USE OF THE ELECTROSPINOGRAM FOR PREDICTING HARMFUL SPINAL-CORD ISCHEMIA DURING REPAIR OF THORACIC OR THORACOABDOMINAL AORTIC-ANEURYSMS

Background. To reduce the incidence of misleading assessments, and to derive criteria for critical spinal cord ischemia during thoracic or t horacoabdominal aortic aneurysm repair, the authors epidurally stimula ted and recorded somatosensory evoked potentials (ESEP) below and abov e, respectively, the spinal segment at risk (electrospinogram). Method s. Epidural somatosensory evoked potentials were analyzed in 100 conse cutive patients undergoing resection of aortic aneurysms using two bip olar catheters (stimulation at the L2 level and recording at the T3 le vel) for the following criteria: 1) the time until ESEP disappeared co mpletely after cross clamping, 2) the duration of complete ESEP loss d uring and after cross clamping, and 3) the time until ESEP recovered a fter declamping. Postoperatively, neurologic deficits were evaluated b y a neurologist who was unaware of the ESEP recordings. Results: Three types of patients could be identified. First, thirty-one patients nei ther showed ESEP loss nor neurologic deficits. Second, ESEP loss occur ring later than 15 min after cross clamping was associated with a neur ologic deficit in 2 of 29 patients (6.9%). And, third, 12 of 40 patien ts (30%) presented a neurologic deficit when ESEP loss occurred within 15 min after cross clamping. Further indicators of an impending risk were a total ESEP loss greater than 40 min (sensitivity 100%, specific ity 68%, positive predictive value [PPV] 35%, and negative predictive value [NPV] 100%), and a recovery of ESEP later than 20 min after decl amping (sensitivity 93%, specificity 86%, PPV 52%, and NPV 99%). Concl usions. Epidural somatosensory evoked potentials appeared to be a reas onable intraoperative predictor of postoperative neurologic outcome, a nd informs surgeons and anesthesiologists about the impending danger a t an early state of the operation.