The purpose of this work was to evaluate the clinical utility of two ASL pulse sequences in pediatric patients. The first sequence is a 3D cylindrically-distributed spiral "static" pCASL protocol used to quantify whole-brain cerebral blood flow (CBF) [1]. The second sequence is a 3D time-resolved "dynamic" approach called CINEMA (Contrast inherent INflow Enhanced Multi phase Angio) [2-4], which provides arterial imaging of the neurovasculature without the need for Gadolinium contrast. With increasing concerns over possible Gadolinium deposition in the brain from contrast-enhanced MRI exams [5-8] in recent months, our pediatric hospital has been actively pursuing non-contrast MRI protocols in neuroimaging [9].

All exams were performed on Philips Ingenia 3 Tesla platforms with R5.1.7 software and 13- or 32-channel head coil arrays with local institutional review board approval. While we have successfully performed both 3D spiral pCASL and CINEMA in more than 10 and 30 patients, respectively, the combination of the two techniques have been particularly useful in assessing Moyamoya disease patients, as the information supplements conventional TOF angiography and post-Gadolinium 3D T1-weighted imaging. In this abstract, we specifically describe two cases from our preliminary experience, (Patient 1) an 8y female, and (Patient 2) a 1y male, both with Moyamoya disease. Diamox (acetazolamide, X-GEN Pharmaceuticals, Inc.), a vasodilator, was intravenously administered (15 mg/kg body weight) to assess CBF reserve response. pCASL data were acquired before Diamox administration and repeated 15 minutes after injection. TOF and CINEMA acquisitions were only acquired prior to Diamox injection.

The technical aspects of both 3D spiral pCASL [1] and CINEMA [2-4] have been described in detail in the accompanying references. Briefly, the 3D cylindrically-distributed spiral pCASL technique utilizes a turbo-spin-echo sequence with a paired spiral-in and spiral-out k-space trajectory such that the spin echo aligns with the center of k-space. Typical imaging parameters used were: 200-260 mm FOV, 3 mm in-plane resolution, 30 transverse slices of 4 mm thickness, no parallel imaging acceleration, TR/TE 4500/20 ms, an echo train length of 9, 7.35 spiral interleaves, 20 ms of spiral readout duration, SPIR fat and FOCI background suppression, 90 mm inferior labeling distance at the level of the carotid bifurcation, and 1600 ms for both label duration and post labeling delay. Approximate scan time was 4:30. Separately, proton-density-weighted data of the same imaging volume was acquired without pCASL magnetization preparation to facilitate CBF computation [10]. White and gray matters voxels were extracted using FSL software to estimate CBF distributions.

For CINEMA, the sequence utilizes a 3D GRE sequence with multi-shot EPI readout coupled to an inversion-recovery Look-Locker scheme. The technique uses STAR – Signal Targeting with Alternating RF for ASL tagging and acquires time-resolved data by varying the interval between labeling and imaging blocks. Typical imaging parameters used in this work were: 160-220 mm FOV, 1.1-1.2 mm in-plane resolution, 80 transverse overcontiguous 1.6 mm slices with 0.8 mm effective thickness, two-fold SENSE in the anterior-posterior direction, TR/TE = 12/5.8 ms for the Look-Locker readout scheme, 10 degrees flip angle, 12 dynamic phases, first phase at 200 ms post labeling delay, 120 ms intervals between consecutive phases, 2 seconds of cycle duration between control and label image sets, 300 mm label thickness, 20 mm inferior label gap, and a scan time of 4-6 minutes.

Figures 1 and 2 illustrate pCASL and CINEMA results from the two patients, along with TOF data. In Patient 1, TOF angiography demonstrates impaired flow in the right middle cerebral (MCA) and internal carotid arteries (ICA). CINEMA images show corroborating data of delayed arterial filling on the right side. pCASL maps pre- and post-Diamox demonstrate very little change in CBF (pre- CBF average: 59.5 ml/100g/min, post- CBF average: 61.9 ml/100g/min), suggesting limited CBF reserve. In Patient 2, the response to Diamox was moderate, with mean whole-brain CBF increasing 11% from 31.4 to 34.9 ml/100g/min. A perfusion defect in the right posterior parietal and temporal lobes is evident and perfusion to it does not seem to improve with Diamox. TOF confirms a compromised right MCA and ICA, and again CINEMA clearly illustrates hindered flow.

Both 3D spiral pCASL and CINEMA provide robust approaches to assess the neurovasculature in pediatric patients without the usage of Gadolinium contrast. Quantitative CBF maps and time-resolve data generated by both ASL pulse sequences yield diagnostically useful information that supports clinical findings from conventional TOF angiography. Further evaluation in patients with seizures, strokes, brain tumors, and arterial-venous malformations is warranted to determine if traditional Gadolinium-enhanced angiography can be supplanted.