Serial determinations of cerebral water content by magnetic resonance imaging after an infusion of hypertonic saline.

OBJECTIVE To determine regional cerebral water content in vivo by magnetic resonance imaging (MRI) after the administration of 7.5% saline in brain-lesioned rabbits. DESIGN Randomized, controlled, intervention trial. SETTING University animal laboratory. SUBJECTS Eighteen male New Zealand white rabbits, randomly assigned to one of three groups. INTERVENTIONS The animals were anesthetized (1% halothane), intubated, and mechanically ventilated to maintain end-tidal CO2 tension between 30 and 35 mm Hg (4 and 4.7 kPa). Arterial and central venous catheters were inserted and arterial blood samples were serially obtained during the experiment. Serum osmolality was measured. A cryogenic cerebral lesion was produced by pouring liquid nitrogen for 1 min into a funnel placed on the intact skull over the right hemisphere. One group of animals received 20 mL of 7.5% saline intravenously 150 mins after the cerebral lesion was generated (7.5% saline group, n = 7). A second group of animals received the same volume of 0.9% saline intravenously (0.9% saline group, n = 7). In a third group of animals (control group, n = 4) no lesion was created and no fluid administered. MEASUREMENTS AND MAIN RESULTS Five spin-echo T2-weighted MRIs of the brain were acquired at 90 mins (Baseline 1), 120 mins (Baseline 2), 150 mins (Infusion), 180 mins (Infusion + 30 mins), and 210 mins (Infusion + 60 mins) after the generation of the cerebral lesion. In the control group, two scans separated by a time interval of 120 mins were performed. The percent changes in signal intensity between the first and the four following scans of a coronal slice of the central region were determined. Analysis of variance and the Mann-Whitney U test were used for statistical analysis. Data are presented as mean +/- SD; p < .05 was considered significant. Serum osmolality increased significantly from 308 +/- 13 mosm/L to 349 +/- 19 mosm/L after the infusion of 20 mL of 7.5% saline, but did not change after the administration of 0.9% saline. Signal intensity in the area between the caudal edge of the core of the lesion and the basal ganglia was 9 +/- 8% higher on the injured side than in the corresponding area on the contralateral side (p < .05). Compared with Baseline 1, signal intensity at Infusion + 60 mins decreased by 26.3 +/- 13.7% in the 7.5% saline group, whereas it decreased by 10.4 +/- 8.6% in the 0.9% saline group (p < .05 between groups). Signal intensity decreased only slightly and nonsignificantly by 0.6 +/- 4.4% between the two scans in the control group. CONCLUSIONS The administration of a 7.5% saline solution causes a prompt and substantial decrease in cerebral water content as assessed by spin-echo T2-weighted MRI. Magnetic resonance imaging offers the opportunity for repeated, noninvasive in vivo determinations of cerebral water content.