Treatment delivery from the patient perspective: potential indication, set-up and organ motion issues at CNAO

1
Treatment delivery from the patient perspective
potential indication, set-up and organ motion
issues at CNAO

• Piero Fossati

ESF Exploratory Workshop on Advanced
Instrumentation for Cancer Diagnosis and
Treatment Oxford, United Kingdom, 23 26
September 2008
2
We all know how nice it is to have the Bragg peak
Goitein et al, Physics today, 2002
Courtesy of Dr. Ando, NIRS
3
We all know how nice it is to have a high RBE
(and only where you need it)
CIRT
7 years later
Vertebral osteosarcoma
Imai et al., Lancet Oncology 2006
4

• We can hope in good clinical results if we are
  able to deliver this radiation, with nice
  physical and biological properties, to the target
• We have to know where the target is
• No matter how good our beam, if the target is not
  where we thought we will not kill the tumor

5
Missing the target
6
Why should the target be elsewhere?

• Mispositioning (interfraction)
• Organ motion (intrafraction)
• Shape change (interfraction)

7
This issues are much more critical in
hadrontherapy (compared to photons RT) because of

• Steeper gradients
• Dose distribution heavily dependent on tissue
  density
• Hypofractionation

8

• Organ motion is even more critical when active
  scanning is used (risk of hot and cold spots)
• For the CNAO delivery system spot size will be
  of millimiters (4 - 10) and time scale will be of
  milliseconds.

9

• What has been done until now ?
• What do we plan to do at CNAO ?
• What would we like to do but do not know how?

10
Delivery system at CNAO

• Fixed vertical and horizontal beams
• Active spot scanning (no passive spread beam)

11
1 - Precise positioning

• Always an issue (for all body sites)
• Need of fixation devices
• Need of position verification devices
• Need of 6 DOF position correction devices
• The only issue for some body district (limb,
  head, spine ?)
• Time consuming

12
Precise positioning

• We believe that for fixed body sites a
  orthogonal KeV X-Ray imaging coupled to a robotic
  positioning device is necessary and sufficient.
• Head and body masks, personalized cushions, bite
  blocks and all such devices employed in high
  precision photons RT can be straightforwardly
  transferd to hadrontherapy

13
A customized Schaer systyem will be used
13
14
14
15
Courtesy of SEAG, RPTC
15
16
16
17
2 Shape change

• Tumor shrinkage
• Weight loss or gain
• Dependent on tumor biology
• Relatively easy to cope with through close
  clinical monitoring of the patient and frequent
  off line imaging (e.g. each week for
  gynecological malignancy)
• Too late if discovered in the treatment room (a
  check is anyway most necessary)

18
3 Organ motion

• The real problem
• Due to peristalsis, swallowing, heartbeat,
  vessels pulsation but especially BREATHING
• Affects lung treatments, liver treatments, and to
  a lesser extent all abdominal and pelvic
  treatments. If the patients is in supine position
  (as may be necessary without a gantry) it may
  also affect retroperitoneal treatments

19
Breathing

• A complex caotic phenomenon
• Intervals of more or less regular breathing that
  each 5- 15 minutes change pattern and through a
  highly unpredictable phase reach a new temporary
  steady state
• No reliable correlation betwen body surface and
  internal displacement during change of patterns
• Frequency, Tidal volume, relative role of
  thoracic muscles and diaphragm and end expiration
  volume are among the parameters that can change
  in minutes
• Elevated inter-patient and intra-patient
  variability
• Pain and emotional status can influence breathing
  pattern
• Displacements up to 3-4 cm can take place
• Not possible to predict a priori direction and
  amplitude of a lung nodule displacement

20
How to deal with respiratory organ motion

• Breathhold, gating, tracking, coaching, abd.
  pressure
• Whichever you choose you need to measure motion
  (target position or surrogate marker)
• Once you know how the target moves you can steer
  the beam, or decide to enlarge the irradiated
  volume (a simple CTV ? PTV approach is not
  adequate)
• It is mandatory to check what you are doing (a
  real time, in vivo dosimetry would be
  appreciated)
• Robustness of the treatment plan should be
  considered (ideally as a parameter for inverse
  planning optimization)
• Everything should happen in real time

21
Strategies

• Breath hold requires active cooperation, you
  cannot avoid measuring target position (to be
  sure it is performed correctly)
• Gating (in my opinion probably the best
  solution)
• Tracking much more complicated with only a small
  gain on time efficiency

22
Measuring target displacement

• Breathing is a caotic phenomenon
• Surrogate markers air flow, body surface
  position (simple 1D marker option cfr NIRS, or
  optoeletronic surface reconstruction with or
  without passive markers)
• Measuring the real thing implanted markers and
  fluoroscopy, or frequent X-ray (how often ?)

23
Surrogate markers
Only the phase probably not enough for spot
scanning (and for stack layer?)
24
More advanced surface reconstruction
Body markers
Surface reconstruction
25

• From the camera images to the surface or marker
  position
• From the marker position to the target position
• How many msec ?
• At the end it is only a guess (caotic behaviour,
  hysteresis)
• NOT ENOUGH

26
You have to look at what you do

• Fluoroscopy (radioprotection issues)
• Periodic x-ray check during dose delivery to
  verify the external ? internal model, i.e. you
  trust the model for a short time between cheks
  (how often have I to check ? Regular interval vs.
  adaptive interval cfr. Isaksson et al, Med Phys
  2005)

27

• What has been done until now (by us)?
• X-ray Exac trac with a custom made BH system
  (together with EIO and Politecnico di Milano).
  Accuray Synchrony (together with CDI) .. ?
• ? not precise enough for spot scanning
• What do we plan to do at CNAO ?? ?? . In the
  beginning only fixed targets

28
What would we like to do but do not know how?

• Approach similar to Accuray synchrony but with
  better surface detection and adaptive chek
  intervals (optimizing what already exist) ?
  Problem our orthogonal X ray flat panels cannot
  be both in position when the terapeutic beam is
  on
• Something new??

29

• Can we detect in real time without extra dose to
  the patient the position of a lung nodule ?
• Can we avoid invasive fiducial marker placement ?
  (I believe NO)
• Could we measure the position of an implanted
  marker in some other way ? Maybe measuring the
  influence of an implanted coil on an external
  magnetic field makes sense or is it only an
  ignorants dreor measuring the change in electric
  capacitance? Is anybody working on this? Does it
  am? Would it work with a ion beam on?

30
If you have an idea you like be sure somebody
else had already had it quite a while ago !
30 mm 20 mm 10 mm
Meas. Sci. Technol. 19 (2008) 024006 (9pp)
doi10.1088/0957-0233/19/2/024006 Tracking of
internal organ motion with a six
degree-of-freedom MEMS sensor concept and
simulation study Manuel Bandala and Malcolm J
Joyce
31
Implantable with 14 gauge needle Accuracy 0.2
mm Read out rate 10 Hz Not yet tested in clinics
Balter et al, IJROBP 2005
32
We are the ones who could benefit most of such
sensors

• Hadrontherapy as a community should support those
  researches
• Probably not too ambitious a goal but investment
  and focusing are required

33

• Steering the beam is easier than measuring the
  motion
• Novel approaches are welcome
• We should strive for a continuous real time
  measurment of the target position
• At present I believe we need implanted markers
• In the end active spot scanning may turn out to
  be simply not the best way to treat lung and
  liver and we may have to go back to passive
  spreading

34
Quality assurance and in vivo dosimetry

• Organ motion is difficult to deal with, we have
  to check what we are doing and how effective it
  is
• In vivo dosimetry is mandatory
• In vivo dosimetry can detect also shape changes
  that eluded clinical monitoring

35
What has been done until now (not by us)?

• In beam PET

What do we plan to do at CNAO ?

• In beam PET (cfr. FP7 call)
• AQUA

36

• Those approaches have much in common
• They mesure something that would have happened
  anyway (and so no extra dose is delivered to the
  patient)
• If the result is positive you can be sure that
  everything went well, if it is negative you are
  not sure of what went bad
• The signal that can be detected bears less
  information respect to the actual dose
  distributions and therefore cold spots may be
  masked by a blurring effect
• You know the result after the treatment session
  is ended, for hypofractionated treatments it may
  be too late

37
Different dose
38
What would we like to do but do not know how?

• When a fiducial marker is inserted, have a
  dosimeter inside it (it should be read without
  extracting the fiducial from the patient body).
  Science fiction ?
• Have a proton portal imaging (or part of it) to
  check position during dose delivery

39
Proton portal imaging ?

• (CNAO can accelerate C12 to 400 MeV/u so protons
  could be much faster)
• Would it be possible to put the Bragg peak beyond
  the patient and measure the range without
  delivering a high dose ?

40
How much dose ?
Can we build this black box ?
41
?

• OK
• Abort
• Correct online?

42
Thank you for your attention