Title: Planning DCE-MRI Study for ACRIN 6676
1Planning DCE-MRIStudy for ACRIN 6676
- Mark Rosen, MD, PhD
- University of Pennsylvania
- American College of Radiology
2Site Qualification
- All DCE-MR sites must complete Form QC and
receive approval of phantom imaging
- Contact the ACRIN MR Core Lab at 215-940-8880 for
details
3Patient Entry Criteria
- Metastatic or locally advanced ccRCC
- Must be enrolled in treatment trial E2804
- Must have at least one qualifying lesion
- based on size, morphology and location
4Definition of Qualifying Lesion
- Lesion must be in anatomic region which can be
covered using a torso array coil - Lower neck through thighs
- Includes upper arms
- Excludes skull, upper neck, distal extremities
- gt 2 cm in fixed sites
- spine, pelvis, extremities
- retroperitoneum, posterior chest wall
- gt 3 cm in mobile sites
- Lungs, liver, kidneys, anterior chest wall
5Identification of Target Lesion
- Review pre-enrollment imaging results
- Select target lesion for DCE-MRI
- Meets size criteria
- Not largely necrotic
- Away from regions of pulsatility
- (such as heart, hila, left lateral segment of
liver)
6Optimizing Choice of Target
- If more than one candidate lesion, choose lesion
in area with least degree of anatomic motion - Pelvis/extremities/spine/post. chest wall
(preferred) - Retroperitoneal nodes (next best)
- Upper lungs (next best)
- Liver/kidney (next best)
- Lower lungs, ant. chest wall (least preferred)
7DCE-MRI Patient Set-up
- IV placement prior to entry onto scanner
- Position patient supine
- Torso coil centered to lesion location
- Arm position above head
- If arms cannot be positioned above head
- If lesion is posterior, then position arms across
chest - keep IV site patent and flowing
- If lesion is anterior, then position arms at
sides - Landmark as close as possible to area of target
lesion - IV KVO with normal saline during DCE-MRI study
Central venous access can be used if this is the
only available venous access, per institutional
policies and in discussion with referring
oncologist. If central access is not rated for
power injection, hand injection at approximately
1cc/sec can be used.
8Sequence List
- 3 Plane Loc
- Axial T1
- Axial T2 SSFSE
- T1 Maps
- 30 Flip (Auto Pre-scan)
- 15 Flip (NO PRE-SCAN)
- 5 Flip (NO PRE-SCAN)
- Dynamic
- 30 Flip (Inject .1mmol/kg _at_1cc/sec, 20cc flush)
(NO PRE-SCAN) - Run 6 minutes with 20-30 seconds of baseline
- phases depends on temporal resolution
- Delayed Coronal in the SAME OBLIQUE PLANE with
FatSat - Delayed Axial of entire anatomic region (i.e.
lungs, pelvis, etc.)
9Preliminary Sequences
- Localizer
- Axial T1
- Recommended Dual-echo FSPGR/FLASH
- Axial T2
- Recommended SS-FSE/HASTE, TE90, no fatsat
- Notes for axial sequences
- Always during end expiration
- May perform with multiple breath holds if
necessary - Slices thickness 5-8 mm
- Cover entire anatomic region
- e.g. lungs, liver/abdomen, pelvis
10Optimizing Choice of Target
- Identify target lesion from axial scans
- If more than one potential lesion
- Choose lesion with following characteristics
- Non-necrotic, non hemorrhagic
- Well-marginated
- 3-6 cm (better than much larger lesions)
11Oblique slab placement
- Parameters for oblique slab (FOV, slice
thickness, of slices, matrix), are determined
from site phantom imaging and Form QC. - From axial image (T1 or T2) identify middle slice
through lesion - Prescribe oblique 3D slab to bisect both lesion
and aorta - Ensure that lesion is in center of FOV in
superior-inferior direction
12Hints for oblique slab placement
- Lower neck, chest, upper arms
- Include descending thoracic aorta
- Abdomen, upper pelvis
- Include abdominal aorta
- Lower pelvis
- Include contra-lateral common iliac artery
- Thigh
- Include ipsi-lateral common femoral artery
13Hints for oblique slab placement
- Neck lesions
- FOV to include upper and mid-descending thoracic
aorta - Upper arm lesions
- Image entire chest
- In upper pelvis
- FOV includes renal vessel origins
- May need to check separate axial slices for
correct slab placement including target lesion
AND vessel - Use coronal localizer to confirm appropriate
superior-inferior placement of oblique slab
14- Avoiding severe phase wrap artifact for DCE-MRI
protocol -
- 1) Patients who can tolerate the arms over head
position should do so for the duration of the
imaging. -
- 2) If patients cannot tolerate this positioning,
arms should be left at the side or lightly
crossed over chest (avoiding kinking of IV line
if placed in antecubital fossa). Arm positioning
in these cases should be based on the following - a) if target lesion is posterior/lateral, place
arms across chest - (Figure a)
-
- b) if target lesion is anterior, place arms at
side - (Figure b)
Figure a
Figure b
15Slab Placement Examples
16Lung metastases
17Peri-aortic LN metastasis
18Gall bladder fossa metastasis (melanoma)
19Colon cancerMetastasis to peritoneum
20Colon cancerMetastasis to liver
21T1 Mapping Series
- Always during suspended end expiration
- Slab prescription as described
- Begin with flip angle 30 degrees
- Perform normal auto pre-scan procedure
- Use 3 NEX for breath-hold (10-15 seconds)
- Inspect for phase encode/wrap artifacts
- Can increase FOV and rerun if required
- Maintain new FOV for all subsequent oblique slab
scans - Repeat above for flip angles 15, then 5
- Do NOT repeat automated pre-scan for flip 15 5
- Do NOT adjust receiver or transmitter gain
22DCE-MRI Series Set-up
- Same slab prescription as T1 mapping
- NEX lt 1 (per phantom imaging)
- Flip angle 30
- Do NOT auto pre-scan between series
- Acquisition for 6 minutes (see chart next page)
- Non breath-held
- Instruct patient on quiet tidal respiration
- Keep images in same series (if possible)
- Can break into several series if needed
- No delay between series once imaging begins
23- After installing parameters, the chart below will
determine the number of scan phases/acquisitions/r
epetitions (see Form QC) - Set multiphase to total of scan
phases/acquisitions/repetitions - Set NO delay between phase acquisitions!
- If machine error for multiphase imaging (as
determined during phantom imaging) - Divide dynamic run into 2 series, 3 min each
- Each runs half the number of repetitions
24DCE-MRI - Injection
- Gadolinium contrast 0.1mg/kg
- Injection rate 1.0 cc/sec
- Saline flush, 20cc _at_ 1.0 cc/sec
- Begin imaging pre gadolinium injection
- Continue for 20-30 seconds before injection
- Inject and continue scan 6 minutes
25Delayed post-gadolinium scans
- Post gad T1 obl. cor. 2D fatsat FSPGR/FLASH
- Oblique coronal, same thickness as DCE-MRI slab
- No inter-slice spacing, interleaved acquisition
- Same FOV and matrix as DCE-MRI slab
- Prescribe at same location/obliquity/orientation
- NEX1, Breath-hold, end tidal expiration
- Post gad T1 axial 2D fatsat FSPGR/FLASH
- Same slice thickness/spacing as pre-gad T1 axial
- Run through entire anatomic region
- e.g. liver, lungs, pelvis
- May run in several breath holds (end tidal exp.)
as required
26Parameter Chartavailable at www.acrin.org
27Cross-reference image of DCE-MRI
- Choose axial image (T1, T2, post-gad) which
demonstrates target lesion - Cross-reference obl. 3D slice locations
- Remove annotations with patient identifiers!
- Screen capture for submission to ACRIN
Ax T2 image w/target lesion
Same with cross referencing
28Dynamic DCE-MRI examples
29Colon cancerMetastasis to peritoneum
30DCE-MRI, peritoneal metastasis
31Kt 0.11889 /- 0.049131 Ve 0.39289 /-
0.14408 Kt 0.1163 units 1/min ve 0.3870
units
Kt 0.057848 /- 0.013971 Ve 0.35429 /-
0.10141 Kt 0.0575 units 1/min ve 0.3505
units
PRE
POST
32Colon cancerMetastasis to liver
33DCE-MRI, liver metastasis
34Kt 0.31476 /- 0.22599 Ve 0.22779 /-
0.10057 Kt 0.3015 units 1/min ve 0.2070
units
Kt 0.18454 /- 0.15173 Ve 0.16865 /-
0.077324 Kt 0.1235 units 1/min ve 0.1246
units
PRE
POST
35Site tumor/artery segmentation
- Site PI must provide their own segmentation of
tumor and artery from each DCE-MRI study - Choose portion of the dynamic gadolinium study
that best demonstrates target lesion - (Often early gadolinium enhancement phase)
- Use repetition with minimal motion artifact
- Refer to pre-treatment study when segmenting
post-treatment study - Attempt to include similar portions of tumor
36Site tumor/artery segmentation
- Identify all slices from DCE-MRI study which
include portions of the target lesion - On local workstation, draw ROI (elliptical or
freehand) to outline entire tumor - Include all tumor-bearing slice locations
- Avoid major vessels or area of artifact
- Repeat for maximally enhancing artery
- Indicate segmentation slices on DP form
- See example on next page
37Representative image from DCE-MRI study (early
enhancement phase)
38(No Transcript)
39Different slice
40Tumor segmentation (10 slices out of 12 slice
DCE-MRI scan)
41Segmentation of Aorta
In general, choose artery on central 2-4 slices
within slab. Choose only maximally enhancing
regions to avoid partial volume effects.
42Complete DP form
Questions 6-7 will capture site Radiologists
analysis information
TECHNOLOGIST
RADIOLOGIST
Page 1
Fax to 215-923-1737
Page 2
43DP page 2 from previous example study 12 slice
DCE-MRI 11/12 slices with tumor ROI 3/12 slices
with arterial ROI
Arterial ROI
Tumor ROI
Leave later rows blank if 3D slab has fewer than
12 slices
44Image Submission
- Complete DP Form
- Submit via email or fax
- Transfer images via secure ftp
- CD-ROM Shipment or TRIAD
- Contact ACRIN at 215-940-8880 for more
information regarding electronic image
submission. - Or contact Triad-Support_at_phila.acr.org
- Include screen captures of the following
- Graphic Rx of Oblique scan cross-referenced on
axial slice - Screen captures with tumor arterial ROIs
45Remember
- Choose optimal lesion for imaging
- Patient positioning and comfort are key!
- Breath-held T1 maps, then
- Instruct for quiet breathing during dynamic
gadolinium DCE-MRI series - Replicate parameters and slab prescription on
follow-up study - Have prior images available for review during
follow-up DCE-MRI scanning session
46Questions
- Donna Hartfiel, data manager
- dhartfiel_at_phila.acr.org
- Jim Gimpel, chief technologist
- jgimpel_at_phila.acr.org
- Mark Rosen, principle investigator
- rosenmar_at_uphs.upenn.edu