Radiation Sensitivity and Resistance Theodore S. Hong, MD Director, - - PowerPoint PPT Presentation

Radiation Sensitivity and Resistance

Theodore S. Hong, MD

Director, Gastrointestinal Radiation Oncology Massachusetts General Hospital Co-Leader, Gastrointestinal Malignancies Progam Dana-Farber/Harvard Cancer Center Associate Professor of Radiation Oncology Harvard Medical School

Disclosures

• Novartis- Research Funding
• Taiho- Research Funding
• Astra-Zeneca- Research Funding
• Bristol Meyers-Squibb- Research Funding
• Clinical Genomics- Advisory Board

“Radiosensitive” or “Radioresistant”

Multiple mechanisms of cell death/toxicity after radiation

• Mitotic

catastrophe

• Stimulation of apoptosis

(or autophagy, or necrosis)

• Irreversible senescence
• Toxic oxidative

modification of biomolecules

• Redox balance

alterations; metabolic derangements Courtesy of Christie Eyler

Gold Standard Radiosensitivity Assay: Colony Formation Assay

• Benefits:

– Incorporates multi- generational cell death and clonogenicity/self- renewal – Accounts for senescence – Accounts for delayed cell death/cell arrest

• Drawbacks:

– Artificial, in vitro system – Time and effort investment Eyler CE, unpublished

Gold Standard Radiosensitivity Assay: Colony Formation Assay

• Benefits:

– Incorporates multi- generational cell death and clonogenicity/self- renewal – Incorporates senescence – Accounts for delayed cell death/cell arrest

• Drawbacks:

– Artificial, in vitro system – Time and effort investment Eyler CE, unpublished

• 4.5
• 4
• 3.5
• 3
• 2.5
• 2
• 1.5
• 1
• 0.5

0.5 2 4 6 8 10

Ln(Surviving Fraction) Dose Radiation (Gy)

Crowley, LC et al. CSHL (Time given for

• ne plate,

usually process 10+ per experiment)

Short term cell viability and proliferation rates do not reflect radiosensitivity

3 Day Viability Assay Colony Formation Assay

20 40 60 80 100 120 Naïve RT exposed Percent Surviving % of 0Gy 0 % of 0Gy 2 % of 0Gy 10 0.01 0.1 1

• 2

3 8

Surviving Fraction Radiation Dose (Gy) RT-Naive RT- Exposed

** ** ** **

Eyler CE, unpublished

Biomarkers for Radiation Oncology

9

Biological Parameter Examples of Candidate Biomarkers Association with Radioresistance or Radiosensitivity Potential Intervention(s) Current Clinical Status Number of Clonogenic Tumor Cells or CSCs Cell surface markers such as CD44 being studied Higher baseline number of clonogenic cells or CSCs correlates with radioresistance Higher radiation dose or radiosensitizer for high CSC number Tumor volume is a surrogate for CSC number, (should) impact RT dosing in clinical practice Accelerated Tumor Cell Repopulation EGFR expression being studied Accelerated repopulation of clonogenic tumor cells or CSCs during RT causes radioresistance Shortening of overall treatment time limits number

• f clonogenic cells that need

to be sterilized by RT HNSCC histology has been used as surrogate in clinical practice to guide accelerated fractionation schemes. Tumor Sensitivity to RT Fraction Size No candidate markers currently exist to predict α/β

• f individual tumors

Some tumors are associated with high sensitivity to RT fraction size (low α/β <10 Gy) Hypofractionation (> 2 Gy daily fraction size) Breast or prostate histology used as surrogate in clinical practice to guide hypo- fractionation schedules Tumor Hypoxia PET/MRI-based imaging markers, hypoxia gene signatures Tumor hypoxia reduces radiation damage to DNA, thereby increasing radioresistance Combination of RT with hypoxic radio-sensitizer or dose increase to hypoxic tumor parts Not yet used in clinical practice HPV Status HPV16 DNA or p16 expression HPV infection causes radiosensitivity, likely through interfering with DNA repair Treatment de-intensification De-intensified treatment to reduce toxicity in HPV+ HNSCC in clinical trials Intrinsic Radiosensitivity DSB repair gene mutations, altered expression, DNA damage foci (eg. γ-H2AX), RSI/GARD, ctDNA ,et cetera Variations in ability of tumor cells to cope with radiation damage may cause radiosensitivity or -resistance Treatment de-intensificaation

• r intensification, respectively

Not yet used in clinical practice Tumor Genotype Mutations in cancer genes such as KRAS, BRAF, EGFR, KEAP1, NRF2 Tumor mutation status may correlate with radiosensitivity

• r -resistance through

several mechanisms Treatment de-intensification

• r intensification, respectively

Not yet used in clinical practice

Adapted from Kirsch et al., JNCI in press

KRAS mutation has long been known to affect cellular radioresistance - but no impact on clinical management yet!

10 Bernhard EJ, et al. Direct Evidence for the Contribution of Activated N-ras and K-ras Oncogenes to Increased Intrinsic Radiation Resistance in Human Tumor Cell Lines. Cancer Research 2000;60:6597-6600.

Courtesy of Henning Willers

Model: Non-canonical pathway of EGFR-PKCα mediated radioresistance in KRASmut cells

Emerging preclinical data on KRAS mutation- dependent pathway of radioresistance

11

C lo n o g e n ic S u rv iv a l F ra c tio n a fte r IR S u rv iv a l fra c tio n a fte r IR

T u m o r C o n t r o l P r o b a b i l i t y ( T C P ) ( % )

Aurora B PKCα EGFR Radioresistance

A B C

Wang et al., Cancer Res 2014, 2017; Gurtner et al., unpublished

KRAS mutation associates with radioresistance in clinical cohorts

12

Reference Institution

• No. of

genotyped pts (n) Cancer type Radiation Dose (Gy) End- point KRAS mut vs wt (%) Garcia- Aguilar et al., Ann Surg 2011 Multi- institutional, prospective 132 Rectal cancer Preop RT+5FU 50.4 pCR 14 vs 33 Mak et al., CCR 2015 DFCI, retro- spective 9 NSCLC SBRT median 54 1-yr LC 57 vs 74 Cassidy et al., Cancer 2017 Emory, retrospective 45 NSCLC SBRT median 50 2-yr LC 44 vs 74 Hong et al., JNCI 2017 MGH, prospective phase II 57 Liver mets SBRT protons median 50 1-yr LC 43 vs 72

Does KRAS status clinically predict response to chemoradiation?

• 132 patients

– Stage II/III rectal cancer – Treated with standard chemoradiation – Evaluated multiple genes – Used Sanger sequencing – Evaluated 23 genes

Garcia-Aguilar J, et al. Ann Surg 2011;254:486-493

Predictors of non-pCR

Garcia-Aguilar J, et al. Ann Surg 2014

MGH- Genotypic Predictors of pCR

• 47 patients
• cT3-4 or N+
• Chemoradiation 50.4 Gy with 5FU
• Genotyping across 15 genes evaluating

140 hotspots

Russo AL, J Gastroint Canc, 2013

Genotype distribution

• KRAS- 43%
• APC- 17%
• BRAF- 4%
• NRAS- 4%
• PIK3CA- 4%
• TP53-4%

pCR

• pCR in WT patients- 23.5%
• 1 patient with a mutation had a pCR- 3.3%

KRAS Status as Predictor Biomarker after Lung SBRT

Mak et al. Clin Lung Cancer, 2015.

Liver Proton SBRT for Hepatic Metastases

• Hepatic metastases
• 50 Gy in 5 fractions
• 91 patients enrolled, 89 analyzed

– 2 did not begin treatment

• Median age - 67 (34-88)
• Male – 56 (62.9%)

Hong TS, et al. JNCI 2017.

Tumor Types

Tumor Type N (%) Colorectal 36 (40.4%) Pancreatic 16 (18.0%) Esophagogastric 12 (13.5%) HCC 9 (10.1%) Lung 4 (4.5%) Gallbladder 3 (3.4%) Breast 3 (3.4%) Small bowel/duodenal 3 (3.4%) H&N 2 (2.2%) Anal cancer 1 (1.1%)

Colorectal Metastases

LC by Mutational Status

Pre-

• perative

CRT Surgical resection

Whole Exome Sequencing Evaluation of Rectal Samples

All FFPE samples (n=34) from 17 patients in our initial cohort were successfully sequenced and passed quality control metrics (out of 175 evaluated)

R

n=8

NR

n=9

Pre-CRT biopsy Post-CRT resection specimen Germline

• 5FU and

RT to 50.4 Gy

• After 8-11

weeks Sequencing

Analysis

• Whole exome

sequencing

• RNA sequencing

Kamran S, ASTRO 2017

TP53/KRAS mutations in pre-CRT samples

Sample TP53 mutation KRAS mutation RC001 p.M237I

• RC003

p.R175H p.G12D RC007

• p.G12S

RC010

• RC012
• p.G13D

RC013 p.R175H

• RC014

p.V274A

• RC017

p.G245S

• Sample

TP53 mutation KRAS mutation RC002

• p.G12V

RC004 p.R248Q p.A146V RC005 p.T125T p.G13D RC006 p.Y107* p.G12V RC008 p.E287* p.L19F RC009

• p.G12V

RC011 p.158H p.G12V RC015 p.R213*

• RC016

p.S241fs

• R

NR

Distribution of KRAS/TP53 co-mutations: R 1/8 (12.5%) NR 5/9 (55.6%)

Co-mutation of TP53 appears to define a particularly radioresistant subset of KRAS mutant cancers

25

A B

KRASmut radioresistant tumors

Hong et al., JNCI 2017 Wang et al., Cancer Res 2017 Kamran et al., unpublished

C

EGFR, PKCα, and Chk1 are potential targets for radiosensitization of KRAS mutant cancers

26

R a d is e n s itiz a tio n F a c to r (S R F 2 G y) R a d is e n s itiz a tio n F a c to r (S R F 2 G y)

A B

AZD7762 (Chk1)

C D

Wang et al., Cancer Res 2014, Liu et al., Mol Cancer Res 2015, Kleiman et al., PLoS One 2013

Rectal cancer(KRAS mut)

Expression Based Testing: Radiation Sensitivity Index

• Built on 10 hub genes associated with SF2

from 48 cell lines obtained from the NCI

Eschrich SA, et al. IJROBP 2009

RSI

• Higher score = more radiation resistant

Correlation of RSI with response

RSI

• Suggests there may be a way to identify

radiation resistant tumors based on expression

• Validated in multiple diseases
• Does not identify a targeted strategy to

radiosensitize

Future Directions-RNA Expression Changes in DDR genes

Courtesy of Darrell Borger

Cell Cycle (n =20): CCND1 CDK6 E2F1 RB1 CDK1 CDKN1A EGFR SMAD3 CDK12 CDKN2A ERBB2 TGFB1 CDK2 CHEK1 FOXM1 TP53BP1 CDK4 CHEK2 MYC WEE1

BAP1 RFWD3 BRCC3 RNF8 FANCL RNF20 HERC2 RNF40 HUWE1 RNF138 KEAP1 RNF168 NEDD8 UBR5 PIAS1 UHRF1 PIAS4 UIMC1 RAD18 USP1 Ubiquination/Neddylation (n=20):

Apoptosis (n =4): BCL2 FAS MYD88 PTEN ATF2 RAD17 FANCG MGMT PLK1 ATM RAD50 FANCI MLH1 PMS2 ATMIN RAD51 FANCM MRE11A PNKP ATR RAD51AP1 ERCC1 MSH2 POLB ATRIP RAD51B ERCC2 MSH6 POLD1 ATRX RAD51C ERCC3 POLE PRKDC BARD1 RAD51D EXO1 POLL RBBP8 BLM RAD52 DCLRE1C POLM RECQL4 BRCA1 RAD54B DDB2 POLQ RECQL5 BRCA2 LIG1 DNA2 MUS81 REV1 BRIP1 LIG3 EME1 NBN RIF1 C9orf142 LIG4 FEN1 NEIL2 RPA2 XRCC1 FAN1 GEN1 NEK1 SFPQ XRCC2 FANCA HELQ NFE2L2 SLX4 XRCC3 FANCB HUS1 NONO SMC5 XRCC4 FANCC INO80 PALB2 SMC6 XRCC5 FANCD2 MCPH1 PARG SPP1 XRCC6 FANCE MDC1 PARP1 TOPBP1 RAD1 FANCF MED1 PARP2 WRN DNA Damage (n =95):

Epigenetic (n =6): ARID1A BMI1 CHD4 CHD6 ING1 KAT5

Autophagy (n =4): ATG7 BNIP3 HMGB1 PINK1

Housekeeping (n =5): RPS3 RPLP0 TBP RPS20 HMBS

Panel of 154 genes associated with DNA damage response

MGH DDR Panel

Luminex RNA expression profiling was utilized where lysates were obtained directly from FFPE tissue and directly hybridized to target Luminex beads followed by signal amplification and reading on a Luminex instrument. This has the advantage of avoiding nucleic acid isolation, cDNA synthesis and PCR amplification to decrease technical bias.

MGH Prospective Studies Resectable

Stage Intervention R0 Resection Rate mDFS (mo) mOS (mo) DFS-2 OS-2 06248- RESECTABLE Short course RT/Adjuvant Gem 62% All Patients 10.4 17.3 20% 40% Resected Patients 14.5 27 25% 53% 11073- RESECTABLE Short Course RT/Adjuvant Gem+ HCQ 72% All Patients 11.7 23.3 32% 43% 11328- BORDERLINE RESECTABLE FOLFIRINOX x 8 Individualized CRT 56% All Patients 14.7 37.7 43% 56% Resected Patients 48.6 Not Reache d 55% 72% 13051- LOCALLY ADVANCED FOLFIRINOX x 8+losartan Individualized CRT 50% All Patients 21 33 32% 67% Resected Patients 28 33 52% 89%

Phase II Study of HCQ and short course protons for resectable PDAC

• PDAC is highly dependent on autophagy
• HCQ is a potent inhibitor of autophagy
• Patients treated with same regimen as

prior study, but with concurrent and maintenance HCQ

Hong TS, et al, ASCO 2017

Kimmelman, Clin Cancer Res, 2015

KRAS driven tumors rely on metabolic scavenging

Kimmelman G&D 2011

CQ HCQ

Autophagy and PDAC

Yang et al, Genes and Development 2011

.0 .1 .2 .3 .4 .5 .6 .7 .8 .9 1 .0 4 0 5 0 6 0 7 0 8 0 9 1 1 1 3 1 5 1 7 1 9

p=0.0012

Surviving

Days

Days Tumor volume (mm3 ) 49 56 63 70 77 84 20 40 60 80 100 120

Ctrl shATG5-2 shATG5-1

* ^ * ^

p=0.02 p=0.06

RNAi to ATG5 suppresses tumor growth Yang et al, Cancer Discovery, 2014 Yang et al, G&D, 2011

Autophagy inhibition has efficacy in PDAC xenografts, PDXs, and GEM Models HCQ treatment of PDAX PDX

autophagy Growth inhibition

Increased ROS Metabolic dysfunction

Cellular consequences of autophagy inhibition

• 1. Miller-Ocuin JL, Bahary NS, Singhi AD, et al: Inhibition of autophagy improves pathologic and biomarker

response to preoperative gemcitabine/nab-paclitaxel in potentially resectable pancreatic cancer: A phase II randomized controlled trial. 2017 Society of Surgical Oncology Annual Cancer Symposium. Abstract 3. Presented March 17, 2017.

HCQ Boosts Antitumor Activity of Neoadjuvant Chemotherapy for Pancreatic Cancer

• Randomized phase 2 trial of preoperative gemcitabine/abraxane +/- HCQ
• Interim analysis presented at the Society of Surgical Oncology
• significantly improved histological response in resected specimens

(also more apoptosis)

• significant decrease in CA-19-9 levels
• decrease in involved lymph nodes
• more T-cell infiltration

HCQ/Preop Short Course CRT

• Pts with radiographically resectable, biopsy-proven

PDAC

• SCRT was 5 Gy x 5 with protons concurrent with Cape

825 mg/m2 BID wk 1 and 2 M-F.

• HCQ was started at 400 mg po BID 1 wk prior to

radiation through SCRT until the day of surgery. Surgery was performed 1-3 wks after completion of SCRT.

• 6 mo of gemcitabine-based chemotherapy after surgery
• Pts resumed HCQ after discharge from surgery and

continued until progression

• Sample size of 50 to evaluate an increase of 2-year PFS

from 30% to 45%

HCQ/Preop Protons results

• 50 pts enrolled
• 46 underwent resection

– No resection- Toxicity -2, intercurrent illness-1 – 38 R0, 8 R1 – 29 patients had positive nodes – 1 pCR, 2 near pCR

• Disease outcomes

– mDFS- 15.7 mo – mOS 19.8 mo – OS-2 43.1%

HCQ/Preop SCRT

OVERALL SURVIVAL PROGRESSION FREE SURVIVAL mPFS -15.7 mo mOS -19.8 mo

Preliminary Data in Pancreatic Cancer- Non-response to Protons

Patient 48 (purple bar) represents a non-responder to chemoradiation + HCQ treatment with a signature of proliferation (CDKN2A) and upregulation of cell cycle checkpoint genes (RAD1 and CCDN1).

Patient 36 (dark blue bar) represents a non-responder to treatment and is characterized by upregulation of various DNA damage repair and checkpoint genes, including FANCM, ERCC3 and

• LIG3. However, this noted upregulation

was not observed throughout the family of genes.

Preliminary Data in Pancreatic Cancer- Non-response to Protons

Expression profiles of the non-responding patients in the cohort are highly diverse across DNA damage response genes. Hierarchical clustering indicates similarities within subsets of non-responders which suggest the possibility of a multi-gene signature across multiple functional pathways.

Conclusions

• Radiation sensitivity varies across and

within cancers

• KRAS remains one of the main culprits of

radiation resistance

• Expression based assays may help

identify resistant tumors and potential future targets.

Acknowlegements

• Christie Eyler, MD, PhD
• Henning Willers, MD
• Sophia Kamran, MD