Embed Size (px)
Citation preview
N (21) %Age (years)
Mean (SD) 64 (13)Median (Min, Max) 66 (33, 82)
SexFemale 10 48
Male 11 52
Acknowledgments & References
Longitudinal changes in cell-free DNA (cfDNA) methylation levels identify early non-responders to treatment in advanced solid tumors
Andrew A. Davis1, Wade T. Iams2, David Chan3, Michael S. Oh1, Robert W. Lentz1, Rohith Srivas4, Nicole Lambert4, Alex Robertson4, Neil Peterman4, Abhik Shah4, Timothy Wilson4, Jason Close4,Peter George4, Haleigh Wood4, Ayse Tezcan4, Ram Yalamanchili4, Ken Nesmith4, John C. Spinosa4, Haluk Tezcan4,#, Young Kwang Chae1,#
1Feinberg School of Medicine, Northwestern University, Chicago, IL, 2Vanderbilt University Medical Center, Nashville, TN3Cancer Care Associates TMPN, Redondo Beach, CA, 4Lexent Bio, Inc., San Francisco & San Diego, CA, #Questions can be addressed to either [email protected] or [email protected]
• We profiled genome-wide methylation levels in cfDNA from serial blood samples in 52 patients undergoing treatment for advanced-stage solid tumors.• Longitudinal changes in cancer-associated methylation patterns in cfDNA were significantly associated with treatment response status at first follow-up imaging and progression-free survival (PFS).• For patients classified as progressors, the cfDNA methylation assay preceded imaging and clinical evaluation by a median of 34 days.
Progression
No Progression
Cycle 1 Cycle 2
ProgressionTake action
Cycle 1 Cycle 2 Cycle 3 Cycle 4Currentapproach
Week 2 5 8 111 4 7 103 6 9 12
With cfDNA methylation
assay
No ProgressionContinue therapy
Patients treated for advanced-stage cancers face considerable burden of cumulative toxicity, out of pocket expenses and loss of opportunity on ineffective treatments. Today, imaging (CT, PET/CT, MRI), the standard for response assessment, typically requires 10-16 weeks or longer on therapy before confident conclusions can be made.
Objective
Highlights
Based on the theory that radiographic progression is preceded by changes in tumor biology that are detectable in peripheral blood, what we are calling “molecular progression”, we have developed a novel approach to quantitatively track changes in cfDNA methylation patterns to monitor response to treatment. Many cancer-specific methylation changes can be detected in cfDNA from plasma1-2, suggesting the viability of such biomarkers in diagnostic applications.
Figure 1. Potential use of cfDNA methylation assay for response monitoring.
Longer term follow-upAssay work flow
We thank all participants and their families. We also thank Dr. Brian McNamee for review of all clinical images.
• cfDNA methylation patterns are informative for treatment response to a new line of treatment in advanced-stage solid tumors.
• Our early data on long-term follow up show that cfDNA methylation changes on treatment are consistent with clinical patterns.
• Integrating methylation-based changes with information about genomic alterations will increase performance of cfDNA-based response monitoring.
Summary of clinical parameters for all patients
We prospectively collected clinical data and blood from 52 patients with advanced stage solid tumors (Table 1 & Fig. 2A,B). Blood was drawn prior to start of a new treatment, after first cycle (median 22 days), and/or second cycle (median 42 days). We also collected longitudinal blood samples from 21 unaffected volunteers without a previous diagnosis of cancer. Longitudinal blood draws were at least one week apart for the unaffected group.
Types of Therapy Lines of Therapy
Table 1. Summary of various clinical parameters for the longitudinal cohort.
1 Obtain serial blood draws 2 Extract cfDNA from plasmaand sequence
Peripheral blood was obtained over time from patients and collected in Streck Cell-Free DNA Blood Collection tubes (Step 1). Plasma was separated from whole blood, after which cfDNA was extracted from 4 mL of plasma (Step 2). Bisulfite-treated (or for longer term follow-up libraries, enzymatic methylation-treated) libraries were prepared using a custom protocol and subjected to whole-genome sequencing (median depth 18X). Finally, by quantifying the cancer-associated methylation using a metric described below, from baseline to subsequent time points, we classified patients as either progressors or non-progressors (Step 3). Treatment response was evaluated by an independent radiologist based on RECIST 1.1 guidelines.
The genome-wide methylation pattern of patients with advanced-stage cancer showed much lower levels of methylation, with localized regions of hypermethylation, relative to a set of 21 unaffected healthy participants (Fig 3A, B)1. This was also reflected in higher levels of variation. The metric is designed to quantify the degree to which a given sequencing library deviates from the expected pattern in a healthy participant based on a set of longitudinal healthy data.
3 Track longitudinal changes in methylation patterns to predict progression
T=0 T=1 T=2
...
Me
Me M
e
Me
Sequence
TimeTreatment
Dev
iatio
n fro
mN
orm
al
A B
C
Figure 2. (A,B) Summary of current line of therapy and number of previous lines of therapy. (C) Timing of sample acquisition and clinical evaluation and imaging.
References1. Xu W, Lu J, Zhao Q, Wu J, Sun J, Han B, Zhao X, Kang Y. Genome-Wide Plasma Cell-Free DNA Methylation Profiling Identifies Potential Biomarkers for Lung Cancer. Dis Markers. 2019 Feb 5.2. Guo S, Diep D, Plongthongkum N, Fung HL, Zhang K, Zhang K. Identification of methylation haplotype blocks aids in deconvolution of heterogeneous tissue samples and tumor tissue-of-origin mapping from plasma DNA. Nat Genet. 2017. Apr;49(4):635-642.3. Lambert N, Robertson, A, et al. Comparison of enzymatic- and bisulfite conversion to map the plasma cell-free methylome in cancer, to be presented at ESMO 2019.
Extract
exampleprogressor
Copies of this poster ob-tained through QR (Quick Response) and/or text key codes are for personal use only and may not be repro-duced without written per-mission of the authors.
Performance of cfDNA methylation assay
A Figure 4. (A) Waterfall plot compares cfDNA methylation-based predictions to imaging at first follow-up, quantified by the sum of longest diameter (SLD) in target lesions by RECIST (n=28). Footnoted cases showed clear clinical progression. Clinical and imaging progressors are marked by black squares.(B) - Histogram of the amount of time by which detection of molecular progression via this assay precedes clinical imaging/evaluation.(C) Progression free survival (PFS) Kaplan Meier plot of the full cohort together and separated by the results of this assay.
(1) - Clinical Progression, death before follow-up imaging (not imaged).(2) - Clinical assessment showed progression.
Sensitivity: 67%Speci�city: 91%
B
C
Genome-wide position Genome-wide position
Nor
mal
ized
met
hyla
tion
leve
l
Nor
mal
ized
met
hyla
tion
leve
l
A BFigure 3. (A) Example traces of normalized methylation levels in megabase bins genome-wide in a unaffected participant (left) and a breast cancer patient (right). (B) CDF of genome-wide methylation levels in 1MB bins averaged over all patient libraries (red) and unaffected control libraries (blue).
Figure 5. For four patients we obtained long term blood samples over the course of many cycles and multiple follow up evaluations. We show here the longitudinal changes in the cancer-associated methylation metric using enzymatic methylation (EM-Seq), a substitute for WGBS3 over the course of treatment. Baseline and follow up clinical/imaging evaluations are shown in boxed text. Collectively the data show that the assay may be of use over longer time periods to track response and identify progression prior to imaging at a subsequent follow up evaluation or provide continued reassurance that the therapy is effective.
Median Methylation Fraction
PFS
PFS
Conflict of Interest DisclosureThis study was sponsored by Lexent Bio. AAD reports travel support (Lexent Bio, Menarini Silicon Biosystems). NP, AR, AS, RS, NL, TW, PG, BW, JC, HW, AT, JCP and HT are current or previous employees of Lexent Bio. WTI reports advisory board membership (Genentech), honoraria (Outcomes Insights), and attended investigator meeting (EMD Serono). YKC has received research grants (Abbvie, BMS, Biodesix, Lexent Bio, Freenome) and honoraria/advisory board membership (Roche/Genentech, AstraZeneca, Foundation Medicine, Counsyl, Neogenomics, Guardant Health, Boehringer Ingelheim, Biodesix, Immuneoncia, Lilly Oncology, Merck, Takeda). DC, RWL and MSO have no conflicts of interest.
N (52) %Age (years)
Mean (SD) 69 (8.6)Median (Min, Max) 70 (43, 83)
SexFemale 30 58
Male 22 42
Cancer SiteLung 25 48
Breast 11 21Other 16 31
Response status at first follow-upProgression 9 17
No Progression 43 83
Treatment Total Patients
Frac
tion
of p
atie
nts
capecitabinepembrolizumab
pembrolizumabpemetrexed & carboplatin
Affected Patients
Unaffected Participants
Conclusions
Can
cer-
asso
ciat
ed m
ethy
latio
n m
etric
Can
cer-
asso
ciat
ed m
ethy
latio
n m
etric
Can
cer-
asso
ciat
ed m
ethy
latio
n m
etric
Can
cer-
asso
ciat
ed m
ethy
latio
n m
etric
