STUDY DESIGN : CASE SERIES OF 19 PATIENTS OPERATED FOR SEVERE SCOLIOSIS AND DIASTEMATOMYELIA SÉRIE

Objective: Diastematomyelia is a rare congenital spine and spinal cord malformation in which the spinal cord is divided into two parts by the osseous or fibrous septum. The incidence of diastematomyelia in patients with the most severe forms of congenital scoliosis is much higher than its general incidence in the population. When performing surgeries to correct scoliotic deformities, the question arises regarding the choice of a strategy for managing the septum. An unambiguous answer to this question does not exist, since the disease is very rare and heterogeneous. The aim was to summarize the data on different surgical strategies for detecting diastematomyelia. Methods: Literature review and retrospective analysis of our own clinical data. Results: We present our own experience of treating 19 patients with diastematomyelia and severe congenital scoliosis. Posture disorder was corrected in all cases; the septum was removed in none of the cases. Significant correction was achieved for all patients, and no neurological complications were observed in the shortand long-term follow-up. Conclusions: Surgical nonremoval of the spur enables compensation to be achieved, without neurological complications either in the immediate postoperative period or in the long-term (more than 2 years) follow-up. Level of Evidence IV; Case seriesh.


INTRODUCTION
Diastematomyelia, also known as split cord malformation (SCM), is a congenital anomaly of the spinal cord and spine in which the spinal cord is split into two parts.As opposed to diplomyelia (the so-called true duplication of the spinal cord), SCM is associated with splitting of the spinal cord into two halves by a bone, cartilage, or fibrous septum located in the sagittal or near-sagittal plane.Diastematomyelia is one of the variants of the closed form of spinal dysraphism. 1 SCM was first described by Olivier in 1837. 2 According to the literature, the rate of SCM in patients with congenital scoliosis is very variable and ranges from 4.9% 3 to 41%. 4 SCM has two aspects: neurosurgical and orthopedic.The neurosurgical aspect is that the risk for the development and progression of neurological symptoms due to traction on the spinal cord by the separating spur increases with the patient's age.The orthopedic aspect involves problems related to full surgical correction of spinal deformities in the presence of the spur separating the spinal cord.
In connection with these aspects, there are various opinions relating to the possible management of patients with diastematomyelia who undergo surgery.There are two dominant viewpoints.4][5][6][7][8][9] The second suggests a differentiated approach, depending on patient's individual characteristics (e.g.] According to the literature, the risk of neurological complications for resecting the spur is relatively high. 5,9However, spur resection is not always beneficial in terms of the extent of the correction.We therefore took the view that the approach of saving the spur during corrective surgery on the spine would be advantageous.
We have the experience of surgical treatment of 19 patients in whom the intraspinal spur was not resected before correction of congenital scoliosis, which had no adverse consequences for the patients.

METHODS
Three hundred fifty eight patients with congenital spinal deformities were treated at the NRITO clinic between 1997 and 2015.In 22 of the patients (6.1%), diastematomyelia was diagnosed during examination.Only 19 of them (13 males and 6 females, with a mean age 11.6 (6−19)) were included in the study group because two patients did not undergo corrective intervention, and one patient had previously been operated on for resection of the intracanal spur.The examination included plain radiographs of the spine in 2 standard upright projections, spondylograms in the lateral bending position, MRI, and MSCT.All the study participants signed an Informed Consent Form.
Scoliosis was diagnosed in 16 patients, and kyphosis was detected in 3 patients.Fourteen patients had bone spur, and 5 patients had fibrous spur.The spur was located in the thoracic spine in eight cases, in the thoracolumbar spine in five cases, and in the lumbar spine in six cases.The spur length varied from 2 to 8 segments and was not longer than three segments in most cases.A concomitant congenital pathology was found in seven patients: syringomyelia (2 cases), joint deformities of the upper and lower extremities (4 cases), Chiari malformation, keeled chest, and kidney duplication.
Preoperatively, 8 patients were neurologically intact, and 11 patients were detected as having neurological symptoms of varying severity: six patients with lower extremity paraparesis or monoparesis (with pelvic dysfunctions in two cases), one patient with pyramidal insufficiency syndrome, and four patients with hypotrophy and hypotension of the lower extremity muscles and lack of Achilles or abdominal reflexes.
The main aim of surgical treatment was to stop the progression of spinal deformity by corrective treatment, and we did not resect the spur in all 19 cases.
Various surgical instrumentations were used to correct the spinal deformity: segmental instrumentation with a hook or hybrid fixation was used in 14 cases; Antares was used in one case; VEPTR instrumentation was used in four children under 10 years of age (the patients underwent 3, 4, 6, and 7 serial distractions).The spinal cord function was monitored using a wake-up test or evoked somatosensory potentials.

RESULTS
The mean Cobb angle of scoliotic deformities in 16 patients was 82.9° (51−170°).In three patients with kyphotic deformities, the Cobb angle was 10° (a 9-year-old patient with the kyphotic apex in the lumbar spine), 118°, and 160°.In the scoliosis patients, the deformity in the lateral bending position was 66.4º (40−105º), on average, i.e. spine mobility was 19.9%.Details of the characteristics of all patients are listed in Table 1.
In the immediate postoperative period, the mean scoliotic deformity was 59.6° (25−104°), i.e. the correction amounted to 23.3°, or 28.1%, of the initial value (table 2).It was not possible to calculate the mean values in patients with kyphotic deformities because control radiography was not conducted in a patient with kyphosis of 118° due to a complication (early abscess).The normal sagittal contour of the lumbar spine was restored in a patient with lumbar kyphosis.Angular kyphosis in the third patient was reduced to 63°.The mean postoperative follow-up period was 32 (24−74) months.At that moment, the mean scoliotic deformity was 67° (31−101°), which means that the total loss of correction was 7.4°.The correction achieved was maintained in patients with angular kyphosis.The patient with lumbar kyphosis developed clinically significant proximal junctional kyphosis (PJK), which required two re-interventions.However, PJK recurred both times.In addition, two early deep abscesses developed, which required removal of an endocorrector in both cases.
No cases of new neurological symptoms or worsening of the preoperative symptoms were observed in the short or long-term postoperative periods.
In our view, our findings on the lack of new neurological symptoms over sufficiently long follow-up periods, in combination with significant correction of deformities, demonstrate that the tactic of saving the spur is an effective and safe approach to minimize risks of neurological symptoms associated with the corrective interventions.

DISCUSSION
Diastematomyelia is a rare disease.However, because of the pathological changes that accompany the disease, the problems associated with the diagnosis and surgical features of SCM are widely discussed in the scientific literature.The issue most actively debated is whether or not the spur should be removed.
We hypothesized that the conservative approach to resection/ non-resection of the diastematomyelic spur, in cases without obvious neurological symptoms caused by the spur, would help minimize the risks.
The orthopedic approach to the problem of diastematomyelia is characterized by wide variability.There are three variants of the surgical approach to the patient with progressive congenital scoliosis who is diagnosed with diastematomyelia: 1. Resection of the intracanal spur in all cases, before the scoliosis correction surgery; 2. Resection of the intracanal spur without scoliosis correction; 3. Scoliosis correction without resecting the intracanal spur.

Group I is represented by the following studies
Winter et al. 3 observed 27 patients with congenital scoliosis and diastematomyelia.Laminectomy and resection of the septum did not provoke worsening of neurological symptoms in 19 patients, in a 6-month follow-up period.The authors concluded that resection of the spur is important to prevent the progression of neurological deficit; therefore, corrective surgery should be delayed until the septum is removed.
McMaster 4    indications for surgical treatment in these patients, preventing the progression of neurological disorders and preventing neurological complications.More than 20 patients underwent preventive removal of the spur without changes in neurological status, achieving very significant correction (from 66º to 14º).Only one neurological complication was observed by the author, in a patient with severe deformities who underwent a massive two-stage intervention with resection of the spur.The motor and sensory functions almost completely recovered with time.The author is convinced that the intraspinal anomaly should be removed in any patient with congenital scoliosis, regardless of the neurological status.
Wenpeng Liu et al. 12 followed-up a group of 48 patients.SCM I was detected in 47 of them.Laminectomy and spur resection were performed in all patients before a corrective intervention.The authors did not report the results and complications but recommended a similar approach in all cases of congenital scoliosis and diastematomyelia because, in their opinion, the intraspinal spur might contribute to the progression of congenital scoliosis.
Ayvaz et al. 13 operated on 32 patients with diastematomyelia and congenital scoliosis.All patients with SCM I (18 people) underwent resection of the spur before the corrective intervention.Only orthopedic surgery (deformity correction and spinal fusion) was performed in SCM II.Correction of scoliosis amounted to 44−47%.Temporary neurological symptoms were detected in two cases of SCM I; no other complications occurred.The authors recommend using this approach for type I and II diastematomyelia.
Hui et al. 14 demonstrated a similar approach to solve the problem.They operated on 45 patients (15 patients with SCM I and 30 patients with SCM II).All the operations were one-stage surgery, but the spur was resected only in patients with type I anomalies.The deformity was corrected from 73.7º to 33.5º (54.5% of correction).All complications (two cases of mild neurological symptoms and one case of liquorrhea) were corrected.The authors believe that spur resection before correction of scoliosis is indicated for SCM I but not for SCM II.
This differentiated approach to types I and II diastematomyelia associated with congenital scoliosis was also demonstrated by Lifeng Lao et al. 15 They operated on five patients.The spine deformity was decreased from 63º to 30.2º (57.2% of correction).No complications were observed.Two patients had a partial regression of preoperative neurological symptoms.The authors recommend a single-stage intervention, with resection of the spur only in the case of SCM I.
Sheng-Li Huang et al. 16 reported on 156 patients operated on for resection of the spur.One patient died due to anesthesia complications (anesthetic accident).Liquorrhea developed in 2 cases, transient neurological symptoms occurred in 4 cases, and epidural hematoma occurred in one patient.There were no persistent neurological complications.The authors suggest that surgery be indicated for all patients with SCM I and neurological symptoms or tethered filum, while conservative treatment be indicated for those with SCM II who are asymptomatic for SCM I.

Group II includes the following studies
Pang et al. 17,18 who suggested the type I and II diastematomyelia classification, depending on the anatomical features, operated on 39 patients and observed an improvement or stabilization of neurological symptoms in 89% of cases.
Schijman 8 performed 17 operations and achieved control of negative neurological changes in 13 patients (76.4%).Of 4 patients who were neurologically intact before surgery, the condition remained unchanged in three patients, and symptoms completely disappeared (after temporary worsening) in one patient.
Massive clinical data are presented by a group of neurosurgeons from New Delhi, 5,20,21 which allowed them to make a significant addition to Pang's classification, by allocating four anatomic subtypes within the SCM I that characterize spatial relationships between the spur and halves of the dural sac.Of 254 patients who underwent spur resection, 68 patients had a partial regression of neurological symptoms; the neurological status remained unchanged in 160 patients; worsening occurred in 15 cases.In addition, 23 patients had liquorrhea, and 6 patients had abscesses.The vast majority of complications were corrected.In a separate paper, the same authors 22 analyzed the treatment outcomes in four patients with the dorsal localization of the intraspinal spur.There was one conclusion in all cases: regardless of the symptoms and diastematomyelia type, the spur should be resected.E.V. Ulrikh, 9 in his monograph, presented the results of surgical treatment of 87 patients.The intraspinal spur was removed in all cases to prevent aggravation or development of neurological symptoms.An improvement in the neurological status was observed in 15 patients (17.3%); worsening of the neurological status was observed in 16 (18.3%)patients, with two of them developing plegia (with partial or full recovery in six cases); the clinical symptoms remained unchanged in 55 (63.2%) patients.One child died due to productive meningoencephalitis.
Borcek et al. 23 reported 34 operated patients with both types of diastematomyelia.An improvement in neurological status was observed in 16 (47%) patients; worsening of the neurological status was observed in four (11.6%) patients; there were no changes in 14 cases.
Only a few authors have demonstrated an ambiguous surgical approach to prevent progression of the pathological process.Cheng et al. 24 analyzed the results of surgery in 112 patients and concluded that the surgery was much more effective in the case of SCM I. Huang et al. 16 operated on 156 patients and concluded that surgery should be performed in patients with SCM I in the presence of progressive neurological symptoms.Conservative treatment is only required in patients with asymptomatic SCM I and SCM II.
Sinha et al. 5 reported a large group of patients (n=203).Motor deficit (weakness and atrophy of limb muscles, gait disorders) was observed in 148 (72.9%) patients, sensory disorders (dysesthesia, hypoesthesia, trophic ulcers, and finger autoamputation) were found in 80 (30.4%) patients.Pelvic organ dysfunction was detected in 66 (32.5%) patients; no neurological symptoms were observed in only 27 out of 203 patients.Among the patients with preoperative neurological symptoms, 60 (40.4%) reported improvement in motor function; 33 (41.2%) patients, improvement in sensory function; 20 (30.3%) patients, improvement in pelvic organ function, and 13 (52%) patients, reduced degree of trophic disorders.Fifteen patients reported aggravation of neurological symptoms immediately after the intervention; no recovery was observed in three of these patients.

Group III includes the following studies
Guixing and Jiaming 10 reported that out of 500 patients with congenital scoliosis, 95 had with anomalies of the spinal cord.Surgical treatment did not involve resection of the diastematomyelic spur.Initial neurological symptoms were found in 41 patients, but no postoperative worsening of the symptoms was observed in any case.
Bollini et al. 11 believe there is no need to resect the spur if the expected correction of scoliosis is "not so significant".
On the basis of the literature data, we conclude that in patients with congenital scoliosis and diastematomyelia, the risk of neurological complications associated with resection of the bone spur is higher than that associated with corrective treatment of the spinal deformity.In most of the 19 patients of our group, the spinal deformity can be assigned to the neglected category; should be changed to "but in all cases we decided not to perform preparatory resection of the diastematomyelic spur.As the main task of surgical treatment, we aimed to stop the progression of spinal deformity.That is why the above correction was relatively small and slightly exceeded the values of preoperative spine mobility, which was determined using spondylograms in the lateral bending position as 23.3º.

CONCLUSION
There is currently a wide range of surgical approaches for diastematomyelia, with no conventional surgical approach, due to the lack of clinical data and in-depth analysis of the issue.Often, the decisions on the choice of approaches were local, based on the personal experience of the researchers.For this reason, the data obtained are highly heterogeneous and difficult to analyze systematically.We believe that further accumulation of clinical data and the development of a differentiated approach to the choice of a surgical approach are required.
Our study included the relatively small number of cases which were heterogeneous by the nature of spine deformity and by duration and severity of disease.However, the chosen surgical approach demonstrates the potential for application in surgical practice.This is confirmed by our high levels of compensation achieved and the lack of neurological complications, both in the immediate postoperative period and in long-term (more than 2-year) follow-up.
All authors declare no potential conflict of interest related to this article.

9
posterolateral wedge-shaped defect of the T5 vertebra; butterfly defect of the T7, T9-T11 vertebrae; concrescence of the arches of the T4-T6, T7-T10 vertebrae; multiple synostosis of the ribs on the right and left sides; spina bifida posterior of the C6-T3, T11-L1 vertebrae No Bony septum of the spinal canal at the apex of kyphosis (T8-T10) lying in the sagittal plane and separating the spinal canal into the right and left lateral 5,10-11, L3-4 vertebral discs; concrescence of spinous processes of the L2-L3 vertebrae, the 2 nd and 3 rd ribs on the left side; spina bifida at the T10-T12, L3-L4 level No Bony septum at the T9-L1, L1-2 vertebral discs; concrescence of the posterior portions of vertebrae at the same level No Bony septum at the T10posterior defect of the L2 hemivertebra, hypoplastic T12-L2 vertebral discs No Splitting of the medullary cone into two halves at the level of the the T3-T8 vertebrae; wedge-shaped hemivertebra between the T2 and T3 vertebrae on the left side; concrescence of the T2-T4, T7-T10 vertebrae at the level of their arches; hypoplastic T3-4, T5-6, and T6-7 intervertebral discs; spina bifida posterior of the T5, T6, T10-T12 vertebrae No Bony septum at the T7-T8 level dividing the spinal canal into the right and left halves 13 Kyphoscoliosis dButterfly defect of the T11 vertebra with the right half of the vertebral body predominating; the left half of the body is fused with the L1 vertebral body; L2 is the right-sided lateral wedge-shaped hemivertebra; concrescence of the bodies of the T12 and L3 vertebrae.Hypoplastic T12-L1 -L2-3 discs.Spina bifida posterior of the T12 -S2 vertebrae.Left-sided concrescence of the ribs A CSF-filled cyst (

Table 1 .
Baseline characteristics of patients.

Table 2 .
Main outcomes on patients.
st -3 rd ribs on the right side Fixed spinal cord, syringomyelia at the T5-7 level.