Introduction

Tethered cord syndrome (TCS) is a neurological disorder caused by tissue attachments of spinal cord at different levels. These attachments cause an abnormal stretching of the spinal cord limiting its movement. In some cases of this congenital deterioration, a spinal cord with conus in normal position can still be tethered by a thick filum or split cord. This limits to the growth of the spinal column. As a result, patients may experience a wide spectrum of clinical presentations ranging from asymptomatic to paraplegic [9, 13]. TCS was described in 1976 by Hoffman et al. [6] after observing that the spinal cord was tethered via a thickened filum terminale to the sacral bones in 31 children and that there was a noticeable neurological improvement following the release of the cord.

This syndrome is not a rare pathological disorder particularly in developing countries. As a result of the development in imaging technologies, the number of adults in whom congenital TCS is diagnosed increases. The rate of TCS incidence is not known. However, the rate of incidence of neural tube defects is estimated as 0.17–6.39 per 1000 live births worldwide [2]. TCS is one of the common spinal dysraphic lesions which have been reduced in developed countries due to the supplementation of folic acid during pregnancy as well as the prenatal diagnosis of dysraphic malformations often leading to termination of the pregnancy.

The diagnosis in adult patients is not well established until back and/or leg pain symptom manifests. Several pathological conditions including fatty and thickened filum terminale, lipomyelomeningocele, meningocele, split cord malformation, diastematomyelia, and post-repair myelomeningocele may be responsible for TCS [5]. It has been well established that early surgical intervention of congenital TCS, whether symptomatic or otherwise, in children prevents additional neurosurgical deficits [4, 9, 10]. With early diagnosis of these conditions, however, there remains an ongoing debate of whether asymptomatic adult patients with a TCS should undergo prophylactic surgery [9]. According to Rajpal et al. [14], the risk of developing new neurological deteriorations in patients with congenital TCS increases with advancing in age. Therefore, Rajpal et al. suggest using surgery for adult patients with TCS to protect them from developing new neurological deficits which may be permanent.

Despite the management of congenital TCS diagnosed in adulthood remains controversial, results of recent clinical studies of surgical intervention in adults are encouraging [8, 9, 11, 14, 16, 20]. Herein, the surgical treatment outcomes of 25 consecutive congenital TCSs diagnosed in adults have been evaluated to find out the factors that may lead to good surgical outcomes.

Material and methods

This retrospective study was approved by the medical ethics committee of our hospital under decision number 507/2015. Written informed consent was obtained from all patients for publication of their cases and accompanying images.

Medical records were retrospectively reviewed for randomly selected 79 cases of congenital TCS which were treated surgically at the Department of Neurosurgery of Bakırköy Research and Training Hospital for Neurology, Neurosurgery and Psychiatry (BRSHH), between the years 2005 and 2015. Only the TCS cases of adults (patients aged 18 years or above) which were diagnosed and surgically treated are included in this study.

The patients underwent full neurological examination especially those who experienced low back and leg pain associated with urination complaints or those with malformations associated with TC. All of our adult TCS patients suffered tolerable pain. They were treated conservatively followed by close follow-ups at 3-monthly intervals to avoid the development of permanent new neurological deficits [16].

Surgical approach was selected according to surgeons’ skills. Intraoperative neurophysiological monitoring (IONM) and laminoplasty were the purpose of the surgery especially during the last 5 years. Laminoplasty is the process involving the bilateral removal of laminas using high-speed drill followed by the untethering (or any other planned surgery) laminas that are bilaterally sutured using a strong nonabsorbable suture to both lamina parts.

Statistics presented hereafter are expressed as the mean ± standard deviation values together with the range between parentheses. Differences among groups were assessed with the one-way analysis of variance (ANOVA) using the SPSS 21.0 statistical package. Significance in the multivariate model was determined using a p value <0.05, and a trend-level effect was assigned to a p = 0.05–0.10. All p values were given with odd ratios (OR).

Surgical procedure

Under general anesthesia and using IONM, the patients are positioned prone using supporting roll on each side. The paramedian vertical midline incision was performed. Hemilaminectomy, laminectomy, or laminotomy (in cases of spina bifida occulta; defective laminotomy was used) was completed after dissecting the paraspinal muscles. Bilateral laminotomy is done using high-speed drills or kerrison rongeurs. Next, ligamentum falvum and the adipose tissue are removed. The spinal cord may continue until the S1 or S2 levels by giving some sacral rootlets. The laminectomy or laminotomy should be performed up to this level. The operative microscope is brought in over the operation field. Thecal sac should be opened in the midline and tacked up bilaterally using strong sutures. After exposing all of nerve roots, filum terminale, and arachnoid bands, the neurosurgeons select the filum terminale using the microscope and IONM. Under the microscope, filum terminale appears darker in color than nerve rootlets. This darker color of filum terminale is related to its fibrovascular tag structure that contains a large vessel, which becomes smaller across through the thecal sac [13, 17, 18]. On the other hand, this vessel is not a reliable landmark for filum terminale since similar vessels can be found on the rootlets or sometimes no vessel can be seen on filum terminale.

It is recommended to use the probe of IONM to check whether the tissue is neural or not. This helps to avoid cutting one of the rootlets instead of the filum terminale to be tethered. The rootlets are retracted laterally, and the filum terminale is coagulated and cut after identification. All connective tissues attached to the caudal part of the spinal cord and conus medullaris should be released. After hemostasis using serum physiologic water, duraplasty is performed using 5.0 absorbable sutures. This was done for the last reported ten cases of our sample to avoid CSF fistula after tight closure of the dura by the surgeons using fibrin sealant products. In cases of bony septum (BS) or fibrous band (FB), resection of BS or FB should be done first before the untethering procedure. In cases of dermal sinus, the tracts may be attached to the thecal sac. Therefore, these structures should also be removed and duraplasty should be done if found necessary.

Illustrative cases

Case 1

A 61-year-old male was referred to BRSHH with low back and right leg pains. He tumbled 6 years earlier. His neurological examination was intact, except for weakness in his right leg distal muscle group of 4/5, L2–5 and S1 dermatomal hypoesthesia, neurogenic claudication, and the Babinski sign being bilaterally evaluated as no response. MRI showed diastematomyelia at L2–4, block vertebra at L3–4, at L3 level split cord malformation as bony septum and dermal sinus tract were apparent. A giant cyst extended between L1 and S2 while filum terminale terminated at S1–2 level (Fig. 1). Urodynamic study showed normocompliance normotonic bladder function. The patient underwent bony septum resection and duraplasty using total L3–4 laminoplasty, followed by cystectomy using total S1 laminoplasty. The histopathological examinations revealed that the cyst is arachnoidal. The patient recovered quite well and was discharged after 3 days without any complications. He was doing well on his postoperative 52nd month doctor visit.

Fig. 1
figure 1

A 61-year-male patient presented with numbness and low back and leg pain. a Sagittal MRI demonstrated cyst extends between L1 and S2. Note that L3 and L4 vertebral fusion anomaly (block vertebra) is apparent. b Axial MRI showed that diastematomyelia extends between both of L2 and L4. Note that split cord malformation as bony septum at L3 level and dermal sinus tract are apparent. c Axial MRI demonstrated that filum terminale terminates at S1–2 level

Full size image

Case 2

A 21-year-old female was referred to our hospital with pain in her low back and both legs for more than 5 years. During the last month before the referral, she developed urinary incontinence. Her neurological examination was intact except for an infrequent loss control of urination. MRI showed that the filum terminale terminated at L5 level (Fig. 2). The patient underwent untethering using L5 laminoplasty. She recovered well and discharged after 2 days without any complications. The patient was doing well on her postoperative 14th month doctor visit.

Fig. 2
figure 2

A 21-year-old female was presented with low back and both leg pain for more than 5 years. Her neurological examination was intact; expect rare loss control of her urination. MRI demonstrated filum terminale terminates at L5 level

Full size image

Results

A sample of 25 adult cases of TCSs that is composed of 16 female (64%) and nine male (36%) patients was treated surgically. The baseline clinical characteristics, co-malformations, treatment, outcomes, and complications of treatment for all cases are given in Table 1. The mean age of the sample is 30.1 ± 10.3 (18–61) years. The mean age of the fully recovered patients is 29.8 ± 10.4 (20–61) years, while that of the remaining patients is 30.5 ± 10.8 (18–48) years. Despite the fact that the recovered group is younger on average, the age factor is not statistically significant (OR 1.02, p = 0.93). The mean follow-up period for all cases is 73.8 ± 43.5 (6–131) months. The duration between the initial symptom and the surgery ranges between 4 months and 25 years with a mean of 6.3 years. The most predominant complaint (68%) is low back pain (Table 2). The most common clinical findings in neurological examinations are urinary incontinence and motor weakness (Table 3). The urodynamic test of 21 patients revealed an overactive detrusor muscle in three patients, post-void residual urine (>100 ml) in nine patients, and normocompliance normotonic bladder function in nine patients. The most associated malformations are diastematomyelia and vertebral fusion anomalies (Table 4). The most common level of tethered cord is L5 level. Untethering was carried out in 88% of the patients while bony septum resection without untethering was applied to 12% of the cases. The involved surgical approaches are laminoplasty (44%), laminectomy (40%), and hemilaminectomy (16%). The surgical morbidity (i.e., new deficit or worsening neurological examination) rate is 4%. The common complication is the CSF leakage (12%). This complication was experienced by the 6th, the 8th, and the 15th patient of our sample (Table 1). None of them developed new neurological deficit or neurologically deteriorated. This rules out the possibility of considering them morbid.

Table 1 Baseline clinical characteristics, co-malformations, treatment, outcomes, and complications of treatment in 25 operated cases
Full size table
Table 2 The presenting symptoms of 25 TCS adult patients
Full size table
Table 3 The clinical findings in neurological examination of 25 TCS adult patients
Full size table
Table 4 Co-malformations that were detected in 25 TCS adult patients
Full size table

Sixty-four percent of the patients had good clinical outcomes (Table 5), with full recovery or resolution of their major symptoms by their last follow-up (mean 73.8 months). The mean length of the hospital stay is 4.76 ± 2.88 (2–12) days. Laminoplasty leads to mean hospital stay of 3.9 ± 2.2 (2–8) days. This is shorter than the hospital stay related to hemilaminectomy and laminectomy approaches which led to mean stays of 4 ± 2.2 (2–7) and 6.4 ± 3.3 (3–12) days, respectively. The difference, though, is not significantly different (p = 0.57 and p = 0.29, respectively). None of the three approaches show any superiority compared to the other two in terms of surgery-related complications.

Table 5 Clinical outcomes according to neurological scoring system
Full size table

A multivariate regression model shows that compared to laminaplasty (OR 2.05, p = 0.047) and hemilaminectomy (OR 1.875, p = 0.049), laminectomy is independent of other risk factors associated with poor or marginally improved (almost unchanged) surgical outcomes. Furthermore, the bladder dysfunction when associated to muscular weakness results in poorer surgical outcomes compared to the other symptoms (OR 7.0, p = 0.014) and long-term (>6 months) symptoms (OR 24.5, p = 0.0001). Cases presented without lipomas are most likely to recover compared to those with lipomas (OR 2.4, p = 0.001). Aggressive surgical treatment of lipoma is the reason of the only recurrence case after 6 years. Presence of lipoma is also associated with motor weakness, atrophy, or bladder dysfunction but only with trend-level significance (OR 1.49, p = 0.09).

When the urodynamic test shows overactive detrusor muscle, no improvement is recorded according to the postoperative urodynamic test. This can be attributed to long-term symptoms. However, when the urodynamic tests show post-void residual urine in excess of 100 ml especially in the patients with short-term (<6 months) symptoms, the postoperative urodynamic tests showed improvement in all of the nine patients (either normal or close to normal residual urine of 50–100 ml) (p = 0.0001).

Discussion

Unless the surgical intervention leads to full recovery or resolution of the major symptoms (leg and/or back pain, motor weakness, sphincter dysfunction, or gait ataxia) that may impact the patient’s life quality, the patient’s satisfaction and their quality of life continue to be poor. Therefore, in this study, we divide the surgical outcomes into two major groups. The first is the “fully recovered group” that consists of the patients who have good clinical outcomes with resolution of their major symptoms. The second is the “others group” which includes all of the other cases with at least one of their major symptom is not resolved.

Gender and age

The female preponderance in adult patients with TCS is a well-established observation in all previously published reports [7, 8, 10, 14, 17]. The female predominance is observed in 64% of our patients in line with the previous reports. The mean age is 30.1 ± 10.3 (18–61) years. At their last follow-up, which is on average after 73.8 months, 64% of patients have good clinical outcomes (Table 5). Out of the 16 recovered patients, there are 11 females and 5 males. The gender factor is not statistically significant (OR 1.76, p = 0.34). Similarly, the mean age of the recovered group is 29.8 ± 10.4 (20–61) years while that of the “others group” is 30.5 ± 10.8 (18–48) years. Despite the fact that the recovered group is younger, on average, the age factor is not statistically significant (OR 1.02, p = 0.93).

Long-term symptoms and neurologic examination on presentation

Most of the adult patients involved with a tethered cord (TC) practice their normal daily activities without any complaints since their childhood despite this congenital disorder. However, during their adulthood, they became more tolerable for any significant symptoms compared to pediatric population. In some cases, the symptoms of the TCS appear after a traffic accident or tumbling. Furthermore, physiological dynamic or static changes occur on spinal column such as those that happen during pregnancy may aggravate the degenerative processes and induce back and leg pains.

Neurological deficits developed in TC adult patients are generally irreversible. Stretching of the conus medullaris and nerve roots may induce back pain, leg weakness, foot deformity, scoliosis, sensory loss, and/or bowel or bladder dysfunction [5, 9, 10, 13, 16]. Untreated TCS can progress in 27.5%, 40%, and 60% of cases in 1, 2, and 5 years after diagnosis, respectively [13]. The duration between the initial symptom and the time of surgical intervention in our sample ranges between 4 months and 25 years with an average of 6.3 years. Bladder dysfunction when associated with muscular weakness (OR 7.0, p = 0.014) and long-term (i.e., >6 months) symptoms (OR 24.5, p = 0.0001) are independent risk factors leading to poor, minimally improved or almost unchanged surgical outcomes even after rehabilitation programs. Therefore, we recommend early surgical intervention in symptomatic patients irrespective of the fact that most of the TC diagnosed in adults are asymptomatic. During the last 11 years, 98 adult patients were diagnosed by our team with TC after incidental permanent or severe pain in the back and or leg or after serious urological complaints. Only 25 adult patients (25.5%) who were diagnosed as symptomatic TC were treated surgically. Klekamp reported that only 50.8% of the adult patients with TC underwent untethering procedure [8]. Asymptomatic TC in our sample was identified in 73 patients (74.5%).

The patients have to undergo full neurological examination especially those who experienced low back and leg pain associated with urination complaints or those with malformations associated with TC (Table 4). The surgical intervention in children and adolescents is recommended to reduce the expected risk of further neurological deficits [16].

Associated malformations

Only two patients in our sample are free of other malformations that are associated with TC (Tables 2 and 4). Therefore, the malformations in patients with classic symptoms of TCS should alert the physicians to refer such patients to neurosurgeons. Split cord malformations (SCMs) are mainly associated with TCS in adults [9]. SCMs are divided into two types based on the state of the dural tube and the nature of the median septum: type I SCM (i.e., diastematomyelia with septum) and type II SCM (i.e., diastematomyelia without septum). The former type usually type I SCM coexists with scoliosis and the TCS. Cutaneous stigmata (hemangioma), skin discolorations, and hypertrichosis are characteristic features of type I SCM. Butterfly vertebrae, block vertebra, hemivertebrae, and spina bifida are the vertebral abnormalities which can be associated with this type [12]. BS or FB midline septum splits the spinal cord into two tubes each containing a hemicord. Beyond the BS, which may occur in either thoracic or lumbar regions, the two hemicords adhere to each other and return to a normal anatomic position.

Some cases of SCM show signs of TC, whereas no spur is found in malformation. This happens in type II SCM. The cleft is generally partial or incompletely split. Spina bifida is often present with the possible formation of hydromyelia and thin FB may form [19]. The management of TCS may change according to the associated SCM type [16]. For type I SCM, the removal of FB or FS may be sufficient. If not, the untethering procedure is necessary. However, for type II SCM, the removal of FB should be followed directly by the untethering procedure.

In our sample, seven out of eight patients with SCM were treated by BS or FB resection. Only four of those seven needed untethering procedure afterwards to relieve their pain or other complaints. In the remaining three patients, the untethering procedure was not applied in order to avoid any new neurological deficits. Out of eight patients associated with lipoma, only two were treated by the removal of lipoma. A third patient was treated by cystectomy. Four out of five patients presented with dermal sinus tract were treated by resection and repairing the skin surgically. The patients with scoliosis and syringohydromyelia did not need additional surgical intervention, whereas 38% of our child TCS patients who were presented with the same malformations underwent additional surgery to treat their complaints [16]. Surgical outcomes in cases without lipomas have better recovery chances compared to those with lipomas (OR 2.4, p = 0.001). Aggressive surgical treatment of lipoma was the reason behind the only recurrence case 6 years later. The presence of lipoma was also associated with motor weakness, atrophy, or bladder dysfunction but only with trend-level significance (OR 1.49, p = 0.09).

Surgical approaches

IONM and laminoplasty were the purpose of the surgery especially in the last 5 years. IONM is useful to reduce the serious complications that may result from TC releasing procedure. Laminoplasty reduces the adhesion that may facilitate recurrence of TCS in adults (even our recurrent patient underwent laminoplasty, but his recurrence is thought to be related to the aggressive removal of lipoma).

Comparing laminoplasty, hemilaminectomy, and laminectomy in our sample, the authors found that laminoplasty leads to mean hospital stay of 3.9 ± 2.2 (2–8) days which is shorter than that of hemilaminectomy and laminectomy approaches which lead to mean hospital stays of 4 ± 2.2 (2–7) and 6.4 ± 3.3 (3–12) days, respectively. The difference, though, is not statistically significant (p = 0.57 and p = 0.29, respectively). The three approaches do not show any superiority from each other in surgery-related complications.

A multivariate regression model shows that compared to laminaplasty (OR 2.05, p = 0.047) and hemilaminectomy (OR 1.875, p = 0.049), laminectomy is independent of other risk factors associated with poor or marginally improved (almost unchanged) surgical outcomes. This statistically significant difference may be related to IONM not to the approach itself.

In cases of lipoma, cyst, or other hamartomas that press on the nerve rootlets and cause narrowing of the thecal sac, hamartomas should be removed too. To avoid the serious complications and to prevent retethering, lipomas should not be removed aggressively. This decision should be taken after the intraoperative evaluation. The operation field should be kept clean and the CSF circulation between the neural elements should be preserved.

In the adult population, the rate of retethering was quoted to be as high as 29% [1–3, 7, 8, 15, 17]. The retethering was developed in one of our male patients 6 years later. This means the retethering rate in our sample is only 4%. The patient underwent untethering procedure and partial surgical resection of his lipoma. This case shows that laminoplasty is a good barrier to protect thecal sac and nerve roots from fibrosis and granulation tissues.

Pre- and postoperative urodynamic studies

The urodynamic test has a predictive value regarding most of the future deteriorations in adult patients with TCS. The patients who were diagnosed with TC incidentally after moderate pain were followed up closely. When their complaints started to be symptomatic and urological complaints were added to their clinical tables, urodynamic studies were reperformed and the surgical intervention decision was taken.

We recommend surgery for the patients whose urodynamic test shows overactive detrusor muscle. No improvement is recorded in the postoperative urodynamic test for those patients. The long-term symptoms can be held responsible for this. On the other hand, when the urodynamic tests show post-void residual urine >100 ml especially in the patients with short-term (<6 months) symptoms, postoperative urodynamic tests show improvement in all nine patients (either normal or close to normal residual urine = 50–100 ml) (p = 0.0001).

The present study suffers from few limitations. The used sample is composed of all documented TC cases in our hospital over 11 years. The sample size is relatively small (n = 25). The sample does not represent a wide geographical area as all the patients are from Istanbul and surrounding area. The results represent a single institute experience. Other institutes may follow different approaches. The study design is retrospective. Further prospective randomized studies with larger samples and long follow-up periods are required to improve the representativeness of the results.

Conclusions

TCS is a congenital deterioration which may lead to serious complaints. A sample of 25 patients was treated surgically, and the results were analyzed in this study. Laminoplasty (or hemilaminectomy), short-term (<6 months) symptoms, patients without lipomas, and presentation with moderate or mild symptoms seem to be proper predictors for good surgical outcomes. Urodynamic study can be used as a predictive tool for diagnosing TCS in asymptomatic adult patients. Further prospective studies are necessary to investigate these findings systematically.