ORIGINAL RESEARCH
Interventional Neuroradiology Cases With Intraprocedural Clots: Initial Experience David S. Gloss* and Brian W. Alkire†
Objective: Intraprocedural clot formation is rare but potentially serious complication of interventional neuroradiology procedures. We investigate if intraoperative monitoring (IOM) has utility to detect such clots. Intraprocedural clot formation is a rare but potentially serious complication of interventional neuroradiology procedures. Intraoperative monitoring detected nearly half of the included cases first. All of the included patients without improvement of the IOM changes were discharged home at best severely disabled. Methods: The study included patients with thromboembolic events during interventional neuroradiology cases at Barrow Neurologic Institute from 2006 to 2010, with prespecified outcomes. Electroneurodiagnostic recordings were reviewed. Results: Twelve patients were included in this study. All showed changes in their IOM. Five showed a change in IOM recording before a change was seen on the angiogram. Two returned to baseline, four improved but not to baseline, and six did not improve at all. All six patients without IOM improvement were discharged at best severely disabled. Ten of the 12 patients with clots had a subarachnoid hemorrhage before treatment. Conclusions: Intraoperative monitoring recording may be a valuable tool in monitoring patients during endovascular treatment to identify intraprocedural thromboembolic events. Intraoperative monitoring may correlate with poor outcomes when the changed responses do not improve. These data might be important when determining how aggressive to be in treating intraprocedural clots. Electroneurodiagnostic seems to be particularly warranted in patients with subarachnoid hemorrhage before treatment. A larger study is needed to validate these findings. Key Words: Intraoperative monitoring, Interventional neuroradiology, Clot. (J Clin Neurophysiol 2014;31: 229–231)
E
ndovascular therapy has significant risk of complications. There is an approximately 5% risk of complications for endovascular embolization of arteriovascular malformations (Lu et al., 2011). The rate of poor outcome after endovascular treatment among ruptured aneurysms is about 20% to 25% from the ISAT and BRAT trials (McDougall et al., 2012; Molyneux et al., 2002). The risk of neurologic complications of endovascular treatment from unruptured aneurysms from a recent meta-analysis is 8% (Hwang et al., 2012). A series investigating the use of intraoperative monitoring (IOM) during neurointerventional procedures had changes occur in 26%, which changed management in 14% (Liu et al., 2003).
From the *Geisinger Health System, Danville, PA; and the †Barrow Neurological Institute, Phoenix, AZ. Presented as a poster at the ACNS annual meeting, February 5–10, 2013, Miami, Florida, U.S.A. Address correspondence and reprint requests to David S. Gloss, MD, Barrow Neurological Institute, 350 W Thomas Rd, 8 HLT, Phoenix, AZ 85013, U.S.A.; e-mail: [email protected]. Copyright Ó 2014 by the American Clinical Neurophysiology Society
ISSN: 0736-0258/14/3103-0229
Intraprocedural stroke is a significant cause of morbidity and mortality from aneurysm surgery (Li et al., 2013). Intraprocedural clot formation is a known risk of interventional neuroradiology (Hahnel et al., 2003; Ries et al., 2009). Because there is evidence that intraoperative monitoring is safe and effective during endovascular coiling of aneurysms, and intraprocedural clot formation occurs, we wanted to investigate the experience at Barrow Neurological Institute with intraprocedural clot formation during neurointerventional procedures.
METHODS We performed an institutional review board approved, retrospective review of the Barrow Neurologic Institute records of intraoperative monitoring of endovascular procedures from 2006 to 2010. We searched to identify cases of intraprocedural clot formation. Inclusion criteria were: there was a clot formed during the procedure, IOM was performed for the procedure, and a Glasgow outcome score (GOS) was obtainable at the end of the hospitalization. There were no exclusion criteria. Our IOM neurointerventional cases are monitored with somatosensory evoked potentials, using Cadwell systems (versions 2.2.17 or 2.5.395). Sometimes, brainstem auditory evoked potentials, also using Cadwell systems, are used. Continuous real time brainstem auditory evoked potentials were obtained by alternately stimulating bilateral ears with a 100 ms square-wave pulse, 90 to 100 dB normal hearing level acoustic click at a repetition rate of 13.11 Hz through Etymotic Research ER-3 insert ear phones. Left and right brainstem auditory evoked potentials were recorded from Cz-A1, Cz-A2, and Cz- posterior cervical using 100 to 1500 Hz filter bandpass. Waves I, III, and V were identified, and the latency and amplitude of wave V were measured. The clinical details, radiologic data, GOS, and IOM changes were collated from the charts and the IOM record.
Alarm Criteria The initial recording, made after induction of anesthesia and before positioning, were used as baseline values. ACNS guidelines were used to define a significant change in somatosensory evoked potentials and brainstem auditory evoked potentials.
RESULTS Fifteen patients were initially identified of 1,211 total patients as having an intraprocedural clot formation. Of the 15 identified, when charts were formally reviewed, 12 were included. Three patients did not meet inclusion criteria. All 12 were neurointerventional angiography procedures done under general anesthesia. The following preprocedure vascular abnormalities were in the process of being treated when a clot formed:
Journal of Clinical Neurophysiology Volume 31, Number 3, June 2014
229
D. S. Gloss and B. W. Alkire
Journal of Clinical Neurophysiology Volume 31, Number 3, June 2014
FIG. 1. A patient with a ruptured right posterior communicating artery aneurysm. During coil embolization, the left upper extremity and left lower extremity somatosensory evoked potentials (SSEPs) became absent. An angiogram was subsequently acquired, which showed a branch of the right middle cerebral artery to be occluded. This clot was treated with alteplase, abciximab, and mechanical thrombectomy. Post treatment, the left lower extremity SSEP showed improvement but the left upper extremity SSEP remained absent. Post procedure, this patient had a right hemisphere stroke that was treated with a right hemicraniectomy. 1. 2. 3. 4. 5.
One ophthalmic artery aneurysm. One angioplasty for stenosis. Two angioplasty for vasospasm. Four posterior communicating artery aneurysms. Four anterior communicating artery aneurysms.
Refer Figs. 1 and 2 for example cases and Table 1 for the results of the 12 patients. All patients with a partial or full recovery of IOM were discharged with a GOS range of 3 to 5. Of the patients without improvement, three of these six patients had a GOS of 1 and none of them had a GOS greater than 3. Ten of the 12 patients had subarachnoid hemorrhage before treatment. Five of the 12 clots were discovered first by changes in IOM, 3 were seen first on the angiogram. In four of the cases, it remains unclear about which showed the clot first. In one case, a clot was identified by IOM changes after the final angiogram was done, so without the IOM change, would not have been identified at all.
DISCUSSION Intraoperative monitoring identified an ischemic change in all 12 patients. Our data may suggest American Academy of Neurology
Class III evidence for the detection of intraprocedural clot formation during neurointerventional procedures. Confirmation of these results with another study rated above class IV may be enough to suggest a community standard. In this series, there was a risk of approximately 1% for intraprocedural clot formation during interventional neuroradiology cases. Ten of the 12 patients with clots had subarachnoid hemorrhage. In all six patients who did not have improvement of IOM, there was at best severe disability at hospital discharge. This information has the potential to change the behavior of the treating physician. This may suggest such patients may benefit from more aggressive treatment of the clot, although this observation should be considered preliminary until there is a dedicated study examining this putative correlation.
CONCLUSIONS Intraoperative monitoring recording identified every included case of intraprocedural thromboembolic events during endovascular treatment. When IOM responses showed a change and did not subsequently improve, all patients had a poor outcome. This might be important when determining how aggressive to be in treating intraprocedural clots. Intraoperative monitoring may be particularly
FIG. 2. Angiogram from a patient being treated for a ruptured left posterior communicating artery aneurysm. During treatment, the right median and right tibial somatosensory evoked potential (SSEP) responses decreased, and a clot was seen in the left MCA. During endovascular treatment of the clot, improvement of the SSEP responses coincided with improved perfusion on the angiogram. Repeated decrease in SSEPs coincided with new clot formation that eventually extended down into the left internal carotid artery. At the end of the procedure SSEPs from the right median and right tibial were absent. Post-procedure the patient had a devastating left hemisphere stroke that resulted in brain death. 230
Copyright Ó 2014 by the American Clinical Neurophysiology Society
Journal of Clinical Neurophysiology Volume 31, Number 3, June 2014
TABLE 1. Individual results of each patient: IOM change, change before seen on angiogram, IOM return to baseline, Glasgow Outcome Score at hospital discharge. Patient Number 1 2 3 4 5 6 7 8 9 10 11 12
IOM Change
IOM Change Before Clot Seen on Angiogram
Did IOM Return to Baseline
Glasgow Outcome Score
Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
No Yes Unclear Unclear Yes Yes No Unclear No Yes Unclear Yes
Partial Partial Yes No Partial No Partial No No Partial No No
5 5 3 1 3 1 3 3 1 3 3 1
warranted in patients with subarachnoid hemorrhage before treatment. A larger, preferably prospective, study is needed to validate these findings.
Copyright Ó 2014 by the American Clinical Neurophysiology Society
Neuroradiology Cases With Intraprocedural Clots
ACKNOWLEDGMENTS The author would like to thank Gary Gronseth, the evidencebased medicine consultant of the American Academy of Neurology for discussions, about choosing an outcome that would be objective, so we could attempt to generate American Academy of Neurology Class III data. REFERENCES
Hahnel S, Schellinger PD, Gutschalk A, et al. Local intra-arterial fibinolysis of thromboemboli occurring during neuroendovascular procedures with recombinant tissue plasminogen activator. Stroke 2003;34:1723–1728. Hwang JS, Hyun MK, Lee HJ, et al. Endovascular coiling versus neurosurgical clipping in patients with unruptured intracranial aneurysm: a systematic review. BMC Neurol 2012;12:99–106. Li H, Pan R, Wang H, et al. Clipping versus coiling for ruptured intracranial aneurysms a systemiatic review and meta-analysis. Stroke 2013;44:29–37. Liu AY, Lopez JR, Do HM, et al. Neurophysiological monitoring of endovascular therapy of aneurysms. AJNR Am J Neuroradiol 2003;24:1530–1537. Lu X, Wu Z, Jiang C, et al. Complication risk of endovascular embolization for cerebral arteriovascular malformation. Eur J Radiol 2011;80:776–779. McDougall CG, Spetzler RF, Zabramski JM, et al. The Barrow Ruptured Aneurysm Trial. J Neurosurg 2012;116:135–144. Molyneux A, Kerr R, Stratton I, et al. International Subarachnoid Aneurysm Trial (ISAT) of neurosurgical clipping versus endovascular coiling in 2143 patients with ruptured intracranial aneurysms: a randomised trial. Lancet 2002;360:1267–1274. Ries T, Siemonsen S, Grzyska U, et al. Abciximab is a safe rescue therapy in thromboembolic event complicating cerebral aneurysm coil embolization: single center experience in 42 cases and review of the literature. Stroke 2009;40:1750–1757.
231
Interventional Neuroradiology Cases With Intraprocedural Clots: Initial Experience David S. Gloss* and Brian W. Alkire†
Objective: Intraprocedural clot formation is rare but potentially serious complication of interventional neuroradiology procedures. We investigate if intraoperative monitoring (IOM) has utility to detect such clots. Intraprocedural clot formation is a rare but potentially serious complication of interventional neuroradiology procedures. Intraoperative monitoring detected nearly half of the included cases first. All of the included patients without improvement of the IOM changes were discharged home at best severely disabled. Methods: The study included patients with thromboembolic events during interventional neuroradiology cases at Barrow Neurologic Institute from 2006 to 2010, with prespecified outcomes. Electroneurodiagnostic recordings were reviewed. Results: Twelve patients were included in this study. All showed changes in their IOM. Five showed a change in IOM recording before a change was seen on the angiogram. Two returned to baseline, four improved but not to baseline, and six did not improve at all. All six patients without IOM improvement were discharged at best severely disabled. Ten of the 12 patients with clots had a subarachnoid hemorrhage before treatment. Conclusions: Intraoperative monitoring recording may be a valuable tool in monitoring patients during endovascular treatment to identify intraprocedural thromboembolic events. Intraoperative monitoring may correlate with poor outcomes when the changed responses do not improve. These data might be important when determining how aggressive to be in treating intraprocedural clots. Electroneurodiagnostic seems to be particularly warranted in patients with subarachnoid hemorrhage before treatment. A larger study is needed to validate these findings. Key Words: Intraoperative monitoring, Interventional neuroradiology, Clot. (J Clin Neurophysiol 2014;31: 229–231)
E
ndovascular therapy has significant risk of complications. There is an approximately 5% risk of complications for endovascular embolization of arteriovascular malformations (Lu et al., 2011). The rate of poor outcome after endovascular treatment among ruptured aneurysms is about 20% to 25% from the ISAT and BRAT trials (McDougall et al., 2012; Molyneux et al., 2002). The risk of neurologic complications of endovascular treatment from unruptured aneurysms from a recent meta-analysis is 8% (Hwang et al., 2012). A series investigating the use of intraoperative monitoring (IOM) during neurointerventional procedures had changes occur in 26%, which changed management in 14% (Liu et al., 2003).
From the *Geisinger Health System, Danville, PA; and the †Barrow Neurological Institute, Phoenix, AZ. Presented as a poster at the ACNS annual meeting, February 5–10, 2013, Miami, Florida, U.S.A. Address correspondence and reprint requests to David S. Gloss, MD, Barrow Neurological Institute, 350 W Thomas Rd, 8 HLT, Phoenix, AZ 85013, U.S.A.; e-mail: [email protected]. Copyright Ó 2014 by the American Clinical Neurophysiology Society
ISSN: 0736-0258/14/3103-0229
Intraprocedural stroke is a significant cause of morbidity and mortality from aneurysm surgery (Li et al., 2013). Intraprocedural clot formation is a known risk of interventional neuroradiology (Hahnel et al., 2003; Ries et al., 2009). Because there is evidence that intraoperative monitoring is safe and effective during endovascular coiling of aneurysms, and intraprocedural clot formation occurs, we wanted to investigate the experience at Barrow Neurological Institute with intraprocedural clot formation during neurointerventional procedures.
METHODS We performed an institutional review board approved, retrospective review of the Barrow Neurologic Institute records of intraoperative monitoring of endovascular procedures from 2006 to 2010. We searched to identify cases of intraprocedural clot formation. Inclusion criteria were: there was a clot formed during the procedure, IOM was performed for the procedure, and a Glasgow outcome score (GOS) was obtainable at the end of the hospitalization. There were no exclusion criteria. Our IOM neurointerventional cases are monitored with somatosensory evoked potentials, using Cadwell systems (versions 2.2.17 or 2.5.395). Sometimes, brainstem auditory evoked potentials, also using Cadwell systems, are used. Continuous real time brainstem auditory evoked potentials were obtained by alternately stimulating bilateral ears with a 100 ms square-wave pulse, 90 to 100 dB normal hearing level acoustic click at a repetition rate of 13.11 Hz through Etymotic Research ER-3 insert ear phones. Left and right brainstem auditory evoked potentials were recorded from Cz-A1, Cz-A2, and Cz- posterior cervical using 100 to 1500 Hz filter bandpass. Waves I, III, and V were identified, and the latency and amplitude of wave V were measured. The clinical details, radiologic data, GOS, and IOM changes were collated from the charts and the IOM record.
Alarm Criteria The initial recording, made after induction of anesthesia and before positioning, were used as baseline values. ACNS guidelines were used to define a significant change in somatosensory evoked potentials and brainstem auditory evoked potentials.
RESULTS Fifteen patients were initially identified of 1,211 total patients as having an intraprocedural clot formation. Of the 15 identified, when charts were formally reviewed, 12 were included. Three patients did not meet inclusion criteria. All 12 were neurointerventional angiography procedures done under general anesthesia. The following preprocedure vascular abnormalities were in the process of being treated when a clot formed:
Journal of Clinical Neurophysiology Volume 31, Number 3, June 2014
229
D. S. Gloss and B. W. Alkire
Journal of Clinical Neurophysiology Volume 31, Number 3, June 2014
FIG. 1. A patient with a ruptured right posterior communicating artery aneurysm. During coil embolization, the left upper extremity and left lower extremity somatosensory evoked potentials (SSEPs) became absent. An angiogram was subsequently acquired, which showed a branch of the right middle cerebral artery to be occluded. This clot was treated with alteplase, abciximab, and mechanical thrombectomy. Post treatment, the left lower extremity SSEP showed improvement but the left upper extremity SSEP remained absent. Post procedure, this patient had a right hemisphere stroke that was treated with a right hemicraniectomy. 1. 2. 3. 4. 5.
One ophthalmic artery aneurysm. One angioplasty for stenosis. Two angioplasty for vasospasm. Four posterior communicating artery aneurysms. Four anterior communicating artery aneurysms.
Refer Figs. 1 and 2 for example cases and Table 1 for the results of the 12 patients. All patients with a partial or full recovery of IOM were discharged with a GOS range of 3 to 5. Of the patients without improvement, three of these six patients had a GOS of 1 and none of them had a GOS greater than 3. Ten of the 12 patients had subarachnoid hemorrhage before treatment. Five of the 12 clots were discovered first by changes in IOM, 3 were seen first on the angiogram. In four of the cases, it remains unclear about which showed the clot first. In one case, a clot was identified by IOM changes after the final angiogram was done, so without the IOM change, would not have been identified at all.
DISCUSSION Intraoperative monitoring identified an ischemic change in all 12 patients. Our data may suggest American Academy of Neurology
Class III evidence for the detection of intraprocedural clot formation during neurointerventional procedures. Confirmation of these results with another study rated above class IV may be enough to suggest a community standard. In this series, there was a risk of approximately 1% for intraprocedural clot formation during interventional neuroradiology cases. Ten of the 12 patients with clots had subarachnoid hemorrhage. In all six patients who did not have improvement of IOM, there was at best severe disability at hospital discharge. This information has the potential to change the behavior of the treating physician. This may suggest such patients may benefit from more aggressive treatment of the clot, although this observation should be considered preliminary until there is a dedicated study examining this putative correlation.
CONCLUSIONS Intraoperative monitoring recording identified every included case of intraprocedural thromboembolic events during endovascular treatment. When IOM responses showed a change and did not subsequently improve, all patients had a poor outcome. This might be important when determining how aggressive to be in treating intraprocedural clots. Intraoperative monitoring may be particularly
FIG. 2. Angiogram from a patient being treated for a ruptured left posterior communicating artery aneurysm. During treatment, the right median and right tibial somatosensory evoked potential (SSEP) responses decreased, and a clot was seen in the left MCA. During endovascular treatment of the clot, improvement of the SSEP responses coincided with improved perfusion on the angiogram. Repeated decrease in SSEPs coincided with new clot formation that eventually extended down into the left internal carotid artery. At the end of the procedure SSEPs from the right median and right tibial were absent. Post-procedure the patient had a devastating left hemisphere stroke that resulted in brain death. 230
Copyright Ó 2014 by the American Clinical Neurophysiology Society
Journal of Clinical Neurophysiology Volume 31, Number 3, June 2014
TABLE 1. Individual results of each patient: IOM change, change before seen on angiogram, IOM return to baseline, Glasgow Outcome Score at hospital discharge. Patient Number 1 2 3 4 5 6 7 8 9 10 11 12
IOM Change
IOM Change Before Clot Seen on Angiogram
Did IOM Return to Baseline
Glasgow Outcome Score
Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
No Yes Unclear Unclear Yes Yes No Unclear No Yes Unclear Yes
Partial Partial Yes No Partial No Partial No No Partial No No
5 5 3 1 3 1 3 3 1 3 3 1
warranted in patients with subarachnoid hemorrhage before treatment. A larger, preferably prospective, study is needed to validate these findings.
Copyright Ó 2014 by the American Clinical Neurophysiology Society
Neuroradiology Cases With Intraprocedural Clots
ACKNOWLEDGMENTS The author would like to thank Gary Gronseth, the evidencebased medicine consultant of the American Academy of Neurology for discussions, about choosing an outcome that would be objective, so we could attempt to generate American Academy of Neurology Class III data. REFERENCES
Hahnel S, Schellinger PD, Gutschalk A, et al. Local intra-arterial fibinolysis of thromboemboli occurring during neuroendovascular procedures with recombinant tissue plasminogen activator. Stroke 2003;34:1723–1728. Hwang JS, Hyun MK, Lee HJ, et al. Endovascular coiling versus neurosurgical clipping in patients with unruptured intracranial aneurysm: a systematic review. BMC Neurol 2012;12:99–106. Li H, Pan R, Wang H, et al. Clipping versus coiling for ruptured intracranial aneurysms a systemiatic review and meta-analysis. Stroke 2013;44:29–37. Liu AY, Lopez JR, Do HM, et al. Neurophysiological monitoring of endovascular therapy of aneurysms. AJNR Am J Neuroradiol 2003;24:1530–1537. Lu X, Wu Z, Jiang C, et al. Complication risk of endovascular embolization for cerebral arteriovascular malformation. Eur J Radiol 2011;80:776–779. McDougall CG, Spetzler RF, Zabramski JM, et al. The Barrow Ruptured Aneurysm Trial. J Neurosurg 2012;116:135–144. Molyneux A, Kerr R, Stratton I, et al. International Subarachnoid Aneurysm Trial (ISAT) of neurosurgical clipping versus endovascular coiling in 2143 patients with ruptured intracranial aneurysms: a randomised trial. Lancet 2002;360:1267–1274. Ries T, Siemonsen S, Grzyska U, et al. Abciximab is a safe rescue therapy in thromboembolic event complicating cerebral aneurysm coil embolization: single center experience in 42 cases and review of the literature. Stroke 2009;40:1750–1757.
231
