Abstract
Purpose
To evaluate the diagnostic utility of four preoperative MRI findings in identifying the location of dural defects associated with spinal extradural arachnoid cyst (SEAC) to optimize surgical planning.
Materials and Methods
Fifteen patients with symptomatic SEAC who underwent surgery between 2003 and 2023 were analyzed retrospectively. The following preoperative MRI features were independently evaluated by two radiologists: the septation sign, extraforaminal extension with or without bone remodeling, the C-sign, and the flow-void sign. The accuracy of each finding in predicting the location of surgically confirmed dural defects was assessed.
Results
The septation sign was observed in all patients (100%), with the accuracy of predicting the location of the dural defect ranging from 80% to 87%. Extraforaminal extension was also identified in all patients, with a similar accuracy of (80%–87%). The C-sign and flow-void signs were less frequently observed (40%–47% and 47%–93%, respectively), with accuracies in predicting the dural defect location ranging from 71%–83% and 71%–79%, respectively. Inter-reader agreement was highest for the C-sign and flow-void sign (100%), followed by extraforaminal extension (93%) and the septation sign (80%).
Conclusion
Recognition of MRI features—the septation sign, extraforaminal extension with or without bone remodeling, the C-sign, and the flow-void sign—may aid in localizing dural defects. These imaging markers may support surgical planning for SEAC.
Keywords: Arachnoid Cysts, Magnetic Resonance Imaging, Spine
Graphical Abstract
Abstract
목적
본 연구는 척추 경막외 수막낭종(spinal extradural arachnoid cyst; 이하 SEAC)에서 경막 결손 위치를 예측함에 있어 수술 전 자기공명영상(MRI)에서 관찰되는 네 가지 소견의 진단적 유용성을 평가하고자 하였다.
대상과 방법
2003년부터 2023년까지 수술을 받은 SEAC 환자 중 15명을 대상으로 후향적 분석을 시행하였다. 한 명의 영상의학과 전문의와 한 명의 영상의학과 전공의가MRI 소견들—격벽 징후(septation sign), 골 변형 유무를 포함한 추간공 외 확장, C-sign, 신호소실(flow-void sign)—의 유무를 독립적으로 판독하였으며, 수술로 확인된 경막 결손 위치와의 비교를 통해 정확도를 평가하였다.
결과
격벽 징후 소견은 모든 환자에서 관찰되었고 수술 소견과의 정확도는 80%–87%였다. 추간공 외 확장도 모든 환자에서 관찰되었고 정확도도 80%–87%였다. C-sign과 신호소실은 각각 40%–47%와 47%–93%의 환자에서 관찰되었고 정확도는 각각 71%–83%와 71%–79%였다. 두 판독자 간의 일치도(percentage agreement)는 C-sign과 신호소실이 가장 높았고(100%), 다음으로 추간공 외 확장(93%), 격벽 징후(80%) 순이었다.
결론
격벽 징후, 골 변형의 유무를 포함한 추간공 외 확장, C-sign, 신호소실의 네 가지 MRI 소견은 척추 경막외 수막낭종에서 경막 결손의 위치를 예측하는 데 유용하며, 수술 전략을 수립하는 데 도움이 될 수 있겠다.
INTRODUCTION
Spinal extradural arachnoid cysts (SEACs) are rare entities, accounting for less than 1% of all space-occupying lesions of the spine (1). They are thought to be linked to congenital abnormalities of the dura mater or acquired dural defects resulting from trauma, inflammation, hemorrhage, or iatrogenic causes, such as surgery (2,3,4). These factors allow the arachnoid mater to herniate through the dural defect, most commonly on the dorsal side of the dural sac in the thoracic or thoracolumbar regions (2,4,5). SEAC may involve a one-way valve-like mechanism that intermittently increases pressure within the extradural cyst (4,5,6). Although SEACs are typically asymptomatic, this dynamic process may cause gradual expansion within the spinal canal or neural foramen, ultimately leading to compression of the spinal cord and/or nerve roots (2,5). Diagnosis of SEAC involves various imaging modalities, including MRI, computed tomography myelography, and digital subtraction cystography (5,7,8,9,10). While the standard surgical approach includes complete cyst resection and dural defect closure (4,6,11), an alternative involves closing the dural defect alone, which can be achieved via selective laminectomy (4,8,12).
SEACs typically communicate with the subarachnoid space through a small dural defect (7). One study found that direct closure of the dural defect without cyst removal—achieved through selective laminectomy—resulted in postoperative neurological recovery comparable to that seen with complete cyst resection (4). Moreover, evidence suggests that SEAC recurrence is more closely linked to failure to close the dural defect than to incomplete cyst resection (13). Given these considerations, accurate preoperative localization of the defect is crucial for guiding the most appropriate surgical approach.
MRI remains the most effective diagnostic modality for SEACs (8,14). However, the accuracy of MRI in localizing dural defects remains a subject of debate (4,9), despite several studies offering valuable insights (8,13). Given the persistent uncertainty regarding MRI’s ability to localize dural defects, this study aimed to assess the diagnostic utility of four preoperative MRI features—the septation sign, extraforaminal extension with or without bone remodeling, the C-sign, and the flow-void sign—in accurately localizing dural defects associated with SEACs to inform optimal surgical planning.
MATERIALS AND METHODS
The Ethical Review Board of our institution approved this study (IRB No. B-2402-880-105), confirming that it adhered to established ethical guidelines and fell within the scope of scholarly research. Due to the retrospective nature of the study and the use of de-identified patient data, informed consent was waived. The execution of all protocols involving human subjects was characterized by a thorough commitment to the ethical standards prescribed by the institutional and national research authorities. These protocols consistently adhered to the principles outlined in the 1964 Helsinki Declaration, its subsequent amendments, and similar ethical frameworks.
STUDY POPULATION
We conducted a retrospective search of all consecutive patients who underwent surgery for symptomatic extradural arachnoid cysts between January 2003 and December 2023. Patients were eligible for inclusion if they had: 1) a confirmed diagnosis of SEAC based on both preoperative MRI and intraoperative confirmation, 2) cyst-related symptoms, 3) in-house surgical treatment with available documentation of the dural defect site, and 4) MRI performed within 60 days before surgery. The exclusion criteria were as follows: 1) insufficient quality of the MRI images for lesion assessment and 2) ambiguity or absence in surgical records regarding the dural defect location. Nineteen patients with symptomatic extradural arachnoid cysts were identified through our clinical data warehouse. Of these, three were excluded due to unavailable data on the dural defect level following surgery at other institutions, and one due to inadequate MRI quality. As a result, 15 patients were included in the final analysis. Clinical and surgical data were collected through a retrospective review of electronic health records and operative notes. Two independent investigators (reader 1: C.G.C. and reader 2: Y.H.) systematically reviewed the data to collect information on demographic variables (age, sex), clinical symptoms, symptom duration, laterality and extent of the cyst, surgical approach, documented dural defect location, and postoperative clinical outcomes. Discrepancies between reviewers were resolved by consensus. All the data presented in Table 1 were derived from this structured chart and operative record review.
Table 1. A Summary of Clinical, Imaging, Surgical, and FU Findings in 15 Patients with SEAC.
| No. of Case | Sex | Age | Symptom | Symptom Duration (Months) | MRI Findings | Surgical Treatment | FU (Months) | Symptom Improvement | Cyst Recurrence | |||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Level | Middle Level of the SEACs | No. of Total Segments | Multiple Cysts | Surgical Method | Defect Level | Defect Laterality | ||||||||
| 1 | M | 14 | Backache | 3 | T2-S2 | T10-11 | 18 | Yes | T9-11 laminotomy + dural defect repair | T7, T8, L2 | NA | 114 | Yes | No |
| 2 | F | 25 | Backache | 24 | T9-L1 | T11 | 5 | No | T10-L1 hemilaminectomy T12-L1 + dural defect repair | T12-L1 | Midline | 58 | Yes | No |
| 3 | F | 18 | Backache, both leg pain | 10 | T12-L3 | L1-2 | 4 | No | T12-L2 laminectomy + cyst excision | T12/L1 | Right | 33 | Yes | No |
| 4 | M | 49 | Right foot weakness | 10 | T11-L2 | T12-L1 | 4 | No | T11-L1 laminotomy, T12 fistulectomy + dural defect repair | T12 | Left | 16 | Yes | No |
| 5 | M | 36 | Lower leg weakness | 2 | T8-T11 | T9-10 | 4 | No | T8-9 laminotomy + cyst excision | T10 | Left | 11.5 | Yes | No |
| 6 | M | 22 | Backache, right leg pain | 12 | T11-L3 | L1 | 5 | No | T12-L2 laminotomy + cyst excision | L1 | Right | 92 | No | No |
| 7 | F | 63 | Backache, right leg pain | 120 | T10-L1 | T11-12 | 4 | No | T10-L1 laminotomy + cyst excision | T12 | NA | 179 | No | No |
| 8 | M | 50 | Both leg numbness and weakness | 5 | T6-T7 | T6-7 | 2 | No | T6-7 laminectomy + cyst excision | T7 | Left | 61.5 | Yes | No |
| 9 | M | 21 | Backache, right leg numbness | 2 | T12-L1 | T12-L1 | 2 | No | T12-L1 laminotomy + cyst excision | T12 | Right | 15 | Yes | No |
| 10 | M | 12 | Gait disturbance | 48 | T9-L2 | T11-12 | 6 | No | T10-L1 laminectomy + cyst excision | T11/12 | Left | 70 | Yes | No |
| 11 | F | 38 | Left leg numbness and weakness | 7 | T12-L2 | L1 | 3 | No | T12-L2 laminectomy + cyst excision | L1 | Right | 19 | Yes | No |
| 12 | F | 69 | Backache, both leg numbness | 120 | T11-L3 | L1 | 5 | No | T12-L1 laminectomy + cyst excision | L1 | Right | 67 | Yes | No |
| 13 | F | 40 | Left leg pain, weakness | 12 | T10/11-L4 | L1 | 6.5 | No | T11-L1 laminectomy + cyst excision | L1 | Left | 8.5 | Yes | Yes |
| 14 | M | 47 | Backache, both leg numbness | 3 | T12-L3 | L1-2 | 4 | No | T12-L1 laminectomy + cyst excision | L1 | Left | 11 | Partial | No |
| 15 | F | 21 | Backache | 2 | T10/11-L2 | T12 | 4.5 | No | T11-L1 laminectomy + cyst excision | T12 | Left | 14.5 | Yes | No |
F = female, FU = follow-up, M = male, NA = not applicable, SEAC = spinal extradural arachnoid cyst
MRI IMAGE PROTOCOLS
MRI examinations were performed using advanced 3T scanners (Ingenia, Philips Medical System, Best, Netherlands), as well as 1.5T MR scanners (Intera, Philips Medical System, Amira siemens/outside Intera Philips/outside Signa GE/outside Magnetom siemens) with patients in the supine position with their arms resting at their sides. Various protocols were used, but all included sagittal and axial T2-weighted image (T2WI), and sagittal and axial T1-weighted image (T1WI). The parameters for individual sequences were as follows: For sagittal T2-weighted turbo spin echo (TSE) images: voxel size = 0.8–0.9 × 1.0 × 3.0 mm3 (cervical) or 0.6–0.9 × 1.0–1.3 × 4.0 mm3 (thoracolumbar); repetition time (TR) = 2,500–4,800 ms; echo time (TE) = 100–120 ms; echo train length = 28–35; slice thickness = 3.0–4.0 mm; slice gap = 0.3–0.4 mm; number of excitations = 2–3. For axial T2-weighted TSE images: voxel size = 0.7 × 0.7 × 3.0 mm3 (cervical) or 0.6 × 0.6 × 4.0 mm3 (thoracolumbar); TR = 2,700–4,500 ms; TE = 120 ms; echo train length = 25–35; slice thickness = 3.0–4.0 mm; slice gap = 1.0 mm; number of excitations = 2. For sagittal enhanced T1-weighted TSE images: voxel size = 0.7 × 1.0 × 3.0 mm3 (cervical) or 0.8 × 1.0 × 4.0 mm3 (thoracolumbar); TR = 500–700 ms; TE = 9–20 ms; echo train length = 5–10; slice thickness = 3.0–4.0 mm; slice gap = 0.3–0.4 mm; number of excitations = 1. For axial enhanced T1-weighted TSE images: voxel size = 0.6 × 0.6 × 3.0 mm3 (cervical) or 0.7 × 0.7 × 4.0 mm3 (thoracolumbar); TR = 500–700 ms; TE = 9–20 ms; echo train length = 5–10; slice thickness = 3.0–4.0 mm; slice gap = 1 mm; number of excitations = 1. Among the 15 patients, 8 underwent MRI evaluations at external institutions using different scanners and protocols.
MRI INTERPRETATION
To minimize potential recall bias, one board-certified radiologist (reader 1, with 9 years of experience in spine radiology) and one radiology trainee (reader 2, with 3 years of experience) independently reviewed the MRI findings to predict SEAC in each patient four weeks after the patient selection process. We assessed four MRI findings: septation sign, level with the most prominent extraforaminal extension with/without bone remodeling, C-sign, and flow-void sign, which may suggest the site of a dural defect. The septation sign is depicted as a thickened septal wall within the SEAC, best demonstrated on T2-weighted axial images (Figs. 1, 2), and if multilevel septations were present, the most prominent level was recorded as the septation sign level. Extraforaminal extension with or without bone remodeling was evaluated on both T2-weighted axial and sagittal images, and the most severe level was determined (Figs. 2, 3). The C-sign was defined as the extension of the SEAC at both the anterior and posterior epidural spaces on axial images (Figs. 2, 4). When present at multiple levels, severity was determined based on anterior epidural space expansion, and the most affected level noted. The flow-void sign, which has been proposed in other studies (8,13,14,15,16), is a cerebrospinal fluid (CSF) flow artifact within the SEAC that may indicate the level of the dural defect (Figs. 2, 4), and the level with the most severe flow-void sign was recorded. The predicted site was considered accurate if it was within one vertebral level of the actual dural defect (e.g., the predicted level of T10 was considered correct if the actual tear was at the T10/T11 disc level). The comparison was confined to the anatomical vertebral level, without distinction of laterality (i.e., left or right). In addition, a subgroup analysis was performed for patients with SEACs extending across more than two vertebral segments, as accurate localization of the dural defect in these cases is of greater clinical relevance for surgical planning. The predictive accuracy of the MRI features was assessed using the same one-level margin threshold.
Fig. 1. SEAC in an 18-year-old female with backache and both leg pain: axial and sagittal T2-weighted 1.5T MR images (case 3).
A. Axial image at the L2 level reveals thickened internal septations within the SEAC, indicative of a positive septation sign (arrow).
B. Sagittal image demonstrates the same septation sign (arrow), consistent with axial findings. The dural defect was surgically confirmed at the T12/L1 disc level.
SEAC = spinal extradural arachnoid cyst
Fig. 2. Spinal extradural arachnoid cyst in a 36-year-old male with lower leg weakness: axial and sagittal T2-weighted 1.5T MR images (case 5).
A. Axial image at the T10 level reveals expansion of the cyst into both the anterior and posterior epidural spaces (arrows), consistent with a positive C-sign. Additionally, a prominent intralesional signal void is noted (open arrow), indicating a flow-void sign.
B. Sagittal image shows thickened internal septations within the extradural arachnoid cyst at the T10 level (arrowhead), corresponding to a positive septation sign.
C, D. Axial image at the T9/10 level and sagittal image demonstrate extraforaminal extension of the cyst, with the most marked protrusion involving the left T9/10 neural foramen (arrowheads).
Fig. 3. Spinal extradural arachnoid cyst in a 40-year-old female with left leg pain and weakness: sagittal T2-weighted 1.5T MR image (case 13) demonstrates multilevel extraforaminal extension with associated bone remodeling. The most prominent lesion is identified at the L1/2 level (arrow).
Fig. 4. SEAC in a 22-year-old male with backache and right leg pain: axial T2-weighted 1.5T MR image (case 6) at the L1 level shows SEAC extension into both anterior and posterior epidural spaces, indicating a positive C-sign (arrows). A prominent signal void is also noted within the cyst, consistent with a flow-void sign (open arrow).
SEAC = spinal extradural arachnoid cyst
STATISTICAL ANALYSIS
All statistical analyses were performed using the SPSS software (version 25.0; IBM Corp., Armonk, NY, USA). Categorical variables included sex, symptom characteristics, the presence or absence of each MRI finding (septation sign, extraforaminal extension with or without bone remodeling, C-sign, and flow-void sign), and laterality of the dural defect (left, right, or midline). Continuous variables included patient age, symptom duration, and length of follow-up (in months). Categorical data were summarized using frequencies and percentages, while continuous variables were presented as mean ± standard deviation. The prevalence of each MRI finding was calculated as the proportion of patients in whom these findings were observed. Accuracy in detecting the dural defect location was defined as the match between MRI-predicted findings and the intraoperatively confirmed defect location. Percentage agreement was defined as the proportion of cases in which both readers predicted the same vertebral level. Interobserver agreement for each MRI feature was assessed using Cohen’s kappa coefficient, which was calculated using the same software. Kappa values were interpreted as follows: <0.20, slight; 0.21–0.40, fair; 0.41–0.60, moderate; 0.61–0.80, substantial; and >0.80, almost perfect agreement.
RESULTS
CLINICAL AND SURGICAL FINDINGS
Fifteen patients were enrolled during the study period (mean age ± standard deviation: 35 ± 18 years; eight males). The presenting symptoms were back pain (60%), numbness (33%), lower-extremity weakness (33%), and leg pain (27%). Symptom duration varied widely, ranging from 2 to 120 months. Dural defects were most frequently found on the left side (seven cases), followed by the right side (five cases) and midline (one case), with two cases lacking documentation of laterality. SEAC was observed as solitary cysts in all but one case, in which the patient had multicentric cysts. Comprehensive details are provided in Table 1, and the frequencies and percentages of clinical characteristics and MRI features are summarized in Supplementary Table 1.
MRI FEATURE ASSESSMENT
As shown in Table 2, four MRI features were evaluated in relation to the level of the surgically identified lesion. For reader 1, the septation sign was observed in all patients, with an accuracy of 87% in predicting the location of the dural defect. Extraforaminal extension with or without accompanying bone remodeling was also consistently observed, with an accuracy of 87% in predicting the location of the dural defect. C-sign and flow-void sign were identified less frequently (40% and 47%, respectively), with accuracies of 83% and 71% in predicting the location of the dural defect, respectively. Reader 2 detected signs of septation and extraforaminal extension in all patients with 80% accuracy in predicting the location of the dural defect. The C-sign and flow-void sign were identified in 47% and 93% of cases, respectively, with accuracies of 71% and 79%, respectively, in predicting the location of the dural defect. Among the 13 patients with cysts extending across more than two vertebral levels, where accurate identification of the surgical level was particularly important, the predictive accuracies of the septation sign, extraforaminal extension with or without bone remodeling, C-sign, and flow-void sign were 85%, 85%, 80%, and 71% for reader 1, and 77%, 77%, 67%, and 77% for reader 2, respectively. The percentage agreement between readers 1 and 2 was 100% for both the C-sign and flow-void sign. Agreement for the septation sign and extraforaminal extension with or without bone remodeling was 80% and 93%, respectively. In addition, interobserver agreement assessed by Cohen’s kappa was almost perfect for the C-sign and flow-void sign (κ = 1.000), substantial for extraforaminal extension with or without (κ = 0.762), and fair for the septation sign (κ = 0.286). The kappa values, along with the percentage of agreement, are summarized in Table 3. Table 4 summarizes the prevalence and predictive accuracy of each MRI feature as assessed by the two readers. Clinical improvement was achieved in 87% of patients (13 cases) following surgery, while persistent back pain and right leg symptoms were reported in two patients. Only one recurrence (7%) was reported during the mean follow-up period of 8.5 months.
Table 2. Assessment of Four MRI Features for Localizing the Dural Defect in 15 Patients with Spinal Extradural Arachnoid Cyst, as Evaluated by Both Eeaders.
| No. of Case | Sex | Age | Defect Level by Surgery | Septation Sign | C-Sign (Anterior Epidural) | Flow-Void Sign | Extraforaminal Extension w/wo Bone Remodeling | ||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Most Severe Level | Detection of Defect Location | Most Severe Level | Detection of Defect Location | Most Severe Level | Detection of Defect Location | Most Severe Level | Detection of Defect Location | ||||
| Reader 1 | |||||||||||
| 1 | M | 14 | T7, T8, L2 | L3 | No | T11 | No | L1 | No | L4/5 | No |
| 2 | F | 25 | T12-L1 | T12/L1 | Yes | T12 | Yes | T11 | No | T11/12 | No |
| 3 | F | 18 | T12/L1 | L2 | No | NA | NA | L1 | Yes | T12/L1 | Yes |
| 4 | M | 49 | T12 | T12 | Yes | NA | NA | NA | NA | T11/12 | Yes |
| 5 | M | 36 | T10 | T10 | Yes | T10 | Yes | T10 | Yes | T9/10 | Yes |
| 6 | M | 22 | L1 | L1 | Yes | L1 | Yes | L1 | Yes | L1/2 | Yes |
| 7 | F | 63 | T12 | T12 | Yes | NA | NA | NA | NA | T12/L1 | Yes |
| 8 | M | 50 | T7 | T7 | Yes | NA | NA | NA | NA | T6/7 | Yes |
| 9 | M | 21 | T12 | T12 | Yes | T12 | Yes | NA | NA | T12/L1 | Yes |
| 10 | M | 12 | T11/12 | T11/12 | Yes | NA | NA | T11/12 | Yes | T11/12 | Yes |
| 11 | F | 38 | L1 | L1 | Yes | NA | NA | L1 | Yes | L1/2 | Yes |
| 12 | F | 69 | L1 | L1 | Yes | NA | NA | NA | NA | L1/2 | Yes |
| 13 | F | 40 | L1 | T12/L1 | Yes | L1 | Yes | NA | NA | L1/2 | Yes |
| 14 | M | 47 | L1 | L1 | Yes | NA | NA | NA | NA | L1-2 | Yes |
| 15 | F | 21 | T12 | T12 | Yes | NA | NA | NA | NA | T12/L1 | Yes |
| Reader 2 | |||||||||||
| 1 | M | 14 | T7, T8, L2 | L3 | No | T11 | No | L1 | No | L4/5 | No |
| 2 | F | 25 | T12-L1 | T12/L1 | Yes | T12 | Yes | T11 | No | T11/12 | No |
| 3 | F | 18 | T12/L1 | T12 | Yes | NA | NA | L1 | Yes | T12/L1 | Yes |
| 4 | M | 49 | T12 | T12 | Yes | NA | NA | T12 | Yes | T11/12 | Yes |
| 5 | M | 36 | T10 | T10 | Yes | T10 | Yes | T10 | Yes | T9/10 | Yes |
| 6 | M | 22 | L1 | L1 | Yes | L1 | Yes | L1 | Yes | L1/2 | Yes |
| 7 | F | 63 | T12 | T12 | Yes | NA | NA | T12 | Yes | T12/L1 | Yes |
| 8 | M | 50 | T7 | T7 | Yes | NA | NA | T7 | Yes | T6/7 | Yes |
| 9 | M | 21 | T12 | T12 | Yes | T12 | Yes | NA | NA | T12/L1 | Yes |
| 10 | M | 12 | T11/12 | T11/12 | Yes | T10 | No | T11/12 | Yes | T11/12 | Yes |
| 11 | F | 38 | L1 | L2 | No | NA | NA | L1 | Yes | L1/2 | Yes |
| 12 | F | 69 | L1 | L1 | Yes | NA | NA | T12 | No | L1/2 | Yes |
| 13 | F | 40 | L1 | T12 | No | L1 | Yes | L1 | Yes | L1/2, L2/3 | Yes |
| 14 | M | 47 | L1 | L1 | Yes | NA | NA | L1/2 | Yes | L2/3 | No |
| 15 | F | 21 | T12 | T12 | Yes | NA | NA | T12/L1 | Yes | T12/L1 | Yes |
F = female, M = male, NA = not applicable, w/wo = with or without
Table 3. Interobserver Agreement for MRI Features Based on Percentage Agreement and Cohen’s Kappa Coefficient.
| MRI Findings | Percent Agreement (%) | Kappa | SE | 95% CI | p-Value |
|---|---|---|---|---|---|
| Septation sign | 80 | 0.286 | 0.307 | -0.316–0.888 | 0.255 |
| Extraforaminal extension w/wo bone remodeling | 93 | 0.762 | 0.223 | 0.325–1.199 | 0.002 |
| C-sign | 100 | 1.000 | 0 | 1.000–1.000 | 0.014 |
| Flow-void sign | 100 | 1.000 | 0 | 1.000–1.000 | 0.008 |
Number of cases used for each analysis: septation sign (n = 15), extraforaminal extension with or without bone remodeling (n = 15), C-sign (n = 6), and flow-void sign (n = 7).
CI = confidence interval, SE = standard error, w/wo = with or without
Table 4. Summary of Prevalence and Accuracy for Each MRI Feature in Predicting the Dural Defect Level in Patients with Spinal Extradural Arachnoid Cyst.
| MRI Findings | Prevalence | Accuracy | |
|---|---|---|---|
| Reader 1 | Septation sign | 100 | 87 |
| Extraforaminal extension w/wo bone remodeling | 100 | 87 | |
| C-sign | 40 | 83 | |
| Flow-void sign | 47 | 71 | |
| Reader 2 | Septation sign | 100 | 80 |
| Extraforaminal extension w/wo bone remodeling | 100 | 80 | |
| C-sign | 47 | 71 | |
| Flow-void sign | 93 | 79 |
Number of cases used for each analysis: for Reader 1—septation sign (n = 15), extraforaminal extension with or without bone remodeling (n = 15), C-sign (n = 6), and flow-void sign (n = 7); for Reader 2—septation sign (n = 15), extraforaminal extension with or without bone remodeling (n = 15), C-sign (n = 7), and flow-void sign (n = 14).
w/wo = with or without
DISCUSSION
The present study demonstrated the potential diagnostic utility of the septation sign and extraforaminal extension, which were observed in all patients (100%) and associated with localization accuracies of 80%–87%. The C-sign and flow-void sign appeared less frequently, in 40%–47% and 47%–93% of cases, respectively, but remained useful complementary indicators with accuracies ranging from 71%–83% and 71%–79%, respectively.
The septation sign was identified in all patients and showed the highest accuracy in predicting the location of the dural defect by both readers (80%–87%). Septations within SEACs, intermittently observed on imaging or during surgery, remain poorly understood etiologically. Gradual cyst enlargement has been proposed to contribute to the development of internal septations (17). Extraforaminal extension with or without bone remodeling was also consistently observed and demonstrated high accuracy in predicting the dural defect location (80%–87%). Larger SEACs have been significantly linked to more pronounced thecal sac compression and a higher frequency of extraforaminal extension (17,18,19,20). Cysts exhibiting bone remodeling tended to be larger, implying a more chronic evolution than those without bony changes. Over time, the increased hydrostatic pressure within larger cysts is thought to lead to gradual remodeling of the adjacent bone (21,22). The C-sign and flow-void sign were less frequently observed, but still showed reasonable accuracy in predicting the location of the dural defect (71%–83% and 71%–79%, respectively). The C-sign may reflect the influence of CSF flow dynamics at the fistula site. It has been postulated that the region of maximal internal pressure corresponds to the location of the dural defect, promoting cyst expansion into both the anterior and posterior epidural spaces. This expansion pattern manifests as a C-shaped configuration in the axial images. The flow-void sign, which reflects turbulent or accelerated CSF movement within the SEAC, may also indicate the location of dural defects (8,23,24). Accelerated CSF flow at the fistula site can generate a prominent signal void on MRI, potentially identifying the level of the dural defect. Previous studies have suggested that evaluating flow voids, axial laterality, maximal cyst size, or the sagittal mid-cyst point could aid in identifying the dural defect site (14).
Numerous investigations have demonstrated that SEACs predominantly occur at the thoracolumbar junction and typically maintain communication with the subarachnoid space through a dural defect (4,5,14). Funao et al. (4) found that 66.7% of their 12 patients had defects located at T12–L1, whereas Xu et al. (8) reported a 100% occurrence at the same level in their 10-patient series. Similarly, the present study identified defects at T12–L1 in 11 of the 15 cases (73.3%). In a review of 145 cases of SEAC, Huang et al. (5) reported that dural defects were mostly located in the mid- or upper-mid portions of the cyst. This distribution may be explained by Pascal’s law, whereby the initially symmetric CSF pressure becomes slightly greater in the inferior regions as the cyst enlarges over time due to subtle hydrostatic differences. Surgical findings in the present study revealed that the dural defect was located at or close to the middle level of the cyst in 11 of 15 cases. This observation aligns with the hypothesis of Huang et al. (5), who suggested that when the defect is not visualized preoperatively, the middle or upper-mid portion of the SEAC is the most probable location.
The majority of prior studies primarily consisted of case reports and literature reviews that documented the presence of these MRI features in SEACs (12,14,22,25,26); however, the specific vertebral or intervertebral level at which these findings appear has rarely been addressed. The present study offers a more targeted approach by assessing the presence and anatomical location of four MRI indicators and comparing them with the operation-confirmed dural defect level in each case, with the goal of refining the preoperative localization of dural defects. Specialized MRI sequences, including cine MRI, constructive interference in steady state (CISS), and time-spatial labeling inversion pulse (Time-SLIP), have been reported to be helpful in identifying dural defect sites (6,7,14,27). However, these advanced techniques are often challenging to implement in routine clinical settings and may require interdepartmental coordination (14). CT myelography is a sensitive tool for detecting fistulas; however, its diagnostic accuracy may decline when CSF flow through the fistula is intermittent. Furthermore, as an invasive modality, it carries associated procedural risks (13,14). Digital subtraction cystography has also been used to localize dural defects, although its clinical utility is limited by the high rate of negative findings (5,9).
This study had several limitations. First, this was a retrospective single-center analysis, which may have limited the generalizability of the results. Second, no external validation was performed. Third, there was heterogeneity in MRI acquisition, as vendors, field strengths, and imaging sequences were not equal in all case series. Fourth, advanced MRI techniques such as cine MRI or time-SLIP were not employed. Fifth, the small sample size may have limited the statistical power and overestimated the diagnostic performance. Lastly, sensitivity, specificity, PPV, and NPV could not be calculated owing to the absence of negative controls, as all patients had confirmed dural defects; thus, only accuracy was assessed.
In conclusion, identifying MRI features such as septation sign, extraforaminal extension with or without bone remodeling, C-sign, and flow-void sign may aid in localizing dural defects and guide surgical strategies for SEAC.
Footnotes
- Conceptualization, C.C.G.
- data curation, H.Y., C.C.G.
- formal analysis, H.Y., L.E.
- investigation, K.Y.
- supervision, C.C.G.
- writing—original draft, H.Y., C.C.G.
- writing—review & editing, all authors.
Conflicts of Interest: The authors have no potential conflicts of interest to disclose.
Funding: None
Supplementary Materials
The Supplement is available with this article at http://doi.org/10.3348/jksr.2025.0045.
Frequency and distribution of clinical characteristics and MRI features in 15 SEAC cases
References
- 1.Fortuna A, La Torre E, Ciappetta P. Arachnoid diverticula: a unitary approach to spinal cysts communicating with the subarachnoid space. Acta Neurochir (Wien) 1977;39:259–268. doi: 10.1007/BF01406736. [DOI] [PubMed] [Google Scholar]
- 2.Qi W, Zhao L, Fang J, Chang X, Xu Y. Clinical characteristics and treatment strategies for idiopathic spinal extradural arachnoid cysts: a single-center experience. Acta Neurochir (Wien) 2015;157:539–545. doi: 10.1007/s00701-014-2278-3. [DOI] [PubMed] [Google Scholar]
- 3.Liu JK, Cole CD, Kan P, Schmidt MH. Spinal extradural arachnoid cysts: clinical, radiological, and surgical features. Neurosurg Focus. 2007;22:E6. doi: 10.3171/foc.2007.22.2.6. [DOI] [PubMed] [Google Scholar]
- 4.Funao H, Nakamura M, Hosogane N, Watanabe K, Tsuji T, Ishii K, et al. Surgical treatment of spinal extradural arachnoid cysts in the thoracolumbar spine. Neurosurgery. 2012;71:278–284. doi: 10.1227/NEU.0b013e318257bf74. discussion 284. [DOI] [PubMed] [Google Scholar]
- 5.Huang H, Wei M, Zhou Q, Peng R, Ding X, Xi J. Spinal extradural arachnoid cysts: a novel formation mechanism and dural defect location technology. Heliyon. 2023;9:e12969. doi: 10.1016/j.heliyon.2023.e12969. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Morizane K, Fujibayashi S, Otsuki B, Sakamoto T, Tsutsumi R, Odate S, et al. Clinical and radiological features of spinal extradural arachnoid cysts: valve-like mechanism involving the nerve root fiber as a possible cause of cyst expansion. J Orthop Sci. 2018;23:464–469. doi: 10.1016/j.jos.2018.01.005. [DOI] [PubMed] [Google Scholar]
- 7.Ishibe T, Senzoku F, Ikeda N, Kamba Y, Mikawa Y. Detection of the communicating hole(s) of spinal extradural arachnoid cysts using time-spatial labeling inversion pulse magnetic resonance imaging. Spine (Phila Pa 1976) 2014;39:E1394–E1397. doi: 10.1097/BRS.0000000000000591. [DOI] [PubMed] [Google Scholar]
- 8.Xu F, Jian F, Li L, Guan J, Chen Z. Surgical treatment of ten adults with spinal extradural meningeal cysts in the thoracolumbar spine. J Korean Neurosurg Soc. 2021;64:238–246. doi: 10.3340/jkns.2020.0244. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Gu K, Kwon JW, Kim ES. Digital subtraction cystography for detection of communicating holes of spinal extradural arachnoid cysts. Korean J Radiol. 2016;17:111–116. doi: 10.3348/kjr.2016.17.1.111. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Patel DM, Weinberg BD, Hoch MJ. CT myelography: clinical indications and imaging findings. Radiographics. 2020;40:470–484. doi: 10.1148/rg.2020190135. [DOI] [PubMed] [Google Scholar]
- 11.Krings T, Lukas R, Reul J, Spetzger U, Reinges MH, Gilsbach JM, et al. Diagnostic and therapeutic management of spinal arachnoid cysts. Acta Neurochir (Wien) 2001;143:227–234. doi: 10.1007/s007010170102. discussion 234-235. [DOI] [PubMed] [Google Scholar]
- 12.Cho HY, Lee SH, Kim ES, Eoh W. Symptomatic large spinal extradural arachnoid cyst: a case report. Korean J Spine. 2015;12:217–220. doi: 10.14245/kjs.2015.12.3.217. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Lee CH, Hyun SJ, Kim KJ, Jahng TA, Kim HJ. What is a reasonable surgical procedure for spinal extradural arachnoid cysts: is cyst removal mandatory? Eight consecutive cases and a review of the literature. Acta Neurochir (Wien) 2012;154:1219–1227. doi: 10.1007/s00701-012-1356-7. [DOI] [PubMed] [Google Scholar]
- 14.Jian Q, Liu Z, Duan W, Jian F, Chen Z. How to locate the dural defect in a spinal extradural meningeal cyst: a literature review. Chin Neurosurg J. 2022;8:27. doi: 10.1186/s41016-022-00291-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Woo JB, Son DW, Kang KT, Lee JS, Song GS, Sung SK, et al. Spinal extradural arachnoid cyst. Korean J Neurotrauma. 2016;12:185–190. doi: 10.13004/kjnt.2016.12.2.185. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Kwon YU, Park DH, Kim DS. [Spinal extradural arachnoid cyst: minimally invasive surgical treatment after localization of dural defect using magnetic resonance myelogram] J Korean Orthop Assoc. 2022;57:436–440. Korean. [Google Scholar]
- 17.Ahmed AK, Anisetti B, Huynh T, Agarwal A, Gupta V, Desai A, et al. Clinical and imaging features of spinal extradural arachnoid cysts: a retrospective study of 50 cases. Neuroradiology. 2022;64:2409–2416. doi: 10.1007/s00234-022-03042-4. [DOI] [PubMed] [Google Scholar]
- 18.Furtado SV, Thakar S, Murthy GK, Dadlani R, Hegde AS. Management of complex giant spinal arachnoid cysts presenting with myelopathy. J Neurosurg Spine. 2011;15:107–112. doi: 10.3171/2011.3.SPINE10672. [DOI] [PubMed] [Google Scholar]
- 19.Hatashita S, Kondo A, Shimizu T, Kurosu A, Ueno H. Spinal extradural arachnoid cyst--case report. Neurol Med Chir (Tokyo) 2001;41:318–321. doi: 10.2176/nmc.41.318. [DOI] [PubMed] [Google Scholar]
- 20.Kahraman S, Anik I, Gocmen S, Sirin S. Extradural giant multiloculated arachnoid cyst causing spinal cord compression in a child. J Spinal Cord Med. 2008;31:306–308. doi: 10.1080/10790268.2008.11760728. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Doita M, Nishida K, Miura J, Takada T, Kurosaka M, Fujii M. Kinematic magnetic resonance imaging of a thoracic spinal extradural arachnoid cyst: an alternative suggestion for exacerbation of symptoms during straining. Spine (Phila Pa 1976) 2003;28:E229–E233. doi: 10.1097/01.BRS.0000065490.71054.2C. [DOI] [PubMed] [Google Scholar]
- 22.Tyagi HR, Kalra KL, Acharya S, Singh RP. Scalloping sacral arachnoid cyst as a cause of perianal pain- a case report. J Orthop Case Rep. 2014;4:28–32. doi: 10.13107/jocr.2250-0685.163. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Miyamoto M, Kim K, Matsumoto R, Isobe M, Isu T. Utility of preoperative magnetic resonance imaging myelography for identifying dural defects in patients with spinal extradural arachnoid cysts: case report. Neurosurgery. 2006;59:E941. doi: 10.1227/01.NEU.0000232659.56174.28. discussion E941. [DOI] [PubMed] [Google Scholar]
- 24.Lee HJ, Cho WH, Han IH, Choi BK. Large thoracolumbar extradural arachnoid cyst excised by minimal skipped hemilaminectomy: a case report. Korean J Spine. 2013;10:28–31. doi: 10.14245/kjs.2013.10.1.28. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Choi JY, Kim SH, Lee WS, Sung KH. Spinal extradural arachnoid cyst. Acta Neurochir (Wien) 2006;148:579–585. doi: 10.1007/s00701-006-0744-2. discussion 585. [DOI] [PubMed] [Google Scholar]
- 26.Min WK, Kim JE. Extensive spinal intradural arachnoid cyst exhibiting a “double cord sign” on magnetic resonance imaging. J Orthop. 2015;13:110–114. doi: 10.1016/j.jor.2015.01.032. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Nakagawa A, Kusaka Y, Jokura H, Shirane R, Tominaga T. Usefulness of constructive interference in steady state (CISS) imaging for the diagnosis and treatment of a large extradural spinal arachnoid cyst. Minim Invasive Neurosurg. 2004;47:369–372. doi: 10.1055/s-2004-830129. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Frequency and distribution of clinical characteristics and MRI features in 15 SEAC cases