Age dependency of the regional cerebral blood volume (rCBV) measured with dynamic susceptibility contrast MR imaging (DSC)

Magnetic

ELSEVIER

l

Resonance Imaging, Vol. 14, No. 2, pp. 157-162, 1996 Copyright 0 1996 Elsevier Science Inc. Printed in the USA. All rights reserved 0730-725X/96 $15.00 + .OO

0730-725X( 95)02058-6

Original Contribution AGE DEPENDENCY OF THE REGIONAL CEREBRAL BLOOD VOLUME (rCBV) MEASURED WITH DYNAMIC SUSCEPTIBILITY CONTRAST MR IMAGING (DSC) FREDERIK WENZ,~- KATR~V FRIEDEMANN G~~cKEL,$

~%EMPP, * THOMAS

GUNNAR BRIX,

* MICHAEL V. KNOPP,* H~l3, * AND GERHARD VAN KAICK *

*Department of Radiological Diagnostics and Therapy, German Cancer Research Center, TDepartment of Clinical Radiology, University of Heidelberg, and *Institute for Clinical Radiology, Mannheim, Germany The changes of the regional cerebral blood volume (rCBV) with age were studied using dynamic susceptibility contrast MRI (DSC). We examined an unselected, random sample of 71 consecutive patients referred for work-up of suspected intracranial tumors (35 normal examinations, 36 tumors) with a standard 1.5 T clinical MR system. Determination of the rCBV was performed with a Z’**-weighted simultaneous dual (SD) FLASH sequence (TR/TEl/TE2/a = 32/25/16/10”, 55 images) after bolus injection of Gd-DTPA. Absolute quantification of the rCBV was achieved by normalizing the measured tissue concentration-time curves with the integrated arterial input function (AIF), which was simultaneously measured in the brain feeding arteries. The rCBV (mean ? SD) was 8.4 t 2.9 ml/100 g and 4.2 + 1.7 ml/100 g in gray and white matter, respectively, with a decline of about 3% and 6% per decade for white and gray matter, respectively. We conclude that DSC using a SD FLASH sequence allows the simultaneous measurement of the AIF and the tissue concentration-time curve and thus an absolute quantification of the rCBV, which is the basis for interperson comparisons and follow-up studies. Keyword: Age.

MRI; Dynamic susceptibility

contrast MRI;

INTRODUCTION

316195;

ACCEPTED

agent; Regional cerebral blood volume;

time curves, from which one may estimate the rCBV. However, the standard DSC technique yields relative rather than absolute rCBV values. We have established a DSC technique’ based on a simultaneous dual (SD) FLASH sequence6 on a standard clinical 1.5 T MR system, which allows the determination of the arterial input function (AIF) and thus enables the absolute quantification of the rCBV without relying on relative values by referencing to an internal standard. This article presents the age dependency of the rCBV in a random sample of 7 1 consecutive patients.

Until recently, the assessment of the cerebral perfusion has been mainly restricted to nuclear medicine tracer techniques. These methods, like single photon emission computed tomography (SPECT) and positron emission tomography (PET), require the application of radioactive tracers. The determination of arterial tracer levels, which is needed for the absolute quantification of the regional cerebral blood volume (rCBV) , is rather invasive, requiring catheterisation of brain supplying arteries. Based on the indicator dilution theory, ’ dynamic susceptibility contrast MR imaging (DSC) allows the assessment of the cerebral hemodynamics.2-5 The passage of a contrast agent bolus through the cerebral capillary network is monitored by a series of T,*-weighted images. The measured signal-time curves are converted into concentrationRECEIVED

Contrast

MATERIALS

AND METHODS

We examined 71 consecutive patients (age 15 -72 yr, mean 43.2 yr) referred for initial work-up of various KIinische Radiologie, Radiologische Universittitsklinik, Neuenheimer Feld 400, 69120 Heidelberg, Germany.

915195.

Address correspondence to Dr. Frederik Wenz, Abteilung 157

Im

Magnetic ResonanceImaging 0 Volume

158

Table 1. Summary of 71 patients Type of tumor

n

Mean age (yr)

Gliomas Meningiomas Hypophyseal adenoma Metastases Craniopharyngioma Chondrosarcoma Colloid cyst Normal MRI Total

1.5 8 5 4 2 1 1 35 71

40.2 54.3 49.6 64.5 20.0 24.0 40 40.6 43.2

intracranial tumors (see Table 1) . The groups of patients with and without tumors had approximately the same mean age. Only patients with arterio-venous malformations (AVM) were excluded from the study. All of the patients were in stable condition, and none of the patients had a history of clinically manifest cerebrovascular disorders or previous cranial irradiation. Informed consent was obtained from all subjects after the nature of the study was explained to them. The methodological details are published elsewhere.5 In short, we used a standard 1.5 T MR system (MAGNETOM SP 4000, Siemens AG, Erlangen, Germany) and a commercially available circular polarized head-coil. The imaging protocol consisted of a sagittal scout image and axial precontrast T,- and T,-weighted spin-echo sequences. A flow-sensitive axial localizer sequence was used to select the appropriate slice position for measuring the AIF. Fifty-five T2*-weighted SD FLASH images (TR/TEl/TE2/a = 31/16/25/ lo”, TH = 5 mm, matrix size 64 X 128, FOV 200 mm) were acquired to monitor the passage of the contrast agent bolus (4-5 s, MAGNEVIST Gd-DTPA 0.1 mmol/kg body weight), which was injected after acquisition of the fourth image into an anrecubital vein. The SD FLASH sequence permits the simultaneous imaging of two independent slices with echo times of 16 and 25 ms during one repetition time of 32 ms. One slice (short TE) was selected to cut the brain feeding arteries perpendicularly; the second slice (long TE) was placed over normal brain tissue. The spatial and temporal resolution are 1.65 X 3.15 X 5 mm3 and 1.9 s, respectively. Postcontrast T,-weighted spin-echo and a T1-weighted 3D MPRAGE concluded the imaging session. The patients were instructed to avoid any head motion during the scanning intervals. Measured signal-time curves were converted into concentration-time curves pixel-by-pixel based on the following logarithmic equation: C(t) = 2.1,

-S(t) ( so )

(1)

14,

Number 2, 1996

where k represents the proportionality factor, and S(t) and So are the signal intensities at time t and t = 0. The AIF was determined using specially developed computer software from the short TE slice as the mean concentration-time curve from pixels in the brain feeding arteries. These pixels were identified based on specific attributes like high and early maximum and short mean transit time. At least six regions of interest (ROI, at least 20 pixels each) were placed at least 2 cm away from detectable lesions in normal appearing gray and white matter for the determination of the rCBV. Correlation with the anatomical images was used to avoid ROI placement over major vessels, areas of previous hemorrhage, or marker clips from stereotactic biopsies, because the underlying theory would not be valid in these situations. The rCBV was then calculated in absolute values (ml/ 100 g) as the area under the measured concentration-time curve normalized to the integrated AIF according to eq. (2) : m k,, rCBV = P

s ,”

s0

Cm(t)dt (2)

AIF( t)dt

k,, corrects for differences in hematocrit

in large and small vessels and p equals the density of brain tissue (1.04 g/ml). RESULTS

All patients tolerated the procedure well; no acute side effects were observed. Figure 1 shows a calculated rCBV map in a patient with a hypophyseal microadenoma. The tumor itself is not included in the imaging section. The higher rCBV in gray matter as compared to white matter results in the brighter shadings. The quantitative evaluation revealed a rCBV (mean ? SD) of 8.4 f 2.9 ml/100 g in gray matter and 4.2 + 1.7 ml/100 g in white matter, with a ratio of 2.1 + 0.6. Assuming a gray-to-white matter ratio of 60% to 40%,7 there is a mean overall rCBV of 6.7 ml/100 g. The age dependency of the rCBV measured in gray and white matter and the gray/white matter ratio can be seen in Figs. 2-4. There is a negative correlation of gray and white matter rCBV with age, which means that there is an age-dependent decrease in rCBV (see Table 2). The age-dependent decline in white matter rCBV of about 3% per decade is less pronounced than in gray matter (about 6% per decade). This is furthermore stressed by the decrease of the gray/white matter ratio with increasing age (Fig. 4). The patients were grouped according to their age,

rCBV and age

l

159

ET AL.

F. WENZ

was statistically significant (Student’s t-test, p = .02 for both variables), whereas the difference in white matter rCBV was not significant (p = .08). The rCBV values in patients with brain tumors away from their lesion showed no significant difference compared to the normal subjects (Table 3). There seems to be no obvious systemic effect of these relatively small tumors on remote regions of the cerebral perfusion. DISCUSSION In recent years, MRI has progressed from providing only precise morphological information to the assessment of functional aspects, especially of the brain. Today, functional and dynamic MRI can answer questions that were previously only approachable with invasive or nuclear medicine tracer techniques. Based on an established physiological principle, namely the indicator dilution theory, dynamic susceptibility contrast MRI allows the assessment of the cerebral hemodynamics. Using paramagnetic contrast agents as indicators and the noninvasive measurement of the AIF with the SD FLASH sequence, one can calculate absolute values for the regional cerebral blood volume. The method is fast (about 2-3 min extra time), free of ionizing radiation from either injected tracers or from percutaneous X-rays, and has a low potential for side effects.* The spatial resolution is comparable to that of emission tomographic methods. The temporal resolution achievable with a standard 1.5 T MR system is sufficient to follow a bolus of 5 s duration. A reduction of the scan time below 1 s will lead to significant

Fig. 1. rCBV image (long TE slice) of a patient with a hypophysealmicroadenoma.IncreasingrCBV is coded in higher signalintensity.

with 42 subjects younger than 50 yr and 29 subjects aged 50 yr and older. The detailed results are shown in Table 3. The difference in gray matter rCBV and in gray/white matter ratio between the two age groups

;

16

0

‘4 10

20

30

40

50

60

70

80

Regression -. 95% confid.

Age [years]

Fig. 2. Age dependencyof the rCBV in gray matter. The leastsquarefit linear regressionline and the 95% confidenceinterval are shown.

Magnetic Resonance Imaging 0 Volume 14. Number 2, 1996

160

............ .......

...................... 0

0

IO

20

30

i

40

:

‘s 50

60

70

80

Regression 95% confid.

Age [Yea4

Fig. 3. Age dependency

of the rCBV

influences of the pulse-wave phase on the signal-time curve, which will make further corrections necessary. Yet, echo-planar techniques will allow a higher spatial resolution and the simultaneous measurement of multiple slices. Particularly the possibility of measuring the AIF from several slices may increase the accuracy of the rCBV determination, although the biologic variability seems to cause most of the variation in our study, as can be seen from the standard deviation of the gray/white matter ratio, which is independent from the AIF determination.

in white matter.

The determination of the rCBV is based on the indicator dilution theory for nondiffusible tracers. However, breakdown of the blood-brain barrier allows the exchange of Gd-DTPA from the intravascular compartment to the extracellular space. A signal increase due to T, effects may interfere with the T2*-induced signal loss, which will lead to inexact calculations.’ Therefore, we performed no measurements within the tumors. However, Gd-DTPA is restricted to the intravascular compartment in regions with an intact bloodbrain barrier; thus, the use of the theory is allowed

0.5 10

20

30

40

50

60

70

1 80

Age [years]

Fig. 4. Age dependency

of the gray/white

matter ratio.

‘s

Regression 95% confid

rCBV and age l F. WENZ

Table 2. Age dependencyof rCBV in gray (GM) and white matter (WM) and the gray/white matter ratio Linear regression GM rCBV (ml/100 g)

WM ICB (ml/100 g) GM rCBV/WM rCBV

Intercept

Slope

11.056 -0.062 4.764 2.475

-0.013 -0.009

r,

perosmotic contrast agents as used for dynamic ultrafast CT may induce hemodynamic alterations.‘* Different groups also reported an age-dependent decline of the I-CBV~~*~*‘~~*~and rCBF.5,7,‘4,20 Marchal et a1.14reported a decline in rCBV of about 6% per decade, which is in very good agreement with our value for gray matter (see Table 2). In our study, there seems to be a subset of patients over 50 yr of age, who have relatively high rCBV values. On first sight, one may speculate that the patients with reduced rCBV have a higher risk for ischemic infarcts, but a decreasedcerebrovascularreservecapacity hasbeen shown I6 to be accompanied by an increased rCBV. The patients in OUTstudy represent an unselected, random sample and therefore have an average risk for cerebrovascular

p

-0.327 -0.112 -0.218

.005 .352 .07

Intercept and slope characterize the least square fit linear regression, r, is the Spearman correlation coefficient, p indicates the significance of the correlation.

and rCBV calculations of normal brain tissue can be performed. For the most important clinical application of the DSC technique today, the early diagnosis of ischemic

diseasecomparableto the whole population. In the group

events, a relative quantification of the rCBV is sufficient.4’10’11 However, referencing to an internal standard

of healthy subjects studied previously,5 there are no old patients with a high rCBV; therefore, the slopes of the least square fit regression lines of gray and white matter and age are somewhat steeper than in this study group. Nevertheless, this issue remains to be clarified in future prospective studies including the assessment of more f&ctional parameters like rCBF and the response to acetazolamide injection. As previously shown,5 this method also offers the possibility to calculate regional cerebral blood flow (KBF) . However, further software development is still necessary to reduce time and computational efforts to enable routine determination of the ICBF. We conclude that DSC MRI using a SD FLASH sequence allows the simultaneous measurement of the AIF and the tissueconcentration-time curve and thus an abso-

(e.g., normal appearing white matter as usually practiced with single slice DSC) may be difficult because age’s7 and radiotherapy may influence the cerebral hemodynamics. The noninvasive determination of the AIF allows the absolute quantification of the rCBV, which is the basis for interindividual comparisons and intraindividual follow-up studies. ‘* The measured rCBV data are in agreement with published DSC data from healthy subjects without intracranial pathology.5 Furthermore, there is a reasonably good similarity, within the range of errors, with rCBV values from healthy volunteers 13-16and hospitalized patients without brain diseaseI measured with

nuclear medicine tracer techniques or ultrafast dynamic CT. However, the rCBV values achieved with DSC MRI are slightly higher than those measured with these techniques. This may reflect different physical and physiological characteristics of the different methods, especially possible differences of the volume of

lute quantification of the rCBV, which is the basis for interpersoncomparisonsand follow-up studies. Acknowledgments-This project is supported by grant #138/94 from the Research Commission, Medical Faculty, University of Heidelberg. The authors gratefully acknowledge the active support of Gerald Weisser, MD, Marco Essig, MD, Frank Bohr, MD, Wolfgang Schreiber, PhD, Isolde Speicher, and Barbara Dillenberger.

distribution within the cerebral circulation. For example, the microsphere or labeled red blood cell tech-

niques assess only DSC MRI follows the whole vascular rently not involved

161

ETAL.

REFERENCES

actually perfused vessels, whereas the passage of Gd-DTPA through system, including those vessels curin gas exchange. Furthermore, hy-

1. Zierler, K.L. Theoretical basis of indicator-dilution methods for measuring flow and volume. Circ. Res. 10:393-407; 1963.

Table 3. Summaryof rCBV in gray (GM) and white matter (WM) and the gray/white

n All subjects <50 yr >=50 yr Without tumor

With tumor

71 42 29 35 36

matter ratio

(yr)

GM rCBV (ml/100 g)

43.2 32.7 58.3 40.2 45.7

8.4 9.1 7.4 7.8 9.0

Age

k ? + + k

2.9 2.7 2.8 2.4 3.2

WM raw (ml/100 g) 4.2 4.3 4.2 3.8 4.6

+ + ? -e +

1.7 1.6 1.8 1.1 2.1

GM rCBV/ WM rCBV 2.1 2.2 1.9 2.1 2.1

? + + i ?

0.6 0.6 0.6 0.5 0.7

162

Magnetic ResonanceImaging 0 Volume 14, Number 2, 1996

2. Rosen, B.R.; Belliveau, J.W.; Chien, D. Perfusion imaging by nuclear magnetic resonance. Magn. Reson. Q. 5:263-281; 1989. 3. Aronen, H.J.; Gazit, I.E.; Louis, D.N.; Buchbinder, B.R.; Pardo, F.S.; Weiskoff, R.M.; Harsh, G.R.; Cosgrove, G.R.; Halpem, E.F.; Hochberg, F.H.; Rosen, B.R. Cerebral blood volume maps of gliomas: Comparison with tumor grade and histologic findings. Radiology 191:4151; 1994. 4. Edelman, R.R.; Mattle, H.P.; Atkinson, D.J.; Hill, T.; Finn, J.P.; Mayman, C.; Rontbai, M.; Hoogewoud, H.M.; K&field, J. Cerebral blood flow: Assessment with dynamic contrast-enhanced T2*-weighted MR imaging at 1.5 T. Radiology 176:211-220; 1990. 5. Rempp, K.; Brix, G.; Wenz, F.; Becker, C.; Gtickel, F.; Lorenz, W.J. Quantification of regional cerebral blood flow and volume by dynamic susceptibility contrast enhanced MR imaging. Radiology 193:637-641; 1994. 6. Perman, W.H.; Gado, M.H.; Larson, K.B.; Perlmutter, J.S. Simultaneous MR acquisition of arterial and brain signal-time curves. Magn. Reson. Med. 28:74-83; 1992. 7. Frackowiak, R.S.J.; Lenzi, G.L.; Jones, T.; Heather, J.D. Quantitative assessment of regional cerebral blood flow and oxygen metabolism in man using I50 and positron emission tomography: Theory, procedure and normal values. J. Comput. Assist. Tomogr. 4:727-736; 1980. 8. Kashanian, F.K.; Goldstein, H.A.; Blumetti, R.F.; Holyoak, W.L.; Hugo, F.P.; Dolker, M. Rapid bolus injection of gadopentetate diglumine: Absence of side effects in normal volunteers. Am. J. Neuroradiol. 11:853-856; 1990. 9. Weiskoff, R.M.; Boxerman, J.L.; Sorensen, A.G.; Kulke, S.M.; Campbell, T.A.; Rosen, B.R. Simultaneous blood volume and permeability mapping using a single Gd-based contrast injection. In: Book of abstracts: Thirteenth Annual Meeting of the Society of Magnetic Resonance in Medicine. Berkeley, CA: SMRM; 1994:279. 10. Maeda, M.; Itoh, S.; Kimura, H.; Iwasaki, T.; Hayashi, N.; Yamamoto, K.; Ishii, Y.; Kubota, T. Tumor vascularity in the brain: Evaluation with dynamic susceptibility-contrast MR imaging. Radiology 189:233-238; 1993.

11. Gtickel, F.; Brix, G.; Rempp, K.; Deimling, M.; Rother, J.; Georgi, M. Assessment of cerebral blood volume with dynamic susceptibility contrast-enhanced gradient echo imaging. J. Comput. Assist. Tomogr. 18:344-351; 1994. 12. Wenz, F.; Rempp, K.; Brix, G.; Hep, T.; Weisser, G.; Debus, J.; Knopp, M.V.; Engenhart, R.; van Kaick, G. Radiation induced rCBV changes of low grade astrocytomas and normal brain tissue. In: Book of abstracts: Thirteenth Annual Meeting of the Society of Magnetic Resonance in Medicine. Berkeley, CA: SMRM; . 1994:523. 13. Grubb, R.L.; Raichle, M.E.; Higgins, C.S.; Eichling, J.O. Measurement of regional cerebral blood volume by emission tomography. Ann. Neurol. 4:322-328; 1978. 14. Marchal, G.; Rioux, P.; Petit-TabouC, M.C.; Sette, G.; Travere, J.M.; LePoec, C.; Courtheoux, P.; Derlon, J.M.; Baron, J.C. Regional cerebral oxygen consumption, blood flow, and blood volume in healthy human aging. Arch. Neurol. 49:1013-1020; 1992. 15. Sabatini, U.; Celsis, P.; Viallard, G.; Rascal, A.; MarcVergnes, J.P. Quantitative assessment of cerebral blood volume by single-photon emission computed tomography. Stroke 221324-330; 1991. 16. Steiger, H.J.; Aaslid, R.; Stooss, R. Dynamic computed tomographic imaging of regional cerebral blood flow and blood volume. A clinical pilot study. Stroke 24:591597; 1993. 17. Greenberg, J.H.; Alavi, A.; Reivich, M.; Kuhl, D.; Uzzel, B. Local cerebral blood volume response to carbon dioxide in man. Circ. Res. 43:324-331; 1978. 18. Hilal, S.K. Hemodynamic changes associated with the intraarterial injection of contrast media. Radiology 86:615-633; 1966. 19. Leenders, K.L.; Perani, D.; Lammertsma, A.A.; Heather, J.D.; Buckingham, P.; Healy, M.J.R.; Gibbs, J.M.; Wise, R.S.J.; Hatazawa, J.; Herold, S.; Beaney, R.P.; Brooks, D.J.; Spit&s, T.; Rhodes, C.; Frackowiak, R.S.J.; Jones, T. Cerebral blood flow, blood volume and oxygen utilization normal values and effect of age. Brain 133:2747; 1990.