Tailoring endovascular strategy to posterior inferior cerebellar artery segment anatomy: Avoiding brainstem infarction in dissecting pseudoaneurysms

Abstract

Background

Dissecting pseudoaneurysms of the posterior inferior cerebellar artery (PICA) are rare, high-risk lesions with outcomes heavily influenced by anatomical location. This study evaluates the relationship between PICA segmental anatomy, endovascular treatment strategy, and clinical outcomes.

Methods

We retrospectively analyzed 21 patients with dissecting PICA aneurysms treated endovascularly between 2013 and 2025. Aneurysms were anatomically classified by segment (P1–P5); P3 was further subdivided into P3A (proximal to the PICA loop) and P3B (distal to the loop). Primary outcome was discharge modified Rankin Scale (mRS ≤ 2; classified as favorable). Secondary outcomes included ischemic stroke and segment-specific treatment trends.

Results

Of 21 patients, 16 (76%) were female and 13 (62%) were White; mean age was 57.3 ± 15.5 years. Most aneurysms were small (<7 mm, n = 16, 76%), ruptured (n = 17, 81%), and located in proximal segments (P1–P3, n = 14, 67%). Hypertension was present in 11 (52%) and intraventricular hemorrhage (IVH) in 12 (57%). Treatments included coiling (n = 10, 48%), Onyx embolization (n = 5, 24%), flow diversion (n = 2, 10%), and adjunctive techniques (n = 3, 14%). Deconstructive strategies were used in 14 (67%). Favorable outcome (mRS ≤ 2) was achieved in 15 (71%). Clinically silent cerebellar strokes occurred in 9 (43%); no brainstem infarctions were seen. IVH was significantly associated with poor outcome (p = 0.043). All P3A cases (n = 3) had favorable outcomes, while 2 of 4 P3B cases (50%) were unfavorable.

Conclusions

PICA segment anatomy predicts treatment safety. Reconstructive strategies are preferred for proximal lesions (P1–P3A); deconstructive methods are safe and effective beyond the PICA loop (P3B–P5).

Highlights

What is already known

Dissecting pseudoaneurysms of the posterior inferior cerebellar artery (PICA) are rare and complex anatomically, with treatment risks—especially of brainstem infarction—largely depending on segmental factors location.

What this study adds

This study presents a practical subdivision of the P3 segment (P3A vs P3B) and shows that customized, segment-specific endovascular approaches—reconstructive for P1–P3A and deconstructive for P3B–P5—improve outcomes and prevent brainstem infarction.

Why it matters

These findings provide actionable, anatomy-driven guidance for treating high-risk PICA dissections, reducing ischemic complications and supporting safe parent vessel sacrifice in select distal lesions.

Introduction

Dissecting distal pseudoaneurysms of the posterior inferior cerebellar artery (PICA) are uncommon vascular lesions that account for a small fraction of all intracranial aneurysms, with an estimated incidence of 0.5–0.7%. ¹ These lesions are underrepresented in the literature. They most commonly present with subarachnoid hemorrhage (SAH), and less frequently with ischemia, but their segmental location has critical implications for management and outcome.^2,3

The anatomical complexity of the PICA, particularly in its proximal trunk segments (P1–P3: anterior medullary, lateral medullary, and tonsillomedullary), poses unique challenges. ⁴ These regions give rise to vital perforating arteries that supply the medulla oblongata, including the dorsal respiratory group, vestibular nuclei, and ascending sensory tracts. Endovascular occlusion of these segments—even when technically feasible—carries a substantial risk of medullary infarction, with resultant brainstem syndromes such as lateral medullary (Wallenberg) syndrome and potentially fatal respiratory compromise.^5,6

While microsurgical trapping with or without bypass has historically been used, recent advances in endovascular therapy—including balloon-assisted or stent-assisted coiling (e.g. using low-profile intracranial stents), flow diversion with devices such as a pipeline embolization device, and liquid embolic agents like n-butyl cyanoacrylate (NBCA) or Onyx—have provided less invasive alternatives. These techniques can facilitate aneurysm occlusion while preserving the parent artery in suitable cases, thereby potentially reducing perioperative morbidity. However, in the context of proximal PICA dissections, parent vessel sacrifice remains controversial. Cases of ischemic brainstem injury despite adequate collateral visualization have been documented, highlighting the importance of preserving perforator-rich zones during treatment.^5,7 Alternative approaches—such as selective stenting, targeted embolization, or flow-modifying strategies—aim to balance aneurysm occlusion with the protection of critical perforators. ⁸

This study evaluates a contemporary endovascular series of dissecting aneurysms involving the distal PICA, anatomically stratified by segments. We aim to assess how aneurysm location influences treatment selection and to determine whether segment-specific endovascular strategies can reduce the risk of brainstem infarction and improve clinical outcomes.

Methods

Study design

A retrospective review was performed using a prospectively maintained institutional database of patients treated for intracranial aneurysms between January 2013 and April 2025. Patients who underwent endovascular treatment for PICA dissecting aneurysms were identified. The study received approval from the institutional review board, and the need for informed consent was waived due to the retrospective nature of the study. This study was reported in accordance with the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines. ⁹

Patient population

Inclusion criteria consisted of adult patients (≥18 years) with dissecting aneurysms located along the PICA trunk who were treated endovascularly. Patients were excluded if they had vertebral artery-PICA junction aneurysms, alternative vascular pathologies including dural arteriovenous fistulas, arteriovenous malformations, or aneurysms unrelated to the PICA trunk.

Aneurysms were anatomically classified using established PICA segment definitions. Segments P1–P3 (anterior medullary, lateral medullary, and tonsillomedullary) were grouped as proximal PICA, while P4 and P5 (telovelotonsillar and cortical segments) were categorized as distal PICA (4). Aneurysm size was categorized as small (<7 mm), medium (7–12 mm), large (13–25 mm), or giant (>25 mm), based on maximal diameter on pre-treatment imaging. ¹⁰

Indications for treatment included SAH, symptomatic unruptured aneurysms, or incidental aneurysms with high-risk morphologic or hemodynamic features as determined by the treating neurointerventionalist.

Study variables and outcomes

Demographic and clinical variables collected included age, sex, race, smoking history, and relevant comorbidities such as hypertension and diabetes mellitus. Aneurysm-specific variables included segment location, side, size, rupture status, and the presence of intraventricular hemorrhage (IVH) or hydrocephalus. Treatment modality was classified as reconstructive (coiling with or without adjuncts, flow diversion) or deconstructive (parent vessel occlusion using coils or liquid embolic agents). Documentation of intimal irregularity or dissection was based on intraoperative digital substraction angiography (DSA) findings and operator assessment.

The primary outcome of the study was clinical outcome at discharge, defined by the modified Rankin Scale (mRS), with favorable outcome classified as mRS ≤ 2 and unfavorable ones mRS > 2. Secondary outcomes included ischemic stroke complications rate (location confirmed by imaging), intimal irregularity treatment, and treatment approach stratified by aneurysm segment.

Statistical analysis

Descriptive statistics were used to summarize demographic, procedural, and outcome data. Normality of continuous variables was assessed with a Shapiro–Wilk test. A Mann–Whitney U test was used for continuous variables, while categorical variables were compared using a χ² test or Fisher's exact test where appropriate. Segment-based analysis (P1–P5) was conducted to assess associations between location, treatment type, and outcome. Multivariate analyses were omitted due to limited sample size. Statistical significance was defined as p < 0.05. Data was organized using Microsoft Excel and analyses using and SPSS v.30.

Results

The cohort consisted of 21 patients (76% female, n = 16), with a mean age of 57.3 ± 15.5 years, predominantly White (62%, n = 13) and Black or African American (38%, n = 8) (Table 1). Most aneurysms were small in size (76%, n = 16), left-sided (57%, n = 12), and ruptured at presentation (81%, n = 17). Hypertension was present in 52% (n = 11), and 57% (n = 12) had IVH. Aneurysms were primarily located in the proximal PICA segment (67%, n = 14) and treated during the acute phase (<72 h) in 71% (n = 15). Endovascular modalities included non-assisted coiling (48%, n = 10), Onyx embolization (24%, n = 5), flow diversion (10%, n = 2), and adjunctive techniques (balloon- or stent-assisted coiling) in three cases. A favorable outcome (mRS ≤ 2) was achieved in 71% (n = 15), and 43% (n = 9) experienced clinically silent ischemic cerebellar strokes.

Table 1.

Overview of the study sample characteristics.

Variable	N (%)
Sex
Female	16 (76.2)
Male	5 (23.8)
Race
White	13 (61.9)
Black or African American	8 (38.1)
Smoking history
Never smoker	13 (61.9)
Smoker	5 (23.8)
Prior smoker	2 (9.5)
Unknown	1 (4.8)
Diabetes mellitus history
Yes	2 (9.5)
No	19 (90.5)
Family history of aneurysm
Yes	1 (4.8)
No	20 (95.2)
Hypertension
Yes	11 (52.4)
No	10 (47.6)
Aneurysm size
Small	16 (76.2)
Medium	4 (19.0)
Giant	1 (4.8)
Side
Left	12 (57.1)
Right	9 (42.9)
Rupture status
Ruptured	17 (81.0)
Unruptured	4 (19.0)
Segment
Proximal	14 (66.7)
Distal	7 (33.3)
Intraventricular hemorrhage
Yes	12 (57.1)
No	9 (42.9)
Hydrocephalus
Yes	11 (52.4)
No	10 (47.6)
Treatment phase
Acute (<72 h)	15 (71.4)
Subacute (72 h to 4 weeks)	2 (9.5)
Elective	4 (19.0)
Endovascular treatment modality
Non-assisted coiling	11 (52.4)
Onyx	5 (23.8)
Flow diverter	2 (9.5)
Balloon-assisted coiling	1 (4.8)
Stent	1 (4.8)
Stent-assisted coiling	1 (4.8)
Stroke outcome
Cerebellar stroke	9 (42.9)
No stroke	12 (57.1)
mRS at discharge
Favorable (mRS ≤ 2)	15 (71.4)
Unfavorable (mRS > 2)	6 (28.6)

mRS: modified Rankin Scale.

Proximal aneurysms were more likely treated with reconstructive methods (18%, n = 4), while distal segments were predominantly managed with deconstructive strategies, including coil sacrifice and Onyx embolization (57%, n = 12). Onyx use was highest in P5 (14%, n = 3), with a non-significant trend toward segment-specific preference (p = 0.397). There were no re-ruptures for those presenting with SAH or ruptures among those treated electively. Treatment of intimal irregularity was more common in P3 (28%, n = 6) and P5 (14%, n = 3), p = 0.162. Cerebellar strokes were more frequent in P3–P5 aneurysms (total nine cases), though not statistically significant (p = 0.084) and all were clinically silent. Favorable outcomes (mRS ≤ 2) were most prevalent in P5 (28%, n = 6) and P3 (23%, n = 5) segments, with unfavorable outcomes clustered in P3 and P4 (each 9%, n = 2), nearing significance (p = 0.096) (Table 2).

Table 2.

Outcome variables, categorized by PICA segments.

				P1	P2	P3	P4	P5	p Value
				N (%)	N (%)	N (%)	N (%)	N (%)
Reconstructive vs deconstructive techniques	Reconstructive	Simple and assisted coiling		2 (9)	2 (9)	0 (0)	0 (0)	0 (0)	0.233
		Flow diversion		0 (0)	2 (9)	0 (0)	0 (0)	0 (0)	0.35
	Deconstructive	Coil sacrifice		0 (0)	1 (4)	5 (23)	1(4)	2 (9)	0.397
		Onyx embolization		0 (0)	0 (0)	1 (4)	1 (4)	3 (14)	0.397
Treating the intimal irregularity/tear	No			2 (9)	3 (14)	1 (4)	1 (4)	3 (14)	0.162
	Yes			0(0)	1 (4)	6 (28)	1 (4)	3 (14)
Ischemic stroke	No stroke			2 (9)	4 (19)	3 (14)	0 (0)	2 (9)	0.084
	Brainstem stroke			0 (0)	0 (0)	0 (0)	0(0)	0 (0)
	Cerebellar stroke			0 (0)	0(0)	4 (19)	2 (9)	3 (14)
Favorable vs unfavorable outcomes	mRS ≤ 2			1 (4)	3 (14)	5 (23)	0 (0)	6 (28)	0.096
	mRS > 2			1 (4)	1 (4)	2 (9)	2 (9)	0 (0)

PICA: posterior inferior cerebellar artery; mRS: modified Rankin Scale.

The P3 segment was anatomically subdivided into P3A and P3B based on the turning point of the PICA loop, which served as a practical landmark for treatment planning. All P3B aneurysms (n = 4), located beyond the loop, were treated with deconstructive methods, whereas P3A aneurysms (n = 3), located proximal to the loop's inflection, included both reconstructive (n = 1) and deconstructive (n = 2) strategies (p = 0.276). Cerebellar strokes occurred equally in both P3A and P3B subgroups (9.5% each), with no cases of brainstem infarction observed (p = 0.659). Favorable outcomes (mRS ≤ 2) were achieved in all P3A cases (14.3%), while two of the four P3B cases resulted in unfavorable outcomes (p = 0.350) (Table 3).

Table 3.

P3 is divided into two sub-segments with relevant outcome parameters.

Variables	P3A	P3B	p Value
Reconstructive	1 (4.8%)	0 (0.0%)	0.276
Deconstructive techniques	2 (9.5%)	4 (19.0%)
Treated intimal tear	2 (9.5%)	4 (19.0%)	0.066
No intimal tear treatment	1 (4.8%)	0 (0.0%)
Ruptured aneurysm	1 (4.8%)	0 (0.0%)	0.499
Unruptured aneurysm	2 (9.5%)	4 (19.0%)
Cerebellar stroke	2 (9.5%)	2 (9.5%)	0.659
Brainstem stroke	0 (0.0%)	0 (0.0%)
No stroke	1 (4.8%)	2 (9.5%)
Favorable outcome (mRS ≤ 2)	3 (14.3%)	2 (9.5%)	0.35
Unfavorable outcome (mRS > 2)	0 (0.0%)	2 (9.5%)

mRS: modified Rankin Scale.

Favorable outcomes (mRS ≤ 2) were achieved in 71% (n = 15), while 29% (n = 6) had unfavorable outcomes (Table 4). All males (24%, n = 5) had favorable outcomes, though the difference by sex was not statistically significant (p = 0.105). Smoking status did not significantly correlate with outcome (p = 0.282), nor did aneurysm rupture (p = 0.16). IVH at presentation, however, was significantly associated with worse outcomes (p = 0.043). No significant differences in mRS outcomes were found based on the PICA segment (p = 1.0), treatment of intimal irregularity (p = 0.89), or treatment method (reconstructive vs deconstructive, p = 1.0). Stroke occurrence was not significantly associated with mRS score (p = 0.676). Figure 1 shows illustrative cases of distal PICA treatment and Figure 2 presents an original illustration depicting the anatomy segments of PICA.

Table 4.

Comparing the outcomes based on modified Rankin score.

		Favorable vs unfavorable outcomes
		mRS ≤ 2	mRS > 2	p Value
		N (%)	N (%)
Sex	Female	10 (47)	6 (28)	0.105
	Male	5 (23)	0 (0)
Smoking history	Never smoker	10 (47)	3 (14)	0.282
	Prior smoker	2 (9)	0 (0)
	Smoker	3 (14)	2 (9)
	Unknown	0 (0)	1 (4)
Ruptured aneurysm status	Yes	11 (52)	6 (28)	0.16
Intraventricular hemorrhage at presentation	Yes	6 (28)	6 (28)	0.043
PICA segment	Proximal segment	10 (47)	4 (19)	1
	Distal segment	5 (23)	2 (9)
Treating the intimal tear	Yes	8 (38)	3 (14)	0.89
Treatment methods	Reconstructive	5 (23)	2 (9)	1
	Deconstructive	10 (47)	4 (19)
Stroke outcome	Yes	6 (28)	3 (14)	0.676

mRS: modified Rankin Scale; PICA: posterior inferior cerebellar artery.

Figure 1.

(a, b) Distal posterior inferior cerebellar artery (PICA) P3 segment in a patient in her 70s demonstrates a preserved, tortuous course medially and posteriorly after origin from the vertebral artery. Despite the presence of a large, unruptured aneurysm in the vicinity, the P3 segment maintains continuity without evidence of occlusion or compromise post-deployment of the flow diverter. The reconstructive approach has effectively redirected flow, sparing the eloquent distal territory of the PICA, which is crucial for medullary and cerebellar perfusion. In the second case, a patient in her 50s with a ruptured dissecting aneurysm involving the distal PICA, the P5 segment demonstrates a significant pre-treatment (c). Following deconstructive Onyx embolization (d), the P5 segment is no longer visualized, indicating successful occlusion of the parent vessel to achieve flow arrest to the ruptured pathology.

Figure 2.

Medical illustration depicting the segments of posterior inferior cerebellar artery (PICA) and its relation to the brainstem and cerebellum.

Discussion

In this retrospective study of 21 patients with dissecting aneurysms of the distal PICA, we explored the interplay between aneurysm location, endovascular treatment strategy, and clinical outcome, with a particular emphasis on preserving brainstem integrity in cases involving proximal PICA segments. The data illustrate several important relationships: (1) deconstructive techniques were more frequently employed in distal aneurysms, (2) reconstructive methods were largely confined to proximal segments, and (3) ischemic complications were most prevalent in P3B and beyond, often associated with liquid embolic or occlusive techniques in anatomically eloquent territories.

Reconstructive versus deconstructive techniques

Of the 21 patients, 67% underwent deconstructive treatment (e.g. coil sacrifice or Onyx embolization), while 33% received vessel sparing or reconstructive therapy (e.g. coiling with or without adjuncts with parent vessel sparing, stent-assisted coiling, or flow diversion). Deconstructive techniques were significantly more common in distal segments (P3B–P5), consistent with the expectation that more distal territories possess richer collateralization and fewer brainstem perforators. This segment-specific predilection has been previously reported in series by Lim et al., ² who recommended parent vessel occlusion in distal dissections to prevent rebleeding without unacceptable ischemic risk and was similarly endorsed in case-based experiences like that of Shintoku et al., where NBCA embolization was safely employed in distal PICA aneurysms. ¹¹

Conversely, reconstructive or parent vessel sparing strategies were almost exclusively used for aneurysms involving P1 and P2. These proximal segments harbor critical medullary perforators, and sacrifice in these regions has been linked to high rates of brainstem infarction and neurological morbidity, as shown in classic surgical series by Jafar et al. ³ and Nakatomi et al. ⁵ In our cohort, patients treated reconstructively had no incidence of brainstem infarct, further validating the necessity of preserving parent vessel patency in these regions.

Despite this anatomical logic, the difference in outcomes between reconstructive and deconstructive groups was not statistically significant (p = 1.0), likely due to appropriate treatment selection based on anatomical risk rather than treatment effect per se. However, this depicts the need for segment-driven strategy selection in PICA dissections rather than a one-size-fits-all approach.

Segmental anatomy and functional outcomes

The segmental distribution of aneurysms followed expected patterns, with 10 cases in proximal segments (P1–P3) and 11 in distal (P4–P5). Favorable outcomes (mRS ≤ 2) were more frequently achieved in patients with distal aneurysms (especially P5, 100% favorable), while worse outcomes clustered in P3B and P4. This trend approached but did not reach statistical significance (p = 0.096), reflecting the impact of ischemic complications in cerebellar and perforator-dependent territories.

This anatomic–outcome relationship has been described in prior series: Yamakawa et al. ¹ found that proximal PICA dissections often caused ischemia due to perforator involvement, while distal dissections more commonly presented with hemorrhage. Nussbaum et al., in their surgical series, observed that even visually normal segments of dissected PICA could harbor subclinical perforator origins and should be handled with caution. ⁷ Our findings mirror this distribution, with brainstem ischemia notably absent and cerebellar infarcts occurring exclusively in P3–P5 lesions.

Intimal irregularity and dissection morphology

Intimal irregularities or suspected underlying dissections adjacent to and not fully incorporated into the pseudoaneurysmal outpouching were identified in 43% of cases and were more frequently present in more distal segments (especially P3B). These findings were based on DSA and operator interpretation, paralleling the diagnostic approach used in prior studies such as Lim et al. ² and Madaelil et al., ¹² where fusiform dilation with irregular lumen morphology was the primary diagnostic marker for dissection.

Interestingly, the presence of intimal irregularity was not statistically associated with poor outcome (p = 0.89), but its segmental distribution overlapped substantially with the occurrence of ischemic infarcts (p = 0.084). This suggests that intimal disruption, while not independently predictive of outcome, may be a marker of higher procedural risk due to vessel fragility or complex flow dynamics. The concept of the “pseudo-lumen” persisting despite proximal clipping, leading to rebleeding or incomplete healing, is also supported by long-term follow-up in the case described by Nakatomi et al., ⁵ underscoring the importance of definitive exclusion of the dissected segment.

The P3 segment was subdivided into P3A and P3B based on the PICA loop inflection point, serving as a landmark for treatment planning (4). All P3B aneurysms (n = 4), located beyond the loop, were treated deconstructively, while P3A aneurysms (n = 3), located proximally, included both reconstructive (n = 1) and deconstructive (n = 2) approaches. Cerebellar strokes occurred equally (9.5%) in both subgroups and were clinically silent, with no brainstem infarctions observed. Favorable outcomes were seen in all P3A cases, while half of the P3B group had unfavorable outcomes. These findings support the PICA loop as a practical boundary for strategy selection, with distal (P3B) aneurysms being safer targets for parent vessel sacrifice due to reduced perforator risk.

Clinical implications for endovascular strategy

The major challenge in managing distal PICA dissections lies in balancing the need for durable aneurysm exclusion with the preservation of critical perfusion territories. While Onyx and NBCA embolization provide rapid and complete occlusion in many cases, their use must be tempered in the proximal trunk due to the risk of non-target embolization. In our series, Onyx was associated with a higher rate of cerebellar infarcts (60%) and was preferentially used in P3–P5 lesions. This echoes the experience of Shintoku et al. ¹¹ and Lee et al., ¹³ who both emphasized the importance of pre-embolization collateral assessment and the use of wedged microcatheter techniques to prevent reflux.

Flow diversion was utilized in only two cases (both in P2) and remains a technically feasible but cautiously applied strategy in PICA dissections pseudoaneurysms. Theoretical benefits include vessel preservation, aneurysm remodeling, and avoidance of perforator sacrifice. However, limitations include the small caliber and tortuosity of PICA rendering access to and navigation within potentially very challenging. In addition, the delivery microcatheter must be navigated sufficiently past the lesion to span the involved segment, requiring very distal access of a larger catheter and stiff Flow diverter (FD) system delivery. Lastly, the necessity for dual antiplatelet therapy in a potentially in the SAH setting, and the potential need for external ventricular drain management introduces the potential for hemorrhagic complications, as well as the uncertain efficacy of flow diversion to acutely treat a ruptured pseudoaneurysm. While early results in selected cases are promising, ⁸ broader application in the PICA remains controversial. Larger studies are needed to define its long-term safety and applicability across different PICA segments.

Overall, our findings support a segment-driven endovascular approach to distal PICA dissections. Reconstructive techniques should be prioritized in proximal segments (P1–P3A) to preserve critical perforators and minimize brainstem infarction risk. In contrast, deconstructive methods are generally safe and effective in distal segments (P3B–P5), where the risk is primarily cerebellar. While intimal irregularity often indicates vessel fragility and may influence procedural planning, it does not independently predict clinical outcome.

Limitations

This study is limited by its retrospective, single-center design and relatively small sample size, which may limit generalizability. Additionally, the diagnosis of dissecting aneurysm was based on qualitative angiographic features and operator judgment rather than histopathologic confirmation. As such, the subjective interpretation of DSA introduces potential classification bias, particularly in borderline or subtle cases. Standardized radiological criteria or vessel wall imaging could enhance diagnostic accuracy in future studies.

Conclusions

Dissecting pseudoaneurysms of the distal PICA require a tailored, segment-based endovascular approach. Proximal trunk aneurysms (P1–P3A) pose a high risk for brainstem infarction due to critical perforator supply and are best managed with reconstructive techniques when feasible. In contrast, deconstructive strategies are generally safe and effective in distal segments (P3B–P5). The presence of intimal irregularity may signal increased procedural complexity but is not associated with worse outcomes. Careful anatomical assessment remains essential to minimize ischemic complications and optimize functional recovery.