mri guided radiotherapy for cervical cancer cervical ...€¦ · mri guided radiotherapy for...
TRANSCRIPT
1
Ina Jürgenliemk-Schulz
MRI guided radiotherapy for cervical cancer UMC Utrecht clinical experience
11.06.2014 Advanced Radiotherapy Techniques
Cervical cancer
• 650 – 700 new patients per year in The Netherland
• About 300 advanced stages
• 20-25 patients treated in UMCU
• 2rd most frequent malignancy in women worldwide
• 1st in India (200 000 new patients per year)
Traditional radiotherapy for advanced cervical cancer
Combination of EBRT and Brachytherapy
• 45 - 50 Gy external beam radiotherapy to the pelvic/PAO region – Conventional or conformal EBRT
• About 25 - 30 Gy brachytherapy to the central part of the pelvis – Point A based treatment planning
• Optional – External boost in case of parametrial invasion or affected lymph nodes
• Overall treatment time about seven weeks
Results of traditional RT not satisfying Logsdon, Int J Radiat Oncol Biol Phys. 1999
• Loco-regional control for locally advanced cervical cancer 53 %
• 5-year disease-free and overall survival less than 45 %
• Grade 3 - 5 toxicity about 17 %
WHO alert in 1999
• Improvement by combining radiotherapy with chemotherapy
• 5 trials published, about 12 % benefit in 5-year DSS
• 2001 Meta-analysis; Green et al. Lancet
• 16 % benefit 5-year DSS
• increased grade 3 and 4 hemato and gastro-intestinal toxicity
Outcome related to periods
RT (± HT) ≤ 1999 RT ± CT 1999-2004
Number of patients 95 54
Median FU in months 49 46
Years of treatment 1983 – 1990 1999-2004
FIGO stages IB1 - IVa IB1 - IVa
Planning aim EBRT Gy 46 45
Planning aim BT (LDR/PDR) Gy 24 35
Total planning aim dose Gy 70 80
Local control % at 3 years 62 81
NED % at 3 years 52 53
Overall survival % at 3 years 47 54
UMCU data Conclusions
Concomitant chemo might not be enough
Can we improve the radiotherapy technique?
Might MRI guidance help?
Aim
increase loco-regional control
increase survival rates
decrease toxicity
2
Challenges
• Combination of macroscopic tumor (primary and lymph node metastases) and a large elective area (uterus, parametria,
vagina, lympatics)
• Internal motion of primary tumor and organs at risk
• Tumor regression during treatment
• Large difference between tolerance dose to organs at risk (small bowel) and required dose to GTV
OAR
OAR
OAR Small bowel
Cervical tumor
MRI guided UMCU Developments from 2004 on
Imaging • From X-ray to MRI guidance
Radiotherapy technique • EBRT: From conventional irradiation to IMRT and IGRT
• Brachy: From point A to GEC ESTRO recommendations
From 2D to 4D adaptive treatment approach !
Impact of MRI on EBRT
EBRT imaging modalities
Clinical investigation at time of diagnosis
Exophytic tumor in the whole cervix
Infiltration of the middle part of the left parametrium
No infiltration into the vagina
Cystoscopy negative
No hydronephrosis
Rectum not involved
FIGO IIB
EBRT Imaging modalities for nodes
Match PET-MRI
Imaging to define tumor spread CT-PET
EBRT Imaging modalities for cervical tumor
Cervical tumor as depicted on CT or MRI
• FIGO IIB, cT2b
Contrast enhanced CT T2 weighted MRI
3
EBRT Imaging modalities
Cervical tumor as depicted on CT or MRI • FIGO IB2, cT4a uterine corpus, rectum and bladder wall
Advantage of MRI to define primary GTV
MRI versus clinical investigation and CT Overall staging accuracy of MRI is superior to clinical examination or CT with respect to
• parametrial infiltration
• infiltration into the uterus
• infiltration into bladder and rectum Boss et al. 2000, Ozarlak et al. 2003, Bipat et al. 2003
Sagittal Axial
GTV = intermediate or high signal on T2 weighted MR images
Diagnostic imaging for cervix
Staging of uterine cervical cancer with MRI:
Consensus guidelines of the European Society of Urogenital Radiology Eur Radiol 2011,21,1102
Morphologic T2 weighted images
a) Hyperintense lesion invading cervical stroma
b) Low signal intensity excluding parametrial invasion
c) Enlarged lymphnode in para-aortic space
c
Functional imaging
DWI images ADC map (B 800) Review: Imaging cervical cancer: recent advances and future directions.
Curr Opin Oncol. 2011 Sep;23(5):519-25.
“Particularly, functional MRI techniques have improved accuracy of disease staging …….”
“Once standardized and validated, the techniques should enable individualized patient treatment and optimization of outcome.”
Treatment volume definition
X-rays CT MRI
Impact of imaging modality on primary tumor depiction
- PTV - CTV - GTV
Contouring guidelines
Consensus contouring guidelines for 3D treatment planning available for CT and MRI approaches (Taylor 2005, Small 2008, Lim 2011)
• Bare chance on more conformity for target delineation and delineation of organs at risk
• Some discrepancies still exist
• At at this moment not clinically proved
Delineation of the targets and OAR
Consensus guidelines Lim et al 2011
4
Consensus guidelines “Small” versus “Taylor” Small, Red Journal, 2008, 71, 428 Taylor, Red Journal, 2005, 63, 1604
Impact of MRI on EBRT treatment techniques
Conventional
Based on bony anatomy
AP/PA fields
3 - 4 treatment fields
Image based
3DC-RT or IMRT
3DC-RT : 3 - 4 treatment fields
IMRT: 5 or more fields
MRI based treatment planning
3DC-RT
4-field box
box
IMRT
7 beam IMRT
IMRT
VMAT with SIB to pathological nodes
MRI based treatment planning
0
100
200
300
400
500
600
700
bowel 30 Gy bowel 42.75 Gy
vo
lum
e in c
c
conventional
conformal
imrt
0
20
40
60
80
100
120
bladder 30 Gy bladder 42.75 Gy
conventional
conformal
imrt
0
20
40
60
80
100
rectum 30 Gy rectum 42.72 Gy
conventional
conformal
imrt
vd Bunt 2006
IMRT allows organ sparing !
Improved Organ sparing
Chance on • Reduced morbidity • Improved QoL
But !!
Impact on treatment margins Patient related:
• Motion and deformation of tumor and OAR • Tumor regression
Externally:
• Contouring uncertainties • Position verification method • Set-up uncertainties
PTV
Courtesy Taran Paulsen-Hellebust
Higher conformity gives higher chance on geographical miss
5
Inter-fraction motion of primary GTV
GTV CTV PTV
• 5 consecutive MRI’s during EBRT • Impact of changes in bladder and
bowel filling on position changes of uterus
• Not only one organ is responsible
Low impact
Low impact
High impact of bladder and bowel
High impact of bladder Van de Bunt et al 2008
Inter-fraction motion of primary GTV
• Extreme difference in bladder filling • Extreme difference in position of the uterus • 6 – 40 mm in different publications
CTV PTV
Pre-treatment MRI MRI in first week of EBRT
Intra-and inter-fraction motion
week1 week4
• 4 consecutive sagittal images taken within 30 minutes of MR imaging
• Impact of patient’s movement en pelvic rotation!
• Impact of bowel filling
Kerkhof et al 2009
week1 week4 Pre-treatment week2
• Impact of changes in bladder and bowel filling, pelvic rotation
• Impact of volume changes of uterus
Kerkhof 2008
Intra-and inter fraction motion
Inter-fraction motion of node GTV
• Affected nodes change their position (< 8 mm) • Position shifts less extensive than for primary GTV
Maximal margins (computer)
0
1
2
3
4
5
6
7
8
9
med lat ant post sup inf
marg
in (
ml)
• 48 nodes, 15 patients • Position shifts in 6 directions
Schippers, submitted
Position shifts of primary tumor and nodes
6
Motion depicted on repetitive MRI
• Organ motion and deformation contribute essentially to the treatment margin concept !
• Order of magnitude between position shifts of primary tumor and affected nodes is different !
MRI to monitor primary tumor regression
• Primary tumors individually shrink during EBRT
• Impact on elective CTV and PTV is low
Before treatment After 30 Gy EBRT
0
100
200
300
400
500
600
700
800
vo
lum
e i
n c
c
GTV CTV PTV
pre-treatment intra-treatment
GTV 49.2%
CTV 13.4%
PTV 9.7%
Average decrease in volume after 30 Gy
MRI to monitor nodal tumor regression
At diagnosis
Enlarged (15 mm) Irregular shape Inhomogeneous
4th week of EBRT
Partial response in week 4 EBRT total dose = 55 Gy (45 + 10)
Node > 10 mm on MRI, PET positive, boost given
Treatment margins
PTV
bowel
rectum
GTVprim
GTV nodes
GTV
CTV
Recommended CTV-PTV margins for elective EBRT
• For primary CTV 15 – 20 mm • For nodal regions 10 mm • Position verification on bony landmarks, off-line protocol • Under investigation for soft tissue contrast, on-line protocols, replanning
Impact of treatment margins on coverage
Planning aim dose:
50 Gy
Comparison:
Four Field Box Large margin IMRT 10-20 mm Small margin IMRT 5 mm For GTV en CTV not that much difference between planning aim and delivered dose
Pelvic Radiotherapy for Cancer of the Cervix: Is What You Plan Actually What You Deliver?
Lim et al 2009
Ongoing
Image based margin concept for highly conformal adaptive EBRT
• Currently subject of investigation (repetitive CT/MRI)
• Promising developments • Image based adaptive “plan of the day” strategies
• Several treatment plans based on repetitive CT with variable bladder filling (Rotterdam group)
• Margin choice depends on position verification method
• Bony anatomy, soft tissue, internal markers • on-line, off-line
7
Ongoing research
Adaptive RT: variable bladder filling scans • Separate movers/non-movers: large/smaller margins →
direct clinical gain
• CBCT based introduction and monitoring
• Repeat MR
30 mm 5 mm
Position verification
Intergration of Cone-beam CT and accelerator
Intergration of MRI and accelerator
EPID • Bony anatomy • Gold markers
Possible gain of an MRI accelerator
Online MRI guidance Reduction CTV-PTV margin to 4 mm Clear reduction of in-field amount of OAR
Kerkhof 2008
MRI in postoperative situation
15 mm 0 mm Superior
28 mm 11.5 mm Posterior
13 mm 4 mm Anterior
10 mm 3 mm Right lateral
20 mm 5 mm Left lateral
max. in 95% median Margins
Endometrium en cervical cancer after hysterectomy 5 consecutive MRI’s during EBRT
5 CTV’s with one encompassing PTV
Toet et al 2007
Impact of MRI on Brachytherapy
Regions of interest for Brachytherapy
• macroscopic tumor after EBRT
• suspected microscopic tumor extension
• pre-treatment tumor extension
• most exposed volume of rectum, sigmoid and bladder
8
From ICRU rules to GEC ESTRO recommendations
According to Manchester System/ICRU 38 • X-ray guided (2D)
• Point A
• ICRU rectum point
• ICRU bladder point
• Point B
• 60 Gy volume
• TRAK
According to GEC ESTRO and ABS
recommendations • MRI/CT guided adaptive
• GTV (macroscopic tumor)
• HR-CTV (GTV + suspected microscopic tumor)
• IR-CTV (pre-treatment extension)
• 2cc rectum
• 2cc sigmoid
• 2cc bladder
• TRAK
Pötter et al., Radiotherapy and Oncology, 2006, 78, 67
ICRU 38, 1985
X-ray guided brachytherapy
Standard type Fletcher application for LDR, PDR or HDR treatments
Standard pear shaped isodose distributions
5.1 cm
4.9 cm
3.1 cm
Iinitial
aEllipsoid formula applied: V=width*thickness*height*0.5
At BT
4.7 cm
2.9 cm
3.6 cm
aV= 24.8 cm3
aV= 38. cm3
MRI guided brachytherapy MRI guided delineation
MRI with applicator in situ Regions of interest delineated
sagittal
axial
coronal
GTV
HR-CTV
IR-CTV
bladder
rectum
bowel
X-ray guided versus MRI guided intracavitary brachy planning
Ke1 hr-ctv bladder rectum sigmoid bowel
in cGy (PD 60 cGy) D90 D2 D2 D2 D2
conventional plan 48.4 60.0 43.1 51.2 49.8
optimized plan 53.1 62.0 49.3 53.6 43.5
Improved dose volume parameters !
Aimed tumor dose EBRT+ BT > 84 Gy (60cGy per PDR pulse)
OAR constraints D2cc • < 90 to Bladder (63 cGy pp)
• < 75 to rectum, sigmoid, bowel (53 cGy pp)
Locations with inadequate coverage by intracavitary applications
A1 A2
Inverse optimization
Extra needles
25% of the
cases
sigmoi
d bowel
s
10% of the
cases
90% of the
cases
Results of planning study with 20 applications
75% of the
cases
9
Adaptation of a the applicator
Applicator produced by Nucletron
150
250
Start clinical introduction July ‘07
Two BT applications First one without needles
Second with needles, if necessary
• Pre-planning for 2nd application based on dose distribution and DVH analysis of first application:
• Needles necessary?
• Where? How deep?
From 3D to 4D
Example
2 needles left side
treatment 3 hr-ctv bladder rectum sigmoid bowel
in cGy (PD 60 cGy) D90 D2 D2 D2 D2
conventional plan 38,3 72,9 39,0 27,6 38,5
optimized plan IC 35,0 64,3 32,3 26,9 38,4
optimized plan IC/IS 3ndl 68,7 56,2 48,5 36,6 32,6
Improved coverage by combining intracavitary and interstitial approach
Optimized IC
Optimized IC/IS 3 needles used
Aimed tumor dose EBRT+ BT > 84 Gy (60cGy per PDR pulse)
OAR constraints D2cc • < 90 to Bladder (63 cGy pp)
• < 75 to rectum, sigmoid, bowel (53 cGy pp)
Effect of IC/IS applicator on HR-CTV dose during learning curve
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
D90
hr-
ctv
in G
yab
10
patient nr
optimized IC
optimized IC/IS
Effect of IC/IS applicator
Optimized plan IC/IS – Optimized plan IC
(in Gyab10)
N=15 avg SD
D90 5.3 3.3
D100 3.9 2.6
EQD2 in Gyab ‘optimized’ plan
average SD
HR-CTV 88 7
Bladder 83 6
Rectum 67 6
Sigmoid 61 5
Bowel 64 9
Dosimetric gain for one application Total dose EBRT + BT
10
D90 HR-CTV as function of Volume HR-CTV
50
55
60
65
70
75
80
85
90
95
100
105
110
115
120
0 20 40 60 80 100 120 140
volume HR-CTV in cc
D9
0 H
R-C
TV
Gyab
filmplan
filmpl scaled
Log. (filmplan)
Log. (filmpl scaled)
50
55
60
65
70
75
80
85
90
95
100
105
110
115
120
0 20 40 60 80 100 120 140
volume HR-CTV in cc
D9
0 H
r-C
TV
Gyab
optimized
IC/IS
Log. (optimized)
MRI guided optimization and additional interstitial component allow to cover larger HR-CTV volumes adequately, has dosimetric benefit
standard optimised
Uncertainties OAR motions and volume changes during HDR
• Repetitive images around HDR fraction • One for treatment planning, one directly before and one after irradiation
• Protocol for bowel preparation
• Protocol to controlled bladder filling
• Differences in bladder filling
• Differences in rectum filling: Gas! • After adaptation comparable DVH parameters
MRI for HDR –treatment planning
MRI pre-irradiation about 1.5 hours later
MRI after insertion of rectal probe
MRI after HDR about 20 min later
Work in progress
Repeated and interventional imaging easier to realize when MR machine is part of brachytheater
• Application, imaging and irradiation in the same room
• Offers possibility to adapt when necessary
• Is less time consuming
• Offers possibility to work with new tools
Courtesy R.Moerland R.Schokker
Difference PDR / HDR
Christel Nomden et al
Treatment periods in UMCU
MRI guided EBRT and BT from 2006 on
46 Gy EBRT 24 Gy LDR-BT 14 Gy Boost (P/N)
Traditional treatment prior 1999
Eventually hyperthermia, IC BT, parametrial boost if indicated
45 Gy EBRT 35 Gy PDR-BT
Traditional treatment 1999-2004
Combined with cisplatinum, ICBT, parametrial/nodal boost if indicated
45 Gy EBRT 40Gy PDR-IGBT 10-14 Gy Boost (N)
MRI guided treatment 2006-2011
Combined with cisplatinum, IC or IC-IS BT, nodal boost if indicated
45 Gy EBRT 40Gy HDR-IGABT 10-14 Gy Boost (N)
MRI treatment 2012-currently
Combined with cisplatinum, IC or IC-IS BT, nodal boost if indicated
14 Gy Boost (P/N)
Retro-_
Impact of MRI guidance on outcome
UMCU RT (± HT) ≤ 1999
RT ± CT 1999-2004
RT ± CT 2006-2009
(Retro-EMBRACE)
RT ± CT 2008-
(EMBRACE)
RT ± CT 2015-
(EMBRACE II)
Number of patients
95 54 46 68
Median FU in months
49 46 41
Years of treatment
1983 – 1990 1999-2003 2006-2008
FIGO stages IB1 - IVa IB1 - IVa IB1 - IVa
Planning aim dose Gy EQD2
> 70 > 80 > 84
Local control % 62 81 93
PFS % (3 years) 52 53 71
OS % (3 years) 47 54 65
Late ≥ G3 morbidity %
ns 20
(RTOG)
9.5
(CTCAE 3.0)
11
Ongoing and future UMCU effort
IGART for primary and post-op treatments • Treatment margin reduction
• SIB techniques (vulva, vagina, pelvic and PAO nodes)
• MRI based daily position verification
• Deformable registration tools
• Adaptive re-planning strategies
Affected nodes/Nodal recurrence • Detection of small nodes possible on anatomical MRI
• Impact of PET-CT, contrast enhanced or functional MRI
• More precise dose calculation from EBRT and BT
• Better understanding of dose response relation
All Node positive Node negative
n=46 n=20 n=26
Three years outcome % % % P
Overall survival 65 50 77 0.032
Cancer specific survival 76 60 89 0.022
Local control 93 94 92 0.799
Pelvic control 84 78 88 0.370
Distant metastasis free survival 77 68 85 0.143
Progression free survival 71 53 85 0.013
Nodal recurrences
13 % UMCU MRI guided treatment 2006-2011
Patterns of regional recurrence after standard chemo-radiotherapy
Beadle et al 2010, MD Anderson
> 1800 cervical cancer patients, definitive radiotherapy • Pelvic control about 82 % • Local control about 86 %
Inside the field = dosimetric problem Outside and marginal = geographical miss
Node example on MRI
At diagnosis
Irregular appearance
and irregular border
4th week of EBRT
Nearly gone
in week 4
EBRT dose 45 Gy
6 weeks after treatment
Nearly no rest
6 weeks after
treatment
12 months after treatment
Nodal failure
Node 8 mm, PET negative, no boost
Summary
• MRI guided treatment for cervical cancer
– Supports better defining the disease at time of diagnosis
– Supports monitoring motion and deformation during treatment
– Supports adaptive treatment approaches
– Supports development of advanced EBRT and BT techniques
– Results in improved local control and survival
– Will hopefully help to improve nodal control
– Will hopefully help to reduce morbidity and improve QoL (also important for intensifying chemo)
UMCU Team
Judith Roesink Robbert Tersteeg Christel Nomden Astrid de Leeuw Petra Kroon -van Loon Rien Moerland Hans de Boer Marielle Phillipens Nico van den Berg Many others Radiologists Gynecologists Medical Oncologists Pathologists