Intraoperative motor-evoked potential monitoring during coil embolization of internal carotid–anterior choroidal artery aneurysms: A comparative case series and lessons on monitoring pitfalls

Abstract

Background:

Ischemic injury to the anterior choroidal artery (AChA) territory during coil embolization of internal carotid artery–AChA (ICA–AChA) aneurysms can be functionally devastating and is not always detected by angiography. We examined the utility and the interpretive pitfalls of intraoperative motor-evoked potential (MEP) monitoring in this specific aneurysm subgroup.

Methods:

We retrospectively reviewed 16 consecutive patients who underwent coil embolization for ICA–AChA aneurysms at a single institution. Seven patients (2013–2025) were monitored with intraoperative MEP (MEP group); 9 patients (2005–2015) treated before monitoring was introduced served as a historical comparison group (non-MEP group). Angiographic AChA changes, MEP waveform changes, intraoperative interventions, and postoperative diffusion-weighted magnetic resonance imaging were analyzed.

Results:

Significant MEP amplitude reduction (≥50%) occurred in two monitored patients, prompting coil retrieval, microcatheter repositioning, or a procedural pause; amplitudes recovered in all. One patient showed transient MEP reduction without angiographic change (possible false-positive). No patient in the MEP group developed postoperative infarction (0%), whereas two of nine non-MEP patients did (22%).

Conclusion:

In this small comparative series, MEP monitoring provided functional information complementary to angiography, with no infarctions in the monitored group. MEP changes must be interpreted alongside angiographic and clinical findings, as both false-positive and false-negative results occur. These hypothesis-generating findings warrant prospective validation.

INTRODUCTION

Internal carotid–anterior choroidal artery (ICA–AChA) aneurysms are uncommon, accounting for approximately 2–5% of intracranial aneurysms.[1] Despite their rarity, they carry disproportionate clinical importance because the anterior choroidal artery (AChA) supplies eloquent structures including the posterior limb of the internal capsule, the optic tract, the lateral geniculate body, and portions of the mesencephalon. Even transient ischemia in this territory can produce dense hemiparesis or visual field loss, making intraoperative protection of the AChA a central concern during endovascular treatment.

Coil embolization is well established for ICA–AChA aneurysms, yet ischemic complications occur in 4–16% of AChA aneurysm interventions, sometimes despite technically successful coiling.[1,4,10] Subtle reductions in AChA flow – from branch occlusion, catheter manipulation, or pressure-related perfusion changes – may not be evident on digital subtraction angiography (DSA). This limitation has driven interest in functional intraoperative monitoring, particularly transcranial motor-evoked potentials (MEPs). Nakagawa et al. studied 164 consecutive patients undergoing coil embolization for intracranial aneurysms of various locations under MEP monitoring and showed that significant MEP changes reliably flagged intraprocedural ischemia and enabled prompt salvage.[6] However, that cohort was anatomically heterogeneous, with only a small fraction of perforator- or branch-related aneurysms, and lacked a non-monitored comparison group. Smaller series have reported MEP changes during AChA aneurysm embolization that preceded or occurred without corresponding angiographic findings,[2,3] suggesting that electrophysiological monitoring may detect functional ischemia earlier than imaging.

The specific value of MEP monitoring in ICA–AChA aneurysms – where the consequences of perforator ischemia are particularly severe – and its incremental benefit over angiography alone remain insufficiently characterized. Equally important, and underemphasized in prior work, is how MEP findings should be interpreted when they conflict with angiography. We, therefore, reviewed our consecutive experience, comparing a monitored cohort with a historical non-monitored cohort, with two aims: to describe the procedural utility of MEP monitoring, and to illustrate the interpretive pitfalls – including false-positive and potential false-negative changes – that bear directly on day-to-day endovascular practice.

MATERIALS AND METHODS

Study design and patient selection

This retrospective review was conducted at a single tertiary neurosurgical center. Consecutive patients with ICA–AChA aneurysms who underwent coil embolization between September 2005 and April 2025 were reviewed. Procedures before July 2015 did not include MEP monitoring; from March 2013, MEP monitoring was incorporated into routine practice. Because monitoring was adopted gradually, its use during the transitional period (March 2013–July 2015) was at the operator’s discretion, which accounts for the partial overlap in treatment eras between the two groups. This yielded a monitored group (n = 7) and a non-monitored group (n = 9). Eligible patients had an aneurysm arising from the AChA origin or the adjacent ICA segment and underwent coil embolization with complete perioperative imaging and clinical data. Patients with fusiform aneurysms, non-AChA territory lesions, or incomplete records were excluded from the study. The study was approved by the Institutional Ethics Committee (approval no. Yokorin-202539); informed consent was waived given the retrospective design.

Endovascular procedure

All procedures were performed under general anesthesia by board-certified neuroendovascular specialists through a transfemoral approach. After systemic heparinization, a guiding catheter was positioned in the cervical ICA and a microcatheter advanced to the aneurysm under biplane fluoroscopy. Coil selection and packing were at the operator’s discretion; balloon- and stent-assisted techniques were not used in the monitored cohort. Angiography was performed at multiple time points to assess AChA patency, particularly after coil insertion or catheter adjustment. Delayed AChA visualization was defined as any new slowing of contrast filling relative to baseline, including delayed branch opacification, distal branch disappearance, or reduced contrast density. When compromise was suspected, microcatheter repositioning, coil retrieval, or intra-arterial pharmacological therapy (e.g., argatroban) was permitted.

Anesthesia and MEP monitoring protocol

In the monitored cohort, total intravenous anesthesia with propofol and remifentanil was used, avoiding neuromuscular blockade after intubation. Transcranial stimulation used screw electrodes at C3/C4 (international 10–20 system); recording electrodes were placed over the bilateral abductor pollicis brevis, tibialis anterior, and abductor hallucis muscles. Stimulation used a train of five pulses, 50-μs pulse width, 2-ms interstimulus interval, and 400–500 V intensity, with a 1.5-kHz high-cut filter. MEPs were assessed before microcatheter navigation, after each coil insertion, after angiographic runs indicating AChA changes, and at procedure end. Amplitude reduction ≥50% from baseline in any monitored muscle was considered significant; on detection, the procedure was paused to evaluate technical or physiological causes followed by targeted intervention.

Clinical and radiographic assessment

Neurological examinations were performed on emergence from anesthesia and throughout hospitalization. Diffusion-weighted magnetic resonance imaging (MRI) was routinely obtained within 24 h to detect AChA-territory ischemia. Extracted variables included age, sex, aneurysm size, rupture status, coil length, intraoperative AChA findings, MEP changes (where applicable), monitoring-prompted procedural modifications, and outcomes. Given the descriptive nature and small cohort, no formal statistical comparison was performed; data are presented as ranges or counts.

RESULTS

Overall findings

Seven patients comprised the MEP-monitored group and nine the non-monitored group; baseline and procedural details appear in Tables 1 and 2. Descriptively, the non-monitored group tended to be older with a higher proportion of unruptured aneurysms, factors that may independently influence outcomes. The monitored cohort (five unruptured, two ruptured) had a mean age of 55.6 years (range 30–71). Significant intraoperative MEP changes (≥50% reduction) occurred in Patients 3 and 7. Angiographic evidence of transient AChA compromise (delayed visualization or branch disappearance) occurred in Patients 2, 3, and 4. No monitored patient developed postoperative infarction on diffusion-weighted MRI; two of nine non-monitored patients did. Monitoring- or angiography-prompted interventions included coil retrieval (Patients 2 and 3), microcatheter repositioning (Patients 3 and 4), and intra-arterial argatroban (Patient 4); all produced angiographic or electrophysiological improvement without lasting deficits.

Table 1:

Patients who underwent coil embolization for ICA–AChA aneurysms without motor-evoked potential monitoring (non-MEP group).^†

Table 2:

Patients who underwent coil embolization for ICA–AChA aneurysms with motor-evoked potential monitoring (MEP group.

Illustrative cases

Concordant electrophysiological and angiographic change (Patient 3). A 30-year-old woman with an unruptured 3 × 2.5 × 2 mm aneurysm developed branch disappearance on DSA after partial coil packing [Figure 1A], with simultaneous 50% amplitude reduction in the left abductor pollicis brevis [Figure 1B]. Coil removal and microcatheter adjustment restored both AChA flow and MEP amplitudes. Postoperative MRI showed no infarction [Figure 1C]. This case exemplifies the intended use of MEP monitoring: corroborating an angiographic warning and triggering timely correction. The angiographic images and the corresponding MEP recordings are now displayed together in a single figure to allow direct temporal comparison.

Figure 1:

Patient 3 – concordant angiographic and electrophysiological change. (A) Intraoperative digital subtraction angiography: Left, pre-embolization; middle, after the first coil, the anterior choroidal artery (AChA) branch is no longer visualized; right, after the third coil. (B) Intraoperative motor-evoked potential recording from the left abductor pollicis brevis, corresponding in time to the angiographic images in (A): The amplitude decreased after Galaxy coil placement and recovered after coil retrieval, with a similar decrease and recovery during target coil placement and retrieval. (C) Diffusion-weighted MRI on the first postoperative day shows no infarction in the AChA territory. Red arrows indicate the AChA and key waveform changes. Red text indicates procedural events.

Angiographic change without MEP change (Patients 2 and 4). In Patient 2 (ruptured 8 × 6 × 3 mm), delayed AChA filling after the fifth coil [Figure 2A] prompted coil retrieval despite stable MEPs [Figure 2B], with normal postoperative MRI. In Patient 4 (unruptured 4 × 4 × 3.5 mm), AChA visualization deteriorated after microcatheter withdrawal without MEP change [Figure 3A]; intra-arterial argatroban improved flow, and MRI was unremarkable; MEP amplitudes remained stable throughout [Figure 3B]. These cases show that not every angiographic change is accompanied by functional compromise, and that stable MEPs may support a more conservative, perfusion-oriented response.

Figure 2:

Patient 2 – angiographic change without motor-evoked potential (MEP) change. (A) Intraoperative digital subtraction angiography: Left, after placement of the fifth coil, visualization of the anterior choroidal artery (AChA) is delayed; middle, unsubtracted image obtained during coil retrieval; right, after coil retrieval the AChA is again visualized (red arrow). (B) Corresponding MEP recordings from the right and left hands over the same period, showing preserved amplitudes throughout. The dissociation between the angiographic and electrophysiological findings is directly appreciable within the single figure. A–H:Eight consecutive MEP recordings acquired sequentially; 1 = right hand (right abductor pollicis brevis), 2 = left hand (left abductor pollicis brevis).

Figure 3:

Patient 4 – angiographic change without motor-evoked potential (MEP) change. (A) Intraoperative digital subtraction angiography: Top, before coil embolization; middle, after placement of the fifth coil (total coil length 25 cm) and microcatheter removal, the anterior choroidal artery (AChA) is no longer visualized; bottom, improved visualization of the AChA after local intra-arterial argatroban infusion. (B) Corresponding MEP recording from the right abductor pollicis brevis, showing stable amplitudes through final (fifth) coil detachment, microcatheter removal, and argatroban administration. J–P (J2–P2) :Seven consecutive MEP recordings from the right abductor pollicis brevis; red labels mark procedural events (J2 final [5th] coil detachment; K2 microcatheter removal; O2 intra-arterial argatroban). Red text indicates procedural events.

MEP change without angiographic change (Patient 7). A 48-year-old woman with an unruptured 2.5 × 3.7 × 2.0 mm aneurysm developed transient MEP reduction in the right abductor hallucis after coil insertion, without angiographic change [Figure 4A-C]. A brief pause without further manipulation allowed amplitudes to recover, and MRI showed no infarction. This may represent a false-positive change from transient hemodynamic fluctuation or catheter-induced microvascular perturbation; however, true ischemia resolving before infarction cannot be excluded. It illustrates the practical difficulty of acting on an isolated MEP change. This patient also harbored a contralateral right ICA aneurysm, which was embolized during the same session after completion of the left ICA–AChA procedure; the coil length given in Table 2 refers to the left ICA–AChA aneurysm alone. The MEP change described here occurred during coiling of the left ICA–AChA aneurysm, and the involvement of the right abductor hallucis is anatomically consistent with the left corticospinal tract and therefore with the left AChA territory rather than with the contralateral procedure.

Figure 4:

Patient 7 – Motor-evoked potential (MEP) change without angiographic change. (A) Intraoperative digital subtraction angiography (DSA) (LAO 0°, caudal 20°) obtained at the time of the MEP change, showing the anterior choroidal artery (AChA), (red arrow) without interval change from baseline; the coil mass within the aneurysm is visible immediately adjacent to it. (B) Corresponding MEP recording from the right abductor hallucis: the amplitude decreased after coil insertion and normalized after a 3-minute pause without further manipulation. Red arrows indicate key waveform changes. (C) Three-dimensional rotational angiographic reconstruction of the same aneurysm (LAO 24°, caudal 26°; dome 3.64 × 3.83 × 3.75 mm, neck width 3.03 mm; the aneurysm is shaded red and the colored lines denote the measurement axes), demonstrating the AChA and its branching pattern close to its origin. The main trunk of the AChA was well visualized and unchanged on intraoperative DSA (A); the caliber and the proximity of its early branches illustrate why compromise confined to a branch or perforator may remain below the spatial resolution of two-dimensional angiography. Scale bars, 5 mm. I4, K4, M4, N4 :consecutive MEP recordings from the right abductor hallucis: I4 baseline; K4 after coil insertion; M4 attenuated amplitudes; N4 improved amplitudes after a brief pause.

The remaining monitored patients (1, 5, and 6) and seven of nine non-monitored patients completed embolization without electrophysiological, angiographic, or clinical abnormality. Two non-monitored patients developed AChA-territory infarction.

DISCUSSION

Preserving the AChA during coil embolization is difficult because perfusion changes may be subtle, transient, or angiographically occult, and ischemic complications have been reported even when the parent artery and AChA appear intact. DSA provides anatomical information, whereas MEPs reflect the functional integrity of corticospinal pathways; the two are complementary. In our monitored cohort, MEP changes occurred in two of seven patients, in both cases coinciding with or preceding angiographic evidence of compromise, and timely intervention restored amplitudes and perfusion. No monitored patient developed infarction, versus two in the non-monitored group. Given the small sample, this is hypothesis-generating rather than confirmatory, but it is consistent with a protective role for functional monitoring in this vulnerable territory.

Our findings align with and extend prior work. Nakagawa et al. demonstrated, in 164 patients across aneurysm locations, that significant MEP changes identified intraprocedural ischemia and enabled salvage, with permanent changes associated with deficits;[6] however, that series lacked a non-monitored comparison and included few AChA-territory aneurysms. Hiraishi et al. reported MEP reductions in three of seven patients during AChA aneurysm embolization despite no corresponding angiographic abnormality,[2] and Ito et al. described MEP decline as the first sign of compromise.[3] Comparable observations exist in open cerebrovascular surgery, where electrophysiological change may precede structural findings.[5] Our series adds a focused ICA–AChA cohort with a historical comparison arm.

Interpretive pitfalls of MEP monitoring. The principal practical lesson of this series concerns the two-directional fallibility of MEP monitoring. Patient 7 illustrates a probable false-positive: a transient amplitude reduction without angiographic correlate that resolved on pausing, which could have prompted unnecessary intervention. Conversely, Takada et al. recently reported a false-negative result during coil embolization of ICA aneurysms (including an AChA aneurysm), in which post-procedural hemiplegia was caused by contrast-induced encephalopathy rather than ischemia, with stable intraoperative MEPs and full recovery within four days.[8] Taken together, these scenarios show that a normal MEP does not guarantee an uncomplicated course, and an abnormal MEP does not always signify true ischemia. MEP findings should therefore be integrated with angiographic appearance, the timing of the change relative to procedural steps, anesthetic and hemodynamic status, and the overall clinical picture, rather than acted on in isolation.

Mechanisms underlying the dissociation between angiographic and electrophysiological findings. Because the intraprocedural decision to continue, pause, or reverse a maneuver often depends on which modality is believed, the discordant cases in this series warrant closer analysis. Patients 2 and 4 showed angiographic AChA compromise without an MEP change, whereas Patient 7 showed the reverse pattern.

In Patient 2, delayed AChA filling appeared after placement of the fifth coil and improved after coil retrieval, while MEP amplitudes in both upper limbs were unchanged throughout. The angiographic abnormality was a slowing of contrast filling rather than a fixed occlusion, and a subtotal, transient reduction in flow may remain above the threshold required for corticospinal conduction. The AChA territory additionally receives collateral supply through anastomoses with the posterior choroidal and adjacent perforating systems, which may sustain perfusion of the posterior limb of the internal capsule while AChA inflow is temporarily reduced.[9,11] In such a setting an angiographic change need not translate into functional compromise, and the absence of an MEP change is not evidence that the angiographic finding was spurious.

In Patient 4, AChA visualization deteriorated immediately after microcatheter withdrawal and improved after local intra-arterial argatroban, while MEP amplitudes remained stable. The temporal relationship to catheter manipulation and the response to a direct thrombin inhibitor implicate platelet–fibrin thrombus or microemboli with subsequent recanalization, rather than mechanical occlusion by the coil mass; thromboembolic events are the principal mechanism of ischemic complication in AChA aneurysm coiling.[1,4,10] Microemboli that lyse or recanalize before the ischemic threshold is reached, or that lodge in branches supplying territory outside the monitored myotomes, would be expected to produce an angiographic signature without an electrophysiological one. The same reasoning indicates that a stable MEP should not be taken as license to leave an angiographically compromised AChA untreated: in both Patients 2 and 4 the angiographic finding, not the MEP, correctly prompted corrective action.

Patient 7 represents the opposite pattern, and three explanations deserve consideration. First, the AChA is a small-caliber vessel with considerable anatomical variability: although usually single, it may arise as two or more branches directly from the ICA, and early branches may originate close to its origin,[9,11] with hyperplastic variants also described.[7] Compromise at the level of one such branch or of its perforators may lie beyond the spatial resolution of DSA while still producing a measurable functional change. This is the mechanism proposed by Hiraishi et al., who observed MEP reduction without angiographic change in three of seven AChA aneurysm embolizations,[2] and an analogous dissociation has been reported in open surgery for AChA aneurysms, in which MEP amplitude fell despite flow that remained visible on indocyanine green videoangiography.[5] Second, because MEPs sample only the muscles included in the recording montage, the correspondence between the ischemic territory and the monitored muscle is imperfect in both directions; a change confined to the abductor hallucis, as here, may reflect involvement of a restricted portion of the corticospinal projection. Third, transient hemodynamic or anesthetic fluctuation can depress MEP amplitude in the absence of ischemia. In this patient, anesthetic depth and blood pressure were stable and amplitudes normalized within three minutes without any corrective maneuver, which is most consistent with a false-positive change, although genuine ischemia resolving before infarction cannot be excluded.

These mechanisms carry direct management implications when the two modalities disagree. When an angiographic change occurs with stable MEPs, a perfusion-oriented response – coil retrieval, microcatheter repositioning, or intra-arterial antithrombotic therapy – together with continued monitoring is appropriate, and the preserved MEP argues against precipitate abandonment of the procedure. When an isolated MEP change occurs with an unchanged angiogram, a brief pause with reassessment of anesthetic depth, blood pressure, and stimulation parameters, followed by repeat angiography, is preferable to reflexive device manipulation, which itself carries thromboembolic risk. In both situations the behavior of the finding over the ensuing minutes – whether it reverses, persists, or progresses – is more informative than the initial observation in either modality alone.

Anatomical variability of the AChA compounds these difficulties. Early branching, hypoplastic or hyperplastic variants, and anomalous origins are well described,[9,11] and some hyperplastic variants resemble a duplicated posterior cerebral artery with potentially higher hemodynamic stress.[7] Such variability means angiographic visualization alone may not exclude territorial ischemia when branch filling is subtle – again arguing for complementary functional monitoring, interpreted with appropriate caution.

This study has important limitations. First, the cohort was small, reflecting the rarity of ICA–AChA aneurysms and the single-center design, precluding statistical analysis and limiting generalizability. Second, and most significantly, the two groups were treated in different eras (non-monitored 2005–2015; monitored 2013–2025). Advances in microcatheter and coil technology and accumulated operator experience may have independently improved outcomes in the monitored cohort, and the absence of a contemporaneous control precludes attributing the outcome difference to monitoring per se. The between-group comparison should be read as hypothesis-generating. Notably, both infarctions in the non-monitored group occurred in comparatively large aneurysms treated with long coil lengths (11 mm and 4.5 mm; total coil length 106 cm and 23 cm, respectively), so part of the between-group outcome difference may reflect aneurysm complexity rather than the absence of monitoring. Third, MEP thresholds are not standardized for endovascular procedures, and amplitude changes may reflect anesthetic or physiological variability; in our two patients with changes, anesthetic depth and blood pressure were stable, but retrospective confounding cannot be excluded. Fourth, angiographic interpretation of small perforators is inherently subjective. Finally, the mechanisms proposed above for the discordant cases are inferential: they are supported by the timing of each event, the response to the intervention used, and the published anatomical and electrophysiological literature, but they cannot be confirmed in a retrospective series without perfusion imaging or histological correlation.

Future work should establish standardized MEP thresholds for endovascular use, incorporate multimodal monitoring (e.g., somatosensory-evoked potentials), and, ideally through a multicenter prospective registry of AChA-region aneurysms, accumulate sufficient cases to determine whether MEP monitoring meaningfully reduces clinically relevant infarction.

CONCLUSION

In this small comparative series of ICA–AChA aneurysm embolization, intraoperative MEP monitoring provided functional information complementary to angiography, and no monitored patient developed postoperative infarction. The series also underscores that MEP monitoring is fallible in both directions: false-positive and potential false-negative results occur, so MEP changes must be interpreted together with angiographic and clinical findings. When the two modalities disagree, the most likely mechanism – transient subtotal flow reduction with collateral support, microembolism with recanalization, or compromise of an angiographically occult AChA branch – should be weighed explicitly, and the reversibility of the finding over the following minutes used to guide management. These hypothesis-generating observations support, but cannot yet confirm, a protective role for MEP monitoring and warrant prospective multicenter study.

Ethical approval:

The research/study approved by the Institutional Review Board at Yokohama General Hospital, number Yokorin-202539, dated March 23, 2026.

Declaration of patient consent:

Patient’s consent is not required as patients identity is not disclosed or compromised. This retrospective study was approved by the Institutional Ethics Committee of Yokohama General Hospital which waived the requirement for individual written informed consent; an opt-out method was used. All images are de-identified and contain no patient-identifying information.

Financial support and sponsorship:

Nil.

Conflicts of interest:

There are no conflicts of interest.

Use of artificial intelligence (AI)-assisted technology for manuscript preparation:

The author(s) confirms that they have used artificial intelligence (AI)-assisted technology. Claude (Anthropic) was used for literature collection and English language proofreading. The authors independently verified all information and take full responsibility for the content and accuracy of the manuscript.

Disclaimer

The views and opinions expressed in this article are those of the authors and do not necessarily reflect the official policy or position of the Journal or its management. The information contained in this article should not be considered to be medical advice; patients should consult their own physicians for advice as to their specific medical needs.