Study design: Controlled correction of scoliosis in a rat model was used to assess the accuracy of intraoperative motor-evoked potential monitoring.
Objectives: The purpose of this study was to develop a model in which motor-evoked potential changes could be compared with neurologic function after surgery, such that a threshold for responding to motor-evoked potential changes may be established.
Summary of background data: Intraoperative motor-evoked potential monitoring has become technically feasible. Clinical application now depends on the development of useful interpretation parameters and correlation with neurologic sequelae.
Methods: Experimental scoliosis was produced in 30 rat pups. After growth, the rats underwent correction of their scoliosis by distraction. Changes in tcMMEP onset latency and amplitude were measured. Distraction was applied either until a 10% delay in tcMMEP onset latency (Group 1), until tcMMEP responses were ablated (Group 2), or for 10 minutes after the loss of transcranial magnetic stimulation response (Group 3).
Results: In Group 1 (n = 10), all animals had tcMMEP with normal onset latency and normal neurologic examinations 24 hours after surgery. In Group 2 (n = 10), tcMMEP were normal in four rats, markedly delayed in three rats, and absent in three rats 24 hours after surgery. neurologic examination was normal in the four rats with normal tcMMEP. Moderate deficit was noted in two of the three rats with prolonged onset latency 24 hours after surgery; the third was intact. Moderate neurologic injury was noted in two of three rats with absent tcMMEP 24 hours after surgery; the third rat was paralyzed. In Group 3 (n = 10), vertebral dislocation was noted on lateral radiographs in eight of 10 animals. Twenty-four hours after surgery, tcMMEP remained absent, and paralysis was noted in the eight rats with dislocation. The two rats without dislocation had delayed tcMMEP but some return of neurologic function.
Conclusions: Comparison of the three groups shows a significant correlation between tcMMEP and endpoint neurologic outcome. None of the rats in Group 1 had a neurologic deficit after surgery as opposed to five of 10 rats in Group 2 and 10 of 10 rats in Group 3 with significant neurologic injury. These findings suggest that a 10% delay in onset latency would be an appropriate threshold for responsing to changes in tcMMEP.