The Cervical Tethered Spinal Cord

Kenji Muro, Arthur J. DiPatri Jr., and David G. McLone

Tethered cord syndrome (TCS) is a well-established and recognized disease entity whose protean manifestations were first described by Hoffman, Hendrick, and Humphreys in 1976.1 That study cohort included cases of caudal TCS caused by a thickened filum terminale with or without associated lipomas. Since this initial report, other investigators have recognized similar clinical scenarios originating from pathological entities in the cervical spinal cord. A spectrum of underlying etiologies has been described in the literature but only case reports and case series exist at present. In addition to the clinical and radiographic features that cervical TCS shares with the more common caudal TCS, other characteristics also deserve mention. The definition, historical perspective, and pathophysiology of TCS as a whole are discussed in greater detail in other chapters within this text.

The hallmarks of TCS are a progressive and often subtle deterioration in motor, sensory, bowel, and bladder function. Evidence of both upper- and lower-extremity involvement may be present with cervical tethering, but the sequence of involvement is unpredictable. Pain along the spinal axis and radicular pain are common presenting complaints,^1–3 whereas bladder and bowel involvement are the most variable. Even though cervical TCS is a less common clinical entity than caudal TCS, prompt recognition and appropriate management of cervical TCS is warranted to avoid potentially irreversible morbidity. This chapter highlights features unique to TCS of the cervical region, with comparison to the more common caudal TCS, where appropriate.

Etiology and Embryology

In our own experience and through a review of the published literature, tethering of the cervical spinal cord occurs much less frequently than at the more caudal levels. Many of the well-described lesions responsible for caudal TCS such as intradural lipomyelomeningocele, fibrous adhesions, split cord malformation, dermal sinus tract, and adhesive arachnoiditis after myelomeningocele repair also apply to the cervical region. One of the more common causes of caudal TCS, the tight filum terminale, for obvious reasons is not relevant to the cervical region; however, we believe like others that the pathophysiology by which the filum terminale effects the onset of caudal TCS also applies to lesions of the cervical spine.⁴ In addition to the pathological entities already listed, other etiologies such as postsurgery spinal arachnoiditis and adhesions that occur following intramedullary spinal cord tumor surgery or after penetrating or nonpenetrating spinal cord injury are being increasingly recognized as a cause for delayed neurological deterioration among these select cohorts. Although cervical spine TCS does seem to be a distinct entity, it shares with caudal TCS the most salient feature, that is, clinical improvement following surgical untethering.^3,5–19

In general there are two broad two categories of lesions that may lead to the development of TCS—congenital and acquired. The cutaneous stigmata associated with congenital lesions are often noted at birth and typically the symptoms associated with TCS manifest early in life, depending on the degree of traction 20 on the spinal cord. The acquired forms occur more frequently in adults. After the initial insult the time that it takes for symptoms attributable to TCS to develop may vary considerably, with reported cases occurring up to 37 years³ after injury.

Congenital Lesions

Almost all of the lesions responsible for tethering of the cervical spinal cord in young patients are congenital. Most occur as a result of defective neurulation. The embryogenesis of these lesions dictates that the epicenter of the tethering structure generally occurs at the dorsal raphe of the spinal cord, and this is an important consideration for preoperative planning and for intraoperative intervention. The defect in primary neurulation that leads to the formation of the cervical myelomeningocele has been well described. This example of disordered embryogenesis ultimately leaves the spinal cord in the form of the neural plate of the embryo. The split cord malformation also originates from defective neurulation, but as detailed by Pang²² in his unified theory, its genesis is from the persistence of an “accessory” or “anomalous” neurenteric canal.²¹ Usually, the neurenteric canal is a developmentally transient communication between the endoderm and ectoderm layers of the embryo. Failure of the accessory neurenteric canal to involute leaves a structure that transfixes the developing neural tube. The resulting fibrous band or bony spur divides and ultimately tethers the spinal cord.²²

Both the lipomyelomeningocele and dermal sinus are defects that occur later in neurulation, at the time when the neural tube undergoes disjunction from the superficial ectoderm. Premature disjunction prior to closure of the neural tube allows mesenchymal cells access to the central portions of the neural tube. Lying in apposition to these mesenchymal cells, the luminal surface of the neural tube induces the mesenchyme to form adipose and connective tissues resulting in a lipoma. In contrast, it is incomplete disjunction, resulting in a persistent connection between the neural and superficial ectoderm, that forms a dermal sinus. Other forms of spinal dysraphism reported to cause cervical TCS include the cervical myelocystocele¹⁸ and meningoceles.^7,8,12 In general cervical meningoceles are not directly responsible for causing the symptoms of a TCS, but rather their presence may indicate the existence of an associated intraspinal lesion such as a lipomyelomeningocele or a split cord malformation,^8,12 which may not necessarily be found at the level of the meningocele.

Acquired Lesions

Among the acquired forms of TCS, the posttraumatic TCS is the most frequently encountered clinicallly and in the literature. A prerequisite for the occurrence of an acquired TCS is some form of intradural pathology. Intradural spinal tumor resections, spinal cord injuries, and previously operated degenerative spine conditions¹⁴ are some of the more common examples. Radiographically, the posttraumatic spinal cord manifests diverse morphological patterns of myelomalacia, namely, progressive posttraumatic myelomalacic myelopathy (PPMM) and progressive posttraumatic cystic myelopathy (PPCM).^{3,12,16,17,23} These two entities are considered by some to be a continuum of the same posttraumatic pathophysiology,²⁴ and risk factors that may increase a patients’ likelihood of developing symptomatic tethering are not clearly defined. It has been hypothesized that in addition to the primary injury to the spinal cord and intradural contents, posttraumatic arachnoiditis and scarring, which are felt to be responsible for the eventual tethering, are promoted by the locally hemorrhagic environment, the chronic recumbent position of many trauma patients, changes in the local cerebrospinal fluid (CSF) dynamics, and chronic ischemia.³

There has not been a correlation between the level of cervical spinal tethering and the time to symptom onset. Furthermore, to explain the variable time period between the initial injury and the onset of TCS, it is likely that the degree of tethering and tension among these patients differs, which results in the variable time to presentation, a theory that Pang and Wilberger proposed in their study of adult patients with caudal TCS.25

Clinical Presentation

The presence of cutaneous stigmata suggestive of an underlying dysraphic lesion is present in almost all congenital lesions. As is frequently the case with the more caudal forms of TCS, it is not uncommon for these structural abnormalities that occur with the congenital etiologies of tethered cervical cord to be noted at birth but ignored until the onset of symptoms later in life. Dermal pits, hemangiomas, meningoceles, lipomas, and a short neck are markers of underlying pathology. Those cases of TCS that are due to spinal injury or are related to prior surgery become apparent after the history and physical exam is obtained.

Regardless of the etiology, the development of symptoms may involve all or relatively few of the neurological modalities traversing the tethered segment. In a caudal tethered cord, onset and worsening of symptoms appear to correlate with periods of rapid growth when it is theorized that increased traction is applied to the spinal cord.¹ Among adult patients, the signs and symptoms of caudal TCS may be aggravated by prolonged sitting or forward bending. No such description exists in cervical TCS except in the posttraumatic literature, where Lee at al¹⁷ describe their patients’ symptoms as “positional” but do not elaborate further. The signs and symptoms of the caudal TCS may also develop acutely after trauma as was noted in 61% of patients in one series.²⁵ In cervical TCS the association with trauma preceding the onset of symptomatic TCS has only been described in one case report of a 34-year-old adult with a split cord malformation of the cervical spine.⁶ Unlike adults, who are able to specify and indicate symptoms, a cautionary note must be voiced in the young infant whose lack of achieving motor and/or sphincter milestones may be perceived as individual variance.

Individuals often give a long history of clumsiness, abnormal gait, and patchy numbness in all four extremities, although initially the symptoms may be limited to only the lower extremities. Bladder and bowel involvement or spinal axis pain is frequently the reason these patients finally seek medical care. Unlike caudal TCS, patients with cervical TCS may also present with involvement of other neurological signs or symptoms, including autonomic dysreflexia, hyperhydrosis, respiratory insufficiency, or even Horner syndrome.^3,17

Neurological examination of the patient often reveals, in addition to the cutaneous manifestation, atrophy of the upper extremity musculature,¹⁵ especially the intrinsic muscles of the hands. Complaints may involve all four limbs, with quadriparesis being a common finding, An asymmetrical motor exam or possibly a hemiparesis may occur. Sensory symptoms may vary from a patchy or asymmetrical sensory loss to a well-defined cervical sensory level. Deep tendon reflexes may range from hypoactive to brisk, and pathological reflexes are not uncommon. A neurogenic bladder can occur as the sole manifesting complaint.

Diagnostic Studies

Magnetic resonance imaging (MRI) has become the first-line modality for screening purposes when concern exists for a tethered spinal cord. Conventional roentgenograms and computed tomographic (CT) scans are indicated for the analysis of the patient’s bony anatomy and may show anomalies such as spina bifida, fused laminae, the bony spur of a split cord malformation, or the segmentation abnormalities of Klippel-Feil syndrome. In most cases, MRI is sufficient to determine the cause and extent of the tethering lesion, and the decision to operate and the design of the procedure can often be made utilizing this study alone. Water-soluble contrast-enhanced CT myelography may be helpful in situations where the MRI is equivocal, and postmyelography tomographic images can sometimes demonstrate the tethering bands and clefts in the spinal cord. 20,23,25 This is especially useful in the situation of prior spinal instrument implantation, which may make MRI difficult to interpret.

Several studies have addressed the cooccurrence of cervical meningoceles with an intraspinal tethering lesion. In five patients with cervicothoracic meningocele, Chaseling et al found “deep extension” of the meningocele at the time of surgery in two patients, and one additional patient had an associated myelographic abnormality, the exact nature of which was not detailed further.7 In another study, Delashaw et al described finding lipomeningomyelocele, tethered cord, and diastematomyelia in a cohort of four patients with cervical meningoceles. Most of these lesions were caudal to the level of the meningocele.⁸

Preoperative somatosensory evoked potentials (SSEPs) are rarely utilized or reported in the literature. Decreases in amplitude and prolonged latencies^10,26 may be noted, although this may also occur in normal patients.¹⁵ The value of this testing may be in its use as a baseline dataset that surgeons would refer to either intraoperatively or in postoperative follow-up. Given the anecdotal nature of obtaining electrophysiological studies preoperatively, there is no consensus on its use.

Treatment Intraoperative Monitoring

Several authors have reported the use of intraoperative anal sphincter electromyography (EMG) and stimulation of the pudendal nerve^2,25,27 and found the technology useful for the determination of functioning neural elements and for operative management of caudal TCS. These studies did not detail the number of instances in which the technique was utilized, nor the subset in which the technique made a difference in intraoperative decision making. Furthermore, they do not compare the outcomes of patients who underwent surgery with or without monitoring.

The use of motor evoked potentials during surgery for patients harboring intramedullary neoplastic lesions, including lipomas, has been reported with 100% sensitivity and 91% specificity of the technique predictive of postoperative motor deficits.^28,29 However, this technique has not yet been reported as an adjunct in the surgical management of cervical tethered cord. Furthermore, there are technical limitations in the ability to perform intraoperative monitoring at the cervical level of certain functions, especially the bladder. Because the bladder is innervated from multiple levels and there are many supraspinal modulators, the accurate identification of bladder efferents remains technically challenging.³⁰

Intraoperative SSEPs are also variably reported in the cervical TCS literature. The most dramatic result was demonstrated in the report by Eller et al, when the SSEPs were shown to have a significant and immediate improvement after releasing the tethering lesion intraoperatively.¹⁰ It is the senior author’s (DGM) experience, however, that SSEPs are not particularly useful in most situations because the act of untethering requires the surgeon to proceed with procedures that must be continued if the goals of the operation are to be realized, despite the monitoring results. Most importantly, there are no prospective studies reporting the use and comparative outcome of patients undergoing surgery for cervical TCS with intraoperative monitoring. Completion of such a study may be logistically difficult, owing to technical considerations and the rarity of the condition; however, it would be a valuable contribution to the field.

Surgical Technique

Almost all publications in the literature, regardless of etiology, report the use of microsurgical technique for untethering the spinal cord. Of utmost importance before embarking on this surgical procedure is the thorough understanding of the anatomy, the tethering lesions and their locations, and anticipation of any pitfalls. This includes the assimilation of all radiographic data, noting the epicenter of the lesion and presence of dysraphic laminae and prior postsurgical interventions, along with the characteristics of the spinal cord in relation to the lesion and whether the tethering is occurring in an eccentric fashion. A detailed discussion with the anesthesiologist is also mandatory to ensure proper handling of the patient at the time of anesthesia induction, intubation, positioning, and intraoperative course, especially if neurophysiological monitoring is utilized.

As with the surgical management of any dysraphic or traumatic lesion of the spine and spinal cord, we prefer to identify and delineate as much normal anatomy and landmarks as possible. It is not uncommon to lose tissue planes when dealing with tethering lesions, so a clear understanding of the local and regional anatomy is requisite before addressing the tethering pathology. Surgical loupe magnification is sufficient for the initial dissection and approach to the spine. At all times, the surgical team must be mindful not only of the responsible lesion that must be completely untethered but also of the exit strategy. By this, especially in patients with spinal dysraphism, it is necessary to assess the quality of the fascial layers and dura for an effective, watertight closure.

At this stage of surgery, in the posttraumatic cervical TCS literature, some authors utilize intraoperative ultrasonography to aid the localization of the tethered segment.17 Once the normal anatomy is delineated and the surgical team is ready to address the tethering lesions, we feel the modern surgical microscope and microsurgical instruments are of paramount importance in the handling of these delicate surgical situations. Clearly, dissection is carried forth with great attention and care of the neural elements, utilizing sharp dissection techniques. Thorough inspection and lysis of all tethering lesions are required; otherwise there remains the risk of postoperative delayed recurrence of symptomatic TCS, as reported by Pang et al with their experience in a cohort of patients with cervical myelomeningocele.²⁵ It is our experience that most congenital and acquired lesions causing TCS, upon close inspection, have a clear plane between the tethering segment and the normal neural tissue. The exception is the lipomyelomeningocele, where the lipoma occupies a juxtamedullary-subpial location with no discernable cleavage plane.³¹ It is our practice that complete untethering with subtotal lipoma resection is the best surgical decision, so as not to jeopardize the patient’s neurological function.

In the senior author’s experience, the use of the handheld CO₂ laser substantially reduces the operative time for management of tethering lesions, with particular reference to lipomyelomeningoceles. Use of the laser also reduces the intraoperative blood loss because the lipoma is vaporized and small vessels are sealed simultaneously.³² Only larger arteries require separate bipolar cauterization. The reduced blood loss is particularly important in infants, whose initial blood volume is small. In fact, these infants do not require intra- or postoperative transfusion. The laser also reduces the need to manipulate the tissues and thus prevents the consequent operative trauma. With 3x loupe magnification, the handheld laser may be played over the exposed surface of the lipoma, reducing it layer by layer, down to the liponeural interface. Loupe magnification and manual laser manipulation are preferred by the senior author to the use of the operative microscope with coaxial laser resection because the surface area of the lipoma is large, and there is need for frequent changes in the angle of approach to the portion of lipoma exposed at any one time.

In the posttraumatic cervical TCS literature, after completion of the untethering procedure, many advocate duraplasty to expand the subarachnoid space. It is argued that this expanded space may result in decreased incidence of retethering by adhesive bands. It is possible that compared with primary closure of the dura, which will always lead to some loss of intrathecal volume, the expansile duraplasty will theoretically result in decreased retethering.17 However, there are no prospective data on this practice, and the definition of retethering, on radiographic or clinical grounds, is also not uniform in the literature.

Outcomes

The most important feature of cervical TCS in common with the more prevalent caudal TCS is that patients derive clinical benefit, either improvement or stabilization in the majority of cases, from surgical intervention. Furthermore, the case reports and series in the literature are encouraging in the paucity of major morbidity and mortality associated with surgery. Therefore, the risk:benefit ratio is in favor of untethering.

In the cervical TCS literature, there are multiple case reports and small series relaying surgical successes^3,5–19; one of the most thorough analyses of patient evaluation, treatment, and outcome was reported by Pang and Dias in 1993. In this cohort of nine patients with cervical myelomeningoceles, the first group of six patients had a simple subcutaneous resection of the sac and ligation of the dural fistula. No lysis of underlying tethering structures was performed. Five of these six patients deteriorated 13 months to 8 years after their initial surgery with worsening hand function and lower extremity spasticity. In contrast, the last three patients in this series underwent complete intradural lysis of tethering lesions at the initial surgery. At 3-year follow-up, two of three patients remained neurologically normal; one of the three patient required an additional surgical intervention due to return of symptoms 4 years after the initial surgery, and workup revealed a residual ventral tethering structure, which was addressed. Reoperations, which were performed in the five patients in the first group, were tolerated well and were successful in improving the patients’ neurological function at 12 to 35 months of additional follow-up.¹⁹

Among the many publications of cervical region dermal sinus tracts,^5,6,11,15 in an interesting retrospective report on their cohort of patients with dermal sinus tracts in the cervical and thoracic spines, Ackerman et al found that all of their patients with dermal sinus tracts over the age of 12 months had a neurological deficit at presentation, whereas patients younger than 12 months were all neurologically normal. The patient ages ranged from 3 days to 55 years at presentation. Eighty percent of the patients presenting with neurological deficit “improved” after surgery, and the remaining patients had stabilization of symptoms.

In the acquired etiologies, the posttraumatic cervical TCS has several larger case series reporting the surgical outcome. In a cohort of 53 patients with PPCM and mean follow-up of 23.9 months, Lee et al report that 56% of patients improved when presenting with motor symptoms, 46% with spasticity, 45% with sensory loss, 47% with gait dysfunction, 36% with axial or radicular pain, and “minimal” relief in those who presented with paresthesias, sphincter dysfunction, or autonomic dysreflexia. Overall, 73% of patients had “satisfactory” results, 21% were clinically stabilized, and 7% were worse after surgery.¹⁷

In a cohort of 40 patients with PPMM, surgery resulted in 79% of patients improving when presenting with motor symptoms, 62% with pain, 50% with sphincter dysfunction, 43% with sensory level loss, and 75% with autonomic dysfunction. Overall, in this study, 83% of patients had “good” results and 17% were clinically stabilized.¹⁸ Supporting these results, 95.8% of patients undergoing untethering felt that surgery was helpful in preventing further neurological deterioration.¹²

Taking the reported literature together with our own experience, the outcome from untethering cervical lesions is favorable, even among symptomatic patients. However, experience has also taught us that the natural history of untreated symptomatic tethered lesions is not favorable for a patient’s good functional outcome, with improvement from surgical intervention dependent on the duration of symptoms. In fact, we believe that prophylactic surgery, when conducted at experienced centers with low associated procedural morbidity and mortality, portends the best outcome in these situations.

Conclusion

Cervical TCS is a distinct clinical entity that shares all of the protean features of caudal TCS. Cutaneous or structural abnormalities are often present at birth with congenital lesions, whereas the patients’ past medical history and thorough physical exam are revealing in acquired cases. At a variable period of months to years, insidious neurological deterioration affects sensory and motor systems, as well as bladder and bowel function; unique to the cervical spine is the manifestation of TCS by autonomic dysregulation, hyperhydrosis, respiratory insufficiency, and Horner syndrome.

Surgical untethering stabilizes the clinical picture, and one can often anticipate significant improvement, especially if performed early in the course. Motor and sensory symptoms benefit the most from untethering, with the response to surgery unpredictable for preoperative bladder and bowel dysfunction. Similar to tethering at the caudal end of the spinal cord, retethering of the cervical spinal cord can occur after surgical untethering, and thus patients must be followed throughout life.

There are no prospective trials assessing the optimum technique and timing for intervention for the management of cervical TCS, owing largely to the diverse etiologies responsible for the tethering. For example, it is not known whether prophylactic surgical intervention is warranted in this situation, although extrapolation from the caudal TCS data would support this practice. Therefore, the literature contains many methodologies for patient management and operative intervention. Also, the long-term outcome of these patients is of interest, especially with the increasing recognition of retethering, whose surgical management has been successful. We recommend the creation of a multiinstitutional, prospective database of this unique disease entity that would facilitate the study of these important questions.