Reports on the feasibility of intraoperative recording of visual evoked potentials (VEP) to monitor visual pathways during surgery revealed conflicting results. Several authors reported favourable findings using VEP monitoring in different surgical procedures [1-5] whereas others observed several difficulties in obtaining consistent results [6-8]. In a recent study  using total intravenous anaesthesia (TIVA) and transient VEP, we observed a considerable intra- and interindividual variability of the recordings in the surgically anaesthetized patient which questions the practical value of VEP monitoring. However, there is an ongoing profound interest in a reliable technique for intraoperative monitoring of visual pathways, because several interventions in neurosurgery as well as in other surgical faculties are inevitably associated with a risk of injury to these structures such as surgery of the anterior skull base, surgery of pituitary tumours and other neoplasms around the sella as well as intra-orbital surgery. This study was designed to evaluate the characteristics of steady-state visual evoked potential (SSVEP) recording in the anaesthetized patient during surgery. The recording technique of SSVEP is basically similar to that of conventional transient VEP, but differs in that a higher stimulation frequency is used which elicits a characteristically sinusoidal response pattern. Thus, SSVEP are potentially advantageous over transient VEP with regard to intraoperative recording because of their shorter acquisition time and their prominent sinusoidal waveform, which may facilitate identification of the potentials. To the best of our knowledge intraoperative recordings of SSVEP with TIVA have not yet been studied.
All patients gave informed consent to participate in this study. The study group consisted of 30 patients. Only patients with non-cranial surgery were included to avoid potentially confounding factors that may affect SSVEP recording. Twenty-four patients underwent a microsurgical procedure for lumbar disc herniation and six patients for cervical disc herniation with ventral interbody fusion. Time of surgery ranged from 52 to 142 min (mean 83 min). No patient had a history of visual problems or clinical signs or symptoms of a lesion around the visual pathways. There were 20 male and 10 female patients with a mean age of 47.9 yr (range 23.3-69.7 yr).
In all patients SSVEP were recorded one or more days before surgery in the neurophysiological laboratory. Recordings in the anaesthetized patients were performed in the operating room after induction of anaesthesia and positioning of the patient. Anaesthesia was induced with propofol 1.5 mg kg−1, fentanyl 5 μg kg−1 and rocuronium 0.4 mg kg−1. Maintenance of anaesthesia was achieved by infusion of fentanyl 4.8 μg kg−1 h−1 and propofol 6 mg kg−1 h−1. This TIVA regimen was applied exclusively in all cases. Routine anaesthesiological monitoring was performed and body core temperature, mean arterial pressure and PaCO2 were documented together with each SSVEP trace. Recordings were performed during the surgical procedure with the patient in deep surgical anaesthesia.
Parameters for stimulation, recording and data processing were basically the same in the awake and anaesthetized patients. Stimulation, recording and data processing were performed with a Nicolet Viking IV System® (Nicolet Biomedical, Madison, Wisconsin, USA). Both eyes were stimulated separately through closed eyelids using goggles with red light-emitting diodes (Nicolet Biomedical) at a stimulus frequency of 8.5 Hz. Silver cup electrodes with adhesive electrolyte paste were placed at Oz referenced to Fz and the ipsilateral earlobe respectively for a two-channel recording with a ground electrode placed at the forehead. Electrode impedance was kept below 5 kΩ. Filters were set at 2-30 Hz. A timebase of 500 ms was used to allow an identification of the typical sinusoidal waveform and 200 trials were averaged for each trace.
The characteristics of the SSVEP traces were analysed visually. The most prominent first main positive peak (termed P1) was identified as well as the following first main negative peak (termed N1). After manual placement of the markers the peak latency of P1 and the peak-to-peak amplitude of P1-N1 were measured electronically. All recordings of the anaesthetized patients were independently analysed by two reviewers (Barbara Fauser and Helmut Wiedemayer) some time after the recording using identical prints of all recorded traces. Each trace was checked for the presence of the typical sinusoidal SSVEP waveform. Both traces of the two-channel recording and consecutive traces were compared to confirm reproducibility. One trace with a clearly identifiable waveform was selected to utilize its measurements for further analysis. For clearness the data obtained after stimulation of the right eye were used in this paper for presentation of latencies and amplitudes values. Analysis of the values obtained after stimulation of the left eye revealed essentially the same results.
Statistical analyses were performed with SPSS® 11.0 (SPSS Inc., Chicago, IL, USA). Normal distribution of the data was assessed using the one-sample Kolmogorov-Smirnov test. Amplitudes and latencies of SSVEP in the awake and anaesthetized patients were assessed by means of a two-tailed, paired t-test. A correlation analysis was used to evaluate the relationship between amplitudes and latencies in both, the awake and anaesthetized patients, and between the P1-N1 amplitude and the rate of identifiable intraoperative SSVEP traces. Data were expressed as mean ± standard deviation if not indicated otherwise.
Data of latencies and amplitudes obtained in the awake and anaesthetized patients are summarized in Table 1. Body temperature (36.1 ± 0.3°C), mean arterial pressure (82.4 ± 7.7 mmHg) and PaCO2 (4.2 ± 0.24 kPa) remained stable during anaesthesia in all patients. Kolmogorov-Smirnov test confirmed the normal distribution of latency and amplitude values. Latencies and amplitudes in both awake and anaesthetized patients, showed a relative large interindividual variability - a fact that is well known from former studies [10,11]. In the anaesthetized patients a prolongation of P1 latency of 16% and attenuation of the P1-N1 amplitude of 67% was observed compared to the awake patients (Table 1). Changes for both, latencies and amplitudes, were statistically significant.
A stable recording of SSVEP with a clearly identifiable sinusoidal waveform was obtained in all 30 awake patients consecutively assessed in the neurophysiological laboratory. In contrast, the recording of SSVEP in the operating room during surgery with the patient anaesthetized revealed inconsistent results. A total of 1360 SSVEP traces recorded intraoperatively (median 48 traces per patient) were analysed by the reviewers. The typical SSVEP pattern was identifiable in only 56% of the traces. In the individual patient the percentage of traces with a recognizable SSVEP waveform recorded intraoperatively ranged from 16.7% to 100%. In only six of the 30 patients (20%) we were able to identify the SSVEP in more than 90% of the intraoperative traces. Recognizable SSVEP was present in less than 75% of the traces in 23 patients (76.7%) and in less than 50% of the traces in 14 patients (46.7%). Examples of different types of findings are given in Figures 1 and 2.
A comparison of awake and anaesthetized patients revealed a significant positive correlation of P1 latencies (r = 0.37, P = 0.045) and P1-N1 amplitudes (r = 0.40, P = 0.027). No significant correlation was found between the P1-N1 amplitudes and the percentage of identifiable intraoperative SSVEP traces. This was true for amplitudes recorded in the awake patients (r = −0.16, P = 0.934) as well as for the anaesthetized patients (r = 0.109, P = 0.567).
The technique of SSVEP recording used in this study was deliberately kept simple in order to cause minimal interference with the anaesthesiological and surgical procedure and to work in clinical routine. TIVA is generally used in all our regular monitoring cases and was also applied in the study patients because several reports suggest, that this anaesthesiological regimen has less adverse effects on the recording of somatosensory evoked potentials (SEP) compared to volatile anaesthetics [12-16]. To avoid any confusion with other factors potentially effecting SSVEP - such as surgery around the visual pathways, cooling of the brain or fluctuation of physiological variables - only patients with non-cranial surgery were selected for this study and special attention was paid to keep the physiological parameters stable during anaesthesia.
A recently performed study at this institution  yielded that using TIVA intraoperative recording of transient VEP is feasible. However, the results were overall disappointing because the recordings were fairly unstable in the surgically anaesthetized patient. As a consequence we proceeded to evaluate the characteristics of intraoperative VEP recording using SSVEP. A feature of SSVEP, which may be favourable in intraoperative monitoring, is the very prominent sinusoidal character of its waveform, which potentially facilitates identification of the peaks in the comparatively electrical noisy environment of the operating room. Furthermore, the relatively high stimulus frequency compared to transient VEP results in a short acquisition time for each trace, a feature that is generally favourable in all intraoperative neurophysiological monitoring procedures.
Analysis of SSVEP is mostly performed by means of a Fourier analysis of the acquired recordings [17-20]. In order to provide a setting similar to intraoperative monitoring techniques used in everyday clinical practice we applied measurements of latencies and amplitudes analogous to those for transient VEP and other evoked potential modalities. Considering these measurements we observed a minor prolongation of the P1 peak latency of 16% and a more pronounced attenuation of the P1-N1 peak-to-peak amplitude of 67% in the surgically anaesthetized patients. Both findings are likely to be affected by the anaesthetic agents as similar changes are well known for SEP recordings and are comparable to our findings in transient VEP. A prominent finding of this study is the high inter- and intraindividual variability in obtaining reproducible SSVEP in the surgically anaesthetized patient. In the awake patient recording of SSVEP caused no problem and except for the variability of latencies and amplitudes all 30 consecutive patients of this study showed well defined typical waveforms. However, using the same equipment and recording parameters, the findings in the anaesthetized patients were different. We observed only three patients where SSVEP were clearly identifiable in all traces during intraoperative recording apart from a minor degradation of the recordings caused by the electrical noisy environment of the operating room (Fig. 1). In the majority of patients (23 of 30 cases) a recognizable SSVEP pattern was present in less than 75% of the intraoperative traces and in about half of the patients we were unable to identify the SSVEP in less than half of the traces (Fig. 2). Correlation analysis suggests that the intraoperative reproducibility of SSVEP is not a function of its amplitude. Neither the amplitudes recorded in the awake patients nor those recorded in the anaesthetized patients showed a significant correlation to the percentage of traces with recognizable SSVEP recorded intraoperatively. Furthermore, confirmed by continuous observation of the patient and the configuration of input and averaged signal during the recording procedure it can be ruled out that the intraindividual variability of SSVEP recording is affected by significant fluctuations of depth of anaesthesia or is a result of electrical artefacts. Compared to our experience with intraoperative transient VEP recording the results with SSVEP are moderately improved concerning the rate of traces with stable intraoperative recordings.
Considering the variability of intraoperative SSVEP recordings we are unable to give a conclusive explanation at this time. Several studies demonstrate that VEP are a result of complex cortical signal processing and are modulated, e.g. by spontaneous electro encephalograph activity , pharmacological agents  and higher cortical functions such as attention [17-20,23]. One may speculate that effects of anaesthetic agents are to some degree variable in different individuals and thus contribute to the variability of SSVEP recordings. Furthermore for recordings of steady-state auditory evoked responses it has been demonstrated that depth of anaesthesia clearly affects amplitudes and waveforms . Our findings do not rule out that different intensities of surgical stimulation together with minor fluctuations of depth of anaesthesia may contribute to the variations in intraoperative SSVEP recordings.
In conclusion, our data suggest that the shorter acquisition time, the characteristic waveform and the slightly enhanced stability are an advantage of SSVEP for intraoperative recordings compared to transient VEP. Nevertheless both, the inter- and intraindividual variability of the recordings are a clear limitation of this technique for intraoperative monitoring of visual pathways in routine clinical use. Further improvement is necessary and our hypothesis for ongoing studies is that a key point lies in the use of a modified anaesthesiological regimen.
1. Handel N, Law J, Hoehn R, Kirsch W. Monitoring visual evoked response during craniofacial surgery. Ann Plast Surg
2. Harding GF, Bland JD, Smith VH. Visual evoked potential monitoring of optic nerve function during surgery. J Neurol Neurosurg Psychiat
3. Herzon GD, Zealear DL. Intraoperative monitoring
of the visual evoked potential during endoscopic sinus surgery. Otolaryngol Head Neck Surg
4. Chacko AG, Babu KS, Chandy MJ. Value of visual evoked potential monitoring during trans-sphenoidal pituitary surgery. Br J Neurosurg
5. Hussain SS, Laljee HC, Horrocks JM, Tec H, Grace AR. Monitoring of intra-operative visual evoked potentials during functional endoscopic sinus surgery (FESS) under general anaesthesia. J Laryngol Otol
6. Raudzens PA. Intraoperative monitoring
of evoked potentials. Ann NY Acad Sci
7. Nau HE, Hess W, Pohlen G, Marggraf G, Rimpel J. Evoked potentials in intracranial operations: current status and our experiences. Anaesthesist
8. Cedzich C, Schramm J, Mengedoht CF, Fahlbusch R. Factors that limit the use of flash visual evoked potentials for surgical monitoring. Electroenceph Clin Neurophysiol
9. Wiedemayer H, Fauser B, Armbruster W, Gasser T, Stolke D. Visual evoked potentials for intraoperative neurophysiologic monitoring using total intravenous anesthesia. J Neurosurg Anesthesiol
10. Fagan Jr JE, Allen RG, Yolton RL. Factors contributing to amplitude variability of the steady-state visual evoked response. Am J Optom Physiol Opt
11. Fagan Jr JE, Downard FG, Yolton RL. Steady-state visual evoked response amplitudes and concurrent electroencephalographic activity. Am J Optom Physiol Opt
12. Sebel PS, Ingram DA, Flynn PJ, Rutherfoord CF, Rogers H. Evoked potentials during isoflurane anaesthesia. Br J Anaesth
13. Scheepstra GL, de Lange JJ, Booij LH, Ros HH. Median nerve evoked potentials during propofol anaesthesia. Br J Anaesth
14. Zentner J, Albrecht T, Heuser D. Propofol increases amplitudes of SEP. Funct Neurol
15. Taniguchi M, Nadstawek J, Pechstein U, Schramm J. Total intravenous anesthesia for improvement of intraoperative monitoring
of somatosensory evoked potentials during aneurysm surgery. Neurosurgery
16. Wiedemayer H, Fauser B, Sandalcioglu IE, Schafer H, Stolke D. The impact of neurophysiological intraoperative monitoring
on surgical decisions: a critical analysis of 423 cases. J Neurosurg
17. Silberstein RB, Ciorciari J, Pipingas A. Steady-state visually evoked potential topography during the Wisconsin card sorting test. Electroenceph Clin Neurophysiol
18. Di Russo F, Spinelli D. Electrophysiological evidence for an early attentional mechanism in visual processing in humans. Vision Res
19. Muller MM, Hillyard S. Concurrent recording of steady-state and transient event-related potentials as indices of visual-spatial selective attention. Clin Neurophysiol
20. Silberstein RB, Line P, Pipingas A, Copolov D, Harris P. Steady-state visually evoked potential topography during the continuous performance task in normal controls and schizophrenia. Clin Neurophysiol
21. Pigeau RA, Frame AM. Steady-state visual evoked responses in high and low alpha subjects. Electroenceph Clin Neurophysiol
22. Thompson JC, Tzambazis K, Stough C, Nagata K, Silberstein RB. The effects of nicotine on the 13 Hz steady-state visually evoked potential. Clin Neurophysiol
23. Di Russo F, Spinelli D. Effects of sustained, voluntary attention on amplitude and latency of steady-state visual evoked potential: a costs and benefits analysis. Clin Neurophysiol
24. Munglani R, Andrade J, Sapsford DJ, Baddeley A, Jones JG. A measure of consciousness and memory during isoflurane administration: the coherent frequency. Br J Anaesth