Spinal cord sodium, potassium, calcium, and water concentration changes in rats after graded contusion injury

Spinal cord Na, K, Ca, and H2O changes were measured 6 h after graded contusion injuries in 40 Sprague-Dawley rats. A 10 g weight was dropped 1.25 cm (n = 6), 2.5 cm (n = 7), 5.0 cm (n = 6), or 7.5 cm (n = 7) onto the thoracic spinal cord of 26 rats. An additional 10 rats served as laminectomy controls and 4 rats were unoperated controls. At 6 h after surgery or injury, the spinal cords were rapidly cut into 4 mm segments, weighed to obtain tissue wet weights (W), dried for 14-16 h at 97 degrees C in a vacuum oven (30 mmHg), and reweighed for tissue dry weights (D). Water concentrations (H2O]d) were estimated from (W-D)/D in units of ml/g D. Ionic concentrations (Na]d, K]d, and Ca]d) of the tissue samples were measured by atomic absorption spectroscopy with units of mumol/g D. Ionic shifts (delta Na]d, delta K]d, delta Ca]d) were calculated by subtracting laminectomy control values from those measured in injured cords. Laminectomy alone significantly increased Na]d and H2O]d compared to unoperated controls. Mean +/- standard deviations of H2O]d, Na]d, K]d, and Ca]d were, respectively, 1.95 +/- 0.07, 182.6 +/- 5.9, 277.2 +/- 11.8, and 12.1 +/- 1.4 in unoperated controls; 2.12 +/- 0.08, 238.6 +/- 9.2, 277.8 +/- 9.2, and 11.7 +/- 1.1 in laminectomy controls. At the impact site, K]d fell by 14-37% and H2O]d rose by 14-24%, Na]d by 13-64%, and Ca]d by 65-137% of laminectomy control values. delta Na]d, delta K]d, and delta Ca]d correlated linearly with impact velocities; Ca]d increased by 1.0% per cm/sec (r = 0.995, p less than 0.005), Na]d increased 0.67% per cm/sec (r = 0.950, p less than 0.01), and K]d decreased 0.34% per cm/sec (r = 0.964, p less than 0.01). Neither delta H2O] nor delta Na]d + delta K]d consistently predicted impact velocity. Na]d + K]d correlated with H2O]d with a slope of 177.4 mumol/ml (r = 0.697, p less than 0.005). Since Na and K constitute greater than 95% of tissue inorganic ions, the slope approximates net ionic shift per ml of water entry or the ionic osmolarity of edema fluid. These results indicate that increasing contusions produce graded ionic shifts and that edema does not predict contusion severity. These data support our hypothesis that net ionic shifts cause edema in injured spinal cords.