Dushyant Sahani, M.DDushyant Sahani, M.DDirector of CT

Associate Professor of Radiology Massachusetts General Hospital

Harvard Medical SchoolEmail-dsahani@partners.org

Director of CTAssociate Professor of Radiology Massachusetts General Hospital

Harvard Medical SchoolEmail-dsahani@partners.org

CT Perfusion Imaging: Technique and Applications in the Body

CT Perfusion Imaging: Technique CT Perfusion Imaging: Technique and Applications in the Body and Applications in the Body

AcknowledgementsAcknowledgementsAcknowledgements• Oncology

– C Willett MD.– A Zhu MD– L Blaszkowsky MD– T Lynch MD– S Yoon MD

• TIMC (3D Lab)– G Harris PhD – A Singh MD – W Cai PhD– X Ma MD

• Oncology– C Willett MD.– A Zhu MD– L Blaszkowsky MD– T Lynch MD– S Yoon MD

• TIMC (3D Lab)– G Harris PhD – A Singh MD – W Cai PhD– X Ma MD

• Correlative Science(Steele Lab)

– R Jain PhD– D Duda PhD– Y Boucher PhD

• Radiology– A Kambadakone MD– O Catalano MD– A Galluzzo MD– H Pien PhD– G Sorensen MD– Sanjay Saini MD– Peter Mueller MD

• Correlative Science(Steele Lab)

– R Jain PhD– D Duda PhD– Y Boucher PhD

• Radiology– A Kambadakone MD– O Catalano MD– A Galluzzo MD– H Pien PhD– G Sorensen MD– Sanjay Saini MD– Peter Mueller MD

RECIST= Response Evaluation Criteria in Solid Tumors

WHO = World Health Organization

18 mm

RECIST

WHO

29 mm

59 mm

Conventional method of monitoring treatment response is change in tumor size

Monitoring Response to Conventional Chemotherapy

Monitoring Response to Conventional Monitoring Response to Conventional ChemotherapyChemotherapy

RECIST 1.010 Target Lesions (>1-2 cm)

5 max in an organ

Non-target lesions

RECIST 1.15 Target Lesions (>1 cm)

2 max in an organ

Short-axis of LN>15 mm

Eisenhauer EA et al. EJC 2009

Therasse P et al. JNCI 2000

Therasse P et al. EJC 2006

RECIST WHOType of metric Uni-dimensional Bi-dimensional (CP)

MAD X LPDCR (Complete Response)

Total disappearance Total disappearance

PR (Partial Response)

30% decrease 50% decrease

PD (Progressive Disease)

20% increase 25% increase

SD (Stable disease)

Neither PR or PD criteria met

Neither PR or PD criteria met

Monitoring Response to Chemotherapy

Monitoring Response to Monitoring Response to ChemotherapyChemotherapy

• Tumor morphology–Confluent, Irregular borders–Unusual configuration;

Circumferential (eg. mesothelioma)–Lesion length > 1.5-2 times lesion

width• Discordant results due to RECIST

technique–Uni-dimensional measurement–Shape changes may confound results

• Tumor morphology–Confluent, Irregular borders–Unusual configuration;

Circumferential (eg. mesothelioma)–Lesion length > 1.5-2 times lesion

width• Discordant results due to RECIST

technique–Uni-dimensional measurement–Shape changes may confound results

Limitations of RECIST guidelinesLimitations of RECIST guidelinesLimitations of RECIST guidelines

Volume: 73.286cm3

5.7 months

13.8 months 18.6 months

2.3 months0 months

19.6 monthsVolume: 161.591cm3Volume: 59.677cm3

Liver tumor treated with chemotherapy

Tumor volumetry is a better representative of tumor burden

Monitoring Response to Chemotherapy: Tumor Volume

Monitoring Response to Monitoring Response to Chemotherapy: Tumor VolumeChemotherapy: Tumor Volume

Prasad SR. Radiology 2002. Husband JE et al. BJC 2004

Good Responders Poor Responders≥ 15% decrease in tumor density <15% decrease in tumor density

Tumor Density: Choi criteriaTumor Density: Choi criteria

* ROI drawn around the margin of the entire tumor† Portal venous phase images for the tumor density measurement in abdomen

¥ Multiple lesions - Mean HU of all the lesions

Choi et al J Clin Oncol 2007, Choi et al The Oncologist 2008

PRE

96HU 54HU 25HU28HU

POST PRE POST

Conventional Imaging LimitationsConventional Imaging LimitationsConventional Imaging Limitations

• Evaluates the gross anatomical change of molecular events

• Time lag- typically weeks to months elapse before change observed

• Cannot measure early changes of disease process

• Changes do not necessarily correlate with disease process

• Cannot measure drug distribution

• Evaluates the gross anatomical change of molecular events

• Time lag- typically weeks to months elapse before change observed

• Cannot measure early changes of disease process

• Changes do not necessarily correlate with disease process

• Cannot measure drug distribution

Novel Oncologic DrugsNovel Oncologic Drugs

AngiogenesisAngiogenesisAngiogenesis• The development of new vessels from preexisting

ones*• Essential step in establishment and growth of

malignancies• Allow tumor progression from in-situ lesion to

invasive

• The development of new vessels from preexisting ones*

• Essential step in establishment and growth of malignancies

• Allow tumor progression from in-situ lesion to invasive

*(Risau W.,1997) *(J. Folkman,1995)Image Courtesy: O Clement MD, Paris, Fr.

Drug DevelopmentDrug DevelopmentDrug Development

–It takes an average of $802M and 12 years to bring a new drug to the market

0 2 4 6 8 10 12

Preclinical

Phase I

Phase II

Phase III

LT animal

NDA

$335M

$142M

$137M

$174M

$14M

From: DiMasi, 2003

SUVs 4.5 1.24 0.75

Baseline 1 month 16 months

Demetri et al N Engl J Med 347:472-480, 2002

FDG-PET Imaging of Imatinib(Gleevec, Novartis) on GIST

FDG-PET Imaging of Imatinib(Gleevec, Novartis) on GIST

BLOOD FLOWFDG-PETENDOSCOPY

baseline

2 weeks following therapy

40 ml/100gm/min

90 ml/100gm/min

Imaging Biomarker Selection: Drug MechanismImaging Biomarker Selection: Imaging Biomarker Selection: Drug MechanismDrug Mechanism

Willett CG et al. Nat Med. 2004 Feb;10(2):145-7

Pre-treatment 10 days Post-Avastin

Image Biomarker-Good ResponseImage BiomarkerImage Biomarker--Good ResponseGood Response

Image Biomarker- Poor ResponseImage BiomarkerImage Biomarker-- Poor ResponsePoor Response

Pre-treatment 10 days post-Avastin

Microvascular StructureMicrovascularMicrovascular StructureStructure

Jain et al., Nat Rev Cancer 38:266, 2002

Normal tissueNormal tissue Tumor tissueTumor tissue

DilatedTortuousSpatially

heterogeneous

DilatedTortuousSpatially

heterogeneous

OrganizedArtery-venous network

Tumor Vasculature Abnormalities Influences Contrast Enhancement

Kinetics on DCE CT/MR

Tumor Vasculature Abnormalities Tumor Vasculature Abnormalities Influences Contrast Enhancement Influences Contrast Enhancement

Kinetics on DCE CT/MRKinetics on DCE CT/MR

• Blood Flow• Blood Volume• Mean Transit Time• Permeability

• Blood Flow• Blood Volume• Mean Transit Time• Permeability

BF

PS

RELIABLE:Iodine Concentration (mg/ml)

CT attenuation

CONVENIENT:• Available technique• High spatial resolution• Low inter-tester variability

• Software is commercially available

RELIABLE:Iodine Concentration (mg/ml)

CT attenuation

CONVENIENT:• Available technique• High spatial resolution• Low inter-tester variability

• Software is commercially available

= linear related

Why CT?Why CT?Why CT?

Miles KA. Acad Radiol 2000;7:840–50

Technique: Site selectionTechnique: Site selection

• Non-contrast CT to cover the entire organ–5mm helical

• 2 cm tumor/4 slices (non-necrotic portion) to be covered for dynamic imaging is selected

• 4 cm with 64-MDCT

• Non-contrast CT to cover the entire organ–5mm helical

• 2 cm tumor/4 slices (non-necrotic portion) to be covered for dynamic imaging is selected

• 4 cm with 64-MDCTSahani DV et al. Radiology 05/07Goh V et al. Radiology 06, ER 07

• Contrast injected at 4-7 cc/ sec• Delay = 5-8 sec (abd) 10 sec (pelvis)• Cine acquzition

–4 contiguous 5 mm slices X 30-120 sec (every 1-2 sec)

–kVp 80-100 and mA 100-160–Limited data –4 slices once every 10-20 sec for 4-6 minutes

• Contrast injected at 4-7 cc/ sec• Delay = 5-8 sec (abd) 10 sec (pelvis)• Cine acquzition

–4 contiguous 5 mm slices X 30-120 sec (every 1-2 sec)

–kVp 80-100 and mA 100-160–Limited data –4 slices once every 10-20 sec for 4-6 minutes

Scaning TechniqueScaning Technique

Sahani DV et al. Radiology 05/07Goh V et al. Radiology 06, ER 07

CTp Technique: Rectum CTp Technique: Rectum

CTp Technique: Abdomen

CTpCTp Technique: Technique: Abdomen Abdomen

CTp Techniques and Protocols

CTp Techniques and Protocols

Technique Protocol Benefits limitations

First pass 20-30 sec cine Breath-holdLess radiation

Inadequate PS measurement

Permeability (PS) Cine 45-120 secLimited scan every 10-20 seconds for 4-6 minutes

Permeability Susceptible to MotionMore radiation

Parameters computedParameters computed

• BF = Blood flow• BV = Blood volume• MTT = Mean transit time• PS = Permeability surface

• BF = Blood flow• BV = Blood volume• MTT = Mean transit time• PS = Permeability surface

Parameters dependent on the scanning technique and mathematic modeling

CT Perfusion (GE)1

Functional CT (Siemens)2

Brilliance (Philips)

Mathematic Model

Deconvolutionmethod

Two-compartment model

Slope method

Principle of the Model

Impulse residue function (IRF) which is time enhancement curve of tissue due to idealized instantaneous injection of one unit of contrast

One way transfer of CM from intra to extra-vascular space proportionate to blood clearance constant, α

Perfusion is ratio of max slope of tissue enhancement curve to max arterial enhancement

Parameters measured

BF, BV, MTT, PS BV and Permeability MTT, time to peak enhancement

Advantages BF, BV, MTT and PS can be calculated using a single CT study

1. Simple analysis2. Efficient in calculation of rate constant K value

1. Short scan duration2. “No venous outflow” is true3. No recirculation

Limitations Partial volume averaging correction required

Assumes that back flux of CM from EVS to IVS is negligible for first 1-2 min

Sensitive to image noise

1Sahani et al, Radiology 2005, 2Ng et al, Radiology 2006

Perfusion CT Parameters and Significance

Perfusion CT Parameters and Significance

Parameter BF BV MTT PS

Definition Flow rate through vasculature in tissue region

Volume of flowing blood within a vasculature in tissue region

Average time taken to travel from artery to vein

Total flux from plasma to interstitial space

Marker Tumor VascularityTumor grade

Mitotic activity and vascularity.

Perfusion pressure

Immature leaky vessels.

Staging, Grading and PrognosisClinical

Application Author

(Journal/Year)Observations

Head and neck Zima et al (Am J Neuroradiol 2007)

Hermans et al (Int J Radiat OncolBiol Phy 2003)

Upper aerodigestive tract cancers with high BF and BV values respond well to induction chemotherapyHead and neck cancers with lower perfusion rate (BF) show poor response to radiotherapy (high local failure).

Lung Li et al (Clinical Radiology 2008) Lung cancers with distant metastases have high BF, BV and and different histological types of lung cancer show no difference in perfusion characteristics

Breast Hirasawa et al(Acad Radiol 2007) Nonscirrhous carcinomas have high BF values compared to scirrhouscarcinomas

Liver Zhu et al (The Oncologist 2008)

Sahani et al (Radiology 2007)

Patients with progressive disease (HCC) had lower baseline MTT values

Well differentiated HCCs show high BF, BV, PS and low MTT values than poorly differentiated HCCs

Pancreas d’Assignies et al (Radiology 2008)

Park et al (Radiology 2009)

Benign endocrine tumors have high BF values. Malignant tumors with liver & lymphnodal metastases have long MTTPancreatic cancers with high baseline KTrans values responded better to concurrent chemoradiation

Colon and Rectum

Sahani et al (Radiology 2005)

Bellomi et al (Radiology 2007)

Rectal cancers with high baseline BF and low MTT responded poorly to chemoradiationRectal cancers with high baseline BF and BV showed good response to chemoradiation

Monitoring Antiangiogenic Response: CT perfusion

Monitoring Monitoring AntiangiogenicAntiangiogenic Response: CT Response: CT perfusion perfusion

Pre- Avastin 10 day Post- Avastin

Favourable ResponseDrop in Blood Flow

Drop in Blood Volume

Monitoring Response to Antiangiogenic(Avastin) Therapy in HCCMonitoring Response to Monitoring Response to AntiangiogenicAntiangiogenic((AvastinAvastin) Therapy in HCC) Therapy in HCC

Parameter Pre Avastin Post Avastin P value

Blood Flow (ml/100mg/min)

105 ± 92.9 50 ± 28.8 0.014

Blood Volume (ml/100mg)

5.4 ± 3.9 2.7 ± 1.1 0.009

Mean Transit time (sec) 7.3 ± 2.8 8.8 ± 2.3 0.009

Permeability Surface (ml/100mg/min)

34.28 ± 14 21.9 ± 8.2 0.003

P value from ‘paired student t test’ between the means of pre and post Avastin

Zhu et al. The Oncologist (2007)

BF = 9.89 ml/100g /min

BV = 0.42 ml/100g

MTT= 11.42 sec

PS= PS= 1.44 ml/100g /min1.44 ml/100g /min

HU= 37,21

BF = 38.1 ml/100g /min

BV = 1.65 ml/100g

MTT=4.38 sec

PS= PS= 5.29 ml/100g /min5.29 ml/100g /min

HU= 49,12

Permeability Surface = 8.57 ml/100g /minBlood Flow = 86.3 ml/100g /min

BaselineBaseline

Lung Cancer Response to CXT

Permeability Surface = 5.24 ml/100g /minBlood Flow = 47.6 ml/100g /min

Post CXTPost CXT

CTP1 CTP2 CTP3

BF 90.3±65 72.65±50 63.5±55

BV 3.36±1.46 2.58±1.4 2.4±2

MTT 4.96±3.3 5.51±3.2 2.9±2.2

PS 13.1±5.87 10.1±6.3 ±6.1

Monitoring Treatment Response

Clinical Application

Author (Journal/Year) Observations

Liver Zhu et al (The Oncologist 2008) Fall in BF, BV, PS and rise in MTT after

antiangiogenic treatment in HCC

Rectum Sahani et al (Radiology 2005)

Bellomi et al (Radiology 2007)

Willett et al (Nature Medicine 2004)

Fall in BF and rise in MTT after chemoradiation in rectal cancerFall in BF, BV and PS after chemoradiation in rectal cancerFall in BF and BV after antiangiogenic treatment in rectal cancer

Pre- Avastin

Post- Avastin

BF MTT

BF MTT

24.3 ml/100g/min

13.9 ml/100g/min

7.9 sec

13.9sec

Sarcoma: Sarcoma: AntiangiogenicAntiangiogenic T/tT/t

BLOOD FLOWPETENDOSCOPY

baseline

2 weeks following therapy

40 ml/100gm/min

90 ml/100gm/min

Monitoring Antiangiogenic (Avastin) Response in Rectal CancerMonitoring Monitoring AntiangiogenicAntiangiogenic ((AvastinAvastin) ) Response in Rectal CancerResponse in Rectal Cancer

Willett CG et al. Nat Med. 2004 Feb;10(2):145-7

Rectal Cancer: CTp changes following Treatment

Rectal Cancer: Rectal Cancer: CTpCTp changes changes following Treatmentfollowing Treatment

MGH Experience

Willett C et al. JCO 2009

Validation and ReproducibilityClinical

Application Author (Journal/Year) Observations

Lung Ma et al (BMC Cancer, 2008) BF, BV and PS values of peripheral lung cancer correlated

positively with MVD

Liver Sahani et al (Radiology 2007) Reproducibility of BF, BV, PS and MTT values with high correlation and variability of 4% in HCC

Pancreas d’Assignies et al (Radiology 2008)4.Abe et al29 (Radiat Med 2005)

BF values of pancreatic endocrine tumors correlated well with

MVDGood linear correlation of BF measured by and CTp in pancreatic tumors

Colon & Rectum

Goh et al (Am J Roentgenol 2006)

Li et al (World J Gastroenterol 2005)

Quantitative perfusion measurements are reproducible in colorectal cancerBF values of colorectal carcinomas did not correlate with MVD

Chan NG et al. CTp JCAT 2009

CTp Challenges CTpCTp Challenges Challenges

• Limited sample volume (2-4 cm)–Choice of location for the

investigation critical • The CTp parameters are

estimates of tissue perfusion • Patient motion can impact

perfusion values

• Radiation dose is an issue

• Limited sample volume (2-4 cm)–Choice of location for the

investigation critical • The CTp parameters are

estimates of tissue perfusion • Patient motion can impact

perfusion values

• Radiation dose is an issue

FDA: High Radiation Doses at Cedars-Sinai Hospital

Radiation Overdoses Point Up Dangers of CT Scans

FDA: High Radiation Doses at FDA: High Radiation Doses at CedarsCedars--Sinai HospitalSinai Hospital

Radiation Overdoses Point Up Radiation Overdoses Point Up Dangers of CT ScansDangers of CT Scans

• Appropriate indication• Appropriate indication

Strategies to Lower CTP DoseStrategies to Lower CTP DoseStrategies to Lower CTP Dose

• Appropriate indication• Compromise resolution and SNR

–5-10 mm thickness–Low dose kVp 80-100 mA100-160–2 second temp resolution–Cine < 40 sec–Appropriate scan delay based on the

circulation time

• Appropriate indication• Compromise resolution and SNR

–5-10 mm thickness–Low dose kVp 80-100 mA100-160–2 second temp resolution–Cine < 40 sec–Appropriate scan delay based on the

circulation time

SummarySummarySummary

• Imaging integral to monitoring treatment response in Oncology trials and decision-making –Expectations are changing –Beyond 2D measurements

• Volume• Density • Function • Combination

• Imaging integral to monitoring treatment response in Oncology trials and decision-making –Expectations are changing –Beyond 2D measurements

• Volume• Density • Function • Combination

SummarySummarySummary

• CT perfusion imaging is an evolving field – redefines CT as a technique that can now depict

vascular physiology in addition to detailed anatomy– Is getting increasingly important in trials for

targeted therapies • Protocol customization is mandatory

– to enable relevant tumor vascular physiology data– Radiation dose optimization

• Data is specific to protocol and processing method

• CT perfusion imaging is an evolving field – redefines CT as a technique that can now depict

vascular physiology in addition to detailed anatomy– Is getting increasingly important in trials for

targeted therapies • Protocol customization is mandatory

– to enable relevant tumor vascular physiology data– Radiation dose optimization

• Data is specific to protocol and processing method

THANK YOUTHANK YOUTHANK YOU