HOW RELIABLE IS THE PREDICTIVE VALUE OF SEP (SOMATOSENSORY-EVOKED POTENTIALS) PATTERNS IN SEVERE BRAIN-DAMAGE WITH SPECIAL REGARD TO THE BILATERAL LOSS OF CORTICAL RESPONSES

Objective: In patients sustaining severe brain damage (SBD), predictio n of later outcome is often very difficult, in particular under condit ions of therapeutic management like relaxation and intravenous phenoba rbital. Somatosensory evoked potentials (SEP) correlated best with lat er prognosis and expected neurological deficit. In detail, the primary bilateral loss of cortical responses (BLCR) is regarded to be a relia ble marker for poor prognosis. The goal of the following prospective s tudy was to reassess the prognostic value of early serial median nerve SEP recording in severe brain damage in comparison to other neurophys iologica, clinical, and neuroradiological parameters and, additionally , to test the hypothesis, that the BLCR-pattern is always associated w ith a poor prognosis and is never reversible. Design: Prospective stud y. Setting: Anaesthesiological hospital intensive care unit. Patients and participants: 42 comatose patients with severe brain damage (29 ma les/13 females), mean age 39.6 +/- 19.3 years, mean initial Glasgow Co ma Scare (GCS) 6.6 +/- 3.1, investigated by means of median somatosens ory evoked potentials (SEP) and brainstem auditory evoked potentials ( BAEP) with serial recordings on day 1, 3 to 4, and 8 to 1, and repeate d cranial computerized tomography. Measurements and results. We classi fied the outcome according to the Glasgow Outcome Scale (GOS). Our dat a showed a high correlation of initial SEP scores (Spearman correlatio n coefficient = - 0.70) with outcome for both a favorable and an unfav orable prognosis (p = 0.0001). The reliability increased with serial r ecordings. The SEPs showed superiority to parallel brainstem auditory evoked potentials (Spearman correlation coefficient = - 0.50, p = 0.00 07), GCS, and standardized neuroradiological criteria. Primary (BLCR) occurred in 16 patients (38 %) and implied a fatal prognosis in all ad ult patients (n = 15, specificity = 93.3 %, sensitivity = 59.3 %). In contrast, a young child with predominant brainstem hemorrhagic contusi ons regained consciousness and developed mild to moderate neurological deficit (GOS 3-4) during long-term follow-up of 4 years. This clinica l improvement paralleled bilateral asymmetric recovery of cortical SEP responses. Conclusions: SEP allow an early reliable assessment of bot h poor and good prognosis in SBD, in particular when applied serially. BLCR does not always imply a fatal diagnosis, as a circumscribed cont usional lesion rarely may lead to selective reversible blockage of asc ending somatosensory pathways in the brainstem.