plasma-derived circulating tumor dna (ctdna) as a ...€¦ · samplecollection §...
TRANSCRIPT
Sample Collection
§ BM aspirates were collected at screening, at the end of cycles 1, 2 and then every other cycle§ Blood samples were collected on Days 1 (pre-dose), 5, 8, 15, 22 of each cycle§ Genomic DNA was extracted from peripheral blood and bone marrow mononuclear cells
(PBMCs & BMMCs, respectively) and ctDNA was isolated from plasmaMolecular Profiling
§ A targeted NGS assay covering 75 genes was used to identify driver mutations§ A ddPCR assay targeting an identified driver mutation was used to enumerate mutant allele
frequencies in Plasma (ctDNA), PBMCs, and BMMCsPatient Characteristics
§ 40 patients enrolled in the Phase 1b were analyzed by targeted NGS, median number of geneswith variants identified was 3 [range 1 - 7]
§ Among them, 20 patients were chosen for subsequent ctDNA analysis§ Clinical response was evaluated by patient’s pathologist using bone marrow aspirate§ Responders were defined as patients achieving complete response (BM blasts <5%) with or
without count recovery (CR+CRi)
Plasma-derived Circulating Tumor DNA (ctDNA) As a Surrogate Biomarker for Treatment Response with the Polo-like Kinase 1 (PLK1) Inhibitor,
Onvansertib, in Combination with LDAC or Decitabine in Acute Myeloid Leukemia (AML)
Errin Samuëlsz1; Latifa Hassaine1; Marion Luebbermann1; Brittany Ross1; Peter Croucher, PhD1, Mark Erlander, Phd1; Maya Ridinger, PhD1
1Trovagene, Inc., San Diego, CA;
Contact: Errin Samuëlsz - (Trovagene, Inc.)[email protected]
Polo-like Kinase 1 (PLK1):
§ Serine/threonine kinase, master regulator of cell-cycle progression § Inhibition of PLK1 causes mitotic arrest in prometaphase and subsequent cell death§ Overexpressed in solid tumors and hematological malignancies, including AML
Onvansertib (also known as PCM-075 and NMS-1286937) :§ Orally-bioavailable, highly-selective PLK1 inhibitor with a ~24-hour half-life§ Potent anti-tumor activity in AML preclinical models1,2
§ Ongoing clinical trials for relapsed or refractory AML, metastatic castration-resistant prostate cancer and metastatic colorectal cancer
Genomics of Acute Myeloid Leukemia (AML)
§ In AML, blast cells carrying a driver mutation in the bone marrow (BM) migrate into circulation in the peripheral blood (PB)
§ Hematologic mutations are detectable in a majority of AML patients, and are reported to persist in over 50% of patients during complete remission (CR)3
Circulating Tumor DNA (ctDNA)
§ Recently ctDNA has been used as a biomarker for monitoring tumor heterogeneity, treatment response, minimal residual disease and disease progression4
Background
REFERENCES:1 Casolaro al., PLoS One, 2013;8(3):e584242 Valsasina et al., Mol Cancer Ther. 2012 Apr;11(4):1006-163 Jongen-Lavrencic et al., N Engl J Med, 2018;378:1189-994 Yeh et al., Blood, 2017;129(12):1685-90
Phase 1b/2 Trial (NCT03303339 ) of Onvansertib in Combination with either Low-Dose Cytarabine or
Decitabine in Patients with Relapsed/Refractory Acute Myeloid Leukemia
Study design
§ Dosing schedule:Onvansertib x 5 days + either Low-dose Cytarabine (LDAC – 20 mg/m2 SC qd x 10d) or Decitabine (20 mg/m2 IV qd x 5d) in a 21 to 28 – day cycle
§ Dose escalation in Phase 1b (3+3 design) with expansion cohort at the maximum tolerated dose (MTD) or recommended phase 2 dose (RP2D) for Phase 2
Study Primary Objectives
§ Phase 1b: Assess safety, define dose-limiting toxicities (DLTs) and MTD/RP2D§ Phase 2: Assess safety, tolerability and preliminary anti-leukemic activity at the MTD (or RP2D)
Key Eligibility Criteria
§ Patients with relapsed/refractory AML who have received ≤3 prior treatment regimens § Treatment-related AML or APL are excluded§ ECOG performance status ≤2
Phase 1b/2 Trial Design
§ Mutant allele frequency in plasma-derived ctDNA was highly correlated with MAF in gDNA from BM cells (R2 = 0.8366)
§ Patients with clinical response (CR/CRi) showed bigger changes in ctDNA MAF than non-responders, suggesting that ctDNA changes can be used as a surrogate fortreatment response
§ Change in ctDNA MAF after the 1st cycle of treatment was highly predictive of response to onvansertib:• All patients achieving CR/CRi (n=6) had a decrease in ctDNA MAF after 1 cycle (Log2(C1/C0) < -0.05)• 86% of the non-responders (12 out of 14) showed no decrease in ctDNA MAF after 1 cycle
Conclusions
Plasma-derived ctDNA MAF Highly Correlates with MAF from Bone Marrow Mononuclear Cells
§ Mutation allelic frequencies (MAF) in ctDNA from plasma, and gDNA from BMMCs and PMBCS were assessed for 20 patients acrosstimepoints (Plasma, BM, PB).
§ Linear regression, and a paired two sample t-test were performed for all timepoints with matched samples.§ When plotted, MAF from plasma and BM showed the highest linearity (R2 = 0.8366) and no significant difference from BM (P=0.2205)
when comparing 49 matched samples.§ These data suggest plasma is most representative of disease state in BM, and may be use to monitor response to treatment
SCREENING
DOSING (DAYS/CYCLE)
Onvansertib
Low-Dose Cytarabine
FOLLOW-UP (DAYS/CYCLE)
Decitabine
1 2 3 4 5 6 7 8 9 10 11-21/28
DOSING (DAYS/CYCLE)
Onvansertib
FOLLOW-UP (DAYS/CYCLE)
1 2 3 4 5 6 7 8 9 10 11-21/28
Bone Marrow
Assessment
Cytogenetic Risk N (%)
Favorable 2 10%Intermediate 6 30%Adverse 9 45%Unknown 3 15%
R² = 0.83660.00
20.00
40.00
60.00
80.00
100.00
0.0 20.0 40.0 60.0 80.0 100.0
Pla
sm
a (
ctD
NA
) M
AF
BMMC MAF
Plasma vs Bone Marrow
R² = 0.71170.0
20.0
40.0
60.0
80.0
100.0
0.0 20.0 40.0 60.0 80.0 100.0
PB
MC
MA
F
BMMC MAF
Peripheral Blood vs Bone Marrow
R² = 0.69150.00
20.00
40.00
60.00
80.00
100.00
0.0 20.0 40.0 60.0 80.0 100.0
Pla
sm
a (
ctD
NA
) M
AF
PBMC MAF
Plasma vs Peripheral Blood
N = 49P = 0.2205
N = 42P = 0.0032
N = 166P = 2.962e-14
Clinical Response
Responder NRPlasma
Prediction
Responder 6 2 PPV 75%NR 0 12 NPV 100%
100% 86%Sensitivity Specificity
90% Diagnostic accuracy7.0 Positive likelihood ratio0.0 Negative likelihood ratio
PatientctDNA MAF
Log2(C1/C0)
BM % Blasts
Log2(C1/C0)Clinical Response
01-021 0.2193 -0.8625 NR01-024 0.0489 -0.3219 NR03-037 -0.0853 0.8301 Responder
05-016 0.5850 3.0704 NR05-030 -0.4258 -2.0000 Responder
05-043 -0.0299 -0.4150 NR07-009 -0.3303 -0.0468 Responder
07-013 0.3182 N/A NR07-018 0.0015 N/A NR07-033 -0.0200 N/A NR07-035 -0.1410 -1.7370 Responder
07-036 0.2940 -1.0000 NR08-027 -0.0096 -0.1043 NR08-058 -1.4399 0.9329 Responder
09-026 0.3531 0.3785 NR09-032 0.2534 0.1454 NR09-064 -1.6881 -3.9069 Responder
11-019 0.0000 0.0000 NR11-040 -0.3411 0.4021 NR12-041 -0.1515 0.6951 NR
N = 20 N = 17 NR = Non-responder
Screening test using ctDNA MAF after 1 cycle
53% Diagnostic accuracy1.10 Positive likelihood ratio0.9 Negative likelihood ratio
Screening test using % BM blasts after 1 cycle
Plasma-derived ctDNA Enables Monitoring of Treatment Response and Disease Progression
§ Twenty (20) patients were chosen for ctDNA monitoring.§ For each, plots were generated overlaying ctDNA MAF and %BM blasts by date of sample collection.§ Of 6 patients who reached CR/CRi ( ), all showed the lowest ctDNA MAF at or before (13-35 days)
determination of clinical response.§ Similarly, 4 patients with Progressive disease (PD, ), displayed a maximum ctDNA MAF at or before
determination of progression.§ Patient 05-030 case:
§ Minimum ctDNA MAF was reached 34 days prior to CR diagnosis§ A spike in ctDNA MAF was detected 53 days before progression, with a maximum MAF at the time of
progression.§ Patients with stable disease also show ctDNA trends resembling the bone marrow§ Patients who responded to treatment showed a ctDNA MAF trend that decreased, while those who
progressed showed an increasing trend.§ ctDNA MAF is capable of monitoring both clinical response and disease progression, and often prior to the
bone marrow biopsy
Clinical Response
Responder NRBM
Prediction
Responder 3 5 PPV 38%NR 3 6 NPV 67%
50% 55%Sensitivity Specificity
0.00
10.00
20.00
30.00
40.00
50.00
60.00
70.00
80.00
90.00
100.00
04/0
4/18
05/0
4/18
06/0
4/18
07/0
4/18
08/0
4/18
09/0
4/18
10/0
4/18
11/0
4/18
12/0
4/18
01/0
4/19
02/0
4/19
03/0
4/19
04/0
4/19
% B
last
s (IH
C) o
r %M
AF (P
lasm
a)
Date of Collection
Patient 07-009
0
5
10
15
20
25
30
35
40
45
50
55
60
10/0
4/18
11/0
4/18
12/0
4/18
01/0
4/19
02/0
4/19
03/0
4/19
04/0
4/19
% B
last
s (IH
C) o
r %M
AF (P
lasm
a)
Date of Collection
Patient 05-030
PatientctDNA MAF Min
(Days before CR/CRi)
07-009 3505-030* 3403-037 2109-064 1307-035 008-058 0
Complete Response (CR) Progressive Disease (PD) Stable Disease (SD)
0
10
20
30
40
50
60
70
7/1/
2018
8/1/
2018
9/1/
2018
10/1
/201
8
11/1
/201
8
12/1
/201
8
1/1/
2019
% B
last
s (IH
C) o
r %M
AF (P
lasm
a)
Date of Collection
Patient 11-019
0
10
20
30
40
50
60
70
80
90
100
6/4/
2018
6/19
/201
8
7/4/
2018
7/19
/201
8
% B
last
s (IH
C) o
r %M
AF (P
lasm
a)
Date of Collection
Patient 05-016
0
10
20
30
40
50
60
8/23
/201
8
8/30
/201
8
9/6/
2018
9/13
/201
8
9/20
/201
8
9/27
/201
8
10/4
/201
8
10/1
1/20
18
% B
last
s (IH
C) o
r %M
AF (P
lasm
a)
Date of Collection
Patient 01-024
0
10
20
30
40
50
60
70
80
90
100
8/26
/201
8
9/10
/201
8
9/25
/201
8
10/1
0/20
18
10/2
5/20
18
% B
last
s (IH
C) o
r %M
AF (P
lasm
a)
Date of Collection
Patient 09-026
PatientctDNA MAF Max
(Days before PD)
11-019 2405-043 14
05-030* 005-016 0
Responders Have a Greater Range of ctDNA MAF than Non-responders
§ As plasma was most correlative with BM, we aimed to explore if changes in plasma MAFdiffered between patients with clinical response (CR/CRi) and non-responders.
§ The change in MAF over all timepoints available for each patients was log transformed:Log2(Max/Min), where Max is highest MAF of all timepoints, and Min is the lowest MAF ofall timepoints.
§ A t-test was performed between responders (n=6) and non-responders (n=14) to determineif ctDNA range distinguished the groups
§ There was a significant difference between the two groups (P=0.002), further supportingctDNA as a biomarker for treatment response. Responders = 6
Non-Responders = 14P = 0.002
Methods and Patients Characteristics
N (%)
Responders 6 30%
Non-Responders 14 70%
ctDNA as a Predictive Biomarker
Plasma-derived ctDNA is a Predictive Biomarker for Treatment Response to Onvansertib
§ We aimed at determining the utility of plasma MAF to predict clinical response after the 1st cycle of treatment§ MAF were assessed in ctDNA before beginning treatment (C0) and after 1 cycle of treatment (C1) when BM aspirates were collected;
and change in MAF calculated as Log2(C1/C0).§ Similarly changes in % BM blasts at end of cycle 1 versus screening were calculated as Log2(C1/C0).§ All patients with clinical response (CR+CRi, n=6) showed a decrease in ctDNA MAF at cycle 1 (Log2(C1/C0) < -0.05), while only 2 of the
14 non-responders (NR) showed a similar decrease.§ Measuring the changes in plasma over the first cycle predicted patient response with 90% diagnostic accuracy, 100% sensitivity and 86%
specificity, with positive predictive value (75%) and negative predictive value (100%) supporting the utility of this analysis§ Conversely, the same analysis in BM Blasts at Cycle 1 were not predictive of clinical response
Earlier Treatment Decisions from Serial Monitoring of ctDNA
% BM blasts cellsddPCR ctDNA % MAFTreatment Cycle
0
10
20
30
40
50
60
70
80
90
100
2/27
/201
9
3/14
/201
9
3/29
/201
9
4/13
/201
9
4/28
/201
9
5/13
/201
9
5/28
/201
9
6/12
/201
9
% B
last
s (IH
C) o
r %M
AF (P
lasm
a)
Date of Collection
Patient 05-043
0
2.5
5
7.5
10
12.5
15
17.5
20
8/26
/201
9
9/26
/201
9
10/2
6/20
19
11/2
6/20
19
% B
last
s (IH
C) o
r %M
AF (P
lasm
a)
Date of Collection
Patient 09-064
0
10
20
30
40
50
60
70
80
90
100
12/1
3/20
18
12/2
8/20
18
1/12
/201
9
1/27
/201
9
2/11
/201
9
2/26
/201
9
3/13
/201
9
% B
last
s (IH
C) o
r %M
AF (P
lasm
a)
Date of Collection
Patient 07-036
• !"#$%&%'%&( = *+(*+-./)
• !1"2%3%2%&( = */(*/-.+)
• 45$%&%'" 46"7%2&%'" 89:;"(448) =*+
(*+-.+)
• <"=9&%'" 46"7%2&%'" 89:;"(<48) = */(*+-*/)
• >%9=#5$&%2 ?22;692( (>?) = *+-*/*+-*/-.+-./
• 45$%&%'" @%A":%ℎ557 C9&%5 (4@C) = DEFGHIHJHIK(LMDNEOHPHOHIK)
• <"=9&%'" @%A":%ℎ557 C9&%5# (<@C) = (LMDEFGHIHJHIK)DNEOHPHOHIK
§ Serial monitoring in plasma provides a less-invasive alternative to bone marrow, often enabling earlier detection of clinical response orprogression:
• Minimum ctDNA MAF was measured at or before remission, while maximum ctDNA MAF was measured at or before progression
§ Patients enrolled in the Phase 2 section will be analyzed to further validate the utility of ctDNA as a biomarker for onvansertib-response and asurrogate for BM biopsy