IMRT: Where next Image guided Radiotherapy

1
IMRT Where next? Image guided Radiotherapy

• Dr. A. W. Beavis PhD
• Principal Physicist,
• Hull and East Yorkshire Hospitals (NHS) Trust
• (Hon) Senior Research Fellow,
• University of Hull

2
Thank you for this honour!

• Mme Naudy and the organisers of this training
  meeting
• Great honour for me and my hospital to be invited
  back to speak on this course for a second time
• Hope you enjoy my lecture!
• www.hullrad.org.uk

3
IMRT is possible in any clinic!

• Now have the weaponry to treat small volumes of
  tissues to high(er) doses
• Need to develop the intelligence to find the
  tissue to target!
• Recall that target volume delineation is as
  bigger part of the prescription as defining the
  doses
• Need to ensure we are treating the right tissues!

4
Limitation of current IMRT planning at PRH and
world-wide

• CT imaging does not show the position or
  viability of tumour
• Solutions
• Advanced MR techniques
• Spectroscopy/ T2 mapping
• DCE
• Also PET imaging

5
CMSFocal FUSION used to fuse T2-weighted images
with CT data set
Use of T2 weighted images routine in our
department for planning purposes
6
Can we just add images together from different
modalities?

• Do we expect the prostate on a CT scan from
  Monday to overlay on the prostate on an MR scan
  acquired the next Thursday?

Courtesy Di Yan, William Beaumont and Marcel Van
Herk, NKI Amsterdam

• No, because organs move, fill-up/ empty and
  generally try to confuse us!

7
GE Lightspeed CT-PET combined scanner
Should we obtain images at same point in time?So
we KNOW the anatomy is registered.
8
But just discussed organs move...?

• So information about position and magnitude of
  motion is required
• So in planning we either need to build in motion
  margins (ICRU50) or know precisely where each
  organ is during treatment
• maybe stratify patients as good/ bad Conformal
  RT candidates

T 0
Z z1
Z z2
Z z3
T t
9
PTV is an envelope within which the CTV is
present with 100 certainty

• Cannot assume any edge of the PTV is less at risk
  from being under-irradiated

DEFINE MARGINS CAREFULLY EVEN WITH ADVANCED
IMAGING!
10
Where are we going? Image guided radiotherapy

• Planning the treatment

11
Defining the Boost target in IMRT ?

• Much interest in Magnetic Resonance Spectroscopy
  MRS for this purpose
• UCSF Pickett et. al. IJROBP 44(4) 921-929
  (1999) and references therein.
• Look for decreased Citrate and enhanced
  Choline traces in the spectra
• MRS is not simple to do and it is time consuming
  to perform necessary studies

12
?

• IMRT offers the capability to BOOST the dose to
  the actual tumour potentially increasing the
  tumour control probability

PTV1 Prostate
PTV2 Boost the conventional dose??
55 Gy covering PTV1
65 Gy covering PTV2 PTV2 is covered by 60 Gy
13
Spectroscopy data for Peripheral Zone (PZ)
Carcinoma and normal PZ data
normal
tumour
T2 image showing MRS sampled voxels
Choline
Citrate
Citrate
Choline

• Single voxel measurements showing normal PZ
  Citrate sample and that for
  diseased (Ca.) tissue

G.P. Liney et al. NMR in Biomedicine 12 39-44
(1999)
14
MR-Spectroscopy Imaging (MRSI)
Tumor
Combination of MRI and 3D-MRSI gtgtgt acquisition
of an image of spectra
Cho
Normal
NAA
Cr
Cho
Necrosis
Courtesy of Prof. Lynn Verhay, UCSF
15
MRSI 3D contours
Abnormality-Index 2,3,4
Voxel size 1cc
Courtesy of Prof. Lynn Verhay, UCSF
16
Our approach

• Though our MR group has experience in single
  voxel MRS they opted for the use of T2 maps
  rather than expanding to multi-voxel MRS
• G.P. Liney et al. Proton MR T2 Maps correlate
  with the Citrate Concentration in the Prostate
  NMR in Biomedicine 9 59-64 (1996)
• T2 map is a plot of the absolute T2 value/
  relaxation time (rather than signal intensity)

17
T2 map .v. spectroscopy (MRS)

• Advantages
• short acquisition time compared to MRS
• produces anatomical image with mm resolution
• no need for registration to T2 weighted image
• modern scanners produces T2 map image as part of
  programmed sequence (Philips Intera 1.5T)
• Disadvantage
• dont see Choline data (not clear if this is
  important)
• biopsy false-negatives - blood and urine in
  Prostate will cause high T2 signal!
• Biopsy false-positives this high signal will
  fade with time (darken image) as fluid disperses

18
T2 maps of volunteer and patient alternative
to MRS and potentially directly applicable into
RT planning.
Dark area indicates disease
T2 map of patient with tumour in Peripheral Zone

• T2 map of healthy volunteer with normal
  Peripheral Zone image

Absolute T2 value .v. Citrate concentration

• Advantages of T2 mapping over MRS
• short acquisition time compared to MRS
• produces anatomical image with mm resolution (no
  registration needed)

G.P. Liney et al. NMR in Biomedicine 9 59-64
(1996) Proton MR T2 Maps correlate with the
Citrate Concentration in the Prostate
19
Dynamic contrast-enhanced-MRI
Another method (relevant to more tumours)

• Gadolinium Gd-DTPA
• Shortens the T1 and T2 signal enhances image on
  a T1 weighted image
• Basically, it enhances the signal from the dense
  vascularisation (and poor integrity) typical of
  neoplastic (tumour) tissue
• Normal tissue does NOT enhance as quickly
• Relevant to most solid tumours
• Citrate chemistry is unique to prostate.

G.P. Liney et al. NMR in Biomedicine 12 39-44.
(1999)
20
T1 weighted images at different time intervals
post contrast injection.
Tumour (volume) in PZ
138
210
t 0
453
Benign disease in CG BPH
Enhancement factor, EF(t)
Normal tissue i.e. RHS of PZ shows no/little
uptake
Dont use rectal probe for Radiotherapy sequences!
21
Transfer of information into the Radiotherapy
Treatment planning process
The data obtained from tracking the progress of
the Contrast during the T1 scan (the DCE scan)
can be analysed in several ways to extract
spatial information to provide the treatment
planning process. The simplest is to threshold
the data set to identify all pixels that
enhanced above a certain (pre-determined) value.
Plot of enhancement factor .v. time for a T1-DCE
sequence obtained for a RT patient. A threshold
analysis has identified those pixels in the
shaded area as being viable tumour
2
22
Dynamic Contrast-Enhancement

• Acquire baseline images pre-contrast
• Multiphase post-contrast
• Uptake of tumour can be analysed
• Parameter map produced

23
Threshold the pixels that indicate tumour, XiO/
FOCAL contouring tools can pick it out
Use automatic contouring tools to pick these
areas out and delineate them
Transfer the information into the Radiotherapy
Treatment planning system
T2-Weighted FSE Image overlay generated from
DCE analysis added on
Can now contour the tumour without further expert
Radiology knowledge
24
Plot of enhancement factor .v. time
ROI 3 (orange) Dark bit of PZ on T2 image could
have interpreted as tumour but DCE indicates
its a benign problem
T2 image with an overlay generated from the DCE
fused on
1
2
1
3
3
2
25

• We are currently already performing (Radiology)
  DCE method on other sites
• Working on their usefulness in Radiotherapy
  planning

• DCE showed nodal involvement and chest wall/
  muscle
• infiltration not seen on planning CT performed 3
  days previously

26
Summary of process 1

• Obtain T2 weighted volumetric data set of images
• Inject contrast
• obtain a time series of T1 weighted images (at
  same location and resolution as above T2 set)
• Analyse data and produce a parameter map
  providing spatial info of viable tumour

27
Summary of process 2

• Fuse the parameter map and the T2weighted images
  to provide tumour and relational anatomy
  information
• DICOM to FOCUS and in FOCALfusion register to
  planning CT
• FOCALSim perform contouring

28
Summary of process 3
PTV2 MRI delineated disease (white CTV)

• Produce inverse plan in CMS FOCUS/ XiO
• Create a simultaneous boost IMRT plan

PTV1 Prostate
60 Gy covering PTV2
52.5 Gy covering PTV1
29
Role of DCE in verification of (conventional)
radiotherapy efficacy?
A
B
Plot of PSA level .v. time. A represents the time
of the pre-Radiotherapy MRI-DCE scan and B the
post-treatment scan.
Enhancement factor plots .v. time for a patient
before conventional radiotherapy.
Plots after treatment
30
Functional MRI for conformal avoidance

• Left handed 21 year old with a glioma in the left
  frontal lobe
• T2-weighted fast spin-echo (TE/TR100/3640 ms) 3D
  T1-weighted fast spoiled gradient-echo
  (TE/TR/flip 4.2/13.4ms/200)
• fMRI acquisition (self-paced finger-thumb
  opposition with the right hand).
  T2-weighted
  (TE/TR/flip50/300 ms/900)

Activation plot
31
(No Transcript)
32
Dose Volume Histogram
MC (Conv)
MC (IMRT_noOAR)
MC (IMRT_OAR)
7.2
34.6
29.6
33
Reproducibility fMRI
Series 2
Series 1
Series 3
0.605
0.605
0.572
Correlation Coefficient
34
Repeated imaging on volunteers

• Area highlighted is consistent with patient data
• Intra session repeats show consistent data
• Inter session repeats show consistency
• Developing this further to create tests to feed
  probability density maps into inverse planning
  driven by biologically defined cost functions
  (EUD approach)
• Wiesmeyer and Beavis AAPM abstract
• ESTRO abstracts .

35
Conclusion

• Inverse planning/ IMRT allows us to deliver
  different dose levels to multiple targets
  simultaneously
• Using MRI - We are developing methods to enhance
  the identification of diseased sites and
  definition of boost target volumes for IMRT
• Concentrating on Prostate, Brain and Head and
  Neck (YCR grant) but any solid tumour is
  possible!

36
Thanks to

• Dr. Gary Liney
• Prof. Lindsay Turnbull MRI/ Radiology
• My colleagues in Hull
• Prof. Lynn Verhay - UCSF
• Dr. Andrea Pirkzall - UCSF

CMS for their interest and financial support in
our research work Also for funding my
participation in this meeting