Transcript Functional Genomics & Systems Biology: Perspectives from

The Molecular Phenotype of Heart
Allograft Biopsies
Mario C Deng
Associate Professor of Medicine
Director of Cardiac Transplantation Research
Center for Advanced Cardiac Care
Department of Medicine
Columbia University
USA
modern medicine & immortality
New York Times Magazine Jan 30, 2000
heart transplant milestones
Animal Htx (Shumway)
(1959)
 First human Htx (Barnard)
(1967)
 Endomyocardial Biopsy (Caves)
(1973)
 Copeland re-Htx (Copeland)
(1974)
 Human HLtx (Reitz)
(1981)
 Baby Htx (Bailey)

Electrical VAD (Portner)

(1984)

(2000)
 Steroids
 ATGAM
 Azathioprin
 CsA
 Fk506
 Sirolimus
(2005)
 Neoral
 MMF
 Chemogenomics
 OKT3
1965
1975
1985
H. Genome (Lander)

Allomap (Deng)
 Daclizumab
1995
2005
2009
heart transplant survival by era
100
All comparisons significant at p < 0.0001
Survival (%)
80
60
1982-1991 (N=18,854)
40
1992-2001 (N=35,146)
2002-6/2006 (N=12,369)
20
HALF-LIFE 1982-1991: 8.8 years; 1992-2001: 10.5 years; 2002-6/2006: NA
0
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Years
ISHLT
Taylor D et al. J Heart Lung Transplant 2008;27: 937-983
heart transplant cause of death
CAUSE OF DEATH
0-30 Days
(N = 3,006)
31 Days –
1 Year
(N = 2,722)
>1 Year –
3 Years
(N = 2,135)
>3 Years –
5 Years
(N = 1,857)
>5 Years –
10 Years
(N = 4,054)
>10 Years
(N = 2,107)
CARDIAC ALLOGRAFT VASCULOPATHY
52 (1.7%)
127 (4.7%)
298 (14.0%)
299 (16.1%)
581 (14.3%)
309 (14.7%)
ACUTE REJECTION
193 (6.4%)
338 (12.4%)
220 (10.3%)
82 (4.4%)
69 (1.7%)
26 (1.2%)
LYMPHOMA
2 (0.1%)
54 (2.0%)
85 (4.0%)
96 (5.2%)
195 (4.8%)
73 (3.5%)
MALIGNANCY, OTHER
1 (0.0%)
57 (2.1%)
218 (10.2%)
340 (18.3%)
749 (18.5%)
392 (18.6%)
CMV
4 (0.1%)
34 (1.2%)
16 (0.7%)
3 (0.2%)
5 (0.1%)
1 (0.0%)
393 (13.1%)
896 (32.9%)
276 (12.9%)
180 (9.7%)
442 (10.9%)
213 (10.1%)
1,257 (41.8%)
500 (18.4%)
499 (23.4%)
379 (20.4%)
765 (18.9%)
353 (16.8%)
TECHNICAL
233 (7.8%)
28 (1.0%)
17 (0.8%)
17 (0.9%)
36 (0.9%)
20 (0.9%)
OTHER
162 (5.4%)
175 (6.4%)
187 (8.8%)
147 (7.9%)
339 (8.4%)
175 (8.3%)
MULTIPLE ORGAN FAILURE
356 (11.8%)
268 (9.8%)
117 (5.5%)
102 (5.5%)
309 (7.6%)
190 (9.0%)
RENAL FAILURE
20 (0.7%)
25 (0.9%)
36 (1.7%)
65 (3.5%)
225 (5.6%)
173 (8.2%)
PULMONARY
133 (4.4%)
108 (4.0%)
96 (4.5%)
85 (4.6%)
172 (4.2%)
99 (4.7%)
CEREBROVASCULAR
200 (6.7%)
112 (4.1%)
70 (3.3%)
62 (3.3%)
167 (4.1%)
83 (3.9%)
INFECTION, NON-CMV
GRAFT FAILURE
ISHLT
Taylor D et al. J Heart Lung Transplant 2008;27: 937-983
allograft rejection
Recipient
immune
response
T-cell activating signals
#1 T-cell receptor
#2 CD28
#2a CD40L
#3 IL2, IL15 etc
Bone
thymus
marrow
TCR
Allo Ag
allo MHC
allo APC
Allo Ag
self MHC
monocyte
lymphnode
B-cell
spleen
self APC
Cardiac allograft
CD3+
T-cell
CD8+
CD4+
rejection
CD8+
Cytotox
CD4+Th1
CD4+/45RO+
tolerance
CD8+28CD8+
suppr
CD4+/25+
CD4+Th2
systems biology strategy
Rejection
Quiescence
phenome
metabolome
proteome
transcriptome
genome
allograft rejection
Recipient
immune
response
Rejection
Quiescence
T-cell activating signals
#1 T-cell receptor
#2 CD28
#2a CD40L
#3 IL2, IL15 etc
phenome
metabolome
TCR
Allo Ag
Allo Ag
allo MHC
self MHC
proteome
transcriptome
allo APC
monocyte
self APC
B-cell
CD3+
T-cell
CD8+
CD4+
rejection
CD8+
Cytotox
CD4+Th1
CD4+/45RO+
tolerance
CD8+28CD8+
suppr
CD4+/25+
CD4+Th
genome
Cardiac allograft
2
organ
systemic
phenome
echo
clinical
proteome
histo, Cd4
biomarkers
transcriptome
RT-PCR, ISH, array
Allomap
genome
DNA-sequencing
DNA-sequencing
endomyocardial biopsy
mild
severe
status quo monitoring






invasive & complication-prone
late-stage cellular rejection diagnosis
insensitive for humoral rejection
significant variability
no insight into molecular mechanisms
resource-intense
rejection grading


0
1





2
3




4

no rejection
A, focal infiltrate without necrosis
B, diffuse sparse infiltrate w/o necrosis
one focus aggressive infiltration/focal myocyte damage
A, multifocal aggr infiltr or myoc damage
B, diffuse inflamm process with necrosis
diffuse, necrosis, edema, hemorrhage

0R
1R

2R

3R

AMR
Quilty
Billingham ME et al. J Heart Lung Transplant 1990;9:587
Stewart S et al. J Heart Lung Transplant 2005;24:1710
current standard of literature
Nielsen H., F.B. Sorensen and B. Nielsen et al., Reproducibility of the acute rejection diagnosis in human cardiac allografts.
The Stanford Classification and the International Grading System, J Heart Lung Transplant 12 (1993), pp. 239–243
Fishbein M.C., G. Bell and M.A. Lones et al., Grade 2 cellular heart rejection does it exist?, J Heart Lung Transplant 13
(1994), pp. 1051–1057
Winters G.L., E. Loh and F.J. Schoen, Natural history of focal moderate cardiac allograft rejection. Is treatment warranted?,
Circulation 91 (1995), pp. 1975–1980
Milano A., A.L. Caforio and U. Livi et al., Evolution of focal moderate rejection of the cardiac allograft, J Heart Lung
Transplant 15 (1996), pp. 456–460
Brunner-La Rocca H.P., G. Sutsch, J. Schneider, F. Follath and W. Kiowski, Natural course of moderate cardiac allograft
rejection early and late after transplantation, Circulation 94 (1996), pp. 1334–1338
Mills R.N., D.C. Naftel and J.K. Kirklin et al., Heart transplant rejection with hemodynamic compromise a multiinstitutional
study of the role of endomyocardial cellular infiltrate. Cardiac Transplant Research Database, J Heart Lung Transplant
16 (1997), pp. 813–821
Winters G.L., B.M. McManus and Rapamycin Cardiac Rejection Treatment Trial Pathologists, Consistencies and
controversies in the application of the ISHLT working formulation for cardiac transplant biopsy specimens, J Heart
Lung Transplant 17 (1998), p. 754
Rodriguez E.R. and International Society for Heart and Lung Transplantation, The pathology of heart transplant biopsy
specimens revisiting the 1990 ISHLT working formulation, J Heart Lung Transplant 22 (2003), pp. 3–15
Marboe CC, Billingham M, ... Berry G. JHLT 2005;24:S219
allograft rejection
Recipient
immune
response
Rejection
Quiescence
T-cell activating signals
#1 T-cell receptor
#2 CD28
#2a CD40L
#3 IL2, IL15 etc
phenome
metabolome
TCR
Allo Ag
Allo Ag
allo MHC
self MHC
proteome
transcriptome
allo APC
monocyte
self APC
B-cell
CD3+
T-cell
CD8+
CD4+
rejection
CD8+
Cytotox
CD4+Th1
CD4+/45RO+
tolerance
CD8+28CD8+
suppr
CD4+/25+
CD4+Th
genome
Cardiac allograft
2
organ
systemic
phenome
echo
clinical
proteome
histo, Cd4
biomarkers
transcriptome
RT-PCR, ISH, array
Allomap
genome
DNA-sequencing
DNA-sequencing
intragraft cytokine expression
Alvarez CM et al. Clin Transplant 2001;15:228
intragraft cytokine expression
Alvarez CM et al. Clin Transplant 2001;15:228
correlation of IL6 with rejection
Author
Year
No
Time
Zhao
Van Hoffen
Baan
Kimball
Van Besouw
1994
1996
1996
1996
1997
21
40
16
62
85
early
first mo‘s
< 12 mo
< 30d
>1y
Method
RT-PCR
ISH/IHC
RT-PCR
ELISA
GIL-ELISA
Result
IL6, TGFß+
IL6,8,9,10+
IL4,IL6+
IL6,8+
IL6+ TxV
no correlation of IL6 with rejection
Author
Year
No
Time
Ruan
Fyfe
Van Besouw
Grant
Lagoo
Grant
George
1992
1993
1995
1996
1996
1996
1997
113
40
49
259
328
187
484
?
var
< 90d
<2y
< 8 wk
<2y
< 8 wk
Deng
Salom
1998
1998
115
22
<3mo
early
Method
Result
IHC
IL6-,IL2+,IFN+
ELISA
IL4, IL6,TNFculture IL6-,IL4-,IL2,IFN+
ELISA
IL6RT-PCR
IL6RT-PC/ELISA
IL6-,IL2+
ELISA
IL6,IL8,TNFELISA
IHC
IL6IL6-, IL1, TNF+
systems biology strategy
phenome/physiome
metabolome
proteome
0
transcriptome
genome
Phenotype 2
top-to-bottom
bottom-to-top
Phenotype 1
pathogenesis-based transcript sets
...We used mouse transplants to annotate pathogenesis-based transcript sets (PBTs) that
reflect major biologic events in allograft rejection—cytotoxic T-cell infiltration,
interferon-γ effects and parenchymal deterioration. We examined the relationship
between PBT expression, histopathologic lesions and clinical diagnoses in 143 consecutive
human kidney transplant biopsies for cause. PBTs correlated strongly with one another,
indicating that transcriptome disturbances in renal transplants have a stereotyped
internal structure. This disturbance was continuous, not dichotomous, across rejection
and nonrejection. PBTs correlated with histopathologic lesions and were the highest in
biopsies with clinically apparent rejection episodes. Surprisingly, antibody-mediated
rejection had changes similar to T-cell mediated rejection. Biopsies lacking PBT
disturbances did not have rejection. PBTs suggested that some current Banff
histopathology criteria are unreliable, particularly at the cut-off between borderline and
rejection...many transcriptome changes previously described in rejection are features of a
large-scale disturbance characteristic of rejection but occurring at lower levels in many
forms of injury. PBTs represent a quantitative measure of the inflammatory disturbances
in organ transplants, and a new window on the mechanisms of these changes.
Mueller TF et al. Am J Transplant 2007;7:1
endomycardial array & rejection
Karason K et al. BMC Cardiov Dis 2006;6:29
endomycardial array & rejection
Karason K et al. BMC Cardiov Dis 2006;6:29
endomycardial array & rejection
Karason K et al. BMC Cardiov Dis 2006;6:29
endomycardial array & rejection
Karason K et al. BMC Cardiov Dis 2006;6:29
endomycardial array & rejection
…Methods: Endomyocardial tissue samples and serum were obtained in connection
with clinical biopsies ... Endomyocardial RNA,..were analysed with DNA
microarray. Genes showing up-regulation during rejection followed by
normalization after the rejection episode were evaluated further with real-time RTPCR…ELISA was performed to investigate whether change in gene-regulation
during graft rejection was reflected in altered concentrations of the encoded protein
in serum…Results…CCL9 was significantly upregulated during rejection (p <
0.05)…There were no changes in CXCL9 and CXCL10 serum concentrations
during cardiac rejection…Conclusion: We conclude, that despite a distinct upregulation of CXCL9 mRNA in human hearts during cardiac allograft rejection, this
was not reflected in the serum levels of the encoded protein. Thus, in contrast to
previous suggestions, serum CXCL9 does not appear to be a promising serum
biomarker for cardiac allograft rejection. The lack of success in the identification of
cardiac rejection biomarkers in the current study indicates that expression profiling
of immunological active cells of the heart recipient may be a better way to identify
cardiac rejection biomarkers.
Karason K et al. BMC Cardiov Dis 2006;6:29
intragraft & PBL expression
…Our results demonstrate that PBL gene expression profiles in acute rejection are distinctly
different from those of normal controls and from patients with well-functioning transplants.
Therefore, acute rejection does influence the gene expression profile of the circulating
lymphocyte pool. Moreover, despite the fact that surprisingly we found very little common gene
expression between PBLs and kidney biopsies, we did identify a large number of lymphocytespecific genes in the kidney tissue. One interpretation is that there are compartment-specific
differences between the PBLs in the circulation and the subset of lymphocytes that are
activated and recruited to the transplant kidney during acute rejection. ..these results…may
explain the failure of more than a decade of work testing PBLs for an array of activation
antigens based on findings in rejecting allografts and other immune models…It is possible that
the gene expression profile of the PBLs represents the adequacy of immunosuppression such
that the rejecting patients reflect the profile of inadequate immunosuppression as compared
with the PBLs sampled from patients with well-functioning transplants... there is a distinct
gene expression profile in the PBL pool that correlates with acute rejection and
immunosuppression. If these results can be confirmed in a large, prospective trial it would
support the use of such profiles as a minimally invasive monitoring strategy for the
immunological status of the graft and support the potential of using them to monitor the
adequacy of immunosuppression…
Flechner SM et al. Am J Transplant 2004;4:1475
current standard of care
…the variability in the grading of heart transplant biopsies suggests the
biopsy itself may not be a true gold standard against which all
subsequent tests should be compared; this has clear implications for the
evaluation of any new molecular diagnostic test if the only end-point is
comparison with biopsy grade. Multifactorial end-points combining
clinical, hemodynamic and biopsy data would provide a better
standard. Indeed, the correlation of these multiple factors and
peripheral blood gene expression with biopsy histology may provide a
basis for further refining of the biopsy grading system by providing
insight into the histologic features that best correlate with immunologic
status and clinical outcomes.
Marboe CC, Billingham M, ... Berry G. JHLT 2005;24:S219
networked alloimmunity









detection of cytokine transcripts does not imply protein
detection of cytokine protein does not imply function
function may vary in different contexts
composite effects of multiple cytokines are rarely tested
unknown cytokines may be involved in rejection
cytokine polymorphisms may explain variations
in-vitro effects may not reflect in-vivo effects
animal data may not translate into in clinical data
> nonreductionist research approach necessary
Orosz CG. J Heart Lung Transplant 1996;15:1063
Baan et al. Transplant Int 1998;11:160
Bumgardner et al. Sem Liver Dis 1999;19:189
allograft rejection
Recipient
immune
response
Rejection
Quiescence
T-cell activating signals
#1 T-cell receptor
#2 CD28
#2a CD40L
#3 IL2, IL15 etc
phenome
metabolome
TCR
Allo Ag
Allo Ag
allo MHC
self MHC
proteome
transcriptome
allo APC
monocyte
self APC
B-cell
CD3+
T-cell
CD8+
CD4+
rejection
CD8+
Cytotox
CD4+Th1
CD4+/45RO+
tolerance
CD8+28CD8+
suppr
CD4+/25+
CD4+Th
genome
Cardiac allograft
2
organ
systemic
phenome
echo
clinical
proteome
histo, Cd4
biomarkers
transcriptome
RT-PCR, ISH, array
Allomap
genome
DNA-sequencing
DNA-sequencing
differential lymphocyte homing
Grant AJ et al. Lancet 2002;359:150
endomyocardial biopsy
mild
severe
status quo monitoring






invasive & complication-prone
late-stage cellular rejection diagnosis
insensitive for humoral rejection
significant variability
no insight into molecular mechanisms
resource-intense
future monitoring







highly sensitive for rejection
strong negative predictive value
positive test >need for further workup
non-invasive
easily repeatable on outpatient basis
low complication rate
decreased costs
systemic IL6 & HTx




15 pts, < 3 mo posttx, at EMB time
IL6 ELISA, RHC, echo
IL2 & rej 2+
IL6 & prognosis
300
250
200
150
100
50
0
IL6
0
5
10
15
RAP
20
25
30
200
150
100
50
0
pg/ml
MOF
stable
1
2
3
4 5 6
Bx time
7
Deng et al. Transplantation 1995;60:1118
8
HTx management
malaise
graft dysfunction
under
immunosuppression
over
infection
cellul rej
SIRS
humor rej
bolus steroids cyclophosph reduced IS antibiotics
Deng et al. Transplantation 1998;65:1255
CARGO clinical study summary
 Candidate gene selection
I
Discovery
~2 years
(microarray)
II
Development
~1 year
(PCR)
III
Clinical
Validation
~1 year
(Molecular Test)
Overview
 Cardiac Allograft Rejection Gene expression
 285 Leukocyte microarray
Observational study = “CARGO”
 Database / literature mining
 8 center, 4-year observational study initiated in
 252 candidate genes
2001 (22% of US HTx).
 629 patients, 4917 post-transplant encounters
 Algorithm development
 Hypothesis
 Real-time PCR
 Gene expression profiling of peripheral blood
 20-gene algorithm to distinguish
mononuclear cells can discriminate ISHLT grade 0
rejection from quiescence
rejection (quiescence) from moderate/severe
(AlloMap molecular testing)
(ISHLT grade ≥ 3A) rejection
 Design & Result
 Validation
Prospective, blinded validation study of 20 gene
 Prospective, blinded, statisticallyalgorithm demonstrated ability to distinguish
powered (n = 270)
Grade 3A rejection from quiescence

 Additional samples tested to further
define performance (n > 1000)
Deng/Eisen/Mehra et al. Am J Transplant 2006;6:150
Invasive Monitoring Attenuation through Gene Expression
IMAGE
Question
How do theGEP-based study restrictions affect clinical implementation?

Study Design
 Prospective
 Multi-center
 Non-blinded
 Randomized
 Non-inferiority

Patients
 2-5 years post-Tx
 ≥ 18 years old
 Stable outpatients
•
Hypothesis
To determine whether the monitoring of acute
rejectionusing GEP is not inferior compared to the
use of the EMB with respect to the event-free
survival
 Decrease in LV function, defined as LVEF
change ≥ 25% compared with the baseline,
or enrollment value, as measured by
echocardiography
 Development of clinically overt rejection
(heart failure, hemodynamic compromise)
 Death from any cause
ClinicalTrials.gov identifier NCT00351559
Pham/Deng/Kfoury et al. J Heart Lung Transplant 2007;26:808
Invasive Monitoring Attenuation through Gene Expression
Gene Profiling Arm
Study End
Year 2 - 3
Enrollment
Visit &
IMAGE
Year 4 - 5
Clinic
x
x
x
x
x
Echo
x
x
x
x
x
Randomization
GEP/EMB
x
x
x
Biopsy Arm
~2 year follow-up
x
x
Allomap implementation milestones

CARGO study start
(2001)

CARGO study completion/CLIA approval
(2005)

Allomap Medicare reimbursement
(2006)

FDA approval IVDMIA
(2008)

CARGOII Study
(2009)

IMAGE Study
(2010)
2000
2005
2010
regulatory transitions – CLIA>FDA





CLIA approval 2005
FDA approval 2008 – safety & efficacy
Center for Devices & Radiological Health
In-vitro diagnostic multivariate index assay IVDMIA
FDA-director Daniel Schultz comment:
“…Allomap is an example of how advancements in
science and technology are leading to new medical
care diagnostics…“
strategies in GEP test development
Phase
Phase 1
Tasks
CLINICAL PHENOTYPE CONSENSUS
DEFINITION
Challenges
• Imperfect Clinical/Phenotype Standards
• Dichotomous vs. Continuous Phenotype Choices
• Multicenter Study
Phase 2
GENE DISCOVERY
• General Gene Discovery Strategy
• Focused vs. Whole Genome Microarray
Phase 3
INTERNAL DIFFERENTIAL GENE LIST
VALIDATION
Phase 4
DIAGNOSTIC CLASSIFIER DEVELOPMENT
• Real time-PCR Validations
• Discriminatory vs. Classifier Genes
• Mathematical Modeling
• Biological Plausibility of the Diagnostic Test Gene List
• Clinical Test Replicability
Phase 5
EXTERNAL CLINICAL VALIDATION
• Independence of Primary Clinical Validation Cohort
• Prevalence Estimation of Clinical Phenotype of Interest
• Regulatory Approval
Phase 6
CLINICAL IMPLEMENTATION
Phase 7
POST-CLINICAL IMPLEMENTATION
STUDIES
• Payer Reimbursement
• Clinical Acceptance
Figure 2: Strategies in gene expression based biomarker test development.
Shahzad ,Sinha , Latif, Deng. Standardized operational procedures in clinical gene expression
biomarker panel development. In: Scherer A (Ed).John Wiley & Sons 2009
dimensionality problem overview
The application of several high-throughput genomic and proteomic technologies generate highdimensional data sets
•The multimodality of high-dimensional expression data can confound both simple mechanistic
interpretations of biology and the generation of complete or accurate gene signal transduction
pathways or networks.
•The mathematical and statistical properties of high-dimensional data spaces are often poorly
understood or inadequately considered, particularly if the number of data points obtained for
each specimen greatly exceed the number of specimens.
•Data are rarely randomly distributed in high-dimensions and are highly correlated, often with
spurious correlations.
•The distances between a data point and its nearest and farthest neighbours can become
equidistant in high dimensions, potentially compromising the accuracy of some distance-based
analysis tools.
•Owing to the ‘curse of dimensionality’ phenomenon and its negative impact on generalization
performance, the large estimation error from complex statistical models can easily compromise
the prediction advantage provided by their greater representation power.
•Some machine learning methods address the ‘curse of dimensionality’ in high-dimensional data
analysis through feature selection and dimensionality reduction, leading to better data
visualization and improved classification.
•It is important to ensure that the generalization capability of classifiers derived by supervised
learning methods from high-dimensional data is independently validated
Clarke R et al. Nat Rev Cancer 2008;8:37
invasive vs noninvasive algorithm
team
patient
assessment
clinical status?
team
patient
assessment
clinical status?
graft function?
graft function?
protocol biopsy grade?
clinical biopsy
treatment & re-biopsy
Allomap-score?
3-6mo <20
7-12 mo < 30
> 12 mo < 34
clinical biopsy
treatment & re-biopsy
no biopsy