IMRT for Gynecologic
Malignancies:
The University of Chicago
Experience
Background
RT has a long history in the treatment of
gynecologic malignancies, notably cervical
and endometrial cancer
The 1st gynecology patient was treated with
RT a century ago
Bulent Aydogan, PhD
The University of Chicago
Gynecologic RT
Highly efficacious and well tolerated
in most patients
Excellent pelvic control particularly
in early stage cervical and
endometrial cancer
Adjuvant RT improves outcome of
women with high risk features
following surgery
GYNGYN-IMRT Rationale
Potential toxicities due to the treatment
of considerable volumes of normal tissues
Small bowel→
bowel→ diarrhea, SBO, enteritis,
malabsorption
Rectum → diarrhea, proctitis, rectal bleeding
Bone Marrow → ↓WBC, ↓platelets, anemia
Pelvic Bones → Insufficiency fractures,
necrosis
Reduction in the volume of normal tissues
irradiated with IMRT may thus ↓risk of
acute and chronic RT sequelae
1
Gynecologic IMRT
flow chart
Patient selection
↓
Simulation – Prone vs. Supine; Type of immobilization
↓
Target and Tissue Delineation – Multiple imaging modalities
↓
Treatment Planning/Optimization – Number of beams/orientation
↓
Plan Evaluation – High conformity vs. dose homogeneity
↓
Quality Assurance – Verification of calculated dose
↓
↓
Treatment Delivery/Verification – Verification scheme / IG
Patient Selection
Poor candidates
Uncooperative patients
Unable to tolerate time on the table
Markedly obese patients not ideal
Inability to capture entire external contour
Difficulties with daily setup
Dosimetric benefits may be less in the
obese*
↓
Imaging for treatment assessment - Tumor response / Adaptive
approach
Simulation and CT Scanning
*Ahamad et al. Int J Radiat Oncol Biol Phys 2002;54:42
Immobilization
University of Chicago
Patients in supine position
Immobilized using a customized device
Patient scanned from L2 to below
ischial tuberosities
Oral, IV and rectal contrast
•Immobilized supine
•Upper and lower
body alpha cradles
indexed to the table
Mell LK, Roeske J, Mundt AJ.
Gynecologic Tumors: Overview Chapter
23 IMRT: A Clinical Perspective BC
Decker, Toronto 2005
2
Contrast Administration
Simulation
University of Colorado
Others favor the prone
position
Helps delineate normal and
target tissues
Data from the U Iowa
suggest dosimetric
benefits with the prone
Position
Oral, rectal and IV contrast
Bladder contrast not needed
*Adli et al
IV contrast is important
(vessels
serve as surrogates for nodes)
*With experience, IV contrast
less needed
.Red J 2003;57:230-238
Schefter T, Kavanagh B.
Cervical Cancer: Case Study. Chapter 23.1
IMRT: A Clinical Perspective 2005
A vaginal marker is also placed
(be careful not to distort)
Note: Not possible in patients treated with pelvic-inguinal
IMRT (frog-leg position)
Target Definition
CTV components depend on the
pathology
In all patients:
Upper ½ of the vagina
Parametrial tissues
Pelvic lymph nodes regions (common,
internal and external iliacs)
In cervical cancer and endometrial
cancer patients with positive cervical
involvement
include the presacral region
CTV and Normal Tissues
postoperative RT
definitive RT
PTV
Bladder
CTV
Small bowel
Large bowel
Rectal wall
P Georg, MD – Med University Vienna
3
Normal Tissues
Normal Tissues
Normal tissues delineated depends on the
clinical case
In most cases, include:
Small bowel, rectum, bladder
In patients receiving concomitant or
sequential chemotherapy, include the bone
marrow
Others include the femoral heads
Kidneys and liver included only if treating
more comprehensive fields
Small Bowel
• Dip small bowel contour into concave CTV
• Conformity reducing small bowel dose
Consistency with contouring helps with
DVH interpretation and outcome studies
Rectum:
Rectum: Outer wall (anus to the sigmoid
flexure)
Small bowel:
bowel: Outermost loops from the L4L45 interspace
Include the colon above the sigmoid flexure as
well in the “small bowel”
bowel” volume
Bone marrow:
marrow: Intramedullary space of the
iliac crests (EASY TO CONTOUR THE BONE)
Stop at the top of the acetabulum
Note that this approach ignores marrow in
other pelvic bones
Treatment Planning
Expand CTV
PTV
To account for setup uncertainty and organ
motion
Appropriate expansion remains unclear
Small Bowel
Jhingran A, et al.
MD Anderson
Endometrial Cancer:
Case Study
Chapter 23.2
IMRT: A Clinical
Perspective BC
Decker 2005
Various expansions have been used for Gyn
IMRT ranging from 0.5 to 1.5 cm
At the U of Chicago, we use 1 cm
Less is known about normal tissue motion
So we don’
don’t expand the normal tissues
Other centers, e.g. MD Anderson, routinely
expand normal tissues
4
Setup Uncertainties
Digitized weekly setup films of 50 patients
Immobilization: Alpha cradle under legs and
upper body with arms above head*
Measured setup position using imageimageregistration interface (Balter, et al.)
σLR = 3.2 mm
σSI = 3.7 mm
σAP = 4.1 mm
Mundt, Roeske and Lujan. Intensity modulated radiation therapy in gynecologic
malignancies. In Medical Dosimetry, June 2002.
Organ Motion
Using this approach, no failures in the
vaginal cuff have been seen in patients
treated with adjuvant IMIM-PRT at our
institution (>100 pts treated)
Nonetheless, tighter volumes could result
in toxicity
Tighter margins are also needed if higher
than conventional doses are used
Organ Motion
A concern in the region of the vaginal
cuff
Two approaches are being studied at our
institution to address this:
IGRT (Varian OBI unit)
Vaginal immobilization (B. Aydogan, Int J Radiat
Oncol Biol Phys 65:266-73, 2006.)
Now we simply avoid tight CTV volumes
and use a 1 cm CTV PTV expansion
Produces very generous volumes around the
vaginal cuff
MD Anderson’
Anderson’s
“Integrated Target Volume”
Volume”
A creative solution to the organ motion
problem developed at MDAH?
Two planning scans: one with a full and one
with an empty bladder
Scans are then fused
An integrated target volume (InTV)
InTV) is
drawn on the full bladder scan
(encompassing the cuff and parametria on
both scans)
PTVInTV
InTV is expanded by 0.5 cm
5
Normal Tissue Changes and
Organ Motion
Bladder and Rectal Volumes
160
Small bowel
Bladder
Bladder
140
120
100
Rectum
Bladder
80
60
40
20
Rectum
Rectum
0
0
Week 3 scan
1
2
Treatment planning scan
DVH Comparisons - Bladder
100
3
4
5
Week
DVH Comparisons - Rectum
100
80
80
Planning
Week 1
Week 2
Week 3
Week 4
Week 5
60
40
20
Planning
Week 1
Week 2
Week 3
Week 4
Week 5
60
40
20
0
0
0
20
40
60
80
Percent Dose
100
120
0
20
40
60
80
100
120
Percent Dose
6
IMRT Planning and delivery
at the University of
Chicago
Initial phase are completed with CORVUS
planning system
Currently we use Varian Eclipse and Pinnacle
7-9 coco-axial beam angles
(equally spaced)
120 Leaf MLC using
stepstep-andand-shoot mode on
a Varian 2100 Ex
Treatment
Planning
Increasing number of
planning systems now
commercially available
Despite inherent
differences, no one
system appears superior
Acceptable gynecologic
IMRT plans have been
produced on all major
planning systems
Treatment Planning
Prescription dose: 4545-50.4 Gy
45 Gy in pts receiving vaginal brachytherapy
50.4 Gy if external beam alone
1.8 Gy daily fractions
Given inherent inhomogeneity of IMRT
Avoids hot spots > 2 Gy
“Dose painting”
painting” (concomitant boosting)
remains experimental
Potentially useful in pts with high risk
factors (positive nodes and/or margins)
IMIM-WPRT Plan Optimization
Current PTVPTV-Specific Criteria
Conformity
PTV Coverage
Hot Spots
Location
Magnitude
Cold Spots
Location
Magnitude
Acceptable
Good
> 98%
Unacceptable
Poor
< 96%
Within CTV
Preferably within GTV
<10% (110% dose)
0% (115% dose)
Edge of PTV
Rectal or bladder
walls in ICB region
>20% (110% dose)
>2% (115% dose)
Edge of PTV
<1% of the total dose
Within CTV or GTV
>1% of the dose
7
DVH Acceptance Criteria for
Small Bowel
IMIM-WPRT Plan Optimization
Normal Tissue Specific Criteria
Dosimetric analysis of acute GI
toxicity in our Gyne IMRT pts was
performed
On multivariate analysis, the
strongest predictor of acute GI
toxicity was the small bowel volume
receiving the prescription dose or
higher (SBvol100%)
A more difficult question is what
makes a normal tissue DVH
acceptable.
IM-WPRT plans achieve better
normal tissue DVHs than WPRT
plans. But how good does a
normal tissue DVH need to be?
The answer is not clear
Roeske et al.
Radiother Oncol 2003;69:201-7.
NTCP Analysis of
Gynecologic IMRT Patients
Isodose Distribution
Comparison
Roeske et al. Radiother Oncol 2003;69:2012003;69:201-7.
1
1+
0.8
0.7
PTV
1
NTCP =
0.9
410
V100
3.2
Conventional
Pelvic RT
PTV
0.6
0.5
100%
IMRT
0.4
70%
0.3
0.2
0.1
0
0
100
200
300
400
500
600
100%
70%
Volume (cc)
8
Absolute Volume (cc) of SBR
Receiving 45 Gy
Small Bowel
100
1200
Conv
80
1000
60
800
40
600
IMRT
20
Conv
IMRT
400
200
0
0
20
40
60
80
100
120
0
1
Percent Dose
IMIM-WPRT
Planning Studies
Author
Roeske
Ahamad
Chen
Selvaraj
2
3
4
5
6
7
8
9
10
Patient Number
Quality Assurance
↓Volume Receiving Prescription Dose
Bowel
Bladder
Rectum
↓50%
↓23%
↓23%
↓40NS
40-63%* NS
↓70%
↓**
↓**
↓51%***
↓31%***
↓66%***
*dependent on PTV expansion used
**data not shown
***reduction in percent volume receiving 30 Gy or higher
Prior to (and throughout) treatment, rigorous QA is
essential
9
Impact of Tumor Regression
in Cervical Cancer Patients
QA
Dose verification
Ion chamber, diode arrays, film, or MU
calculation.
Verify setup accuracy on day 1 and then
weekly with orthogonal xx-ray films
Role of daily CBCT is still not clear
Special QA problem is that field sizes
may exceed MLC travel limits
Fields must be split into
movements
2 carriage
Kamath S et al. Med Phys 2004;31:3314
Hong L et al. Int J Radiat Oncol Biol Phys 2002;54:278
Bladder
Tumor
Rectum
Week 1
Tumors
Shrink
Plan
adapts
14 cervical cancer pts
MRI before RT and after 30 Gy
46% GTV
Van de Bunt et al. Int J Radiat Oncol Biol Phys 64(1):189-96, 2006.
Clinical Experience
Bladder
Tumor
Rectum
Prescription
Week 3
Isodose
Between 2/00 and 7/05, >150 women were
treated with IMIM-WPRT in our clinic
Most had cervical cancer, primarily stage
IB
Most underwent definitive RT and, in
stages IB2IB2-IIIB, concomitant cisplatincisplatinbased chemotherapy
Endometrial cancer patients were treated
following primary surgery
ICB was administered in ~50% of women
following IMIM-WPRT
Mundt, Roeske, et al. Gyne Oncol 82(3): 456-463, 2001.
Mundt et al. Int J Radiat Oncol Biol Phys 52(5):1330-1337, 2002.
10
Acute GI toxicity in IMIMWPRT Patients vs. WPRT
Clinical Experience
How do results compare to conventional
treatments?
Acute GI toxicities (Grade 2)
WPRT:
91%
IM60%
p = 0.002
IM-WPRT:
Acute GU toxicities (Grade 2)
WPRT:
20%
IM10%
p = 0.22
IM-WPRT:
100
90
80
70
60
IM-WPRT
WPRT
50
40
30
20
10
0
Grade 0
Grade 1
Grade 2
Grade 3
Mundt et al. Int J Radiat Oncol Biol Phys 52(5):1330-1337, 2002.
Chronic GI Toxicity
Excellent Pelvic Control Rates
90%
80%
70%
60%
50%
IM-WPRT
WPRT
40%
30%
Cervical Cancer
Kochanski J, Mundt AJ. ASCO (2004) 34 stage I-II
cervical cancer pts
21 intact uterus,
13 postoperative Median follow-up = 26.2 months
3-year actuarial pelvic control = 92%
20%
10%
0%
0
1
2
3
On multivariate analysis controlling for age, chemo, stage and site,
IMRT remained statistically significant ( p = 0.01; odds ratio 0.16, 95%
confidence interval 0.04, 0.67)
Endometrial Cancer
Knab B, Mundt AJ. ASTRO (2004) 31 stage I-III
endometrial cancer pts treated postoperatively
Median follow-up = 24.1 months
3-year actuarial pelvic control = 100%
11
Pelvic Control
While encouraging, followfollow-up remains
relatively short and the number of
patients treated remains small
Only with longer followfollow-up and larger
patient cohorts can more definitive
statements be made
Cooperative groups (RTOG, GOG) are
currently developing protocols to
evaluate IMRT in gynecology patients
Conclusions
IMRT is a useful means of reducing the
volume of normal tissues irradiated in
gynecologic patients receiving WPRT
Our initial evaluation indicate a significant
reduction in GI toxicity relative to patients
receiving conventional therapy
Continued followfollow-up and critical evaluation are
required to validate the long term merits of
this approach
Future Directions
Bone marrow sparing IMRT
IGRT and adaptive radiotherapy in
GYNGYN- IMRT
IMRT as a replacement of or
complimentary to brachytherapy
What about the negatives?
IMRT results in higher volumes of normal
tissue receiving lower doses
Increased MUs result in higher total body
doses
Target and tissue delineation are timetime-
consuming
LongLong-term followfollow-up is not available assessing
tumor control and unexpected sequelae
Clinical data are available from only one
institution and while prospective no
randomized comparisons have been performed
12
Acknowledgements
Acknowledgements
J Roeske,
Roeske, PhD – Loyola University
AJ Mundt,
Mundt, MD – Univ of California, San
Diego
LK Mell,
Mell, MD – Univ of California, San Diego
P Georg, MD – Med University Vienna
X. Allen Li, PhD – Med College of Wisconsin
R Miralbell,
Miralbell, MD – Instituto Oncologico
Teknon,
Teknon, Barcelona and Hopitaux
Universitaires,
Universitaires, Geneva Switzerland
Tuning Structures
An anterior structure (AVOID
(AVOID)) to
reduce dose to the small bowel
A SHELL around the PTV to force
conformity
First a 0.5 cm expansion is made on the
PTV (GAP
(GAP))
The SHELL is then a 2 cm expansion
around the GAP
A posterior structure (Rectum
(Rectum--PTV)
PTV) to
reduce the dose to the rectum
13