Image-Guided Radiation Therapy: Image Guided Radiation Therapy: A Refresher D.A. Jaffray, Ph.D. Radiation Therapy Physics Princess Margaret Hospital/Ontario Cancer Institute Professor Departments of Radiation Oncology and Medical Biophysics University of Toronto
106
Embed
Image-Guided Radiation Therapy:Guided Radiation Therapy · PDF fileImage-Guided Radiation Therapy:Guided Radiation Therapy: A Refresher ... imaging technology reported here and ...
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Image-Guided Radiation Therapy:Image Guided Radiation Therapy: A Refresher
D.A. Jaffray, Ph.D.
Radiation Therapy Physicspy yPrincess Margaret Hospital/Ontario Cancer Institute
ProfessorDepartments of Radiation Oncology and Medical Biophysics
University of Toronto
DisclosureDisclosure
Presenter has a financial interest in some of thePresenter has a financial interest in some of the imaging technology reported here and
research collaborations with Elekta Philipsresearch collaborations with Elekta, Philips, IMRIS, Modus Medical, and Raysearch.
Results from studies using investigational devices will be described in this presentation.
Learning Objectives• Learn the challenges in achieving precision
and accuracy in treatment sites influenced by organ motion.
• Identify the technologies available or being y g gdeveloped for image-guided RT.
• Understand the benefits and limitations ofUnderstand the benefits and limitations of current solutions.
• Be aware of quality and safety issues related• Be aware of quality and safety issues related to IGRT practice
Pre- Test QuestionPre Test Question
• What type of geometric targeting uncertaintyWhat type of geometric targeting uncertainty has the biggest impact on the PTV margin in fractionated RT?fractionated RT?1. The Random-type Errors
2 The Systematic type Errors2. The Systematic-type Errors
3. The Residual Errors
4 The Absol te Error4. The Absolute Error
5. Not Sure
IGRT: Targeting vs Image-based M tManagement
Therapy Design
Intervention
IGRT
Clinical ObjectiveMonitoring
Inter- and Intra-fraction Uncertainties
Local Therapy and Image-Guidancepy g• Radiation therapy is a proven local therapy. • Precise radiation therapy offers:Precise radiation therapy offers:
– Reduce severity and risk of therapy-induced complications.– Increase both quality and probability of success.
• Further potential:– Broaden application of proven therapies.
Permit new therapies that are intolerant to geometric– Permit new therapies that are intolerant to geometric imprecision.
• Addressing geometric uncertainties may expose other g g y pfactors determining outcome.Co-localization of the target and therapeutic dose distribution within the human body is a
significant technical challenge.
Targeting Uncertainty in RTTargeting Uncertainty in RT• Setup Variation
• Internal Organ Displacement
• Volume Change and Deformation
Motion: Respiration-Induced
Normal Breathing
Deep Breathingg Breathing
Breath-hold Exhale
Breath-hold Inhale
Siemens 1.5T (2000)
FAST, ‘PERIODIC’
Motion: Bladder Filling
1 hr cine MR (sagittal, TRUFISP sequence)
MODERATE, NOT REPRODUCIBLE
Motion: Bowel Effects“Full” Rectum “Empty” Rectum
FAST, NOT PERIODIC
Motion: Bowel Effects
Measured Prostate and Rectal Wall Motion
Movie loop derived from 17Movie loop derived from 17 volumetric CT scans.
P t t R dProstate - RedRectum - Green
Left Lateral – Treatment Beam Perspective
Cancer f hof the
Cervix: Therapy-induced Changes Week 1 Week 2
S i lSagittal Images
Week 3 Week 4Chan, Dinniwell et al., PMH
FAST + SLOW, UNPREDICTABLE
IG Technologies - Generation IIG Technologies Generation I
Cyberknife
kV Radiographic
Ultrasound Portal Imaging
Markers (Active and
Passive)
IG Technologies - Generation IIIG Technologies Generation II
Faster imaging, higher CNR, more responsive in delivery.T. Pawlicki – Sunday 11:15-12:15 am
Ultrasound-based 4D Prostate Tracking
St t f th t US I Q lit• State-of-the-art US Image Quality• Does not require bladder fill for prostate
imaging• Potential application in prostatectomy and
GYN ti tGYN patients• Mechanical 3D probe operated remotely• Probe outside of radiation fields
• no effect on dosimetry orcollisionN i i d i i i• Non-invasive and non-ionizing
Exterior view of the system. The O-ring is skewed in the
counterclockwise direction.
O-ring Structure (d = 3.3 m) Gimbal Mounted Treatment Head (c MLC)Two kV x-ray Tubes for Stereo LocalizationTwo Flat-panel Detectors
(a) Cone beam computed tomography image of the pelvis for a prostate case. The X-ray parameters were 120 kVp, 200 mA, 10 ms, and 800 mAs. The total monitoring dose was 19.4 mGy. (b) The conventional X-ray computed tomography image of the same area of the same patient.
Kamino et al. IJORBP, 2006
MRgRT External Beam WorkflowResearch Facility - under construction – operational spring 2013
IGRT-guided pre-localization of MR Imaging FOV
Confirmation of delivery i bilit
Robotic control of MR, table and Shielding System
Linear motion of magnet ti t
Critical time specification (<90s) from end of imaging
to beam-on.
I i (di t tiviability
Reference CBCT for MR-guidance
over patient.
RT present for movement.
Pre-stored MR configuration f MR i l ti St
Image processing (distortion correction, calibration) and
planning (adaptation).
Generation of couch orfrom MR-simulation Stage Generation of couch or machine adjustment.
*MR can begin to image within 5s of stopping.
MRgRT Pelvis Coil: Volunteer Images
Prostate
1.5 T, T2 weighted images of two volunteers.
March 2009First Images March 2009
MR imaging without 6 MV irradiation
MR imaging during 6 MV irradiation ofMR imaging during 6 MV irradiation of
object imaged (no FF)
Courtesy of G. Fallone, Cross Cancer Institute, Edmonton, Canada
Courtesy of J. Lagendijk, Utrecht, Netherlands
Courtesy of J. Lagendijk, Utrecht, Netherlands
Courtesy of J. Lagendijk, Utrecht, Netherlands
ICRU 50 and the Development of IG in Radiation Therapy
• International Commission of Radiation Units and Measurements (ICRU) Report #50 & #62#50 & #62
• Formalisms developed to identify and explicitlyidentify and explicitly communicate the intentions of the RT intervention:– Radiation Dosage– Geometric Extent (the need to
constrain the geometric extent for gthe avoidance of normal structures)
– Uncertainty in the intervention
Prescription: Merits of Embracing the ICRU Formalism
Cl i i f h h i• Clear communication of the therapeutic intention to the rest of the technical staff– Extent of microscopic disease and normalExtent of microscopic disease and normal
structure• Quality and Uniformity of Practice• Created a fulcrum in the system to
explicitly accommodate the geometric uncertainties inherent in the processuncertainties inherent in the process.
• Driven the development of “margins”– Margins to guarantee coverageMargins to guarantee coverage– Margins to guarantee avoidance (no fly zone)
The Role of the ICRU FormalismThe Role of the ICRU Formalism in Advancing RT Practice.
• Explicit accommodation of the geometric uncertainty associates it with the unnecessary irradiation of a structure due to t h i l li it ti f th tistructure due to technical limitations of the practice.
• Creates a clear and distinct association between technology and toxicity can be of great advantage intechnology and toxicity can be of great advantage in formalizing arguments for (i) advancing the technological state of the field, and (ii) standardizing th tithe practice.
• With standardization comes opportunity for demonstration of the improvement associated withdemonstration of the improvement associated with technological advancement.
IntentIntent
‘Actual’
4-field box CRT 3D-CRT IG-IMRT
Images courtesy Dr. John Schreiner
1
Prostate
2
1
Weekly/Pelvis
Skin Mark
od RL
4
3
Daily Ultrasound
Daily/Pelvis
ctio
n M
etho
RL AP SI
6
5
Daily CBCT/Tissue
Daily Radiograph/Fiducial
Cor
rec
7
0 2 4 6 8 10
Daily CBCT/Fiducial
Comparison of residual errors for different image-guided correction techniques in prostate in the left right (LR) anterior
Residual Error (mm)
correction techniques in prostate, in the left-right (LR), anterior-posterior (AP), and superior-inferior (SI) directions.
Mageras GS, Mechalakos J. Semin Radiat Oncol. 2007 Oct;17(4):268-77
• Lexicon for Image-guidance• PTV is a conceptual ‘device’ created to assure
dosimetric coverage of the CTV• Therefore must accommodate both geometry and• Therefore, must accommodate both geometry and
dosimetry in determination of the PTV margins.• Fractionation also plays a role in marginFractionation also plays a role in margin
determination
PTV Design with IGRTPTV Design with IGRT
• IGRT reduces but does not eliminateIGRT reduces, but does not eliminate, geometric uncertainties
• Reduced geometric uncertainties may allow• Reduced geometric uncertainties may allow reduced PTV marginsP ll l/ l d• Parallel/related processes:– Image guided radiation therapy– Quality assurance / uncertainty management
• Need to be aware of other uncertainties (i.e. (target definition)
“Off-line Corrections”: Increasing Accuracy and Measuring Precision for Reduced Margins
Individual UncertaintyPopulation UncertaintySI
Lat
SystematicCorrection
Requires time, new information, and effort to estimate the
systematic correction and random uncertainty for a patient.
Population Margins Individual Margins
“On-line Corrections”: Increasing Accuracy and Increasing Precision for Reduced Margins
Individual Uncertainty – Off-lineg g
Reduced Uncertainty – On-lineSI
Lat
SystematicCorrection Multiple
On-line
Concerns:(i) Residual Uncertainties (e.g. intra-fractionOn line
Corrections(i) Residual Uncertainties (e.g. intra fraction
motion, measurement and correction errors)(ii) Strength of IG Surrogates (e.g. markers, bones)
Reduced MarginsIndividual Margins(iii) Stability of Systematic Error (e.g. time trends)
PTV Margin Recipes• Most popular formulation is 2.5Σ + 0.7σ, where:
Σ is the standard deviation of systematic uncertaintyΣ is the standard deviation of systematic uncertaintyσ is the standard deviation of random uncertainty
A ifi f i i d di d• Assumes specific fractionation, dose gradient and dose coverage objectives.
E.g: “We have measured the random uncertainty for our prostate patients to be a standard deviation of 4mm and the systematic uncertainty to be a4mm and the systematic uncertainty to be a standard deviation of 2mm.Using (2.5 x 2mm) + (0.7 x 4mm), this results in 7mm
M Van Herk, Seminars in Radiation Oncology, 14(1) 2004 52-64
Using (2.5 x 2mm) + (0.7 x 4mm), this results in 7mm PTV margins.”
Rational Margin Design vs Misplaced g g pUncertainty – ‘The PTV Carpet’
PTV
Impact of adopting information that b ‘ b i l b ’appears to be ‘obviously better’
IS3R’11 Geets et al. 2005Particularly concerning for systematic changes in target
volume delineation.
4D IGRT Brings More Variables4D IGRT Brings More Variables• Phase-specific treatment approaches
– Gating, tracking, breath-hold• Recognize phase and amplitude as positioning
i blvariables.• Examine the random and systematic errors in
positioning the target with these variablespositioning the target with these variables.• Stability of these parameters over time will challenge
current off-line strategies.g• Deformation….many more patient-specific variables.
• T4N2 NPC for combined modality therapy. GTV ‘hugging’ brainstem and chiasm
Organs at Risk
hugging brainstem and chiasm• Clivus and cavernous sinus chosen as region of
interest (ROI) / ’clipbox’ for matchinginterest (ROI) / clipbox for matching
Head and Neck Group, PMH
Hitting the Target and Avoiding Organs at Risk
• Spine curvature not reproducible
Organs at Risk
• Non-uniform margins are appropriate
Head and Neck Group, PMH
Image-guided Radiation TherapyImage guided Radiation Therapy is Quality Assurance
• Streamlined measurement of patient/anatomy position within the treatment room.– Quantitative, accurate scale– Minimally invasivey– Large field-of-view
• Verify consistency of planned and actualVerify consistency of planned and actual geometry
• Feasible to integrate into current• Feasible to integrate into current cost/operational paradigm
Impact of IGRT on Localization ErrorsImpact of IGRT on Localization ErrorsGradual adoption of on line CBCT
IGRTof on-line CBCT
since 2005.
15/16 Units now equipped with kV
cone-beam CT.
The rate of reportable incidents (i.e. non-near miss) has p ( )decreased by 50% since introduction of IGRT (2005-2007).
Internal Data - PMH, Toronto
Are we applying IGRT technologies properly? Are we inadvertently doing harm?
IGRT is currently a solid tool to tackle the problem of radiotherapy accuracy (reduction in systematic errors).
h ( d d li i ) li i h i d i fOther errors (read: target delineation) limits the PTV margin reduction for most RT treatments.
facilitate implementation of new RT techniques (eg liver and lung SBRT)… facilitate implementation of new RT techniques (eg, liver and lung SBRT) and in selected sites reduce toxicity and improve local control.
The whole chain of interventions in the RT process should be prospectivelyThe whole chain of interventions in the RT process should be prospectively assessed. This is particularly important because other steps in the RT process (eg, contouring or valid measurements of toxicity) are at least as important as high geometric precision.
Bujold et al. Semin Radiat Oncol 22:50-61 (2012)
g g p
Unlikely we will see RCT for IGRT technology in isolation.
There are clear technical advantages, but there are also costs (more than just $$$)…
Bujold et al. Semin Radiat Oncol 22:50-61 (2012)
th f ll i i bl i ifi tl l t d t FFBF…the following variables were significantly related to worse FFBF: risk group according to the NCCN (high- to very high risk vs. intermediate- to low-risk), dose (70 vs. 78 Gy), average cross-sectional area (>16 vs. <16 cm2) and, unexpectedly, the use of implanted markers as opposed to bony structures for patient positioning. In retrospect, the margins around the clinical target
Int. J. Radiation Oncology Biol. Phys., Vol. 74, No. 2, pp. 388–391, 2009
positioning. In retrospect, the margins around the clinical target volume appeared to be inadequate in the cases in which markers were used.
…arbitrarily chosen non-uniform margins of 3 mm LR and 5 mm (AP; CC)…. narrow LR margin may lead to an underdosage. Second, …the planning system is unable to handle CC margins that are not a multiple of the planning CT slice thickness (2 mm); … the real CC margin was
Int. J. Radiation Oncology Biol. Phys., Vol. 74, No. 2, pp. 388–391, 2009
p g ( ); g4 mm instead.
SD-IGRT: MetastasesAn IGRT enabled activity.
(all lesions) (bone only)
MR Imaging for IGBTIGBT
• Intrinsic ContrastsIntrinsic Contrasts– Structural and Functional
• High Spatial Resolutiong p• No Known Toxicities• Volumetric Image
Acquisition• Concerns of Distortion• Access/Devices
MRI AnatomyMRI Anatomy -
Cervix CancerCervix Cancer
The mean D90 for HR CTV was 6 Gy higher when using one plan than when using individual treatment plans. The D2cc increased 3.5 Gy for the bladder, 4.2 Gy for the rectum and 5.8 Gy for the sigmoid.
Th f l t t t l ld h lt d i 8/14 diThe use of only one treatment plan would have resulted in 8/14 cases exceeding constraints for bladder, rectum or sigmoid.
MR-Guided Brachytherapy: IGRT Success Story
Are we applying IGRT technologies properly? Are we inadvertently doing harm?
Imaging, Adjustment and DeliveryTissue Definition Treatment Planning Delivery
QA Transfer of IGRT Parameters to Tx Unit !
Selection of PTV Margins Consistent with IGRT Procedure and Data
Consistent IGRT Procedure at the Tx Unit
Delineation of Target and Normal Structures -
Peer Review!!!
Review of IG Results to
! Documentation of IG Performance Data
! Confirm Coverage/Avoidance
!
!
Education and Training
Quality Assurance, Documentation, Evidence of IG Performance
! Nomenclature and IG Lexicon
Recommendation (1 thru 5) Reference( )IGRT Infrastructure 1. Establish a multi-professional team responsible for IGRT
ti iti(White and Kane 2007)
activities.2. Establish and monitor a program of daily, monthly, and
annual QA for all new or existing IGRT sub-systems.(Klein, Hanley et al.
2009; Yin, Wong et al. 2009)
3. Provide device and process-specific training for all staff operating IGRT systems or responsible for IGRT delivery.
(Yin, Wong et al. 2009)
4. Perform ‘end-to-end’ testing for all new IGRT procedures (from simulation to dose delivery) and document performance prior to clinical release.
(Yin, Wong et al. 2009)
p p5. Establish process-specific documentation and procedures
for IGRT.(Hendee and Herman ;
Yin, Wong et al. 2009)
Recommendation (6 thru 10) Reference( )IGRT Infrastructure 6. Clearly identify who is responsible for approval of IGRT
ti d i i d th h b thi(Potters, Kavanagh et
al.)correction decision and the process whereby this decision is made and documented.
)
7. Establish and document site-specific planning d ifi ll h d f d fi i
(ICRU50 1993; ICRU62 1999;procedures, specifically, the procedure for defining PTV
margins. Link these planning procedures to IGRT procedures.
ICRU62 1999; Keall, Mageras et al. 2006)
Patient-Specific Procedures8. Multi-professional peer-review of PTV volumes. Peer-
review of GTV/CTV volumes by RadOncs.(Adams, Chang et al.
2009)y9. Verify proper creation and transfer of IGRT reference
data (PTV, OARs, DRRs etc.) to IGRT system.(Potters, Gaspar et al.)
10 Establish a reporting mechanism for IGRT related (Hendee and Herman ;10. Establish a reporting mechanism for IGRT-related variances in the radiation treatment process.
(Hendee and Herman ; CAPCA 2006)
How far should we go inHow far should we go in responding to ‘displacements’ d d i h i iddetected with image-guidance
“Adaptive radiotherapy has been introduced as a feedback controlAdaptive radiotherapy has been introduced as a feedback control strategy to include patient-specific treatment variation explicitly in the control of treatment planning and delivering during the treatment
” D Ycourse.” D. Yan
4D IGRT Ch i T t d Ti4D IGRT – Changes in Targeted Tissues
Detectable Changes in Volume of ‘Something’ Courtesy of Gerald Lim, PMHNSCLC
IGRT Detected Changes in Lung T G id /Di iTargets – Guidance/Discussion
• Siker ML et al “Tumor volume changes on serial• Siker ML et al. Tumor volume changes on serial imaging with megavoltage CT for non-small-cell lung cancer during intensity-modulated radiotherapy: how g y pyreliable, consistent, and meaningful is the effect?
– Int J Radiat Oncol Biol Phys. 2006 1;66(1):135-41• Kupelian PA et al. “Serial megavoltage CT imaging
during external beam radiotherapy for non-small-cell lung cancer: observations on tumor regression during treatment.
I t J R di t O l Bi l Ph 2005 N 15 63(4) 1024 8– Int J Radiat Oncol Biol Phys. 2005 Nov 15;63(4):1024-8
Courtesy of J-J. Sonke, NKI
RT-induced Change and the Domain of Adaptation
• Major concern regarding the true extent of disease (in• Major concern regarding the true extent of disease (in planning and during RT).
• Is there opportunity to reduce target volumes as therapyIs there opportunity to reduce target volumes as therapy progresses?– Depends on what is changing – normal tissue vs target
lvolumes• Need to reflect on the definition of GTV and CTV
– State-of-the-art IGRT imaging systems are not necessarilyState of the art IGRT imaging systems are not necessarily standard of care for use in target definition
– Redefinition of target volumes is not standard of care in RTAdaptive Approaches – with the exception of ‘systematic
shifts’ are in the domain of clinical research.
IG Technologies Needed for 4D
Deformable R i t ti
Auto-Registrationsegmentation
Dose Tracking
Re-planningTracking
These systems are available in the research setting, but not mature in commercial systems.
8 GyPre-TxCa Cervix: Tumour Shrinkage
& Deformation
20 Gy 28 GyDuring RT
What is the actual dose delivered?
Can we
38 Gy 48 Gy
accommodate these changes to reduce
l inormal tissue dose?
SLOW, UNPREDICTABLE
Methods - Deformation Vector Fields (5 wks)
Week 1 Week 2 Week 3
Planning (pre-treatment)
Week 4 Week 5
Brock, et al., Medical Physics 2005;32:1647-1659.
Adaptive Benefit - Conventional vs On-line 1) IMRT with uniform 3mm PTV margin, no re-planning
2) Automatic re-plan with pre-treatment optimization function
Planning DeliverIMRT Plan Optimization Function
Automatic Weekly Replan
Stewart, et al., IJORBP, 2009Orbit Workstation, Raysearch Labs
Research Results – Target Coverage (N=33 Pats)GTV CTV
5151GTV CTV
49
50
Gy)
100%
98%49
50
Gy)
100%
98%
48
98%
Vol
ume
(
95%
48
98%
Vol
ume
(
95%
47
Dos
e to
9 95%
47
Dos
e to
9 95%
8 (24%)
2(6%)
Pl d N R l A W kl45
46
Pl d N R l A W kl45
46(24%)
Planned No Replan Assess WeeklyPlanned No Replan Assess Weekly
Message: A large fraction of patients would maintain coverage with a 3mm margin! But, who are they? We can’t tell until the 2nd-3rd week of treatment.
Summary• IGRT seeks to address geometric uncertainties in
dose placement for target and normal tissuesdose placement for target and normal tissues.• It has become a routine part of current RT
practice.p• Safe application of IGRT technology requires
additional training and careful integration into the li i lclinical process.
• Numerous new IGRT devices in development.IGRT l h i d i RT h• IGRT reveals changes in anatomy during RT that challenge conventional practices – this is an area of on-going researchof on going research.
Post- Test QuestionPost Test Question
• What type of geometric targeting uncertaintyWhat type of geometric targeting uncertainty has the biggest impact on the PTV margin in fractionated RT?fractionated RT?1. The Random-type Errors
2 The Systematic type Errors2. The Systematic-type Errors