Assessment of cerebrovascular reactivity with functional magnetic resonance imaging: comparison of CO 2 and breath holding Andreas Kastrup*, Gunnar Kru ¨ger, Tobias Neumann-Haefelin, Michael E. Moseley Department of Radiology, Stanford University, Stanford, CA, USA Received 31 December 1999; accepted 15 December 2000 Abstract Cerebral blood flow (CBF) and oxygenation changes following both a simple breath holding test (BHT) and a CO 2 challenge can be detected with functional magnetic resonance imaging techniques. The BHT has the advantage of not requiring a source of CO 2 and acetazolamide and therefore it can easily be performed during a routine MR examination. In this study we compared global hemodynamic changes induced by breath holding and CO 2 inhalation with blood oxygenation level dependent (BOLD) and CBF sensitized fMRI techniques. During each vascular challenge BOLD and CBF signals were determined simultaneously with a combined BOLD and flow-sensitive alternating inversion recovery (FAIR) pulse sequence. There was a good correlation between the global BOLD signal intensity changes during breath holding and CO 2 inhalation supporting the notion that the BHT is equivalent to CO 2 inhalation in evaluating the hemodynamic reserve capacity with BOLD fMRI. In contrast, there was no correlation between relative CBF changes during both vascular challenges, which was probably due to the reduced temporal resolution of the combined BOLD and FAIR pulse sequence. © 2001 Elsevier Science Inc. All rights reserved. Keywords: Functional magnetic resonance imaging; Human brain; Hypercapnia; Blood oxygenation; Cerebral blood flow; Cerebrovascular reactivity 1. Introduction With the advent of functional neuroimaging techniques the field of human brain mapping and neuroscience has experienced tremendous changes in the past decade. Stand- ing out in this context are newly developed functional mag- netic resonance imaging (fMRI) techniques, which uniquely combine the capabilities of high resolution anatomic imag- ing with noninvasive mapping of hemodynamic responses underlying neuronal events. Most fMRI techniques are based on the blood oxygenation level dependent (BOLD) contrast, using paramagnetic deoxyhemoglobin as an en- dogenous contrast agent [1]. A focal neuronal activation causes an increase in cerebral blood flow (CBF) without commensurate increase in oxygen extraction and changes the absolute deoxyhemoglobin concentration, which in turn alters the local magnetic field susceptibility and T2*. Since the BOLD technique does not measure tissue perfusion directly, a number of CBF-based fMRI techniques have been developed, such as the flow-sensitive alternating in- version recovery technique (FAIR) [2,3], which allow to determine relative and absolute CBF changes during brain activation. Besides solely mapping cognitive brain functions an in- creasing number of studies indicate that fMRI techniques could potentially become useful clinically [4]. In the past few years, these techniques have been employed to map BOLD signal intensity and CBF changes during vascular challenges, such as inhalation of CO 2 gas mixtures or in- jection of acetazolamide [5–9]. Although the cerebrovascu- lar reserve capacity can also be measured with positron emission tomography (PET), single photon emission-com- puted tomography or xenon computed tomography, these methods suffer from several disadvantages such as the ne- cessity to use radioisotopes, poor spatial and/or temporal resolution, limited availability and costly examination. In contrast to transcranial Doppler sonography (TCD) fMRI techniques have a high spatial resolution, so that impaired reactivity can be identified in very small brain regions. The measurement of cerebrovascular reserve potentially permits clinicians to identify subgroups of patients with carotid * Corresponding author. Neurologische Universita ¨tsklinik Tu ¨bingen, Hoppe-Seyler-Str. 3, Tu ¨bingen D72076, Germany. Fax: +49-7071- 295260. E-mail address: andreas.kastrup@uni-tuebingen.de (A. Kastrup). Magnetic Resonance Imaging 19 (2001) 13–20 0730-725X/01/$ – see front matter © 2001 Elsevier Science Inc. All rights reserved. PII: S0730-725X(01)00227-2