Treatment Planning and ProtonsTreatment Planning and Protons

Radhe Mohan PhDRadhe Mohan, PhD

1

SWAAPM Meeting, Houston, TXOctober 9 - 10, 2009

2

The more you learn, the more you realizeThe more you learn, the more you realize how little you know …

Attributed to Socrates by Platoy

3

OutlineOutline

Characteristics of protonsCharacteristics of protonsDose calculationsUncertaintiesPassively scattered protons strategiesy p gScanning beam and IMPT strategiesLimitation of the concept of PTVLimitation of the concept of PTVPlan robustnessSome clinical resultsProton therapy research at MDACC

4

py

Characteristics of ProtonsCharacteristics of Protons

Loose energyPristine Bragg peak due Loose energy continuouslyDeposit less energy

Pristine Bragg peak due to a pencil beam of

protons in homogeneous Deposit less energy per unit pathlength in plateau and more

p gphantom

p ateau a d o enear the end of the range

Dos

e

Have a finite range –they stop

D

5Depth

To Make Protons Useful for Radiotherapy …

Thin pencil beams must be spread laterally and longitudinally

The Passive Scattering Mode of Proton Beam DeliveryBeam Delivery

7

Dose Calculations for Passively Scattered Proton Therapy

SOBP is decomposed into pencils and

Proton Therapy

SOBP is decomposed into pencils and layers of energyIntegral of each pencil across a planeIntegral of each pencil across a plane for each energy is its depth doseLateral dose distribution approximated by multiple Gaussians to account for MCS, nuclear i i

Xinteractions, etc.Ray tracing used to approximately account for inhomogeneities

2400

2600

2800

2400

2600

2800

account for inhomogeneitiesEmpirical corrections and fitting of parameters leads to acceptable 1600

1800

2000

2200

1600

1800

2000

2200

8

parameters leads to acceptable accuracy in most situations

50 100 150 200 250 300

1200

1400

Y distance [pixels]50 100 150 200 250 300

1200

1400

Y distance [pixels]

Uncertainties in Proton TherapyUncertainties in Proton Therapy

Inter-and intra-fractional variationsInter-and intra-fractional variationsDosimetric : CT numbers, high-Z artifacts,

i t t i dconversion to stopping powers, dose computation approximationsAssumption of Radiobiological Effectiveness (RBE) to be 1.1…

9

Proton Therapy of Moving Lung TumorsProton Therapy of Moving Lung Tumors

Prescription Dose Line

Treatment planned based on single free-breathing CT

The same treatment plan calculated on 10 phases of

10

image (perceived dose distribution)

pthe 4D CT image

Dong

Inter-Fractional Variations

IMRT: 7 beams plan

Original Plan After Two weeks of Radiotherapy

p

Proton: 3 beams planProton: 3 beams plan

11Zhang

Vulnerability of Protons to UncertaintiesVulnerability of Protons to Uncertainties

Protons more sensitive to uncertainties thanProtons more sensitive to uncertainties than photonsI t i d l (i th 1980 )Importance recognized early (in the 1980s)

Limited by tools by today’s standard (repeat CTs, 4D CT) t bt i tit ti i f ti b t4D CT) to obtain quantitative information about the consequences and find solutions

Rudimentary early solutions were developedRudimentary early solutions were developed …

12

Accounting for UncertaintiesAccounting for Uncertainties

Margins, Apertures and g , pCompensators

13

Beam Specific Margins for UncertaintiesBeam-Specific Margins for Uncertainties -

Proximal Margin

Protons

Proximal Margin

CTV Lateral Margin =Lateral Margin PTV Margin

14

Distal Margin

Compensator “Smearing” to Account for Lateral P iti Ch i I h itiPosition Changes in Inhomogeneities

15 Urie, et al, Phys. Med. Biol., 1983.

Scanning Beams and Intensity Modulated Proton TherapyIntensity Modulated Proton Therapy

16

Scanning Beams and IMPTScanning Beams and IMPTBeam 174 MeV

3.2mBeam Profile Monitor

Beamlet

Scanning Magnets

Helium Chamber

S t P iti M itSpot Position MonitorDose Monitor 1, 2 20 cm

17

Isocenterσ = 0.8 cm

Scanning Beams and IMPTScanning Beams and IMPT

Range of energies (72 – 221 MeV for Hitachi)Range of energies (72 – 221 MeV for Hitachi)Spots positioned by magnetic steeringDiscrete spot scanning or raster scanning Spots from placed in layers in the target for p p y geach of multiple fieldsSpot intensities determined usingSpot intensities determined using optimization techniques

18

Layer-by-Layer Scanning of Proton Beamlets for IMPTProton Beamlets for IMPT

19 Spot positionMovie.10Movie.10 Cumulative dosedoseMovie.9Movie.9

Dose Calculations for Intensity Modulated Proton TherapyProton Therapy

Beamlets are assumed to be GaussianBeamlets are assumed to be GaussianApproximations related to

Scattering from beam line componentsNuclear interactions

Contributions from thousands of “spots” of specified energies and intensities constituting the scanned beam are summedMinor inaccuracies can add up to significant p gdiscrepancies

20

Intensity Modulated Proton TherapyIntensity Modulated Proton Therapy

Three possible IMPT optimization strategiesThree possible IMPT optimization strategiesSingle field uniform dose (SFUD)

Robustness similar to PSPTRobustness similar to PSPT

3D IMPTIn-between robustnessIn between robustness

Distal edge trackingMost vulnerable to uncertaintiesost u e ab e to u ce ta t es

No need for compensatorNormally no apertureNormally, no aperture

21

Accounting for Uncertainties in IMPTAccounting for Uncertainties in IMPT

No such thing as smearing - consequences ofNo such thing as smearing - consequences of lateral variations in inhomogeneities ignoredF SFUD di t l d i l iFor SFUD, distal and proximal margins = those for PSPTFor 3D IMPT and DET, no current solutionFuture solution- Robust optimizationp

22

Evaluation of Proton Plans

23

The PTV Concept is Not Valid for ProtonsThe PTV Concept is Not Valid for Protons

Main reason: Planning margins to CTV areMain reason: Planning margins to CTV are different in different directions and for different beamsdifferent beamsThen how do we

Evaluate plans using DVHsCompare proton plans with photon plansAdd proton dose distributions to photon dose distributions

24

PTV Concept is Not Adequate Even for PhotonsEven for Photons

PTV

CTVCTV

DVH for PTV represents the “worst” case scenarioDVH for CTV represents the “best” case scenariao

25

p

Levegrun et al, IJROPB 2000: No prostate dose response

One Solution

“Uncertainty” (or “Worst Case”) Analysis

26

Parallel Opposed Prostate – TransversePrescription (75 6 CGE) Isodose

5 mm Post.Prescription (75.6 CGE) Isodose

Range –3.5%g

Range +3.5%

Nominal

275 mm Ant.

Some Clinical ResultsSome Clinical Results

28

Planning Study (23 patients) to Design Randomized PSPT vs IMRT NSCLC TrialRandomized PSPT vs. IMRT NSCLC Trial

0 001

(%)

Aut oPl an Dosi mpl an Pr ot onpl an

0 820. 00

0. 0790. 015

0. 0010. 001

0. 230. 4010. 0010. 0

0. 00030. 005

60708090

100

ativ

e Vo

lume

0. 9

0. 820. 450. 079

0. 0

0. 4010. 069

01020304050

[Gy]

or

Rela

0

PTVD

at95

%V

PTVD

2cc

Lung

Vat 5

Gy

Lung

Vat 2

0Gy

MLD

Cor d

Dose

at1%

V

EsoD

1/3G

y

Hear

t D1/

3Dose

Dose-volume indices are averaged over 23 patient

29

D-V Indices of the First 5 Patients Enrolled on R d i d PSPT IMRT NSCLC T i lRandomized PSPT vs. IMRT NSCLC Trial

8000

Aut oPl an Cl i ni cal Phot on Cl i ni cal Pr ot on

6000

7000

3000

4000

5000

Dose

[cG

y]

1000

2000

3000Do

0

1000

PTVD95 LungV20 MLD Cor dD1% EsoD1/ 3 Hear t D1/ 3

30Dose-volume indices are averaged over 5 patient

Proton Therapy vs. 3D CRT & IMRT of Locally Advanced NSCLCLocally Advanced NSCLC

Median Total DoseMedian Total Dose3D CRT & IMRT: 63 GyProton therapy: 74 CGEProton therapy: 74 CGE

Esophagitis Gr ≥ 3 Tx Related Pneumonitis Gr ≥ 3

40.00%45.00%

Esophagitis Gr ≥ 3

30 00%

35.00%

Tx Related Pneumonitis Gr ≥ 3

3D CRTIMRTP t

25.00%30.00%35.00%

20.00%25.00%30.00%

Protons

10.00%15.00%20.00%

5 00%

10.00%15.00%

310.00%5.00%

0.00%5.00%

Reduction of Bone Marrow Toxicity inReduction of Bone Marrow Toxicity inReduction of Bone Marrow Toxicity in Reduction of Bone Marrow Toxicity in Patients with LA NSCLC treated with Patients with LA NSCLC treated with Concurrent ChT/ PBT compared to IMRTConcurrent ChT/ PBT compared to IMRTConcurrent ChT/ PBT compared to IMRTConcurrent ChT/ PBT compared to IMRT

Ritsuko Komaki,M.D,

32

Tumor Volume and BM toxicityTumor Volume and BM toxicity

GTV (cc)GTV (cc) NN % of % of >>2 Grade 2 Grade Hemoglobin Hemoglobin

EventsEvents

P valueP value % of % of >>2 Grade 2 Grade PlateletsPlateletsEventsEvents

P valueP value

(%)(%) (%)(%)

>>100100 PBT = 19PBT = 19IMRT= 64IMRT= 64

1 (5%) 1 (5%) 24 (38%)24 (38%)

<0.001<0.001 1 (5%)1 (5%)3 (5%)3 (5%)

0.0290.029IMRT= 64 IMRT= 64 24 (38%) 24 (38%) 3 (5%)3 (5%)

>>200200 PBT = 7PBT = 7IMR 43IMR 43

0 (0%)0 (0%)19 (44)19 (44)

0.0010.001 1 (14%)1 (14%)2 (5%)2 (5%)

0.0560.056IMR= 43IMR= 43 19 (44)19 (44) 2 (5%)2 (5%)

>>300300 PBT = 5PBT = 5 0 (0%)0 (0%) 0.720.72 0 (0%)0 (0%) 0.3070.307IMR= 30IMR= 30 14 (47)14 (47) 2 (7%)2 (7%)

>>400400 PBT = 3PBT = 3 0 (0%)0 (0%) 0.1980.198 0 (0%)0 (0%) 0.7540.754IMR= 26IMR= 26

( )( )7 (27%)7 (27%)

( )( )2 (8%)2 (8%)

>>500500 PBT = 2PBT = 2 0 (0%)0 (0%) 0.3960.396 0 (0%)0 (0%) 0.3660.366

33

>>500500 PBT 2PBT 2IMR = 19IMR = 19

0 (0%)0 (0%)7 (37%)7 (37%)

0.3960.396 0 (0%)0 (0%)1 (5%)1 (5%)

0.3660.366

Tumor Volume and BM toxicityTumor Volume and BM toxicity

GTV GTV (cc)(cc)

NN % of % of >>2 2 Grade Grade

NeutrophilNeutrophil

P P valuevalue

% of % of >>2 2 Grade Grade WBCWBC

P P valuevalue

% of % of >>2 2 Grade Abs. Grade Abs.

LymphocyteLymphocyte

P valueP value

Neutrophil Neutrophil EventsEvents

(%)(%)

WBCWBCEvents Events

(%)(%)

Lymphocyte Lymphocyte EventsEvents

(%)(%)

>>100100 PBT = 19PBT = 19IMRT=62 IMRT=62

554444

0.0280.028 26266767

0.0140.014 1001009797

<0.001<0.001

>>200200 PBT = 7PBT = 7IMRT= 42IMRT= 42

004343

0.270.27 14146363

0.1820.182 1001009898

0.0010.001

>>300300 PBT = 5PBT = 5IMRT = 29IMRT = 29

004848

0.3280.328 20206363

0.2740.274 1001009696

<0.006<0.006

>>400400 PBT = 3PBT = 3IMRT= 25IMRT= 25

004040

0.7040.704 33335757

0.4280.428 1001009696

0.6120.612

34>>500500 PBT = 2PBT = 2

IMRT = 18IMRT = 1800

33330.8130.813 5050

44440.5790.579 100100

1001000.8810.881

Current Research at MDACC (In Collaboration with MGH)(In Collaboration with MGH)

Reduction in uncertainties andReduction in uncertainties and accounting of residual uncertainties

Stopping powers improvementStopping powers improvementGating and breath-holdAdaptive replanningp p gDose calculation accuracy

IMPT developmentpRobust optimizationEvaluation of robustness of treatmentEvaluation of robustness of treatment plansDose accumulation

35

Dose accumulationTrials and studies

SummarySummary

Abundant room for improvement ofAbundant room for improvement of efficiency, cost-effectiveness, accuracy and quality of treatmentsand quality of treatmentsA boon and a bain for physicistsEven in the current state of the art PT has the potential to be significantly superior for certain sites

LungPediatrics…

36Future potential is great

However, Be Mindful of Irrational Exuberance LestExuberance Lest …

37

Parachute use to prevent death and major trauma l t d t it ti l h ll t ti i frelated to gravitational challenge: systematic review of

randomised controlled trials – Smith et al, BMJ 2003;327:1459-14612003;327:1459-1461

Conclusion:As with many interventions intended to prevent ill health, the effectiveness of parachutes has not been subjected to rigorous evaluation by usingbeen subjected to rigorous evaluation by using randomised controlled trials. Advocates of evidence based medicine have criticised the adoption of interventions evaluated by using only observational data. We think that everyone might benefit if the mostWe think that everyone might benefit if the most radical protagonists of evidence based medicine organised and participated in a double blind,

38

randomised, placebo controlled, crossover trial of the parachute.

Thank YouThank You

39