1. Patrick J. Bolan Assistant Professor
Center for Magnetic Resonance Research, Dept Radiology
University of Minnesota
Radiology/BME collaborators: Mike Garwood, Kamil Ugurbil, Greg Metzger, Tommy Vaughan
Clinical collaborators: Mike Nelson, Tim Emory, Doug Yee, radiology residents
My group: Timo Liimatainen (postdoc), Leighton Warmington (BPhys MS), Avani Chandresekaran (CS MS)
Center for MR Research
UM Medical Center – Fairview

2. Areas of Research Breast MR imaging and spectroscopy
Evaluating cancer treatment response
High-field Body MR Technique 3T 4T 7T

3. Breast MR Spectroscopy
Single-Voxel 1H MRS
Contrast-enhanced MRI
suppressed
water
lipids
Choline
compounds
(tCho)
lipid
invasive ductal carcinoma
Frequency (ppm)
High [tCho]  cancer
(proliferation, cell density, upgragulated transport & kinase activity)

4. Diagnosing Suspicious Lesions at 4T10
ROC cutoff = 1.0 mmol/kg
sensitivity 72%
specificity 83% 8 6
[tCho] (mmol/kg) 4 2
Malignant (n=58)
Benign (n=54)
Normal
(n=5)
Haddadin et al., NMR Biomed 2009

5. Treatment Monitoring with MRS
Size changes takes weeks, metabolic take hours
[tCho] as a Predictive Biomarker
switch drugs / treatment strategy
evaluate new drugs / therapies with short exposure
More robust than diagnostic setting
Lesions are bigger  more SNR
Longitudinal data  self normalizing

6. Objective Responder Baseline LD0 = 2.7 cm Vol0 = 20 cc1 2 3 4 5 6
Baseline
LD0 = 2.7 cm
Vol0 = 20 cc
[tCho]0 = 8.4 mmol/kg
AC X 1
LD24 = 2.7 cm
Vol24 = 20 cc
[tCho]24 = 7.4 mmol/kg B 1 2 3 4 5 6
AC X 4
LDf = 1.5 cm
Volf = 2 cc
[tCho]f = 0 mmol/kg C 1 2 3 4 5 6
ppm
Patient 284 = 8/4/04 – 8/12/04 – 9/23/04
%ΔtCho24 = -12%, %ΔLD = -44%, %ΔVol = -90%
Meisamy et al., Radiology 2004

7. Objective Responders (N=17/21) Nonresponders (N=13/15)
Baseline
Day 1 1 2 3 4 5 6 7 8 9
[tCho] (mmol/kg)
Baseline
Day 1 1 2 3 4 5 6 7 8 9
[tCho] (mmol/kg)
Day-1 ∆[tCho] predicts response:
accuracy 83%
PPV 90%
NPV 77%
Haddadin et al., NMR Biomed 2009

8. I-SPY / ACRIN 6657: Multi-site breast MRS trial
Taxane
Surgery
MRI/MRS #2
MRI/MRS #3
pre-op
MRI/MRS #1
+ 4 cores
Nola Hylton & Laura Esserman, UCSF
Monitoring neoadjuvant chemo with DCE-MRI + MRS
Single voxel MRS, water as internal reference (T2-corrected)
Stratified by field strength (1.5T/3T) & MR2 timing (1 day / 2-4 days)
7 sites accruing, 33% done (March 2009)
UMN
1.5T Siemens UC
1.5T Philips
3T Philips UW
1.5T GE
3T Philips
MSKCC
1.5T GE
UPenn
3T Siemens
UCSF
1.5T GE
Georgetown
1.5T Siemens
UTSW
1.5T Philips
3T Philips
Mayo
1.5T GE

9. ACRIN 6657 MRS Analysis Feedback for training/technical QC
subcontracted ACRIN CORE Lab
QC Phantom analysis

10. Open Breast MRS Trials Pfizer Phase I, CP-751,871 (Yee)
anti IGF-1R
4T CMRR
Novartis Phase II, RAD001/Everolimus (Yee)
mTOR inhibitor
Komen, RFA Ablation (Tuttle)
Dual contrast MRI pre- and post-RFA
I-SPY/ACRIN 6657 (Peterson/Nelson)
AC/Tax
Fairview 1.5T / 3T

11. Methods for High-Field MR: acquisitionx y z
Pulse sequences
-500 Hz
-300
-100
100
300
TE (ms) 45 57
sidebands
water
Novel Acquisition Strategies
Spectroscopic imaging
Novel Hardware
Single-voxel

12. Methods for High-Field MR: analysis
Post-processing, artifact correction
Quantification
data
residual
baseline
model
Spectral Fitting

13. Methods for High-Field MR: imaging
Novel Sequences
Parallel Imaging
Simulation, Optimization
time (s)
A) GRE signal strength
B) Simulated Bolus Gd Concentrations
C) Arterial Signal Intensity
D) Tissue Signal Intensity 1 2 3 4 5 6 7 8
0.05
0.1
0.15
0.2
0.25
[Gd] (mM)
Signal intensity (a.u.)
No T2* effect
1.5T blood
3T blood
7T blood
1.5T tissue
3T tissue
7T tissue 50
100
150
200
250
300
AIF
cancer
normal PZ
0.02
0.04
0.06
0.08
0.12
0.14
0.16
0.18
Ignoring T2*
T2* 7T
0.01
0.03
0.07
Cancer, ignoring T2*
Cancer, w/tissue T2*
Normal PZ, ignoring T2*
Normal PZ, w/tissue T2*
Figure 11 – Simulations showing that R2* effects of Gd-based contrast agents increase at ultra-high field, and substantially impact both blood and tissue signal intensity. See text for details.

14. 7T Body MR from CMRR Prostate Cardiac Whole-body Liver MSK Axial PCr
PCho
PCr Pi
ATP