Delayed myelopathy caused by cerebrospinal fluid pseudocyst following decompression for thoracic ossification of the ligamentum flavum: a case report and literature review

Abstract

Background

Delayed neurological deterioration caused by a postoperative high-pressure cerebrospinal fluid (CSF) collection is an uncommon but clinically recognized complication after thoracic decompression complicated by incidental durotomy. We report a case of delayed thoracic myelopathy due to a tension subfascial CSF collection and summarize relevant literature.

Case presentation

A 55-year-old woman with thoracic myelopathy due to ossification of the posterior longitudinal ligament at T2–4 and ossification of the ligamentum flavum at T5–6 underwent posterior decompression and instrumented fusion (T2–6 laminectomy with T1–6 instrumentation). An incidental dural tear occurred at T5–6 and was managed with an onlay artificial dural patch reinforced with fibrin glue and gelatin sponge. The wound drain was removed on postoperative day 7. On postoperative day 9, she developed acute left lower-extremity weakness. MRI demonstrated a large subfascial CSF collection causing significant spinal cord compression, consistent with a tension pseudomeningocele. Bedside subfascial incision and continuous closed low-level drainage (with the drainage bag kept at bed level) resulted in rapid neurological improvement. After recurrence following drain removal, ultrasound-guided drainage was performed, leading to sustained symptom relief. At 3-month follow-up, she had full motor recovery and follow-up MRI confirmed resolution of the collection with adequate decompression.

Conclusions

A tension subfascial CSF collection after incidental durotomy can mimic recurrent stenosis and cause delayed thoracic myelopathy. When imaging confirms a compressive CSF collection, continuous closed low-level subfascial drainage may be considered before revision surgery.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12891-026-09786-x.

Introduction

Thoracic Ossification of Ligamentum Flavum, TOLF, is a degenerative disease characterized by heterotopic ossification of the ligamentum flavum and is one of the most common causes of thoracic spinal stenosis [1–4]. The pathological process involves the gradual transformation of the fibrous tissue of the ligamentum flavum into a bony structure, which in turn compresses the spinal cord or nerve roots, triggering motor-sensory deficits, sphincter dysfunction, and even paraplegia in the lower limbs [5, 6]. In recent years, with the aging of the population and the development of imaging technology, the clinical diagnosis rate of TOLF has increased significantly; however, its complex pathogenesis, diversified clinical manifestations, and high-risk surgical treatment remain major challenges in spine surgery. Although surgical decompression remains the mainstay of treatment, intraoperative dural tears and postoperative cerebrospinal fluid (CSF) leakage are major concerns [1, 7]. Typically, incidental durotomies are managed intraoperatively with primary repair or patch techniques, and postoperative CSF leakage resolves within a few days. However, in some instances, persistent CSF leakage or a hypothesized valve-like mechanism at the repair site can lead to the formation of a tension pseudomeningocele. Unlike routine CSF leaks, this condition allows fluid to accumulate under high pressure, causing delayed spinal cord compression and catastrophic neurological deterioration days or weeks after the initial recovery.

This phenomenon represents a serious iatrogenic complication that mimics recurrent stenosis but requires a fundamentally different management strategy. In this report, we present a case of delayed thoracic myelopathy caused by a tension subfascial CSF pseudocyst following TOLF decompression. We discuss the potential pathogenesis of this complication, including progressive CSF accumulation, and evaluate the efficacy of continuous subfascial drainage as a practical and frequently employed management option to avoid the risks associated with high-risk revision surgery.

Case presentation

A 55-year-old woman presented to our outpatient clinic with a 3-month history of progressive numbness and weakness in both lower extremities, accompanied by an unsteady gait. The symptoms had an insidious onset and gradually worsened over time.

On admission, neurological examination was consistent with thoracic myelopathy. Motor testing demonstrated symmetric paraparesis with Medical Research Council (MRC) grade 4/5 strength in both lower limbs. Deep tendon reflexes were brisk in the knees and ankles bilaterally. Sensory examination revealed an abnormal pinprick sensation and hyperalgesia over the trunk and both lower extremities. The tandem gait was unstable.

Computed tomography with three-dimensional reconstruction demonstrated degenerative changes in the thoracic spine, including osteophytic proliferation along the posterior aspect of T2–4 and degenerative changes involving the right facet joints at T5–6, with bony spinal canal stenosis. Magnetic resonance imaging (MRI) revealed ossification of the posterior longitudinal ligament at T2–4 and TOLF at T5–6, resulting in thoracic spinal cord compression (Fig. 1). Under general anesthesia, the patient was positioned prone. Through a posterior midline approach, the posterior elements from T1 to T6 were exposed. Bilateral pedicle screws were inserted from T1 to T6 under fluoroscopic guidance as a standard surgical precaution to prevent iatrogenic spinal instability and progressive post-laminectomy kyphosis following the extensive decompression. A T2-6 laminectomy was performed, and the ossified ligamentum flavum at T5-6 was resected using rongeurs and an ultrasonic bone knife. During decompression at T5-6, severe adhesions and friable dural tissue led to an incidental dural tear. This resulted in a visible non-watertight defect measuring approximately 0.5–0.8 mm. Although the initial tear appeared small, the surrounding dural tissue was severely thinned and compromised due to the chronic inflammatory adhesions associated with TOLF. Consequently, the functional defect was likely larger or highly prone to expansion under continuous postoperative CSF hydrostatic pressure, rendering primary watertight suturing unfeasible. Primary watertight suturing was judged not feasible because the dura mater is ossified and severely adhered. An onlay dural substitute patch (BIOFILM; Tianxinfu Medical Appliance Co., Ltd., Beijing, China) was applied to the defect site and reinforced with fibrin sealant (Human Freeze-Dried Fibrin Sealant; Shanghai RAAS Blood Products Co., Ltd., Shanghai, China) and a gelatin sponge (Hemostatic Sponge; Guilin Fukangsen Medical Devices Co., Ltd., Guilin, China). A standard closed-suction Jackson-Pratt (JP) type drain (16-Fr, NANTONG SANLI MEDICAL DEVICES CO., LTD., NANTONG, CHINA) was placed and the wound was closed in layers. Operative time was 263 min and estimated blood loss was 350 mL.

Fig. 1.

Preoperative imaging. Mid-sagittal CT reconstruction showing the “beak-type” ossification of the ligamentum flavum (OLF) at T5-6 level causing stenosis. Sagittal T2-weighted MRI demonstrating spinal cord compression (high signal intensity) corresponding to the ossified segment

Postoperatively, neurological function improved. To manage the subfascial collection, the drain was maintained strictly in a gravity drainage mode, with the collection bag positioned at bed level according to our specific care protocol. Daily drain outputs and related clinical symptoms are summarized in Table 1. The drain was subsequently removed on postoperative day (POD) 7 once the output criteria of < 50 mL/24 h were met [8]. Upon drain removal, formal suture closure of the drainage tract was performed to prevent persistent cerebrospinal fluid fistulas. Following this procedure, the patient initiated ambulation with a brace on POD 7, demonstrating improved gait stability. However, on postoperative day 9, she developed new-onset left lower extremity weakness, with hip flexor strength decreased to MRC grade 2 and knee extensor strength to MRC grade 3. She was treated with methylprednisolone (40 mg daily) and mannitol (125 mL every 8 h). Physical examination revealed a palpable, fluctuant swelling at the incision site. Repeat thoracic MRI demonstrated a large subfascial fluid collection with significant spinal cord compression (Fig. 2), consistent with delayed myelopathy secondary to postoperative fluid accumulation.

Table 1.

Daily drain output and clinical status

POD	Volume of drainage (ml)	Appearance	Wound Fluctuation	Clinical Manifestation
1	150	Blood-tinged	No	Regular
2	100	Blood-tinged	No	Emesis
3	110	Blood-tinged	No	Regular
4	100	Blood-tinged	No	Regular
5	80	Blood-tinged	No	Regular
6	30	Blood-tinged	No	Regular
7	50 (Drain Removal)	Blood-tinged	No	Regular
8	-	-	No	Regular
9	300 (Re-inserted)	Blood-tinged	Yes	Neurological deterioration
10	50	Blood-tinged	No	Regular
11	80	Blood-tinged	No	Regular
12	150	Blood-tinged	No	Fever
13	180	Clear	No	Fever
14	220	Clear	No	Regular
15	300	Clear	No	Headache
16	280	Clear	No	Regular
17	270	Clear	No	Regular
18	250	Clear	No	Regular
19	260	Clear	No	Regular
20	220	Clear	No	Regular
21	250	Clear	No	Regular
22	280	Clear	No	Regular
23	190	Clear	No	Regular
24	80	Clear	No	Regular
25	50 (Drain Removal)	Clear	No	Regular
26	100 (Re-inserted)	Blood-tinged	Yes	Hypokinesia
27	110	Clear	No	Regular
28	50	Clear	No	Regular
29	50	Clear	No	Regular
30	40	Clear	No	Regular
31	70	Clear	No	Regular
32	40	Clear	No	Regular
33	40 (Drain Removal)	Clear	No	Regular

Fig. 2.

Postoperative sagittal T2-weighted MRI obtained on day 9 after symptom deterioration. The image reveals a massive, hyperintense fluid collection (pseudocyst) located dorsal to the thecal sac at the laminectomy site. Note the significant mass effect compressing the spinal cord anteriorly, confirming the diagnosis of a tension pseudomeningocele causing delayed myelopathy

At the bedside, following disinfection, the incision was made using a scalpel No. 11, scissors, and hemostat at the point of maximal fluctuation on the right side of the surgical wound under local anesthesia (5 mL lidocaine). The deep fascia was opened, and the collection was accessed in the subfascial plane over the laminectomy bed. A standard closed-suction Jackson-Pratt (JP) type drain (16-Fr, NANTONG SANLI MEDICAL DEVICES CO., LTD., NANTONG, CHINA) was inserted to a depth of 6 cm. Upon opening the deep fascia, there was an immediate high-pressure release of blood-tinged fluid, and approximately 300 mL (estimated based on the volume collected in the drainage pack) drained rapidly within approximately 10 min. Immediately thereafter, drainage was converted to a controlled closed system using gravity-only, with the collection bag maintained at bed level to limit siphoning and an intermittent clamping schedule. The patient was continuously monitored for neurological status and symptoms of intracranial hypotension (postural headache, nausea/vomiting, altered consciousness, and pupillary changes). The following day, left knee extensor strength improved to MRC grade 4 (hip flexor strength remained MRC grade 2). The drain was maintained and subsequently removed 17 days later (left lower-limb strength had recovered to MRC grade 5). About 10 h after the drainage was removed, the patient developed recurrent symptoms (hip flexor strength decreased to MRC grade 4 and knee extensor strength to MRC grade 4). Emergency ultrasound-guided percutaneous catheter drainage was performed. A standard closed-suction Jackson-Pratt (JP) type drain (16-Fr, NANTONG SANLI MEDICAL DEVICES CO., LTD., NANTONG, CHINA) was inserted to a depth of 5.5 cm into the collection under ultrasound guidance and managed with gravity, bag position maintained at bed level, and an intermittent clamping schedule. Daily outputs are shown in Table 1. The catheter was removed after 7 days when < 50 mL/24 h were met. Thereafter, back pain was relieved, and gait as well as lower-limb numbness/weakness improved substantially.

At the 3-month follow-up, no recurrence of neurological symptoms was noted. The patient reported no back pain, bilateral lower-limb strength had recovered to MRC grade 5, and follow-up MRI confirmed the resolution of the fluid collection with adequate spinal cord decompression (Fig. 3). A timeline of key events is provided in Fig. 4.

Fig. 3.

Follow-up MRI at 3 months postoperatively. The fluid collection has completely resolved, and the spinal cord is adequately decompressed with no signs of recurrence

Fig. 4.

Timeline of the clinical course. The timeline summarizes surgical intervention, postoperative recovery, neurological deterioration, drainage management, and eventual full recovery during the patient’s clinical course

Discussion

Postoperative cerebrospinal fluid leakage after surgery for TOLF is a non-negligible complication in clinical practice, with a significantly higher incidence than that after cervical and lumbar spine surgeries. In general, cerebrospinal fluid leakage resolves within 4–5 days after surgery. This phenomenon is corroborated by an in vitro histological study, which found that dural healing takes 4 days [9]. A literature review of postoperative CSF leak and pseudomeningocele management strategies identified a broader array of recent high-impact studies, revealing considerable variation in drainage protocols. Reference lists of included articles were manually screened. The nine included publications are summarized in Table 2 [10–18]. A primary divergence exists regarding the use of negative pressure. While Mayeku et al. [19] advocated for subfascial drains with partial suction for 48 h to manage intradural tumor surgeries, recent evidence increasingly warns against active suction. For instance, Zhai et al. [20], Ropper et al. [18], and Bauerle et al. [15] utilized closed subfascial drains without suction, regulating flow passively to prevent exacerbating the dural tear. Bauerle et al. [15] explicitly utilized a passive bile bag system, demonstrating safety and a remarkably low reoperation rate. In our initial management, we adopted a gravity-only, non-suction closed system utilizing a standard Jackson-Pratt type drain, aligning with this passive drainage philosophy to avoid enlarging the fistula.

Table 2.

Summary of the current case and reviewed literature regarding symptomatic spinal pseudocysts causing delayed neurological deterioration

Author/Year	Patient Age/Sex	Diagnosis (TOLF/OPLL)	Onset Time (Days post-op)	Mechanism	Treatment (Drainage vs. Revision)	Outcome	Ref
Current Case	55 / F	OPLL T2–4 + OLF T5–6	9 days	Ball-valve + Tension Pseudocyst	Bedside Drainage	Recovery	-
Wada (2024)	73 / F	Thoracic Tumor	60 days	Ball-valve	Revision Surgery	Recovery	[13]
Macki (2014)	57 / F	Thoracic Stenosis T10-11	60 days	Pseudomeningocele	Revision Surgery (Primary Repair)	Recovery	[12]
Weng (2010)	Series (n = 11)	Cervical Herniation	mean 50.6 days	Giant Pseudomeningocele	Revision Surgery+ subarachnoid drainage	Recovery	[14]
Enke (2018)	64 / F	Lumbar Stenosis	30 days	Giant Pseudomeningocele	Revision Surgery+ Prolonged Drainage	Recovery	[10]
Helle (1981)	48 / M	Cervical Disc	21 days	Ball-valve	Surgical Excision	Recovery	[11]
Bauerle (2024)	Series (n = 108)	Spine Surgery	Intraoperative	Incidental Durotomy	Subfascial Passive Bile Bag Drainage	Recovery	[15]
Jang (2023)	Series (n = 14)	Cervical OPLL	Intraoperative	Intraoperative CSF Leak	Pump-Regulated Continuous Lumbar Drainage	Recovery	[16]
Lenschow (2022)	Series (n = 17)	Intradural Spinal Surgery	Early post-op	Planned Durotomy	Epidural drainage + Early mobilization	Recovery	[17]
Ropper (2018)	Series (n = 12)	Extradural Spinal Tumor	Intraoperative	Intraoperative Durotomy	Closed Subfascial Gravity Drainage	Recovery	[18]

Abbreviations: TOLF Thoracic ossification of the ligamentum flavum, F Female, M Male, CSF Cerebrospinal fluid

The table highlights the onset time of delayed myelopathy and the comparison of management strategies. While revision surgery (e.g., primary repair, fat graft) is the most common treatment reported in the literature, the current case demonstrates successful resolution using a minimally invasive bedside drainage technique

Another major divergence involves continuous versus intermittent drainage. Mammadkhanli et al. [21] successfully utilized an intermittent clipping technique, clamping the drain for 4 h and opening it for 10 min, to allow passive drainage while minimizing intracranial hypotension. In contrast, cases requiring complete CSF diversion often rely on continuous lumbar subarachnoid catheters [13, 14]. Recently, Jang et al. [16] introduced pump-regulated volumetric continuous lumbar drainage to strictly prevent the abrupt large-volume overdrainage inherent to gravity-dependent systems. During our emergency bedside rescue, we initially allowed brief decompression of 300 mL to immediately relieve spinal cord compression; however, we swiftly transitioned to a controlled gravity system combined with an intermittent clamping schedule similar to that described by Mammadkhanli et al. [21], safely converting the high-pressure tension cyst into a low-pressure subfascial environment.

Initially, our drain management relied on a globally recognized output-driven protocol, removing the subfascial drain on postoperative day (POD) 7 based on a < 50 mL/24 h criterion. This specific threshold is widely accepted as the standard of care in routine spinal surgeries to balance the prevention of compressive hematomas with the mitigation of retrograde infection risks [22–24]. In standard cohorts, the complication or failure rate of this protocol is exceptionally low (typically < 2–5%) [23].

However, the catastrophic neurological decline observed in our patient mandates a critical analysis of why this standard protocol proved insufficient. We attribute this failure not to the drainage duration itself, but to the unique pathoanatomy of this specific case. The severe ‘beak-type’ TOLF morphology and chronic inflammatory adhesions resulted in an irregular, non-watertight dural defect with extremely poor local tissue quality. Unlike simple, linear incidental durotomies that can achieve adequate fascial scarring within 7 days, this compromised defect was biologically incapable of healing in that standard timeframe. Consequently, once the drain was removed based on the < 50 mL criterion and the patient mobilized, the unhealed fascial tract was exposed to a sudden spike in local hydrostatic pressure. Once the superficial skin and fascial layers sealed, the persistent CSF leakage naturally accumulated within the closed subfascial space, progressively enlarging and eventually producing spinal cord compression. Furthermore, we hypothesize that a one-way ‘ball-valve’ mechanism may have also contributed to entrapping CSF and rapidly forming a high-pressure tension pseudocyst. This case highlights a critical caveat for spine surgeons: standard volume-driven drain removal criteria (e.g., < 50 mL/24 h) are unreliable and potentially dangerous in the presence of complex, non-healable dural defects caused by severe ossified pathologies. Enke et al. [10] experienced a similar massive recurrence when removing subcutaneous drains on POD 5, ultimately requiring 4 weeks of prolonged suction drainage with stricter removal criteria (< 15 mL/24 h).

Regarding mobilization, Mayeku et al. [19] and Lenschow et al. [17] suggested that early ambulation (as early as POD 1) might not increase CSFL rates in selected uncomplicated cases. However, for complex or recognized leaks, extended bed rest remains a cornerstone of therapy. Milton et al. [25] and Bauerle et al. [15] emphasize strict head-of-bed flat protocols to manage hydrostatic pressure. Drawing upon our patient’s rapid neurological decline upon mobilization and the strict two-week supine bed rest protocol mandated by Wada et al. [13]. Regarding mobilization, previous prospective data [25] suggesting the feasibility of early ambulation must be cautiously contextualized. Based on the neurological decline observed in our patient, we suggest that the high-pressure nature of a leak through a non-watertight defect necessitates extreme caution regarding early upright mobilization. The hydrostatic pressure generated by an upright posture easily overwhelms the fragile, unhealed fascial and artificial dural layers, directly precipitating tension pseudocyst formation.

Therefore, we clarify that the ‘selected cases’ suitable for early mobilization are strictly limited to uncomplicated cases with small, linear incidental durotomies, in which primary, definitively watertight suture closure was achieved intraoperatively. Conversely, cases involving complex pathologies, such as TOLF, in which severe adhesions and fragile tissue preclude watertight closure and necessitate on-lay patching, require strict recumbency (flat bed rest). Prolonged immobilization, combined with controlled low-level drainage, is essential to minimize hydrostatic pressure and allow profound fascial scarring before the patient resumes an upright posture.

Although our drainage management achieved rapid neurological improvement, we do not advocate rapid or uncontrolled decompression as a routine approach, nor do we present bedside drainage as a universal first-line treatment. Instead, the clinical message is twofold: (1) achieving the most watertight dural repair feasible at the index surgery should remain the primary goal; and (2) when watertight closure cannot be achieved, clinicians should anticipate delayed high-pressure compressive collections and adopt explicit postoperative protocols (controlled drainage configuration, objective removal criteria supplemented by clinical assessment, and close neurological monitoring), with early surgical re-exploration remaining warranted when neurological deterioration persists, a CSF fistula is uncontrolled, or infection is suspected.

Conclusion and future directions

Delayed tension pseudomeningocele is an uncommon but clinically recognized and serious postoperative complication after surgery for TOLF. Because this was a single-case report, our observations should not be interpreted as establishing the safety, efficacy, or generalizability of any specific drainage strategy. The principal clinical message is that non-watertight dural repair may predispose to the delayed formation of a high-pressure compressive subfascial CSF collection, and prompt imaging and timely decompression are critical when neurological deterioration occurs. Controlled closed drainage may be considered only as a temporizing, individualized option in selected patients under close monitoring, whereas early surgical re-exploration remains appropriate when deficits persist, CSF fistulas are uncontrolled, or infection is suspected.

Authors’ contributions

All authors contributed to the conception and design of the study. Zheng and Wang performed material preparation, data collection, and analysis. The first draft of the manuscript was written by Shuxin Zheng, and all the authors commented on the previous versions of the manuscript. All authors have read and approved the final manuscript. Shuxin Zheng and JianZhi Wang contributed equally to this work and shared first authorship. Lei Wang and Linnan Wang shared their corresponding authors.

Funding

Screening System for scoliosis in Children and Adolescents in Ngari, Tibet (JH2024047).

Data availability

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

Declarations

Ethics approval and consent to participate

This study was conducted as a retrospective review of anonymized clinical data. Formal ethical approval was waived by the Ethics Committee of West China Hospital, as the study involved no intervention and posed minimal risk to participants. Written informed consent for participation was obtained from the patient.

Consent for publication

Written informed consent for publication of the patient’s clinical details and accompanying images was obtained from the patient. A copy of the written consent is available for review by the Editor-in-Chief of this journal.

Competing interests

The authors declare no competing interests.

Clinical trial number

Not applicable.