chapter27-bwiney · imrt 26.8 34
TRANSCRIPT
6/5/15
1
Small Field Dosimetry, SRS and Eyes Brian Winey, PhD
Massachusetts General Hospital Harvard Medical School
Small Field Dosimetry
§ Small Field Definition: § Primary limitation is lateral
disequilibrium
§ Range dependency
§ Shallow ranges (<4 cm), lateral disequilibrium occurs d < 1 cm
§ Deeper ranges, lateral disequilibrium occurs at larger field sizes
Small Field Dosimetry
§ Measurements are more difficult § Volume effects and increase penumbra
§ Multiple proposed methods
§ Ion chambers, MLICs, parallel chambers, diamond, film, diodes, etc.
§ Limitations of devices already presented
6/5/15
2
Most Common Measurements
§ Small/Micro IC (ex: PTW Pinpoint, others) § PDD (CAX, OAX), Absolute Dose
§ Cross calibrated against standards
§ Saturation in Scanned beams
§ Parallel chambers: § Integrated Depth Dose
§ MLIC: § Integrated Depth Dose
Most Common Measurements
§ Film: § Planar Doses
§ LET and Calibration
§ Diodes and Diamond: § Point Doses, especially for very small fields
§ Cross calibration with IC in larger Field at same depth
§ LET and drifting
Small Field Chambers/Diodes:
6/5/15
3
Small Field Characteristics
§ Penumbra Increases with range
§ Increased lateral penumbra affects the CAX for smaller fields
§ Reduces the Bragg peak due to multiple coulomb scatter and geometry/beam optics for aperture collimation
Penumbra:
§ Sharper at Shallow Depths
§ Less Sharp at Greater Depths
Small Field Dosimetry
§ All 16cm/1cm (R/M) and entrance dose
§ Penumbra Increases
6/5/15
4
Small Field Dosimetry: PDD
Small Field Dosimetry: IDD
Similar to Plane Parallel Chamber Measurements
Field Size Effects: Output
J. Daartz, MGH
Greatest effects: Large Depth, Small Field
6/5/15
5
SRS History
§ Gamma Knife original photon treatment (1950’s)
§ Ten years later (1960’s): proton radiosurgery
§ Linac based begun in 1980’s and Cyberknife later
§ Thousands of patients treated with Photon SRS—clinically proven technique
Why Proton SRS?
§ Generally with respect to photon SRS § Distal Edge
§ Integral Dose
§ Higher TCP/Lower NTCP
History of Proton Radiotherapy § 1947: R. Wilson proposes
using protons for radiotherapy
§ 1954: First patient treated with protons at Berkeley
§ 1960+: Kjellberg studies and applies protons for neurological tumors § 1st HCL patient: 2yo with supra-
pituitary mass
§ Many reports since…
R. Wilson, Ph.D.
6/5/15
6
History of Proton Radiotherapy
§ HCL II
§ Physics Lab
R. Wilson, Ph.D.
HCL in the 1960-70’s
R. Wilson, Ph.D.
§ Seated patients, orthogonal kV imaging
§ Invasive Frame
STAR § STereotactic Alignment for Radiosurgery
§ Alignment system designed for fixed beam delivery system at HCL (1991)
6/5/15
7
STAR § STereotactic Alignment for Radiosurgery
§ Positioner was moved to MGH in 2001
§ New Beam Line: § Single Scattering System, optimized for
Cranial Proton Radiotherapy
§ Single Scattering: § Smaller Penumbra
§ Higher Dose Rates § Limited Field Size
§ Also treat on Gantries!
Proton Therapy Center at MGH
4.5m
Francis H. Burr Proton Therapy Center - STAR beamline
6/5/15
8
Dose Calculations
§ TPS (XiO): RC/Aperture Calculations § Range/Mod Calculations
§ Laminate Program Calculates Feasible R/M with BABS
TARGET
Patient Specific Target Range
Modulation d90-p90
§ Generally Use 90% Normalization § Sharper Penumbra § Less Ripples (Hot/Cold Spots)
§ Like Photon SRS: No PTV—discussion topic
§ Doses (CGyE) similar to Photon SRS (except AVM)
Patient and field specific hardware
+ =
Aperture Range Compensator
Lateral conformation
Distal conformation
Martijn Engelsman, Ph.D.
6/5/15
9
Dose Calculations
§ TPS: RC/Aperture Calculations § Range/Mod Calculations
§ Laminate Program Calculates Feasible R/M with BABS
§ Laminate Program Calculates MU using output model
§ Field Size Correction applied to output
STAR QA/QC Weekly: Range, Modulation, Output, Alignment
Ψ = ao ( 1 + a1 (R/M – 1)a2 ) × ( 1 + a3 R2 + a4 R + a5 )
Zebra (MLIC)
STAR QA/QC Monthly: Flatness, Symmetry, Comprehensive Alignment, X/P Coincidence
-60 -40 -20 0 20 40 600
10
20
30
40
50
60
70
80
9095
100
lateral distance [mm]
rela
tive
dose
horizontalvertical
100 150 200 250
40
60
80
100
120
140
160
180
200
220
240-
0
2
4
6
8
1
1difference:hor: 0.19mmver: 0.10mm
red: xhair center, black: proton field center
6/5/15
10
STAR Daily QA/QC Daily: Range, Modulation, Output, Beam Steering Constancy
IC1 IC2/3 Steering/hardware
T1 Output
RV Range
ALL BABS Layers Tested Weekly
STAR Daily QA/QC Daily: Range, Modulation, Output, Beam Steering Constancy
Invasive Fixation (1960’s-2004)
6/5/15
11
Non-Invasive Fixation
Immobilization: Conventional GTC vs mGTC
-2 -1 0 1 20
20
40
60
80
100
120
CC (mm)
Counts
-2 -1 0 1 20
20
40
60
80
100
LAT (mm) -2 -1 0 1 20
20
40
60
80
AP (mm)
-2 -1 0 1 20
50
100
150
200
Rotation (deg)
Counts
-2 -1 0 1 20
20
40
60
80
Roll (deg)-2 -1 0 1 20
10
20
30
40
50
Pitch (deg)
Intrafraction Motion
B Winey, J Daartz, F Dankers, M Bussiere
J Appl Clin Med Phys 2012 May 10; 13(3)
Immobilization precision of a modified GTC frame
Immobilization: mGTC Efficacy
6/5/15
12
Patient Alignment: Fiducials
Most difficult point in the patient workflow.
Patient Setup § Implanted Fiducials Localized in TPS § Iso and Fiducials sent to Imaging
System
§ Simple Ray-Tracing backprojection
Why Proton SRS?
§ Costs?
§ Uncertainties?
§ Benefits?
6/5/15
13
Costs?
§ Billing is the same as photon SRS
§ Add a proton modifier
§ No Additional cost versus photon
Uncertainties?
§ Range uncertainties (CT, SPR, Motion, Setup, Geometric Patient Daily Variations)
§ Motion-Miss Targets
§ Field Size Effects
§ Penumbra
§ Online Imaging Limited
§ ∴ Affect the conformality (Rx dose)
Proton range changes (Cranial SRS)
§ Fluids in sinuses § Scattering from heterogeneities § Setup Uncertainties § Air gap
Lei Dong, Ph.D.
6/5/15
14
Proton range changes: Cranial SRT § Fluids in sinuses § Scattering from heterogeneities § Setup Uncertainties § Air gap § Onyx for AVM
§ Artifacts § WET
Lei Dong, Ph.D.
Patient Setup § Implanted Fiducials Localized in TPS § Simple Ray-Tracing backprojection
§ Or 2D/2D anatomic markers
Setup Uncertainties
§ Externally verified 2D/3D registration algorithm (Reg23)
Steininger, et.al. PMB, 2012, 57, 4277-4292.
6/5/15
15
-1 0 10
5
10
15
20
25
30
Δ X (mm)-1 0 10
5
10
15
20
25
30
Δ Y (mm)-1 0 10
5
10
15
20
25
30
Δ Z (mm)0 1 20
5
10
15
20
25
30
Δ R (mm)
-1 0 10
10
20
30
40
Δ RX (deg)-1 0 10
10
20
30
40
Δ RY (deg)-1 0 10
10
20
30
40
Δ RZ (deg) !
-2 0 20
10
20
30
40
Δ X (mm)-2 0 20
10
20
30
40
Δ Y (mm)-2 0 20
10
20
30
40
Δ Z (mm)0 1 20
10
20
30
40
Δ R (mm)
-2 0 20
20
40
60
80
Δ RX (deg)-2 0 20
20
40
60
80
Δ RY (deg)-2 0 20
20
40
60
80
Δ RZ (deg)
-5 0 50
10
20
30
40
50
60
Δ X (mm)-5 0 50
10
20
30
40
50
60
Δ Y (mm)-5 0 50
10
20
30
40
50
60
Δ Z (mm)0 5 100
20
40
60
80
Δ R (mm)
-5 0 50
10
20
30
40
50
60
Δ RX (deg)-5 0 50
10
20
30
40
50
60
Δ RY (deg)-5 0 50
10
20
30
40
50
60
Δ RZ (deg)
STAR: Fiducials Gantry: Fiducials
Gantry: Anatomy
Intrafractional Motion
!
Cranial Intrafractional Motion
Impact on MFO Planning Less impact on Passive Scattered
Lei Dong, Ph.D.
Positioning Summary
Patient positioning coordinate
Fiducial-based patient positioning
Anatomy-based patient positioning
left/right 0.74 1.12superior/inferior 0.73 1.57
anterior posterior 0.85 1.18pitch 0.64 1.02roll 0.55 1.64yaw 0.39 0.89
Positioning uncertainty (mm / deg)
Patient setup positioning uncertainties at the Francis H. Burrproton therapy center of the MGH.
6/5/15
16
1.5 mm setup error
Worst Case
Pituitary Adenoma: Effects were due to lateral shifts and range uncertainties.
Uncertainty Mitigation § What do we do with all of this
information: § Margins: Distal/Proximal
§ Beam angle selection
§ Smearing
§ Feathering
§ Gating
§ OARs
6/5/15
17
Typical Planning (PBS)
§ Multiple Studies to account for uncertainties in planning
§ Robust optimization
§ Beam specific margins (only SFUD) § Preprocessing
§ Online
Beam Angle Selection
1. Avoid beam entrance angles along and through heterogeneous boundaries 2. Avoid distal edge sparing. 3. Use multiple beams to reduce uncertainty of a single beam!
OARs § AVOID distal edge sparing!
§ If unavoidable, use multiple fields to spread the risk and reduce the dose to the OAR if there is an error.
6/5/15
18
Using Multiple Beams
§ Spreads uncertainty due to range, patient setup, LET, and patient motion
§ Difference in lateral and distal uncertainties
§ Increases conformality for both scanned and scattered delivery
§ Increased Robustness
Proton SRS Treatment Planning Summary
§ Field Size/MCS
§ Beam positions
§ Heterogeneities
§ Penumbra Regions
§ Distal Positions
§ LET/RBE
§ More beams à More Conformal/Less Uncertainties from single beam
Why Proton SRS?
§ Benefits: § Distal Edge
§ Penumbra
§ Integral Dose
§ Higher TCP/Lower NTCP
6/5/15
19
Still have the Bragg Peak! § Primary ‘physics’ advantage over photons:
Bragg Peak
0
5
10
15
20
25
30
35
40
0 50 100 150 200 250 300
Energy (MeV)
Ran
ge
(cm
)
Source: NIST database
p+ Beam Range = 15 cm
Integral Dose
Integral Dose § The V40% for protons is smaller
than photons
§ Due to the incorporation of uncertainties in planning, the conformality is tighter with photons for most SRS targets
§ Abnormally shaped targets or targets close to an OAR can have tighter conformality
§ Clinical Significance?
6/5/15
20
Proton SRS for Benign Cases: Secondary Cancer Risks
Risk of 2nd cancer Clinical symptomsEUD (Gy) NTCP (%) EUD (Gy) NTCP (%) NTCP (%)
SRT 32.1 23 <0.1 28 <0.1 <0.12-field photon 5.7 48 1.3 48 1.2 133-field photon 11.2 38 <0.1 40 0.1 2IMRT 26.8 34 <0.1 37 <0.1 12-field proton 1.5 30 <0.1 35 <0.1 <0.13-field proton 4.3 29 <0.1 35 <0.1 <0.14-field proton 6.1 27 <0.1 34 <0.1 <0.15-field proton 6.8 26 <0.1 34 <0.1 <0.1
Right temporal lobe Left temporal lobeRadiographic changes
Acoustic àà Sarcomatous Hanabusa, 2001 Acoustic àà Glioblastoma Shamisa, 2001 AVM àà Glioblastoma Kaido, 2001 Acoustic àà Meningiosarcoma Thomsen, 2000 NF2 àà Malig n. sheath (3 cases) Baser, 2000 NF2 àà malignant meningioma Baser, 2000 NF2 àà Malignant ependymoma Baser, 2000 Mening àà Glioblastoma Yu, 2000 Acoustic àà Malig Schwannoma Shih, 2000 Cav hem àà Glioblastoma Salvati, 2003 Acromeg àà Meningioma Loeffler, 2003 Acromeg ààVestibular Schwannoma Loeffler, 2003 AVM àà Meningioma Sheehan 2006 Many more studies…
Winkfield, et al, 2011
Triage: Which Patients to Treat
§ Benign Neoplasms: § Acoustic Neuromas § Meningiomas § Pituitary Adenomas
§ Arteriovenous Malformations § Metastatic Lesions
§ Multiple Lesions § Close proximity to surface or critical
structures (optics, brainstem)
PROTON SRS CASES Caseload
28%
11%
15%11%
33%
2%
AVMAcousticPituitaryMeningiomaOtherExtracranial
6/5/15
21
Proton SRS Examples
§ Pituitary Adenoma § Atypical Meningioma (multiple sites) § Single Meningioma § Acoustic Neuroma § Arteriovenous Malformation § Large Met Near Optics § Multiple Mets § Comparisons to Photons are only for Linac
Pituitary Adenomas, 18-20Gy
Endocr Pract 2007 Nov-Dec 13(7)
Proton Stereotactic radiosurgery in management of persistent acromegaly
Petit, Biller, Coen, Swearingen, Ancukiewicz, Bussiere, Chapman, Klibanski, Loeffler
Clin Neurosurg 2008; 55
Management of recurrent and refractory Cushing’s disease with reoperation and/or proton beam radiosurgery
Aghi, Petit, Chapman, Loeffler, Klibanski, Biller, Swearingen
Patient 1 – Pituitary Adenoma
Patient 2 – multiple atypical meningioma
Protons: 10 fields
X-Rays: 12 dynamic arcs
Atypical Meningioma, 12-18 Gy
6/5/15
22
Protons X-Rays
Patient 2 – multiple atypical meningioma
Atypical Meningioma, 12-18 Gy
Meningioma, 12Gy
Int J Radiat Oncol Biol Phys 2011 Dec 1; 81(5)
Proton stereotactic radiosurgery for the treatment od benign meningiomas
Halasz, Bussiere, Dennis, Niemierko, Chapman, Loeffler, Shih
Patient 3 – meningioma
Protons X-Rays
Patient 4 –meningioma
Meningioma, 12Gy
6/5/15
23
Patient 5 – 4.2 cc acoustic neuroma ( 30 x 180 cGy)
Protons: 4 fields 6 MV X-Rays: 3 dynamic arcs
Acoustic Neuroma, 54Gy in 30 fx
Patient 5 – 4.2 cc acoustic neuroma ( 30 x 180 cGy) Protons X-Rays
Acoustic Neuroma, 54Gy in 30 fx
Acoustic Neuroma, 12Gy in 1 fx
Int J Radiat Oncol Biol Phys 2002 Sep 1; 54(1)
Proton beam stereotactic radiosurgery of vestibular schwannomas
Harsh, Thornton, Chapman, Bussiere, Rabinov, Loeffler
Neurosurgery 2003 Sep; 53(3)
Proton beam radiosurgery for vesibular schwannoma: tumor control and cranial nerve toxicity
Weber, Chan, Bussiere, Harsh, Ancukiewicz, Barker, Thornton, Martuza, Nadol, Chapman, Loeffler
Patient 6 – Acoustic Neuroma (Single Fx)
6/5/15
24
AVM, 12-16Gy in 1 or 2 fx’s
J Neurosurg 2003 Aug 99(2)
Dose-volume prediction of radiation-related complications after proton beam radiosurgery for cerebral arteriovenous malformations
Barker, Butler, Lyons, Cascio, Ogilvy, Loeffler, Chapman
Int J Radiat Oncol Biol Phys 2012 Jun 1; 83(2)
Planned two-fraction proton beam stereotactic radiosurgery for high-risk inoperable cerebral arteriovenous malformations
Hattangadi, Chapman, Bussiere, Nimierko, Ogilvy, Rowell, Daartz, Loeffler, Shih
Astro Abstract 2012
Proton Beam Stereotactic Radiosurgery for inoperable cerebral ateriovenous malformations.
Hattangadi
Patient 7 – Arteriovenous Malformation
Patient 8 – multiple mets
Protons: 5 fields
X-Rays: 11 arcs
Brain Mets, All, 18Gy in 1 fx
Protons X-Rays
Patient 8 – 4 mets
Brain Metastasis, 18Gy in 1 fx
6/5/15
25
Patient 9 – 5 cc met close to optics
Protons: 3 fields 6 MV X-Rays: 6 dynamic arcs
Large Met, 4x 5 Gy
Protons X-Rays
Patient 9 – 5.0 cc met close to optics
Large Met, 4x 5 Gy
Brain Metastasis, 18Gy in 1 fx
Patient 10: Met Close to Optics
6/5/15
26
Scanning Example
Eyes
§ Excellent local control, especially when tumors are too deep for brachytherapy
§ Slightly higher vision preservation
Eyes
§ Many ocular proton therapy patients
§ Requires lower energies and smaller accelerators
§ Many centers use EyePlan
§ Discussions regarding the future of ocular treatment planning
6/5/15
27
Ocular Therapy
§ Tumor dimensions obtained by ultrasound and ophthalmoscopy
§ Some centers evaluating CT and MR
§ Localized based upon anatomic landmarks (ex. optic disk or limbus)
§ Dimensions of the eye
also recorded
Ocular Therapy
§ Radio-opaque clips are attached to the outside of the eye in close proximity to the target
§ The locations are documented
Ocular Therapy
§ Dimensions transferred to EyePlan
§ Margins are added (2-4 mm)
§ Aperture and R/M
calculated
6/5/15
28
Ocular Therapy
§ Treatment setup is with clips and radiographs
§ Monitor the retina with video
Ocular Therapy
§ Measurements: Typically with micro ICs, diodes, and MLICs
§ Minimal field size effects due to shallow ranges
§ Diodes for small fields and cross calibrated at larger fields
Ocular Therapy
§ CT based treatment planning
§ Less special hardware
§ Routine TPS
§ Challenges: Reproducibility, eye positioning, intra-fx monitoring, beam commissioning, small range (absorbers)
6/5/15
29
Onward… § More Robust Treatment Planning
§ Better Imaging: § Real Time
§ Proton Range Information
§ Better probabilistic models
§ Clinical Trials
Robustness
§ Include probability estimates in the treatment planning optimization
§ Reduce high gradients in close proximity to OARs
§ Include Range Uncertainties, Setup Uncertainties, and Motion
= + +
5 mm overshoot
C2 chordoma: Rx = 77.4 Gy(RBE)
Trofimov et al
6/5/15
30
Robust optimization (illustration)
robust IMPT Standard IMPT
(Trofimov et al 2012)
Robust optimization (illustration)
no-uncertainty plan § high gradients
robust against range errors
§ reduced gradients
Unkelbach et al 2009
Imaging
§ IBA CBCT
§ MedPhoton
§ 2D/3D
6/5/15
31
PET Range Verification
0 50 100 150 200 250 3000
200
400
600
800
1000
Act
ivity
(Bq/
ml)
Depth (mm)
MC_PA MC_LL MC_PA+LL NeuroPET
Protons
Beam Range (155 mm)
Protons activate carbon and oxygen nuclei that decay by positron emission.
Chul Hee Min, MGH
Summary § Small Fields can be useful for multiple
treatment sites in proton therapy
§ Challenges remain for accurate measurements
§ Proton SRS is a viable option for cranial SRS
§ Benign cases probably have the most benefits with protons à Integral Dose
§ Currently, less conformal due to uncertainties: § Online range verification § Robust planning
§ Patient Imaging
§ Ocular Proton Therapy has high LC § Challenges of TPS and clinical deployment
Thank You!
http://gray.mgh.harvard.edu