Microsurgical management of brainstem cavernous malformations: a 10-year single-surgeon experience

Abstract

Background

This study primarily reports a single-surgeon microsurgical experience emphasizing anatomical considerations and operative strategies.

Methods

We retrospectively reviewed 17 patients who underwent microsurgical resection of brainstem cavernous malformations at a single institution over ten years. Data included demographics, clinical presentation, lesion characteristics, surgical approach, and outcomes. Functional status was assessed using the modified Rankin Scale. The Lawton grading system was applied for preoperative stratification.

Results

Gross total resection was achieved in the majority of cases. Twelve of 17 patients (70.6%) developed no new neurological deficits, and most postoperative deficits were transient. At last follow-up, 10 patients (58.8%) achieved functional independence (modified Rankin Scale 0–1), while 5 (29.4%) remained mildly disabled. One patient experienced sustained worsening. Exploratory analyses showed no significant predictors of favorable outcome, reflecting the limitations of the small cohort. Long-term follow-up (≥ 12 months) was available in only six patients (35.3%).

Conclusions

Microsurgical resection of selected brainstem cavernous malformations is feasible with acceptable morbidity in experienced hands. However, given the small cohort size, retrospective design, and limited follow-up, these findings should be interpreted cautiously. Larger, multicenter studies are needed to validate prognostic factors and refine surgical indications.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12893-026-03703-8.

Introduction

Brainstem cavernous malformations (BSCMs) are rare, angiographically occult vascular lesions composed of clusters of thin-walled capillaries without intervening brain parenchyma [1]. They account for approximately 10–15% of all intracranial cavernous malformations but are of particular concern because of their eloquent location and the significant morbidity associated with symptomatic hemorrhage. Clinical manifestations include cranial nerve palsies, hemiparesis, sensory deficits, and ataxia, and repeated bleeding events can lead to cumulative neurological deterioration [2, 3].

Natural history studies suggest that once a BSCM bleeds, the risk of recurrence is substantial, with five-year rebleed rates reported above 30%. The combined risk of rehemorrhage or progressive neurological decline is even higher, emphasizing the importance of weighing the risks of intervention against conservative management [4].

Microsurgical resection remains the only definitive treatment option for BSCMs, but surgery carries considerable risk given the dense concentration of vital neural structures in the brainstem. Advances such as intraoperative neurophysiological monitoring, diffusion tensor imaging (DTI), and neuronavigation have improved surgical safety and precision [2, 5, 6]. Despite these developments, patient selection and the optimal timing of surgery remain subjects of ongoing debate.

To guide operative decision-making, grading systems such as the Lawton classification have been proposed, incorporating lesion size, location, midline extension, and presence of developmental venous anomalies [7]. While early studies suggest that these frameworks may help predict outcomes, further validation across different institutional settings is needed [8, 9].

Our aim is to contribute an institutional perspective to the growing body of literature on BSCM surgery and to highlight practical considerations in operative management, while acknowledging the limitations inherent to a small retrospective series.

Materials and methods

Study design and setting

We conducted a retrospective case series of patients with brainstem cavernous malformations treated surgically at our institution between January 2015 and December 2024. All operations were performed by the senior author (AY). The study protocol was reviewed and approved by the institutional review board and informed consent for surgery was obtained from all patients.

Eligibility criteria

Patients were eligible if they had a cavernous malformation located within the brainstem (midbrain, pons, or medulla), with or without limited extension into the thalamus. Lesions confined entirely to the thalamus or basal ganglia were excluded. Only patients who underwent microsurgical resection were included; conservatively managed cases were not part of this analysis.

Data collection

We extracted clinical, radiological, and surgical data from electronic medical records and operative notes. Variables included age, sex, presenting symptoms, neurological deficits, hemorrhage history, lesion size and location, axial and midline extension, presence of developmental venous anomalies (DVAs), surgical approach, use of intraoperative neurophysiological monitoring (IONM), and perioperative complications.

Hemorrhage definition and timing

Hemorrhagic events were defined as acute neurological symptoms associated with MRI evidence of recent hemorrhage [2]. Hemorrhages were classified as acute (≤ 3 weeks), subacute (3–8 weeks), or chronic (> 8 weeks) according to Porter et al. (1999) [10]. Surgical intervention was generally considered after at least one symptomatic hemorrhage, with preference for lesions that were acute, exophytic, or in contact with a pial surface.

Surgical technique

The surgical approach was tailored to lesion location and safe-entry zones. Standard microsurgical principles were followed, and neuronavigation and multimodal IONM were used in all cases. DVAs were preserved whenever present. Postoperative MRI was performed within 24 h to assess extent of resection.

Outcome assessment

Functional status was assessed using the modified Rankin Scale (mRS) preoperatively, at discharge, and at the most recent follow-up. Follow-up evaluations were conducted in outpatient clinic visits or during inpatient rehabilitation, depending on patient availability.

Statistical analysis

Given the small cohort, analyses were descriptive. Continuous variables are reported as means with ranges, and categorical variables as counts and percentages. Exploratory correlations between mRS outcomes and clinical/radiological variables were examined. A logistic regression model using age and Lawton grade as covariates was attempted but interpreted with caution given the limited sample size. Statistical analysis was performed using IBM SPSS Statistics version 28.0 (IBM Corp., Armonk, NY, USA).

Results

Patient characteristics

Seventeen patients underwent microsurgical resection of BSCMs during the study period. The mean age was 42.8 years (range 20–71), with 11 males and 6 females (M: F ratio 1.8:1).

Clinical presentation and preoperative deficits

The most frequent presenting symptom was body numbness (6 patients, 35%), followed by facial numbness (4, 24%), limb weakness (4, 24%), and headache (4, 24%). Preoperative neurological deficits were dominated by limb weakness (8 patients, 47%) and facial weakness (7, 41%). Ataxia and cranial nerve palsies were also common. Detailed symptoms and deficits are summarized in Table 1.

Table 1.

Clinical Presentation and Preoperative Deficits (n = 17)

Symptom / Deficit	n (%)
Body numbness	6 (35)
Limb weakness	8 (47)
Facial numbness	4 (24)
Facial weakness	7 (41)
Ataxia / unsteady gait	5 (29)
Cranial nerve palsy*	5 (29)
Headache	4 (24)
Other symptoms**	5 (29)

*Includes 3rd, 5th, or 6th nerve palsies

**Includes dizziness, visual disturbance, nystagmus, dysarthria

Lesion characteristics

Lesion size ranged from 9 to 30 mm (mean 16.4 mm). The pons was the most common location (9 patients, 53%), followed by the midbrain (4, 24%) and medulla (2, 12%). Two patients had multisegmental involvement. Midline extension was present in 9 patients (53%), and developmental venous anomalies were identified in 2 (12%). The average distance to the pial surface was 1.5 mm. These findings are summarized in Table 2.

Table 2.

Lesion Characteristics

Variable	Value
Mean size (mm, range)	16.4 (9–30)
Pons	9 (53)
Midbrain	4 (24)
Medulla	2 (12)
Multisegmental involvement	2 (12)
Midline crossing	9 (53)
Developmental venous anomaly	2 (12)
Mean distance to pial surface	1.5 mm (0–4.3)

Hemorrhage history and lawton grades

Twelve patients (71%) presented after a single hemorrhage, and 5 (29%) after two. At the time of surgery, 10 lesions (59%) were acute, 5 (29%) subacute, and 2 (12%) chronic. Lawton grades ranged from 0 to 7, with nearly half of patients (8, 47%) classified as low grade (0–2), 7 (41%) as intermediate (3–4), and 2 (12%) as high grade (5–7).

Surgical approaches and outcomes

The most common approach was suboccipital craniotomy (7 patients, 41%), followed by Kawase (3, 18%), subtemporal transtentorial (3, 18%), occipital interhemispheric (3, 18%), and far lateral (1, 6%). Gross total resection was achieved in all cases as confirmed on early postoperative MRI.

Postoperatively, 12 patients (71%) developed no new neurological deficits. Five patients (29%) experienced new or worsened deficits. To this study, “transient” deficits were defined as new or worsened neurological impairments resolving to baseline or improving to mRS ≤ 2 within 3 months postoperatively. Four of the five affected patients met this definition and demonstrated clinical improvement during early follow-up and rehabilitation. One patient had persistent neurological worsening (mRS 3 at last follow-up). Complications included one case of atrial fibrillation and one case of meningitis, both resolving without permanent sequelae (Table 3).

Table 3.

Surgical Approaches and Outcomes

Lesion Location	n (%)	Surgical Approach(es) Used	Gross Total Resection n (%)	mRS 0–1 at Last Follow-up n (%)	New/Worsened Deficit n (%)
Midbrain (n = 4)	4 (24%)	Subtemporal trans tentorial (3)	4 (100%)	3 (75%)	1 (25%)
		Occipital interhemispheric (1)
Pons (n = 9)	9 (53%)	Suboccipital (5)	9 (100%)	5 (56%)	3 (33%)
		Kawase (3)
		Subtemporal trans tentorial (1)
Medulla (n = 2)	2 (12%)	Suboccipital (1)	2 (100%)	1 (50%)	1 (50%)
		Far lateral (1)
Multisegmental (n = 2)	2 (12%)	Occipital interhemispheric (2)	2 (100%)	1 (50%)	0 (0%)
Total	17 (100%)	—	17 (100%)	10 (59%)	5 (29%)

At discharge, 8 patients (47%) had stable function compared to baseline and 6 (35%) showed early improvement. At last follow-up, 10 patients (59%) achieved functional independence (mRS 0–1), 5 (29%) had mild disability (mRS 2), and 1 patient (6%) experienced sustained worsening (mRS 3). The mean hospital stay was 9.4 days (range 1–19). The patient with sustained worsening (mRS 3) had a pontine lesion with midline extension requiring a suboccipital approach; postoperative imaging confirmed complete resection without hemorrhagic complication. The persistent deficit was characterized by residual hemiparesis that did not fully recover during the available follow-up period.

Follow-up and exploratory analysis

Mean follow-up duration was 14.5 months (range 2 weeks–42 months); however, only 6 patients (35%) were available for ≥ 12 months of follow-up, limiting interpretation of long-term outcomes. Exploratory correlations between age, lesion size, Lawton grade, and mRS outcomes showed no statistically significant associations, reflecting the limited cohort size.

Illustrative cases

Case 1 – lateral pontine cavernoma

A 44-year-old man presented with hemiparesis, facial weakness, and diplopia. MRI revealed a hemorrhagic pontine lesion with lateral extension and corticospinal tract displacement on DTI. A subtemporal transtentorial approach was chosen to avoid the floor of the fourth ventricle. Gross total resection was confirmed, and the patient recovered without new deficits (Fig. 1; Video 1). At 12-month follow-up, the patient had returned to independent daily activities with complete resolution of diplopia and only minimal residual facial weakness (mRS 1).

Fig. 1.

A+B: a right-sided posterior lateral pontine expansile hemorrhagic lesion suggestive of cavernoma. C: Right side CST is pushed medially by the lesion (red arrow)

Case 2 – pontomedullary cavernoma

A 22-year-old man with facial weakness, diplopia, and hemiparesis was found to have a pontomedullary cavernoma. Given the anterolateral location, a far-lateral approach was selected. The lesion was completely resected with preservation of critical neurovascular structures. The patient was discharged with only mild residual abducens palsy (Fig. 2; Video 2).

Fig. 2.

Left pontomedullary cavernoma. A: Axial Susceptibility-Weighted sequence. B: Sagittal usceptibility-Weighted Imaging. The cavernoma is in the left pontomedullary area (hypointense)

Discussion

Multiple large single-center series and meta-analyses have reported outcomes following microsurgical resection of BSCMs, including cohorts exceeding 100 patients [2, 10, 11]. In comparison, our study represents a smaller, contemporary single-surgeon experience over a defined 10-year period. While it does not aim to redefine outcome benchmarks, it provides a granular institutional perspective reflecting modern adjuncts such as routine DTI integration and multimodal intraoperative monitoring. Rather than offering novel statistical associations, our findings contribute descriptive data on approach selection, perioperative morbidity, and short-term functional outcomes within a consistent surgical philosophy.

Patient demographics, presentation and preoperative neurological deficits

The demographic profile in our cohort (mean age 43 years, slight male predominance) was consistent with prior studies [2, 11]. Clinical presentations mirrored known patterns, most commonly hemiparesis, sensory changes, and cranial neuropathies. These symptoms reflect disruption of corticospinal and brainstem sensory pathways and emphasize why even small hemorrhages in this region may cause disabling deficits.

Compared with cavernous malformations located in the cerebellum or other eloquent supratentorial regions, brainstem cavernous malformations (BSCMs) carry a distinct risk profile because of the compact arrangement of long tracts and cranial nerve nuclei within the brainstem. While cavernomas in eloquent supratentorial locations can also produce substantial morbidity, surgical series have shown that many non-brainstem eloquent lesions can be managed with acceptable long-term functional outcomes after microsurgical resection [12]. Similarly, cerebellar cavernous malformations often allow more favorable surgical trajectories and postoperative recovery profiles than lesions intrinsic to the brainstem [13]. In contrast, even small brainstem lesions may produce disproportionate neurological deficits because of the density of critical neuroanatomical structures, and the threshold for intervention is therefore more conservative. Our functional independence rate of 59% is broadly comparable to previously reported brainstem-specific series but remains lower than that reported in many non-brainstem cavernoma cohorts, underscoring the distinctive surgical challenges posed by BSCMs.

Most lesions were ≤ 20 mm and located in the pons, consistent with published series [3, 14]. More than half crossed the midline, a factor known to increase surgical complexity [7]. Most patients were operated after a single hemorrhage, reflecting our institutional practice of favoring early intervention in symptomatic cases. This approach is supported by reports that rebleeding substantially increases long-term disability [4, 10].

Selection of the surgical approach was guided by lesion location, pial or exophytic presentation, and safe-entry zones. Suboccipital and telovelar routes were most frequently used for posterior pontine and medullary lesions, whereas lateral and transtentorial approaches were reserved for anterolateral or superiorly placed lesions. In all cases, neuronavigation and intraoperative neurophysiological monitoring were essential adjuncts. Preservation of associated DVAs was achieved in every case, consistent with the principle that their sacrifice may lead to venous infarction [15].

Our perioperative outcomes were encouraging, with gross total resection in all patients and the majority avoiding new permanent deficits. Nevertheless, five patients developed new or worsened deficits postoperatively, underscoring the inherent risk of operating in the brainstem. Most deficits improved with rehabilitation, but one patient had sustained worsening at follow-up. Functional independence (mRS 0–1) was achieved in 59% of patients, comparable to other institutional series [2, 9]. Importantly, only 35% of our cohort had ≥ 12 months of follow-up, limiting conclusions about long-term outcomes.

We attempted exploratory analysis of potential predictors of outcome, including age and Lawton grade, but no statistically significant associations were identified. Given the very small sample, these results should not be over-interpreted. Instead, they reinforce that individual institutional experiences are primarily descriptive.

Alternative management strategies, including stereotactic radiosurgery (Gamma Knife), have been proposed for select brainstem cavernous malformations (BSCMs), particularly in surgically inaccessible locations or in patients considered high risk for open resection. Meta-analyses and systematic reviews suggest that stereotactic radiosurgery may reduce post-treatment hemorrhage rates in patients with deep-seated cavernous malformations, including those located in the brainstem, although the evidence is largely derived from retrospective series and heterogeneous observational cohorts [16, 17]. Moreover, while pooled analyses indicate a reduction in hemorrhage rates following radiosurgery, concerns remain regarding radiation-induced adverse effects and delayed neurological injury, even though the reported incidence of permanent radiation injury is relatively low [17, 18]. Consequently, microsurgical resection remains the only treatment capable of immediate lesion removal, whereas radiosurgery may be considered in carefully selected cases in which the anticipated surgical risk is prohibitive.

Strengths and limitations

The strengths of this study include detailed characterization of clinical, radiological, and surgical variables and the use of standardized outcome measures. However, the limitations are substantial: retrospective design, small sample size, single-surgeon experience, and incomplete follow-up. These factors limit generalizability and preclude strong statistical conclusions.

In summary, microsurgical resection of brainstem cavernous malformations can be performed with acceptable morbidity in carefully selected patients, particularly when guided by modern imaging and intraoperative adjuncts. Our experience highlights the importance of individualized surgical planning and the role of safe-entry zones. Larger, multicenter studies with long-term follow-up are needed to define optimal management strategies and validate grading systems for risk stratification. Accordingly, this experience should be viewed as descriptive and hypothesis-generating rather than confirmatory.

Conclusion

Microsurgical resection of selected brainstem cavernous malformations can be performed with acceptable morbidity. Our 10-year institutional experience demonstrates the importance of individualized surgical planning, adherence to safe-entry zones, and preservation of developmental venous anomalies. However, the small cohort size, retrospective design, and incomplete long-term follow-up limit the generalizability of our findings. These results should therefore be regarded as descriptive and hypothesis-generating. Larger, prospective multicenter studies are needed to refine surgical indications, validate grading systems, and establish long-term outcome benchmarks for this challenging pathology.

Abbreviations

BSCMs

Brainstem cavernous malformations

mRS

Modified rankin scale

DTI

Diffusion tensor imaging

DVA

Developmental venous anomaly

IONM

Intraoperative neurophysiological monitoring

Authors’ contributions

All authors contributed to study conception and design. Data collection and surgical procedures were performed by AY. AM analyzed the data and drafted the manuscript. All authors critically revised the manuscript and approved the final version.

Funding

No external funding was received for this study.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study was conducted in accordance with the Declaration of Helsinki. Ethical approval was obtained from the Medical Research Committee (MRC), Hamad Medical Corporation, Doha, Qatar (Approval No: MRC-01-25-1137). The requirement for informed consent to participate was formally waived by the MRC due to the retrospective design of the study and the use of anonymized data.

Consent for publication

Written informed consent for publication of clinical data and any potentially identifiable images was obtained from all patients or their legal guardians.

Competing interests

The authors declare no competing interests.