The Evolving Landscape of Post-Therapy Brain Tumor Imaging - PowerPoint Presentation

Slide1

The Evolving Landscape of Post-Therapy Brain Tumor Imaging

Jeffrey Ware, MD, Ronald Wolf, MD, PhD, Harish Poptani, PhD, Donald O’Rourke, MD, Suyash Mohan, MDNeuroradiology DivisionDepartment of Radiology University of Pennsylvania

ASNR 2015 Annual Meeting Chicago, IL April 25 – April 30, 2015

Tumor or treatment? Stay tuned…

Educational Exhibit: eEdE-62

Presentation Number: 673

 Slide2

Disclosure Statement

Neither the authors nor their immediate family members have a financial relationship with a commercial organization that may have a direct or indirect interest in the contentSlide3

Background

Improvements to the standard treatment of brain gliomas are resulting in prolonged survival and better overall outcomesNew treatments are also bringing about new and complex imaging manifestationsThis exhibit reviews key concepts in neuro-oncologic imaging essential to accurate radiological assessment in the perioperative and postoperative setting

Preoperative planningPreoperative mappingRisk-benefit assessmentPost-therapy imagingPerioperative infarctPseudoprogressionTrue progression

Radiation necrosis

PseudoresponseTreatment

response criteriaEmerging therapies

Laser-induced thermal therapy

Vaccine

and

immunotherapies

Tumor-treating

fields

OutlineSlide4

Extent of Resection

Maps of language cortex (left) and white matter tracts (right) overlaid on anatomic images depict the anatomic relationship of tumor to nearby eloquent structures

1

Sanai

 N, et al (2011) An extent of resection threshold for newly diagnosed glioblastomas

. J Neurosurg 

115:3–82

McGirt

 MJ, et al.

(2009) Association of surgically acquired motor and language deficits on overall survival after resection of

glioblastoma

ultiforme

. 

Neurosurgery 65:463–469

Preoperative mapping of eloquent cortex and white matter tracts with

fMRI

and

diffusion

tractography

, respectively, allows more informed surgical planning

Mortality

benefits

of cytoreduction

1

can be more accurately weighed against

morbidity of

postoperative neurologic deficits

2Slide5

Cytoreduction

vs. Postoperative DeficitsPreoperative MRI in a 57 year-old man with GBM. Diffusion tractography (DTI) of the corticospinal tract (CST) superimposed on anatomic imagesFLAIR & DTIT1+C & DTIT1+C &

DTIClose

proximity of tumor to the right CSTSlide6

Early follow-up MRI shows

thick, nodular enhancement about the posterior and medial margin of the resection cavity, consistent with residual neoplasm which was unable to be resected due to proximity to the CST. The patient did not experience new neurologic deficits following surgery.T1+CFLAIRT1

Follow-up MRIEarly

recurrence/True Progression (TP)

Subsequent follow-up MRI

FLAIR

T1+CSlide7

Cytoreduction

vs. Postoperative DeficitsPreoperative MRI in a 44 year-old man with GBM. DTI of the CST and fMRI of hand motor areas superimposed on anatomic images show proximity of tumor to eloquent structuresT1+CFLAIR & fMRI

FLAIR & DTIPostoperative MRI shows complete resection of enhancing tumor, however the patient developed new right-sided weakness and aphasia

T1+C

Postoperative

tractography

was performed, demonstrating the left CST

remained intact

The patient experienced near-complete functional recovery, as the symptoms were due to a supplementary motor area syndromeSlide8

Preoperative Mapping: Benefits

Eloquent cortical mapping (fMRI)1Tractography for operative planning & intraoperative navigation (DTI)2,3

1Petrella JR., et al., “Preoperative fMRI Localization of Language and Motor Areas: Effect on Therapeutic Decision Making...” Radiology (2006)2Wu JS, et al. “Clinical Evaluation and Follow‐Up Outcome of DTI‐Based Functional Neuronavigation,” Neurosurgery (2007)

Smaller craniotomy

Reduced duration of surgery

Selection of more aggressive surgical strategy

Greater extent of resection

Reduced duration of surgery

Decreased seizure incidence

Improved postoperative motor performance

Higher median survival in high grade gliomasSlide9

Post-Therapy Imaging

Early postoperative MRI – within 24H (maximum 72H) of surgeryAssess extent of resection & residual tumorIdentify perioperative ischemia – include DWIIndications for follow-up imagingAssess prognosis of postoperative neurologic deficitsBefore chemotherapy and/or radiationChange in clinical statusRoutine surveillance

Techniques

Conventional MRI

Diffusion MRIMR

perfusion and permeabilityMR spectroscopy

Positron emission tomography (PET)

Post-therapy imaging overview at the University of PennsylvaniaSlide10

Preoperative images show enhancing & non-enhancing components of a recurrent glioma, which underwent re-resection

PreoperativePostoperativeFollow upT1+CT1+CT1+C

T1DWIDWIFLAIR

Perioperative Ischemia

Recognition

of perioperative infarct is important to avoid misclassification of

enhancement in an evolving infarct

as recurrent

neoplasm on follow-up imaging

Immediate postoperative images

reveal new

restricted diffusion adjacent to the resection cavity, indicating

a perioperative

infarct

On follow-up,

gyriform

enhancement about the resection cavity represents an evolving

subacute

infarct rather than recurrent

tumorSlide11

Temozolomide

(TMZ) and MGMTTMZ induces tumor cell death by causing DNA damageMGMT (O-6-methylguanine-DNA methyltransferase) is involved in DNA repairTumor cells expressing a silencing MGMT promoter methylation exhibit inhibited DNA repair and are more sensitive to TMZMGMT promoter methylation is associated with improved survival with TMZ treatment

Pseudoprogression (PsP) – refers to clinical & radiologic manifestations of an

exaggerated but favorable treatment response

On MRI, manifests as progressive edema and enhancement,

thereby mimicking true tumor progressionOccurs

in the early post-treatment period (<6 months)

Mechanisms

Pseudoprogression

(

PsP

)Slide12

Pseudoprogression

Follow-up MRI in a patient status post recent chemoradiation for glioblastoma shows new focal enhancement adjacent to the resection cavityT1+CFLAIRPre-radiationFollow up

Distinguishing PsP from true progression is critical to avoid discontinuing effective therapy, as

PsP

represents a favorable treatment response

?

Does this represent true tumor progression or

pseudoprogression

?Slide13

PsP

vs. True ProgressionConventional MRIClinical evaluation and conventional MRI have significant limitations in distinguishing treatment effects from early tumor progression (EP)

Most conventional MRI features are not specific to either PsP or EP11Young RJ, et al. Potential utility of conventional MRI signs in diagnosing pseudoprogression in glioblastoma.

Neurology

. 2011; 31;76(22):1918-24.

T1+C images obtained immediately following tumor resection (left), and at 4 and 8 months following resection (middle and right, respectively) show

progressive nodular enhancement about the resection cavity with

subependymal

extension

Progressive

subependymal

enhancement

, however, is

relatively specific

for

true tumor

progression

The

findings proved

to

represent

tumor progression

at

re-resection

.Slide14

PsP

vs. True ProgressionMR PerfusionTPPsP

TUMOR

High blood volume (

rCBV

)

due to high vascular density

High vascular permeability (

K

trans

)

due to immature

neovasculature

PsP

Lower blood volume (

rCBV

)

due to relative absence of

neoangiogenesis

MR perfusion measurements of blood volume differ significantly between

PsP

and TP (right). There is some overlap, however, due to pitfalls and the fact that treatment effects may coexist with residual/recurrent tumor.

Pitfalls

Low

rCBV

in tumor may result from necrosis &

edema resulting in vessel compression and subsequent

hypoperfusion

High

rCBV

in

PsP

may result from aneurysm, telangiectasia, vascular elongation, or endothelial

proliferationSlide15

PsP

vs. True ProgressionMR Perfusion

rCBV

K

trans

v

p

T1+C

PsP

TP

ADC

MR perfusion measurements can aid in distinguishing

pseudoprogression

from true tumor progression. Note areas of

relatively higher cerebral blood volume (3

rd

column) and permeability (5

th

column)

in the case of true progression.Slide16

PsP

vs. True ProgressionPost-opFLAIRT1+CrCBV3 month Follow up

Progressive edema and enhancement about the resection cavity is concerning for recurrent glioma

rCBV

remains low, however, suggesting

pseudoprogression

Slide17

p53

Ki-67H&E (10X): Reactive brain tissue, hyalinized vessels

H&E (20X): Reactive brain tissue, hyalinized vesselsFinal Pathology Diagnosis: Predominantly

treatment effect with small proportion (<10%) of residual

gliomaSlide18

MR Spectroscopy

Normal Brain Glioma

Glial tumors demonstrate a higher choline peak and a lower N-acetylaspartate (NAA) peak

compared to normal brain

MR

spectroscopy (MRS) allows characterization and quantification

of certain metabolites within a region of

interest

Cho

NAA

NAA

Cho

Courtesy:

Gaurav

VermaSlide19

MR Spectroscopy – EPSI

3D mapping of choline-to-creatinine ratio in a patient with a glioma

Multislice

echo-planar spectroscopic imaging sequences (EPSI) utilize very short echo times to construct a high-resolution metabolite map of the entire

brain from

a single

15 minute examination

Courtesy:

Gaurav

VermaSlide20

MR Spectroscopy – EPSI

Metabolite Maps

Cho/CrCrCho/NAA

NAA

Cho

Courtesy: Gaurav VermaSlide21

MR Spectroscopy – EPSI

T1+C

FLAIR

Segmentation

Enhancing

Immediate

Peritumoral

Distant

Peritumoral

Overlaying MRS data on anatomic segmentations allows metabolite profiles of different regions (tumor,

peritumoral

, and distant

peritumoral

) to be assessed and compared

Courtesy:

Gaurav

Verma

EPSISlide22

PsP

vs. True ProgressionMR SpectroscopyTrue Progression Pseudoprogression

Distant

Peritumoral

Immediate

Peritumoral

Enhancing

Cases of true progression show higher choline across all three regions, and particularly in the enhancing region

Courtesy:

Gaurav

VermaSlide23

PsP

vs. True ProgressionIntegration of Multiparametric ImagingPost-op4 month follow-upFLAIR

T1+CrCBVKtrans

Vp

Follow up MRI reveals

progressive enhancement adjacent to the

glioma

resection cavity

Perfusion parameters

rCBV

,

Ktrans

, and

Vp

are

elevated

. Spectroscopy

demonstrates an

elevated Cho/NAA

.

Integrated findings are highly concerning for

true progression

.

NAA

ChoSlide24

H&E (2.5x): Residual/recurrent

gliomaH&E (20x): Residual/recurrent gliomap53Ki-67Final Pathology Diagnosis:

Predominantly (75%) recurrent/residual gliomaSlide25

PsP

vs. True ProgressionDifferentiation – Summary Specific clinical & imaging features do, however, greatly influence the likelihood that imaging findings in the post-treatment setting represent either PsP or tumor progressionPredictorPseudoprogressionTrue progression

Clinical symptomsNo new symptomsNew neurologic symptomsMGMT promoter statusMethylatedUn-methylatedLesion number

Single

MultipleEnhancement pattern

Rim-enhancement, central necrosisSolid, nodular,

subependymalMR perfusion

Low,

stable, or decreasing

rCBV

High or increasing

rCBV

MR diffusionHigh

ADC (necrosis)

Low ADC (cellularity)

MR spectroscopy

Depressed

Cho, Cr, NAA peaks

High lipid/lactate

High

Cho/Cr

High Cho/NAA

PET

Low metabolic activity

High metabolic

activity

Pseudoprogression

remains a clinical diagnosis which requires

integration of clinical data and radiologic

findings; thereby necessitating a team-based approach including radiologists,

neurosurgeons, and neuro-oncologists

No single

predictor has

sufficiently high sensitivity and specificity to reliably

differentiate

PsP

from true progression on a routine

basisSlide26

Radiation Necrosis (RN)

White matter hypoperfusion and necrosis resulting from radiation-induced vasculopathyIn contradistinction, PsP represents treatment effect on tumor and blood-brain barrier and should replace the outdated term early radionecrosis

Progressive edema and enhancement, thereby mimicking true tumor progression

Occurs months to years following

therapy

Classic “soap bubble” or “swiss cheese” MRI appearance, however

no imaging features are widely accepted as specific for RN, which remains a pathologic

diagnosis

Mechanisms

Imaging FeaturesSlide27

Radiation Necrosis (RN)

FLAIRT1+CrCBV

NAACrChoLipid/lactate

Follow up MRI (bottom row) reveals

progressive enhancement and edema

about the resection cavity The corresponding region, however, demonstrates

relatively low rCBV. MR

spectroscopy reveals

a

high lactate peak, low Cho/NAA, and a relative decrease in all normal metabolites

.

Integrated findings

suggest radiation necrosis rather than progressive neoplasm.

Advanced

MR techniques

hold

the potential to better characterize RN

, but

this remains an area of active

researchSlide28

Antiangiogenic Therapy

Glial tumors rely on dysregulated angiogenesis, leading to the formation of abnormal, permeable blood vesselsAntiangiogenic therapy refers to antibody-mediated inhibition of vascular endothelial growth factors (VEGF)Inhibition of VEGF results in vascular normalization (↓ vessel diameter, ↓ permeability, and basement membrane thinning)

Antiangiogenic therapy results in rapid reduction of enhancement and edema

Little survival

benefit without concomitant cytotoxic

therapyMay promote the

development of an invasive, non-enhancing phenotype

Pseudoresponse

(

PsR

) – progression of non-enhancing tumor despite ↓ enhancement

and

vessel permeability

Mechanisms

Imaging ManifestationsSlide29

Pseudoresponse

Pre-AvastinPost-Avastin 1

Post-Avastin 2T1+CFLAIR

rCBV

Pre-therapy scan shows an enhancing mass in the right temporal

lobeFollowing initiation of antiangiogenic therapy, there is

marked decrease in enhancement and adjacent edema

Subsequent MRI shows

increasing T2 signal and

rCBV

adjacent to the lesion

despite

minimal enhancement,

suspicious

for progression of the non-enhancing tumor componentSlide30

Treatment Response Criteria

1Macdonald DR, et al. Response criteria for phase II studies of supratentorial malignant glioma. Journal of Clinical Oncology 8, 1277–1280 (1990)2Wen PY, et al. Updated Response Assessment Criteria for High-Grade Gliomas: Response Assessment in Neuro-Oncology Working Group. Journal of Clinical Oncology 28, 1963–1972 (2010)McDonald Criteria

1Most widely used guidelineRelies on the product of largest diameters of enhancing tumor

Key Limitations

Enhancement is not

specific to

tumor

Doesn’t

account for non-enhancing

tumor

Doesn’t

account for

multifocality

Doesn’t

address

PsP

or

PsR

RANO Criteria

2

(

R

esponse

A

ssessment in

N

euro-

O

ncology)

Both clinical and imaging components

Enhancing and non-enhancing tumor

Measurable and non-measurable disease

Accounts for

PsP

and

PsR

First progression following therapy (<12 weeks)

New enhancement outside radiation field

Pathologic confirmation of progression

First progression following therapy (>12 weeks)

New

lesion outside treatment

field

Increase

in size of ≥ 25% (accounting for

steroids)

Increase

in non-enhancing tumor

on

anti-angiogenic

RX

Clinical

deterioration not otherwise explained

RANO Criteria should be considered when evaluating glioma treatment responseSlide31

Emerging Therapies

Laser-Induced Thermal TherapyVaccine & Immune-Based Therapy

Tumor-Treating FieldsSlide32

Laser-Induced Thermal Therapy

30 year old woman with a non-enhancing right frontal mass lesion,

rCBV is not elevated

FLAIR

T1+C

FLAIR/

rCBV

Norred

SE, & Johnson JA. Magnetic Resonance-Guided Laser Induced Thermal Therapy for Glioblastoma

Multiforme

: A Review.

BioMed

Research International

2014,

1–9 (2014

)

Laser

ablation was elected over

craniotomy and was

performed at the time of stereotactic

biopsy, which confirmed the diagnosis of WHO grade II

glioma

. MRI images confirm fiber placement within the lesion and are subsequently used to monitor tissue heating

Optical fiber introduced into tumor is used to deliver direct laser energy, inducing thermal damage & necrosis

Less invasive

treatment approach

Performed concurrently with MRI

Confirm correct fiber location

Real-time monitoring of tissue heatingSlide33

Vaccine & Immune Therapies

Ongoing clinical trials for recurrent GBM:Vaccines against tumor-specific antigens (rindopepimut, HSPPC-96)Stimulated autologous dendritic cells pulsed with immunogenic tumor antigens (ICT-107)Redirected autologous T-cells engineered to target tumor-specific moieties (CART-EGFRvIII)Experimental treatments are resulting in new and complex imaging manifestations in the post-treatment setting

Surveillance examinations in a patient with recurrent GBM on a vaccine therapy show nodular enhancement and elevated

rCBV

, suggesting tumor progression

T1+C

rCBV

6 months

12 months

Biopsy, however, revealed a predominance of necrosis

80% necrosis

6

0-70% necrosisSlide34

Tumor-Treating Fields

Novel therapeutic methodologyDirect application of a non-uniform electric field Disrupts mitotic activity within actively dividing cellsPreferentially induces apoptosis within tumor cellsCourtesy: NovaCare

1Wong ET et al. Response assessment of NovoTTF-100A versus best physician’s choice chemotherapy in recurrent glioblastoma. Cancer Medicine. 2014 Jun;3(3):592–602.In

a phase III clinical trial for recurrent GBM

1, demonstrated equivalent efficacy and less toxicity when

compared to chemotherapySlide35

Conclusions

RANO criteria should be considered to evaluate treatment responsePseudoprogression: CommonSuspect with  of enhancing lesion ~ 3 monthsMore often seen with methylated MGMTImproved overall survivalMR perfusion and spectroscopy can help differentiate from true progression

Pseudoresponse:

Rapid decrease in

contrast enhancement with a high response rate

Progression of non-enhancing tumor componentModest effects on overall survival

Emerging brain tumor therapies are resulting in new and complex imaging patterns in the post-treatment setting

Advanced

multi-parametric

imaging techniques

hold promise

for improved characterization of

post-treatment patterns

, however this remains an active area of research

&

will likely require more prospective testing &

standardized

quantitationSlide36

Acknowledgements

Gaurav VermaSumei WangSulaiman SheriffAndrew MaudsleyR21 Grant: 1R21CA170284EPSI Development GroupSlide37

PENN RADIOLOGY

THE ROOTS OF RADIOLOGICAL EXCELLENCETHANKS FOR VIEWING OUR EXHIBIT!jeffrey.ware2@uphs.upenn.edusuyash.mohan@uphs.upenn.edu