Liquid Biopsy in the clinic: currentstatus and future development

(1)

Liquid Biopsy in the clinic: current status and future development

Marina Chiara GARASSINO

(2)

I will focus my attention on

ctDNA and clinical data

(3)

Unknown KRAS

EML4-ALK

HER2 BRAF PIK3CA AKT1 NTRK ROS1

NRAS KIF5B-RET EGFR

Target therapy 2016

Pao, Hutchinson. Nat Med 2012;18:349–351

(4)

Kris M, JAMA 2014

(5)

Two different worlds and approaches

WILD TYPE TARGET THERAPIES

TARGET THERAPIES CHEMOTHERAPY

IMMUNOTHERAPY CHEMOTHERAPY

“MUTATED”

(6)

Two different worlds and approaches

“MUTATED” WILD TYPE

TARGET THERAPIES TARGET THERAPIES

CHEMOTHERAPY

IMMUNOTHERAPY

CHEMOTHERAPY

(7)

Resistance to EGFR TKIs

FIRST AND OBVIOUS APPLICATION

(8)

Erlotinib

Erlotinib v v Chemotherapy in EGFR Driven Chemotherapy in EGFR Driven Advanced NSCLC

Advanced NSCLC

Oxnard et al CCR 2011

Rosell et al. Lancet Oncology 2012

EGFR driven NSCLC

TKI treatment

(9)

Resistance to EGFR TKIs

• Camidge DR et al. Nat Rev Clin Oncol.

2014;11:473–81

(10)

T790M: Severe structural changes in ATP-binding pocket

Michalczyk A et al. Bioorg Med Chem. 2008 Apr 1;16(7):3482-8.

Structural changes of the ATP pocket for greater steric hindrance of methionine compared to threonine

Treonine (~116 A) Metionine (~163 A)

T790M Exon 20 substitution

(11)

Treatment at

Progression

(12)

Systemic PD: therapeutic approaches

Continue targeting of sensitive cells

• Continue same TKI (slow PD) Target resistant cells

• Switch therapy to – Chemotherapy

– Third-generation TKI

Target sensitive and resistant cells

• Continue TKI + add new agent – Concurrent or intercalated

chemotherapy

– Additional targeted agent

Sensitive cell Resistant cell

Tumour(s)

at baseline Tumour(s)

at PD

(13)

T790M Inhibitors

(14)

T790M Inhibitors

(15)

– EGFR mutant-specific, i.e. sparing wild type EGFR

– Activity against sensitizing EGFR mutations (Del19, L858R, L861Q etc)

– Activity against T790M resistance mutation

Inhibitory Concentrations of TKIs Against EGFR Proteins

Zhou W et al. Nature 2009

(16)

T790M EGFR mutant-specific kinase inhibition

Zhou W et al. Nature 2009

(17)

Phase I/II studies of osimertinib…AURA and AURA2

data cut off

date

Meeting/J

ournal Year Author N

ORR for T790M

mPFS for T790M

Waterfall plot

AURA Ph 1 Various

20-240mg 9.27.201

3 WCLC 2013 Ranson 34 50% N/A

AURA Ph 1 Various

20-240mg 4.27.201

4 ASCO 2014 Janne 107 64%

AURA Ph 1 Various

20-240mg 8.1.2014 NEJM 2015 Janne 253

(138 T790M) 61% 9.6m

AURA Ph 1 Various

20-240mg 12.2.201

4 ELCC 2015 Janne 283 59% 13.5m

AURA extension 80mg 5.1.2015 WCLC 2015 Yang 201 61% Not

reached

AURA 2 80mg 11.1.201

5 Lancet Oncol 2016 Goss 210 70% 9.9m

AURA ext + AURA 2 80mg 11.1.201

5 ELCC 2016 Yang 397 66% 11.0m ^*********

Presented by T. Mitsudomi at the World Conference on Lung Cancer 2016

Plenary Session: PL03 Presidential Symposium Discussion about abstract PL03.03

(18)

Overview of plasma analyses of ctDNA in AURA trials

Across the AURA trials, plasma was collected for analyses to determine whether genotyping of plasma ctDNA could identify patients who gain clinical benefit from osimertinib

• BEAMing, beads, emulsion, amplification, and magnetics; ctDNA, circulating tumor deoxyribonucleic acid; ddPCR, droplet digital polymerase chain reaction; NGS, next- generation sequencing, QD, once daily.

Phase III study:

AURA3¹ Phase II studies:

AURA extension and AURA2² Phase I study: AURA³ Treatment/dosing Osimertinib 80 mg QD vs platinum

pemetrexed Osimertinib 80 mg QD Osimertinib dose-escalation and dose- expansion cohorts (20–240 mg QD)

Tissue T790M status T790M-positive T790M-positive T790M-positive and -negative cases

Analysis Pre-planned analysis; plasma collected contemporaneous with tissue and tested

retrospectively

Pre-planned for regulatory

submission Exploratory post hoc analysis

Plasma assay cobas^® cobas^® BEAMing

Method of

comparison cobas^®FFPE tissue NGS ddPCR or cobas^®FFPE tissue

Number of patients

(n=399 T790M positive by tissue test; 399 n=184 plasma T790M positive;

n=175 T790M plasma negative;

n=40 missing/invalid)

(n=401 AURA extension; n=472 in 873

AURA2) 216

(19)

216 comprised the eligible study population

Data cutoff: May 1, 2015.

EGFR, epidermal growth factor receptor.

Oxnard GR, et al. J Clin Oncol. 2016;34(28):3375-3382.

19

308 patients eligible for this biomarker analysis

71 patients with no central tumor genotyping results

37 patients with no central plasma genotyping results

216 patients eligible for diagnostic comparison, with both central tumor and

plasma genotyping available 237 patients eligible for analysis of

tumor genotype and outcome

271 patients eligible for analysis of plasma genotype and outcome

94 patients excluded:

• 60 previously untreated

• 9 with a known EGFR mutation other than L858R or exon 19 deletion

• 25 with no known EGFR mutation by tissue or plasma genotyping

402 patients enrolled onto AURA Phase I escalation and expansion cohorts

(20)

Sensitivity/specificity of plasma genotyping revealed a higher rate of false positives for T790M

ctDNA, circulating tumor deoxyribonucleic acid.

20

• Sensitivity was 82%–86% for sensitizing mutations and 70% for T790M mutation

• False-positive rate was 2%–3% for sensitizing mutations but higher (31%) for T790M, perhaps due to heterogeneous presence of a resistance mutation missed in the reference tumor biopsy

• Sensitivity for T790M was highly associated with detection of a sensitizing mutation in ctDNA

• The specificity was 97.5% and 96.5% for exon 19 and L858R mutations, respectively.

100 10

1 0.1 0.01 0.001 N/D

Sensitivity of each assay

82.3% 86.3% 70.3%

0.06 0.04

19 del n=136

L858R n=73

T790M n=158

Allelic fraction (%)

Specificity of each assay

100 10 1 0.1 0.01 0.001 N/D

97.5% 96.5% 69.0%

19 del n=80

L858R n=143

T790M n=58

Sensitivity of T790M assay

plasma sens positive

n=137

plasma sens negative

n=21

80.3% 4.8%

100 10

1 0.1 0.01 0.001 N/D 0.06

0.04

0.06

(21)

High ORR in patients with tumor- or plasma-positive T790M

CI, confidence interval; ORR, objective response rate.

21

**

100 60

–20 –60 –100 20 80 40

–40 –80 0

Plasma T790M positive Plasma T790M negative Plasma T790M unknown ORR (95% CI): 62% (54, 70) Tumor T790M positive (n=173)

*

100 60

–20 –60 –100 20 80 40

–40 –80 0

Tumor T790M positive Tumor T790M negative Tumor unknown ORR (95% CI): 63% (55, 70) Plasma T790M positive (n=164)

* * * * *

100 60

–20 –60 –100 20 80 40

–40 –80 0

Plasma T790M positive Plasma T790M negative ORR (95% CI): 26% (15, 39) Tumor T790M negative (n=58)

* * * ** *

100 60

–20 –60 –100 20 80 40

–40 –80 0

ORR (95% CI): 46% (36, 56) Plasma T790M negative (n=102) Tumor T790M positive Tumor T790M negative Tumor unknown

ORR (95% CI) P value Plasma T790M positive 63% (55, 70)

0.011 Plasma T790M negative 46% (36, 56) ORR (95% CI) P value

Tumor T790M positive 62% (54, 70)

< 0.001 Tumor T790M negative 26% (15, 39)

(22)

PFS by tumor and plasma T790M status: T790M negative via plasma testing better than expected attributed to false negatives

Data cutoff: May 1, 2015. Multiple doses included.

CI, confidence interval; PFS, progression-free survival.

• Median PFS is dramatically different based on tumor T790M status (P<0.001):

– 9.7 months in tumor T790M positive – 3.4 months in tumor T790M negative

• Plasma T790M-positive status predicts for a prolonged PFS (9.7 months) on osimertinib, similar to what is seen based on central tumor genotyping

• However, median PFS is not significantly different based on plasma T790M status (P=0.188)

– 9.7 months in plasma T790M positive – 8.2 months in plasma T790M negative

22

0 3 6 9 12 15 18 21 24

100 80 60 40 20 0

All patients with plasma T790M results

Time from first dose (months) Probability of progression-free survival

Plasma T790M negative (n=104) Plasma T790M positive (n=169)

Median PFS (95% CIs) Plasma T790M positive 9.7 (8.3, 11.1) Plasma T790M

negative 8.2 (5.3,

10.9) Log-rank test P=0.188

24 80

60 40 20 0

0 3 6 9 12 15 18 21

All patients with tumor T790M results

Probability of progression-free survival

Tumor T790M negative (n=58) Tumor T790M positive (n=179)

Median PFS (95% CIs) Tumor T790M positive 9.7 (8.3, 12.5) Tumor T790M negative 3.4 (2.1, 4.3)

Log-rank test P<0.001

Time from first dose (months)

(23)

Detection of sensitizing mutation as a control creates separation of PFS curves

• In the 104 patients with T790M-negative plasma genotyping, detection of the sensitizing mutation was studied to help inform true negative versus false negative

– Plasma T790M negative/sensitizing positive: 38% ORR, 4.4 month median PFS – Plasma T790M negative/sensitizing negative: 64% ORR, 15.2 months median PFS

• If plasma T790M negative/sensitizing negative are excluded from PFS analysis reflecting their unknown plasma T790M mutation status, a significant difference is seen between T790M positive and T790M negative

23

• Data cutoff: May 1, 2015.

• CI, confidence interval; PFS, progression-free survival.

• Oxnard GR, et al. J Clin Oncol. 2016;34(28):3375-3382.

24

0 3 6 9 12 15 18 21

100

Plasma T790M negative (n=104) Plasma T790M positive (n=169)

Median PFS (95%

CIs) Plasma T790M positive 9.7 (8.3, 11.1) Plasma T790M negative 8.2 (5.3, 10.9)

Log-rank test P=0.188

Median PFS (95% CIs)

Plasma T790M positive 9.7 (8.3, 11.1)

Plasma T790M negative/sensitizing positive 4.4 (2.8, 6.8) Plasma T790M negative/sensitizing

negative 15.2 (11.0, 17.9)

Log-rank test P=0.002

24

0 3 6 9 12 15 18 21

100 80 60 40 20 Probability of progression-free survival 0

Plasma T790M negative / sensitizing positive (n=69) Plasma T790M positive (n=169)

Plasma T790M negative / sensitizing negative (n=35)

80 60 40 20 Probability of progression-free survival 0

(24)

Proposed paradigm for use of plasma genotyping at relapse

These data support consideration of a paradigm where plasma genotyping is used as a screening test for T790M, prior to performing an EGFR resistance biopsy

24

• EGFR, epidermal growth factor receptor; FDA, Food and Drug Administration; FFPE, formalin-fixed, paraffin-embedded;

TKI, tyrosine kinase inhibitor. Oxnard GR, et al. J Clin Oncol. 2016;34(28):3375-3382.

A. Current paradigm

Acquired resistance to EGFR-TKI

All patients undergo biopsy, FDA-approved FFPE assay for T790M

Third-gen. EGFR-TKI

Chemotherapy

B. Proposed paradigm for use of plasma diagnostics Acquired resistance to EGFR-TKI

FDA-approved plasma assay for T790M and

sensitizing mutations T790M negativ

e T790M positiv

e

Skip biopsy, start third-gen. EGFR-TKI

Third-gen. EGFR-TKI

Chemotherapy Biopsy, FDA-approved

FFPE assay for T790M

T790M positive

T790M negative T790M

negative T790M positiv

e

(25)

• Analysis of PFS by BICR was consistent with the investigator-based analysis: HR 0.28 (95% CI 0.20, 0.38), p<0.001;

median PFS 11.0 vs 4.2 months.

Population: intent-to-treat

Progression-free survival defined as time from randomisation until date of objective disease progression or death. Progression included deaths in the absence of RECIST progression.

Tick marks indicate censored data; CI, confidence interval

AURA3 primary endpoint:

PFS by investigator assessment

1.0

0.8

0.6

0.4

0.2

0

0 3 6 9 12 15 18

Probability of progression-free survival

No. at risk Osimertinib Platinum-pemetrexed

Months 279

140

240 93

162 44

88 17

50 7

13 1

0 0

Median PFS, months (95%

CI)

HR (95% CI)

10.1 (8.3,

12.3) 0.30 (0.23,

0.41) p<0.001 4.4 (4.2, 5.6)

Osimertinib Platinum-pemetrexed

(26)

AURA3: plasma sample collection

Data cut-off April 15, 2016.

EGFR, epidermal growth factor receptor; FFPE, formalin-fixed, paraffin-embedded.

1. Wu Y-L, et al. Presented at: IASLC 17^thWorld Conference on Lung Cancer; December 4-7, 2016; Vienna, Austria. Abs MA08.03. 2. Mok TS, et al. Supplementary material. N Engl J Med. 2017; 376:629-640.

26

N=1036² Screened patients

• n=184 plasma T790M positive

• n=172 randomized to treatment n=359

Tumor T790M positive • n=175 plasma T790M negative

• n=168 randomized to treatment

Matched plasma sample

n=47 plasma T790M positive n=205

Tumor T790M negative n=158 plasma T790M negative

•

Contemporary FFPE tissue biopsies were prospectively tested with the cobas

^®

EGFR Mutation Test and matched plasma

samples were collected at screening and retrospectively tested with the cobas

^®

EGFR Mutation Test v2

¹

(27)

AURA3: T790M mutation is detected in plasma of ~50% of patients with

T790M in tumor tissue

Data cut-off April 15, 2016.

*Percent agreement of the cobas^®plasma test with the cobas^®tissue test. Positive percent agreement and negative percent agreement are used here as measures of test sensitivity and specificity, respectively, and calculated with invalid results excluded.

Wu YL, et al. J Thorac Oncol. 2017; 12(S1), abstract MA08.03.

27

Plasma ctDNA test results, n Tissue T790M positive

(n=399) Tissue Exon 19 deletion

positive (n=427) Tissue L858R positive (n=253)

Plasma positive 184 273 139

Plasma negative 175 60 67

No plasma test / invalid 37 / 3 91 / 3 47 / 0

Percent agreement using tissue test as

reference, % (95% CI)^* T790M Exon 19 deletion L858R

Positive percent agreement (sensitivity) 51 (46, 57) 82 (77, 86) 68 (61, 74)

Negative percent agreement (specificity) 77 (71, 83) 98 (96, 100) 99 (98, 100)

Overall concordance 61 (57, 65) 89 (86, 91) 88 (85, 90)

• Patients with tissue sample available at screening (n=756)

• 51% sensitivity and 77% specificity for T790M detection using cobas

^®

tissue test as reference

• High sensitivity and specificity is observed for Exon 19 deletion and L858R

(28)

65

AURA3: osimertinib benefit in patients with plasma T790M-positive status is similar to patients with tumor

tissue T790M-positive status ^1-3

• Data cut-off April 15, 2016. Tick marks indicate censored data.

• PFS is defined as time from randomization until date of objective disease progression or death. Progression included deaths in the absence of RECIST progression. Osimertinib administered 80 mg orally once daily. Platinum-pemetrexed group treatment consisted of:

pemetrexed 500 mg/m²+ carboplatin AUC5 or cisplatin 75 mg/m²Q3W for up to 6 cycles + optional maintenance pemetrexed for patients whose disease had not progressed after 4 cycles of platinum-pemetrexed. RECIST v1.1 assessments performed every 6 weeks until objective disease progression.

• *PFS adjusted for ethnicity. All patients were selected using a tumor tissue test for EGFR T790M (by cobas^®EGFR Mutation Test) from a biopsy after disease progression prior to study entry. ^†Response did not require confirmation per RECIST v1.1.

• CI, confidence interval; HR, hazard ratio; ORR, objective response rate; PFS, progression-free survival; RECIST, Response Evaluation Criteria In Solid Tumors

• 1. Mok TS, et al. N Engl J Med. 2017; 376:629-640. 2. Suppl. Info for: Mok TS, et al. N Engl J Med. 2017; 376:629-640. 3. Wu YL, et al. J Thorac Oncol. 2017; 12(S1), abstract MA08.03.

Tumor T790M-positive (intent-to-treat)* Plasma T790M-positive status

Osimertinib Platinum- pemetrexed Median PFS, months (95% CI) 8.2 (6.8, 9.7) 4.2 (4.1, 5.1)

PFS HR (95% CI) 0.42 (0.29, 0.61)

ORR,^†% (95% CI) 77 (68, 84) 39 (27, 53)

Osimertinib Platinum- pemetrexed Median PFS, months (95% CI) 10.1 (8.3, 12.3) 4.4 (4.2, 5.6) PFS HR (95% CI) 0.30 (0.23, 0.41),* P<0.001

ORR,^†% (95% CI) 71 (65, 76) 31 (24, 40)

No. at risk Osimertinib

1.0 0.8 0.6 0.4 0.2

0 0 3 6 Months9 12 15 18

279 140

240 93

162 44

88 17

50 7

13 1

0 0 Osimertinib (n=279) Platinum-pemetrexed (n=140)

No. at risk Osimertinib

0.8 0.6 0.4

0.2

0

0 3 6 9 12 15 18

116 56

95 39

63 13

35 5

20 2

5 1

0 0 Months

Osimertinib (n=116) Platinum-pemetrexed (n=56)

Probability of progression-free survival Probability of progression-free survival

Platinum- pemetrexed Platinum-

pemetrexed

(29)

Wu and colleagues: conclusions

• In AURA3, T790M mutation is detected by cobas® EGFR Mutation Test v2 in plasma of approximately half (51%) of those patients with T790M in tumor tissue

– There are differences in the detection of T790M using tissue and plasma samples. These differences may reflect tumour biology; however, further studies are warranted

• Patients with a plasma T790M positive status show

comparable responses to osimertinib with patients who are tissue T790M positive (AURA3: intent-to-treat

population)

• Biopsy testing is recommended for patients with a plasma T790M-negative test where feasible

29

• EGFR, epidermal growth factor.

• Wu YL, et al. J Thorac Oncol. 2017; 12(S1), abstract MA08.03.

(30)

EGFR Mut+ NSCLC Clonal Evolution Under TKI Pressureisease to AZD9291 in EGFR

DS Costa & Kobayashi et al. Nature Medicine 2016

(31)

Mechanisms of Resistance to AZD9291 in EGFR T790M Positive Lung Cancer

• 15 (22%) out of 67 patients, had detectable C797S, all with detectable T790M

• C797S was more common with EGFR exon 19 del (13/43, 30%) vs those with L858R (2/24, 8%, p=0.06)

• 32 of 67 (48%) had no detectable T790M in plasma despite presence of the EGFR-TKI-sensitizing mutation, suggesting overgrowth of an alternate resistance mechanism, such as MET or HER2 amplification or BRAF V600E

Thress KS et al. Nature Medicine 2015

Oxnard G et al. MINI 17.07

(32)

32

Long-Hua Guo, Xu-Chao Zhang, Zhi-Hong Chen, Jian Su, Jin-Ji Yang, Chong-Rui Xu, Zhi Xie, Wei-Bang Guo, Hong- Hong Yan, Xue-Ning Yang, Wen-Zhao Zhong, Qiu-Yi Zhang, Yi-Long Wu, Qing Zhou

Intratumor Heterogeneity of EGFR Activating Mutations Analyzed in Single Cancer Cells in

Advanced NSCLC Patients

Guangdong Lung Cancer Institute, Guangdong General Hospital &

Guangdong Academy of Medical Sciences, Guangzhou/China

ORAL 16.07 – Q Zhang

MEK Inhibitors can overcome MAPK pathway

re-activation

(33)

TATTON study – ongoing

NCT02143466

EGFR-TKI naïve:

AZD9291 + durvalumab Dose 2

AZD9291 (qd) + durvalumab (q 2 weeks)

Dose 2 – continuous

AZD9291 (qd) + selumetinib (bid) Asia

Dose 2 – continuous

AZD9291 (qd) + selumetinib (bid) ROW

Dose 2 – intermittent: 4 days on / 3 days off AZD9291 (qd) + selumetinib (bid) ROW

Dose 2

AZD9291 (qd) + savolitinib (qd)

Acquired resistance to initial EGFR-TKI, cMET negative:

AZD9291 + selumetinib

Acquired resistance to initial EGFR-TKI, cMET positive:

AZD9291 + savolitinib Dose 2

AZD9291 (qd) + durvalumab (q 4 weeks)

Dose 2

AZD9291 (qd) + durvalumab + tremelimumab (q 4 weeks) Part A – Dose escalation

(all with acquired resistance to EGFR-TKI)

Part B – Dose expansion (different lines of treatment)

Acquired resistance to T790M-directed EGFR-TKI, cMET negative:

AZD9291 + selumetinib

Acquired resistance to T790M-directed EGFR-TKI, cMET positive:

AZD9291 + savolitinib

(34)

Algorithm at PD during EGFR TKIs

PD during EGFR TKIs tx

T790M+ T790M-

AZD9291 or clinical trial Chemotherapy

(+TKI) Targeted agents in clinical trials T790M- / other mutations

Mutational status evaluation

(35)

Resistance to ALK TKIs

(36)

Oncogene driven NSCLC tailored treatment

Erlotinib vs. chemotherapy EGFR-driven advanced in

NSCLC ¹

1. Rosell R, et al. Lancet Oncol 2012;13:239–46;

2. Shaw T, et al. N Engl J Med 2013;368:2385–94.

Crizotinib vs.

chemotherapy in ALK+

NSCLC ²

(37)

Acquired resistance in EGFR mutant and ALK positive NSCLC

EGFR mutant

¹

ALK positive

²

Common themes

Second site mutations in target (e.g. T790M / L1196M) Use of alternative signalling pathways (e.g. MET / EGFR)

1. Camidge DR,et al. Nat Rev Clin Oncol 2014;11:473–81;

2. Shaw AT & Engelman JA. J Clin Oncol 2013;31:1105–11.

(38)

Approximately 30% of crizotinib-resistant tumours harbour resistance mutations

Doebele RC, et al. Clin Cancer Res 2012;18:1472–1482.

ALK TKI resistance mechanisms

mutation ALK

ALK copy number gain Alternate

oncogene

Unknown

(39)

Gainor et al. Cancer Discov 2016

Acquired Resistance to 2nd Generation TKIs

(40)

Second Generation

ALK Inhibitors

(41)

Next-generation ALK inhibitors

Marsilje, et al. J Med Chem 2013

Crizotinib Ceritinib TAE684

Alectinib

Brigatinib ASP3026

X-396

New ALK inhibitors should be active against resistance mutations, have proven

CNS activity, improved systemic efficacy and have an acceptable safety profile

(42)

Crizotinib-resistance: Sequential treatment with next-generation ALK inhibitors

1. Kim DW, et al. Lancet Oncol 2016;17:452–463;

2. Ou S-HI, et al. J Clin Oncol 2016;34:661–668;

3. Shaw AT, et al. Lancet Oncol 2016;17:234–242;

4. Gettinger SN, et al. WCLC 2015 (Abstr. 2125);

5. Bauer TM, et al. WCLC 2015 (Abstr. 295).

n Countries ORR mPFS (mos)

Ceritinib ASCEND-1

¹

163 Global 56% 6.9

Alectinib NP28673

²

138 Global 50% 8.9

NP28761

³

87 US/

Canada 48% 8.1

Brigatinib Phase 1/2

⁴

70 US/Spain 71% 13.4

Lorlatinib Phase 1

⁵

43 Global 47% NR

(43)

Safety profile of second-generation ALK inhibitors

Perol et al., ESMO 2016.

0 10 20 30 40 50 60 70 80 90

Crizotinib

Grade 1–2 Crizotinib

Grade 3–4 Ceritinib

Grade 1–2 Ceritinib

Grade 3–4 Alectinib

Grade 1–2 Alectinib

Grade 3–4 Brigatinib

Grade 1–2 Brigatinib Grade 3–4

Fatigue Diarrhoea Nausea

Transaminases

Crizotinib PROFILE 1014

Ceritinib ASCEND 2

Alectinib NP28673

Brigatinib 180 mg ALTA

AE (% o f pa tien ts )

(44)

Gainor et al. Cancer Discov 2016

Acquired Resistance to 2nd Generation TKIs

(45)

…

CRIZOTINIB NEXT-GEN ALK TKI

NEXT-GENERATION ALK TKI

10-11 months 9-12 months

≈ 20 months

PFS1 PFS2

PFS

…

??

+

Should we give the best drug first?

(46)

Re-biopsy may be key for optimal sequencing strategies

• Shaw AT, et al. N Engl J Med 2016;374:54−61.

Sequencing strategies should be flexible; in some cases revisiting previous agents may be the best approach

Crizotinib Chemotherapy Lorlatinib Crizotinib

Ceritinib AUY922 Crizotinib

Biopsy Biopsy Biopsy Biopsy

Months: 6 12 18 24 30 36 42 48

Effect of therapy

Before lorlatinib Response to lorlatinib Resistance to lorlatinib Response to crizotinib

C1156Y L1198F

C1156Y

(47)

Resistance to ROS1 TKIs

(48)

Drilon et al, CCR 2016

ROS1 D2033N mutation: resistance to crizotinib

can be overcome by cabozantinib

(49)

Mutation Location ROS1 fusion

Active next generation inhibitor

G2032R

¹

solvent front CD74-ROS1 cabozantinib, lorlatinib, foretinib, brigatinib (in vitro)

⁴

cabozantinib, lorlatinib (patient)

D2033N

²

solvent front CD74-ROS1 cabozantinib (in vitro, patient)

²

L2155S (cell line)3

n.r. SLC34A2-

ROS1

n.r.

L2026M

⁴

gate-keeper CD74-ROS1 cabozantinib, brigatinib, certinib, foretinib, lorlatinib

⁴

S1986Y/F

⁵

double mutation

EZR-ROS1 lorlatinib (patient)

⁵

L1951

⁶

solvent front cabozantinib (in vitro, pat.-derived cells)

⁶

Mutations in the ROS1 kinase domain conferring crizotinib resistance

1

Awad et al, NEJM 2013;

²

Drilon et al, 2015;

³

Song et al, 2015;

⁴

Chong et al,

CCR 2016;

⁵

Facinetti et al., CCR 2016,

⁶

Katayama et al, CCR 2015

(50)

ROS1 as a Therapeutic Target – Jürgen Wolf

Frequency of crizotinib-resistance mutations in ROS1 + NSCLC

10 ROS1 + patients undergoing rebiopsy after acquired resistance to crizotinib 1 patient with 2 biopsies

• 5 / 11 (45%) ROS1 resistance mutation found 5 / 8 (62.5%) non-CNS metastases

• 4 / 11 (36%) G2032R

• 1 / 11 (9%) D2033N

• 3 / 6 no resistance mutation: no addition aberration using Snapshot

Gainor et al., ASCO 2016, #9072

(51)

Bypass pathway activation as mechnism of crizotinib resistance: KIT activation

Identification of KIT p.D816G in crizotinib-resistant

ROS1 + NSCLC

Dziadziuszko et al, JTO 2016

Overcoming KIT-mediated resistance by

combining ponatinib and

crizotinib

(52)