Microsoft PowerPoint - 01-04-Avet-Loiseau.ppt [Compatibility Mode]

Genetic Changes in My
gyeloma :
Prognostic Implications
Hervé AVETLOISEAU, MD, PhD
Hematology
gy Laboratory
University Hospital
Nantes, France

Cytogenetics & Prognosis in Myeloma:
Current Knowledge
Poor prognosis
pr
· Non-Hyperdiploidy
· t(4;14)(p16;q32)
· del(17p)
· 1q g
qgains
· del(13)
Good prognosis
· Hyperdiploidy

Overall Survival

Analysis using genome-wide SNParrays
Patients
· 192 patients < 65 years old
· Newly diagnosed symptomatic myeloma
· Treated with double high-dose
g
Melphalan
p
· Median follow-up = 66 months
· Highly purified plasma cells
·500K
·
Affymetrix SNP arrays
arrays
Results
· All the patients display abnormalities
(vs 10-30% with karyotype)
· Two clusters: hyperdiploidy vs non

Prognostic value of hyperdiploidy (karyotype)
Smadja et al., Blood 2001;98:2229

Copy number summary plot

Prognostic value of chr. 5 gain vs ploidy
1.00
0.75
alivv
P= .0014
SurSllra
P< .0001
ve 0.50
Ofo
ility
ab
oba
Pr 0.25
hyperdiploid amp(5q)
hyperdiploid no amp(5q)
non hyperdiploid
Global P= .0001
0.00
0
20
40
60
80
100
Time (Months)
Patients at risk :
69
63
53
24
3
0
30
21
17
6
3
0
87
68
49
17
4
0
Validation by FISH in an independent series (273 patients)

Univariate Analyses
Factor
HR
95% CI
P
-
Age (< 50 years)
-
<.01
Albumin (< 35 g/L)
2.38
1.33 to 4.17
<.05
Creatinine (> 180 µmol/L)
2.20
1.15 to 4.21
-
Hemoglobin (< 10 g/dL)
-
-
Platelets (< 130 109/L)
-
<.001
S2M ( 5.5 mg/L)
2.97
1.91 to 4.60
<.001
ISS
2.04
1.52 to 2.73
<.05
Del(13)
1.01
1.00 to 1.01
<.01
t(4;14)
2.27
1.33 to 3.87
-
t(11;14)
-
-
<0
<. 5
05
Del(17p) ( 60% )
2.26
1.61 to 4.39
=.001
Hyperdiploidy ( 47 chr)
0.48
0.31 to 0.75
<.05
Loss 1p
1.72
1.1 to 2.7
<.01
Gain 1q
1.85
1.21 to 2.85
<.05
Gain 3
0.56
0.35 to 0.90
<.001
Gain 5
0.43
0.27 to 0.70
-
Gain 6p/6 *
-
Loss 6q
-
Gain 7
-
Loss 8p/ 8§
<.001
Gain 9
0.45
0.28 to 0.72
-
Gain 11
-
Gain 11q only
<.001
Loss 12p
2.56
1.49 to 4.35
-
Loss 13
-
Loss 14q (1vs>1)
<.05
Gain 15
0.61
0.39 to 0.93
-
Loss 16p
-
Loss 16q
-
Gain 18
<.05
Gain 19
0.57
0.37 to 0.90
-
Loss 20p
-
Gain 21
-
Loss 22

Multivariate Analyses

Multivariate Analyses
1.00
al 0.75
v
rviruSllraevO0.50f
yo
ilityba
ob
Pr 0.25
amp(5q) nodel(12p) noamp(1q)
others
amp(5q) amp(1q) del(12p)
or no amp(5q) and[amp(1q) and/or del(12p)]
P<.0001
0.00
0
20
40
60
80
100
Time(months)
Patients at risk :
54
51
42
20
3
0
80
68
57
24
6
0
58
37
23
4
1
0

Multivariate Analyses with IFM model
Multivariate analysis (Cox model)
Prognostic
g
variables
Hazard ratio [95% CI*]
[]
P
S M(5.5 mg/L)
2.99 [1.88 - 4.74]
<.0001
2
16
1. 6
66 [09
[0. 7
97 - 2.86]
.066
t(4;14)_del(17p)
Amp(5q31.3)
0.45 [0.25 - 0.80]§
.0070
Del(12p
12 13
p
.31)
31
26
2. 1[
61 14
[1. 4
44 - 47
4. 2]§
72]
.0015

Multivariate Analyses (5q, 12p, s2m)
1.00
la 0.75
vivvurSll
eravO 0.50
of
litybi
oba
Pr 0.25
amp(5q) nodel(12p) lowS2M
others
amp(5q) highS2Mdel(12p)
or noamp(5q) and[highS2Mand/or del(12p)]
P<.0001
0.00
0
20
40
60
80
100
Time(months)
Patientsat risk:
53
50
44
22
2
0
86
74
55
21
4
0
46
26
17
3
3
0

Putative Targets
Chr 5 gains: usually the whole chromosome...
Del(12p): centered on CD27/TNFRSF7
E
s
t 1.00
im
a
t
p(Logrank) <.0001
e
p
r 0.75
o
b
a
b
i
l
i
CD27 expression
0.50
t
y
o
f
s
u 0.25
r
v
i
CD27Q4
v
CD27Q3
a
CD27Q2
l
CD27
CD2 Q1
0.00
0
20
40
60
80
100
patientsat risk:
Overall Survival (Months)
Q1
43
24
18
6
1
0
Q2
42
37
26
10
3
0
Q3
43
38
29
7
1
0
Q4
42
37
33
17
2
0

Other Genomic Models: GEP
Identification of a 15-gene set
C
B
All patients (IFM data set)
Test group (IFM data set)
A
Training group (IFM data set)
F
E
Validation (Mulligan's
D
Validation (Mayo Clinic data set)
Validation (UAMS data set)
data set

bortezomib

alone)

Decaux O et al., JCO 2008

Genomic Models: Multivariate Analyses
Univariate Analysis
Adjusted Cox
(log-rank test)
HR
95% CI*
P
HR
95% CI
P
Predictor
3.16
1.98-5.04
<.0001
2.08
1.29-3.34
.003
15-gene model
2.94
2.17-3.99
<.0001
2.77
2.03-3.78
<.0001
CNA-based model

Genomic Models: Multivariate Analyses
1.00
lva
urvi 0.75
P=.46
P<.0001
lSal
ver
P<.0001
0.50
fO
P<.029
lityo
abi 0.25
0.
ba
Pro
Global P<.0001
0.00
0
20
40
60
80
Time(months)
Patients at risk
Bothlow
96
88
75
33
4
15-genehi h
g 25
19
10
3
0
CN
hi As
gh high
26
17
12
1
1
Bothhigh
16
6
3
1
1

Conclusions & Perspectives
Copy number analyses reveal novel prognostic regions
Copy number analyses may dissect hyperdiploidy
Copy number analyses enable prognostication
Independent prognostic value of GEP & SNParray
SNParrays + Expression arrays should enhance myeloma
oncogenesis understanding, and ultimately treatment.

Stéphane MINVIELLE
Florence MAGRANGEAS
Catherine CHARBONNEL
Wilfried GOURAUD
Loïc CAMPION
DANA-FARBER
CANCER INSTITUTE
Nikhil MUNSHI
Cheng LI
Kenneth ANDERSON
Michel ATTAL
Thierry FACON
Claire MATHIOT
Gérald MARIT
Supported by
by a grant of
of a
Jean-Luc HAROUSSEAU
Specialized Program Of Research
Philippe MOREAU
Excellence (SPORE)