone resorption
Ann Oncol 2005;16:1223-31
NTX, ICTP, CTX
Why do we use bone markers for the assessment
d
an monitit i
or ng MM
MM b
d
one i
di
?
sease
· Bone lesions
do
not
normally
heal
even
if
MM
goes into remission (osteoblast dysfunction).
· Radiographs frequently do
not
indicate
increased bone resorption in MM progression.
· BMD measurements are often not informative for
bone disease status in MM.
· Bi
Bioch
i
em cal mark
f
ers o b
one metabolilism have
been used in an effort to better monitor the
myeloma bone disease and
improve
assessment of disease progression.
iochemical markers of bone remodeling
Formation
Resorption
Calcium
bALP
TRACP
Collagen type I
BSP
propeptides
OH-proline
OH-lysine-glycosides
Pyridinium crosslinks
Osteocalcin
Collagen type l degradation
products:
p
- and C- terminal cross-linking
telopeptides of collagen type I
(X-CTX)
Osteoblasts
C-terminal cross-linking
telopeptide o
f
of t ype
type-I c
ollagen
collagen
Bone matrix
generated by MMPs (ICTP)
Osteoclasts
Adapted from Lipton et al, Clin Lymphoma Myeloma 2007;7:346-53
Assembly of pro-a collagen and processing into
tropocollagen
Assembly of tropocolla
p
gen into fibers
g
Bone collagen degradation products
Due to bone specificity and their unique
characteristics NTX, ICTP, and CTX have almost
totally replaced the use of older resorption indices
in the diagnostic assessment of bone diseases.
Urinary NTX in myeloma bone disease
250
38
p<0.001
87
200
18
34
r
eatinine)
150
93
150
108
BCE/mM c
100
21
2
(
nM
NTX
50
0
N =
45
10
28
83
Controls
group A (no lytic lesions)
group B (1-3)
group C (>3 or #)
GROUP
Terpos et al, Blood 2003;102:1064-9
Serum TRACP-5b and myeloma bone disease
20
15
p<0.0001
L
)
10
ACP-5b (U/
R
T
5
0
A (n=10)
B (n=28)
C (n=83)
Bone Disease
Terpos et al, Int J Cancer 2003;106:455-7
Serum ICTP and urinary NTX are elevated
prior to SRE in
p
patients with
p
pro
p gressive MM
g
No new SRE
New SRE - PD
· No Serum ICTP or
uNTX increases in
patients with no
progression
· ICTP and/or NTX
elevated before
development of
osteolytic events in
all but 1 case
Abildgaard et al, Haematologica 2004;89:567-77
MM patients with moderate or high urinary NTX
levels are
a
t
at H
igher
Higher Risk of SREs
P value
1 7
. 5
75
All SREs
.016
2.25
.032
1st SRE
1.87
.083
3.01
.008
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
Relative risk of event versus low NTX group
NTX 50 - 100 vs. NTX < 50 nM BCE/mM creatinine
NTX > 100 vs. NTX < 50 nM BCE/mM creatinine
Coleman et al J Clin Oncol 2005;23:4925-35;
Lipton et al, Clin Lymphoma Myeloma 2007;7:346-53.
Elevated baseline NTX correlates with shorter
time to
1st
1 SRE a
mong
among ZOL
-
ZOL treated MM
Patients
NTX, E (128 AT RISK, 67 EVENTS)
70
NTX, N (82 AT RISK, 34 EVENTS)
60
P = .0152
50
SRE, %
1
40
s with
30
20
Patient
10
0
0
3
6
9
12
15
18
21
24
Time since randomization, months
E = NTX > 50 nmol/mmol creatinine; N = NTX 50 nmol/mmol creatinine.
Unpublished data; Analysis by R. Cook and Drs J. Berenson and E. Terpos.
Elevated baseline NTX correlates with shorter
survival among
ZOL
treated
-
MM
Patients
NTX, E (128 AT RISK, 48 DIED)
NTX N
, (82
A
T
AT R
ISK
RISK 1
, 7
17 DIED)
DIED)
30
P = .0263
s
ed, %
20
ts decea
n
10
Patie
0
Time since randomization, months
E = NTX > 50 nmol/mmol creatinine; N = NTX 50 nmol/mmol creatinine.
Unpublished data; Analysis by R. Cook and Drs J. Berenson and E. Terpos.
Baseline NTX is an Independent Prognostic Indicator of
Death (Dichotomous and Continuous Variable)
P value
NTX 50
2.41
Univariate
.0057
2.59
Reduced Multivariate
.0033
NTX (per +100-unit increment)
1.79
Univariate
0120
.
1.93
Reduced Multivariate
.0103
0
1
2
3
4
5
6
Relative risk of death
Pts had to have a complete set of data for all variables assessed; NTX in nmol/mmol creatinine.
Reduced multivariate model included age, myeloma Ig type, NTX level, hemoglobin level, and SGOT level.
Unpublished data; Analysis by R. Cook and Drs J. Berenson and E. Terpos.
Markers of bone resorption and
sRANKL/OPG ratio in MM
2.8
1.8
2.6 r = 0.69,
1.6
r = 0.87,
p < .0001
1.4
p < .0001
2.4
b 1.2
2.2
1.0
2.0
.8
LG NTX 1.8
.6
1.6
LG TRACP-5
.4
1.4
.2
1.2
0.0
-1.5 -1.0
-.5
0.0
.5
1.0
1.5
2.0
-1.5 -1.0
-.5
0.0
.5
1.0
1.5
2.0
2.0
30
1.5
121
p<0.0001
36
1.0
52
KL/OPG
.5
N
0.0
Log sRA
-.5
68
-1 0
159
134
.0
149
158
-1.5 N =
45
10
28
83
Control
A
B
C
Bone Disease
Terpos et al, Blood 2003;102:1064-9
Bone formation markers
· bone alkaline
phosphatase
(bALP)
· Osteocalcin
(OC o
r
or BGP)
BGP)
· C- or N-
propeptides of
of
procollagen
(PICP PINP)
,
Serum bALP in MGUS and MM
2.5
2.0
130
159
1.5
151
149
1.0
35
90
67
122
.5
116
p<0.001
L
P
Lg bA
0.0
N =
45
40
35
86
Controls
MGUS
MM Stage1/2
MM Stage 3
GROUP
Politou, et al. Br J Haematol. 2004;126:686-9
bALP: negative results
g
· Negative results of BALP as a prognostic indicator
BALP did not show statistical correlations with the
presence of lytic lesions or with clinical or laboratory
parameters1
BALP levels did not correlate with disease stage or extent
f l i di
2
of lytic disease
BALP levels were not significantly different between
ti t
ith MGUS
d MM3
patients with MGUS and MM
A trend toward lower BALP values in patients with early
bone affectation
af
was
as observed
1. Corso, et al. Haematologica. 2001;86:394;
2. Terpos, et al. Blood. 2003;102:1064;
3. Hernandez, et al. Hematol J. 2004;5:480.
Monitoring bisphosphonate treatment with NTX
change
g (%
( ) NTX
)
150
controls (n=32)
100
p<0.001
50
pamidronate 90mg (n=30)
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
months post therapy
Terpos et al, EJH 2000;65:331-6
Pamidronate produced a greater reduction of NTX
than ibandronate
i n
in patients
who
received
VAD
120
pamidronate 90 mg
100
ibandronate 4mg
) 80
60
h
ange %
40
NTX (c
p<0.01
20
0
0
2
4
6
8
10
12
Months
Terpos et al, EJH 2003;70:34-42
ZOL significantly reduces NTX levels vs. PAM in
patients with
bone
lesions
from
breast
c
ancer
cancer or
MM
0
Zoledronic acid 4 mg
%
%
10
Pamidronate 90 mg
20
aseline,
b
30
40
ge from
n
50
60
P = .001
P = .083
P = .005
e
dian cha
70
P = .001
P = .039
P = .075
M
80
0
1
3
6
9
13
End of
t
s d
u y
T ime, months
Adapted from Lipton et al. Clin Lymphoma Myeloma 2007;7:346-53.
Thal/Dex effect on bone marker levels*
5
5
4
4
/L)
L
)
U
P<
P .001
U
3
P<.001
3
*
2
P=.005
2
P=.001
CTX (ng/m
1
RACP-5b ( 1
T
0
0
N = 30 35 35 35
N = 30 35 35 35
Controls Baseline 3 mo 6 mo
Controls Baseline 3 mo 6 mo
5
)2
60
-
2
4
r = 0.541,
(x10
40
P = .006
3
ratio
*
G
20
G 2
g
e (%),
/OPG ratio
P<.01
P<.0001
1
0
Chan
sRANKL
0
sRANK/OP
-20
N =
30 35
35
35
-3 -2 -1 0 1 2 3 4
35
Controls Baseline 3 mo 6 mo
Change (%) of CTX
*All patients were received zoledronic acid.
Terpos et al, Leukemia 2005;19:1969-76
Thal/dex effect on bone metabolism of newly
diagnosed MM
g
patients
p
Responders (77%)
P<0.0001 P<0.0001
P<0.0001
P=0.007
80
60
40
20
0
NTX
CTX
PYD
DPYD
OC bALP
Non-responders (23%)
100
P=0.04 P=0.006
80
60
40
20
0
NTX
CTX
PYD
DPYD
OC bALP
Tosi et al, EJH 2006;76:399-404
Bortezomib affects markers of bone formation
and osteoblast stimulators
200
200
) 150
L
150
P <.01
L
150
P =.035
100
(U/L) 100
P <.001
P =.007
K-1 (ng/m
K
50
ALP
50
P =
.001
b
50
D
0
0
P <.0001
70
11
10
60
9
P <.01
P =.24
P =.15
50
8
P <.01
P =.23
40
7
6
(ng/mL)
30
(pmol/L)
G
5
OC
20
OP
4
10
3
2
0
1
P <.001
-10
0
N = 33 34 32 19
N = 33 34 32 19
C PT 4* 8*
C PT 4* 8*
C= Control; PT = Prior treatment; *After cycle number.
Terpos et al, BJH 2006;135:688-92
Bortezomib affects markers of bone
resorption and osteoclast stimulators
p
1.4
5
1.2
4
l/L) 1.0
P =.01
o
a
tio
01
o 0.8
3
P <.001
P <.0001
P =.01
(pm 0.6
2
P <.001
P =.01
0.4
ANKL
1
NKL/OPG r
R 0 2
A
s
R 0.2
R
0
0
-0.2
-1
2.0
5
5
1.5
4
P <.0001
P <.001
P <.001
4
P <.01
3
1.0
P =.001
X (ng/mL)
CP-5b (U/L) 3
T
P <.0001
A
C
2
TR
0.5
2
1
1
-0.5
0
N = 33 34 32 19
N = 33 34 32 19
C PT 4* 8*
C PT 4* 8*
C= Control; PT = Prior treatment; *After cycle number.
Terpos et al, BJH 2006;135:688-92
Zoledronic Acid increases serum Osteoprotegrin
in early
stage
-
Multiple
Myeloma
Methods
· Pilot study comparing OPG, RANKL, and MIP-1 levels in patients
with stage IA MM with healthy patients
Results
· Patients with smoldering MM (n = 26) had significantly decreased
OPG levels (P < .05) and increased RANKL/OPG ratios
· Patients treated with zoledronic acid (n = 16) had increased levels
of OPG, decreased RANKL/OPG ratios and a 3.8% increase in BMD
(P < .05)
Conclusion
· Zoledronic acid increases levels of OPG, decreases RANKL/OPG
ratio, and increases BMD in patients with MM
Martini et al, Haematologica 2006;91:1720-1
Conclusions for Bone Markers in MM
· Biochemical markers of bone turnover and osteoclast function
reflect t he
the extent of
bone
disease in
in M
M
MM
(NTX, ICTP > CTX >DPD & PYD)
· NTX and ICTP correlates with SREs and disease pro
p gression
g
· NTX is an independent prognostic factor for survival
· There is a dichotomy re
y
gardin
g
g the
g
pro
p gnostic value of
g
bALP
that depends on disease stage
· May be useful in monitoring bisphosphonate treatment
· Chemotherapeutic and novel anti-myeloma agents affect
levels of bone markers in patients with multiple myeloma
· Marker (NTX)-related therapy ?
· Novel markers (RANKL, OPG) need further evaluation
Side-effects of bisphosphonates
Long-term safety: zoledronic acid is as safe and well
tolerated as pamidronate
Patients, n (%)
Zoledronic acid 4 mg
Pamidronate
(n = 563)
(n = 556)
Bone pain
325 (57.7)
316 (56.8)
Nausea
270 (48.0)
266 (47.8)
Fatigue
241 (42.8)
240 (43.2)
Pyrexia
213 (37.8)
172 (30.9)
Vomiting
187 (33.2)
183 (32.9)
Anemia
181 (32.1)
175 (31.5)
Di
D a
i rrhe
a
a
rrhe
163 (29.0
(29. )
0
162 (29.1)
(29.1
Dyspnea
158 (28.1)
155 (27.9)
Myalgia
153 (27.2)
143 (25.7)
Constipation
150 (26.6)
162 (29.1)
Cough
150 (26.6)
129 (23.2)
Arth
t r
h a
r l
a gi
g a
i
148 (26 3
. )
3)
131 (23 6
. )
6)
Weakness
127 (22.6)
108 (19.4)
Headache
123 (21.8)
149 (26.8)
*Regardless of study drug relationship.
Rosen et al, Cancer 2003;98:1735-44
Zoledronic acid vs. pamidronate: time to first
ti
serum crea i
n
i
ne ncrease
100
80
out
h
,
% 60
s
wit
a
se
t
s
40
e
n
c
r
e
a
ti
in
n Hazard
ratio
P value
20
Pa
Zol 4 mg
272
1.057
.839
Pam
Pa
9
0
90 mg
mg
268
0
0
120
240
360
480
600
720
Time, days
Zol 4 mg
272
226
197
152
81
68
30
Pam 90 mg
268
213
182
138
73
59
27
*Post15-minute infusion amendment.
After start of study drug.
Rosen et al, Cancer 2003;98:1735-44
Osteonecrosis of the Jaw
Clinical features of suspected ONJ
·
Exposed bone in maxillofacial
area that occurs in association
with dental surgery or occurs
spontaneo sl
u
ith
y, w no
no
evidence of healing
Worki
ki
D
ng i
Diagnosi
f
s o O
NJ
ONJ
·
No evidence of healing after
6 w
eeks
weeks o
f
of appropriate
evaluation and dental care
·
No evidence of metastatic
disease in the jaw or
osteoradionecrosis
ONJ: characteristics
Symptoms
· "heav
·
y j aw", a dull aching
aw", a dull achin sensation
g
· numbness/tingling of
·
the
numbness/tingling of
jaw
the
· tooth pain
·
· undiagnosed
undiagnosed oral pain
Signs
· roug
·
h area on
h
th
th j
e
b
aw one
b
· soft
·
tissue
soft
swelling, drainage or
tissue
infection
· expose
·
d
d bone
b
in th
th
l
e ora cavitity
· sudden change in the
·
health of
sudden change in the
periodontal tissue
f
·
f
ailure o oral mucosa to heal
· loosening of
·
teeth
loosening of
Greek Myeloma Study Group experience (1)
November 1991 - January 2005
303 patients
with MM
49 (16%): no bisphosphonates
91 (30%): zoledronic acid 4mg
78 (25%): pamidronate 90mg
85 (28%) : pamidronate 90 mg
zoledronic acid 4mg
Zervas et al, BJH 2006;134:620-3
Results (
1)
(1)
· The analysis was prospectively after the presence
of our first case (Feb 2001)
· The median numb
f
er o bi
h
sp
h
osp onate i f
n usions
administered to all patients was 15 (range: 4-77)
· The median time of exposure was 24 m (4-120 m)
· No myeloma patitient h
w
t
o was no given
bisphosphonates developed or present symptoms
of ONJ (p<0.001)
Results (2)
28 out of 254 (11.02%)
developed ONJ
11 diagnosi
f
s con ifi
d
rme
17 diagnosis b
d
ase
R
on x
by biopsy
and CT criteria
Variables with independent prognostic
importance for ONJ develo
p
pment
p
Parameter
Relative Risk
P-value
(95%CI)
Pamidronate alone, 0 cycles of
1.00
administra
administr tion and no thal
Zoledronic acid alone
9.5
0.042
(1.0-83.2
(
)
Subsequent use of pamidronate
2.1
0.495
and zoledronic acid
(0.2-18.0)
Cycles of bisphosphonate
4.9
0.012
administration
(1.9-9.9)
Thalidomide use
use
2 4
.
0 043
.
(1.0-5.7)
Bisphosphonate type and time to ONJ
· ONJ occurred significantly earlier among patients receiving only
zoledronic acid
· Median exposure was 27.5 months vs. 67.5 months, p=0.001
Ti
t
me o ONJ
ONJ d
evelopment with
ith dif
f
dif erent
Survival Function for patterns 1 - 3
bisphosphonates
drug
1,0
aredia
NJ
both
O
0,8
zometa
e
lop
0,6
v
i
v
al
id not dev Sur
d
0,4
ho Cum
0,2
roportion w
P
0,0
0,00
20,00
40,00
60,00
80,00
100,00
120,00
diarkeia
months
Greek Myeloma Study Group experience (2)
April 1995 - June 2005
202
ti
pa
t
en
i
s w th
ith MM
2 (1%): ibandronate
93 (46%): zoledronic acid 4mg
4: ZA + ibandronate
33 (16%): pamidronate 90mg
2: pamidronate + ibandronate
1: clodronate + ZA
1: residronate + ZA
66 (32%): pamidronate
p
90 mg
zoledronic acid 4mg
Dimopoulos et al,
Haematologica 2006;91:968-71
Results-Patients
15 out of 202 (7.42%)
developed ONJ
7 r
eceiv
r
e
eceiv d
e Z
A
ZA a
lone
alone
1 r
eceiv
r
e
eceiv d
6 sequential PA + ZA
pamidronate alone
1 sequential ZA + IB
q
Results (2)
Osteonecrosis
Yes
Ye
No
p
Median exposure (months)
All patients
39 (n=15)
28 (n=187)
.048
Zoledronic acid
30 (11-53) (n=7)
19 (5-85) (n=86)
.185
Pamidronate
68 (n=1)
19 (n=32)
· No patient who received fewer than 9 treatments with
biphosphonates developed osteonecrosis.
· ONJ occurred significantly earlier among patients receiving
only zoledronic acid:
median exposure: 30 months for ZA vs. 62 months
for all others p=0.009.
Results (3)
0,7
Zoledronic acid
0,6
alone ( )
r
d 0,05
others (
)
0,4
p=0.009
Haza
m
Cu 0,3
0.2
0,1
0,0
0,
0 00
0
00
25,
25 00
50,
50 00
75,
75 00
100,
100 00
125
,
125 00
Exposure (months)
Relative risk for ONJ development
p
15 developed
ONJ
Relative risk
(7.4%)
12
24 m
36 m
48 m
months
%
95%
%
95%
%
95%
%
95%
CI
CI
CI
CI
All (n=202)
1
0-2
3
1-4
6
2-10
13
5-21
Zoledronic acid (n=93)
1
0-3
5
0-11
15
3-27
15
3-27
PA (n=33)
0
0
1
0-3
1
0-3
5
0-11
ASCO guidelines
· The U
pdate
Update Committee
Committee suggests
suggests that
bisphosphonate treatment continue for a period
of 2
y
ears
years.
· At 2 years, physicians should seriously consider
discontinui
b
ng i
bi h
sp
h
osp
t
ona es i
t
n pa iti t
en
ith
s w
responsive or stable disease, but further use is
at the discretion of the treating physician.
· Re-initiation at relapse.
Kyle et al, JCO 2007;25:2464-72
Conclusions for ONJ (1)
· ONJ is a complication of bisphosphonate treatment,
associated i
w h
t h
t
i
e t
f
me o exposure i
i
n pat
i
ents w h
t
MM.
· The risk seems to be higher when zoledronic acid
(compared to pamidronate) is used.
· Thalidomide may be implicated in the pathogenesis of
ONJ observed
in
in M
M
MM patients w
ho
who received zoledronic
acid
· No satisfactory
t herapy
therapy is
currently
available
.
Conclusions (2)
· Although the association of ONJ with the use of
bisphosphonates seems well established, several
issues need to be clarified:
the pathophysiological mechanism
Other possible risk factors,
prevention and appropriate treatment.
· The diagnosis of ONJ, in most cases, has been made
retrospectively based on the review of medical records
rather than
by
a
s
pecialist
specialist
· Further studies are needed to establish the time for
bisphosphonates administration
with
t he
the better
r
esults
results
for the patient.
ZOMETA Trial 2105: Trial Design
· Patients with
multiple
T G
x
A
roup , n = 36
36
ZOMETA
ZOMET q 4 wk
myeloma and
breast cancer
+ bone mets
· Pretreated
t
ion 1:1
Pretreated
with 9 to 12
doses of
ZOMETA
during
g prior
p
a
ndomiza
year
R
N = 72
Tx Group B, n = 36
ZOMETA q 12 wk
0
4
8 12 16 20 24 28 32 36 40 44 48 52
First study
Study drug infusions every 4
Last study
Drug infusion
weeks
Drug infusion
Acknowledgments
Hammersmith, London
Greek Myeloma Study Group
Amin Rahemtulla
M.A. Dimopoulos
p
(Athens)
Marianna Politou
K. Zervas (Thessaloniki)
Richard Szydlo
E. Katodritou (Thessaloniki)
Tasos K
aradimitris
Karadimitris
A. Anagnostopoulos (Athens)
Jane F. Apperley
A. Pouli (Athens)
Sh ffi ld
K T
. sionos
Tsionos (Athens)
Sheffield
Peter Croucher
D. Christoulas (Athens)
Debby Heath
E. Verrou (Thessaloniki)
Rob Coleman
K. Anargyrou (Athens)
E. Vervessou (Athens)
Richard Cook
Cook (Ontario Canada)
,
James Berenson (LA, CA, USA)
Allan Lipton (Hershey, PA, USA)