Quantitative MR Imaging of Acute Stroke

1
Quantitative MR Imaging of Acute Stroke
Risto Kauppinen
2
Acute Stoke Advances since 1990

• Unambiguous diagnosis of acute ischemia by MRI
  (1990)
• Monitoring expansion of ischaemic damage by MRI
  (1991)
• rtPA introduced as a therapeutic agent (1996)
• Availability of CT scans for stroke A E
  (since mid 90s)
• Shift to 3T MR scanners in Clinical Radiology
  (since 2005)
• Computing power has increased and become
  cheaper

3
MRI translation to clinic
DWI (45 min) T2w (65 min)
4
MRI translation to clinic
DWI (45 min) T2w (65 min)
Kuharchyck et al. 1991 Warach et al. 1992
5
Quantitative MRI (qMRI)

• Each pixel in an image is represented by a
  physically
• meaningful number
• - Relaxation times (T1, T2, T1r), ADC,
  haemodynamics etc.
• Normative values
• - Requires acquisition of multiple data points

Griffin JL et al. Cancer Res 633195, 2003
6
qMRI in clinical settings
Griffin JL et al. Cancer Res 633195, 2003
7
qMRI in clinical settings
Griffin JL et al. Cancer Res 633195, 2003
8
Expectations from imaging in clinics
Griffin JL et al. Cancer Res 633195, 2003
Radiol Clin N Am 491-26 (2011)
9
Goals in acute stroke management

• Rescue the penumbra
• by maximising use of available
• treatment strategies
• Patient specific management
• Guide patient triaging
• for investigational therapies

10
qMRI Diffusion MRI in acute ischaemia

• DWI ADC image

-Catastrophic drop in CBF -Energy
failure -Depolarisation -Disturbance in water
homeostasis
11
ADC and Blood Flow
Ischaemia compromised normal blood flow
Grohn et al. J Cereb Blood Flow Metab 20 316,
2000
12
ADC and Blood Flow
Ischaemia penumbra normal blood flow
ADC/Trace decrease in acute ischaemia is not an
ON-OFF event motivation for qMRI
Grohn et al. J Cereb Blood Flow Metab 20 316,
2000
13
qMRI pixelwise histogram of ADCs
ADC
ADC map
Normal Ischaemic Ischaemic compromised
V O L U M E
14
qMRI pixelwise histogram of ADCs
ADC
ADC map
Normal Ischaemic Ischaemic compromised
V O L U M E
15
qMRI Potentials of (q)ADC

• State of tissue beyond perfusion-diffusion
  mismatch as assessed
• by volume (mis)match
• Degree of ischaemia in parenchyma
• Guide patient selection for reperfusion therapy

16
qMRI Potentials of (q)ADC

• State of tissue beyond perfusion-diffusion
  mismatch as assessed
• by volume (mis)match
• Degree of ischaemia in parenchyma
• Guide patient selection for reperfusion therapy

17
MR relaxometry

• Image pixels are either absolute T1 or T2
  relaxation times
• Absolute T1/T2 are much more sensitive to
  parenchymal alterations than either T1w or T2w
  images

18
T1 and T2 in acute stroke
19
Multiparametric qMRI in acute ischaemia
Gröhn O.H.J. et al. MRM 42 268, 1999
20
Acute ischaemia
Transition to irreversible
4
0.2
0.1
2
20
40
60
80

0
40
60
80
20
DT2 (ms)
DDav (10-3 mm2/s)
0
-0.1
-0.2


-2






-0.3


-4


-0.4

-0.5
-6
Time post-ischaemia (min)
Time post-ischaemia (min)
Gröhn O et al. JCBFM 18911 (1998)
21
Time of Stroke Onset by MRI
Vertebral artery occlusions
Remote control graded occlusion
2 days later remote controlled gradual occluder
Jokivarsi et al. Stroke 41 2335-40, 2010
22
Time of Stroke Onset
Vertebral artery occlusions
Remote control graded occlusion
2 days later remote controlled gradual occluder
a) Controllable forebrain ischaemia b) Cortical
hypoperfusion (misery perfusion) c) Middle
cerebral artery (MCA) occlusion
Jokivarsi et al. Stroke 41 2335-40, 2010
23
Time of Stroke Onset
Vertebral artery occlusions
Remote control graded occlusion
2 days later remote controlled gradual occluder
a) Controllable forebrain ischaemia b) Cortical
hypoperfusion (misery perfusion) c) Middle
cerebral artery (MCA) occlusion
Jokivarsi et al. Stroke 41 2335-40, 2010
24
Time of Stroke Onset
Vertebral artery occlusions
Remote control graded occlusion
2 days later remote controlled gradual occluder
Calibration for human brain parenchyma not trivial
a) Controllable forebrain ischaemia b) Cortical
hypoperfusion (misery perfusion) c) Middle
cerebral artery (MCA) occlusion
Jokivarsi et al. Stroke 41 2335-40, 2010
25
Absolute T2 in Acute Stroke
lt3 hours of stroke qT2 _at_1.5T Cut-off
7.5ms Sensitivity 0.824 Accuracy 0.794 ROC
0.757 (ROC(ADC) 0.635)
Siemonsen et al. Stroke 40 1612, 2009
26
T1 in Acute Stroke
30 min 2.5 h 24 h
Very early increase in T1 in Str by 6316ms
(6) Able to discriminate lesion expansion and
non-damaging cortex, despite similar CBF values
in early stroke
27
Multi-parametric MRI of Acute Stroke
Jokivarsi et al. MRM under revision
28
Multi-parametric MRI of Acute Stroke
30 min of ischaemia 24 hours of
ischaemia
Jokivarsi et al. MRM under revision
29
qMR spectroscopy (qMRS)

• Focus on endogenous metabolites

30
State-of-the-art 1H MRS
NAA
Cr
(lac)
1H MRS neurochemical profile at 3T Wilson et
al. Magn Reson Med 65 1 (2011)
31
1H MRS Metabolites in Stroke
NAA
Cr
Lac
Van der Toorn et al. MRM 32 865 (1994)
32
1H MRS in Acute Stroke
Saunders et al. JMRI 7 1116 (1997)
33
Reduced NAA clinical stroke syndrome, more
extensive infarction, severe drop in blood
flow, presence of lactate Increased lactate
large infarcts and reduced NAA
34
Reduced NAA clinical stroke syndrome, more
extensive infarction, severe drop in blood
flow, presence of lactate Increased lactate
large infarcts and reduced NAA
35
Low NAA and high lactate predicted expansion of
DWI lesion
36
qMR in acute stroke Conclusions

• Potentials to provide clinically important data
  from a single exam
• Objective assessment of tissue status (early on)
• Potentially guides clinical management of
  patients
• Aids to maximise use of available therapies
• Allows patient specific treatment protocols

37
qMRI/S in clinical setting

• Standard clinical hardware
• Requires expertise and commitment
• Standardised MR protocols
• Regular QA according to appropriate procedures
• Automated on-line data processing
• Computer-assisted decision making tools

38
Stroke Management in the 21st Century

• Active prevention
• Thrombolysis
• Minocyclin
• Hematopoetic growth factors
• Hypothermia
• Remote preconditioning

39
MR in Evaluation of Acute Stroke Patients in 2020s

• Application specific scanners
• Lower capital costs
• Lower running costs
• Increased availability at AE
• Faster through-put
• Improved data quality
• Tissue status assignment for therapeutic
  procedures
• Improving overall outcome of stroke patients