To the Editor:—

In their confirmatory study, Kottenberg-Assenmacher et al.  concluded that surface hypothermia to a core temperature of 32°C did not depress median nerve somatosensory-evoked potentials (SSEP) amplitude, but did prolong latency. 1This is an important observation, as patients undergoing certain orthopedic and vascular surgical procedures that may result in spinal cord ischemia have SSEPs monitored. Because these patients are often mildly to moderately hypothermic, it is important to know if the spinal cord temperature is at the same core temperature and, if so, what affect spinal cord hypothermia has on SSEPs.

In their study, the authors induced hypothermia, and although the mean total duration of the surface hypothermia was not stated, the article inferred that it was acute and transitory. To determine whether there were effects of hypothermia on the spinal cord that affected the SSEPs, we recommend the following:

1. Measure the temperature of the spinal cord. This can be approximated clinically by inserting a thermistor-tipped 4-French pulmonary artery catheter via  a needle into the subarachnoid space at the lumbar level. Using this technique, the lowest mean cerebrospinal fluid temperature in a case series of patients undergoing spinal cord cooling for thoracoabdominal aneurysm repair was 26.7 ± 3.2°C. 2This technique will provide an indirect, but clinically relevant, measurement of temperature at the spinal cord level. If this were undertaken, the tibial nerve SSEP would need to be measured to determine the effect of spinal cord hypothermia on the SSEP.

2. Measure SSEP in patients with acquired poikilothermia in whom steady-state hypothermia can be maintained. MacKenzie et al.  published a case series of four patients with acquired poikilothermia who had spontaneous hypothermia maintained at 33.5 ± 0.3°C for a minimum of 4 days. 3The inference is that in poikilothermia, the patient’s central core steady-state temperature is more likely to reflect the local spinal cord temperature. In these patients, hypothermia increased the latency of the central conduction times; however, in this group, only the latencies could be reliably identified.

As interest in moderate hypothermia as a neuroprotective technique increases, it is important to definitively answer the question concerning the effect of spinal cord hypothermia on SSEPs.

Kottenberg-Assenmacher E, Armbruster W, Bornfeld W, Peters J: Hypothermia does not alter somatosensory evoked potential amplitude and global cerebral oxygen extraction during marked sodium nitroprusside–induced arterial hypotension. A nesthesiology 2003; 98: 1112–8
Murray MJ, De Ruyter ML, Torres NE, Lunn JJ, Harrison BA Thoracoabdominal aortic aneurysm repair: Reducing the incidence of paraplegia. Semin Cardiothorac Vasc Anesth 1999; 3: 30–3
MacKenzie MA, Vingerhoets DM, Colon EJ, Pinckers AJLG, Notermans SLH: Effect of steady hypothermia and normothermia on multimodality evoked potentials in human poikilothermia. Arch Neurol 1995; 52: 52–8