New Insights In to the Molecular Basis of MM
Pathogenesis & Prognosis
John D. Shaughnessy, Jr.
and Bart

Barlogie
Myeloma Institute for Research and Therapy
Can Genomics Aid in Risk Assessment and
Help Guide

Ther

ap
Ther
y
ap Decisions
Common Histology/
Variable Outcome
Pt X
100%
80%
Aliven 60%
Pt Y
o
40%
Proporti
20%
0%
0
24
48
72
96
120 144 168
192
Months from Start of Therapy
Integrating High Resolution Genomics with Outcome in
Large
g Uniformly
y Treated Patient Cohorts
Agilent 244K
aCGH platform
CD138+
DNA
Plasma cells
~ 800

ne

wly
ne

diagnosed
MM treated
TT
Affymetrix U133Plus2
on TT
Affymetrix
GeneChip
mRNA
Defining HighRisk by Correlating
Gene Expr

ession
Expr
Extr

emes
Extr
with

Survival
iv
351 Pa
P tien
a
ts
tien
25,000 Genes
Log Rank Test
Log Rank Test
Q1 vs. Q2


Q2 Q4
Q4 vs. Q1


Q1 Q3
19 Q1 Genes
51 Q4 Genes
Correla
Corr
t
ela ed
t
with
Correla
Corr
t
ela ed
t
with
Short Survival
Short Survival
70 Gene Model Predominated by Elevated Expression
of 1q and Reduced Expression of 1p Genes
1q21
70Gene Risk and GEP PI Correlated +1q & 1p
Risk Score
1.0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16 17 18 19 20 2122
X
Y
t
.5
ficien
0
ff
Coen
0.0
latiore 5
Cor
0.
1.0
Proliferation Index
1.0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16 17 18 19 20 2122
X
Y
nte
.50
ffici
coe
0.0
ation
5
el
0.
Corr
1.0
GEP70Gene Model Identifies 13% of Newly
Diagnosed MM with High Risk
30
High Risk
52
cy
en
20
uq
15
Fre
01
05
10
1
2
0.66
Shaughnessy et al. 2007
70Gene High Risk is an Independent Variable in TT2
OS
EFS
**MVA includes all standard and molecular variables including GEP PI
GEP Risk Accounts for 25% of Outcome Variability in TT2.
Additional Vi
Variabl
bles At
Ano h
ther 15%
Over
Ov all
er Survival
viv
(%)
40
ility
30
riab
20
Vaem 10
coutO 0 70 CA FL LDH
GEP
p1q21
MRI
Other
mA
Shaughnessy et al., BJH, 2007
Therapy Reveals Biology: Overcoming HighRisk
Define GEP/Molecular Basis of Long
Ter
Te m
r
Disease

Cont
Con ro
r l
o in

High


High Risk
Validation of 70Gene
High Risk

Model
70Gene HiR, but Not MS, is Independent Variable in TT3
MVA
MV
OS
EFS
EF
70Gene Model Validated in Relapsed Refractory MM
Tre
Tr a
e t
a e
t d
e with Single

Ag

en
Ag
t
en Bzb or High

Dose

De

x
De
Velcade Arm
High Dose Dex Arm
1.0
1.0
ases 0.8
0.8
ases
C
C
of 0.6
of 0.6
ortionp
ortion
p 04
0.4
p
o
04
0.4
o
Pr
Pr
0.2
OS P = 0.0495, HR = 2.34
0.2
OS P = 0.0174, HR = 2.55
LowRisk 34/70
LowRisk 29/65
0.0
HighRisk 6/10
0.0
HighRisk 8/11
012
24
36
012
24
36
Mont
Mon hs Fr
om
Fr
Start
St
of
Therapy
Ther
Mont
Mon hs Fr
om
Fr
Start
St
of
Therapy
Ther
Zhan et al., Blood 2008, 111: 968969
70Gene Model Defines High Risk in Newly
Diagnosed MM

Treated with ASCT at Mayo Clinic
100%
Deaths/N
Low Risk

17/40
High Risk
23/31
80%
ve
60%
Ali
40%
ortionpoPr 20%
P= 6e04
HR=3.07
0%
0
20
40
60
80
100
120
Months from Start of Therapy
Chng et al., Can Res. 2007, 67:2982
70Gene Model Validated in IFM Trials
Key findings:
· IFM 15gene model, reflecting proliferation, defines highrisk
· UAMS 70gene model validated in IFM data set and visa versa
· In MVA UAMS model significant in all trials; IFM model only a subset
· IFM model significant for Bzb not High Dose Dex in APEX
Molecular Characterization of
Focal Tumor Growth
Osteolytic Lesions Adjacent to Focal Tumor Growth:
Dist
Dis inct Biology?

XRAY
Foc
Fo a
c l Lesion
MRI
Aim: Molecular and Clinical Significance of Intra and
Int
In e
t r
e
r Pat
Pa ien
e t
n V
a
V riability in

Random

and

Fo
c
Fo a
c l Disease
Random
Aspirate
(RS)
(R f
o
CTGuided
Iliac
Aspirate of
Crest
Foc
Fo a
c l
a Lesion

Log2 MAS5.0 DKK1
6
10
12
814
BM Bcells
BM PC

MGUS PC
WM PC
MMCL
MM PC
MM PC FL
RL MM PC
RL MM

PC FL
BM BX
MGUS BX
WM BX


MM BX
MM BX FL
RL MM BX
RL MM

BX FL

Abnormal Cytogenetics in both Random and Focal Lesion
Confers Poor Prognosis in 70Gene Low Risk Disease
1.0
0.8
Cases 06
0.6
of
0.4
ortionpoPr 0.2
0.0
0
1224
3648
60
72
84
96
108 120
Months From Start of Therapy
Zhou et al., 2009, In Press, BJH
Distribution of 70Gene Risk Scores within CA Subgroup:
RS+/FL
F +
L Low Ri k
s is Diff

er
t
en
3.0
RS /FL+

2.5
20
2.0
RS/FL
1.5
ensity
RS+/FL+
D
1.0
RS+/FL
0.5
00
0.0
3
2
1
0
1
2
3
Low Risk Score
High Risk Score
RS+/FL+ like?
CA/Risk
(N)
CA/Risk
(N)
All
290
All
290
1.0
RS+/FL+, low
25
1.0
RS+/FL, low
36
RS+/FL+, high
29
RS+/FL, high
13
0.8
0.8
0.6
0.6
ensityD 0.4
Density 0.4
D
0.2
0.2
0.0
0.0
3

2

1

0
1
2
3
3

2

1

0
1
2
3
Risk Score
Risk Score
CA/Risk
(N)
CA/Risk
(N)
All
290
All
290
3.0
RS/FL+, low
46
2.0
RS/FL, low
130
RS/FL+, high
4
RS/FL, high
7
2.5
1.5
2.0
y
ty
1.5
1.0
Densit
Densi
1.0
0.5
0.5
0.0
0.0
3
2
1
0
1
2
3
3
2
1
0
1
2
3
Risk Score
Risk Score
CA, CA1 and CA13 in RS+ / FL+ Disease
70
70 Gene Risk

Model &

Molecular Subgroups
Distinct Gene Expression Signatures Define Six
Reproducible Subgroups
100 up
p & 100
down per
subgroup
n
essio
Expren
Ge
Zhan et al., 2006
Molecular Subgroups in MGUS, MM, and Relapse MM
40
MGUS (103)
35
MM (803)
30
RELAPSED MM (131)
s
25
e
20
Cas
of
15
%
10
5
0
CD1CD2HY
LB
MF
MS
MY
PR
HighRisk MM in all Molecular Subtypes
100
LR
LoR
90
80
HiR
70
Cases
60
50
r of
40
TT2
30
20
Numbe
10
0
MY
PR
LB
MS
HY
CD1
CD2
MF
(N=95) (N=29) (N=31) (N=44) (N=66) (N=22) (N=43) (N=21)
50
45
40
35
30
f Cases o 25
20
TT3
15
10
Number
5
0
MY
PR
LB
MS
HY
CD1
CD2
MF
(N=50) (N=16) (N=21) (N=19) (N=41) (N=5) (N=15) (N=14)
MS Subtype and 70Gene Risk in TT2
100%
ts
80%
n
atie
60%
P
ofno 40%
oporti
20%
Pr
0%
0
2
4
6
8
Years from start of treatment
Median =
63 mo
Group
Deaths/N 1YearEstimates
Low/nonMS
78 / 270
94% (92,97)
P = 0.008
P < 0 0001
.
Low/MS
18 / 35
94% (85,100)
High/nonMS
26 / 37
78% (65,92)
P = 0.0002
P < 0.05
High/MS
8 / 9
67% (34,99)
Bzb Added to TT3: CA, 70Gene, & LDH Remain, but t(4;14)
& B2M, No Longer Independent Risk Factors
MVA
OS
EFS
TT2
TT3
Defining
De
Role of

Thalidomide


in Risk Groups
Thal Improves Outcome in
Cytogenetic Abnormality (CA) Positive Disease
NS
No CA
vs.
P = 0.02
CA
70Gene LoR
NS
NS
vs.
70Gene HiR
Thal Elicits No CALike Outcome in CA+ / Low Risk
NS
No
CA
P = 0.003
CA
TT2 No
CA1, 13, or
Ni
Ne th
Thal
ither
/Low
CA1 & CA13
Risk
CA1, 13, or
Neither
TT2 Thal
CA1 & CA13
/Low Risk
CA1, 13, or
Ni
Neither
TT3 /
CA1 & CA13
Low Risk
Note:
No CA high risk
No CA13

high risk
Global GEP Differences in
CA1+ & CA13+ vs. CA1+

or CA13+ or CA
8000
TT2 +Thal
7000
TT2 Thal
6000
Gene
TT2
1
5000
1q2
4000
3000
osomem 2000
Chro
1000
0
CA1/CA13
Other
CA1/CA13
Other
Median
Mean
Disruption of G1 > S Checkpoint in MM:
Cooperation of Amp1q21 and del13?
Universial 1o Events
Cdk1/Cyc
Cdk1/Cy lin B

G2
M
CCND1
CCND2
CCND3
pSRbE2F + CDK6
13, Rb
S
G1
pSTRb
2o Event
E2F
A1
Amp1q21
21
CKS1B
Serial GEP Analysis Reveals
Risk Often Increases at
Relapse
R
Figure 2
PC
GUS
N
MYELOMA
N
M
HMCL
Relative Risk
Longitudinal Studies Reveal GEP-Defined
High-Risk Score Reaches 76% at Relapse
3
2.5
2
1.5
skScoreRi 1
High Risk
ene
0.5
Cut Point
G
0
70-
-0.5
baseline
-1
relapse
-1.5
-2
Patients
Change
g in 70Gene Risk Score Over Time
Length of time to relapse generally inversely
proportional to BL GEP score
2
e 1
or
Sc
iskR 0
1

N = 71 paired samples
0
500
1000
1500
2000
2500
Days From Baseline Sample
High Risk Score at Relapse Predicts
Post
Po
st Rel
Re aps
p e
s Sur

vival
iv
100%
BL /
RL
Dea h/
ths/N Median ()
(Mo)
Low / Low
10 / 38
53 (47, 56)
Low / High
14 / 20
13 (8,22)
ve 80%
v
High /
Low
3 /
3
20 (16

,20)
Ali
High / High
14 / 18
6 (4,8)
60%
rtion 40%
opoPr 20%
0%
0
1
2
3
4
5
6
7
8
Years from Relapse
Both 1q21 Copy Number and Percentage of
Cells with

Amp1q21

Increase
Incr
at Rel
Re aps
p e
s
3 copies
>= 4
copies
100
1q21p
80
Am
60
withls
cel
40
of%
20
0 1 2 3 4 5 6 7 8 9 10 1112 1314 1516 1718 1920 2122 2324 2526 27 28
Pi
Paired Bl
Base ili
/
ne R
/R l
e apse Sl
Samples
Hanamura et al. Blood 2006
High Risk

in MM

Is
Evolving
GEP Can

be

Used

to

Cr
ea
Cr t
ea e
t
Virtual Karyotypes
Gene Spikes Define Recurrent Translocations in MM
no
essi
Expr
ne
Ge
Zhan et al., 2006
Relationship Between Copy Number Changes and
Molecular Subgroups
21 X
12
3
4
5
6
7
8
9
10
11
12
1
13 4 15 16 17 181920 22 Y
PR
MS
MF
LB
HY
CD2
CD1
6
4
2
0
-2
-4
-6
Hyperdiploidy in MM Generally Characterized by
Tr
T isomies
r
of

Odd

Chr

o
Chr mosomes
o
2 X
12
3
4
5
6
7
8 9
10
11
12
1
13 415 16 17 1 1 2 2 Y
1
8 9 0 2
PR
MS
MF
LB
HY
CD2
CD1
k1
k2
k3
k4
Chr 1
Chr 3
Chr 5

Chr 7
Chr 9
Chr 11

Chr 13
Chr 15
Chr 19
Chromosome 11
Chromosome 1q
Gi
Gain
Gain
Cutoff for
FISH
FISH
aCGH signal
l
Normal
Norma
aCGHbased copy number
aCGHbased copy number
Chro
Chr mosome
o
13

Loss
FISH
Normal
aCGHbased copy number
True Negative
Training Set Test Set
e
False Negative
score
atur
Chr1p
True Positive
sign
GEP
False Positive
aCGHbased copy number
False Positive
Tr
T aining
r
Set
Se Test
s Se
t
Se
True Positive
e
score
tur
Chr1q
signa
GEP
False Negative
aCGHbased copy number
True Negative
Chromosome 11
Training Set Test Set
eor
e sc
tur
gna si
GEP
aCGHbased copy number
Chromosome 13
Training Set Test Set
eor
e sc
tur
gna si
GEP
aCGHbased copy number
John Crowley
Bart Barlogie
Antje Hoering
Jeff Haessler
Research Faculty
Pinqping Qu
& Fellows
Research Support
Clinical Faculty
Jackie Szymonifka
Frits Van Rhee
Vanessa Bolejack
Fenghuang Zhan
Austin Porter
Elias Anaissie
Joshua Epstein
Owen St
h
ep ens
Yazan Ald
lsayed
SWO
SW G
Erming Tian
David R. Williams
Elias Kiwan
Yawei Qiang
Xiaosong Wu
Mauricio PinedaRoman
Madhav Dhodapkar
Shmuel Yaccoby
Yan Xiao
Bijay Nair
Ricky Ed
E mondson
d
Nathan Brown
Rahnuma Wahid
Damir Herman
Hongwei Xu
Sarah Johnson
Bob Kordsmeier
Radiology
James P. Stewart
Susan Ward
Ron Walker
Major
Major Support

Benzhenng Chen
Edgardo Angtuaco
i
Marie As
trid
As
Catr
Ca on
Rudy Van Hermert
Igor Entin
tron
·NCI
Bo Hu
Twyla B Bartel
Tracy LY Brown
·MMRF
Yiming Zhou
·
Pathology
SWOG
Jeffrey Sawyer
·Grand
Gr
F
o
F unda
o
tion
unda
Bill Bellamy
Madhumita Santra
·Lebow
Lebow Fund

Robert Lorsbach