E. Pedroni Paul Scherrer Institute SWITZERLAND PSI...
Transcript of E. Pedroni Paul Scherrer Institute SWITZERLAND PSI...
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
PSI
25.01.2009
Session: delivery systems and gantries
E. Pedroni
Paul Scherrer Institute
SWITZERLAND
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
First presentation: Scholz GSI
• Biological treatment planning (Local Effect Model L EM)
• Conformal scanning provides variable modulation of the range
– this gives laterally-varying mixtures of LET and fragmentation depth-profiles
– Need to adjust physical dose to obtain an homogeneous equivalent dose
• Model constituents
– Alpha-beta model response to X-ray doses
– Statistical fluctuations of the microscopic dose around the single ion tracks
– Average of dose effect give the effective RBE at a point
– Biological data were presented, which confirm the method
• The LEM model is a big step forward for ion therapy
– The established standard for ion therapy in the context of scanning beams
– With scanning one can obtain a more homogeneous equivalent dose than with scattering
• Why not starting using this model for protons as well ?
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
• Ballistic properties of ions are superior to protons• Comment: for discussion later
– Gain in healthy tissue sparing (in terms of dose distribution)
• Step going from photons-to-protons = factor 2.5
• Step going from protons-to-carbon = factor 1.2-1.3
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
Second presentation: Pedroni PSI
In room
Sliding CT
New proton
scanning
gantry for
treating
moving targets
New gantry 2
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
• Second generation scanning gantry – based on Gantry 1 ex perience
– Combination of proton treatments with a sliding CT
• For image guided proton therapy
– Option to use BEV X-ray simultaneous to proton beam delivery
• For QA of moving targets
– System designed for developing very fast dose painting
• Spots – lines –contours ( > 1 cm/ms)
• Repainting and gating
– Fast energy changes (80 ms)
– Fast 2d parallel scanning
– Simulated parallel scattering with variable range modulation
– The complement to this talk was given in another session by David Meer
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
• Link to other sessions on scanning advancements
– An excellent presentations on scanning and scattering was given by J. Flanz
– Impressive: the work on advanced developments on scanning with the NIRS synchrotron
in Chiba by Dr. Noda
• Synchrotrons can provide complex delivery patterns during slow extraction
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
45° dipoles
90° dipole
scannermagnets
treatmentroom
absorber
Third presentation: Weinrich GSI Gantry optics
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
The new point of reference for ion gantries
• Ion optical properties of the ion gantry at GSI
– The first ion gantry of the world – quite impressive to see the system rotating (video)
• Major problem encountered: cabling drum (a solvable problem)
• Description of the ion optical parameters
– Tolerances were very stringent - what has been achieved is probably OK
– Coupling of vertical and horizontal plane in order to achieve a round beam at the
isocenter
• By equalizing the divergence of the beam at gantry entrance
– Corrections as a function of the gantry angle
– Parameterization for 5 different spot sizes
– Commissioning work: A lot of work – in principle: no essential differences to the
commissioning work for a proton gantry
• Link:
– the very informative HIT presentation of T. Haberer in another session
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
Fourth presentation: Trbojevic BNL
FFAG permanent magnets gantry
Workshop on Hadron Beam Therapy of Cancer, Erice 9
78o
r=2.71 m
h1=1.58 m
13 cells - 25 cells
150o
h2=2.42 m
Orbits magnified10 times
From a density of the No-Fe-B 11.7 gr/cm3
The weight of the whole gantry ~ 500 kg.
(Eberhard Keil)
h3=4.29 m
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
• FFAG light weight gantry – static solution – no power consumption– Only 500 kg structure (effective bending radius of 1.7 nm) for protons
– Alternating permanent magnets (Halbach round cells of 18 cm)
– Repeated very strong focusing over very short distances (60 cm)
• A gantry dealing with any energy [70, 250 MeV] • and with any time structures of the injected beam
– Scanning can be made very fast in all 3 dimensions;
• in range and laterally at wish
• Ideal to track organ motion laterally and in range - equally quick
• Solution for ions (superconducting windings) – superconducting windings around a pipe – homogeneous resultant field
– 1500 kg
• Pictorial description– The “fiberglass cable” for charged particle beam?
• When the first magnet prototype?
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
• A New Tracking Gantry-Synchrotron Idea
– Basis of the new idea is use of the following:
• simple, synchrotron and gantry magnet lattices (small aperture)
• series connection of magnets for 5 Hz tracking
• one power supply for main ring/gantry magnets
– Scanning idea (H and C)
• Scan full length tumour on a line in depth at each 5 Hz cycle
– stripping foils for wide energy-range extraction ( I(t) within a pulse)
• Extract slowly beam while accelerating
• the full beam current is delivered in every scanning cycle
• Transverse motion and depth-scans in subsequent cycles
• Fast scans for least effect of tumour-movement (10x10x10 in 20 s) 80s?
• Scanning started upstream of triplet?
Fifth presentation: Rees, RAL, UK.
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
Conceptual gantry design for C 6+& H+ ions
(300°)
~10 m
H+
C6+
small aperture
low rf volts
5 Hz rings
foils
~5 m
2 nA (aver), 250 MeV H +
0.1 pnA, 400 MeV/u C 6+
10 x 10 x 10 cm 3 tumourscanned in ~20 secondswith a dose of 2.5 Grays(for one gantry position)
C = 43.68 m & 49.92 m
H¯
C4+
(small apertures)
(elliptically shaped central support )
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
“CONTROVERSIAL POINTS” FOR DISCUSSION
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
TOPIC 1: TYPE OF PARTICLE
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
Rationale for using protons
• Low LET
– Same radiobiology as photons (fractionation to protect vital tissues within the target)
– Good for treating large targets
• Advantage: avoid the dose bath of the photons outside the target …
– Not so good for small tumors (AVMs > 2-3 cm)
• Due to MCS in the Bragg peak –> less good lateral dose fall-off
– Carbon has a better ballistic property
• Needs to be compared with Gamma knife etc.
• The decision to use protons can be decided solely on the comparison of the protons- with the photons dose distribution pla ns
– Decision within treatment planning
• In principle: protons always superior to photons (a question of costs)
– Gain: lower integral dose outside the target: by a factor of 2 to 5
• A question of cure, complications (and quality of life after surviving cancer)
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
Rationale for using carbon ions
• Arguments in favor of carbon
– The high LET of heavy ions should be an advantage for treating
• Radio-resistant tumors
• Poorly oxygenated tumors (OER oxygen enhancement ratio)
– Better precision of the beam - Less MCS - Sharper Bragg peaks (He – Li –C )
• But with fractionation tails – Dosage uncertainties due to the varying RBE (LEM)
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
X-rays
neutrons
X-rays
neutrons
• Difference between high-LET and low-LET is expressed at low doses
– At high doses the differences tend to disappear
– Should one then uses carbon with low doses (high fractionation)?
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
Possible disadvantages of high LET
• Too simplistic
– Carbon is better because is “more effective” (higher RBE)
• What matters is the ratio of surviving healthy cells to tumor cells (target and plateau)
– Therapeutic ratio T
– Potentiation with fractionation Tn use protons if T> 1 (within target)
• High LET can be a contra-indication
– Absence of repair of radiation damage in the healthy tissues within the target
• important for large tumors (cases infiltrating vital structures)
– Risk of late effects (important for pediatric treatments)
• Poor experience with neutrons (and pions) in the 80’s
– Bladder shrinking after 2 years pion therapy delivery at PSI
• Higher magnetic rigidity of the beam (x3)
– Size of facility doubled -> Costs doubled (huge gantry or no gantry)
– Needs substantial investments in radiobiology (should be the main interest)
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
Controversial – a too simple statement
• Carbon therapy “needs” few fractions – is therefore less expensive
– Nothing forbids to deliver protons with a low number of (high dose) fractions
• In fact
– Good experience with small tumors
• radio-surgery
• eye treatments (melanomas)
• At a very high dose per fraction protons behave similarly to high-LET radiation
• Repair of radiation damage in the cell is suppressed
• Double Strand Breaks gets enhanced
– If we deliver carbon ions with only a few fractions of high doses we loose the beneficial
effect of having low LET in the plateau region of the beam
• We shoudl not sell the argument twice
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
Research issues for ion therapy
• Boost therapy?
– Carbon boost followed by photons (IMRT) ?
– Why not carbon ions followed by proton therapy? (the best possible boost treatments)
• But HIMAC has shown very good results with carbon ion s
– With hypo-fractionation (lung tumors)
• Comparison with protons is however missing
• May by there is really something new there - when using high LET with ions
• Helium
– As the low-LET radiation source for highest precision?
• We have learned from A. Brahme
– Today we have to look also at the new molecular radiobiology endpoints
• Lithium and others light ions – best for apoptosis within the Bragg peak region
• We certainly need much more research with ions and protons!
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
TOPIC 2: GANTRY or NO GANTRY?
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
The rationale for using a gantry ?
• To apply several fields (“beam incidences”) typically 3 (1-3)
– To distribute the plateau dose over several tissues (in the same session)
– To stay below organ tolerance
• Due to the Bragg peak protons need a lower number of fields as photons (3-9)
• With the patient treated in supine position
– To keep the position of internal organs unchanged (soft tissues in the body)
• Same position as at the time of CT-data-taking for treatment planning
– For best comfort of the patient
• To offer maximum flexibility of choosing the beam di rections
– Avoid beam going through sensitive organs
– Avoid beam going through complex density heterogeneities in the patient body
• Dose errors due to interplay of MCS and range
– Shadows at interfaces parallel to the beam (bones and metal implants)
– Select possibly angles with low ranges (keep the energy – plateau dose length - small)
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
• Disadvantages ?
– None - A gantry can do more than an horizontal beam line
– With a gantry we can deliver many fields (2-3) in sequence without entering the room
(and without changing the room)
• But it implies additional costs
• For protons: the established standard solution
– Which is the proper mixture of gantries and horizontal beam lines?
– How many horizontal beam lines do we need, if we don’t treat mainly prostate? eyes?
– If we choose a few fixed angles
• How severe are the logistics problems?
• For carbon facilities
– If too bulky…
– Why not having few dedicated proton gantries in the facility?
– To avoid suboptimal treatments with protons (Medaustron poster)
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
• If you use a gantry;
– Don’t forget space for new future in-room diagnostics
– Leave enough space in the room to adapt your system to the future developments
coming from conventional therapy
• CT
• CT-PET?
• Out-of-room positioning with MRI?
– (Position reproducibility?)
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
• If you use a gantry for carbon ions and protons togethe r
– You should optimize the nozzle first for the protons
• The most sensitive in terms of beam quality (MCS)
– To obtain a small enough beam size
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
TOPIC 3: SCANNING
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
Why scanning ?
• Best way to automates treatment
– Multiple fields in one go (avoid changes of individualized hardware) (gantry)
• Flexibility to deliver non-homogeneous dose distribut ions
– IMPT ( = simultaneous optimization of dose fields) – compete with IMRT
– Biological targeting (image guided proton-ion therapy)
• Variable modulation of the range (conformity)• Best use of the beam
– Less neutrons (~ factor of 5?)
– Less activation
• Better dose fall-off (in theory)
– Dose edge enhancement with sparse scanning
– Collimation as an option
• The only major problem ; organ motion sensitivity
– Fast scanning helps
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
Do we need to have scattering as well?
• Dose precision
– Scanning alone is superior for most deep seated tumors
– Scattering is in practice better for lower ranges (but not necessarily simpler – Optis 2)
– Scanning with optional collimation is superior everywhere
• More equivalent to single scattering than double scattering
• Best strategy?
– Build facility for scattering and then add scanning …
• Doubling of the efforts …
– Or rather build a facility for scanning and then just simulate scattering …?
• Simulated scattering (very fast - uniform scanning)
– With high repainting number
– Parallel scattering no dose errors from compensators
– Variable modulation of the range shrinking shape of iso-energy-layers
– Less neutron dose
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
Organ motion errors with scanning
• Disturbance of the lateral dose fall-off(same problem scattering and scanning)
• Add safety margins
• Reduce with Gating or Tracking
• Disturbance of the dose homogeneity
– Scattering – highly repainted - insensitive
– Single painted scanning - very sensitive
– Repainted scanning
• Alone – for medium motion
• With Gating or Tracking – for large motion
• The experience of treating moving targets with scanning is still inexistent
– WE HAVE TO LEARN HOW TO COPE WITH
THAT
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
Choice of delivery techniques matters
Fast energy change Slow energy change
- G2 lines, scaled - G2 spots, iso-layer - G2 spots, scaled - G1 spots, scaled
• data points: Median of the RMS spectrum• shaded area upper boundary: 75 % of the dose distributions are above.• shaded area lower boundary: 25 % of the dose distributions are below.
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
Step 1 - Repainting (repainting strategies)
• Scaled repainting
• Repeat same scan n times with dose divided by n
• Layered repainting
• Deliver partial scans with constant max. partial dose
• Revisit only those spots which need more dose
• Using spots or lines or contours
beam
0 % missing dose25 % missing dose
75 % missing dose50 % missing dose
100 % missing dose
Break possible synchronism of scanning and respiration periods
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
Step 2 - Gating + Repainting
• Treatment of moving targets
– Which gating signal should we use?
• breath holding ?
• external signal (like thoracic strip ?)
• 3-d stereo viewing of external marks on patient skin?
• Whole energy layer delivered within a gate? (200 ms pe r plane)
Pla
n1
Beam
Target at times ti
Iso-energylayers E at ti
E
Residual motion?
Desynchronize?
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
Step 3 - Tracking?
Nodule
Rib
Lung
• Lateral shift applied to the beam as a function of t he target displacement
– Given for example by BEV X-ray images
• Problem:
– Density heterogeneities in the beam path
• example bone in front of a lung tumor
– Does tracking need the inclusion of dynamic range corrections?
• Probably - See thesis work of C. Bert at GSI
Beam
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
Step 3.bis - Why not “Multi-gating”?
• Associate several time- gates with corresponding gated-CT plans – Deliver multiple time-instances treatment plans simultaneously
• Full energy layer of a plan is delivered within its ow n gate– A better alternative to tracking?
• No need for energy corrections - No target motion-deformation model needed
• Repainted by definition – Fast scanning - Good duty factor of using of the beam
t1 t2
Pla
n1
Pla
n3
Pla
n2
Beam
Target at times ti
Iso-energylayers E at ti
E
t3
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
• If the previous methods don’t work …
– Use scattering as a backup solution for treating moving organs
• Simulation on the scanning nozzle
• Fast uniform scanning is at the moment an important issue (“safety net”)
Step 0 - Simulated scattering
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
TOPIC 4: ACCELERATOR TYPE
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
Specifications for the accelerator ?
• Small pencil beam (larger beam are easy to produce)– Good lateral and distal dose fall-off – 3 mm sigma
– Absolute beam positioning precision better than 1 mm
• DC beam– Ideally: 100% duty factor for (repainted) scanning
• 2d - fast magnetic scanning– Motion faster than 1cm / ms
• Continuous fast scanning with dynamic beam intensit y modulation?– For 1cm wide pulsed beam - corresponds to a repetition rate of 1 KHz
• With a few % precision of the control of the dose per pulse
• Dose control precision per spot– 1 % spot dose precision of average spot 0.2 % for absolute dosimetry
– a challenge for pulsed systems? (Control of the spot dose within the pulse)
• Dose spot dynamics for conformal therapy: 30:1 (dista l proximal)– Within a layer of a conformal treatment we have a non-homogeneous proton fluence
Really Needed?
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
• Fast energy changes? How much?
– For volumetric repainting (necessary?)
• Sychrotrons – 1 s
• Cyclotron + degrader + gantry - 80 ms (gantry limited)
• Cyclinac - 1ms
– But gantry limited (if one uses a gantry) … unless
• Gantry with a large dp / p acceptance (U. Amaldi - good for tracking)
• FFAG – energy pulse by pulse – not gantry limited (if a FFAG gantry)
• There is no experience up to yet in treating moving t argets with beam scanning – we have to work on it
• The new dream machines?
– Dielectric wall accelerator (a few meters proton linac) could be the revolutionary invention
for the whole field – and steer radiotherapy away from photons to protons in general
– Laser based acceleration – probably a longer perspective ?
E. Pedroni CPT - Paul Scherrer Institute - Erice 20 -04-2009
• THANK YOU
A very exciting field also in the future…
Based on a beautiful idea