If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Address for reprints: Ibrahim Sultan, MD, Division of Cardiac Surgery, Department of Cardiothoracic Surgery, UPMC Center for Thoracic Aortic Disease, University of Pittsburgh, Heart and Vascular Institute, University of Pittsburgh Medical Center, 5200 Centre Ave, Suite 715, Pittsburgh, PA 15232.
Division of Cardiac Surgery, Department of Cardiothoracic Surgery, University of Pittsburgh, Pittsburgh, PaHeart and Vascular Institute, University of Pittsburgh Medical Center, Pittsburgh, Pa
Division of Cardiac Surgery, Department of Cardiothoracic Surgery, University of Pittsburgh, Pittsburgh, PaHeart and Vascular Institute, University of Pittsburgh Medical Center, Pittsburgh, Pa
To evaluate the ability of intraoperative neurophysiologic monitoring (IONM) during aortic arch reconstruction with hypothermic circulatory arrest (HCA) to predict early (<48 hours) adverse neurologic events (ANE; stroke or transient ischemic attack) and operative mortality.
Methods
This was an observational study of aortic arch surgeries requiring HCA from 2010 to 2018. Patients were monitored with electroencephalogram (EEG) and somatosensory evoked potentials (SSEP). Baseline characteristics and postoperative outcomes were compared according to presence or absence of IONM changes, which were defined as any acute variation in SSEP or EEG, compared with baseline. Multivariable logistic regression analysis was used to assess the association of IONM changes with operative mortality and early ANE.
Results
A total of 563 patients underwent aortic arch reconstruction with HCA and IONM. Of these, 119 (21.1%) patients had an IONM change, whereas 444 (78.9%) did not. Patients with IONM changes had increased operative mortality (22.7% vs 4.3%) and increased early ANE (10.9% vs 2.9%). In multivariable analysis, SSEP changes were correlated with early ANE (odds ratio [OR], 4.68; 95% confidence interval [CI], 1.51-14.56; P = .008), whereas EEG changes were not (P = .532). Permanent SSEP changes were correlated with early ANE (OR, 4.56; 95% CI, 1.51-13.77; P = .007), whereas temperature-related SSEP changes were not (P = .997). Finally, any IONM change (either SSEP or EEG) was correlated with operative mortality (OR, 5.82; 95% CI, 2.72-12.49; P < .001).
Conclusions
Abnormal IONM events during aortic arch reconstruction with HCA portend worse neurologic outcomes and operative mortality and have a negative predictive value of 97.1%. SSEP might be more sensitive than EEG for predicting early ANE, especially when SSEP changes are permanent.
Intraoperative neurophysiologic monitoring (IONM) can be used to guide neuroprotective decision-making during aortic arch surgery. IONM changes were significantly associated with operative mortality and postoperative adverse neurologic events, with permanent SSEP changes in cortical amplitude or prolonged latency being the most predictive type of IONM change. These findings suggest that IONM should be routinely used during aortic arch surgery.
See Commentaries on pages 1982 and 1984.
Adverse neurologic events (ANE) are potentially catastrophic complications of open aortic arch surgery because they are associated with in-hospital mortality and diminished quality of life.
ANE might be embolic phenomena, caused by debris from atherosclerotic brachiocephalic vessels or arising from surgical instrumentation. ANE might also be related to hypoperfusion, caused by malperfusion from static or dynamic obstructive dissection flaps impacting the carotids or caused by inadequate cerebral protection during circulatory arrest. Hypothermia constitutes the basis of neuroprotection during arch reconstruction by reducing cerebral metabolism. However, the risk of ANE and operative mortality increases as the duration of hypothermic circulatory arrest (HCA) increases because hypothermia incompletely suppresses cerebral metabolic demand.
Therefore, cerebral adjuncts, such as antegrade cerebral perfusion (ACP) and retrograde cerebral perfusion (RCP), have been used in conjunction with HCA to improve neuroprotection.
Similar cerebral protective effectiveness of antegrade and retrograde cerebral perfusion combined with deep hypothermia circulatory arrest in aortic arch surgery: a meta-analysis and systematic review of 5060 patients.
Although these adjunctive strategies might mitigate the risk of ANE compared with straight HCA, the risk of ANE remains, highlighting the continued opportunity for improving neurologic outcomes of open aortic arch surgery.
This has promoted the American Heart Association in a scientific statement to suggest the use of intraoperative neurophysiologic monitoring (IONM) when available to detect ANE during cardiac and thoracic surgery.
Considerations for reduction of risk of perioperative stroke in adult patients undergoing cardiac and thoracic aortic operations: a scientific statement from the American Heart Association.
IONM with electroencephalography (EEG) and somatosensory evoked potentials (SSEP) has been used to help manage cerebral perfusion strategies and guide decision-making during aortic surgery. IONM (specifically SSEP and/or motor evoked potentials) has been evaluated and widely accepted for use during open and endovascular repair of the descending aorta
Despite the evidence that IONM can be used to detect and treat large vessel occlusion strokes after cardiac surgery, its utility in a large data set has not been explored.
At the University of Pittsburgh Medical Center, SSEP and EEG are used during all surgeries involving the aortic arch, including all type A aortic dissections.
In this study we sought to evaluate the ability of IONM changes during aortic arch reconstruction with HCA to predict early (<48 hours) ANE and operative mortality. We expect the results of our study will provide scientific evidence for recommendations made by the American Heart Association to detect stroke using IONM and improve its utilization from “when available” to clinical standard of care.
Methods
Patient Population and Study Design
This was an observational study, using a prospectively maintained institutional database of all consecutive aortic surgeries performed at the University of Pittsburgh Medical Center between July 2, 2010 and June 30, 2018. Definitions and terminology were consistent with the Society of Thoracic Surgeons database. All patients who underwent aortic arch surgery requiring HCA were included for analysis, whereas patients who underwent open thoracoabdominal aortic repair or thoracic endovascular aortic repair were excluded. Patients were included if they underwent HCA with dual modality neurophysiologic monitoring (combined SSEP and EEG). Patients were subsequently dichotomized into 2 groups, according to the presence or absence of significant IONM changes (defined in “Intraoperative Neuromonitoring”). Baseline characteristics and postoperative outcomes of patients who had significant IONM changes were compared with those who did not experience any significant IONM changes. The primary aim of this study was to evaluate the ability of IONM changes to predict early ANE and operative mortality. Operative mortality included 30-day and/or in-hospital mortality, whereas ANE included stroke and transient ischemic attack. ANE were further divided into early (<48 hours after surgery) and late (>48 hours after surgery), with early ANE being the primary outcome of interest. This study was approved by the institutional review board of the University of Pittsburgh on April 17, 2019 (STUDY18120143), with written consent being waived.
Intraoperative Neuromonitoring
To generate SSEPs, subdermal needle electrode pairs were used to independently stimulate the left and right median or ulnar nerves at the wrist, and the left and right tibial nerves. Constant current stimulation was used, at intensities sufficient to evoke a consistent and supramaximal response. To record the thalamocortical and cortical potentials, or N20-P30 SSEPs, scalp electrodes were placed at P4/Fz and P3/Fz (according to the International 10-20 System
). The dorsal column nucleus, or brainstem potential, was recorded using an electrode, which was referenced to Fz and localized on the mastoid. The peripheral potential generated in the brachial plexus was recorded using electrodes placed at the bilateral erbs point and referenced to each other. To compute the averages in each recording, at least 128 trials were used. Significant SSEP changes were defined as a persistent and consistent ≥50% decrease in the cortical amplitude and/or ≥10% prolongation of latency from baseline values in ≥2 averaged trials.
An example of a normal SSEP is shown in Figure 1, and an example of an abnormal SSEP is shown in Figure 2. EEG was recorded using electrodes placed on the scalp (according to the International 10-20 System
). EEG was recorded using 8 channels: F3-P3, P3-O1, F3-T3, T3-O1, F4-P4, P4-O2, F4-T4, and T4-O2. To be considered a significant EEG change, the EEG recording must display a ≥50% decrease in the amplitude of the fast frequency or a ≥50% increase in theta or delta activity.
Permanent SSEP or EEG change is defined as any significant change that did not recover to nonsignificant change or baseline by the end of the procedure, whereas SSEP and EEG changes that returned to baseline by the end of the procedure were deemed reversible. Temperature-related SSEP or EEG change is defined as any significant change that occurred within 10 minutes of rewarming or cooling, or any significant change recorded in the intraoperative log as being directly related to rewarming or cooling.
Figure 1This flowsheet shows left and right upper extremity somatosensory-evoked potential (SSEP) waveforms. To generate upper extremity SSEPs, subdermal needle electrode pairs were used to independently stimulate the left and right median or ulnar nerves at the wrist. Scalp electrodes were placed at P4/Fz and P3/Fz (according to the International 10-20 system). Constant current stimulation was employed, at intensities sufficient to evoke a consistent and supramaximal response. The Fz-P4 <US> and Fz-Cv2 <US> channels represent the cortical and subcortical (brainstem) responses from the left ulnar nerve. The Fz-P3 <UD> and Fz-Cv2 <UD> channels represent the cortical and subcortical (brainstem) responses from the right ulnar nerve. These waveforms demonstrate no significant changes in somatosensory response throughout the procedure.
Figure 2This flow sheet shows left and right lower extremity somatosensory evoked potential (SSEP) waveforms. To generate lower extremity SSEPs, subdermal needle electrode pairs were used to independently stimulate the left and right tibial nerves. Scalp electrodes were placed at P4/Fz and P3/Fz (according to the International 10-20 System). Constant current stimulation was used, at intensities sufficient to evoke a consistent and supramaximal response. The Fz-Pz and P4-P3 channels represent the cortical responses from the left and right tibial nerves. Significant SSEP changes were defined as a persistent and consistent ≥50% decrease in the cortical amplitude and/or ≥10% prolongation of latency from baseline values in ≥2 averaged trials. The red box highlights a significant decrease in amplitude of the response in the left tibia, which later returns to baseline.
During the study's time frame at this institution, hemiarch replacement with HCA and RCP was the preferred operation for all patients with type A aortic dissection or aneurysmal dilation of the ascending aorta, which has been previously described.
However, total arch replacement was performed with HCA and ACP if there was: (1) an intimal tear involving the aortic arch, (2) aneurysm of the aortic arch, (3) circumferential dissection of the aortic arch, or (4) carotid dissection causing cerebral malperfusion with or without carotid thrombosis.
Thus, all hemiarch replacement occurred with RCP and all total arch replacement occurred with ACP. Central aortic cannulation was performed via the modified Seldinger technique and with transesophageal echocardiographic guidance, and was the preferred arterial cannulation strategy for all patients, except in the presence of complete circumferential arch dissection. To avoid anesthetic-induced IONM changes, inhalational anesthetics are avoided as much as possible and a total intravenous anesthetic paradigm is used. Whenever inhalational anesthesia is needed, the upper limit of 0.5 end tidal isoflurane is used as a marker.
Routine neurocerebral monitoring was used during all aortic arch surgery via SSEP and EEG, which were monitored by a dedicated and certified neurophysiologist. The attending surgeon interpreted the abnormal IONM events in concert with the neurophysiologist and decided the appropriate intraoperative intervention, in conjunction with the neurophysiology, perfusion, and anesthesia teams (see Figure 3). Although the data might be confusing when presented directly to the surgeon, the most important aspect in patients with acute aortic dissection is to differentiate between peripheral malperfusion and a cortical event. This can generally be done using IONM data. When this is determined, the second aspect is to determine temporary versus permanent IONM change and determine if any interventions (raising perfusion pressure, increasing pulsatility) might help correct IONM changes. If these IONM abnormalities continue to be permanent, the final step is to image the patient as soon as the patient leaves the operating room before arrival to the cardiac service intensive care unit if safe to do so.
Figure 3Interventions performed to reverse intraoperative neuromonitoring abnormalities. IONM, Intraoperative neurophysiologic monitoring.
Primary stratification was between the group of patients with IONM changes (either SSEP changes or EEG changes) and the group of patients without IONM changes. Continuous variables are presented as mean ± standard deviation for normally distributed data, or median and interquartile range (IQR) for non-normally distributed data. Categorical variables are summarized using frequency and percentage. Student t test was used to compare normally distributed continuous variables between groups, and the nonparametric Mann-Whitney U test was used for non-normally distributed continuous variables. The χ2 or Fisher exact test was used to compare categorical variables between groups, as appropriate. All tests were 2-sided with an α level of 0.05 considered to indicate statistical significance.
Differences between the group with IONM changes and the group without IONM changes were described for baseline demographic, clinical, and operative variables (Table 1). Patients with IONM changes were compared against patients without IONM changes for univariate differences in short-term postoperative outcomes (Table 2). Next, 2 multivariable logistic regression models were built to identify variables associated with early postoperative ANE (<48 hours after aortic arch surgery; Table 3). In the first model (Table 3), any significant SSEP change and any significant EEG change were included in the model to assess differences between each neuromonitoring modality. Conversely, in the second model (Table 3), any permanent IONM change (SSEP and/or EEG) and any temperature-related IONM change (SSEP and/or EEG) were included in the model to assess differences between each type of IONM event. However, for both models predicting early ANE, age, cerebrovascular disease, type A aortic dissection, redo sternotomy, hemiarch (vs total arch) replacement, concomitant aortic root replacement, surgical status, cardiopulmonary bypass time, circulatory arrest time, and nadir bladder temperature were considered for multivariable adjustment. Backward stepwise estimation was used for both models, with a threshold of inclusion in the model of P < .020, and the IONM/SSEP/EEG related variables being treated as locked terms. Finally, a multivariable logistic regression was built to identify variables associated with operative mortality for patients who received dual modality neuromonitoring (Table 4). Any significant IONM change (either SSEP or EEG) was included in the model for operative mortality, and all variables listed in Table 1 were considered for multivariable adjustment. Again, backward stepwise estimation was used, with a threshold of inclusion of P < .020, and the IONM-related variable was treated as a locked term. Finally, the Hosmer-Lemeshow test statistic was performed to assess goodness-of-fit for all logistic regression models (Tables 3 and 4). All statistical analyses were performed using STATA, version 15.0 (Stata Corp, College Station, Tex).
Table 1Baseline characteristics, of patients with a significant (IONM) event and patients without a significant IONM event
All hemiarch replacement occurred with retrograde cerebral perfusion and all total arch replacement occurred with antegrade cerebral perfusion.
284 (64.0)
47 (39.5)
<.001
Aortic root replacement
95 (21.4)
17 (14.3)
.084
Surgical status
.036
Elective
184 (41.4)
34 (28.6)
Urgent
65 (14.6)
20 (16.8)
Emergent/salvage
195 (43.9)
65 (54.6)
Cardiopulmonary bypass time, minutes
190 [160-230]
229 [190-270]
<.001
Circulatory arrest time, minutes
23 [16-32]
30 [22-47]
<.001
Cerebral perfusion time, minutes
22 [15-30]
31 [21-45]
<.001
Nadir bladder temperature, °C
21.1 ± 2.7
21.1 ± 3.2
.978
Aortic (vs peripheral) cannulation
429 (96.6)
109 (92.4)
.042
Data are presented as n (%), mean ± standard deviation, or median [interquartile range], except where otherwise noted. IONM, Intraoperative neurophysiologic monitoring.
∗ All hemiarch replacement occurred with retrograde cerebral perfusion and all total arch replacement occurred with antegrade cerebral perfusion.
Table 2Perioperative outcomes of patients with a significant (IONM) event and patients without a significant IONM event
Variable
No significant IONM event (n = 444)
Significant IONM event (n = 119)
P value
Operative mortality (STS definition)
19 (4.3)
27 (22.7)
<.001
ICU length of stay, hours
45 [25-100]
84 [33-158]
<.001
Stroke/TIA
18 (4.1)
13 (10.9)
.004
Early, (<48 h after surgery)
13 (2.9)
13 (10.9)
<.001
Late, (>48 h after surgery)
5 (1.1)
0 (0.0)
.245
Prolonged ventilation, >24 h
92 (20.7)
43 (36.1)
<.001
New dialysis requirement
34 (7.7)
24 (20.2)
<.001
Reexploration for bleeding
24 (5.4)
7 (5.9)
.839
Postoperative blood product transfusion
192 (43.2)
65 (54.6)
.027
Data are presented as n (%) or median [interquartile range], except where otherwise noted. IONM, Intraoperative neurophysiologic monitoring; STS, Society of Thoracic Surgeons; ICU, intensive care unit; TIA, transient ischemic attack.
Table 4Multivariable logistic regression model for operative mortality (STS definition for in-hospital and/or 30-day mortality), for the effect of any IONM change
The overall model has a Hosmer-Lemeshow goodness-of-fit χ2 statistic of 5.71 (P = .680). CI, Confidence interval; IONM, intraoperative neurophysiologic monitoring.
Baseline Demographic, Clinical, and Operative Variables
A total of 563 patients who underwent aortic arch surgery with HCA and IONM were identified. All patients had dual modality neuromonitoring with SSEP and EEG. One hundred nineteen (21.1%) of these patients had either SSEP changes and/or EEG changes. These 119 patients constituted the group with IONM changes, whereas the remaining 444 constituted the group without IONM changes. Table 1 includes the baseline characteristics for the entire cohort (n = 563), analyzed according to the presence or absence of IONM changes. Patients with IONM changes had greater body mass index (32.6 ± 23.0 vs 29.1 ± 5.9) but were otherwise similar across demographic and clinical variables. Of note, patients with IONM changes were as likely to have had aortic dissection etiology, as they were to have had aneurysmal disease (P = .791). Patients with IONM changes were more likely to have had a redo sternotomy (29.4% vs 14.2%), were less likely to have had hemiarch (vs total arch) replacement (39.5% vs 64.0%) and were more likely to have had emergent/salvage surgery (54.6% vs 43.9%).
Perioperative Outcomes
Intraoperatively, 102 patients (18.1%) had a significant SSEP change, whereas 62 patients (11.0%) had a significant EEG change. There were 66 (11.7%) permanent IONM changes (SSEP and/or EEG), and there were 20 (3.6%) temperature-related IONM changes (SSEP and/or EEG). Of the 102 patients with SSEP changes, 53 (52.0%) had permanent changes, whereas 49 (48%) had reversible changes. Sixteen (15.7%) patients had temperature-related SSEP changes (either during rewarming or cooling), 9 of which were permanent changes and 7 of which were reversible. Of the 62 patients with EEG changes, 34 (54.8%) had permanent changes and 9 (14.5%) had temperature-related changes (either during rewarming or cooling). Finally, patients with IONM changes had significantly longer cardiopulmonary bypass time (229 [IQR, 190-270] vs 190 [IQR, 160-230] minutes) and longer circulatory arrest time (30 [IQR, 22-47] vs 23 [IQR, 16-32] minutes), whereas nadir bladder temperature was similar across each group (P = .978).
Postoperatively, patients with IONM changes were more likely to have had operative mortality (22.7% vs 4.3%; Figure 4 and Video 1) and were more likely to have had ANE (10.9% vs 4.1%). Considering the timing of postoperative ANE, there were more early (<48 hours) ANE in the group with IONM changes (10.9% vs 2.9%), compared with the group without IONM changes. There was no statistical difference between each group for late ANE (>48 hours; P = .245), with a timing of onset that ranged from postoperative day 4 to day 25. Patients with IONM changes also had longer intensive care unit length of stay (84 [IQR, 33-158] vs 45 [IQR, 25-100] hours), were more likely to have had prolonged mechanical ventilation (36.1% vs 20.7%), were more likely to have had new-onset renal failure requiring hemodialysis (20.2% vs 7.7%), and were more likely to have had postoperative blood product transfusions (54.6% vs 43.2%). However, the incidence of reexploration for excessive bleeding was similar across each group (P = .839). Table E1 shows perioperative outcomes compared across patients without a significant IONM event, patients with a reversible IONM change, and patients with a permanent IONM change, which shows that patients with permanent IONM changes had increased rates of operative mortality, early ANE, prolonged ventilation, and new dialysis requirement.
Figure 4In an observational study of 563 patients who received aortic arch surgery with hypothermic circulatory arrest and intraoperative neurophysiologic monitoring (IONM) from 2010 and 2018, all patients were monitored using electroencephalogram and somatosensory evoked potentials. Baseline characteristics and postoperative outcomes were compared according to presence or absence of IONM changes, which were defined as any acute variation in somatosensory evoked potentials or electroencephalogram, compared with baseline. One hundred nineteen (21.1%) patients had an IONM change, whereas 444 (78.9%) did not. Patients with IONM changes were more likely to have had operative mortality (22.7% vs 4.3%; P < .001) and were more likely to have had early adverse neurologic events, <48 hours after surgery (10.9% vs 2.9%; P < .001). Abnormal IONM events during aortic arch reconstruction might portend worse neurologic outcomes and increased operative mortality.
With 431 patients having had no IONM changes as well as no early ANE and with 444 total cases without IONM changes, the negative predictive value (NPV) of IONM changes is 97.1% for early ANE. Sensitivity, specificity, and positive predictive value (PPV) were not calculated for this cohort. As an observational study, in which the surgeons, perfusionists, and anesthesiologists might have collectively opted to intervene on an IONM change, some patients might have had an intraoperative ischemic insult that was reversed by that very intervention. Thus, some patients might have shifted from a disease-positive state to a disease-negative state within the group with IONM changes, thereby rendering sensitivity, specificity, and PPV inaccurate for this observational cohort. However, in this study, 18 patients (4.1%) developed ANE in the absence of a significant IONM change, of which 13 were early (<48 hours after surgery) and 5 were late (>48 hours after surgery).
Radiologic Imaging and IONM Events
Considering the 13 patients with IONM events who ultimately developed an early ANE, 9 patients had postoperative radiologic findings that perfectly correlated with their intraoperative neuromonitoring changes in terms of laterality: 2 patients had bilateral lesions with global IONM changes, whereas 7 had unilateral lesions with a corresponding IONM change. Conversely, 2 patients had lesions on imaging that partially correlated with their IONM findings: 1 patient had bilateral lesions, but only left-sided IONM changes, whereas another patient had a right-sided lesion, but had global IONM changes. Finally, 2 patients with IONM events and an early ANE had missing radiologic imaging data, and could not therefore be assessed for radiologic correlation with their IONM events. Considering the 13 patients without IONM events who ultimately developed an early ANE, 8 patients had small lesions (<1 cm) and/or subcortical infarcts. However, 5 of these patients had large lesions in large vessel distributions, which were undetected during IONM.
Multivariable Analysis for Operative Mortality and ANE
Table 3 shows 2 multivariable logistic regression models used to assess the relationship between IONM changes and early ANE (<48 hours after surgery). Table E2 shows all variables considered in these models, as well as their univariable logistic associations with early ANE. In multivariable analysis, SSEP changes were associated with early ANE (odds ratio [OR], 4.68, 95% confidence interval [CI], 1.51-14.56; P = .008), whereas EEG changes were not (P = .532; Table 3). In multivariable analysis, permanent SSEP changes were associated with early ANE (OR, 4.56; 95% CI, 1.51-13.77; P = .007), whereas temperature-related SSEP changes were not (P = .997; Table 3). Other risk factors for early postoperative ANE included increasing age and increasing cardiopulmonary time (Table 3). Finally, Table 4 shows a multivariable logistic regression model used to assess the relationship between IONM changes and operative mortality. Table E3 shows all variables considered in this model, as well as their univariable logistic associations with operative mortality. Any significant IONM change was associated with operative mortality (OR, 5.82; 95% CI, 2.72-12.49; P < .001). Please see Video 1 for a discussion of results.
Discussion
IONM has been used to detect ischemia of the central nervous system and, consequently, to guide neuroprotective decision-making during aortic surgery, especially during open and endovascular repair of the descending aorta. However, in the setting of aortic arch reconstruction, there is a paucity of data on the ability of IONM to predict ANE and operative mortality. A few notable findings are evident from this observational analysis of IONM changes during aortic arch reconstruction with HCA. First, IONM changes were associated with operative mortality and early postoperative ANE. Second, SSEP changes might be more associated with early postoperative ANE than EEG changes. Third, permanent SSEP changes correlated with the occurrence of early postoperative ANE, which highlights the continued opportunity to optimize strategies for reversing neurologic insult during aortic arch surgery. Finally, with an NPV of 97.1%, the absence of IONM changes during arch surgery informs the surgeon that there is a low risk of early ANE. Alternatively, IONM changes, particularly permanent SSEP changes, should provoke immediate cross-sectional cerebrovascular imaging (eg, computed tomographic angiography) to determine whether potential endovascular neurointervention might be indicated.
Is two really better than one? Examining the superiority of dual modality neurophysiological monitoring during carotid endarterectomy: a meta-analysis.
Monitoring electrophysiologic function during carotid endarterectomy: a comparison of somatosensory evoked potentials and conventional electroencephalogram.
During repair of the descending aorta, spinal cord ischemia secondary to segmental artery disruption is the most feared neurologic complication. Some data suggest that transcranial motor evoked potentials are more sensitive than SSEPs for detecting spinal cord ischemia, possibly because the motor pathways are more sensitive to malperfusion; however, most data support the use of at least 1 IONM modality during descending aorta repair.
During carotid endarterectomy, however, postoperative stroke secondary to cerebral malperfusion is the most feared neurologic complication. Some data suggest that dual-modality neuromonitoring (SSEP and EEG together) is more sensitive for detecting postoperative stroke than SSEP or EEG alone.
Is two really better than one? Examining the superiority of dual modality neurophysiological monitoring during carotid endarterectomy: a meta-analysis.
Monitoring electrophysiologic function during carotid endarterectomy: a comparison of somatosensory evoked potentials and conventional electroencephalogram.
In concert, each neuromonitoring modality compensates for the other's deficiencies, with EEG being more sensitive for cortical changes, whereas SSEP is more capable of detecting subcortical changes.
Is two really better than one? Examining the superiority of dual modality neurophysiological monitoring during carotid endarterectomy: a meta-analysis.
Monitoring electrophysiologic function during carotid endarterectomy: a comparison of somatosensory evoked potentials and conventional electroencephalogram.
In this study of aortic arch reconstruction, IONM changes were associated with early postoperative ANE. Similarly, Ghincea and colleagues reported that IONM (EEG and SSEP with or without transcranial motor evoked potentials) has high sensitivity and specificity for detecting postoperative stroke after aortic arch surgery.
In this study, all patients with available imaging data had postoperative radiologic findings that correlated with their intraoperative neuromonitoring events, which is reassuring concerning the accuracy of IONM. However, considering the 13 patients without IONM events who ultimately developed an early ANE (2.3% of the overall cohort), 5 of these patients had large lesions in large vessel distributions, which were undetected during IONM. This is a concerning finding that deserves further investigation. Conversely, 8 of these 13 patients had small lesions (<1 cm) and/or subcortical infarcts, for which IONM has reduced sensitivity. Although this is predictable considering the current technology, it also deserves further investigation. Nevertheless, 10.9% of patients with IONM events ultimately had an early ANE, whereas only 2.9% of patients without IONM events developed early ANE, which suggests that IONM changes were highly correlated with early ANE. And, with an NPV of 97.1%, the absence of IONM changes during arch surgery informs the surgeon that there is a low risk of early ANE. Moreover, patients in our study with any IONM change had 5.8 times higher odds of operative mortality, compared with patients without any IONM changes, highlighting the clinical importance of limiting perioperative ANE during aortic arch surgery. Taken together, these findings suggest that IONM should be routinely used during aortic arch surgery.
We acknowledge that in this observational study we did not explicitly compare the effect of IONM versus not using IONM. In contradistinction, Ghincea and colleagues reported that the incidence of postoperative stroke was not statistically different for patients who received IONM, compared with patients who did not receive IONM.
By implication, IONM might not have any demonstrable effect on postoperative stroke. However, the 2 groups in their study were significantly different across numerous baseline variables, suggesting substantial selection bias for the use of IONM in their cohort. Thus, future studies ought to compare IONM with a protocol-driven treatment algorithm for IONM alerts, against no IONM. Meanwhile, we have learned that the use of IONM can inform the surgical team of a potential ANE well before patients can provide an adequate neurologic exam and thus provide earlier indication to obtain diagnostic computed tomographic angiography to reveal a potentially intervenable cerebrovascular occlusion.
SSEP changes might be more correlated with early postoperative ANE than EEG changes. In this study, patients with intraoperative SSEP changes had 4.7 times higher odds of developing early postoperative ANE, compared with patients without SSEP changes, whereas intraoperative EEG changes were not associated with early postoperative ANE. To be sure, EEG was correlated with ANE in univariable analysis, but was not correlated with ANE when adjusted for SSEP changes. This might be because SSEP can detect more subtle cerebral insults, including subcortical changes. Others have argued that SSEP is less sensitive than EEG when the patient is under anesthesia and have promoted EEG as a critically necessary complement to optimizing the efficacy of intraoperative monitoring that includes SSEP.
Is two really better than one? Examining the superiority of dual modality neurophysiological monitoring during carotid endarterectomy: a meta-analysis.
Monitoring electrophysiologic function during carotid endarterectomy: a comparison of somatosensory evoked potentials and conventional electroencephalogram.
Permanent IONM changes were highly associated with early postoperative ANE, whereas temperature-related IONM changes were not associated with early ANE. Brief, temperature-related IONM changes (during rewarming and cooling) might constitute false positive results and therefore might be less concerning than sustained IONM changes. Moreover, patients with permanent IONM changes showed a 4.6 times higher odds of developing early postoperative ANE, compared with patients with temporary IONM changes.
It is likely that the protective association of temporary IONM changes in relation to early postoperative ANE reflects intraoperative interventions that were used to reverse neurologic insults (see Figure 3). Symmetric IONM changes during cooling are expected and should not be responded to with any immediate intervention. Furthermore, EEG and SSEP are isoelectric during the time of HCA; therefore, any neurologic injury during distal reconstruction, as well as any immediate benefits of cerebral perfusion, might not be readily apparent during the time of HCA. However, asymmetric changes while rewarming can be concerning and might be a symptom of malperfusion or neurologic injury. The typical first response is to increase perfusion pressure to mitigate effects of hypoperfusion. At the same time, ultrasound of both carotid arteries is performed by our anesthesia colleagues to ensure good flow in the carotid arteries distally. The brachiocephalic anastomoses are carefully inspected to ensure no anastomotic kinks or narrowing. Finally, when the heart is adequately rested after the cross-clamp is removed, we allow the patient to be pulsatile while maintaining hypertension to help with the same. Patients with carotid artery dissection and thrombus burden might have IONM changes when cardiopulmonary bypass is initiated. In that situation, our technique is to expose the cervical carotid artery, clamp the distal carotid artery, and revascularize the carotid artery in the neck. In general, it should be noted that the neurologic benefit of RCP is derived from flushing gaseous and particulate matter out of the cerebral vasculature. Hence, when hemiarch replacement can be performed, RCP is the preferred method of cerebral perfusion at our institution because it can be used to prevent and reverse embolism in the cerebral vasculature. When performing selective ACP during total arch replacement, prompt initiation of unilateral, then bilateral, ACP often corrects IONM abnormalities as well. Some groups have advocated routine RCP before or after HCA to flush out embolic debris.
For all IONM abnormalities that persist at the end of the procedure, we advocate immediate head and neck computed tomography angiogram with brain perfusion to look for any large vessel occlusion and to determine whether endovascular neurointervention might be indicated. Regardless, the recognition that permanent IONM changes can indicate early ANE highlights the need for determining strategies for reversing neurologic insult during or immediately after aortic arch surgery—the time when one would most likely be able to positively alter the outcome.
Limitation
There are several important limitations to this study. First, this was an observational study in which we did not explicitly compare the effect of using versus not using IONM. Future studies that compare IONM with a protocol-driven treatment algorithm for IONM alerts versus no IONM might provide more impactful data and might further demonstrate the efficacy of IONM-based intraoperative interventions for preventing early ANE. Second, in this study we were unable to determine the sensitivity, specificity, and PPV of IONM for detecting ANE. To accurately determine the diagnostic efficacy of IONM, universal neuroimaging for all patients receiving IONM (with or without changes) would be necessary. Moreover, to accurately assess diagnostic capability, the outcome (ANE) should optimally not be biased by intervention (eg, mean arterial pressure augmentation to reverse IONM changes). Obviously, this would be decidedly unethical for patients who develop IONM changes who might be experiencing cerebral ischemia. Regardless, the strong association of permanent SSEP changes with ANE suggests that SSEPs have some predictive capacity. Third, because of the small number of outcomes (31 ANE and 46 operative mortality), the multivariable analysis was potentially underpowered and therefore limited in the number of risk factors that could reasonably be analyzed in the regression model. Fourth, this series of aortic surgeries was performed by a small group of surgeons in a single, high-volume institution thereby limiting the generalizability of the findings. Finally, the study lacks granular data on what intraoperative interventions (on the basis of IONM changes) were performed (eg, mean arterial pressure augmentation and cerebral perfusion flow rate alterations). As a corollary, there might have been variability among surgeons with respect to how IONM changes were interpreted and action taken, suggesting some level of heterogeneity and selection bias within the cohort.
Conclusions
In this study of the role of IONM during aortic arch reconstruction with HCA, IONM changes were correlated with operative mortality and early postoperative ANE. Along with a NPV of 97.1%, this suggests that IONM should be routinely used during aortic arch surgery. SSEP changes were more associated with early postoperative ANE than EEG changes, and permanent IONM changes were highly correlated with early postoperative ANE. This study highlights the continued need to optimize strategies for mitigating or reversing neurologic insults during aortic arch surgery.
Dr Sultan receives institutional research support from Medtronic and AtriCure and consults for Medtronic Vascular. Dr Kilic consults for Medtronic and Abiomed. These are unrelated to this report. All other authors reported no conflicts of interest.
The Journal policy requires editors and reviewers to disclose conflicts of interest and to decline handling or reviewing manuscripts for which they may have a conflict of interest. The editors and reviewers of this article have no conflicts of interest.
Précis of the outcomes (stroke and mortality) of intraoperative neuromonitoring events during aortic arch reconstruction with hypothermic circulatory arrest. Study design, results, and clinical implications are discussed. Video available at: https://www.jtcvs.org/article/S0022-5223(21)01124-7/fulltext.
Précis of the outcomes (stroke and mortality) of intraoperative neuromonitoring events during aortic arch reconstruction with hypothermic circulatory arrest. Study design, results, and clinical implications are discussed. Video available at: https://www.jtcvs.org/article/S0022-5223(21)01124-7/fulltext.
Appendix E1
Table E1Perioperative outcomes, across patients without a significant IONM, patients with a reversible IONM event, and patients with a permanent IONM event
Data are presented as n (%), except where otherwise noted. IONM, Intraoperative neurophysiologic monitoring; STS, Society of Thoracic Surgeons; TIA, transient ischemic attack.
Similar cerebral protective effectiveness of antegrade and retrograde cerebral perfusion combined with deep hypothermia circulatory arrest in aortic arch surgery: a meta-analysis and systematic review of 5060 patients.
Considerations for reduction of risk of perioperative stroke in adult patients undergoing cardiac and thoracic aortic operations: a scientific statement from the American Heart Association.
Is two really better than one? Examining the superiority of dual modality neurophysiological monitoring during carotid endarterectomy: a meta-analysis.
Monitoring electrophysiologic function during carotid endarterectomy: a comparison of somatosensory evoked potentials and conventional electroencephalogram.
Intraoperative neuromonitoring for both arch surgery and descending/thoracoabdominal interventions remains debated. While opponents complain about the common false-positive findings, the authors from Pittsburgh describe the predictive value of intraoperative neurophysiologic monitoring (IONM) during aortic arch surgery with hypothermic circulatory arrest.1 Importantly, the results demonstrated that patients with IONM changes had an increased mortality and adverse neurologic events, with somatosensory-evoked potentials being more sensitive than electroencephalogram in predicting neurologic events.
Aortic arch surgery, both in acute and elective settings, remains one of the most complex cardiac procedures. Despite the efforts that have been undertaken to reduce complications, high mortality and morbidity are still associated with this kind of surgery, as reported by the largest international registries.1,2 Particularly, neurologic injury has historically been the most-feared complication. Strategies for improving mortality and preventing neurologic events have developed over time. In a contemporary practice, hypothermic circulatory arrest with antegrade or retrograde cerebral perfusion, neurophysiologic intraoperative monitoring, and pharmacologic strategies have become ways to reduce the incidence of neurologic injury of aortic arch surgery.
01124-7&id=fx1.jpg){kind=link}
01124-7&id=fx2.jpg){kind=link}
01124-7&id=gr1.jpg){kind=link}
01124-7&id=gr2.jpg){kind=link}
01124-7&id=gr3.jpg){kind=link}
01124-7&id=gr4.jpg){kind=link}
01124-7&id=fx4.jpg){kind=link}