De novo epilepsy after the occipital transtentorial approach: association with occipital lobe injury and persistent FLAIR lesions

Abstract

Background

The occipital transtentorial approach (OTA) is a useful surgical corridor for deep-seated lesions involving the pineal region, thalamus, ventral temporal structures, and posterior fossa. Although occipital lobe injury is a recognized complication of OTA, its relationship with postoperative epilepsy has not been systematically investigated. We examined the association between postoperative occipital lobe injury and de novo epilepsy following OTA.

Methods

Thirty-two consecutive OTA procedures performed in 31 patients between 2013 and 2025 were retrospectively reviewed. Clinical records, operative videos, and magnetic resonance imaging studies were analyzed. Postoperative occipital injury was assessed using the final follow-up fluid-attenuated inversion recovery (FLAIR) MRI. Patients were classified according to postoperative FLAIR lesion volume (0 mL, < 2 mL, and ≥ 2 mL). Factors associated with de novo epilepsy and postoperative FLAIR lesions ≥ 2 mL were evaluated.

Results

De novo postoperative epilepsy developed after 4 of 32 procedures (12.5%). All patients with epilepsy demonstrated intraoperative occipital cortical contusion and postoperative FLAIR lesions ≥ 2 mL. Both factors were significantly associated with epilepsy (p = 0.0058 and p = 0.0010, respectively). Factors associated with postoperative FLAIR lesions ≥ 2 mL included compression of basal occipital veins (p = 0.0044), sacrifice of bridging veins (p = 0.0423), and intraoperative occipital cortical contusion (p = 0.000036). Notably, all FLAIR lesions ≥ 2 mL occurred in procedures complicated by approach-related occipital lobe injury. No patient without a FLAIR lesion ≥ 2 mL developed postoperative epilepsy.

Conclusions

Occipital cortical contusion and persistent postoperative occipital FLAIR lesions ≥ 2 mL were strongly associated with de novo epilepsy following OTA. Venous compromise, including compression of basal occipital veins and sacrifice of bridging veins, was associated with occipital lobe injury. These findings suggest that meticulous preservation of venous structures and avoidance of occipital lobe injury are important surgical principles for reducing postoperative epilepsy after OTA.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1007/s00701-026-07003-1.

Introduction

Postoperative epilepsy is one of the most consequential complications following neurosurgical procedures [3, 16]. Although neurological deficits such as motor/language impairment, visual disturbance, and cognitive dysfunction are generally recognized as major surgical complications, the incidence and clinical impact of epilepsy caused by approach-related brain injury have received comparatively little attention. Even when seizures are adequately controlled with antiseizure medications, the diagnosis of epilepsy may impose substantial restrictions on daily life, including long-term medication use, driving limitations, and occupational constraints [4, 17]. Therefore, minimizing approach-related brain injury remains an important goal in neurosurgical practice.

The contribution of surgical approach–related injury to postoperative epilepsy is often difficult to evaluate because many intracranial lesions themselves possess epileptogenic potential. For example, seizures are common in patients with supratentorial gliomas, making it challenging to distinguish epilepsy caused by the underlying disease from that caused by surgical manipulation. Consequently, the specific role of approach-related cortical injury in the development of postoperative epilepsy remains poorly understood.

The occipital transtentorial approach (OTA) is a well-established surgical corridor for lesions located in the pineal region [13], midbrain [5], thalamus [10], ventral temporal structures including the parahippocampal and fusiform gyri [11], and posterior fossa [7, 14]. A fundamental principle of OTA is to obtain a sufficient operative corridor while minimizing injury to the occipital lobe. Depending on the location of the target lesion, OTA may be performed either through the interhemispheric fissure [1] or by superior retraction of the basal occipital surface, as originally described by Poppen [8]. In either technique, adequate mobility of the occipital lobe is essential for safe surgical exposure.

Importantly, many lesions commonly treated through OTA—including posterior fossa lesions, such as hemangioblastomas, pilocytic astrocytomas, and arteriovenous malformations, as well as pineal region tumors including germ cell tumors—rarely cause epilepsy by themselves [2]. Therefore, the occurrence of de novo epilepsy after OTA provides a unique clinical setting in which the contribution of approach-related occipital lobe injury can be evaluated with relatively limited confounding from the intrinsic epileptogenicity of the underlying lesion.

Several anatomical and technical factors may contribute to occipital lobe injury during OTA. Adequate cerebrospinal fluid drainage is crucial for reducing brain volume and minimizing occipital lobe retraction. However, retraction of the occipital lobe may still be required. The steep tentorial angle may limit the superior operative corridor toward the pineal or tectal region [6]. Previous studies have suggested that bridging veins near the torcular Herophili are relatively uncommon and rarely constitute a major obstacle during OTA [7, 11, 13]. However, when present, such veins may restrict occipital lobe mobility, resulting in increased focal retraction and a narrower operative corridor. Compression or sacrifice of occipital venous structures may also compromise regional venous outflow and contribute to local ischemic injury. Such occult injuries may subsequently appear as hyperintense lesions on postoperative fluid-attenuated inversion recovery (FLAIR) magnetic resonance imaging.

Traditionally, complications of OTA have been discussed primarily in terms of postoperative hemorrhage and visual field deficits [6]. The OTA may affect the occipital lobe during resective surgery, and seizures arising from the occipital lobe often propagate to the mesial temporal lobe, which can cause focal impaired awareness seizures [15]. However, the relationship between postoperative occipital lobe injury and the development of de novo epilepsy has not been systematically investigated. Furthermore, factors associated with the development of postoperative occipital FLAIR lesions remain poorly understood.

In the present study, we examined the association between postoperative occipital FLAIR lesion volume and de novo epilepsy following OTA. We also explored anatomical and surgical factors associated with the development of occipital lobe injury after OTA.

Methods

Patient population

Among 68 OTA procedures performed by the senior author (T.K.), 32 consecutive OTA procedures performed in 31 patients at Kitasato University Hospital between April 2013 and July 2025 were included in the present study. Procedures performed at previous institutions were excluded because complete clinical records, operative videos, magnetic resonance imaging (MRI) studies, electroencephalographic data when available, and long-term follow-up imaging were not uniformly available for detailed retrospective review. Patient demographics, extent of resection, lesion location, histopathological diagnosis, operation time, tentorial angle [9], cerebrospinal fluid (CSF) drainage via ventricular puncture, excessive occipital lobe retraction, compression of basal occipital veins, the presence of bridging veins within 3 cm of the torcular Herophili (Fig. 1), sacrifice of bridging veins (Fig. 1), intraoperative occipital cortical contusion (Fig. 2), postoperative occipital FLAIR lesions, de novo postoperative epilepsy, and follow-up duration were analyzed.

Fig. 1.

Representative case demonstrating the combined influence of a steep tentorial angle and complex bridging venous anatomy on occipital lobe mobilization (Case 27). A 32-year-old male with a tectal ganglioglioma underwent partial resection through a right occipital transtentorial approach. Cerebrospinal fluid drainage through ventricular puncture was successfully achieved. The operation time was 8 h 5 min. The patient subsequently developed a persistent postoperative occipital FLAIR lesion (≥ 2 mL) and de novo epilepsy during follow-up. A Preoperative axial FLAIR image demonstrating a lesion involving the tectal-pineal region. B Immediate postoperative axial FLAIR image demonstrating partial tumor removal and a newly developed hyperintense lesion in the right occipital lobe (circle). C Follow-up axial FLAIR image demonstrating a persistent occipital hyperintense lesion (circle). D Measurement of the tentorial angle (69.6°) on a midsagittal T2-weighted image. The steep tentorial angle substantially limited the superior operative corridor toward the tectal region. E Intraoperative photograph obtained after dural opening. The solid arrow, dashed arrow, and arrowhead indicate bridging veins located near the confluence of sinuses (torcular Herophili) and draining the occipital lobe. The combination of a steep tentorial angle and complex bridging venous anatomy markedly restricted occipital lobe mobilization and narrowed the operative corridor. F Intraoperative photograph obtained after partial tumor removal. The solid arrow and arrowhead indicate the same bridging veins shown in panel E. Because adequate surgical exposure could not be achieved through the limited operative corridor, the small bridging vein indicated by the dashed arrow in panel E was sacrificed

Fig. 2.

Representative case demonstrating occipital cortical blanching associated with compression of a basal occipital vein during OTA (Case 31). A 73-year-old man with a hemangioblastoma arising from the superior surface of the cerebellar hemisphere underwent gross-total resection through a left occipital transtentorial approach. The operation time was 4 h 21 min. Cerebrospinal fluid drainage through ventricular puncture was successfully achieved, and no bridging vein was encountered or sacrificed. Although marked occipital lobe retraction was required temporarily, compression of a basal occipital vein was observed during the procedure. The patient subsequently developed a persistent postoperative occipital FLAIR lesion (≥ 2 mL) and de novo epilepsy during follow-up. A Preoperative contrast-enhanced T1-weighted magnetic resonance image demonstrating a cerebellar tumor with marked enhancement. B Immediate postoperative contrast-enhanced T1-weighted image confirming gross-total tumor removal. C Axial FLAIR image obtained 15 months after surgery demonstrating a persistent hyperintense lesion in the left occipital lobe (circle). D Measurement of the tentorial angle (56.7°) on a midsagittal magnetic resonance image. E Intraoperative photograph obtained after gross-total resection of the tumor. Arrowheads indicate focal cortical blanching surrounding a basal occipital vein (arrow), suggestive of impaired venous drainage caused by vein compression during occipital lobe retraction

This study was approved by the Institutional Review Board of Kitasato University School of Medicine (approval No. G23-08, “Clinicopathological and Molecular Significance of Brain Tumor Surgery Cases at Kitasato University: A Single-Center Retrospective Study and Prospective Database Construction”). The study was conducted in accordance with the principles of the Declaration of Helsinki.

Surgical technique

At our institution, OTA is routinely used not only for pineal and posterior fossa lesions but also for selected lesions involving the posterior parahippocampal and fusiform gyri, where it provides a favorable superior operative trajectory while minimizing temporal lobe manipulation [11]. All procedures were performed with the patient in the prone position. OTA was performed through a unilateral occipital craniotomy to expose the confluence of sinuses, transverse sinus, and superior sagittal sinus. Before dural opening, ventricular puncture through the occipital horn was routinely performed to achieve adequate cerebrospinal fluid drainage and brain relaxation [5].

The approach was performed according to the principles of Poppen’s occipital transtentorial approach [8]. After gentle superior elevation of the basal occipital surface, the quadrigeminal cistern was opened to allow further CSF release. The tentorium was then incised approximately 1.5 cm lateral to the straight sinus and extended toward the free tentorial edge. When necessary, the tentorial incision was extended toward the transverse sinus to obtain adequate exposure of posterior fossa lesions.

When bridging veins were encountered within the operative corridor, the surrounding arachnoid membrane was dissected to maximize venous mobility, and the veins were protected using oxidized cellulose and fibrin glue whenever possible.

Radiological and intraoperative assessments

Tentorial angle was defined as the angle between the tentorium and a line drawn along the hard palate on midsagittal magnetic resonance imaging and was measured for each patient in the perioperative period [9] (Fig. 1, 2).

Excessive occipital lobe retraction was defined as temporary or sustained bending of the occipital cortex exceeding approximately 60 degrees at the outer edge of the brain retractor during surgery, as determined by retrospective review of operative videos.

Intraoperative occipital cortical contusion was defined as visible cortical blanching, discoloration, or apparent mechanical contusion of the occipital cortex observed under the operating microscope (Fig. 2).

Follow-up duration was defined as the interval between surgery and the final MRI examination.

Postoperative occipital injury was assessed using the final available follow-up MRI rather than immediate postoperative imaging in order to evaluate persistent occipital lobe damage. Hyperintense lesions located within the approach-side occipital lobe on FLAIR imaging were identified and their volumes measured using OsiriX MD software (Pixmeo, Switzerland). Lesion volumes were obtained by manual slice-by-slice segmentation of persistent postoperative FLAIR hyperintensities on DICOM images, and total lesion volumes were automatically calculated from the summed segmented regions. All measurements were performed by a single investigator (T.K.) using a predefined measurement protocol. Patients were classified into three groups according to postoperative FLAIR lesion volume: 0 mL, < 2 mL, and ≥ 2 mL. The threshold of 2 mL was predetermined before analysis of the relationship between FLAIR lesion volume and postoperative epilepsy. Because no established threshold exists for postoperative occipital FLAIR lesions, a cutoff value of 2 mL was prospectively selected to distinguish focal postoperative signal changes from more substantial structural injury.

Definition of postoperative epilepsy

De novo postoperative epilepsy was diagnosed in patients without a preoperative history of epilepsy who developed postoperative epileptic seizures or clinical findings strongly suggestive of non-convulsive seizure activity after surgery, and whose events were considered attributable to approach-related occipital lobe injury. Acute symptomatic seizures occurring during the early postoperative period were not classified as postoperative epilepsy. The diagnosis was based on the overall clinical course, including seizure semiology, prolonged impaired consciousness suggestive of non-convulsive seizures, electroencephalographic findings when available, neuroimaging findings supporting epileptic activity, response to antiseizure medication, and long-term clinical follow-up. Both focal seizures and focal-to-bilateral tonic–clonic seizures were included. A single seizure occurring during the early postoperative period in the setting of acute structural brain injury was classified as an acute symptomatic seizure rather than de novo postoperative epilepsy.

Patients who developed seizures clearly attributable to tumor-related epileptogenic foci outside the occipital lobe were not classified as having OTA-related postoperative epilepsy. Specifically, two patients with supratentorial temporal lobe lesions (Supplementary Table 1, Cases 1 and 2) developed seizures during follow-up; however, these events were excluded because electroencephalography demonstrated seizure onset arising from the temporal lobe rather than the occipital lobe.

Detailed clinical characteristics of all epilepsy cases are provided in Supplementary Table 2.

Statistical analysis

Associations between clinical variables and de novo postoperative epilepsy were evaluated using Fisher’s exact test for categorical variables and the Mann–Whitney U test for continuous variables.

Factors associated with postoperative occipital FLAIR lesions (≥ 2 mL versus < 2 mL or no lesion) were analyzed using the same statistical methods.

Because of the limited sample size and non-normal distribution of the data, nonparametric statistical methods were used throughout the study. All statistical tests were two-sided, and a p value < 0.05 was considered statistically significant.

Given the exploratory nature of this study and the limited sample size, no adjustment for multiple comparisons was performed.

Results

Patient characteristics

The median follow-up duration was 85.2 months (range, 0.6–143.5 months). Patient characteristics are summarized in Table 1 and Supplementary Table 1.

Table 1.

Patient characteristics

Variable	Value
Number of patients	31
Number of OTA procedures	32
Age, mean ± SD, years	38.3 ± 23.6
Age, median (range), years	31.5 (7–75)
Sex distribution (31 patients)	20 males, 11 females
Extent of resection
Partial Removal	3 (9.4%)
Subtotal Removal	8 (25.0%)
Total Removal	21 (65.6%)
Lesion location
Cerebellum	15 (46.9%)
Midbrain	7 (21.9%)
Pineal	4 (12.5%)
Thalamus	3 (9.4%)
Parahippocampal gyrus	2 (6.2%)
Tentorial edge	1 (3.1%)
Pathology
Hemangioblastoma	7 (21.9%)
Rosette-forming glioneuronal tumor	3 (9.4%)
Pilocytic astrocytoma	2 (6.2%)
Diffuse astrocytoma	2 (6.2%)
Others	18 (56.2%)
Operation time, mean ± SD	7 h 20 min ± 2.76 h
Operation time, median (range)	6 h 36 min (4 h 21 min–15 h 27 min)
Tentorial angle, mean ± SD, degrees	56.2 ± 6.8
Tentorial angle, median (range), degrees	56.2 (44.7–73.1)
CSF drainage via ventricular puncture	31 (96.9%)
Strong occipital lobe retraction	14 (43.8%)
Occipital basal vein compression	5 (15.6%)
Bridging vein within 3 cm of torcular	5 (15.6%)
Bridging vein sacrifice	2 (6.2%)
Intraoperative occipital cortical contusion	10 (31.2%)
Postoperative occipital FLAIR lesion
0 mL	14 (43.8%)
< 2 mL	11 (34.4%)
≥ 2 mL	7 (21.9%)
De novo postoperative epilepsy	4 (12.5%)
Follow-up duration, median (range), months	85.2 (0.6–143.5)

Baseline demographic, radiological, and operative characteristics of the 32 occipital transtentorial approach (OTA) procedures performed in 31 patients. Data are presented as number (%) unless otherwise indicated. Continuous variables are presented as mean ± standard deviation (SD) or median with range, as appropriate. Tentorial angle was defined as the angle between the tentorium and a line drawn along the hard palate on midsagittal magnetic resonance imaging and was measured for each patient in the perioperative period [6]. Follow-up duration was defined as the interval between surgery and the final magnetic resonance imaging examination. FLAIR fluid-attenuated inversion recovery, CSF cerebrospinal fluid

Posterior fossa lesions accounted for nearly half of all procedures (15/32, 46.9%), including seven hemangioblastomas. Lesions were also located in the midbrain (7 cases), pineal region (4 cases), thalamus (3 cases), parahippocampal gyrus (2 cases), and tentorial edge (1 case).

The median operation time was 6 h 36 min (range, 4 h 21 min–15 h 27 min). The median tentorial angle was 56.2° (range, 44.7°–73.1°).

Intraoperative occipital cortical contusion was identified in 10 procedures (31.2%). The measured postoperative occipital FLAIR lesion volume for each procedure is provided in Supplementary Table 1.　Postoperative occipital FLAIR lesions were classified as 0 mL in 14 procedures, < 2 mL in 11 procedures, and ≥ 2 mL in 7 procedures (Supplementary Fig. 1). De novo postoperative epilepsy developed after 4 of 32 procedures (12.5%). Detailed clinical characteristics of these patients are summarized in Supplementary Table 2.

Factors associated with de novo postoperative epilepsy

Comparisons between procedures with and without de novo epilepsy are summarized in Table 2.

Table 2.

Factors associated with de novo postoperative epilepsy

Variable	Epilepsy (+), n = 4	Epilepsy (−), n = 28	p value
Lesion location: cerebellum	2/4	13/28	1.0000
Operation time, median (IQR)	7 h 42 min (6 h 35 min–9 h 03 min)	6 h 22 min (5 h 20 min–8 h 23 min)	0.6691
Tentorial angle, median (IQR), degrees	55.5 (52.2–59.9)	56.2 (51.9–58.3)	1.0000
CSF Drainage via Ventricular Puncture	3/4	28/28	0.1250
Strong occipital lobe retraction	3/4	11/28	0.2951
Occipital basal vein compression	2/4	3/28	0.1053
Bridging vein within 3 cm of torcular	2/4	3/28	0.1053
Bridging vein sacrifice	1/4	1/28	0.2379
Intraoperative occipital cortical contusion	4/4	6/28	0.0058
Postoperative FLAIR lesion ≥ 2 mL	4/4	3/28	0.0010

Comparison of clinical, radiological, and operative variables between procedures with and without de novo postoperative epilepsy following the occipital transtentorial approach. Categorical variables were analyzed using Fisher’s exact test and continuous variables using the Mann–Whitney U test. Statistically significant p values are shown in bold. IQR interquartile range, FLAIR fluid-attenuated inversion recovery, CSF cerebrospinal fluid

All four procedures complicated by de novo epilepsy demonstrated intraoperative occipital cortical contusion and postoperative FLAIR lesions ≥ 2 mL (Fig. 3). Both factors were significantly associated with postoperative epilepsy (p = 0.0058 and p = 0.0010, respectively).

Fig. 3.

Association between postoperative occipital FLAIR lesion volume and de novo postoperative epilepsy. The incidence of de novo postoperative epilepsy according to postoperative occipital FLAIR lesion volume on the final follow-up MRI examination. No patient without a postoperative FLAIR lesion or with a lesion volume < 2 mL developed epilepsy. In contrast, de novo epilepsy occurred in 4 of 7 procedures (57.1%) with postoperative FLAIR lesions ≥ 2 mL. FLAIR, fluid-attenuated inversion recovery

No significant associations were observed for lesion location, operation time, tentorial angle, CSF drainage via ventricular puncture, excessive occipital lobe retraction, compression of basal occipital veins, the presence of bridging veins within 3 cm of the torcular Herophili, or sacrifice of bridging veins.

Factors associated with postoperative FLAIR lesions ≥ 2 mL

Factors associated with postoperative occipital FLAIR lesions ≥ 2 mL are summarized in Table 3.

Table 3.

Factors associated with postoperative occipital FLAIR lesions ≥ 2 mL

Variable	FLAIR ≥ 2 mL, n = 7	FLAIR < 2 mL/none, n = 25	p value
Lesion location: cerebellum	3/7	12/25	1.0000
Operation time, median (IQR)	8 h 05 min (6 h 07 min–11 h 14 min)	6 h 12 min (5 h 24 min–7 h 51 min)	0.3158
Tentorial angle, median (IQR), degrees	56.6 (55.4–62.1)	55.6 (51.7–57.9)	0.4938
CSF Drainage via Ventricular Puncture	6/7	25/25	0.2188
Strong occipital lobe retraction	4/7	10/25	0.6691
Occipital basal vein compression	4/7	1/25	0.0044
Bridging vein within 3 cm of torcular	3/7	2/25	0.0566
Bridging vein sacrifice	2/7	0/25	0.0423
Intraoperative occipital cortical contusion	7/7	3/25	0.000036

Comparison of clinical, radiological, and operative variables between procedures with postoperative occipital FLAIR lesions ≥ 2 mL and those with FLAIR lesions < 2 mL or no lesion. Postoperative occipital injury was assessed on the final follow-up magnetic resonance imaging examination. Categorical variables were analyzed using Fisher’s exact test and continuous variables using the Mann–Whitney U test. Statistically significant p values are shown in bold. IQR interquartile range, FLAIR fluid-attenuated inversion recovery, CSF cerebrospinal fluid

Compression of basal occipital veins was significantly associated with the development of FLAIR lesions ≥ 2 mL (4/7 vs. 1/25, p = 0.0044). Intraoperative occipital cortical contusion was also strongly associated with FLAIR lesions ≥ 2 mL (7/7 vs. 3/25, p = 0.000036).

Although sacrifice of bridging veins was observed in only two procedures, both developed FLAIR lesions ≥ 2 mL, resulting in a significant association (2/7 vs. 0/25, p = 0.0423).

No significant associations were observed for lesion location, operation time, tentorial angle, CSF drainage via ventricular puncture, excessive occipital lobe retraction, or the presence of bridging veins within 3 cm of the torcular Herophili.

Notably, all procedures with intraoperative occipital cortical contusion demonstrated postoperative FLAIR abnormalities. Among these, seven procedures developed FLAIR lesions ≥ 2 mL, whereas no FLAIR lesion ≥ 2 mL was observed in procedures without intraoperative cortical contusion (Fig. 4).

Fig. 4.

Relationship between intraoperative occipital cortical contusion and postoperative occipital FLAIR lesion volume. Distribution of postoperative occipital FLAIR lesion volumes according to the presence or absence of intraoperative occipital cortical contusion. All procedures with postoperative FLAIR lesions ≥ 2 mL were associated with intraoperative cortical contusion, whereas no procedure without cortical contusion developed a FLAIR lesion ≥ 2 mL. FLAIR, fluid-attenuated inversion recovery. These findings suggest that intraoperative occipital cortical contusion represents a major determinant of persistent postoperative occipital lobe injury

Discussion

De novo postoperative epilepsy developed after 4 of 32 procedures (12.5%). The diagnosis of postoperative epilepsy was supported by long-term clinical follow-up, seizure semiology, electroencephalographic findings when available, and favorable responses to antiseizure medications in all patients. In one patient, prolonged postoperative impaired consciousness associated with focal hyperperfusion in the approach-side occipital lobe on arterial spin labeling MRI, strongly suggestive of non-convulsive seizure activity despite the absence of electroencephalographic confirmation, was classified as an early postoperative acute symptomatic seizure rather than de novo postoperative epilepsy according to the current ILAE concepts and was therefore not included in the de novo postoperative epilepsy cohort.

To our knowledge, the incidence of de novo epilepsy after OTA has not previously been reported. The present study demonstrated a significant association between postoperative occipital lobe injury and de novo epilepsy following the OTA. All patients who developed de novo epilepsy exhibited both intraoperative occipital cortical contusion and postoperative occipital FLAIR lesions ≥ 2 mL. Furthermore, factors associated with postoperative occipital injury included compression of basal occipital veins, sacrifice of bridging veins, and intraoperative cortical contusion. These findings suggest a sequential relationship in which surgical manipulation causes occipital lobe injury, resulting in persistent structural damage visible on FLAIR MRI and ultimately leading to postoperative epilepsy.

Postoperative epilepsy has traditionally received less attention than major neurological complications such as hemorrhage, infarction, visual field deficits, or motor/language dysfunction. However, epilepsy may have a substantial impact on long-term quality of life. Even when seizures are adequately controlled, patients may require lifelong antiseizure medication, face restrictions on driving, and experience limitations in employment and social activities. Several studies have demonstrated that epilepsy is associated with increased rates of unemployment, early retirement, and occupational disadvantage, even among patients with otherwise preserved neurological function. Frequent seizures have been identified as a major predictor of loss of employment, whereas seizure freedom is associated with successful job retention [4, 17]. Therefore, postoperative epilepsy should not be regarded as a minor complication. The occurrence of de novo epilepsy after surgery for otherwise non-epileptogenic lesions may be particularly detrimental because the complication arises not from the disease itself but from treatment intended to improve the patient’s condition.

A unique feature of the present study is that many lesions treated through OTA rarely cause epilepsy by themselves. Posterior fossa and pineal region lesions, including hemangioblastomas, pilocytic astrocytomas, germ cell tumors, and arteriovenous malformations, are generally not considered intrinsically epileptogenic. Although cerebellar epilepsy has been described [2], it is a rare entity that typically presents during infancy and is most commonly associated with developmental or low-grade cerebellar tumors such as gangliogliomas, gangliocytomas, ganglioneurocytomas, hamartomas, and pilocytic astrocytomas. Seizure onset after early childhood is uncommon. Consequently, the occurrence of de novo epilepsy after OTA provides a relatively pure model for evaluating the contribution of approach-related brain injury to postoperative seizure development. In contrast to studies involving supratentorial gliomas or cortical lesions, the confounding influence of lesion-related epileptogenicity is relatively limited in the present cohort.

Our findings indicate that persistent occipital FLAIR lesions are strongly associated with postoperative epilepsy. Importantly, all patients who developed epilepsy demonstrated FLAIR lesions ≥ 2 mL, whereas no patient without a substantial FLAIR lesion developed epilepsy. These observations suggest that postoperative FLAIR abnormalities are not merely radiological findings but may represent clinically meaningful structural injury with epileptogenic potential. Although the precise pathological substrate underlying these lesions remains uncertain, venous congestion, ischemic injury, cortical contusion, and subsequent gliosis may contribute to epileptogenesis.

In the present series, occipital cortical contusion appeared to represent the final common pathway leading to both persistent postoperative FLAIR abnormalities and de novo epilepsy. As illustrated in Fig. 2, intraoperative cortical blanching surrounding a compressed basal occipital vein provides visual evidence that venous compromise may contribute to focal occipital cortical injury during OTA. All patients who developed epilepsy exhibited intraoperative cortical contusion, and all procedures with FLAIR lesions ≥ 2 mL demonstrated evidence of cortical injury. These findings suggest that venous compromise alone may not be sufficient to produce clinically significant sequelae unless it ultimately results in structural cortical damage. From a practical standpoint, occipital cortical contusion may serve as an intraoperative surrogate marker of excessive local retraction or venous insufficiency and should be regarded as a warning sign for subsequent postoperative morbidity.

Previous studies have suggested that a steep tentorial angle may increase the risk of occipital lobe injury during the Poppen approach. Li et al. reported that larger tentorial angles were significantly associated with postoperative occipital lobe damage in patients undergoing surgery for tentorial or pineal region tumors and proposed that a steep tentorium may increase the difficulty of cerebrospinal fluid release and occipital lobe mobilization during the approach [6]. However, tentorial angle was not significantly associated with postoperative FLAIR lesions or de novo epilepsy in the present series. This discrepancy may reflect differences in lesion location. Tentorial alignment is likely most relevant for lesions located in the pineal region or in the rostral portion of midbrain, where the operative corridor is directly influenced by the slope of the tentorium. In contrast, for more laterally located lesions, such as tumors arising from the superior surface of the cerebellar hemisphere, the tentorium becomes progressively flatter and the influence of tentorial angle may be less pronounced. Furthermore, even a favorable tentorial angle may not guarantee adequate occipital lobe mobility when bridging veins are present near the torcular Herophili. In one representative patient with a tectal ganglioglioma and a steep tentorial angle of 69.6°, a bridging vein near the torcular significantly restricted occipital lobe mobilization, resulting in a technically demanding procedure complicated by postoperative occipital lobe injury (Fig. 1). These observations suggest that venous anatomy may outweigh the protective effect of tentorial alignment in selected cases. Recent simulation studies have further emphasized the importance of preoperative evaluation of tentorial anatomy and venous structures when planning OTA [12].

The present study also highlights the importance of venous anatomy during OTA. Compression of basal occipital veins was significantly associated with postoperative FLAIR lesions ≥ 2 mL but not directly with de novo postoperative epilepsy. In addition, sacrifice of bridging veins was associated with the development of larger FLAIR lesions. These findings suggest that venous compromise may contribute to occipital lobe injury even in the absence of obvious arterial ischemia. Preservation of venous drainage and minimization of focal occipital lobe compression may therefore represent important technical considerations for reducing approach-related morbidity.

Several technical implications arise from these findings. Adequate cerebrospinal fluid drainage remains essential for reducing brain volume and minimizing occipital lobe retraction. Careful arachnoid dissection around bridging veins may increase venous mobility and expand the operative corridor without excessive cortical pressure. Because OTA is performed through a relatively narrow operative corridor, there is a tendency for the retractor to remain fixed at the same location for extended periods while deep microsurgical manipulation is performed. Under such circumstances, even moderate retraction may result in focal cortical injury. Surgeons should therefore strive to minimize retraction pressure, frequently reassess the degree of cortical compression, and periodically reposition the retractor whenever possible. Surgeons should also remain aware that even subtle occipital cortical injury, which may not produce immediate neurological deficits, can result in delayed epilepsy and long-term functional consequences.

This study has several limitations. First, it was a retrospective single-center study with a relatively small sample size. Second, all procedures were performed by a single surgeon, which may limit the generalizability of the findings. Third, postoperative epilepsy was identified primarily through comprehensive clinical assessment, and electroencephalography was not available in every patient. Although this reflects the retrospective nature of the study, diagnoses were based on the overall clinical course, seizure semiology, electroencephalography findings when available, long-term follow-up, and response to antiseizure medications. Nevertheless, the possibility of significant underdetection is likely low because patients were followed for a relatively long period by the operating surgeon, who routinely inquired about seizure-related symptoms, and electroencephalography was performed whenever clinically indicated. Finally, because only four epilepsy events were observed and the study was exploratory in nature, multivariable analysis could not be performed. In addition, only procedures performed at the current institution were included because complete source data from previous institutions were unavailable for detailed retrospective review. This may have introduced selection bias.

Despite these limitations, the present study provides evidence that postoperative occipital lobe injury is closely associated with de novo epilepsy following OTA. To our knowledge, this is the first study to demonstrate an association between persistent postoperative occipital FLAIR lesions and epilepsy after OTA. Although the sample size was limited, the remarkable consistency of the clinical, radiological, and surgical findings across all epilepsy cases strengthens the validity of our observations. Future multicenter studies are warranted to validate these findings and to further clarify strategies for preventing approach-related occipital lobe injury. Nevertheless, OTA remains a valuable approach for deep-seated lesions, and the present findings emphasize that meticulous preservation of the occipital lobe and its venous drainage should be regarded as an essential component of this approach to minimize the risk of postoperative epilepsy.

Conclusions

Occipital cortical contusion and postoperative occipital FLAIR lesions ≥ 2 mL were strongly associated with de novo epilepsy following OTA. Venous compromise, including compression of basal occipital veins and sacrifice of bridging veins, was associated with postoperative occipital lobe injury. These findings suggest that meticulous preservation of venous structures and avoidance of occipital lobe injury are important surgical principles for reducing postoperative epilepsy after OTA.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

H.H. collected and analyzed the data and drafted the manuscript.

T.K. conceived the study, performed all surgical procedures, supervised the project, and revised the manuscript.

T.O., D.Y., and H.K. contributed to data interpretation and manuscript revision. 

I.S. supervised the study and revised the manuscript.

All authors approved the final manuscript.

Funding

No funding was received for this study.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Informed consent

Informed consent was obtained using an institutional opt-out procedure.

Conflict of interest

The authors declare that they have no conflict of interest.

Use of artificial intelligence

The authors used ChatGPT (OpenAI) for language editing and manuscript refinement. The authors reviewed and approved all content and take full responsibility for the final manuscript.