8/12/2011
Disclosure
Patient Positioning in Head
& Neck Treatment
Lei Dong, Ph.D.
Dept. of Radiation Physics
University of Texas MD Anderson Cancer
Center, Houston, Texas
Sponsored Research: Varian
Sponsored Research: Philips
 NIH Grant: Optimization of Proton Therapy
 CPRIT Grant: Advanced Volumetric
Imaging and Adaptive RT for Lung Cancer


AAPM Therapy Educational Symposium
Aug. 2nd, 2011
Outline
Learning Objectives
•
•
•
Learn about essential components of
Head & Neck Immobilization
Learn about IGRT applications and
limitations
Suggest IGRT/ART strategies
Immobilization
 IGRT
 Adaptive RT

1
8/12/2011
Immobilization Principles for H&N

Head

Aquaplastic mask




Bite block
Mouth-piece

Neck

Clavicles



Thermal Mask Immobilization
Cutout over eyes and mouth, bolus when needed
Chin
Custom head mold
Aquaplastic mask
Shoulder restraining methods
Mouth-piece/Bite-block for
Tongue/Chin Immobilization
2
8/12/2011
Civico
How effective is an
immobilization device?
How much margin do we need if
IGRT is not used
Orfit
Daily Setup Shifts Evaluated by
IGRT
CIVICO
ORFIT
10/12 patients and 350 measurements
3
8/12/2011
Mask doesn’t properly fit
Margin Considerations
Direction
Systematic
Error (cm)
Random Error
(cm)
Van Herk’s
Margins (cm)
Civco
Orfit
Civco
Orfit
Civco
Orfit
AP
0.163
0.202
0.124
0.125
0.494
0.593
SI
0.156
0.122
0.156
0.117
0.499
0.387
ML
0.127
0.161
0.107
0.109
0.392
0.479
Planning CT
Daily CT09
Measuring set up using MatriXX system
Mask doesn’t properly fit
• Base Frame
• 9o Wedge
• Foam Head Rest (Medium
Density):
 1 to 6
B
C
Energy: 6MV
F
Daily CT09
Planning CT
Author: Luis Fong, Ph.D.
Mayo Clinic
4
8/12/2011
No. 1 MatriXX Measurement
Rt
From IMRT to IGRT
% Transmission
98.0-100.0
96.0-98.0
94.0-96.0
92.0-94.0
90.0-92.0
Inf
Transmission through Central Area
Average
95.8
Max
93.7
Min
97.0
Author: Luis Fong, Ph.D.
Mayo Clinic
Image Guidance Technologies for H&N IMRT
Direct Portal Verification
Daily orthogonal KV x-rays
 Implanted fiducials and kV x-rays
 Optic-based surface alignment
 In-room volumetric imaging

Tomotherapy MV CT
 Cone-beam CT (KV and MV)
 kV CT-on-rails
 MR (in development)

To Isocenter Verification
5
8/12/2011
Isocenter Verification
KV DRR
MV
Isocenter Verification
KV DRR
MV
AHE enhancement
KV-KV
What is the planning CT scan
slice spacing in your practice?
1.
2.
3.
4.
5.
kV DRR
5 mm
3 mm
2.5 mm
2 mm
< 1.5 mm
On-board kV Imaging
6
8/12/2011
2D planar x-rays = volume alignment?
•
Traditional alignment assumes rigid-body for the entire H&N region
•
Can you combine multiple 2D ROI for a 3D ROI?
•
Setup Uncertainties In Head & Neck Treatment
2 x 2D = 3D?
Elapsed
Days
Patient with Tongue Base Carcinoma
19 CT Scans over 47 Days
Lei Dong et al. (MDACC)
DailyVoluntary
Setup Uncertainties
JunctionCaptured
Shifts by
Orthogonal kV Images
2D vs. 3D Alignment
7
8/12/2011
Weekly film discovered:
IGRT is off?
3D vs. 2D
In-room
CT
Plan
Daily IGRT Reduces Margins by Half
Imaging Dose
Planning CT
28 consecutive patients 1013 CBCT scans
Pre-alignment margins: ~ 4-5 mm
 Post-alignment margins: ~ 1.5 – 2.5 mm
 More margins may be needed larynx
cancers or patients with significant weight
loss
Den et al., IJROBP v76 No. 5, 2010
CBCT “Standard Head” 0.4 cGy


Varian OBI ver 1.4
8
8/12/2011
Planning
Planning
Daily
Daily
3D IGRT Challenge
Which structure to align?
How to handle non-rigid setup
error?
Examples of anatomy rotation between the planning CT (first row) and the daily CT (second row). The axial
CT images shown on the left indicate a roll in the patient’s head; the coronal CT images on the right show
yaw (rotation of the spinal column).
Change in the Neck Curvature
Planning CT
PPM
Daily Cone-beam CT with
planning contour overlay
C2
C6
L. Zhang et al. "Multiple regions-of-interest analysis of setup uncertainties for head and
neck cancer radiotherapy," Int. J. Radiat. Oncol. Biol. Phys. 64 (5), 1159-69 (2006).
9
8/12/2011
40
Histogram of PPM Measured Random Shifts
in AP Direction
1 .5
0 .5
0 .0
-0 .5
C6
PPM
-1 .0
-1 .5
SI
1 .0
0 .5
0 .0
-0 .5
C6
PPM
-1 .0
-1 .5
-1 .5
-1 .0
-0 .5
0 .0
0 .5
1 .0
1 .5
C2 Shifts (cm)
-1 .5
-1 .0
-0 .5
0 .0
0 .5
1 .0
1 .5
C2 Shifts (cm)
RL
1 .0
Zhang LF et al. Multiple regionsof-interest analysis of setup
uncertainties for head-and-neck
cancer radiotherapy. Int J Radiat
Oncol Biol Phys 2006;64:15591569.
0 .5
0 .0
-0 .5
C6
PPM
-1 .0
-1 .5
-1 .5
-1 .0
-0 .5
0 .0
0 .5
1 .0
30
20
10
Stent (AP)
Without Stent (AP)
0
-1.0
-0.8
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
40
Histogram of PPM Measured Random Shifts
in SI Direction
30
20
10
Stent (SI)
Without Stent (SI)
0
Red: no mouth-piece
-1.0
Blue: with mouth-piece
1 .5
C2 Shifts (cm)
-0.8
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
40
Percent
C6/PPM Shifts (cm)
1 .5
Histogram
comparison of
PPM measured
random shift
between stent
group and nonstent group
Percent
1 .0
Percent
AP
C6/PPM Shifts (cm)
C6/PPM Shifts (cm)
1 .5
Histogram of PPM Measured Random Shifts
in RL Direction
30
20
10
Stent (RL)
Without Stent (RL)
0
-1.0
-0.8
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
PPM Measured Random Shifts (cm)
Setup
uncertainties
in subregions of
H&N
anatomy
Increased
deformation with
longer distances
from the reference
(C1-C3)
Kranen et al., IJROBP, V73, pp1566 (2009)
Kranen et al., IJROBP, V73, pp1566 (2009)
10
8/12/2011
Impact of ROI Selection for Patient Setup
A Region-of-Interest based
image guidance strategy is
preferred
The selection of ROI depends on
the clinical case
Kranen et al., IJROBP, V73, pp1566 (2009)
Primary Tumor Response To RT
140
HN01
120
HN03
100
HN04
HN05
Volume (cc)
Anatomy Changes During
Radiotherapy
HN02
HN06
80
HN07
HN08
60
HN09
HN10
HN13
40
HN14
HN15
20
0
0
7
14
21
28
35
42
49
Elapsed Treatment Days
Barker et al. Int J Radiat Oncol Biol Phys 2004; 59 (4):960-970.
11
8/12/2011
Significant Anatomic Variations
Planning CT
Lee, C., K. M. Langen, et al. (2008). Int J Radiat Oncol Biol Phys 71(5): 1563-71.
A RTOG H0022 case
During Treatment
CTV is in the air!
C2-bone Setup and Anatomy
Change
Image-guided Adaptive
Radiotherapy for H&N
Dong/MDACC
12
8/12/2011
Benefit of IGRT Alone



Parotid gland mean dose above the planned dose by 5
to 7 Gy in 45% of the patients
Bone (C2) alignment led to reductions relative to BB
alignment in 91% of patients (median, 2 Gy; range, 0.3–
8.3 Gy).
The parotid dose from bone alignment was still greater
than planned (median, 1.0 Gy, p = 0.007). Neither
approach affected tumor dose coverage.
Who is going to draw these
contours in adaptive
radiotherapy?
O’Daniel et al. IJROBP vol.69 No.4 (2007)
What is Deformable Image
Registration?

Auto-Segmentation of H&N Anatomy
Planning CT Image
Mid-Course CT Image
Geometric transformation that brings one
image in precise spatial correspondence with
another image.
Registration
Planning Contours
Deformed Contours
13
8/12/2011
Head and Neck Case
Replanning Challenges
upper
Planning contours were
automatically transformed to
subsequent CT images
upper
lower
lower
Volume
DVH
Dong/MDACC
Re-Plan Activities
20%
What are the benefits of
adaptive replanning?
Re-Plan1
Re-Plan2
10%
0%
-10%
-20%
CTV70 % Change
CTV63 % Change
CTV57 % Change
-30%
Left Parotid % Change
Right Parotid % Change
-40%
1
2
3
4
5
6
7
8
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
14
8/12/2011
Re-Plan for Fraction #11
Re-Plan for Fraction #11
Fraction #11: 1st replan
Fraction #11: 1st replan
Solid lines: 1st replan
Dotted Lines: original plan
Target Coverage is not
a problem
Original Plan Re-calculated on Fraction #11
Replan on Fraction #11 CT
15
8/12/2011
Fraction #11: 1st replan
Solid lines: 1st replan
Dotted Lines: original plan
Fraction #11: 1st replan
Contra-lateral parotid
can be reduced by
replanning
Fraction #11: 1st replan
Solid lines: 1st replan
Dotted Lines: original plan
Solid lines: 1st replan
Dotted Lines: original plan
Replanning benefits
both parotids
Fraction #23: 2nd replan
Solid lines: 2nd replan
Dotted Lines: 1st replan
Oral cavity
16
8/12/2011
Fraction #23: 2nd replan
Tumor Progression During Treatment
Planning CT
1st replan re-calculated on Fraction #23
2 weeks into treatment
Replan on Fraction #23 CT
Practical Issues in Adaptive RT

Mask is not fitting
How do you improve immobilization?
Planning
First Fraction
Last Fraction
17
8/12/2011
H&N Setup Uncertainties
Adaptive Mask?
1.0
pat1
pat2
pat3
pat4
pat5
pat6
Linear (pat1)
Linear (pat2)
Linear (pat3)
Linear (pat4)
Linear (pat5)
Linear (pat6)
0.9
0.8
3D Shifts (cm)
0.7
Cranial Stopper
0.6
0.5
0.4
0.3
0.2
0.1
0.0
0
5
10
15
20
25
30
35
Fraction #
Summary for H&N patients
• Setup uncertainties
– Need to worry about residual setup error at different parts of
the H&N anatomy
Acknowledgement
• MD Anderson H&N Team
• ROI-based alignment is strongly recommended
– Immobilization technique is important to minimize non-rigid
changes.
– Systematic uncertainties appear to be larger than random
setup uncertainties.
• Offline correction strategy is beneficial.
• Image Guidance
– 3D alignment can resolve out-of-plane rotations and 3D
shifts of ROI better than 2D projection method
• Trends in volumetric and positional variations
– Adaptive radiotherapy strategy
18