Identification of Brain Tissue at Risk for Delayed Cerebral Ischemia After Aneurysmal Subarachnoid Hemorrhage

Inclusion criteria

Exclusion criteria

Age ≥ 18 years old

Nonaneurysmal hemorrhage

Confirmed aSAH

Unstable clinical condition

Aneurysm secured (clipping or coiling)

Aneurysm not secured (clipping or coiling)

WFNS grade 1 or 2

Poor clinical grade (WFNS grade ≥ 3)

Acute or chronic respiratory condition

Untreated hydrocephalus

Confusion, agitation

Vasospasm on recent imaging

Table 18.2

Demographics and clinical characteristics at presentation and type of treatment

CASE #	Age	Sex	WFNS grade	Fisher grade	Aneurysm treatment
1	42	Male	I	II	Coil
2	57	Male	II	III	Coil
3	61	Male	II	III	Clipping
4	43	Female	II	III	Coil
5	42	Female	II	IV	Coil

Imaging Protocol and Data Analysis

MR imaging was performed on a 3.0-T scanner (Signa; GE Healthcare, Milwaukee, WI) using an eight-channel phased-array head coil. Functional MRI-based blood oxygen level-dependent (BOLD) volumes were acquired during CO₂ challenge using a model-based prospective end-tidal targeting (MPET) algorithm [3] with the RespirAct^TM (Thornhill Research Inc, Toronto, Canada). The technique has been described in greater detail elsewhere [4]. During the acquisition of the BOLD sequences, subjects’ P_ETCO₂ levels were set to 40 mmHg for 2 min (step 1), 45 mmHg for 2 min (step 2), and then back to 40 mmHg (step 3). The fraction of inspired oxygen (FiO₂) was kept at 100 %.

To obtain the cerebrovascular reactivity (CVR) maps, from the tidal pCO₂ waveforms generated by the RespirAct™, the end-tidal points were manually selected, generating end-tidal CO₂ waveforms. MR and P_ETCO₂ data was imported into the software AFNI (Analysis of Functional Neuroimaging). The first raw images of each BOLD-MRI acquisition were reviewed and the first three volumes discarded to allow for magnetization equilibration. To correct for motion, up to 9 (of 72) volumes in which there was appreciable change in head position between the anatomical acquisition and the BOLD-MRI acquisition were excluded before generating maps of CVR. To account for the artifact generated by coils and clips, the source images were evaluated for signal loss when examining the CVR maps to identify regions where the CVR maps were valid and where they were unreliable. A linear slope of best fit approximated the percentage of BOLD signal change per mmHg change in end-tidal CO₂. Confidence of this fit was assessed with an r-value (Pearson product–moment correlation coefficient). CVR maps were generated by least-squares fitting of the BOLD-MRI signal waveform to the P_ETCO₂ waveform on a voxel-by-voxel basis. From the fitted data, percentage MRI signal change per mmHg P_ETCO₂ change on a voxel-by-voxel basis was calculated (= CVR index).

Results

We recruited 3 men and 2 women, mean age 49 (range 42–61) years. World Federation of Neurological Surgeons (WFNS) and Fisher grades are shown in Table 18.1. Despite the severity of their illness, good quality anatomical and functional images were obtained in all patients. No harmful events related to the MRI examination or CO₂ challenges were observed. CVR studies were obtained between 1 and 8 days after SAH (average 3.6 days). Three patients were scanned within 72 h after onset of the SAH, 1 patient at day 4, and 1 patient at day 8. All patients were in stable neurological condition, without clinical evidence of DCI before and during the MRI exam. Initial imaging before or during aneurysm treatment and repeated cerebrovascular imaging at the time of the CVR test showed no evidence of vasospasm.

Imaging results and clinical findings are shown in Table 18.2. As mentioned, the initial imaging showed no vasospasm in any patient. Delayed imaging showed vasospasm in four patients (80 %), three (60 %) of whom had DCI of varying severity. CVR–MRI test results were considered abnormal in two patients, and both developed DCI. One of these patients had a minor stroke documented in a follow-up MRI, 9 days after hemorrhage in the same territory (severely abnormal CVR suggestive of paradoxical flow with hypercapnia, the “steal phenomenon”). A third patient with a normal CVR-MRI result had DCI. This patient developed progressive confusion and decreased level of consciousness and was taken to the interventional suite for balloon angioplasty. Interestingly, despite the patient’s neurological deterioration, the angiogram demonstrated mild-to-moderate vasospasm in the anterior circulation but no significant spasm in the posterior fossa, site of the aneurysm. Another patient with a normal CVR result had radiological evidence of vasospasm in routine follow-up imaging but no DCI. No mortality occurred in this series.

Only gold members can continue reading. Log In or Register to continue