1

Slide  #1          ©  DSL    

David  S  Liebeskind,  MD  Professor of Neurology & Director of Stroke Imaging Associate Neurology Director, UCLA Stroke Center

No potential conflicts to disclose

   Principles  and  Poten:al  of  Perfusion  Imaging:  

Realizing  Cerebral  Blood  Flow    from  Hemodynamics  to  Permeability    

Slide  #2          ©  DSL    

Objectives

!   Introduce basic concepts in imaging cerebral blood flow !   Describe current and evolving perfusion modalities !   Explore mathematics of hemodynamic measures

Slide  #3          ©  DSL    

Perspective

!   Technical specifications of perfusion imaging require standardization (eg acquisition, processing, interpretation) spearheaded by ongoing work of STIR collaboration

!   Limitations of oversimplified clinical impression (eg ‘MTT is best’, ‘is perfusion proven?’)

!   Balanced perspective on current perfusion imaging tools, adapting use to ultimately improve clinical practice !   Understanding foundation of techniques & pathophysiology !   Using acute MCAO as a model

Wintermark,  M  et  al.  AJNR  Am  J  Neuroradiol.  2008;29:e23-­‐30.  Wintermark,  M  et  al.  Stroke.  2008;39:1621-­‐8.  

2

Slide  #4          ©  DSL    

Practical

!   Perfusion based on theoretical concepts, evident on routine imaging and not technique specific

!   CT/MRI examples as most common modalities !   Relative advantage of multimodal imaging approaches !   Goal is not pretty pictures alone! !   Goal is not debate of CT versus MRI, CBV versus MTT, Tmax

thresholds, mismatch size !   Tools to understand specific pathophysiology and

hemodynamics apparent at the bedside

Slide  #5          ©  DSL    

Vascular transit

Slide  #6          ©  DSL    

Parameters…

!   CBF (cerebral blood flow) – volume of blood flowing through a given unit volume of brain per unit time

!   CBV (cerebral blood volume) – total volume of blood in a given unit volume of brain

!   MTT (mean transit time) – average transit time of blood through a given brain region

!   …many others

3

Slide  #7          ©  DSL    

Gamma variate?

!   BAT !   TTP !   Peak !   FWHM !   CBV !   CBF

Slide  #8          ©  DSL    

Perfusion imaging modalities

!   Ultrasound !   CTP !   XeCT – inhaled xenon and oxygen !   DSC MRI !   ASL MRI !   SPECT – intravenous radiopharmaceutical !   PET – intravenous H2

15O !   DSA

Slide  #9          ©  DSL    

“Perfusion” – CT and MRI

!   Arteries, capillaries and veins !   Ischemic conditions – delay and dispersion as hallmark of

collateral perfusion

4

Slide  #10          ©  DSL    

Dynamic first-pass methodology

!   Exogenous intravascular tracer (eg, iodinated contrast or gadolinium)

!   Assumptions – non-diffusible, neither metabolized nor absorbed

!   BBB permeability derangements require modification

Slide  #11          ©  DSL    

Nondeconvolution approach

!   AIF – arterial input function !   VOF – venous output function

!   CBV derived whole-brain perfused blood volume

Slide  #12          ©  DSL    

Deconvolution

!   AIF – arterial input function !   Signal intensity non-linear with gadolinium concentration !   SVD – singular value decomposition

!   Delay and dispersion of contrast bolus !   Bolus-delay correction !   Calculation of the residue function

!   CBF proportonial to residue height !   CBV proportional AUC

!   Central volume principle !   MTT=CBV/CBF

5

Slide  #13          ©  DSL    

Pitfalls

!   Poor cardiac output or low concentration !   Limited acquisitions that do not capture entire CBV !   Patient motion !   Permeability

Slide  #14          ©  DSL    

CT perfusion

Slide  #15          ©  DSL    

Location, location, location

6

Slide  #16          ©  DSL    

CBV MTT CBF

Slide  #17          ©  DSL    

CT hypodensity

Parsons,  MW  et  al.  Neurology  2007;68:730-­‐736.  

Slide  #18          ©  DSL    

CTP subtleties (ROI, large vessels)

7

Slide  #19          ©  DSL    

Perfusion software

Slide  #20          ©  DSL    

DSC MRI (PWI)

!   Gradient echo (GRE) or spin echo (SE) acquisition !   Measures ΔT2* or ΔT2 !   Indicator dilution theory !   Assumes intact blood-brain barrier !   Preload contrast dose prior to bolus to saturate !   Time-intensity curves

Slide  #21          ©  DSL    

Mismatch

8

Slide  #22          ©  DSL    

ASL

!   Arterial spin labeling – detection of magnetically labeled water in arterial routes

!   Pulsed or continuous !   Rapid relaxation of longitudinal magnetization !   Varying postlabeling delay times

Wintermark,  M  et  al.  Stroke.  2005;36:e83-­‐99.  

Slide  #23          ©  DSL    

Slide  #24          ©  DSL    

Perfusion angiography

!   Perfusion Angiography™ from film or digital copies of angiography performed at any site in a multicenter study, allowing for retrospective comparison of perfusion measures even when multimodal CT/MRI is unavailable.

!   Advantages over CT/MRI perfusion methods include: !   Identification of arterial inflow and venous outflow routes !   Known signal intensity versus contrast concentration !   Absence of recirculation effects !   High resolution !   Multiplanar projections !   Contemporaneous with endovascular procedures !   No additional contrast or radiation exposure

9

Slide  #25          ©  DSL    

Wintermark,  M  et  al.  Stroke.  2005;36:e83-­‐99.  

Slide  #26          ©  DSL    

Wintermark,  M  et  al.  Stroke.  2005;36:e83-­‐99.  

Slide  #27          ©  DSL    

Novel perfusion parameters

10

Slide  #28          ©  DSL    

Delay and dispersion

Slide  #29          ©  DSL    

ASL

Slide  #30          ©  DSL    

SASL

Hendrikse,  J  et  al.  Stroke  2004;35.  

11

Slide  #31          ©  DSL    

Perfusion imaging – CT/MRI

!   Tissue concentration curves provide a wealth of data

Slide  #32          ©  DSL    

Time-to-peak (TTP)

Slide  #33          ©  DSL    

Relative perfusion measures

12

Slide  #34          ©  DSL    

Low perfusion hyperemia

!   1968 – Waltz and Sundt !   1980 – Tomita !   Temporal profiles in CBV during MCAO in cats !   Arteriolar vs venous

Slide  #35          ©  DSL    

Hemodynamic stages

Nemoto, E. M. et al. Stroke 2003;34:2-3

!   CBV as critical parameter !   Bidirectional changes have

defied mathematical models

Slide  #36          ©  DSL    

Wintermark,  M  et  al.  Stroke.  2005;36:e83-­‐99.  

13

Slide  #37          ©  DSL    

Myth of autoregulation

!   In acute MCAO, autoregulation is exhausted before collaterals are recruited

Slide  #38          ©  DSL    

Autoregulation

!   Vast majority of acute ischemic stroke cases involve thromboembolic occlusion, pure hypoperfusion exceedingly rare

!   Distal arterial bed dilates instantaneously following proximal occlusion

!   Autoregulatory changes exhausted before collateral recruitment, distal arterial bed pressure-passive zone

!   Autoregulatory “failure” takes place before 911 called, likely never relevant in clinical practice

Slide  #39          ©  DSL    

Low perfusion hyperemia

!   Hyperemia, or increases in relative CBV values, were evident in regions of “oligemia” in all cases (p<0.001).

!   CPP was decreased, exhibiting a gradient extending deep from the cortical surface in all cases (p<0.001).

Stroke 2008;39:572.

14

Slide  #40          ©  DSL    

Cerebral venous diversion

!   CBV (predominantly venous) diversion to periphery from core/penumbra/”benign oligemia”

!   Starling resistors in parallel, collapsible tubes !   Although incipient arterial insufficiency due to retrograde

collateral flow, predominantly venous process

Slide  #41          ©  DSL    

Venous pathophysiology

Marginal arterial inflow causes venous collapse

Venous blood is displaced (peripheral displacement of CBV)

Downstream resistance increases due to venous collapse

Resistance limits further arterial inflow (reduction in CPP)

Progressive ischemia causes infarct evolution

Infarct grows until CBV can maintain venous patency and preserve CPP (border of penumbra and benign hyperemia)

Slide  #42          ©  DSL    

Benign oligemia revisited

!   CBV underestimates final infarct !   CPP (1/MTT) overestimates final infarct !   CBV x CPP

15

Slide  #43          ©  DSL    

Resistance!

!   Pressure and

Slide  #44          ©  DSL    

FH MRI

Stroke 2008;39:574.

Slide  #45          ©  DSL    

Recanalization and reperfusion

Stroke 2008;39:576.

16

Slide  #46          ©  DSL    

No reflow

!  Distal emboli? !  Complication of

reperfusion therapies?

!  Resistance due to venous collapse?

!  Is complete reperfusion (TICI 3) impossible?

Slide  #47          ©  DSL    

Permeability

Ann Neurol 2007;62:170-6.

Slide  #48          ©  DSL    

CT permeability images

17

Slide  #49          ©  DSL    

DEFUSE permeability

Slide  #50          ©  DSL    

Conclusions

!   Perfusion imaging techniques offer tremendous potential to probe , the critical variable in acute ischemic stroke

!   Ultimate goal to maximize use of perfusion imaging to exploit often subtle pathophysiology and optimize patient care