Clinical outcomes of resection of thoracic arachnoid webs: patient series

Abstract

BACKGROUND

Spinal arachnoid web (SAW) is a rare pathology of the intradural arachnoid layer that can cause progressive neurological symptoms including pain, numbness, weakness, imbalance, and incontinence. Surgical fenestration is the primary treatment and has shown generally favorable outcomes.

OBSERVATIONS

This retrospective case series includes 18 patients who underwent SAW fenestration at a tertiary care center between 2015 and 2024. The mean age was 54.7 years, and the mean BMI was 34.0 kg/m². Preoperative symptoms included back pain (n = 15), sensory changes (n = 15), gait instability (n = 14), and lower extremity weakness (n = 14). Most patients experienced symptom improvement postoperatively, although 5 (27.8%) reported worsening in at least one domain. Of the 15 patients with a preoperative scalpel sign, 46.7% showed radiographic resolution following surgery.

LESSONS

The authors present one of the largest single-center series of surgically treated SAW to date. While a subset of patients experienced persistent or worsened symptoms, the majority demonstrated both clinical and radiological improvement. These findings support surgical fenestration as a generally effective intervention for symptomatic SAW. Given the progressive nature of the condition and its potential impact on function, early recognition and surgical referral are critical to optimize patient outcomes.

https://thejns.org/doi/10.3171/CASE25910

Spinal arachnoid webs (SAWs) are rare abnormalities of the intradural arachnoid tissue that can cause spinal cord compression and myelopathy.¹ These webs are abnormal thickening of the arachnoid membrane that can obstruct the subarachnoid space.^1,2 Patients with SAWs often complain of pain, weakness, incontinence, and sensory changes in the distribution of the affected spinal cord, but SAWs may also present asymptomatically as an incidental finding.^1,3,4 When properly diagnosed and surgically managed, more than 90% of patients experience improvement in their neurological function.⁵

The pathogenesis of SAWs is not completely understood, but associations have been made with trauma, inflammation, syringomyelia, and congenital abnormalities. SAWs may follow a clear history of trauma, supporting the possibility that SAWs are a variant or collapsed arachnoid cyst.^1,6,7 While arachnoid webs and cysts share overlapping clinical and imaging features, arachnoid webs are typically defined as focal bands of thickened arachnoid tissue that distort the dorsal spinal cord without forming a discrete fluid-filled cavity. In contrast, arachnoid cysts are enclosed CSF-filled sacs. Because of frequent imaging and intraoperative ambiguity, these entities may exist on a pathological continuum.⁸ SAWs have also been associated with thickened ligamentum flavum, supporting an alternative theory of a congenital origin.^9,10 Recent research has also identified an association between SAWs and systemic inflammation, spurring a theory that SAWs are a secondary proliferation of the leptomeninges following inflammatory disruption.^2,11

Syringomyelia is a condition characterized by the development of a fluid-filled cavity within the spinal cord called a syrinx and is commonly associated with SAWs.² Previous case series have shown that 67%–85% of SAWs occur adjacent to a syrinx.^2,5 Similarly, the association with SAWs is not completely understood but seems to be caused by an alteration of CSF flow dynamics.¹² Development of syringomyelia has been well documented as an increase in pulse pressure causing a low-pressure, high-velocity channel that distends the spinal cord.¹³ Some reports have postulated a one-way ball-valve mechanism that causes obstruction and fluid accumulation.⁷

Diagnosing SAWs is challenging due to their rarity, nonspecific symptoms, and subtle MRI findings. SAWs are classically associated with the scalpel sign (Fig. 1C), which consists of an extramedullary transverse band of arachnoid tissue and a dorsal indentation of the spinal cord.^14,15 However, like other focal arachnoid pathologies, these findings are subtle and difficult to visualize, requiring an experienced radiologist to diagnose. Additionally, not all patients present with the scalpel sign.⁴ Spinal cord herniation is important in the differential diagnoses, with some radiographic similarities to SAW. CT myelography is often used to confirm the diagnosis and distinguish between these entities.

FIG. 1.

Diagnostic preoperative images obtained in patients 3 and 17. Patient 3. A: CT myelogram showing anterior displacement and flattening of the spinal cord at T1–2 due to arachnoid web with syrinx-related expansion at C6 and above. B: Correlative findings on T2-weighted MR image. Patient 17. C:Myelogram showing septations and loculations in the subarachnoid space in the thecal sac at T9 and above. D:Correlative findings on T2-weighted MR image.

Given the favorable surgical outcomes, accurate characterization and diagnosis are key to identifying and treating SAWs. In the present study, we retrospectively analyzed the medical records of 18 adult patients with surgically treated SAWs from one tertiary care center.

Study Description

Study Population and Selection

Following institutional review board approval, we retrospectively acquired the medical records of patients who underwent resection of SAW at The Ohio State University Wexner Medical Center between 2015 and 2024. Resection was performed by spine fellowship-trained neurosurgeons. Chart review was conducted for demographic information, comorbidities, past medical and surgical history, neurological symptoms, pre- and postoperative imaging, surgical and hospital course, and clinical progression. SAWs were considered idiopathic in the absence of trauma, thoracic spine surgery, spinal tumor, spinal infection, inflammatory conditions, arachnoid cysts, syringomyelia, congenital anomalies, or vascular malformations. Patient inclusion criteria included preoperative MRI, surgical diagnosis and resection of SAW, and age greater than 17 years. For the purposes of this study, we did not distinguish between arachnoid webs and arachnoid cysts due to a lack of definitive radiographic and clinical points of differentiation. Descriptive statistics were utilized due to the relatively small cohort size. The paired t-test was used to evaluate pre- and postoperative radiographic outcomes.

Demographics and Clinical Presentation

Demographic, comorbidity, and presenting symptom data are described in Table 1. Eighteen of 28 patients met inclusion criteria. The mean age and BMI were 54.7 years and 34.0 kg/m², respectively. Active smokers comprised 38.9% (n = 7) of the cohort. Patient comorbidities included diabetes (n = 3) and autoimmune conditions (n = 2). Idiopathic SAWs were present in 13 patients. The remaining patients had risk factors including prior thoracic SAW resection or thoracic spinal surgery, with one instance of staphylococcal abscess in the midthoracic spine. Additionally, 2 patients had a history of prior CNS or systemic infection, and 1 had history of minor lumbar trauma secondary to a motor vehicle accident. All 18 patients in this cohort were found to have thoracic SAWs, and 5 experienced the concomitant presence of a syrinx.

TABLE 1.

Patient demographics, medical history, and presenting symptom (n = 18)

	Value
Age, yrs	54.7 (10.8)
BMI, kg/m2	34.0 (7.8)
Male sex	12 (66.7)
White race	14 (77.8)
Black race	4 (22.2)
Active smoker	7 (38.9)
Diabetes	3 (16.7)
Active steroid use	1 (5.6)
Autoimmune condition	2 (11.1)
Prior spinal surgery	5 (27.8)
Prior infection	2 (11.1)
Prior spinal trauma	1 (5.6)
Thoracic SAW	18 (100.0)
Idiopathic SAW	13 (72.2)
Syrinx	5 (27.8)
Presenting symptom
Back pain	15 (88.3)
Lower extremity weakness	14 (77.8)
Sensory symptoms	15 (83.3)
Bowel or bladder dysfunction	6 (33.3)
Hyperreflexia	10 (55.6)
Clonus	9 (50.0)
Gait instability	14 (77.8)

Values are given as number of patients (%) or mean (SD).

No surgeries were performed via a minimally invasive or endoscopic approach. Additionally, no cases consisted of an intramedullary approach to the syrinx. There was 1 case (patient 6) of shunt placement indicated for drainage of a syrinx. Postoperative hospital courses were uncomplicated for all patients (94.4%) except one (patient 6), who experienced a pulmonary embolism in the days following surgery.

Table 1 further depicts the symptoms noted on initial presentation to the clinic. All patients (n = 18) experienced myelopathic symptoms including back pain (n = 15); sensory symptoms including numbness, paresthesia, proprioceptive difficulty, or vasomotor symptoms (n = 15); gait instability (n = 14); and lower extremity weakness (n = 14). Less frequently encountered symptoms included hyperreflexia (n = 10), clonus (n = 9), and bowel or bladder dysfunction (n = 6). An additional 3 patients presented with unilateral or bilateral foot drop.

Diagnostic Protocol

While diagnostic steps varied slightly between patients, the diagnostic workup of a newly discovered SAW is described here. On presentation to the clinic with myelopathic symptoms such as those described in Table 1, the initial workup includes physical examination and MRI. Further imaging for preoperative planning includes CT myelography (Fig. 1). Myelography aids in the diagnosis of SAW via visualization of CSF flow, spinal cord morphology, and meningeal abnormalities; in addition, myelography is useful for preoperative surgical planning when an arachnoid web is suspected. The presence of a fluid-filled sac adjacent to the cord may be suggestive of cyst-like architecture, while thin, fibrous bands suggest a web-like architecture. Additionally, ventral cord herniation can be ruled out with the guidance of myelography.

Symptomatic Course

Presentation and symptomatic courses are detailed in Table 2. Most patients and symptom parameters improved from the time of surgery to the last known follow-up. Nine (50%) patients had less than 12 months of follow-up, and postoperative imaging was not routinely performed at standardized intervals. The median follow-up was 357.5 days (range 25–2015 days). However, some patients experienced persistent or worsening symptoms. Patient 15 reported worsening bilateral lower extremity numbness, tingling, and balance with continued pain; further workup with MRI showed a new spinal cord enhancement suspicious for an underlying inflammatory or autoimmune etiology. Patients 3, 4, 5, and 11 experienced new-onset neurological symptoms such as incontinence, numbness, and pain that were unrelated to the primary pathology of interest or the surgical encounter. Among the 18 patients, symptomatic improvement varied based on the symptom itself. Pain improved in 11 of 15 (73%) patients who originally presented with pain. Weakness improved in 10 of 14 (71%) patients, and sensory symptoms improved in 10 of 15 (67%) patients. Bowel and bladder symptoms improved in 5 of 6 (83%) patients, and new symptoms developed in 1 patient. Radiographic resolution of the scalpel sign was noted in 5 of 15 (33.3%) patients, and syrinx improvement was noted in 6 patients with compatible postoperative imaging.

TABLE 2.

Neurological symptom information for individual patients

Patient No.	Levels	Pain	Preop Symptoms				Scalpel Sign	Syrinx Level	Pain	Postop Symptoms				Scalpel Sign	Syrinx Resolution	FU (days)
			LE Weakness	Sensory	BB	TB				LE Weakness	Sensory	BB	TB
1	T4–5	Yes	Yes	Yes	No	No	No		+	+	+		No	No		512
2	T2	Yes	Yes	Yes	No	No	Yes		+	+	+		Yes	Yes		391
3	T1–2	Yes	Yes	Yes	No	Yes	Yes	C6–T1	=	+	+	−			Partial	330
4	T9–11	Yes	Yes	Yes	No	No	No		+	=	−		No	Yes		543
5	T5–6	Yes	No	Yes	No	No	Yes	T5–6	−		=				Partial	1285
6	T6–10	Yes	Yes	Yes	Yes	No	Yes	T7–T9	+	+	+	+	No	No	Partial	385
7	T5–6	Yes	No	Yes	Yes	Yes	Yes		+		+	+				151
8	T4–5	Yes	Yes	No	No	No	Yes		+	+			No	No		2015
9	T5–7	Yes	Yes	Yes	Yes	Yes	Yes		+	=	=	=				658
10	T5–7	Yes	Yes	No	No	Yes	Yes		+	=			No	No		527
11	T7–8	No	Yes	Yes	Yes	Yes	Yes		−	+	+	+	No	No		540
12	T2–5	No	Yes	Yes	Yes	No	No			+	=	+	No	Yes		224
13	T2–3	Yes	Yes	Yes	No	No	Yes	C7–T3	+	+	+		No	No	Yes	174
14	T6	Yes	Yes	Yes	No	No	Yes	Medulla T5–6	+	+	+		No	No	Partial	106
15	T5–6	Yes	Yes	Yes	No	No	Yes		=	−	−		No	No		75
16	T7–8	No	No	Yes	No	No	Yes				+		No	Yes		153
17	T9–10	Yes	No	Yes	No	No	Yes	T9–12	+		+		No	Yes	Partial	55
18	T3–4	Yes	Yes	No	Yes	No	Yes			+		+				25

BB = bowel or bladder dysfunction; FU = follow-up; LE = lower extremity; TB = transverse band; + = improvement of symptoms; − = worsening of symptoms; = = no change in symptoms.

Operative and Radiographic Variables

The operative steps for SAW resection were similar for all patients. Each case featured intraoperative fluoroscopy, ultrasound, microsurgical techniques, and neuromonitoring.

First, somatosensory and motor evoked potentials were established by a neuromonitoring team. Fluoroscopy was then utilized for anatomical localization and incision planning. Following dorsal midline laminectomy, intraoperative ultrasound was utilized to visualize the location and extent of the web. Additionally, it confirmed the presence of adequate bony exposure. At this point, the microscope was brought into the operative field and the remainder of the case was performed utilizing microsurgical techniques. The dura was opened sharply and tacked bilaterally to grossly visualize the web. Following this step, the web was gently elevated and separated from the cord and dissected with microscissors. The resected tissue was sent for pathological analysis. At this point, dentate ligaments could be severed unilaterally to mobilize the spinal cord for visualization of the ventral aspect to rule out ventral cord herniation. Finally, osteoplastic reconstruction of dorsal spinal elements could be performed. Among the included patients, 12 underwent laminoplasty and 6 laminectomy.

In Table 2, pre- and postoperative radiographic findings are described. Sagittal and axial MRI was available for all patients preoperatively and for 72.2% of patients postoperatively, with the remainder receiving no imaging or CT. Most webs were multilevel, with only two single-level webs (patients 2 and 14). Preoperatively, transverse bands and scalpel sign were present in 5 (27.8%) and 15 (83.3%) of patients, respectively. Two patients demonstrated resolution of the transverse bands. Seven instances of dorsal indentation were resolved on follow-up, with 2 patients demonstrating new postoperative indentation. The syrinx was present in 6 (33.3%) patients preoperatively, with partial resolution in 6 patients and complete resolution in 1 patient. In all patients but one, the syrinx expanded beyond the level(s) of the associated SAW.

Imaging and intraoperative photographs are shown in Fig. 2. The dorsal indentation caused by an SAW, commonly referred to as the scalpel sign, is visible in Fig. 2A. In Fig. 2B, resolution of the scalpel sign is seen indicating resection of the SAW. Figure 2C and D demonstrates pre- and postoperative imaging for patient 13 with visible resolution of the syrinx. Figure 2A and B provides an example of a thoracic SAW prior to resection; in Fig. 2B, the same patient’s anatomy is seen post–SAW resection. Figure 3C–F demonstrates intraoperative ultrasound images in the axial and sagittal planes. No radiographic recurrences or overt cord tethering were identified on follow-up imaging, which was limited by the short follow-up period.

FIG. 2.

Patient 12. A and B: Pre- and postoperative thoracic (A) and cervical (B) MR images showing evidence of thoracic SAW resolution. A scalpel sign is visible at T3–4 in panel A. Patient 13. C: Preoperative sagittal MR image featuring irregular cord expansion due to syrinx formation with ventral cord compression. D: Evidence of partial syrinx resolution is evident with central canal expansion still appreciable at the distal segment.

FIG. 3.

Intraoperative photographs displaying the arachnoid web (A) and cord postresection (B) for patient 13. Intraoperative axial (C and E) and sagittal (D and F) ultrasound images obtained in patients 12 (C and D) and 14 (E and F).

Informed Consent

The necessary informed consent was obtained in this study.

Discussion

Observations

In this case series, we present 18 cases with surgically fenestrated SAW and their respective clinical outcomes to provide standardized diagnostic and operative protocols. Several single-center case series on SAWs have been published in the last decade, with the largest describing the courses of 26 surgical patients.^2,16–19 A larger, multicenter study of 85 patients managed medically and surgically found that back pain and weakness of the lower extremities were chief complaints of most included subjects, as was the case in our study.²⁰ The rates of other presenting features in our series, such as sensory loss, bowel or bladder dysfunction, and gait abnormalities, were similar to those in the existing literature. Additionally, the multicenter study found that nearly 90% of SAWs were located between T2 and T6. In our case series, a similar proportion of patients (77.8%) had SAWs involving T2–6, with the remaining located between T7 and T11. The upper and middle thoracic spinal cord is characterized by a physiological narrowing of the canal, limited vascular supply, and substantial mechanical stress; however, further investigation is needed to understand the consistent localization of SAWs in this region.

SAWs are associated with trauma and may arise from a collapsed arachnoid cyst; they also may be associated with a congenitally thickened ligamentum flavum.^1,6,7,9,10 Inflammatory conditions such as infection, prior spine surgery, and myelofibrosis (as described in a case report) may be an additional potential causative factor.^2,11 However, a series of 17 patients featuring pathology sample analysis did not find evidence of inflammation; rather, samples consisted of connective, collagenous tissue.¹⁹ In our study, few patients presented with comorbidities associated with systematic inflammation. As detailed in Table 2, patient 13 experienced worsening symptoms postoperatively correlated with new spinal cord enhancement suggestive of an inflammatory or autoimmune etiology.

Evidence of SAW on MRI is sometimes subtle, and care must be paid to differentiate web or cystic arachnoid structures from ventral transdural herniation of the spinal cord.²¹ Differentiation may be possible in the presence of the scalpel sign and myelography showing the absence of a cyst.²² This sign is classically present in Fig. 2A. The scalpel sign is identified by dorsal indentation of the spinal cord and is associated with transverse bands of arachnoid tissue in adjacent thecal space. Indeed, the presence of the scalpel sign is highly correlated with SAW, as demonstrated in a case series of 12 patients with a 100% rate of scalpel sign presence.¹⁶ Although there is evidence of SAW without the scalpel sign in the literature, 83% of our patients demonstrated a scalpel sign preoperatively on sagittal MRI.^4,14 In our series, only 1 patient of the 5 with persistence of this sign on postoperative MRI experienced symptomatic worsening. This suggests that while the scalpel sign is useful for diagnosis, it is not necessarily correlated with the severity of symptoms. In contrast, the presence of transverse bands in our patients was less common, with 5 cases seen on preoperative imaging and only 1 on postoperative imaging. It should be noted that surgical changes could have made visualization of arachnoid transverse bands more difficult, with scarring and inflammation potentially obscuring identification.

Pathogenesis of secondary syrinx formation has been speculated to be caused by chronic derangement of CSF dynamics.^12,23 A case report found that SAWs presenting concomitantly with syringomyelia may feature rostral, progressive syrinx expansion, a finding potentially explained by continuously abnormal CSF flow.²⁴ In Fig. 1, myelograms demonstrate subarachnoid space abnormalities and impeded CSF flow secondary to the arachnoid web. SAWs have been observed to move in a pulsatile, balloon-like fashion by ultrasound and microscopy.²⁵ In Fig. 3C–F, examples of ultrasound imaging from 2 of our patients are seen; in both, pulsatile CSF motion was noted. Furthermore, previously conducted case series have demonstrated that 67%–85% of SAWs occur with an associated syrinx.^2,5 However, only 6 (33.3%) of our patients had syrinx preoperatively. This finding suggests that SAW presents heterogeneously, and clinical suspicion should remain high even in the absence of classic sequelae. Additionally, it should be noted that in patients with syringomyelia and SAW, residual symptoms can be seen even with radiographic syrinx resolution.

It has been shown that more than 90% of SAW patients respond positively to surgical intervention.⁵ Surgical technique for SAW resection has been detailed in the literature.²⁶ We detail our surgical protocol utilized in this series; for all patients, a combination of intraoperative fluoroscopy, ultrasound, and microsurgical techniques were used to facilitate resection. A 9-patient series found that surgical intervention was beneficial and the risk of postoperative neurological deficits was low even at long-term follow-up (3.2 years).^27,28 In our series, all patients were managed surgically, and 5 (27.7%) experienced worsening of one or more symptoms. However, 3 of these patients had new symptom worsening or new neurological deficit due to pathology unrelated to their operation. Therefore, in comparison to the existing literature, our case series supports the conclusion that surgery is beneficial for most patients. Furthermore, one of the variables differentiating our patients’ operations was laminectomy versus laminoplasty; however, our results do not suggest that this choice changes postoperative symptomatic course. Published evidence suggests that laminoplasty results in fewer CSF leaks and better range of motion after tumor resection, but these have not yet been compared in the context of arachnoid web or cyst resection.²⁹ In this study, most patients received laminoplasty, and there were no cases associated with CSF leakage. Further investigation would benefit the understanding of differences in operative technique. Lastly, given the low number of complications in the postoperative period and positive clinical courses after resection, limited long-term follow-up is indicated for these patients.

This study was limited by a low number of patients that precluded in-depth analysis due to low statistical power. Additionally, operations were performed by four different neurosurgeons, and operative techniques varied slightly case to case. Five patients were without postoperative MRI, although all had postoperative follow-up in the clinic.

Limitations

This study has several limitations inherent to its retrospective, single-institution design. First, the overall sample size was small, which limited the statistical power and precluded subgroup analyses. Additionally, nearly half of the cohort (8/18 patients) had less than 1 year of follow-up, which restricts our ability to assess long-term outcomes, including the potential for recurrent arachnoid webs, spinal cord tethering, or delayed neurological decline. Postoperative imaging was not performed in a standardized fashion and was often obtained based on surgeon preference or new or persistent symptoms, introducing variability in radiographic assessment.

Furthermore, standardized patient-reported outcome measures and other objective outcome measures were not collected, limiting our ability to objectively quantify symptom severity and response to surgery.

Lessons

SAWs are an uncommon but clinically significant cause of progressive myelopathy. Despite diagnostic challenges and radiographic subtleties, timely diagnosis and resection can yield substantial neurological and radiographic improvement. Our case series supports prior findings that microsurgical resection is a safe and effective treatment, even in patients with associated syringomyelia. While some patients experienced symptom persistence or unrelated neurological decline, most had stable or improved outcomes. These findings highlight the importance of recognizing SAW in patients with thoracic cord compression and incorporating myelography and ultrasound into preoperative and intraoperative workflows. Further multicenter studies are needed to refine the etiology of SAW, diagnostic criteria, and operative strategies.

Disclosures

Dr. Viljoen reported consulting fees and grants from Medtronic outside the submitted work. Dr. Khalsa reported personal fees from Globus outside the submitted work. Dr. Grossbach reported personal fees from Medtronic and grants from Medtronic and DePuy outside the submitted work.

Author Contributions

Conception and design: Ritchey, Eaton, Ladd, Viljoen, Khalsa, Xu, Grossbach. Acquisition of data: Ritchey, Jacob, Ladd, Khalsa, Xu. Analysis and interpretation of data: Ritchey, Jacob, Weinberg, Ladd, Viljoen, Khalsa, Grossbach. Drafting the article: Ritchey, Jacob, Weinberg, Khalsa. Critically revising the article: Ritchey, Weinberg, Ladd, Viljoen, Khalsa, Grossbach. Reviewed submitted version of manuscript: Ritchey, Weinberg, Ladd, Viljoen, Khalsa, Xu, Grossbach. Approved the final version of the manuscript on behalf of all authors: Ritchey. Administrative/technical/material support: Eaton. Study supervision: Weinberg, Eaton, Grossbach.

Correspondence

Nathan P. Ritchey: The Ohio State University College of Medicine, Columbus, OH. nathan.ritchey@osumc.edu.