Treating difficult-to-diagnose tight filum terminale: our experience with four patients

Abstract

Tight filum terminale (TFT) is a general term for pathological conditions that result in abnormal tension on the spinal cord, pulling the conus medullaris caudally. Because symptoms can vary, we aim to review the usefulness of Komagata’s criteria in our experience with four patients who had TFT that was missed in prior workups. We performed a retrospective review of the medical records of four patients who underwent resection of the filum terminale for TFT. A total of four patients underwent surgery. The patients’ chief complaints were lower back pain, lower limb pain and numbness. All patients met the Komagata diagnostic criteria for TFT and also had neurological abnormalities of the upper limbs, such as numbness and pathological reflexes. We resected the filum terminale in all patients, and achieved resolution of their preoperative symptoms. Komagata’s diagnostic criteria are seemingly useful for the diagnosis of TFT.

Background

It is not uncommon that patients with low back pain or lower limb pain have no visible abnormalities in the intervertebral discs or spinal canal on MRI or other imaging, making diagnosis difficult. Tethered cord syndrome is described in the literature as a condition in which the conus medullaris is pulled caudally due to abnormal tension of the filum terminale, causing symptoms such as pain, numbness and bladder-rectal dysfunction.^{1 2} Tight filum terminale (TFT) is a type of tethered cord syndrome, in which the conus medullaris is not anchored to the caudal side of the spinal cord and is present at the same level (L1/L2) as in healthy individuals. There is no anatomic abnormality on MRI, making diagnosis difficult even after the onset of symptoms.³

The onset of TFT is roughly divided into childhood onset and adult onset. Most patients with childhood onset have urological complaints such as incontinence, but almost all adult patients report low back pain and lower limb pain that does not correspond to a dermatome. In addition, adult patients often have symptoms related to the lumbosacral spinal cord, such as upper and lower limb numbness, bladder-rectal dysfunction and disorders of trunk flexion.^{1 4} There are few imaging features associated with TFT, making it difficult to reach a diagnosis through imaging alone.³ Komagata et al noted that cervical and trunk flexion in the sitting position exacerbates symptoms in patients with TFT, and that deflexion of the cervical spine alleviates symptoms in almost all patients. These findings are rarely seen in lumbar disc herniation (LDH) or lumbar spinal canal stenosis, leading the authors to propose the TFT provocation test (figure 1).¹ Komagata et al also created diagnostic criteria for TFT consisting of five items, including some frequently observed clinical features that are specific to TFT (table 1).¹ Patients with four or five positive findings may be confidently diagnosed with TFT. In a study of 134 patients who were diagnosed with TFT using these criteria, all had good results after resection of the filum terminale.¹

Figure 1.

TFT provocation test. 96.4% of TFT patients test positive. TFT, tight filum terminale.

Table 1.

Diagnostic criteria for TFT (Komagata et al)

1	Back or lower limb pain
2	Restriction of trunk anteflexion: Finger-floor distance ≥20 cm
3	Bladder-bowel dysfunction: (eg, micturition ≥8 times a day）
4	Neurological symptoms inconsistent with imaging findings
5	Positive TFT provocation test (figure 1)

The presence of ≧4 criteria confirms the diagnosis.

TFT, tight filum terminale.

We examine herein our experience with four patients who visited our hospital after having difficulty obtaining a diagnosis at outside institutions. We diagnosed TFT using Komagata’s criteria and performed resection of the filum terminale to improve patients’ symptoms. We report our experience with these patients and the results of a review of the literature.

Case presentation

Our patients were relatively young: a 26-year-old man, a 42-year-old man, a 36-year-old man and a 17-year-old woman. All patients visited our hospital in 2017 or 2018. All patients reported low back pain, lower limb pain and numbness in the upper and lower extremities persisting for more than 6 months. All patients had undergone MRI at an outside hospital with no findings consistent with their symptoms, making diagnosis difficult.

All patients had significantly compromised trunk flexion, frequent urination (≥8 times/day), exacerbation of symptoms with cervical anteflexion and trunk flexion, and relief of pain when cervical anteflexion was released. All patients had neurological symptoms of the upper limbs and two patients had weakness of the lower limbs that could not be explained based on their imaging findings. All patients satisfied Komagata’s diagnostic criteria for TFT. After making the diagnosis, we performed resection of the filum terminale in all four patients. All had good resolution of their symptoms by 2 years after surgery. Table 2 shows a summary of all patients.

Table 2.

Patient summary

	Patient 1	Patient 2	Patient 3	Patient 4
Age/sex	26 years/male	42 years/male	36 years/male	17 years/female
Symptoms	LBP, left lower limb pain, bilateral upper limb numbness	LBP, bilateral upper/lower limb numbness	Left hip joint pain with flexion	LBP, left lower limb muscle weakness/numbness
Time of symptom onset	six mo	10 year	1 year	6 months
Straight leg raise test	+/+	+/-	-/+	-/-
Femoral nerve stretch test	-/-	-/-	-/-	-/-
Lower limb MMT	Right: full/Left: all 4	Full	Full	Right: full/Left: all 2
Patellar tendon reflex	2/2	3/3	3/3	3/3
Achilles tendon reflex	2/2	4/4	4/4	2/0
Upper limb tendon reflex	N.P.	Hyper-reflexia	Hyper-reflexia	Hyper-reflexia
Pathological reflex of upper limb	N.P.	N.P.	Scapulohumeral reflex +/+ Wartenberg’s sign +/-	Wartenberg’s sign +/+ Hoffmann sign +/-
Frequent urination (>8 times/day)	+	+	+	+
Body stiffness (FFD >20 cm）	+	+	+	+
TFT provocation test	+	+	+	+
APR	0.27	0.43	0.28	0.09
MRI findings	No abnormalities
Postoperative symptoms at 2 years follow-up
Pain/numbness/frequent urination	Resolved
Reflex aberrance	Normalised
Lower limb MMT recovery	Left. 4→5	Full→full	Full→full	Left 2→5
Complication/recurrence	None

APR, anterior–posterior ratio; DTR, deep tendon reflex; FFD, finger floor distance; LBP, low back pain; MMT, manual muscle testing; TFT, tight filum terminale.

Representative case

A 36-year-old man had the chief complaints of pain with left hip flexion and urination more than eight times a day. He had a history of surgery at the L5/S1 level for LDH when he was 21 years of age. He began experiencing his current symptoms about 1 year prior to his presentation to our hospital, visiting multiple orthopaedic outpatient clinics where detailed examinations of the lumbar spine and hip joint were performed but no abnormalities were noted. On presentation to our hospital, examination revealed no lower limb weakness but he did have a positive straight leg raise test of the left leg, hyper-reflexia of the patellar tendons of both lower limbs, and clonus of the Achilles tendons on both side. Pathological reflex of upper limb such as Wartenberg sign and Scapulohumeral reflex was enhanced. The scapulohumeral reflex on both sides was hyper-reflexic. His finger-floor distance (figure 2) was greater than 20 cm, and his TFT provocation test was positive. Lumbar MRI showed disc degeneration at L5/S1, but there was no evidence of spinal canal stenosis or nerve root compression (figure 3A, B). Although the Patrick sign was positive at the left hip, MRI showed no lesions inside or outside the joint (figure 3C). No spinal cord compression lesions were observed on cervical and thoracic MRI (figure 3D–F). The anterior–posterior ratio (APR; figure 3F) of the subarachnoid spaces at the apex of the thoracic kyphosis curve on thoracic MRI was as low as 0.28, and the ‘short cut sign’ was observed in the spinal cord. Based on these findings, we suspected TFT and performed a resection of the filum terminale. After surgery, his symptoms resolved and the tendon reflexes of the upper and lower limbs normalised. He has not experienced any postoperative complications or recurrence of his symptoms.

Figure 2.

Finger floor distance (FFD). FFD is defined as the distance between the longest finger on the right side of the patient and the floor.

Figure 3.

A 36-year-old man with tight filum terminale (patient 3). (A) Lumbar spine MRI reveals degenerative disc findings after disc herniation at L5/S1, but no obvious spinal canal stenosis (B) MR neurography. The lower lumbar nerve roots follow a normal course, and no obvious foraminal stenosis is observed (C) left hip MRI. No obvious lesions are present inside or outside the hip joint (D), (E) cervical and thoracic spinal MRI. There is no obvious spinal canal stenosis. The anterior–posterior ratio (APR) of the subarachnoid spaces at the apex of the thoracic kyphosis on thoracic MRI (E, F; arrows) is as low as 0.28, indicating a spinal cord ‘short cut sign’ (F) APR is defined as ratio of the width of the anterior subarachnoid space (A) to the width of the posterior subarachnoid space (B).

TIP:

• All investigations that create a background (baseline) picture are relevant.

• All investigations that are crucial to management decisions should be discussed in full—include the limitations of investigations.

Choose appropriate images and videos to illustrate your point (maintaining patient confidentiality).

Differential diagnosis

TIP: Please do not list diagnoses. We want to understand how the final diagnosis was teased out. This is often the most important section and needs to be substantially discussed.

All working diagnoses need to be substantiated.

Treatment

Procedure for filum terminale resection

1. A lead electrode for somatosensory evoked potentials (SEP) is placed in the epidural space of the conus medullaris at the L1 level. Since the purpose of this surgery is to reliably identify the filum terminale and to avoid accidental resection of the cauda equina, placing the guiding electrode at the level of the Conus allows for the most reliable detection of SEP, as the stimulation and recording points are closer together than at any other site.¹

2. L5/S1 fenestration is performed in the prone position, and the exposed dura is incised (figure 4A).

3. The filum terminale is identified using SEP (the cauda equina nerve generates an SEP, but the filum terminale does not). Macroscopically, the filum terminale appears as a reddish cord (figure 4B).

4. The identified filum terminale is resected. To prevent haemorrhage and adhesion formation after surgery, bipolar electrocautery is used to achieve haemostasis of the filum terminale and to transect it on the caudal side. Rapid retraction of the separated end of the filum terminale to both poles is often observed (figure 4C).

5. To prevent postoperative leakage of cerebrospinal fluid, the dura mater is tightly sutured and an adhesive with fibrinogen and coagulation factor XIII (Beriplast P Combi-Set; CSL Behring, King of Prussia, Pennsylvania, USA) is applied (figure 4D).

Figure 4.

Surgical findings during resection of the filum terminale (patient 3) (A) Durotomy at the L5/S1 level. (B) Somatosensory evoked potentials are performed to detect the filum terminale. The filum terminale is observed as a reddish cord where somatosensory evoked potentials do not appear. The arrowheads indicate the filum terminale (C) dissection of the filum terminale (the arrowheads). To prevent bleeding and adhesions, the filum terminale is cauterised and then transected (D) closure of dura mater. The dura and soft tissues are sutured tightly to prevent postoperative leakage of cerebrospinal fluid.

Outcome and follow-up

All four patients experienced marked improvement of their preoperative low back pain, lower limb pain, and both upper and lower limb numbness. Their urinary frequency and pathological reflexes including upper limb all normalised during the postoperative course. In the two patients who had weakness of the lower limbs, their muscle strength was restored to normal by the time of final follow-up, 2 years after surgery (table 2).

Discussion

All four of our patients had TFT that developed in late adolescence or adulthood. Adult-onset TFT reportedly develops as follows. (1) A result of fibrosis of the filum terminale, gradually progressing as the body matures; elasticity of the filum terminale is gradually lost. (2) When height increases rapidly during adolescence, tension in the filum terminale increases and TFT develops. (3) The condition can also be triggered by increased physical activity that puts strain on the spinal cord, such as sports and work activities. (4) Or it can develop as a result of age-related spinal degeneration that limits vertical movement of the spinal cord and filum terminale.^{3 4} Some of our patients were engaged in relatively heavy-duty work such as truck driving and dairy farming, and others developed TFT after playing intense sports such as basketball. Our observations support the above purported mechanisms of TFT development.

Patients with TFT experience exacerbation of symptoms with cervical flexion and trunk flexion and experience relief of symptoms when cervical flexion is released. The specificity of these findings is higher when patients are sitting than when they are standing. The TFT provocation test, which uses this phenomenon to induce symptoms, is a simple and highly accurate diagnostic test devised in 1996 by Komagata et al. They conducted this test in multiple patients with TFT, LDH and lumbar spinal canal stenosis, and in healthy subjects. The positive rate in the sitting position was 96.4%, 3.3%, 0% and 0%, respectively, with excellent sensitivity and specificity for TFT.¹ The TFT provocation test in the sitting position was positive for all 4 of our patients, and we considered it useful for diagnosis.

In the clinical practice, distinguishing between TFT and LDH is often problematic. Both conditions have a relatively young onset, and both often feature low back pain and lower limb pain characteristic of sciatica, along with a positive straight leg raise test. However, TFT is distinguished by a high frequency of dysuria, less disc degeneration and swelling on MRI, and a positive TFT provocation test.¹ One of our patients had mild LDH seen on MRI at another hospital, but we were able to make the diagnosis of TFT based on the results of the TFT provocation test.

Half of our patients had pathological reflex of upper limb such as Wartenberg sign, Scapulohumeral reflex and Hoffmann sign, that was normalised after the surgery. In the literature, the frequency of occurrence of these abnormal reflexes in normal subjects is about 3%–5%,^{5 6} but in this study, about half of them were observed, and it was confirmed that the pathological reflex findings of upper limb were specific to TFT.

Researchers speculate that these findings result from the pathological condition of traction myelopathy, mainly in the conus medullaris but also extending to the superior spinal cord due to the reduced reserve capacity for spinal cord traction caused by the lack of elasticity in the filum terminale.¹ When TFT is suspected, it is important to include the upper limbs in the physical and neurological examination.

Although many attempts have been made to establish an imaging diagnosis of TFT, no findings have proven definitive.³ In patients with tethered cord syndrome, the filum terminale is often thickened to a diameter of 2 mm or more.⁷ Thickening of the filum terminale is observed in patients with TFT, a subtype of tethered cord syndrome, whose conus medullaris is at the normal level.^{2 8} However, Komagata et al note that the diameter of the filum terminale was 1.5 mm in 70% of their patients who underwent filum terminale resection.^{1 3} This indicates that the onset of symptoms may be due to the lack of elasticity of the filum terminale, regardless of its thickness.^{1 3} Komagata et al divided their patients into two groups: those with a filum terminale diameter of less than 1.5 mm and those with a diameter of more than 1.5 mm observed at surgery. They compared the severity of the preoperative JOA (Japan Orthopaedic Association) score with the Japan Orthopaedic Associatio 6 months after surgery and found no correlation in symptom improvement with the thickness of the filum terminale.¹ These results suggest that the difference in filum terminale findings on lumbar spine MRI and macroscopic findings have little diagnostic significance.

The spinal cord ‘short cut sign’ has been reported as a useful diagnostic imaging finding for TFT. In this phenomenon, the spinal cord is pulled by the TFT and appears to be cut short at the apex of the thoracic kyphosis, the shortest distance in the spinal canal (figure 3E, F). Komagata et al calculated the APR, the ratio of the anterior and posterior subarachnoid space at the apex of the thoracic kyphosis on thoracic MRI, and compared the results in a large number of patients with and without TFT. They found that the short cut sign is a significant finding for TFT: the spinal cord is significantly anteriorly displaced in the TFT group, with an average APR of 0.28 vs 0.83 in the non-TFT group (p=0.03). Use of the APR can lead to a more accurate diagnosis when combined with clinical symptoms.¹ In our patients, the average APR at the thoracic kyphosis level was 0.27, supporting the results of Komagata et al.

Patients with TFT experience symptoms because of ischaemic changes when the blood flow in the spinal cord is decreased due to traction on the spinal cord.⁹ Remission of symptoms can be easily achieved by adjusting the posture. Irreversible changes are unlikely to occur, and postoperative improvement is equally good regardless of the age of onset or length of illness.¹ One of our patients had 10 years of symptoms before finding relief after surgery.

Each of our four patients experienced difficulty with diagnosis at other hospitals, leading to a delay in treatment. They had severe pain, urinary incontinence and lower limb weakness that made it impossible to continue working or attending school. Some of our patients were forced to retire or take leaves of absence. Providing appropriate and prompt treatment after obtaining a definite diagnosis is important for improving the pathological and social prognosis of the disease.

Patient’s perspective.

I had severe hip pain when flexed, so I went to several hospitals but was not diagnosed. I've been suffering from pain for over a year. The pain was so strong that I was unable to continue working, but I was diagnosed with Tight Filum Terminale, and for the first time I knew my disease name. I had an operation and fortunately the preoperative pain disappeared completely. Three years have passed since the operation, but I am grateful to have been able to spend my days happy without any symptom recurrence. I eagerly hope that this disease will be recognised and that patients with the same disease will be saved. (Patient 3)

Learning points.

• When a young patient has low back pain and lower limb pain, no abnormalities on imaging, and frequent urination, the diagnosis of tight filum terminale (TFT) should be taken into consideration.

• It seems that Komagata’s diagnostic criteria, especially the TFT provocation test, are useful for the diagnosis of TFT.

• Many patients with TFT also have numbness and pathological reflexes in the upper limbs; together with the spinal short cut sign observed on thoracic MRI, these findings are useful to suggest the diagnosis of TFT.