Arteriovenous fistula of the filum terminale masqueraded as a failed back surgery syndrome - A case report and review of literature.

BACKGROUND: The filum terminale arteriovenous fistulas (FTAVFs) are a very rare type of spinal vascular malformation. Clinically, these lesions could present with a progressive ascending myelopathy also called FoixAlajouanine syndrome. Due to the rarity of these vascular malformation, some can be misdiagnosed, submitted to unnecessary spinal surgery, and even masqueraded as a failed back surgery syndrome. Based on the present case and related literature, we review all the cases with similar history and describe factors that should raise awareness for diagnosis of this spinal vascular malformation. CASE DESCRIPTION: We present a case of a patient with a FTAVF at the level of L5-S1 that presented with a FoixAlajouanine syndrome. He had been previously submitted to a lumbar decompressive laminectomy without sustained improvement. After the identification and surgical treatment of the vascular malformation, he had progressive neurological improvement.
CONCLUSION: FTAVF is a very rare spinal intradural spinal vascular malformation that can be masqueraded as a failed back surgery syndrome. In these cases, signs of ascending myelopathy should prompt awareness and vascular voids must be carefully evaluated in MRI. Copyright:

INTRODUCTION

The filum terminale arteriovenous fistulas (FTAVFs) are a very rare type of spinal vascular malformation accounting for approximately 3% of cases.[2,6,9,22] They are generally characterized by a single direct shunt between the artery of the filum terminale (distal prolongation of anterior spinal artery [ASA]) and a draining vein.[6,9,24,27] These lesions cause a congestive myelopathy and progressive spinal cord ischemia due to spinal cord venous congestion, vascular hypertension, and a putative chronic steal phenomenon.[6,14,24] Clinically, these lesions could present with a progressive ascending myelopathy also called Foix-Alajouanine syndrome.[3,6,9,14,16] This syndrome can encompass motor weakness, sensory disturbances, back pain, sphincter disturbances, and sexual dysfunction.[3,6,14,16,18] Due to the rarity of FTAVF and the nonspecific clinical manifestations, some of these cases have been misdiagnosed, undergone lumbosacral surgery, and subsequently, labeled as failed back surgery syndrome.[10,11,14,15,25,28,29,31,32]

We describe a case of a patient with a FTAVF presented with a Foix-Alajouanine syndrome that had been previously misdiagnosed and had undergone a lumbar laminectomy for lumbar stenosis. We review all the FTAVF with similar history described in the literature. This literature search was performed on PubMed using the key words “arteriovenous fistula of the filum terminale,” “FTAVF,” “previous surgery,” “lumbosacral surgery,” and “failed back surgery syndrome.” We also describe factors that should raise awareness for diagnosis of this spinal vascular malformation.

CASE DESCRIPTION

A 73-year-old man was evaluated at the neurosurgical outpatient clinic with an 18-month history of intermittent back pain and progressive gait disturbance due to diminished strength in both legs. Seven years before, the patient had similar neurogenic claudication as main symptom and a lumbar CT scan was done solely, which showed stenosis at the L4-L5 level as the key feature. He underwent lumbar decompressive laminectomy in another institution, with transient improvement at that time.

During the following months, the symptoms slowly progressed and the patient became unable to walk without assistance. He also developed urinary sphincter disturbance with incontinence episodes. The neurological examination showed a bilateral lower extremity motor weakness with a motor strength grade 3/5 bilateral in hip flexion and extension, and a Grade 4/5 bilateral in knee extension, ankle dorsiflexion, knee flexion, great toe extension, and ankle plantar flexion. The patient also had hypoesthesia from L4 dermatome downward bilaterally, abnormally increased patellar and ankle reflexes bilaterally, and extensor plantar reflexes.

A lumbar MRI was performed, demonstrating prominent vascular voids near conus medullaris but no significant high T2 signal involving the spinal cord [Figure 1a and b]. Further evaluation was undertaken with CT angiogram and digital subtraction angiography (DSA). These examinations identified a FTAVF which was fed by the filum terminale artery that arose from the ASA and were supplied by the left D10 segmental artery [Figure 1c and d]. The patient also had an electromyographic study done that reveals chronic neurogenic changes of L3, L4, L5, and S1 bilaterally.

Figure 1:

Preoperative images: (a) MRI T2 sagittal showing multiple vascular voids near conus medullaris (arrow). (b) MRI T2 axial revealing two hypointense structures on the left (arrow) that were confirmed to be the arterial feeder and draining vein of the filum terminale arteriovenous fistula (FTAVF) after the DSA. (c) CT angiogram showing the contrast enhancement of the FTAVF (arrow) and multiple vascular voids around the conus medullaris and lower spinal cord (arrow head). (d) Dorsolumbar DSA showing the FTAVF at L5-S1 (arrow) with a descending arterial feeder from left D10 and an ascending draining vein.

Therapeutic options were discussed with the patient, who decided to undergo surgery. Reopening and extension of the previous laminectomy down to S1 was performed, to expose the fistula. After durotomy, the direct connection between the artery of the filum terminale and the arterialized vein was evident, forming the arteriovenous fistula (AVF) at the level of L5-S1 [Figure 2a]. With the aid of intraoperative neurophysiological monitoring, we confirm the filum terminale due to the absence of sensory-motor potentials on it [Figure 2b and c]. An intraoperative angiography with indocyanine green (ICG) was performed to identify the exact shunt location [Figure 2d]. After confirmation of the AVF, a clip was placed interrupting the fistulous path, followed by the observation that the draining vein was no longer arterialized [Figure 3a]. This interruption was confirmed with the ICG video angiography [Figure 3b].

Figure 2:

Intraoperative images: (a) image of the artery of the filum terminale directly connected to an arterialized vein, which corresponds to the arteriovenous fistula (arrow). The normal filum terminale (asterisk) is seen distally (left correspond to cranial and right is caudal). (b) Intraoperative neurophysiological examination testing a cauda equina root and (c) the filum terminale. (d) Intraoperative ICG video angiography confirming the exact shunt location of the filum terminale arteriovenous fistula.

Figure 3:

(a) After clip placement on the filum terminale arteriovenous fistula, the draining vein is no longer arterialized. (b) Intraoperative indocyanine green video angiography confirming the interruption of the fistulous shunt.

Surgery and immediate postoperative period were uneventful. Over the following weeks, both sensory and motor deficits, progressively improved. Ten days after surgery, the patient recovered the voluntary control of the urinary sphincter and the urinary catheter was removed. He underwent physiotherapy and was able to walk unassisted 6 months after surgery.

DISCUSSION

Spinal intradural AVFs represent a rare spinal vascular malformation category.[25] According to Kim and Spetzler classification, those can be divided into dorsal and ventral intradural AVFs, being the last one further subclassified into type A, B, and C depending on their size, arterial feeders, and dilatation of draining veins.[13] The FTAVFs are a rare type of spinal intradural AVFs, accounting for approximately 3% of spinal intradural vascular malformations.[2,6,9,22] These are generally characterized by a single direct shunt between the artery of the filum terminale (distal prolongation of ASA) and a draining vein.[1,3,6,8,9,12,18,29,33] Rarely, these FTAVF can have multiple shunt points.[6,8,9] Because of its different morphology and the unique location along the filum terminale, some authors advocate that FTAVF should be classified as a distinct form of intradural arteriovenous shunt.[6]

The etiology of FTAVFs is unknown. Some authors consider it as a congenital malformation,[11,23] others believe that it is an acquired lesion.[16,31,32] Supporting the congenital nature is that some FTAVFs are associated with neural tube defects such as tethered cord,[28] diastematomyelia,[19] syringomyelia,[1] hypoplastic sacrum,[1] and spina bifida[18] suggesting an association of these during the embryonic period. In our review, we found 20 cases associated with neural tube defects.[5,19,28] The acquired etiology is supported by the fact that most patients are too old to considerer FTAVF a congenital lesion[28,31,32] In fact, some consider that the neural tube defects are contributing factors for the formation of this malformation at older ages.[28] Noteworthy that 21 of the cases reported have concomitant lumbar degenerative changes like spinal canal stenosis or lumbar disc herniation according to our review.[2,7,8,10,14,24,25,30] It is postulated that the repetitive microtrauma and chronic inflammation within the filum terminale, caused by long-standing degenerative stenosis, can induce a positive feedback loop and eventual fistula formation.[10,24] In the remaining cases of FTAVF reported, we found neither congenital nor acquired lesions described that can be associated with this vascular malformation. We conclude that probably FTAVFs are caused by an association of factors that can be congenital, acquired, or an aggregation of both.

FTAVFs can present with an ascending congestive myelopathy;[3,5,16,27] as was seen in our case. Theoretically, first, there is an increased venous pressure secondary to the augmented blood flow through the fistulous shunt. This will cause a venous hypertension which is highlighted by the venous varices. Second, a prolonged steal phenomenon, driven by the chronic low-pressure shunt through the fistula, will induce a chronic arterial steal and decreased cord perfusion. The persistent venous congestion and spinal cord ischemia will induce a progressive spinal cord dysfunction.[6] Regardless of the location of the spinal intradural AVF, it is the lower part of the cord that is commonly affected first.[14] This phenomenon is explained by two motives: on the one hand, the intraspinal venous system is valveless, and due to gravity, the conus and lower cord are more congested; on the other hand, lower cord has physiologically less collateral vascularity.[11]

The ascending congestive myelopathy caused by spinal AVF is also known as Foix-Alajouanine syndrome.[11,14,16] This syndrome was initially described in 1926 by Charles Foix and Théophile Alajouanine and originally called subacute ascending necrotizing myelitis.[4] It is clinically expressed by lower limb weakness, ascending hypoesthesia, loss of sphincter control, and areflexia or hyperreflexia.[16] Due to the rarity of these vascular malformations, particularly FTAVF, and the low specificity of its presenting clinical manifestations, they are often misdiagnosed as polyradiculopathy, polyneuropathy, intramedullary tumor, or even masqueraded as lumbosacral degenerative disease or neural tube defects.[11,14,15,25,26] This can lead to delayed diagnosis or unnecessary spinal surgery.[24,25]

According to our review, there are 66 cases of FTAVF reported in the literature.[6,10,17,24] Reviewing all FTAVF cases reported, 11 cases were submitted to previous lumbosacral surgery and subsequently masqueraded as a failed back surgery syndrome [Table 1]. Of these, 10 had lumbar degenerative pathologies and 1 had a tethered cord syndrome with a sacral lipoma (the youngest of all). Ten patients were male and one female; they had a median age of 59 years old. The average length of symptoms was 37 months. All cases presented with motor symptoms that impaired autonomous gait, hypoesthesia, and sphincter dysfunction (not stated sensory symptoms in 1 case and sphincter functionality in 2 cases). In most cases (7 cases), the reflexes were not mentioned. After the first surgery, only 3 cases had partial improvement. MRI and selective arteriography were diagnostic exams in all cases. After the treatment of FTAVF, most cases showed sustained improvement, excepting 2 cases (patient previous submitted to partial removal of sacral lipoma; patient that was submitted to unsuccessful embolization and then refused the surgery). Our case was initially misdiagnosed as lumbar stenosis, having been submitted to a decompressive laminectomy. Besides being the oldest patient reported, he meets the other clinical features. He had transient improvement after the first surgery, but sustained improvement was only verified after the FTAVF disconnection.

Table 1:

Patients with filum terminale arteriovenous fistulas previously operated for lumbar degenerative disease reported in the literature.

Like the other cases with similar history, our case highlights the importance of clinical awareness to signs such as paraparesis with gait disturbance and sphincter disturbance, in patients with previous surgery for lumbar degenerative disease. Due to the fact that these neurological signs are very rare in this type of pathology, the MRI scans must be carefully analyzed, especially the T2 sequence to search for vascular voids. Although rare, FTAVF is a diagnose that should be considered in cases of failed back surgery syndrome.

The interruption of the FTAVF can be done by both surgical and endovascular techniques.[10,20] The type of treatment depends on numerous factors such as the patient preoperative morbidity, functional status, angioarchitecture and anatomy of the fistula, local expertise, and patient preference.[6] In general, microsurgical interruption of FTAVF has been the favored intervention.[27] Since these malformations are fed by a small caliber, long tortuous, and low-lying ASA, with fistulous point close to the arterial supply of spinal cord, microsurgical treatment is considered to be safer for avoiding arterial ischemic complications.[5,28,31,34] Another advantage of the surgical treatment is the possibility to treat both the fistulous shunt and the lumbar degenerative disease in the same surgical session in cases where both are concomitant.[10] It should be noted that in cases with multiple feeding vessels, endovascular treatment may not be a viable curative option.[9,21,27] The treatment key factor is safe isolation of the feeding artery, fistulous point, and draining vein to interrupt the arteriovenous shunt and preserve the normal vascularization of the nervous structures.[6]

The outcome of interrupted FTAVF is dependent on several factors such as duration of symptoms, pretreatment disability, and success of the procedure to close the fistula.[11] After shunt obliteration, the venous congestion tends to decrease and results in progressive improvement or, at least, symptoms stabilization.[1] In our review, almost all the patients that were successfully treated to the FTAVF had neurological improvement. The exception was the case with a tethered cord syndrome with a lipoma that had only symptomatic stabilization,[28] probably due to inherent spinal cord compromise.

CONCLUSION

FTAVF is a very rare spinal intradural spinal vascular malformation that can be misdiagnosed, submitted to unnecessary spinal surgery, and even masqueraded as a failed back surgery syndrome. Signs of ascending myelopathy should prompt awareness and the MRI must be carefully evaluated for vascular voids in these cases.