Use of lumbar laminoplasty vs. laminotomy for transection of the filum terminale does not affect early complication rates or post-operative course

Abstract

Introduction

Various techniques are used for spinal cord untethering. The purpose of this study was to compare patient characteristics, post-operative course, and early complications after laminotomy versus laminoplasty for transection of the filum terminale for tethered cord release.

Methods

Retrospective analysis of clinical and magnetic resonance imaging data was undertaken for all patients (< 18 years) who underwent tethered cord release by transection of the filum terminale at Oregon Health & Science University, Doernbecher Children’s Hospital, from 2000 to 2011.

Results

Data from two hundred and forty-eight patients were analyzed. Mean age was 5.2 years (range: 0.3 to 16.8 years). Access to the thecal space during surgery was achieved using laminotomy or laminoplasty in 82 (33.1%) and 166 (66.9%) patients, respectively. Laminoplasty patients were significantly younger than laminotomy patients (3.2 vs. 9.3 years, p < 0.0001); other clinical and radiographic characteristics were similar between the groups. Nine patients (3.6%) experienced early complications, including cerebrospinal fluid leak (n=2), suprafascial infection requiring surgical management and intravenous (IV) antibiotics (n=3) or IV antibiotics alone (n=1), a small area of peri-incisional cutaneous necrosis (n=1), peri-operative seizures (n=1), and mild, transient malignant hyperthermia (n=1). There was no difference in the number of early complications between the two groups. Univariate and multivariate analyses revealed no significant risk factor for post-operative complication associated with technique. As judged by caregivers, independent of surgical technique, 97% of patients improved after surgery.

Conclusion

There was no difference in complication risk when performing transection of the filum terminale for tethered cord release using laminotomy or laminoplasty.

Pediatric neurosurgeons commonly transect the filum terminale to treat clinical signs and symptoms of tethered cord syndrome (TCS), including pain, weakness, sensory loss, musculoskeletal deformity, and/or voiding dysfunction, associated with magnetic resonance (MR) imaging abnormalities of the filum and/or low position of the conus medullaris [1–6]. Early complications after surgery for filum transection have been reported to occur in 0% to 6% of cases [7–10,4].

Systematic information is not available, however, regarding the association of specific operative techniques, such as laminectomy versus laminoplasty, with patient clinical characteristics, post-operative course, and early complications. In this study, we retrospectively reviewed these variables in 248 consecutive patients undergoing tethered cord release by filum terminale transection at a single academic pediatric hospital. We hypothesized that laminoplasty was used more commonly in younger children based on their more favorable anatomy for this technical maneuver, and that laminoplasty might prevent CSF leak and pseudomeningocele formation by re-establishing the bony spinal canal immediately dorsal to the intended site of iatrogenic durotomy.

Methods

Patients who underwent a tethered cord release by filum terminale transection at Oregon Health & Science University (CPT code 63200) between 2000 to 2011 were identified using surgeon logs and billing databases, which were cross-correlated for accuracy. Data collected included age, gender, clinical presentation, operative technique, MR imaging findings, length of stay, bed rest, early post-operative complications, and clinical outcome. Patients were seen in clinic at a minimum approximately 2 weeks (by a nurse practitioner) and three months (by the primary surgeon) after surgery. Patients with more complex forms of tethered spinal cord, such as myelomeningocele, lipomyelomeningocele, and meningocele, were excluded.

Laminotomy involved an interlaminar approach at a single level (with exposure of the interlaminar space and removal of the interlaminar ligament and ligamentum flavum, with a rim of bone in the inferior aspect of the superior adjacent lamina and of the superior aspect of the inferior adjacent lamina. Laminoplasty involved bilateral transection of a single lamina just medial to the facet joints on each side, transection of the caudal interspinous ligament with lateral rotation of the lamina based on the cranial interspinous ligament, and later replacement of the lamina in anatomical position using permanent suture. Mid- to lower- lumbar exposures were used based on the level of the conus medullaris. The operative microscope was used for dural opening and closure and untethering in all cases. Untethering was accomplished by resecting a small section of the filum terminale for pathological analysis [11]. The dura was closed with a 6–0 prolene running suture.

A univariate analysis was performed to compare demographics from the laminoplasty and laminotomy groups. Age, sex, surgical technique, conus level, and MR imaging variables (i.e. thickened filum > 2 mm, fatty filum, and conus level), length of stay, and bed rest (independent variables) were used to predict complications (dependent variable) in a multiple logistic regression analysis. Statistical analysis was performed using the SPSS statistical package software (SPSS, Inc. version 16, Chicago, IL, USA) and statistical significance was based on a value of p < 0.05.

Results

Patient demographics and clinical characteristics

Retrospective database searches produced a list of 293 potentially eligible patients. Twenty-five patients with myelomeningocele, lipomyelomeningocele, or meningocele and 10 patients with age > 18 years were excluded from the study. Ten patients with incomplete records were also excluded. Two hundred and forty-eight eligible patients form the basis of this report (Table 1). Mean age at untethering was 5.2 years (range: 0.3 to 16.8 years). One hundred thirty-five patients were female (54.4%) and 113 male (45.6%).

Table 1.

Comparison between laminotomy and laminoplasty groups

Patient Variables	Exposure Technique		p value
	Laminotomy (n=82)	Laminoplasty (n=166)
Age (years)	9.3	3.2	<0.0001*
Gender			0.787#
Female (n, %)	46, 56.1	89, 53.6
Male (n, %)	36, 43.9	77, 46.4
Fatty filum (n, %)	46, 63.9	110, 70.5	0.359#
Thickened filum (n, %)	35, 51.5	76, 57.6	0.454#
Conus level (median)	L2	L2	0.508*
Duration of bed rest (mean, median; days)	1.6, 1.0	1.4, 1.0	0.161*
Length of stay (mean, median; days)	2.0, 1.0	1.8, 2.0	0.155*
Complications (n, %)	1, 1.2	8, 4.8	0.279#
CSF leak (n)	0	2
Suprafascial infection requiring surgical management (n)	0	4
Small area of cutaneous necrosis (n)	0	1
Peri-operative seizure (n)	0	1
Mild malignant hyperthermia (n)	1	0

^*Student t-test

^#2×2 contingency table chi-square test

Two hundred and four patients presented with orthopedic, neurologic, urologic symptoms, with or without cutaneous findings such as a sacral dimple, deformed gluteal fold and/or sacrum, characteristic hemangioma, and/or associated congenital deformity such as imperforate anus or VATER syndrome. An additional 33 patients presented with cutaneous or associated congenital findings, alone. The indication for MR imaging diagnosis of tethered cord was insufficiently documented in the remaining 11 cases.

MR imaging demonstrated a normal appearing conus medullaris (CM) (above the mid L2 level)[12] in 82 patients (35.5%), a low-lying CM (at or below mid L2) in 149 patients (64.5%) and a very low-lying CM (at or below the L2–3 level) in 76 patients (32.9%). Imaging demonstrated a thickened filum (> 2mm diameter) in 111 patients (55.5%), and a fatty filum in 156 patients (68.4%). The mean and median period of post-operative bed rest was 1.6 days and 1 day, respectively, for the laminotomy group and 1.4 days and 1 day, respectively, for the laminoplasty group. The mean and median length of stay was 2 days and 1 day, respectively, for the laminotomy group and 1.8 days and 2 days, respectively, for the laminoplasty group.

Clinical outcome

Of 204 patients who were symptomatic pre-operatively, 198 patients (97.1%) demonstrated clinical improvement at the most recent post-operative follow-up (defined as an improvement in the presenting symptom(s) according to the patient and/or primary caregiver), with no difference between operative technique in clinical outcome (X² (1, n = 204) = 2.97, p = 0.183).

Operative technique and complications

A total of 82 patients (33.1%) underwent a laminotomy and 166 patients (66.9%) underwent a laminoplasty (Table 1). As hypothesized, univariate analysis revealed a significant difference in age laminoplasty vs. laminotomy (3.2 vs. 9.3 years, p < 0.0001).

Nine patients experienced complications (3.6%). Five underwent revision surgery for repair of CSF leak (n=2) or surgical management of suprafascial infection (n=3). There was no association between surgical revision and operative technique (X² (1, n = 248) = 2.52, p = 0.174). One additional patient experienced a suprafascial infection requiring IV antibiotics alone, one a small area of peri-incisional cutaneous necrosis, one transient peri-operative seizures, and one mild, transient malignant hyperthermia. A multiple logistical regression analysis was performed to correlate different variables including age, sex, surgical technique, conus level, MR imaging variables, bed rest, and length of stay with the incidence of post-operative complications. None of the independent variables reached statistical significance.

During the study period, no patient underwent early or late re-operation for spinal fusion (with the exception of children undergoing de-tethering prior to previously planned scoliosis surgery).

Discussion

Complications of filum terminale transection generally result from the surgical approach, rather than from filum transection, itself. CSF leak and infection may complicate any intrathecal surgical procedure, including filum transection. Chern and colleagues determined that a longer hospital stay for maintaining patients horizontally positioned after filum terminale transection does not prevent CSF leak [8]. They also concluded that the use of a dural sealant did not affect the incidence of CSF leak [8].

The present study fails to identify any significant clinical risk factor, including surgical technique (laminoplasty vs. laminotomy) for the occurrence of early surgical complications after filum transection. Although 8 of 9 complications occurred in the laminoplasty group (including all 7 surgical site related complications), this observation failed to reach statistical significance, given the very low overall rate of complications in both groups. Furthermore, the laminoplasty technique was utilized much more frequently in younger children, in whom the laminar bone cuts are much easier to perform. As a result, surgical technique strongly co-varied with age, such that young age could instead account for a slightly increased complication rate in the laminoplasty vs. the laminectomy group. Whether an even larger series of patients would reveal significant differences based on age or technique, and which of these variables is the primary driver of surgical risk, is presently uncertain.

This study found that the occurrence of complications was not associated with a lengthened primary post-operative hospital stay. This is likely because many of these complications presented after hospital discharge or in clinical follow-up. Nevertheless, the overall complication rate seen in the present large patient series is comparable to previous reports [7–10,4] and is low.

There was no occurrence of serious or permanent complication or death. In 248 cases, a high rate of symptomatic improvement was achieved with either surgical technique (> 97%).

This study is limited by retrospective data collection, the limited number of complications allowing for analysis of risk, and the strong surgeon bias in selecting surgical technique based on age. The study also did not address other common techniques, such as filum terminale transection via sacral level approach [13]. Clinical outcomes, although not the principal focus of this study, were reported globally and subjectively by patients and caregivers without a validated instrument. Finally, patients were not followed for a sufficient duration of time to identify any potential late complications of lumbar spinal surgery, such as deformity or instability, which could also vary according to surgical techniques utilized.

No case of confirmed spinal cord re-tethering was identified. Although this complication is relatively common in more complex forms of dysraphism [14–16], it has also recently been reported at surprisingly high rates after filum terminal transection [17]. Although other authors might consider the 3% of patients in the present series with poor clinical response to primary detethering as suffering from ‘re-tethering’, we did not observe radiographic or progressive clinical changes in even these patients suggestive of such a diagnosis.

Conclusions

Use of distinct operative techniques for filum terminale transection (laminoplasty or laminotomy) result in similar clinical outcomes and surgical complication rates. In all patients, complications were rare and temporary. Although laminoplasty was often selected for use in younger patients, for whom it is technically easier, both techniques appear to be acceptable and safe options for use in filum terminale transection procedures.

Figure 1.

Illustration of the two approach techniques utilized for filum transection. The top panels illustrate the incision location (red) and bony removal (green) for the interlaminar (left) and laminoplasty (right) techniques. The bottom panels illustrate the dural exposure achieved and planned durotomy using each approach, interlaminar (left) and laminoplasty (right).