Effect of hyperglycemia on progressive paraparesis in a rat etastatic spinal tumor model

Hyperglycemia has been shown to potentiate ischemic injury of the spinal cord by quenching vasodilators and potentiating tissue acidosis and free radical production. Steroid-induced hyperglycemia is a common event in the surgical management of metastatic epidural spinal cord compression (MESCC). The goal in this study was to determine whether experimentally induced hyperglycemia accelerates neurological decline in an established animal model of MESCC.

Sixteen Fischer 344 rats underwent a transabdominal approach for implantation of a CRL-1666 breast adenocarcinoma cell line within the vertebral body of L-6. After 72 hours of recovery from tumor implantation, the animals received intraperitoneal injections every 12 hours of either 2 g/kg dextrose in 5 ml 0.09% saline (hyperglycemia, 8 rats) or 5 ml 0.09% saline alone (normoglycemia, 8 rats). Weights were taken daily, and the hindlimb function was tested daily after tumor implantation by using the Basso-Beattie-Bresnahan (BBB) scale (score range 1-21). Animals were killed at time of paralysis (BBB Score < 7), and the volume of epidural tumor growth within the spinal canal was measured. To determine the degree of hyperglycemia induced by this dextrose regimen, a surrogate group of 10 Fischer 344 rats underwent intraperitoneal injections of 2 g/kg dextrose (5 rats) or 0.09% saline (5 rats) every 12 hours, and serum glucose levels were assessed 1, 3, 6, 8, 10, and 12 hours after injections for 24 hours.

Dextrose versus saline injections resulted in elevated mean serum glucose at 3 (259 vs 103 microg/dl), 6 (219 vs 102 microg/dl), 8 (169 vs 102 microg/dl), and 10 hours (118 vs 99 microg/dl) after injection, returning to normal levels by 12 hours (96 vs 103 microg/dl) just prior to subsequent injection. All rats had normal hindlimb function for the first 8 days after tumor implantation. Hyperglycemic versus normoglycemic rats demonstrated a worsened median BBB score by postimplantation Day 9 (Score 20 vs 21, p = 0.023) through Day 16 (Score 8 vs 12, p = 0.047). Epidural tumor volume demonstrated a near-linear growth rate across both groups; however, hyperglycemic rats developed paralysis earlier (median 15.5 vs 17.5 days, p = 0.0035), with significantly less epidural tumor volume (2.75 +/- 0.38 cm(3) vs 4 +/- 0.41 cm(3), p < 0.001) at time of paralysis.

In a rat model of metastatic epidural spinal cord compression, rats maintained in a hyperglycemic state experienced accelerated time to paralysis. Also, less epidural tumor volume was required to cause paralysis in hyperglycemic rats. These results suggest that hyperglycemic states may contribute to decreased spinal cord tolerance to compression resulting from MESCC. Clinical studies evaluating the effect of aggressive glucose control in patients with MESCC may be warranted.