1. Perfusion Imaging Grand Rounds January 18, 2017
Gloria J. Guzmán Pérez-Carrillo, MD, MSc
Neuroradiology Section, Department of Medical Imaging

2. Nothing to disclose

3. Overview
Clinical uses of perfusion imaging, in particular stroke and tumor evaluation
CT perfusion
MR perfusion
Dynamic Contrast-Enhanced (DCE)
Dynamic Susceptibility Contrast (DSC)
Arterial Spin Labeling (ASL)
Intravoxel Incoherent Motion (IVIM)
Summary of MR perfusion

4. Learning objectives
To define the different types of perfusion techniques
To understand their clinical applications in stroke and tumor evaluation
To define their technical limitations 

5. Perfusion modalities and clinical uses
Stroke work-up
CT Perfusion
MR Dynamic Susceptibility Contrast (DSC)
Tumor imaging MR
Dynamic Contrast-Enhanced (DCE)
Dynamic Susceptibility Contrast (DSC)
Arterial Spin Labeling (ASL)
Intravoxel Incoherent Motion (IVIM)
Multiple other clinical indications: Vasospasm, Moya Moya, etc.
Briefly

7. CT Perfusion

8. CT perfusion, how to perform?
Arterial input function (AIF)-A2 and venous output function (VOF)-superior sagittal sinus after contrast

9. CT perfusion, how to perform?
Arterial input function (AIF)-A2 and venous output function (VOF)-superior sagittal sinus

10. CT perfusion parameters
Cerebral blood flow (CBF):
Measured in units of milliliters of blood per 100 g of brain tissue per minute and is defined as the volume of flowing blood moving through a given volume of brain in a specific amount of time
Cerebral blood volume (CBV):
Measured in units of milliliters of blood per 100 g of brain and is defined as the volume of flowing blood for a given volume of brain
Mean transit time (MTT):
Measured in seconds and defined as the average amount of time it takes blood to transit through the given volume of brain
Time to peak (TTP):
Measured in seconds as the time it takes for the bolus to reach the highest HU
CBF= CBV/MTT

11. CT perfusion parameters

12. Tmax
Tmax is theoretically the arrival delay between the AIF and [contrast] 
Used either as an independent measure or equivalent to MTT
Multiple factors can affect its calculation (deconvolution method,bolus delay, bolus dispersion, etc.)

13. CT perfusion penumbra CBF= CBV/MTT
rCBF
CBF= CBV/MTT
Thus, when there is increased MTT, there is decreased CBF
Decreased CBF may or may not be associated with decreased CBV. If mismatched = PENUMBRA
rCBV
MTT

14. CT perfusion, quantitative?
In the clinical setting, for now only qualitative due to differences between vendors and post-processing software
Quantitative
MTT increase > 145%
CBF < 25 mL x 100 g-1 min-1
CBV < 2 mL x 100 g-1
TTP > 6 seconds

15. CT perfusion pearls
In normal patients, cortex ALWAYS has higher CBV and CBF as compared to white matter
TTP is not a reliable measurement due to hemodynamic variability
MTT is the most sensitive parameter
Posterior circulation ischemia harder to image due to prolonged time of bolus arrival
Only a few levels acquired, not entire brain volume, be aware perfusion defects could be missed

16. CT perfusion, pitfalls of acquisition
Suboptimal contrast bolus
Cardiac arrhythmia
Severe intracranial vascular narrowing
Multiple intracranial emboli
Lack of adequate flow within the vessels of the circle of Willis
Inappropriate ROI over vessels

18. MR Perfusion

19. Why perfusion in tumors
Differentiation between tumor recurrence and post-treatment change. Seminal article Sugahara et al., AJNR 2000

20. Why perfusion in tumors
Predictive of survival: Seminal article Meng Law et al., Radiology 2008

21. Dynamic Contrast-Enhanced (DCE)

22. DCE, what is it? “Permeability” MRI
Takes advantage of increased T1 relaxivity of Gd
Based on a two-compartmental (plasma space and extravascular-extracellular space) pharmacokinetic model
Resulting signal intensity–time curve reflects a composite of tissue perfusion, vessel permeability, and extravascular-extracellular space
Depicts the wash-in, plateau, and washout contrast kinetics of the tissue

23. DCE, why it works
Increase in the rate of T 1 relaxation is proportional to the concentration of the contrast agent.
Relate signal enhancement in T 1 -weighted images to the tissue contrast agent concentration. 
Extracellular contrast media diffuse from the blood into the EES of tissue at a rate determined by tissue perfusion and permeability of the capillaries and their surface area.

24. DCE, parameters Transfer constant (k-trans)
Fractional volume of the extravascular-extracellular space (ve)
Rate constant (kep, where kep = k trans /ve)
Fractional volume of the plasma space (vp)

25. DCE, parameters Transfer constant (k-trans)
Fractional volume of the extravascular-extracellular space (ve)
Rate constant (kep, where kep = k trans /ve)
Fractional volume of the plasma space (vp)

26. DCE, parameters Transfer constant (k-trans)
Fractional volume of the extravascular-extracellular space (ve)
Rate constant (kep, where kep = k trans /ve)
Fractional volume of the plasma space (vp)

27. DCE, K-trans
K-trans most commonly used. Due to BBB, normally there is very low permeability (i.e. leakiness) in the cerebral blood vessels. Thus, any increased in permeability is due to increase in number of abnormal BV’s due tumor angiogenesis (VGEF pathway)
Tumor vessels are further characterized by the absence of muscularis propria, widened interendothelial junctions, and a widely discontinuous or absent basement membrane, results in increased permeability
Increased abnormal BV’s are usually insufficient to feed tumor, thus regional area of necrosis and hypoxia

28. DCE, how to perform? Perform baseline T1 mapping
Acquire DCE MR perfusion images
Convert signal intensity data to gadolinium concentration, determine the vascular input function, and perform pharmacokinetic modeling

29. DCE, baseline T1 mapping
Assumed that the change in T 1 is directly proportional to the tissue contrast agent concentration, however, this is not always true
Baseline mapping of T 1 and equilibrium magnetization as a preliminary step in DCE-MR imaging analysis needed to correct for the nonlinear relationship between signal intensity on dynamically acquired DCE-MR imaging images and underlying gadolinium-based contrast agent concentration
In our case, we do variable angle technique

30. DCE, dynamic imaging
Done with spoiled GRE to correct for high T2 component of GRE alone, with SNR trade-off
We use three-dimensional time-resolved imaging of contrast kinetics and time-resolved angiography with stochastic trajectories are K-space subsampling techniques that (in contrast to older rapid imaging techniques such as keyhole imaging) sample both the center and periphery of K-space during the dynamic image acquisition, allowing for higher spatial resolution

31. DCE, how to perform?

32. DCE, how to interpret
Qualitative: higher leak on K-trans maps on affected side vs. contralateral tissue
Quantitative:
K-trans: high-grade tumors (median 0.89, 25th to 75th quartiles 0.46 –2.67) vs. low-grade tumors (median 0.09, 25th to 75th quartiles 0.04 –0.13).
Kep: high-grade tumors (median 6.76, 25th to 75th quartiles 3.77–16.88) vs. low-grade tumors (median 0.66, 25th to 75th quartiles 0.29 –1.04).
Ve: high-grade tumors (median 0.12, 25th to 75th quartiles 0.11–0.15) vs. low-grade tumors (median 0.23, 25th to 75th quartiles 0.19 –0.26).

33. DCE, how to interpret
Dynamic curves, akin to breast cancer MR studies

34. DCE, pitfalls
If there is a lot of underlying tissue damage from surgery/treatment, there will be a lot of “leakiness” and K-trans might be spuriously elevated
Complexity in image acquisition and pharmacokinetic model post-processing, user-dependence, and lack of widely available and easy-to-use post-processing software are all drawbacks
Lower temporal and spatial resolution compared to DSC

35. Dynamic Susceptibility Contrast (DSC)

36. DSC, what is it?
First pass of a bolus of gadolinium-based contrast agent through brain tissue is monitored by a series of T2*-weighted MR images
Susceptibility effect of the paramagnetic contrast agent leads to a signal loss in the signal intensity– time curve, which is converted into a concentration-time curve on a pixel-by-pixel basis

37. DSC for stroke, similar to CT
rCBF
rCBV
MTT
TTP

38. ADC
Flair
rCBV
T1Gd
DSC, how to perform

39. DSC, how to measure
Use rCBV, do ROI over lesion (A) and mirrored on contralateral side (B) and do a ratio (A/B)

40. DSC for tumor vs. radiation necrosis
Threshold for radiation necrosis rCBV <1.5, with median <1.0
Threshold for tumor rCBV>1.5 to 2.0 (sens vs. spec)
AJNR February 2009 30: 

41. DSC for tumor vs. radiation necrosis
Threshold for radiation necrosis rCBV <1.5, with median <1.0
J Neurooncol (2010) 99: 81-88

42. DSC, pitfalls Inappropriate bolus: evaluate time-concentration curve
rCBV of up to 2.4 has been reported in radiation necrosis, although rare
DSC MR perfusion does not allow a robust absolute quantification, mainly because of the lack of a direct linear relationship between contrast concentration and signal changes
Susceptibility artifacts affect signal, thus always correlate with SWI/GRE images

43. DSC, pitfalls
In area of susceptibility, you have no T2* signal, thus it will show up as a black hole in DSC

44. DSC, pitfalls Partial volume effect from low spatial resolution
Stenosis will rate-limit contrast and result in underestimation of CBF and overestimation of MTT, must always check graph
Non-uniform T2* relaxivity of tissues

46. Arterial Spin Labeling (ASL)

47. ASL, what is it?
Uses magnetically labeled blood as an endogenous tracer, without need to use Gd
Two main types of ASL technique: continuous ASL and pulsed ASL
Continuous: prolonged radiofrequency pulse that continuously labels arterial blood water below the imaging slab until a steady state of tissue magnetization is reached
Pulsed: short radiofrequency pulse is used to label a thick slab of arterial blood at a single point in time and imaging is performed after a period of time to allow distribution in the tissue of interest

48. In tagged images, magnetization is inverted
ASL, how is it done?

49. ASL for tumor
ASL
DSC

50. ASL for tumor Measures CBF
Ratio of tumor vs. normal contralateral site is performed
For high-grade tumors a 3-4 ratio is used as a cutoff, but there is a wide range, thus it not used clinically

51. ASL, pitfalls
Significantly lower SNR and spatial/temporal resolution compared to DSC
Labeling techniques / Continuous technique: Increases magnetization transfer effect. If the magnetization transfer effects are present only during the labeling scheme, perfusion may be overestimated because the saturation effect of the macromolecular pool will result in reduced signal of the free water pool from the tissue of interest
Complex mathematical kinetic modeling
Arterial transit time varies significantly across BV, ASL uses only a single time point to decrease acquisition time
Partial volume errors similar to DSC

52. Intravoxel Incoherent Motion (IVIM)

53. IVIM, what is it? Diffusion-based perfusion
Fast diffusion in intravascular component (f) and fast decay (D*)*

54. IVIM, how to perform? Acquire multiple b values between b0-b100
Between 3-5 values, typically b0, 10,20,40, 80
No Gd used, use endogenous diffusivity of perfusion

55. IVIM, how to perform?
Derive f, or the perfusion fraction

56. IVIM, how to measure Measure f, or the perfusion fraction
 American Journal of Neuroradiology 35.2 (2014):

57. IVIM, pitfalls Relatively new, not enough data
Since based on diffusion values, results vary by machine and vendor
Vendor platforms are available, but not a lot of clinical experience with them
Complex mathematical modeling, has not yet been determined whether Gaussian, Bayesian, Kurtosis, Skweness is best

59. Not all black or white with tumors
Important to recognize that treated tumors are dynamic and heterogeneous
Both radiation necrosis/treatment change and tumor are often present
Multi-parametric tools that include physiological and molecular markers hold promise as complementary tools to anatomic and perfusion imaging for tumor evaluation

60. In summary
Korean J Radiol 2014;15(5):

61. In summary
Clinical perfusion CT and MR are mainly performed for stroke and tumor evaluation
MR DSC is the workhorse of tumor evaluation, except in the setting of susceptibility artifacts
DCE is a powerful perfusion tool, but technically challenging
ASL and IVIM are alternatives in patients that have contraindications to Gd studies
Further work needs to be done to refine the techniques

62. In summary We have reviewed:
The different types of perfusion techniques
The clinical application of perfusion in stroke and tumor evaluation
The technical limitations of perfusion imaging

63. Korean J Radiol 2014;15(5):

64. References
Barajas, R. F., et al. "Distinguishing recurrent intra-axial metastatic tumor from radiation necrosis following gamma knife radiosurgery using dynamic susceptibility-weighted contrast-enhanced perfusion MR imaging." American Journal of Neuroradiology 30.2 (2009):
Essig, Marco, et al. "Perfusion MRI: the five most frequently asked technical questions." AJR. American journal of roentgenology 200.1 (2013): 24.
Federau, Christian, et al. "Perfusion measurement in brain gliomas with intravoxel incoherent motion MRI." American Journal of Neuroradiology 35.2 (2014):
Jahng, Geon-Ho, et al. "Perfusion magnetic resonance imaging: a comprehensive update on principles and techniques." Korean journal of radiology 15.5 (2014):
Law, Meng, et al. "Gliomas: predicting time to progression or survival with cerebral blood volume measurements at dynamic susceptibility-weighted contrast-enhanced perfusion MR imaging 1." Radiology 247.2 (2008):
Mitsuya, Koichi, et al. "Perfusion weighted magnetic resonance imaging to distinguish the recurrence of metastatic brain tumors from radiation necrosis after stereotactic radiosurgery." Journal of neuro-oncology 99.1 (2010):
Niesten, Joris M., et al. "Optimisation of vascular input and output functions in CT-perfusion imaging using 256 (or more)-slice multidetector CT." European radiology 23.5 (2013):
Stroke CTP Vanderbilt Protocol and Workflow Manual.
Sugahara, Takeshi, et al. "Posttherapeutic intraaxial brain tumor: the value of perfusion-sensitive contrast-enhanced MR imaging for differentiating tumor recurrence from nonneoplastic contrast-enhancing tissue." American Journal of Neuroradiology 21.5 (2000):
Tofts, Paul S., et al. "Estimating kinetic parameters from dynamic contrast-enhanced T 1-weighted MRI of a diffusable tracer: standardized quantities and symbols." Journal of Magnetic Resonance Imaging 10.3 (1999):

65. Thank you for your attention!