Transcript Molecularly Targeting Therapy for Metastatic Colorectal Cancer

ADVANCES IN PERSONALIZED
AND PRECISION MEDICINE
Jeffrey Meyerhardt, MD, MPH
Dana-Farber Cancer Institute
Boston, MA
WHAT IS CANCER?
CANCER CELLS DON’T FOLLOW NORMAL RULES

Gene mutations, chromosome alterations, and changes in
other molecules enable them to replicate endlessly
Ignore signals ordering them to self-destruct
 Migrate to distant sites in the body, such as the bones, liver, or
brain, where they set up satellite tumors


Usually, several genomic abnormalities are driving the cancer
These can be different in different cancers that otherwise seem to
be the same.
 Two people with the same type of colon cancer, for example, may
not respond to treatment in the same way if their cancers are
driven by different combinations of mutations.

CANCER CELLS DON’T FOLLOW NORMAL RULES
 Cancer-driving
abnormalities may occur spontaneously
(de novo) or be passed from parent to child (hereditary).
Spontaneous changes happen all the time – thousands of times
a day in single cell – often due to DNA damage caused by
factors like sunlight, diet, or smoking
 Some abnormalities are harmless, some cause the cell with the
mutation to die, and some are fixed by the body’s elaborate
repair mechanisms or are eliminated before cell replication
occurs
 A very small minority of these changes, however, can evade
repair and alter the cell in a way that, rather than die, develops
an improved ability to grow or survive. These alterations lead to
cancer.

GENE ALTERATIONS THAT CAN DRIVE CANCER
 Single
nucleotide variation (SNV) or “point
mutation”


A single base nucleotide
May cause the gene to direct the cell to make an
abnormal protein or may prevent the gene from
making a protein
GENE ALTERATIONS THAT CAN DRIVE CANCER
 Insertions
and Deletions
GENE ALTERATIONS THAT CAN DRIVE CANCER
 Copy
number variation (CNV)
Each of us typically has two
identical copies of each gene.
 A copy number alteration
(CNV) is when one or both of
those copies are lost
(“deletion”) or when extra
copies are acquired
(“amplification”).
 CNVs can activate a cancerrelated gene or silence genes
that suppress tumors

GENE ALTERATIONS THAT CAN DRIVE CANCER
 Structural


alteration (SV)
A chromosome may have broken and segments become rearranged or
swapped with pieces of other chromosomes (“translocation”) or the
same chromosome (“inversion”).
When the chromosome breaks and rejoins (“breakpoints”), genes
around the break are altered which can change behavior and function
TRADITIONAL CHEMOTHERAPY
VERSUS TARGETED THERAPIES
 Standard
chemotherapy

Destroys rapidly
dividing cells, normal
and cancerous in a
wide range of tissues,
often causing side
effects like nausea,
mouth sores, and hair
loss.
TRADITIONAL CHEMOTHERAPY
VERSUS TARGETED THERAPIES
 Targeted
therapies
Engineered to attack tumor cells with specific
abnormalities.
 Block the growth and spread of cancer while leaving
normal cells largely unharmed.
 May strike directly at cells with specific genetic changes
that drive tumors’ development and survival, or to
inhibit overactive signaling pathways that allow cancer
cells to grow and divide uncontrollably
 Other treatments enlist the immune system to identify
and fight the cancer cells.

www.usatoday.com
METHODS TO TEST FOR GENETIC VARIATIONS

Single mutation tests


DNA microarray testing




Multiplex of key potential prognostic and predictive markers
Actionable genetic mutations
Next generation sequencing test can test for base pair substitutions,
insertions/deletions, copy number variations and rearrangements (depending on
which test used)
Whole exome sequencing


Eg. RAS or BRAF for colorectal cancers
Sequencing of the expressed genes in genome
Whole genome sequencing
SOMATIC V GERMLINE MUTATIONS
 Somatic
– within the tumor
 Germline – born with
 Adenomatous polyposis coli gene  familial adenomatous polyposis
 Mismatch repair genes  hereditary non-polyposis colorectal cancer (Lynch)
 MYH – MYH-associated polyposis
 LKB1 – Peutz-Jeghers syndrome
 SMAD4, BMPR1A -Juvenile polyposis
 PTEN _Cowden disease
DNA MICROARRAY
From National Human Genome Research Inst website
DNA MICROARRAY
DFCI PROFILE
DFCI PROFILE
DFCI PROFILE
DFCI WHOLE EXOME SEQUENCING – CANSEQ
CHANGE IN COST OF DNA SEQUENCING
CHANGE IN COST OF WHOLE GENOME SEQUENCING
CHROMOSOMAL VERSUS MICROSATELLITE INSTABILITY
LEADING TO COLORECTAL CANCER
Markowitz and Bertagnolli. NEJM 361: 2449-60
MISMATCH REPAIR MECHANISM
Banno K, Yanokura M, Kobayashi Y, Kawaguchi M, Nomura H, Hirasawa A, Susumu N, Aoki D - Curr. Genomics (2009)
MOLECULAR DIVERSITY OF COLORECTAL CANCER
(16%)
(84%)
TCGA Network. Nature. 2012; 487: 330-337.
MOLECULAR DIVERSITY OF COLORECTAL CANCER
TCGA Network. Nature. 2012; 487: 330-337.
TCGA BOTTOM LINE
 Colorectal
 No
cancers are molecularly very diverse
single driver that leads to cancer
 Backup
 One
pathways when one pathway not working right
strategy will not work for everyone’s cancer even
with a particular known mutation
EXAMPLES OF PRECISION MEDICINE
FOR COLORECTAL CANCER
PREDICTORS OF EPIDERMAL GROWTH
FACTOR RECEPTOR INHIBITORS
EPIDERMAL GROWTH
FACTOR RECEPTOR IN CRC
Ligand
Extracellular
EGFR
PTEN
PI3K
Akt
Ras
Raf
MEK
MAPK
Cell survival
Proliferation
DNA
Angiogenesis
Intracellular
Cell motility
Metastasis
PHASE 4 STUDY: CETUXIMAB VS BEST SUPPORTIVE CARE
• Patients
previously
treated with
fluoropyrimidine,
oxaliplatin,
irinotecan
• EGFR expression
required
R
A
N
D
O
M
I
Z
E
Cetuximab
400 mg/m2
loading then
250 mg/m2
weekly
N = 287
Stratification based on
ECOG score (0/1 v 2)
and center
Best supportive
care
N = 285
Jonkers et al. N Engl J Med 2007; 357:2040-2048
Kaplan–Meier Curves for Overall Survival (Panel A) and Progression-free Survival (Panel B).
PHASE
4 STUDY: CETUXIMAB VS BEST SUPPORTIVE CARE
Jonkers et al. N Engl J Med 2007; 357:2040-2048
FINDING A PREDICTIVE MARKER
Khambata-Ford S et al. J Clin Oncol. 2007;25:3230-3237. \
for Progression-free Survival According to Treatment.
PHASE Kaplan–Meier
4 STUDY:Curves
CETUXIMAB
VS BEST SUPPORTIVE CARE
KARAPETIS CS ET AL. N ENGL J MED 2008;359:1757-1765.
Curves for Progression-free Survival According to Treatment.
FKaplan–Meier
IRST LINE
PANITUMUMAB: PRIME STUDY
Douillard J Y et al. Ann Oncol 2014;25:1346-1355
BEYOND KRAS CODON 12 AND 13
Data from 773 patients treated with cetuximab at 11 European centres
De Roock et al. Lancet Oncology. 2010. Volume 11, Issue 8, Pages 753-762
OTHER RAS MUTATIONS: CRYSTAL TRIAL
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
HR (95% CI) 0.70 (0.56–0.87)
FOLFIRI + cetuximab
FOLFIRI
0
4
8
12
16
RAS wild-type
Probability of PFS
Probability of PFS
KRAS codon 12/13 wild-type
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
HR (95% CI) 0.56 (0.41–0.76)
FOLFIRI + cetuximab
FOLFIRI
20
0
1
0
Number of patients at risk
178
153
114
75
189
154
92
44
3
6
9
Months
Number of patients at risk
316
227
350
237
128
111
40
22
8
4
12
Months
31
11
15
18
21
24
8
5
4
3
0
0
0
0
Ciardiello et al. ASCO 2014
BRAF MUTATIONS IN LATER LINE THERAPY
Author
Lambrechts1†
Ruzzo2‡
Di Nicolantonio3‡
Treatment
BRAF
wt : mt
BRAF mt
frequency
[in KRAS wt]
Response Rate
BRAF
mt vs wt
Cetuximab + irinotecan in chemorefractory pts
540 : 26
4.6%
[not reported]
8 vs 26%
Cetuximab + irinotecan in
irinotecan-refractory pts
57 : 9
8%
[14%]
0 vs 33%‡
Panitumumab
or cetuximab monotherapy or
cetuximab + CT
68 : 11
9.7%
[13.9%]
0 vs 32%‡
*Significant; †All patients; ‡KRAS wt patients only (Ruzzo, n = 66 of 117; Di Nicolantonio, n = 79 of 113); §Data not available.
1Lambrechts
D et al. J Clin Oncol. 2009;27:15s: Abstract 4020.
A et al. J Clin Oncol. 2009;27:15s: Abstract 4058.
3Di Nicolantonio F et al. J Clin Oncol. 2008;26:5705-5712.
2Ruzzo
BRAF MUTATIONS IN 1ST LINE THERAPY
Probability of OS (%)
KRAS WT/BRAF WT
No. of patients
CT + cetuximab
CT
CT + cetuximab
CT
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
HR = 0.840; 95% CI, 0.710-0.993
P = 0.041
FOLFIRI/FOLFOX4 + cetuximab: (n = 349)
FOLFIRI/FOLFOX4: (n = 381) median
KRAS WT/BRAF MT
HR = 0.633; 95% CI, 0.378-1.060
P = 0.079
FOLFIRI/FOLFOX4 + cetuximab: (n = 32)
FOLFIRI/FOLFOX4: (n = 38)
0
6
12
18
24
349
381
32
38
317
350
25
24
268
283
16
14
225
212
12
6
163
149
8
6
30
36
42
48
54
60
120
107
5
3
80
63
2
3
63
46
2
1
19
17
2
0
4
2
0
0
0
0
0
0
Time (months)
Bokemeyer et al. European Journal of Cancer, 2012-07-01, Volume 48, Issue 10, Pages 1466-1475
POTENTIAL UPFRONT RESISTANCE TO ANTI-EGFR MABS
IN KRAS WT CRC
Sensitive
population
Dienstmann, Salazar & Tabernero, ASCO Educational Book 2014
MORE ON BRAF MUTATED TUMORS
SINGLE AGENT ACTIVITY OF VEMURAFENIB IN
BRAF-MUTANT MCRC
Modest activity in BRAF-mutant mCRC compared with BRAF mutant melanoma
(Response rate of ~ 5% vs ~ 60%-80%[2])
%Change From Baseline
(Sum of Lesion Size)
100
mCRC (N = 19)[1]
75
50
25
0
-25
-50
-75
(RECIST cutoff for PR, 30%)
-100
1. Kopetz S, et al. ASCO 2010. Abstract 3534. 2. Flaherty KT, et al. N Engl J Med. 2010;363:809-19.
LABORATORY DISCOVERIES ON WHY BRAF INHIBITORS
DON’T WORK BY THEMSELVES IN COLORECTAL CANCER
 Need
to target more than one pathway

BRAF inhibitors increase level of EGFR receptors
 Strategy - combine BRAF and EGFR Inhibitor

Lack of sustained suppression of another protein
(ERK) with BRAF inhibitors alone
 Combined inhibition of BRAF and MEK may lead to
superior suppression of this signal
Prahallad A, et al. Nature. 2012;483:100-103. Corcoran RB, et al. Cancer Discov. 2012;2:227-235.
Signaling in BRAF
mutated CRC
Reactivation of EGFR
signaling upon BRAF
inhibition
Robust inhibition of
MAPK pathway
signaling with
inhibition of BRAF,
MEK, EGFR
Bendell. et al. ASCO 2014.
COBRIM UPDATED PFS: ASCO ‘15
Survival Distribution Function (%)
Kaplan-Meier Plot for PFS
Intent-to-Treat Population
100
+
+ +
80
+
++
+
60
20
+
Vemurafenib + cobimetinib
Vemurafenib + placebo
+
0.58b
(0.460-0.719)
++++
+ ++
Median follow-up was 14.2 months
+
+++++++ +++
+
+++
Data cutoff of January 16, 2015 was 1 year
++
++ +++
+ + +
from enrollment of last patient
+++ + ++++
++
++++ ++
+++
+++ +
++
+
+++
+ +++++
Cobimetinib + vemurafenib (n=247)
+
+
Placebo + vemurafenib (n=248)
Censored
1 Months
No. of patients at risk
HRa
(95% CI)
+
+
40
0
+
5 Months
9 Months
13 Months
17 Months
21 Months
25 Months
Time
238
240
215
205
190
150
168
115
142
87
116
67
79
45
46
30
21
17
8
3
1
Larkin et al ASCO ‘15 Abs 9006
D+P+T (N = 35)
CR+PR: 9 (26%)
Stable disease: 18 (51%)
Color: confirmed response
Height of bar: best unconfirmed response
Maximum % Change from Baseline
D+P (N=20)
CR+PR: 2 (10%)
Stable disease: 16 (80%)
100
80
60
40
20
0
-20
-40
-60
-80
-100
Maximum % Change from Baseline
DABRAFENIB (BRAF-I) + PANITUMUMAB (EGFR-I) +/TRAMETINIB (MEK-I) IN BRAFV600-MUTANT MCRC
100
80
60
40
20
0
-20
-40
-60
-80
-100
Progressive disease
Stable disease
Partial response
Complete response
+
+
*Maximum reduction from baseline is 0%
+
+
+ + + +
*
+
+ + + +
+ +
+ + + +
+ + +
+ + +
+
*Maximum reduction from baseline is 0%
+RP2R cohort
Atreya et al ASCO ‘15 Abs 103
EFFICACY OF DABRAFENIB (BRAF-I) + PANITUMUMAB (EGFR-I) +/TRAMETINIB (MEK-I) IN BRAFV600-MUTANT MCRC – TIME ON STUDY
Bendell. et al. ASCO 2014.
CHECKPOINT INHIBITORS AND
MISMATCH REPAIR DEFICIENT CRC
PD1/PDL1 AND INHIBITORS
DURABLE RESPONSE WITH ANTI-PD1 ANTIBODY
Lipson E J et al. Clin Cancer Res 2013;19:462-468
Pembrolizumab Single Agent in MSI and MSS CRC
LE DT ET AL. N ENGL J MED 2015;372:2509-2520.
LE DT ET AL. N ENGL J MED 2015;372:2509-2520.
HER2 TUMORS – ROLE OF
TRASTUZUMAB/LAPATINIB
HER2 PATHWAY
HERCALES TRIAL FOR HER2 +
METASTATIC COLORECTAL CANCER
849 patients with mCRC KRAS exon 2 WT
Trastuzumab iv 4mg/kg load
and then 2mg/kg/qw
Lapatinib po 1000 mg/qd
803 HER2-negative
46 HER2+ (5.4%)
22 not eligible because PS ≥2 or tumor-related
comorbidities
24 enrolled
PD
1 too early for safety & efficacy assessment
23 evaluable for response
HERCALES TRIAL FOR HER2+ METASTATIC CRC
Best Response
RECIST 1.1 by centralized revision
N
%
Responses (PR+CR)
8
34.7
Complete Response
1
4.3
Partial Response
7
30.4
Stable Disease >4 mos
7
30.4
Stable Disease <4 mos
3
13.0
Progressive Disease
5
21.7
23
100
Total
Primary endpoint met in advance with 8/23 objective responses
(as per protocol, 6/27 needed to declare the study positive)
78%
disease
control
HERCALES TRIAL FOR HER2+ METASTATIC CRC
RESPONSE BY HER2 SCORING
Spaghetti plot
Waterfall plot
NEW LESION
Change in target lesion from baseline (%)
Patients on treatment
PD
HER2 2+
*3 patients are not shown: 122026 (IHC 2+), not assessed yet; 121011 (IHC 3+) and 121013 (IHC 3+) early clinical PD.
56
Change in target lesion from baseline (%)
HER2 3+
CONCLUSIONS FOR MOLECULAR DEFINED
COLORECTAL CANCER THERAPY


Colorectal cancer is a molecularly diverse disease
Predictive markers growing






RAS testing should be performed for all metastatic CRC patients
BRAF may not predict anti-EGFR therapy but should be tested for
potential trials
MSI testing for genetic screening and for checkpoint inhibitor trials
HER2 testing may be important as well
While mutations are found, finding agents to target most of
what is found has been a challenge
I would recommend next generation sequencing for most
patients – role of whole exome and whole genome testing is
more uncertain – lots of data, need to understand how to use it