- Email: [email protected]
Can somatosensory and visual evoked potentials predict neurological outcome during targeted temperature management in post cardiac arrest patients?
Accepted Manuscript Title: Can somatosensory and visual evoked potentials predict neurological outcome during targeted temperature management in post cardiac arrest patients? Authors: Seung Pill Choi, Kyu Nam Park, Jung Hee Wee, Jeong Ho Park, Chun Song Youn, Han Joon Kim, Sang Hoon Oh, Yoon Sang Oh, Soo Hyun Kim, Joo Suk Oh PII: DOI: Reference:
S0300-9572(17)30265-4 http://dx.doi.org/doi:10.1016/j.resuscitation.2017.06.022 RESUS 7226
To appear in:
Resuscitation
Received date: Revised date: Accepted date:
30-3-2017 30-5-2017 20-6-2017
Please cite this article as: Choi Seung Pill, Park Kyu Nam, Wee Jung Hee, Park Jeong Ho, Youn Chun Song, Kim Han Joon, Oh Sang Hoon, Oh Yoon Sang, Kim Soo Hyun, Oh Joo Suk.Can somatosensory and visual evoked potentials predict neurological outcome during targeted temperature management in post cardiac arrest patients?.Resuscitation http://dx.doi.org/10.1016/j.resuscitation.2017.06.022 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Can somatosensory and visual evoked potentials predict neurological outcome during targeted temperature management in post cardiac arrest patients? Seung Pill Choi1, Kyu Nam Park1,*, Jung Hee Wee1, Jeong Ho Park1, Chun Song Youn1 Han Joon Kim1, Sang Hoon Oh1, Yoon Sang Oh2, Soo Hyun Kim1, Joo Suk Oh1
1
Department of Emergency Medicine, College of Medicine, The Catholic University of
Korea, Seoul, Republic of Korea; 2Department of Neurology, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea
Word count : 2824
Address for correspondence Kyu Nam Park, M.D., Ph.D. Department of Emergency Medicine Seoul St. Mary’s Hospital 222, Banpo-daero, Seocho-Gu Seoul, 137-701, Republic of Korea Tel: 82-2-2258-1987 Fax: 82-2-2258-1997
E-mail: [email protected]
Competing interests The authors have no competing interests to declare.
Can somatosensory and visual evoked potentials predict neurological outcome during targeted temperature management in post-cardiac arrest patients?
Purposes: In cardiac arrest patients treated with targeted temperature management (TTM), it is not certain if somatosensory evoked potentials (SEPs) and visual evoked potentials (VEPs) can predict neurological outcomes during TTM. The aim of this study was to investigate the prognostic value of SEPs and VEPs during TTM and after rewarming. Methods: This retrospective cohort study included comatose patients resuscitated from cardiac arrest and treated with TTM between March 2007 and July 2015. SEPs and VEPs were recorded during TTM and after rewarming in these patients. Neurological outcome was assessed at discharge by the Cerebral Performance Category (CPC) Scale. Results: In total, 115 patients were included. A total of 175 SEPs and 150 VEPs were performed. Five SEPs during treated with TTM and nine SEPs after rewarming were excluded from outcome prediction by SEPs due to an indeterminable N20 response because of technical error. Using 80 SEPs and 85 VEPs during treated with TTM, absent SEPs yielded a sensitivity of 58% and a specificity of 100% for poor outcome (CPC 3-5), and absent VEPs predicted poor neurological outcome with a sensitivity of 44% and a specificity of 96%. The AUC of combination of SEPs and VEPs was superior to either test
alone (0.788 for absent SEPs and 0.713 for absent VEPs compared with 0.838 for the combination). After rewarming, absent SEPs and absent VEPs predicted poor neurological outcome with a specificity of 100%. When SEPs and VEPs were combined, VEPs slightly increased the prognostic accuracy of SEPs alone. Although one patient with absent VEP during treated with TTM had a good neurological outcome, none of the patients with good neurological outcome had an absent VEP after rewarming. Conclusion: Absent SEPs could predict poor neurological outcome during TTM as well as after rewarming. Absent VEPs may predict poor neurological outcome in both periods and VEPs may provide additional prognostic value in outcome prediction.
Key words: somatosensory evoked potentials; visual evoked potentials; cardiac arrest; targeted temperature management
Introduction Despite advances in cardiopulmonary resuscitation (CPR) and critical care medicine, early prediction of neurological outcome for patients resuscitated from cardiac arrest is still challenging. Before targeted temperature management (TTM) is incorporated as a standard therapy for cardiac arrest patients, bilateral absence of the N20 component in the median nerve somatosensory evoked potentials (SEPs) recorded on days 1 to 3 or later after CPR has been reported to accurately predict poor outcome.1 However, the predictive value of SEPs recorded within 24 hours of initiating TTM has not been fully revealed. Visual evoked potentials (VEPs) are reliable prognostic indicators for perinatal asphyxia.2 Normal and abnormal/absent VEPs are associated with normal and abnormal outcome, respectively.3,4 Yet VEPs have not been used as an outcome predictor for adult cardiac arrest patients, and it is not known whether VEPs are influenced by TTM. Therefore, we investigated the prognostic value of SEPs and VEPs during TTM and after rewarming and whether VEPs can provide additional prognostic value to SEPs for patients treated with
TTM.
Methods Patients This study was reviewed and approved by the local ethics committee of our hospital. Between March 2007 and July 2015, we retrospectively studied adult patients (≥ 18 years) admitted to the emergency department of Yeouido St. Mary’s Hospital (a university teaching hospital in Seoul, Korea) who were treated with TTM after successful resuscitation from out-of- or in-hospital cardiac arrest. All unconscious patients who were resuscitated from non-traumatic cardiac arrest were eligible for the TTM irrespective of initial rhythm. The exclusion criteria for the TTM were trauma, intracranial haemorrhage, haemodynamic instability unresponsive to volume resuscitation and vasopressor treatment, known terminal illness, poor pre-arrest neurologic status and a “do not attempt resuscitation” order. Cooling was initiated to achieve a target temperature of 33°C using cold saline and an endovascular cooling device (CoolGard Thermal Regulation System, Alsius Corporation, Irvine, CA, USA) or external cooling device using self-adhesive gel-coated pads (Arctic Sun, Bard Medical, Louisville, CO, USA). Sedation (midazolam, 0.08 mg/kg intravenously) and paralysis (rocuronium, 0.8 mg/kg intravenously) were administered for shivering control, followed by continuous infusion of midazolam (0.04-0.2 mg/kg/h) and rocuronium (0.3-0.6 mg/kg/h). After a 24 hour period of maintenance, patients were slowly rewarmed to normothermia at a rate of 0.25°C/h. Sedation and paralysis were stopped at 35°C. SEPs and VEPs were recorded during TTM and after rewarming in these patients following the clinical pathways and the protocol of our hospital. The patients were evaluated in terms of age, gender, cause of death, if the collapse was witnessed, if a bystander performed CPR, the initial electrocardiogram (ECG), the initial Glasgow Coma Scale (GCS) score, the initial neurologic examination, the body temperature when an EP was performed, the time between evoked potentials and ROSC, and the cerebral performance category (CPC) score at discharge (Table 1). Good neurologic outcome was defined as CPC 1-2 (no or moderate
disability) and poor neurologic outcome as CPC 3-5 (severely disabled, comatose, or dead).
SEP recordings SEPs were recorded using a Viking IV (Nicolet Instruments, Madison, WI, USA) during TTM and after rewarming of the patients. EEG disk electrodes were placed at both Erb’s points, spinous process C7 and C3’ and C4’ following the International 10-20 system with skin electrode impedance below 5 kΩ. They were referenced to the mid frontal region at Fz. All patients were examined bilaterally. Electrical stimulation (impulse duration 0.2 ms; stimulation rate 3-5 Hz; and intensity 7-15 mA) was delivered overlying the median nerve at the level of the wrist. At least 200 stimulations were averaged per recording. In each SEP, the latencies to N9, N13 and N20 and the cervicocortical conduction time and N20-P23 amplitude were evaluated. SEPs were classified as SEP absent (bilaterally absent cortical N20 responses after left and right median nerve stimulation in the presence of a cervical potential), SEP present (cortical N20 response present on at least one side), or indeterminable (technically insufficient recording).
VEP recordings VEPs were recorded from a single active electrode at Oz referenced to Fz. The stimulus was a binocular flash from light-emitting diode (LED) goggles that were placed over the patient’s closed eyes. Impedances were kept below 5 kΩ. A band pass of 0.5-100 Hz was employed. Stimulation frequency was 1.3 Hz and two averages of 100 stimulations were repeated to yield reproducible results for each patient. In each VEP, the latencies to N75, P100 and N145 and N75-P100 amplitude were evaluated. VEPs were classified as VEP absent (bilaterally absent cortical P100 responses) or VEP present (cortical P100 response present on at least one side).
Statistical analysis Quantitative data was given as the mean and standard deviation or the median and interquartile range and qualitative data as counts and percentages. Chi-square test was used
to assess qualitative data. T-test or Wilcoxon signed rank sum test was used to compare the quantitative data. To evaluate the prognostic value of SEPs and VEPs, receiver operating characteristic (ROC) analysis was performed. Means of left and right evoked potentials were used for calculations of latencies and amplitudes. Amplitudes and latencies were calculated from cases having N20 waves in the SEPs or P100 waves in the VEPs during TTM and after rewarming. In case of asymmetry between left and right on SEPs or VEPs, the better response was used for statistical comparisons of each component between the good outcome (CPC 1-2) and the poor outcome (CPC 3-5) group. A p value of < 0.05 was considered significant. Statistical analysis was performed using the Statistical Package for Social Sciences (SPSS) version 18.0 (SPSS Inc., Chicago, IL, USA).
Results In total, the study included 115 comatose patients who were treated with TTM after cardiac arrest between March 2007 to July 2015 (Supplemental Fig. 1). Among them, 30 patients during TTM did not record SEPs and VEPs because recordings were performed during normal working hours. Fourteen patients after rewarming died early before SEPs and VEPs could be repeated. Per physician decisions, 11 SEPs and 36 VEPs were not recorded after rewarming (Supplemental Table 1). Eventually, 85 SEPs and 85 VEPs were performed during TTM and 90 SEPs and 65 VEPs after rewarming. Five SEPs during TTM and nine SEPs after rewarming were excluded from outcome prediction by SEPs due to an indeterminable N20 response because of technical error. Demographic and clinical characteristics of the 115 patients are presented in Table 1. The mean core temperature was 33℃ for EPs during TTM and 36.8℃ for EPs after rewarming. The mean time from ROSC to EPs during TTM was 21.0 hours, while for EPs after rewarming, the mean time from ROSC was 70.2 hours. Thirty-one patients (27%) achieved a good neurological outcome at discharge. Outcome prediction using SEPs and VEPs. Using 80 SEPs and 85 VEPs during TTM, absent SEPs yielded a sensitivity of 58% and a specificity of 100% for poor neurological
outcome and absent VEPs predicted poor neurological outcome with a sensitivity of 44% and a specificity of 96% (Table 2). The AUC of combination of absent SEPs and absent VEPs was 0.838 (95% C.I. 0.749-0.927) which was also superior to either absent SEPs alone or absent VEPs alone (p < 0.001 and p = 0.002, respectively) (Fig 1). Out of eight patients with SEPs detected, seven were correctly predicted to have a poor neurological outcome by absent VEPs except for one patient, increasing the sensitivity by 13%. For one patient with indeterminable SEPs, a poor neurological outcome could also be predicted with absent VEPs (Table 3). After rewarming, absent SEPs predicted poor neurological outcome with a sensitivity of 59% and a specificity of 100% and absent VEPs with a sensitivity of 47% and a specificity of 100% (Table 2). When SEPs and VEPs were combined, VEPs slightly increased the prognostic accuracy of SEPs alone (Fig. 1). Three patients with present SEPs and six patients with indeterminable SEPs could be predicted to have poor neurological outcome using absent VEPs. Although one patient with absent VEPs during TTM had a good neurological outcome, none of the patients with a good neurological outcome had absent VEPs after rewarming (Table 3). Latency and amplitude of SEPs and VEPs. The amplitudes and latencies were calculated only from cases having N20 waves in the SEPs or P100 waves in the VEPs both during TTM and after rewarming. Mean values of latency and amplitudes of SEP and VEP components are presented in Supplemental Table 2. In the SEPs, the latencies to N13 responses and to early cortical N20 responses were significantly longer during TTM than after rewarming while cervicocortical conduction time and N20-P23 amplitude were not significantly different between both periods. With regard to the VEPs, the latencies to N75 responses, to P100 responses, and to N145 responses were not significantly different between during TTM and after rewarming. The left-sided N75-P100 amplitude was significantly higher after rewarming compared to during TTM, but the right-sided N75-P100 amplitude was not. When each component of SEPs was compared between the good and the poor outcome during TTM and after rewarming, the N20 peak-to-peak amplitude was significantly higher in the good outcome group than the poor outcome group, although the cut-off value for outcome could not be determined (p=0.018 during TTM and
p=0.012 after rewarming) (Table 4) (Fig. 2). However, there was no significant difference in all components of VEPs in both periods between the two outcome groups.
Discussion The results of our study indicate that in post-cardiac arrest patients treated with TTM, absent SEPs could predict poor neurological outcome during TTM and after rewarming. Absent VEPs may predict poor neurological outcome in both periods and may increase the predictive power of SEPs. To the best of our knowledge, this is the first report using VEPs for outcome prediction in adult post-cardiac arrest patients, although VEPs have been commonly used to predict neurological outcome and visual prognosis in perinatal asphyxia.2-6 A VEP has several positive and negative peaks observed at various times after presentation of the stimuli. Of these, P100 is relatively stable and the most useful clinically. The P100 is thought to arise 100 msec after the stimuli in the striated and parastriated visual cortex.7 Hildebrandt et al. reported that in 8 patients who recovered from vegetative state and 13 who remained in vegetative state, brain perfusion and VEP reactivity in the occipital and parietal areas were associated with recovery from hypoxic vegetative state, showing that the recovered patients had no absent VEP and higher perfusion in the visual cortex and in the precuneus.8 In this study, all 10 patients with absent VEP remained in the vegetative state, and 7 of 10 patients with present VEP recovered from vegetative state. Absent VEPs, therefore, can become a predictor of poor outcome for post-cardiac arrest patients treated with TTM. McCulloch et al. showed that in 25 children with perinatal asphyxia, a strong association was found between normal, abnormal, or absent VEP in the early postnatal period and long-term visual outcome; only one of six patients with an absent VEP recovered normal vision, and blindness and visual impairment were always associated with abnormal neurologic outcomes.5 Taylor et al. described that in the 57 asphyxiated term infants who had SEPs and VEPs recorded during the first three days of life, the first week and at follow-up visits, normal SEPs virtually guaranteed normal outcome and abnormal
VEPs guaranteed abnormal outcome.6 Both together had a higher predictive power than either alone. The combination of VEPs and SEPs yields a powerful means of prognostication for full-term asphyxiated infants. In our study, one of 26 every patients who had bilaterally absent P100 on VEPs during TTM recovered with a good outcome after rewarming, resulting in a 96% specificity and 4% false positive rate (FPR) for poor outcome. Twenty three patients with absent VEPs after rewarming did not regain consciousness. For effects of low temperature on VEPs, Russ et al. reported progressive latency prolongation and amplitude reduction of VEP with hypothermia, leading to a complete loss of waves at 25-27℃.9 Burrows et al. reported that in neonates and infants undergoing profound hypothermic circulatory arrest, VEPs disappeared at a nasopharyngeal temperature of 18.9±2.8℃ and reappeared at a nasopharyngeal temperature of 32.8±1.4℃. With faster cooling, VEPs disappeared at a higher temperature than with slower cooling.10 VEPs, therefore, may disappear at a core temperature of 33℃ if the patient initially has VEPs with a low amplitude of P100. Therefore, we suggest that an absence of VEPs during TTM should not be used for predicting a poor outcome and subsequent withdrawal of a life-sustaining treatment but instead be used as a supportive predictor of SEPs, which would call for further VEP recordings after rewarming. In TTM-treated patient, a bilaterally absent SEPs accurately predicts poor neurological outcome both during TTM (FPR≤2) and after rewarming (FPR≤1)11-13 although two cases of false positive prediction have been reported at the 48-72 hours after ROSC.14,15 Our results showed all patients with absent SEPs during TTM (n=30) and after rewarming (n=32) had poor neurological outcome, indicating 0% FPR, which is in line with a few studies.16,17 In a prospective study by Grippo et al., 60 patients had SEPs during hypothermia and after rewarming and none of the patients with absent SEPs, including 24 patients during hypothermia and 20 patients after rewarming, recovered consciousness (GOS 1-2).17 They insisted that absent SEPs during hypothermia retain a prognostic value for poor neurological outcome, as in normothermic patients. Bowes et al. noted that 40 of 43 patients with absent SEPs during hypothermia and 42 patients with absent SEPs after
rewarming had poor neurological outcome (GOS 1-3).18 When the 3 patients with absent SEPs during hypothermia who recovered to a good outcome were reevaluated by 2 experienced neurophysiologists, the results were undeterminable because of too much noise. When the recordings are obtained under uncertain technical conditions, SEPs should not be classified as absent, and further recordings are essential.17,18 Considering our results of recording SEPs, bilaterally absent N20 peaks remained a strong noninvasive predictor for poor neurologic outcome during TTM as well as after rewarming. Our study included 13 indeterminable SEPs (5 during TTM and 8 after rewarming), which were excluded from outcome prediction due to not having cervical waves. In these cases, VEPs performed together with SEPs could contribute prognostic information about the patients. With regard to the location of cortical injury in hypoxic ischaemic encephalopathy, the Rolandic, parietal, and occipital cortices are more frequently injured than the frontal and temporal cortices.19,20 Considering that SEPs and VEPs reflect the functions of the primary sensory cortex of the parietal lobe and the visual cortex of occipital lobe, respectively, both absent SEPs and absent VEPs can increase the accuracy of poor outcome prediction with FPRs close to 0%. Interestingly, in 3 patients, cortical responses of SEPs were absent during TTM but recovered after rewarming, and none of them recovered consciousness. Recently, a few studies reported that there was no reappearance of cortical response on SEPs after rewarming with initially absent SEPs during TTM.17,21 The results of our study indicate that irrespective of the reappearance of cortical responses on SEPs after rewarming, absent SEPs during TTM may retain prognostic power for poor neurological outcome, with need of further confirmation in larger study. Latencies of different SEP components increased during TTM but the amplitudes of cortical SEPs was not significantly different between both periods, supporting prior studies .16,17 Reduced SEP amplitude is associated with a poor outcome.22,23 Endisch et al. demonstrated that the lowest amplitude of SEP in a survivor with a good outcome (CPC 1-3) was 0.62 µV22 and Logi et al. noted that none of the patients with an SEP amplitude of < 1.2 µV recovered consciousness.23 In contrast, our study showed that the lowest amplitudes in a patient with a good outcome (CPC 1-2) were
0.726 µV during TTM and 0.465 µV after rewarming. Therefore, we think that the cut-off of SEP amplitude cannot be determined for predicting poor outcome because the recordings are performed in different conditions between studies. There are some limitations in this study. First, it is a retrospective study and although our patients were treated in accordance with the clinical pathways and the protocol for post-cardiac arrest patients of our hospital, there may be some selection bias because some patients did not have SEPs and VEPs. Second, in Korea, by law, we do not withdraw a life-sustaining treatment, and the treating physician was not blinded to
the results of
the SEP and VEP recordings, which could influence patient outcomes. Third, only 34 of 115 patients in this study had SEPs and VEPs together in both periods, so we had difficulty identifying the relationship of changes in SEPs and VEPs during TTM and after rewarming to neurological outcome. Therefore, a further prospective study that includes a larger number of patients is necessary. In conclusion, in post-cardiac arrest patients treated with TTM, absent SEPs could predict poor neurological outcome during TTM as well as after rewarming. Absent VEPs may predict poor neurological outcome in both periods and VEPs may provide additional prognostic value in outcome prediction.
References 1.
Wijdicks EF, Hijdra A, Young GB, Bassetti CL, Wiebe S; Quality Standards Subcommittee of the American Academy of Neurology. Practice parameter: prediction of outcome in comatose survivors after cardiopulmonary resuscitation (an evidence-based review): report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology 2006;67:203-10.
2.
van Laerhoven H, de Haan TR, Offringa M, Post B, van der Lee JH. Prognostic tests in term neonates with hypoxic-ischemic encephalopathy: a systematic review. Pediatrics 2013;131:88-98.
3.
Whyte HE. Visual-evoked potentials in neonates following asphyxia. Clin Perinatol 1993;20:451-61.
4.
Muttitt SC, Taylor MJ, Kobayashi JS, MacMillan L, Whyte HE. Serial visual evoked potentials and outcome in term birth asphyxia. Pediatr Neurol 1991;7:86-90.
5.
McCulloch DL, Taylor MJ, Whyte HE. Visual evoked potentials and visual prognosis following perinatal asphyxia. Arch Ophthalmol 199;109:229-33.
6.
Taylor MJ, Murphy WJ, Whyte HE. Prognostic reliability of somatosensory and visual evoked potentials of asphyxiated term infants. Dev Med Child Neurol 1992;34:507-15.
7.
Chi OZ, Subramoni J, Jasaitis D. Visual evoked potentials during etomidate administration in humans. Can J Anaesth 1990;37:452-6.
8.
Hildebrandt H, Happe S, Deutschmann A, Basar-Eroglu C, Eling P, Brunhöber J. Brain perfusion and VEP reactivity in occipital and parietal areas are associated to recovery from hypoxic vegetative state. J Neurol Sci 2007;260:150-8.
9.
Russ W, Kling D, Loesevitz A, Hempelmann G. Effect of hypothermia on visual evoked potentials (VEP) in humans. Anesthesiology 1984;61:207-10.
10.
Burrows FA, Hillier SC, McLeod ME, Iron KS, Taylor MJ. Anterior fontanel
pressure and visual evoked potentials in neonates and infants undergoing profound hypothermic circulatory arrest. Anesthesiology 1990;73:632-6. 11.
Kamps MJ, Horn J, Oddo M, Fugate JE, Storm C, Cronberg T, Wijman CA, Wu O, Binnekade JM, Hoedemaekers CW. Prognostication of neurologic outcome in cardiac arrest patients after mild therapeutic hypothermia: a meta-analysis of the current literature. Intensive Care Med 2013;39:1671 -82.
12.
Sandroni C, Cariou A, Cavallaro F, Cronberg T, Friberg H, Hoedemaekers C, Horn J, Nolan JP, Rossetti AO, Soar J. Prognostication in comatose survivors of cardiac arrest: an advisory statement from the European Resuscitation Council and the European Society of Intensive Care Medicine. Resuscitation 2014;85:1779-89.
13.
Nolan JP, Soar J, Cariou A, Cronberg T, Moulaert VR, Deakin CD, Bottiger BW, Friberg H, Sunde K, Sandroni C. European Resuscitation Council and European Society of Intensive Care Medicine Guidelines for Post-resuscitation Care 2015: Section 5 of the European Resuscitation Council Guidelines for Resuscitation 2015. Resuscitation 2015;95:202-22.
14.
Leithner C, Ploner CJ, Hasper D, Storm C. Does hypothermia influence the predictive value of bilateral absent N20 after cardiac arrest? Neurology 2010;74:965-9.
15.
Arch AE, Chiappa K, Greer DM. False positive absent somatosensory evoked potentials
in
cardiac
arrest
with
therapeutic hypothermia. Resuscitation
2014;85:e97-8. 16.
Tiainen M, Kovala TT, Takkunen OS, Roine RO. Somatosensory and brainstem auditory evoked potentials in cardiac arrest patients treated with hypothermia. Crit Care Med 2005;33:1736-40.
17.
Grippo A, Carrai R, Fossi S, Cossu C, Mazzeschi E, Peris A, Bonizzoli M, Ciapetti M, Gensini GF, Pinto F, Amantini A. Absent SEP during therapeutic hypothermia did not reappear after re-warming in comatose patients following cardiac arrest. Minerva Anestesiol 2013;79:360-9.
18.
Bouwes A, Binnekade JM, Kuiper MA, Bosch FH, Zandstra DF, Toornvliet AC, Biemond HS, Kors BM, Koelman JH, Verbeek MM, Weinstein HC, Hijdra A, Horn J. Prognosis of coma after therapeutic hypothermia: a prospective cohort study. Ann Neurol 2012;71:206-12.
19.
Choi SP, Park KN, Park HK, Kim JY, Youn CS, Ahn KJ, Yim HW. Diffusion-weighted magnetic resonance imaging for predicting the clinical outcome of comatose survivors after cardiac arrest: a cohort study. Crit Care 2010;14:R17.
20.
Järnum H, Knutsson L, Rundgren M, Siemund R, Englund E, Friberg H, Larsson EM. Diffusion and perfusion MRI of the brain in comatose patients treated with mild
hypothermia
after
cardiac
arrest:
a
prospective
observational
study. Resuscitation 2009;80:425–430. 21.
Cloostermans MC, van Meulen FB, Eertman CJ, Hom HW, van Putten MJ. Continuous electroencephalography monitoring for early prediction of neurological outcome in postanoxic patients after cardiac arrest: a prospective cohort study. Crit Care Med 2012;40:2867-75.
22.
Endisch C, Storm C, Ploner CJ, Leithner C. Amplitudes of SSEP and outcome in cardiac arrest survivors: A prospective cohort study. Neurology 2015;85:1752-60.
23.
Logi F, Fischer C, Murri L, Mauguière F. The prognostic value of evoked responses from primary somatosensory and auditory cortex in comatose patients. Clin Neurophysiol 2003;114:1615-27.
Fig. 1. Receiver operating characteristic curve for the prediction of neurological outcome according to SEPs, VEPs, and both during TTM (A) and after rewarming (B)
Fig. 2. N20 peak-to-peak amplitude (µV) of good outcome group (◇) and poor outcome group (◆) in the SEP during TTM (A) and after rewarming (B) in the same patients
Table 1. Patients demographics and clinical characteristics. Total patients
Good neurological
Poor neurological
(n=115)
outcome (n=31)
outcome (n=84)
Male
77 (67.0%)
23 (74.2%)
54 (64.3%)
0.316
Age
54.7 ± 16.6
50.2 ± 16.3
56.4 ± 16.5
0.077
Witnessed arrest
70 (60.9%)
25 (80.6%)
45 (53.6%)
0.008
Bystander CPR
41 (35.7%)
15 (48.4%)
26 (31.0%)
0.083
Shockable rhythm
17 (14.8%)
11 (35.5%)
6 (7.1%)
<0.001
48 (39%)
25 (80.6%)
23 (27.3%)
<0.001
109 (94.8%)
29 (93.5%)
80 (95.2%)
0.718
3 (3, 3)
3 (3, 5)
3 (3, 3)
<0.001
Absent PLR after ROSC
63 (54.7%)
8 (25.8%)
55 (65.5%)
<0.001
Absent CR after ROSC
104 (90.4%)
23 (74.2%)
81 (96.4%)
0.001
Time from ROSC to EP during TTM (n=85) (h)
21.0 ± 6.7
19.9 ± 7.3
21.5 ± 6.4
0.299
Time from ROSC to EP after rewarming (n=90) (h)
70.2 ± 20.5
74.3 ± 23.1
68.3 ± 19.2
0.204
Body temperature when EP during TTM (n=85) (℃)
33.0 ± 0.6
33.3 ± 0.5
32.9 ± 0.7
0.027
Body temperature when EP after rewarming (n=90) (℃)
36.8 ± 0.7
36.9 ± 0.5
36.7 ± 0.8
0.247
CPC 1
27 (23.5%)
27 (87.1%)
0 (0.0%)
CPC 2
4 (3.5%)
4 (12.9%)
0 (0.0%)
CPC 3
1 (0.9%)
0 (0.0%)
1 (1.2%)
CPC 4
35 (30.4%)
0 (0.0%)
35 (41.7%)
CPC 5
48 (41.7%)
0 (0.0%)
48 (57.1%)
Cardiac cause OHCA GCS after ROSC
Neurological outcome at discharge
p
Values are presented as mean ± SD, median (IQR), and frequency (%). CPR: cardiopulmonary resuscitation, OHCA: out-of-hospital cardiac arrest, GCS: Glasgow Coma Scale, PLR: pupillary light reflex, CR: corneal response, ROSC: return of spontaneous circulation, EP: evoked potential, TTM: targeted temperature management.
Table 2. Statistical evaluation of somatosensory and visual evoked potentials for the prediction of poor neurological outcome in 115 patients
Modality
n†
Good/poor
Sensitivity
Specificity
PPV
NPV
SEP
80
28/52
58
100
100
54
(43-71)
(85-100)
(86-100)
(41-70)
44
96
96
46
(31-58)
(80-100)
(78-100)
(33-59)
71
96
97
64
(57-82)
(80-100)
(85-100)
(48-78)
67
96
97
59
(53-78)
(80-100)
(85-100)
(43-73)
59
100
100
55
(45-72)
(87-100)
(89-100)
(40-69)
47
100
100
38
(33-62)
(79-100)
(85-100)
(24-54)
63
100
100
52
(47-77)
(76-100)
(89-100)
(33-69)
During TTM (n=85)
VEP
SEP+VEP SEP+VEP††
85
80
85
28/57
28/52
28/57
After rewarming (n=90)
SEP
VEP
SEP+VEP
81
65
57
27/54
16/49
16/41
SEP+VEP††
65
16/49
65
100
100
48
(50-78)
(76-100)
(87-100)
(31-66)
95% Confidence interval in parentheses, †: Number of patients with available data in each subgroup, ††: When SSEP is indeterminable, neurological outcome is predicted by VEP. PPV: positive predictive value, NPV: negative predictive value, FPR: false positive rate, SEP: somatosensory evoked potentials, VEP: visual evoked potentials, TTM: targeted temperature management
Table 3. Relationship between SEPs and VEPs and neurological outcome
TTM (n=85)
After rewarming (n=90)
SEP
VEP
Outcome (n)
Total (n)
SEP
VEP
Outcome (n)
Total (n)
SEP (Ind)
VEP (-)
PO (1)
1
SEP (Ind)
VEP(-)
PO (6)
6
SEP (Ind)
VEP (+)
PO (3)
4
SEP (Ind)
VEP(+)
PO (2)
2
8
SEP (+)
VEP(-)
PO (3)
3
42
SEP (+)
VEP(+)
PO (14)
31
GO (1) SEP (+)
VEP (-)
PO (7) GO (1)
SEP (+)
VEP (+)
PO (15) GO (27)
GO (17)
SEP (-)
VEP (-)
PO (17)
17
SEP (-)
VEP(-)
PO (14)
14
SEP (-)
VEP (+)
PO (13)
13
SEP (-)
VEP(+)
PO (9)
9
SEP (Ind)
Unchecked
PO (1)
1
SEP (+)
Unchecked
PO (4)
15
GO (11) SEP (-)
Unchecked
PO (9)
9
SEP: somatosensory evoked potentials, VEP: visual evoked potentials, Ind: indeterminable, (+): presence, (-): absence PO: poor neurological outcome, GO: good neurological outcome
Table 4. The comparisons of each component between both outcome groups in the same patients who have N20 wave in the SEP (n=30)
TTM
After rewarming
N13 lat
N20 lat
CCT
N20-P23
N13
(ms)
(ms)
(ms)
Amp (µV)
(ms)
Good
14.71 ±
22.10 ±
7.25 ±
(n=22)
1.12
1.48
1.16
Poor
14.61 ±
22.78 ±
7.98 ±
(n=8)
1.18
2.85
2.16
p
0.909
0.982
0.597
2.10 ± 1.13
1.08 ± 0.91 0.018
N20 lat
CCT
N20-P23
(ms)
(ms)
Amp (µV)
19.14 ±
5.71 ±
1.11
1.27
21.98 ±
8.93 ±
1.66
4.35
5.12
0.534
0.170
0.298
13.25
lat
±
0.99 13.03
CCT: cervicocortical conduction time, lat: latency, Amp: amplitude
±
2.48 ± 0.96
1.33 ± 1.10 0.012
S0300-9572(17)30265-4 http://dx.doi.org/doi:10.1016/j.resuscitation.2017.06.022 RESUS 7226
To appear in:
Resuscitation
Received date: Revised date: Accepted date:
30-3-2017 30-5-2017 20-6-2017
Please cite this article as: Choi Seung Pill, Park Kyu Nam, Wee Jung Hee, Park Jeong Ho, Youn Chun Song, Kim Han Joon, Oh Sang Hoon, Oh Yoon Sang, Kim Soo Hyun, Oh Joo Suk.Can somatosensory and visual evoked potentials predict neurological outcome during targeted temperature management in post cardiac arrest patients?.Resuscitation http://dx.doi.org/10.1016/j.resuscitation.2017.06.022 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Can somatosensory and visual evoked potentials predict neurological outcome during targeted temperature management in post cardiac arrest patients? Seung Pill Choi1, Kyu Nam Park1,*, Jung Hee Wee1, Jeong Ho Park1, Chun Song Youn1 Han Joon Kim1, Sang Hoon Oh1, Yoon Sang Oh2, Soo Hyun Kim1, Joo Suk Oh1
1
Department of Emergency Medicine, College of Medicine, The Catholic University of
Korea, Seoul, Republic of Korea; 2Department of Neurology, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea
Word count : 2824
Address for correspondence Kyu Nam Park, M.D., Ph.D. Department of Emergency Medicine Seoul St. Mary’s Hospital 222, Banpo-daero, Seocho-Gu Seoul, 137-701, Republic of Korea Tel: 82-2-2258-1987 Fax: 82-2-2258-1997
E-mail: [email protected]
Competing interests The authors have no competing interests to declare.
Can somatosensory and visual evoked potentials predict neurological outcome during targeted temperature management in post-cardiac arrest patients?
Purposes: In cardiac arrest patients treated with targeted temperature management (TTM), it is not certain if somatosensory evoked potentials (SEPs) and visual evoked potentials (VEPs) can predict neurological outcomes during TTM. The aim of this study was to investigate the prognostic value of SEPs and VEPs during TTM and after rewarming. Methods: This retrospective cohort study included comatose patients resuscitated from cardiac arrest and treated with TTM between March 2007 and July 2015. SEPs and VEPs were recorded during TTM and after rewarming in these patients. Neurological outcome was assessed at discharge by the Cerebral Performance Category (CPC) Scale. Results: In total, 115 patients were included. A total of 175 SEPs and 150 VEPs were performed. Five SEPs during treated with TTM and nine SEPs after rewarming were excluded from outcome prediction by SEPs due to an indeterminable N20 response because of technical error. Using 80 SEPs and 85 VEPs during treated with TTM, absent SEPs yielded a sensitivity of 58% and a specificity of 100% for poor outcome (CPC 3-5), and absent VEPs predicted poor neurological outcome with a sensitivity of 44% and a specificity of 96%. The AUC of combination of SEPs and VEPs was superior to either test
alone (0.788 for absent SEPs and 0.713 for absent VEPs compared with 0.838 for the combination). After rewarming, absent SEPs and absent VEPs predicted poor neurological outcome with a specificity of 100%. When SEPs and VEPs were combined, VEPs slightly increased the prognostic accuracy of SEPs alone. Although one patient with absent VEP during treated with TTM had a good neurological outcome, none of the patients with good neurological outcome had an absent VEP after rewarming. Conclusion: Absent SEPs could predict poor neurological outcome during TTM as well as after rewarming. Absent VEPs may predict poor neurological outcome in both periods and VEPs may provide additional prognostic value in outcome prediction.
Key words: somatosensory evoked potentials; visual evoked potentials; cardiac arrest; targeted temperature management
Introduction Despite advances in cardiopulmonary resuscitation (CPR) and critical care medicine, early prediction of neurological outcome for patients resuscitated from cardiac arrest is still challenging. Before targeted temperature management (TTM) is incorporated as a standard therapy for cardiac arrest patients, bilateral absence of the N20 component in the median nerve somatosensory evoked potentials (SEPs) recorded on days 1 to 3 or later after CPR has been reported to accurately predict poor outcome.1 However, the predictive value of SEPs recorded within 24 hours of initiating TTM has not been fully revealed. Visual evoked potentials (VEPs) are reliable prognostic indicators for perinatal asphyxia.2 Normal and abnormal/absent VEPs are associated with normal and abnormal outcome, respectively.3,4 Yet VEPs have not been used as an outcome predictor for adult cardiac arrest patients, and it is not known whether VEPs are influenced by TTM. Therefore, we investigated the prognostic value of SEPs and VEPs during TTM and after rewarming and whether VEPs can provide additional prognostic value to SEPs for patients treated with
TTM.
Methods Patients This study was reviewed and approved by the local ethics committee of our hospital. Between March 2007 and July 2015, we retrospectively studied adult patients (≥ 18 years) admitted to the emergency department of Yeouido St. Mary’s Hospital (a university teaching hospital in Seoul, Korea) who were treated with TTM after successful resuscitation from out-of- or in-hospital cardiac arrest. All unconscious patients who were resuscitated from non-traumatic cardiac arrest were eligible for the TTM irrespective of initial rhythm. The exclusion criteria for the TTM were trauma, intracranial haemorrhage, haemodynamic instability unresponsive to volume resuscitation and vasopressor treatment, known terminal illness, poor pre-arrest neurologic status and a “do not attempt resuscitation” order. Cooling was initiated to achieve a target temperature of 33°C using cold saline and an endovascular cooling device (CoolGard Thermal Regulation System, Alsius Corporation, Irvine, CA, USA) or external cooling device using self-adhesive gel-coated pads (Arctic Sun, Bard Medical, Louisville, CO, USA). Sedation (midazolam, 0.08 mg/kg intravenously) and paralysis (rocuronium, 0.8 mg/kg intravenously) were administered for shivering control, followed by continuous infusion of midazolam (0.04-0.2 mg/kg/h) and rocuronium (0.3-0.6 mg/kg/h). After a 24 hour period of maintenance, patients were slowly rewarmed to normothermia at a rate of 0.25°C/h. Sedation and paralysis were stopped at 35°C. SEPs and VEPs were recorded during TTM and after rewarming in these patients following the clinical pathways and the protocol of our hospital. The patients were evaluated in terms of age, gender, cause of death, if the collapse was witnessed, if a bystander performed CPR, the initial electrocardiogram (ECG), the initial Glasgow Coma Scale (GCS) score, the initial neurologic examination, the body temperature when an EP was performed, the time between evoked potentials and ROSC, and the cerebral performance category (CPC) score at discharge (Table 1). Good neurologic outcome was defined as CPC 1-2 (no or moderate
disability) and poor neurologic outcome as CPC 3-5 (severely disabled, comatose, or dead).
SEP recordings SEPs were recorded using a Viking IV (Nicolet Instruments, Madison, WI, USA) during TTM and after rewarming of the patients. EEG disk electrodes were placed at both Erb’s points, spinous process C7 and C3’ and C4’ following the International 10-20 system with skin electrode impedance below 5 kΩ. They were referenced to the mid frontal region at Fz. All patients were examined bilaterally. Electrical stimulation (impulse duration 0.2 ms; stimulation rate 3-5 Hz; and intensity 7-15 mA) was delivered overlying the median nerve at the level of the wrist. At least 200 stimulations were averaged per recording. In each SEP, the latencies to N9, N13 and N20 and the cervicocortical conduction time and N20-P23 amplitude were evaluated. SEPs were classified as SEP absent (bilaterally absent cortical N20 responses after left and right median nerve stimulation in the presence of a cervical potential), SEP present (cortical N20 response present on at least one side), or indeterminable (technically insufficient recording).
VEP recordings VEPs were recorded from a single active electrode at Oz referenced to Fz. The stimulus was a binocular flash from light-emitting diode (LED) goggles that were placed over the patient’s closed eyes. Impedances were kept below 5 kΩ. A band pass of 0.5-100 Hz was employed. Stimulation frequency was 1.3 Hz and two averages of 100 stimulations were repeated to yield reproducible results for each patient. In each VEP, the latencies to N75, P100 and N145 and N75-P100 amplitude were evaluated. VEPs were classified as VEP absent (bilaterally absent cortical P100 responses) or VEP present (cortical P100 response present on at least one side).
Statistical analysis Quantitative data was given as the mean and standard deviation or the median and interquartile range and qualitative data as counts and percentages. Chi-square test was used
to assess qualitative data. T-test or Wilcoxon signed rank sum test was used to compare the quantitative data. To evaluate the prognostic value of SEPs and VEPs, receiver operating characteristic (ROC) analysis was performed. Means of left and right evoked potentials were used for calculations of latencies and amplitudes. Amplitudes and latencies were calculated from cases having N20 waves in the SEPs or P100 waves in the VEPs during TTM and after rewarming. In case of asymmetry between left and right on SEPs or VEPs, the better response was used for statistical comparisons of each component between the good outcome (CPC 1-2) and the poor outcome (CPC 3-5) group. A p value of < 0.05 was considered significant. Statistical analysis was performed using the Statistical Package for Social Sciences (SPSS) version 18.0 (SPSS Inc., Chicago, IL, USA).
Results In total, the study included 115 comatose patients who were treated with TTM after cardiac arrest between March 2007 to July 2015 (Supplemental Fig. 1). Among them, 30 patients during TTM did not record SEPs and VEPs because recordings were performed during normal working hours. Fourteen patients after rewarming died early before SEPs and VEPs could be repeated. Per physician decisions, 11 SEPs and 36 VEPs were not recorded after rewarming (Supplemental Table 1). Eventually, 85 SEPs and 85 VEPs were performed during TTM and 90 SEPs and 65 VEPs after rewarming. Five SEPs during TTM and nine SEPs after rewarming were excluded from outcome prediction by SEPs due to an indeterminable N20 response because of technical error. Demographic and clinical characteristics of the 115 patients are presented in Table 1. The mean core temperature was 33℃ for EPs during TTM and 36.8℃ for EPs after rewarming. The mean time from ROSC to EPs during TTM was 21.0 hours, while for EPs after rewarming, the mean time from ROSC was 70.2 hours. Thirty-one patients (27%) achieved a good neurological outcome at discharge. Outcome prediction using SEPs and VEPs. Using 80 SEPs and 85 VEPs during TTM, absent SEPs yielded a sensitivity of 58% and a specificity of 100% for poor neurological
outcome and absent VEPs predicted poor neurological outcome with a sensitivity of 44% and a specificity of 96% (Table 2). The AUC of combination of absent SEPs and absent VEPs was 0.838 (95% C.I. 0.749-0.927) which was also superior to either absent SEPs alone or absent VEPs alone (p < 0.001 and p = 0.002, respectively) (Fig 1). Out of eight patients with SEPs detected, seven were correctly predicted to have a poor neurological outcome by absent VEPs except for one patient, increasing the sensitivity by 13%. For one patient with indeterminable SEPs, a poor neurological outcome could also be predicted with absent VEPs (Table 3). After rewarming, absent SEPs predicted poor neurological outcome with a sensitivity of 59% and a specificity of 100% and absent VEPs with a sensitivity of 47% and a specificity of 100% (Table 2). When SEPs and VEPs were combined, VEPs slightly increased the prognostic accuracy of SEPs alone (Fig. 1). Three patients with present SEPs and six patients with indeterminable SEPs could be predicted to have poor neurological outcome using absent VEPs. Although one patient with absent VEPs during TTM had a good neurological outcome, none of the patients with a good neurological outcome had absent VEPs after rewarming (Table 3). Latency and amplitude of SEPs and VEPs. The amplitudes and latencies were calculated only from cases having N20 waves in the SEPs or P100 waves in the VEPs both during TTM and after rewarming. Mean values of latency and amplitudes of SEP and VEP components are presented in Supplemental Table 2. In the SEPs, the latencies to N13 responses and to early cortical N20 responses were significantly longer during TTM than after rewarming while cervicocortical conduction time and N20-P23 amplitude were not significantly different between both periods. With regard to the VEPs, the latencies to N75 responses, to P100 responses, and to N145 responses were not significantly different between during TTM and after rewarming. The left-sided N75-P100 amplitude was significantly higher after rewarming compared to during TTM, but the right-sided N75-P100 amplitude was not. When each component of SEPs was compared between the good and the poor outcome during TTM and after rewarming, the N20 peak-to-peak amplitude was significantly higher in the good outcome group than the poor outcome group, although the cut-off value for outcome could not be determined (p=0.018 during TTM and
p=0.012 after rewarming) (Table 4) (Fig. 2). However, there was no significant difference in all components of VEPs in both periods between the two outcome groups.
Discussion The results of our study indicate that in post-cardiac arrest patients treated with TTM, absent SEPs could predict poor neurological outcome during TTM and after rewarming. Absent VEPs may predict poor neurological outcome in both periods and may increase the predictive power of SEPs. To the best of our knowledge, this is the first report using VEPs for outcome prediction in adult post-cardiac arrest patients, although VEPs have been commonly used to predict neurological outcome and visual prognosis in perinatal asphyxia.2-6 A VEP has several positive and negative peaks observed at various times after presentation of the stimuli. Of these, P100 is relatively stable and the most useful clinically. The P100 is thought to arise 100 msec after the stimuli in the striated and parastriated visual cortex.7 Hildebrandt et al. reported that in 8 patients who recovered from vegetative state and 13 who remained in vegetative state, brain perfusion and VEP reactivity in the occipital and parietal areas were associated with recovery from hypoxic vegetative state, showing that the recovered patients had no absent VEP and higher perfusion in the visual cortex and in the precuneus.8 In this study, all 10 patients with absent VEP remained in the vegetative state, and 7 of 10 patients with present VEP recovered from vegetative state. Absent VEPs, therefore, can become a predictor of poor outcome for post-cardiac arrest patients treated with TTM. McCulloch et al. showed that in 25 children with perinatal asphyxia, a strong association was found between normal, abnormal, or absent VEP in the early postnatal period and long-term visual outcome; only one of six patients with an absent VEP recovered normal vision, and blindness and visual impairment were always associated with abnormal neurologic outcomes.5 Taylor et al. described that in the 57 asphyxiated term infants who had SEPs and VEPs recorded during the first three days of life, the first week and at follow-up visits, normal SEPs virtually guaranteed normal outcome and abnormal
VEPs guaranteed abnormal outcome.6 Both together had a higher predictive power than either alone. The combination of VEPs and SEPs yields a powerful means of prognostication for full-term asphyxiated infants. In our study, one of 26 every patients who had bilaterally absent P100 on VEPs during TTM recovered with a good outcome after rewarming, resulting in a 96% specificity and 4% false positive rate (FPR) for poor outcome. Twenty three patients with absent VEPs after rewarming did not regain consciousness. For effects of low temperature on VEPs, Russ et al. reported progressive latency prolongation and amplitude reduction of VEP with hypothermia, leading to a complete loss of waves at 25-27℃.9 Burrows et al. reported that in neonates and infants undergoing profound hypothermic circulatory arrest, VEPs disappeared at a nasopharyngeal temperature of 18.9±2.8℃ and reappeared at a nasopharyngeal temperature of 32.8±1.4℃. With faster cooling, VEPs disappeared at a higher temperature than with slower cooling.10 VEPs, therefore, may disappear at a core temperature of 33℃ if the patient initially has VEPs with a low amplitude of P100. Therefore, we suggest that an absence of VEPs during TTM should not be used for predicting a poor outcome and subsequent withdrawal of a life-sustaining treatment but instead be used as a supportive predictor of SEPs, which would call for further VEP recordings after rewarming. In TTM-treated patient, a bilaterally absent SEPs accurately predicts poor neurological outcome both during TTM (FPR≤2) and after rewarming (FPR≤1)11-13 although two cases of false positive prediction have been reported at the 48-72 hours after ROSC.14,15 Our results showed all patients with absent SEPs during TTM (n=30) and after rewarming (n=32) had poor neurological outcome, indicating 0% FPR, which is in line with a few studies.16,17 In a prospective study by Grippo et al., 60 patients had SEPs during hypothermia and after rewarming and none of the patients with absent SEPs, including 24 patients during hypothermia and 20 patients after rewarming, recovered consciousness (GOS 1-2).17 They insisted that absent SEPs during hypothermia retain a prognostic value for poor neurological outcome, as in normothermic patients. Bowes et al. noted that 40 of 43 patients with absent SEPs during hypothermia and 42 patients with absent SEPs after
rewarming had poor neurological outcome (GOS 1-3).18 When the 3 patients with absent SEPs during hypothermia who recovered to a good outcome were reevaluated by 2 experienced neurophysiologists, the results were undeterminable because of too much noise. When the recordings are obtained under uncertain technical conditions, SEPs should not be classified as absent, and further recordings are essential.17,18 Considering our results of recording SEPs, bilaterally absent N20 peaks remained a strong noninvasive predictor for poor neurologic outcome during TTM as well as after rewarming. Our study included 13 indeterminable SEPs (5 during TTM and 8 after rewarming), which were excluded from outcome prediction due to not having cervical waves. In these cases, VEPs performed together with SEPs could contribute prognostic information about the patients. With regard to the location of cortical injury in hypoxic ischaemic encephalopathy, the Rolandic, parietal, and occipital cortices are more frequently injured than the frontal and temporal cortices.19,20 Considering that SEPs and VEPs reflect the functions of the primary sensory cortex of the parietal lobe and the visual cortex of occipital lobe, respectively, both absent SEPs and absent VEPs can increase the accuracy of poor outcome prediction with FPRs close to 0%. Interestingly, in 3 patients, cortical responses of SEPs were absent during TTM but recovered after rewarming, and none of them recovered consciousness. Recently, a few studies reported that there was no reappearance of cortical response on SEPs after rewarming with initially absent SEPs during TTM.17,21 The results of our study indicate that irrespective of the reappearance of cortical responses on SEPs after rewarming, absent SEPs during TTM may retain prognostic power for poor neurological outcome, with need of further confirmation in larger study. Latencies of different SEP components increased during TTM but the amplitudes of cortical SEPs was not significantly different between both periods, supporting prior studies .16,17 Reduced SEP amplitude is associated with a poor outcome.22,23 Endisch et al. demonstrated that the lowest amplitude of SEP in a survivor with a good outcome (CPC 1-3) was 0.62 µV22 and Logi et al. noted that none of the patients with an SEP amplitude of < 1.2 µV recovered consciousness.23 In contrast, our study showed that the lowest amplitudes in a patient with a good outcome (CPC 1-2) were
0.726 µV during TTM and 0.465 µV after rewarming. Therefore, we think that the cut-off of SEP amplitude cannot be determined for predicting poor outcome because the recordings are performed in different conditions between studies. There are some limitations in this study. First, it is a retrospective study and although our patients were treated in accordance with the clinical pathways and the protocol for post-cardiac arrest patients of our hospital, there may be some selection bias because some patients did not have SEPs and VEPs. Second, in Korea, by law, we do not withdraw a life-sustaining treatment, and the treating physician was not blinded to
the results of
the SEP and VEP recordings, which could influence patient outcomes. Third, only 34 of 115 patients in this study had SEPs and VEPs together in both periods, so we had difficulty identifying the relationship of changes in SEPs and VEPs during TTM and after rewarming to neurological outcome. Therefore, a further prospective study that includes a larger number of patients is necessary. In conclusion, in post-cardiac arrest patients treated with TTM, absent SEPs could predict poor neurological outcome during TTM as well as after rewarming. Absent VEPs may predict poor neurological outcome in both periods and VEPs may provide additional prognostic value in outcome prediction.
References 1.
Wijdicks EF, Hijdra A, Young GB, Bassetti CL, Wiebe S; Quality Standards Subcommittee of the American Academy of Neurology. Practice parameter: prediction of outcome in comatose survivors after cardiopulmonary resuscitation (an evidence-based review): report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology 2006;67:203-10.
2.
van Laerhoven H, de Haan TR, Offringa M, Post B, van der Lee JH. Prognostic tests in term neonates with hypoxic-ischemic encephalopathy: a systematic review. Pediatrics 2013;131:88-98.
3.
Whyte HE. Visual-evoked potentials in neonates following asphyxia. Clin Perinatol 1993;20:451-61.
4.
Muttitt SC, Taylor MJ, Kobayashi JS, MacMillan L, Whyte HE. Serial visual evoked potentials and outcome in term birth asphyxia. Pediatr Neurol 1991;7:86-90.
5.
McCulloch DL, Taylor MJ, Whyte HE. Visual evoked potentials and visual prognosis following perinatal asphyxia. Arch Ophthalmol 199;109:229-33.
6.
Taylor MJ, Murphy WJ, Whyte HE. Prognostic reliability of somatosensory and visual evoked potentials of asphyxiated term infants. Dev Med Child Neurol 1992;34:507-15.
7.
Chi OZ, Subramoni J, Jasaitis D. Visual evoked potentials during etomidate administration in humans. Can J Anaesth 1990;37:452-6.
8.
Hildebrandt H, Happe S, Deutschmann A, Basar-Eroglu C, Eling P, Brunhöber J. Brain perfusion and VEP reactivity in occipital and parietal areas are associated to recovery from hypoxic vegetative state. J Neurol Sci 2007;260:150-8.
9.
Russ W, Kling D, Loesevitz A, Hempelmann G. Effect of hypothermia on visual evoked potentials (VEP) in humans. Anesthesiology 1984;61:207-10.
10.
Burrows FA, Hillier SC, McLeod ME, Iron KS, Taylor MJ. Anterior fontanel
pressure and visual evoked potentials in neonates and infants undergoing profound hypothermic circulatory arrest. Anesthesiology 1990;73:632-6. 11.
Kamps MJ, Horn J, Oddo M, Fugate JE, Storm C, Cronberg T, Wijman CA, Wu O, Binnekade JM, Hoedemaekers CW. Prognostication of neurologic outcome in cardiac arrest patients after mild therapeutic hypothermia: a meta-analysis of the current literature. Intensive Care Med 2013;39:1671 -82.
12.
Sandroni C, Cariou A, Cavallaro F, Cronberg T, Friberg H, Hoedemaekers C, Horn J, Nolan JP, Rossetti AO, Soar J. Prognostication in comatose survivors of cardiac arrest: an advisory statement from the European Resuscitation Council and the European Society of Intensive Care Medicine. Resuscitation 2014;85:1779-89.
13.
Nolan JP, Soar J, Cariou A, Cronberg T, Moulaert VR, Deakin CD, Bottiger BW, Friberg H, Sunde K, Sandroni C. European Resuscitation Council and European Society of Intensive Care Medicine Guidelines for Post-resuscitation Care 2015: Section 5 of the European Resuscitation Council Guidelines for Resuscitation 2015. Resuscitation 2015;95:202-22.
14.
Leithner C, Ploner CJ, Hasper D, Storm C. Does hypothermia influence the predictive value of bilateral absent N20 after cardiac arrest? Neurology 2010;74:965-9.
15.
Arch AE, Chiappa K, Greer DM. False positive absent somatosensory evoked potentials
in
cardiac
arrest
with
therapeutic hypothermia. Resuscitation
2014;85:e97-8. 16.
Tiainen M, Kovala TT, Takkunen OS, Roine RO. Somatosensory and brainstem auditory evoked potentials in cardiac arrest patients treated with hypothermia. Crit Care Med 2005;33:1736-40.
17.
Grippo A, Carrai R, Fossi S, Cossu C, Mazzeschi E, Peris A, Bonizzoli M, Ciapetti M, Gensini GF, Pinto F, Amantini A. Absent SEP during therapeutic hypothermia did not reappear after re-warming in comatose patients following cardiac arrest. Minerva Anestesiol 2013;79:360-9.
18.
Bouwes A, Binnekade JM, Kuiper MA, Bosch FH, Zandstra DF, Toornvliet AC, Biemond HS, Kors BM, Koelman JH, Verbeek MM, Weinstein HC, Hijdra A, Horn J. Prognosis of coma after therapeutic hypothermia: a prospective cohort study. Ann Neurol 2012;71:206-12.
19.
Choi SP, Park KN, Park HK, Kim JY, Youn CS, Ahn KJ, Yim HW. Diffusion-weighted magnetic resonance imaging for predicting the clinical outcome of comatose survivors after cardiac arrest: a cohort study. Crit Care 2010;14:R17.
20.
Järnum H, Knutsson L, Rundgren M, Siemund R, Englund E, Friberg H, Larsson EM. Diffusion and perfusion MRI of the brain in comatose patients treated with mild
hypothermia
after
cardiac
arrest:
a
prospective
observational
study. Resuscitation 2009;80:425–430. 21.
Cloostermans MC, van Meulen FB, Eertman CJ, Hom HW, van Putten MJ. Continuous electroencephalography monitoring for early prediction of neurological outcome in postanoxic patients after cardiac arrest: a prospective cohort study. Crit Care Med 2012;40:2867-75.
22.
Endisch C, Storm C, Ploner CJ, Leithner C. Amplitudes of SSEP and outcome in cardiac arrest survivors: A prospective cohort study. Neurology 2015;85:1752-60.
23.
Logi F, Fischer C, Murri L, Mauguière F. The prognostic value of evoked responses from primary somatosensory and auditory cortex in comatose patients. Clin Neurophysiol 2003;114:1615-27.
Fig. 1. Receiver operating characteristic curve for the prediction of neurological outcome according to SEPs, VEPs, and both during TTM (A) and after rewarming (B)
Fig. 2. N20 peak-to-peak amplitude (µV) of good outcome group (◇) and poor outcome group (◆) in the SEP during TTM (A) and after rewarming (B) in the same patients
Table 1. Patients demographics and clinical characteristics. Total patients
Good neurological
Poor neurological
(n=115)
outcome (n=31)
outcome (n=84)
Male
77 (67.0%)
23 (74.2%)
54 (64.3%)
0.316
Age
54.7 ± 16.6
50.2 ± 16.3
56.4 ± 16.5
0.077
Witnessed arrest
70 (60.9%)
25 (80.6%)
45 (53.6%)
0.008
Bystander CPR
41 (35.7%)
15 (48.4%)
26 (31.0%)
0.083
Shockable rhythm
17 (14.8%)
11 (35.5%)
6 (7.1%)
<0.001
48 (39%)
25 (80.6%)
23 (27.3%)
<0.001
109 (94.8%)
29 (93.5%)
80 (95.2%)
0.718
3 (3, 3)
3 (3, 5)
3 (3, 3)
<0.001
Absent PLR after ROSC
63 (54.7%)
8 (25.8%)
55 (65.5%)
<0.001
Absent CR after ROSC
104 (90.4%)
23 (74.2%)
81 (96.4%)
0.001
Time from ROSC to EP during TTM (n=85) (h)
21.0 ± 6.7
19.9 ± 7.3
21.5 ± 6.4
0.299
Time from ROSC to EP after rewarming (n=90) (h)
70.2 ± 20.5
74.3 ± 23.1
68.3 ± 19.2
0.204
Body temperature when EP during TTM (n=85) (℃)
33.0 ± 0.6
33.3 ± 0.5
32.9 ± 0.7
0.027
Body temperature when EP after rewarming (n=90) (℃)
36.8 ± 0.7
36.9 ± 0.5
36.7 ± 0.8
0.247
CPC 1
27 (23.5%)
27 (87.1%)
0 (0.0%)
CPC 2
4 (3.5%)
4 (12.9%)
0 (0.0%)
CPC 3
1 (0.9%)
0 (0.0%)
1 (1.2%)
CPC 4
35 (30.4%)
0 (0.0%)
35 (41.7%)
CPC 5
48 (41.7%)
0 (0.0%)
48 (57.1%)
Cardiac cause OHCA GCS after ROSC
Neurological outcome at discharge
p
Values are presented as mean ± SD, median (IQR), and frequency (%). CPR: cardiopulmonary resuscitation, OHCA: out-of-hospital cardiac arrest, GCS: Glasgow Coma Scale, PLR: pupillary light reflex, CR: corneal response, ROSC: return of spontaneous circulation, EP: evoked potential, TTM: targeted temperature management.
Table 2. Statistical evaluation of somatosensory and visual evoked potentials for the prediction of poor neurological outcome in 115 patients
Modality
n†
Good/poor
Sensitivity
Specificity
PPV
NPV
SEP
80
28/52
58
100
100
54
(43-71)
(85-100)
(86-100)
(41-70)
44
96
96
46
(31-58)
(80-100)
(78-100)
(33-59)
71
96
97
64
(57-82)
(80-100)
(85-100)
(48-78)
67
96
97
59
(53-78)
(80-100)
(85-100)
(43-73)
59
100
100
55
(45-72)
(87-100)
(89-100)
(40-69)
47
100
100
38
(33-62)
(79-100)
(85-100)
(24-54)
63
100
100
52
(47-77)
(76-100)
(89-100)
(33-69)
During TTM (n=85)
VEP
SEP+VEP SEP+VEP††
85
80
85
28/57
28/52
28/57
After rewarming (n=90)
SEP
VEP
SEP+VEP
81
65
57
27/54
16/49
16/41
SEP+VEP††
65
16/49
65
100
100
48
(50-78)
(76-100)
(87-100)
(31-66)
95% Confidence interval in parentheses, †: Number of patients with available data in each subgroup, ††: When SSEP is indeterminable, neurological outcome is predicted by VEP. PPV: positive predictive value, NPV: negative predictive value, FPR: false positive rate, SEP: somatosensory evoked potentials, VEP: visual evoked potentials, TTM: targeted temperature management
Table 3. Relationship between SEPs and VEPs and neurological outcome
TTM (n=85)
After rewarming (n=90)
SEP
VEP
Outcome (n)
Total (n)
SEP
VEP
Outcome (n)
Total (n)
SEP (Ind)
VEP (-)
PO (1)
1
SEP (Ind)
VEP(-)
PO (6)
6
SEP (Ind)
VEP (+)
PO (3)
4
SEP (Ind)
VEP(+)
PO (2)
2
8
SEP (+)
VEP(-)
PO (3)
3
42
SEP (+)
VEP(+)
PO (14)
31
GO (1) SEP (+)
VEP (-)
PO (7) GO (1)
SEP (+)
VEP (+)
PO (15) GO (27)
GO (17)
SEP (-)
VEP (-)
PO (17)
17
SEP (-)
VEP(-)
PO (14)
14
SEP (-)
VEP (+)
PO (13)
13
SEP (-)
VEP(+)
PO (9)
9
SEP (Ind)
Unchecked
PO (1)
1
SEP (+)
Unchecked
PO (4)
15
GO (11) SEP (-)
Unchecked
PO (9)
9
SEP: somatosensory evoked potentials, VEP: visual evoked potentials, Ind: indeterminable, (+): presence, (-): absence PO: poor neurological outcome, GO: good neurological outcome
Table 4. The comparisons of each component between both outcome groups in the same patients who have N20 wave in the SEP (n=30)
TTM
After rewarming
N13 lat
N20 lat
CCT
N20-P23
N13
(ms)
(ms)
(ms)
Amp (µV)
(ms)
Good
14.71 ±
22.10 ±
7.25 ±
(n=22)
1.12
1.48
1.16
Poor
14.61 ±
22.78 ±
7.98 ±
(n=8)
1.18
2.85
2.16
p
0.909
0.982
0.597
2.10 ± 1.13
1.08 ± 0.91 0.018
N20 lat
CCT
N20-P23
(ms)
(ms)
Amp (µV)
19.14 ±
5.71 ±
1.11
1.27
21.98 ±
8.93 ±
1.66
4.35
5.12
0.534
0.170
0.298
13.25
lat
±
0.99 13.03
CCT: cervicocortical conduction time, lat: latency, Amp: amplitude
±
2.48 ± 0.96
1.33 ± 1.10 0.012