Comparative study of complex spina bifida and split cord malformation
http://www.100md.com
《美国医学杂志》
1 Department of Neurosurgery, Sanjay Gandhi Post Graduate Institute of Medical Sciences & King Georges Medical University, Lucknow, India
2 Department of Neuroanestheliology, Sanjay Gandhi Post Graduate Institute of Medical Sciences & King Georges Medical University, Lucknow, India
Abstract
OBJECTIVE: To see the difference in clinical profiles, radiological findings and surgical outcome of the group 1 split cord malformation and meningomyelocele (SCM with MMC) from group 2 (SCM without MMC). METHODS: 46 patients of SCM were selected from a total of 138 cases of spinal dysraphism. They were divided into two groups, based on presence or absence of MMC. Group I (SCM with MMC) n =19 patients and Group II (SCM without MMC) n=27 patients. A detail clinical evaluation and MR screening of whole spine of all cases was performed. All patients underwent surgical detethering of cord. After an average follow-up of 1.7 years, the operative results were clinically assessed and statistical significance was calculated. RESULTS: Male to female ratio was1:09. Mean age of presentation was 3.6 years. Cutaneous markers like tuft of hair, cutaneous haemangioma, etc, had a higher incidence in group II in comparison to group I (50% vs 10.5%). The incidence of motor deficits was significant in group I in comparison to group II (63% vs 40%). The incidences of sensory loss, trophic ulcers, sphincteric dysfunction and muscle atrophy were relatively more common in group I patients, while neuro-orthopedic deformities such as congenital telepes equinovarus (CTEV), scoliosis and limb shortening were more frequent (67%) in group II children as compared to group I (53%). Type I SCM has higher incidence in group I children. Low lying conus were found in 47% patient of group I, while in group II it was noticed in 69%. The associated cranial anomalies like hydrocephalus, ACM and syrinx, were slightly higher in group I patients. At surgery, dysgenetic nerve roots, neural placode, arachnoid bands and atrophic cord were seen mainly in group I. Postoperative complications like, CSF leak, pseudomeningocele and meningitis were more commonly encountered in group I patients. The patients of group II showed better operative outcome compared to group I cases. CONCLUSION: Incidence of SCM with MMC amount to 41% of total SCM cases. Progressive neurological deficit was higher in this group (SCM with MMC) in comparison to the group harboring SCM without MMC. In view of a significant association of SCM in MMC cases, associated with other craniospinal anomalies, a thorough screening of neuraxis (by MRI) is recommended to treat all treatable anomalies simultaneously for desired outcome.
Keywords: Split cord malformation; Complex spina bifida; Spinal dysraphism
Split Cord Malformation (SCM) is the most common underlying anatomical abnormality of occult spinal dysraphism.[1],[2] Understanding of the split cord malformation has become better after the paper of Pang in 1992.[3],[4] SCM may present with MMC (complex spina bifida)[5] or without MMC as an occult anomaly.[3],[6] SCM commonly presents with various cutaneous markers like hairy patch, subcutaneous lipoma, cutaneous capillary hemangioma and uncommonly with meningomyelocele sac.[5],[7] Children with SCM may not have any neurological deficit at the time of birth but deficits may appear with growth of the child, attributed to spinal cord tethering.[3] Once neurological deficits have occurred, often they do not improve completely even after detethering.[5] MMC is a crippling condition manifesting with mild to severe neurological deficit and usually with less number of cutaneous markers other than overt MMC sac. Hence, the clinical spectrum of SCM and MMC is relatively different. With the advent of MRI, the diagnosis of SCM has become relatively frequent. MRI not only deciphers the precise anatomy but also helps in detecting the associated lesions. Though, SCM has been variably described in recent literature, presently very scarce material is available regarding the problem of SCM, when it is associated with meningomyelocele.
Present study was carried out to compare the clinical profiles, radiological findings and surgical outcome of patients having SCM with MMC (Group I) and SCM without MMC (Group II).
Materials and methods
Forty six patients of SCM were selected and analyzed from a total of 138 cases of spinal dysraphism, studied prospectively and retrospectively (126/12), in the Neurosurgery department of Sanjay Gandhi Post Graduate Institute of Medical Sciences, Lucknow, India between June 1989 and June 2001. The study was carried out to compare the clinical profiles, radiological findings and surgical outcome of the patients having SCM with MMC and occult SCM alone. Forty six cases were divided into two groups, based on presence or absence of MMC; Group I (SCM with MMC) n=19 patients and Group II pure occult SCM (SCM without MMC) n=27 patients. Male to female ratio was 1: 09. Mean age of presentation was 3.6 years for all cases (range 2 days to 14 years). A detailed clinical examination of each case was recorded. Eight patients were referred with plain X-ray spine (6 of group I and 2 of group II). One patient of group I came with plain CT of spine. Spinal MRI was performed in all 46 cases, where entire spine was screened. However, cranio-spinal MRI could be possible in 27 patients on account of financial reasons. SCM was classified into type I and Type II depending on the presence of bony spur or fibrous septum respectively. This typing was based on radiological and operative findings Figure1 and Figure2. All the patients underwent surgical excision of bony spur/fibrous septum, and repair of MMC depending on the type of anomaly. Postoperatively the patients were examined on 7th day and then followed up periodically at 1 months, 3 and 6 months interval. After an average follow-up of 1.7 years (range 6 weeks to 6 years), the operative results were clinically assessed.
Clinical findings
Bar diagram 1 to 5 and table1, table2 show comparison of the incidence of cutaneous markers, neuro-orthopedic syndromes, associated anomalies, clinical profiles and outcome in both group I and II.
GROUP I (SCM WITH MMC)
Group I comprised of 19 children having congenital meningomyelocele, or lipomeningomyelocele along with SCM. Fifteen children of this group presented with overt meningomyelocele, where MRI showed underlying split cord malformation. Progressive sensory and motor deficit was found in 9 and 12 cases respectively. Out of 12 children with lower limb weakness, 3 had paraplegia with complete anesthesia. Autoamputation of little toe was seen in two children and sphincter disturbance was noted in 6. Three children showed non-healing trophic ulcers on their feet. Neuro-orthopedic syndrome (NOS) was observed in the form of scoliosis/hemivertebrae and congenital talipes equino varus in five each, congenital dislocation of hip in two, rib cage anomaly in one and leg length discrepancy in two cases. Two children with MMC also had a patch of hair near the sac. Children having MMC and SCM simultaneously were labeled as harboring the "complex spina bifida".[5]
Four of 19 children were operated for MMC at peripheral hospital at birth by general surgeon without adequate investigation, where incomplete surgery was performed. These children had progressive deterioration in neurological function within 6 years of their primary repair, the oldest child being 6-years-old-male with non-healing trophic ulcer over the right great toe and 50% hypoaesthesia on dorsum of foot. Another 4 years old boy, whose meningomyelocele was operated, following the shunt insertion for hydrocephalus at one month of life, presented with rapid onset of paraplegia with sphincteric dysfunction and anesthesia below L-1 dermatome. One patient had progressive bilateral congenital talipes equino varus and another had history of frequent falls and weakness of both lower limbs, after two years of surgery. MR scan of these children revealed underlying split cord malformation.Figure3, Figure4, Figure5, Figure6, Figure7
GROUP II (SCM WITHOUT MMC)
Group II, consisting of 27 children, presented with cutaneous stigmata, progressive neurological deterioration or both, leading to clinical suspicion of tethered cord syndrome. Of cutaneous markers, hypertrichosis was present in thirteen children, cutaneous capillary hemangioma in four and dermal sinus with skin dimple in three. A 14-year-old boy had multiple neurofibromas all over the body with cervical cord splitting and anamolous vertebral bodies. Two patients in this group presented with low backache paraesthesia and progressive weakness of lower limbs. Eleven of the twenty seven children manifested mainly with progressive and asymmetric motor deficit, one child had left leg weakness only. Ten patients had graded or patchy hypoaesthesia below the level of lesion. One child presented with non-healing trophic ulcer on the foot at the pressure points. Seven patients had sphincteric involvement in the form of urinary incontinence and constipation of varying severity.
Amongst the neuro-orthopedic syndrome (NOS), congenital talipes equino varus (one bilateral) was seen in five children and nine had progressive scoliosis. Rib cage anomaly in 1, limb length discrepancy in three, high arch foot in 1 and valgus deformity of foot in 1 were other types of NOS seen in this group of patients.
Radiological Findings
Plain X-ray spine of two patients of group I had bony spur visible at the lumbar and dorsal spine, two had hemivertebrae in dorsal spine and 4 patients had deficient posterior arch.
Out of total 46 cases of SCM, 38 patients had type I and 8 had type II SCM. Fifteen children of group I had type I SCM, while four had type II SCM. In group II patients SCM type I was present in 23 of 27 and SCM type II in four cases.
Bar diagrams Figure6 and Figure7 show the incidence of various associated spinal and cranial lesions in both group I and II cases of SCM. Only nine patients of group II, who had split cord in dorsal region, had normal position of the conus. The other eighteen cases had their conus ending between L2 and S2. Of 19 patients of group I, nine had low-lying cord ending below L2, whereas in ten patients conus was lying at the level of L2. Hydrosyrinx was observed in seven patients of group I and in 4 of group II children. Hydrosyrinx in both the groups remained rostral to the defects and extent was 2-4 vertebral level except in one of group I patient, who had holocord syrinx. None of these patients warranted surgery for the syrinx in isolation. Neural placode was seen in all patients of group I, whereas intraspinal lipoma in 4, dermoid in 1, arachnoid cyst in 2 and thick/fatty filum seen in 5 cases. In group II other associated tethering lesions were intraspinal lipoma in 7, intraspinal dermoid in 3 patients and neurenteric cyst in only 1 patient. Neural placode was not found in any of the children of group II.
Five patients of group I had hydrocephalus, of which two were symptomatic to have enlargement of head and bulging fontanelle. These two required shunt CSF diversion before definitive surgery. Two children became symptomatic on long-term follow-up with progressive enlargement of head, proved on serial scans and required ventriculoperitoneal shunting while one other child remained asymtomatic. Three children of group II had hydrocephalus, two of them had symptomatic hydrocephalus and required shunt before definitive surgery. One patient, whose preoperative scan showed minimal ventricular dilatation, required shunt in follow-up, as he developed clinically significant increase in head size.
In group I, two children had dysgensis of corpus-callosum and another two showed Arnold Chiari malformation More Details. Aqueductal stenosis was noticed in 2 patients. In group II, none of the patients had Arnold Chiari Malformation (ACM), but aqueductal stenosis was noted in two children.
Surgical procedure and findings
Surgical Procedure
Bony spur in type I SCM was extradural, to lie between two dural tube. After laminectomy, bony spurs were identified following separation of two dural sleeves. Complete excision was done with small rongeur or micro drill. Both dural tubes were opened in their midline; usually there are fibrous adhesions with hemicords and dura on medial side. After opening dura, a thorough adhesiolysis was done and dural closure performed posteriorly in such a manner to form a common tube for both the cords. In type II SCM, the two cords were exposed and identified till the two become single rostrally and caudally. Fibrous septum between two was excised, arachnoids adhesions and anomalous nerve roots were also divided to allow free mobility of both the cords. A tight filum terminale, if diagnosed was also divided and a watertight dural closure done. Excision and repair of MMC sac were performed simultaneously in cases having meningomyelocele. At times, fascial graft or synthetic dura was used for lax dural closure. Associated tethering lesions like dermoid, lipoma were also removed simultaneously.
Surgical Findings
In group I, SCM was found one to two vertebral levels rostral in 14 cases, while it was at the level of meningomyelocele in five cases. In none of the case, SCM was caudal to MMC. While in group II, SCM was found either beneath the cutaneous marker or in close proximity, coinciding with bony anomalies of spina bifida.
Various Image findings (as described under radiological findings) were confirmed during surgical procedure. However, four of group I and one of group II patients macroscopically showed that some of the nerve roots were seen very thin and atrophic as compared to others. In five of group I and three of group II children, few nerve roots were seen significantly adherent to dura with arachnoid bands. Considering these as additional tethering lesions, these were also divided. One case of group I had intramedullary abscess without the prior evidence of infection, dermal sinus or previous surgery.
Surgical Outcome
Pre-operative sphincter dysfunction was present in 13 patients (group I = 6, group II = 7) Five patients belonging to group II showed satisfactory improvement subjectively, whereas only four children of group I had improvement in the bladder function with increase in their dry periods. Neuro-orthopedic syndrome (NOS) became stabilized in all patients of group I and group II, no further improvement or deterioration was noticed in follow-up. However, we feel that the physiotherapy and management of neuroorthopedic deformities were not adequate in most of these children.
Post operative complications
Post-operative CSF leak occurred in eight patients of group I and three of group II. Five of group I and one of group II required re-exploration and repair of dural defect with facia lata patch or synthetic graft (Gore tex), whereas 3 of group I and 2 of group II were treated conservatively with pressure dressing and acetazolamide therapy and they responded well. Pseudomeningocele was noticed in five of group I and three of group II patients. One patient of group I required thecoperitoneal shunt. Two patients required reexploration and repair, whereas in two patients only aspiration and pressure dressing were needed. In group II, one child required reexploration and repair, whereas in two patients only aspiration and pressure dressing were done. Four patients in group I and two of group II developed meningitis. All of them responded to antibiotic except one patient of group I who died due to fulminant meningitis. One child of cervical complex spina bifida expired while recovering in post-operative period due to sudden respiratory arrest. Instability of cervical spine was blamed for such catastrophe.
Discussion
SCM is not an uncommon finding in spinal dysraphism.[1],[8],[9],[10],[11] On review of our total 138 patients of spinal dysraphism SCM constituted 46 of all, reflecting an incidence of SCM to be 33 % (46/138), slightly more than reported by others, which may be on account of referral bias.[3] Nineteen children of MMC had SCM, which represent 14% of total spinal dysraphism cases. Previous pathological studies of children with MMC revealed a significant incidence of associated SCM. Campbell et al[8] found 36 cases of SCM in 100 infants of open spinal dysraphism and Emery & Lendon[9] reported 78 cases amongst 100 spinal dysraphic patients. In the present study nineteen children of MMC with unsuspected SCM could not have been diagnosed without an awareness of such coexistence. Hence, it is apparent from the above observations that the association of SCM in MMC is not an infrequent finding, which needs a special attention while screening and planning to treat these children for favourable results.
SCM is more common in females according to western literature[2],[3],[11] but male predominance is seen in Indian subcontinent, including the present study.[5],[6],[12]
The skin manifestations in occult dysraphism represent minor aberrations in the development of the surface ectoderm, brought about by the adverse influences of a dorsal endomesenchymal tract, but these changes are overshadowed by changes in the surface ectoderm in case of an associated meningomyelocele.[4],[5],[13],[14],[15] The skin markers are quite frequent in occult spinal dysraphism i.e. in SCM. In the present series, the incidence of cutaneous marker is 90% in group II cases as compared to only 10.4% in group I,[3],[4],[10],[16] signifying a low incidence of cutaneus markers in SCM presenting with meningomyelocele. It is an important observation because these cases of SCM have a higher chance of missed diagnosis without awareness. Amongst these markers the hypertrichosis was the commonest, being present in 50% of the cases of group II. The other markers like capillary hamangiomas and dermal sinus were seen in group II only and not in group I. This agrees to the consensus of incidence of cutaneous markers in occult dysraphism varying between 20 to 75%.[4],[7],[17],[18],[19],[20],[21]
Neurological deficits, as such are common in meningomyelocele, these become further dense in presence of other tethering lesions like SCM in same patients.[13],[22] Considering the high possibility of progressive and often irreversible neurological deterioration, it is logical to advocate the screening of entire spine with MRI in every spina bifida child. If SCM is found in association with MMC, it should be treated simultaneously to circumvent its deleterious effect on the developing cord.[4] While comparing both the groups, it was observed that the incidence of neurological deficit was more significant in group I as compared to group II: motor weakness (63.1% vs 40.7%), sensory loss (47.3% vs 37%) and sphincter involvement (31.5% vs 25.9%) table1. Following surgery the improvement in neurological functions were more marked in group II cases compared to group I: motor weakness (54.4% vs 33.3%), sensory functions (70% vs 66.6%) and bladder control (71.4 vs 66.6%) table2. While the incidence of neuro-orthopedic syndrome was 66.6% in group I and 96% in group II patients, this is in accordance with other reported series.[1],[4]
Children with MMC or previously operated for open defect tend to have more wound complication and CSF leakage, because of the thin skin covering the lesion.[4] Similar observations were made in the present study (42.7%).
Amongst associated craniospinal anomalies, hydrosyrinx was seen in 7 of 19 (36.8 %) cases of group I and 4 of 27 (14.8 %) of group II, which is within the reported range of 25 to 40%.[11],[15],[20] but relative incidence is higher in group I cases in this series. Amongst the other anomalies Chiari malformation (ACM) was present in 13% of cases of group I and none in group II suggesting an association, more with MMC and least with SCM. Moreover the incidence of clinically significant ACM in cases of spinal dysraphism ranges from 30 to 50% in existing literature,[14] but no comparative data is available for two groups of SCM reported by us. The overall incidence of hydrocephalus in meningomyelocele is reported as 86% in western literature;[15] in the present study, it was encountered in 5/19 (26.3%) of group I patients and 3/27 (11.11%) of group II patients, which is 17.39% of total number of cases. Our previous studies[5],[6] and other studies[12] from Indian subcontinent revealed almost similar low incidence of hydrocephalus in cases of SCM. Though the incidence of hydrocephalus is relatively more in cases of SCM with MMC, a larger study would be required to establish it.
Incidence of low lying cord in SCM with MMC is reported 30% by Iskander whereas in pure SCM it was found in 40.5% of cases as reported by Ersahin et al.[10] In our series, it was noticed in 47.3%, and 66.6% of group I and group II patients respectively. We encountered relatively higher incidence of low lying cords in both groups of cases. Intraspinal lipoma was the significant finding in the present study, both in group I, 4/19 (21.0%), and group II 7/27 (25.9%). Barring two patients in group I, in whom lipoma was rostral to the split cord, all the patients had distal lipomas in both the groups. This coincides with the reported incidence of terminal lipomas by Pang[4] who encountered 6 lipomas in total 37 cases of SCM. Other associated lesions with SCM are dermoid, epidermoid, arachnoid cyst and neurenteric cyst, as reported in literature.[6],[12],[23]
Type I SCM was encountered in 38 of 46 (82%) of our cases, while type II was seen in rest of 8 children. As an additional observation, 15 of 19 (79%) children of group I (SCM with MMC) had type I SCM signifying a higher incidence of bony spur in cases of SCM associated with meningomyelocele. It was interesting to note that 14 of 19 split cord malformations were located one or two vertebral levels rostral to myelomeningoceles and remaining 5 were found at the level of meningomyeloceles itself. It is obvious that no SCM was found below the level of MMC. The type I SCM is reported to have higher incidence of 75% in comparison 25 of type II in literature.[10],[11],[12] A higher incidence of type I SCM is reported in cases associated with MMC by Pang et al[10] and Ersahin et al.[10]
Occasionally, the shortened nerve roots, conjoined nerve roots, angulated nerve roots and arachnoid bands are seen at the level of dysraphic state, causing tethering of cord. [24],[25] Lassman and James have devised the term 'meningocele manque' for the cases in whom adhesion of the conus, filum, cauda equina roots to the inner aspect of the posterior dura occur.[26] The present series also witnessed the thin and aberrant nerve roots in 4 of goup I and1 of group II patients, whereas nerve roots were adherent to dura with arachnoid bands in 5 of group I and 3 of group II cases. The presence of such structures, most probably signify the additional tethering elements, need to be catered during surgery.
Conclusion
The incidence of SCM with meningomyelocele being 41% of total SCM cases is obviously significance. The prognosis of SCM with MMC is not as good as noted in pure SCM cases. Progressive sensorimotor deficit also has higher incidence in cases of SCM with MMC, whereas the incidence of NOS is relatively more in pure SCM cases. A screening of entire neural axis is seemingly indispensable, in order to cater to the occult associations simultaneously (in cases of overt spina bifida) for better outcome. Because of significant occurrence of MMC in SCM and concurrent hydrocephalus, it seems appropriate to investigate these children thoroughly by cranio-spinal MRI. In case of equivocal radiological findings, SCM should be searched one to two level higher than MMC, if not present at same level in view of its higher incidence of rostral location.
Acknowledgement
The authors are extremely grateful to Mr. A.P. Dhar Dwivedi for preparation of this manuscript.
References
1. Andar UB, Harkness WFJ and Hayward RD: Split cord malformation of the Lumbar region. Pediatr Neurosurg 1997;26: 17-24.
2. Anderson FM: Occult spinal dysraphism: A series of 73 cases. Pediatrics 1975; 55: 826 - 835.
3. Pang. D, Dias MS & Ahab-barmada M: split cord malformation: Part I; unified theory of embryogenesis for double spinal cord malformation, Neurosurg 1992 ; 31: 3, 451-480.
4. Pang. D; split cord malformation: Part II: clinical syndrome, Neurosurg 1992; 31: 3, 481-500.
5. Kumar R, Bansal KK, and Chhabra DK: Occurrence of Split Cord Malformations in Meningomyelocele: Complex Spina Bifida. Pediatr Neurosurg 2002 ; 36: 119 - 127.
6. Kumar R, Bansal KK, and Chhabra DK: Split cord malformation in pediatric patients: out come of 19 cases. Neurology India 2001; 49: 128-133.
7. Eid K, Hochberg J, Saunders DE. Skin abnormalities of the back in diastematomyelia. Plast Reconstr Surg 1979; 63: 534- 549.
8. Campbell LR, Dayton D and Sohal GS. Neural tube defects: A review of human and animal studies on the etiology of neural tube defects. Teratology 1986; 34: 171-187.
9. Emery JL and Lendon RG. The local cord lesion in neurospinal dysraphism (Meningomyelocele). J Path 1973; 110: 83-96.
10. Ersahin Y, Mutluer S, Kocaman S, and Demirtas E: Split cord malformations in children. J Neurosurg 1998 ; 88: 57 - 65.
11. Iskander BJ, Mclaughlin C and Oakes WJ. Split cord malformation in myelomeningocele patients. Br J Neursurg 2000 ; 14(3): 200-203.
12. Jindal A, Mahapatra AK. Split cord malformation - a clinical study of 48 cases. Indian Pediatrics 2000 ; 37: 603 - 607.
13. Miller A, Guille JT, Brown Jr: Evaluation and treatment of diastematomyelia. J Bone Joint Surg (Am) 1993; 75: 1308-1317.
14. Reigel DH, Rotenstien D. Spina bifida. In pediatric neurosurgery by section of pediatric neurosurgery of American association of neurological surgeon, 3rd edition, USA; WB Saunders, 1994; 51-76.
15. Rokos J. Pathogenesis of diastematomyelia and spina bifida. J Path 1975; 117: 155-161.
16. Ersahin Y, Demirtas E, Mutluer S, Tosun A & Saydam: Split cord malformation: Report of three unusual cases. Pediatr Neurosurg 1996; 24: 155-159.
17. Dale AJD. Diastematomyelia. Arch Neurol 1969; 20: 309 - 317.
18. Guthkelch AN. Diastematomyelia with median septum. Brain 1974; 97: 729-742.
19. Hood RW, Riseborough EJ, Nehme AM, Micheli LJ, Strand RD, Neuhauser EB: Diastematomyelia and structural spinal deformities. J Bone Joint Surg 1980; 62 A: 520-528.
20. Humphreys RP. Spinal dysraphism. In Rengachary SS, Wilkins RH, eds. Neurosurgery, 2nd edn. Mc Graw - Hill, 1996.
21. James CCM, Lassman LP. Spinal dysraphism. The diagnosis and treatment of progressive lesions in spina bifida occulta. J Bone Joint Surg 1962; 44B: 828-840.
22. Meacham WF. Surgical treatment of diastematomyelia. J Neurosurg 1967; 27: 78-85.
23. Kumar R, Jain R, Rao KM, Nuzhat H. Intraspinal neurenteric cysts. Child Nerv Syst 2001; 17: 584-588.
24. McLone DG. Technique for closure of meningomyelocele. Childs Brain 1980; 6: 65-73.
25. Guthkelch AN, Pang D, Varies JK. Influence of closure technique on results in myelomeningocele. Childs Brain 1981;8: 350-355.
26. Lassman LP, James CCM. Meningocele manque. Child Nerv Syst 1977; 3: 1.(Kumar Raj, Singh SN, Bans)
2 Department of Neuroanestheliology, Sanjay Gandhi Post Graduate Institute of Medical Sciences & King Georges Medical University, Lucknow, India
Abstract
OBJECTIVE: To see the difference in clinical profiles, radiological findings and surgical outcome of the group 1 split cord malformation and meningomyelocele (SCM with MMC) from group 2 (SCM without MMC). METHODS: 46 patients of SCM were selected from a total of 138 cases of spinal dysraphism. They were divided into two groups, based on presence or absence of MMC. Group I (SCM with MMC) n =19 patients and Group II (SCM without MMC) n=27 patients. A detail clinical evaluation and MR screening of whole spine of all cases was performed. All patients underwent surgical detethering of cord. After an average follow-up of 1.7 years, the operative results were clinically assessed and statistical significance was calculated. RESULTS: Male to female ratio was1:09. Mean age of presentation was 3.6 years. Cutaneous markers like tuft of hair, cutaneous haemangioma, etc, had a higher incidence in group II in comparison to group I (50% vs 10.5%). The incidence of motor deficits was significant in group I in comparison to group II (63% vs 40%). The incidences of sensory loss, trophic ulcers, sphincteric dysfunction and muscle atrophy were relatively more common in group I patients, while neuro-orthopedic deformities such as congenital telepes equinovarus (CTEV), scoliosis and limb shortening were more frequent (67%) in group II children as compared to group I (53%). Type I SCM has higher incidence in group I children. Low lying conus were found in 47% patient of group I, while in group II it was noticed in 69%. The associated cranial anomalies like hydrocephalus, ACM and syrinx, were slightly higher in group I patients. At surgery, dysgenetic nerve roots, neural placode, arachnoid bands and atrophic cord were seen mainly in group I. Postoperative complications like, CSF leak, pseudomeningocele and meningitis were more commonly encountered in group I patients. The patients of group II showed better operative outcome compared to group I cases. CONCLUSION: Incidence of SCM with MMC amount to 41% of total SCM cases. Progressive neurological deficit was higher in this group (SCM with MMC) in comparison to the group harboring SCM without MMC. In view of a significant association of SCM in MMC cases, associated with other craniospinal anomalies, a thorough screening of neuraxis (by MRI) is recommended to treat all treatable anomalies simultaneously for desired outcome.
Keywords: Split cord malformation; Complex spina bifida; Spinal dysraphism
Split Cord Malformation (SCM) is the most common underlying anatomical abnormality of occult spinal dysraphism.[1],[2] Understanding of the split cord malformation has become better after the paper of Pang in 1992.[3],[4] SCM may present with MMC (complex spina bifida)[5] or without MMC as an occult anomaly.[3],[6] SCM commonly presents with various cutaneous markers like hairy patch, subcutaneous lipoma, cutaneous capillary hemangioma and uncommonly with meningomyelocele sac.[5],[7] Children with SCM may not have any neurological deficit at the time of birth but deficits may appear with growth of the child, attributed to spinal cord tethering.[3] Once neurological deficits have occurred, often they do not improve completely even after detethering.[5] MMC is a crippling condition manifesting with mild to severe neurological deficit and usually with less number of cutaneous markers other than overt MMC sac. Hence, the clinical spectrum of SCM and MMC is relatively different. With the advent of MRI, the diagnosis of SCM has become relatively frequent. MRI not only deciphers the precise anatomy but also helps in detecting the associated lesions. Though, SCM has been variably described in recent literature, presently very scarce material is available regarding the problem of SCM, when it is associated with meningomyelocele.
Present study was carried out to compare the clinical profiles, radiological findings and surgical outcome of patients having SCM with MMC (Group I) and SCM without MMC (Group II).
Materials and methods
Forty six patients of SCM were selected and analyzed from a total of 138 cases of spinal dysraphism, studied prospectively and retrospectively (126/12), in the Neurosurgery department of Sanjay Gandhi Post Graduate Institute of Medical Sciences, Lucknow, India between June 1989 and June 2001. The study was carried out to compare the clinical profiles, radiological findings and surgical outcome of the patients having SCM with MMC and occult SCM alone. Forty six cases were divided into two groups, based on presence or absence of MMC; Group I (SCM with MMC) n=19 patients and Group II pure occult SCM (SCM without MMC) n=27 patients. Male to female ratio was 1: 09. Mean age of presentation was 3.6 years for all cases (range 2 days to 14 years). A detailed clinical examination of each case was recorded. Eight patients were referred with plain X-ray spine (6 of group I and 2 of group II). One patient of group I came with plain CT of spine. Spinal MRI was performed in all 46 cases, where entire spine was screened. However, cranio-spinal MRI could be possible in 27 patients on account of financial reasons. SCM was classified into type I and Type II depending on the presence of bony spur or fibrous septum respectively. This typing was based on radiological and operative findings Figure1 and Figure2. All the patients underwent surgical excision of bony spur/fibrous septum, and repair of MMC depending on the type of anomaly. Postoperatively the patients were examined on 7th day and then followed up periodically at 1 months, 3 and 6 months interval. After an average follow-up of 1.7 years (range 6 weeks to 6 years), the operative results were clinically assessed.
Clinical findings
Bar diagram 1 to 5 and table1, table2 show comparison of the incidence of cutaneous markers, neuro-orthopedic syndromes, associated anomalies, clinical profiles and outcome in both group I and II.
GROUP I (SCM WITH MMC)
Group I comprised of 19 children having congenital meningomyelocele, or lipomeningomyelocele along with SCM. Fifteen children of this group presented with overt meningomyelocele, where MRI showed underlying split cord malformation. Progressive sensory and motor deficit was found in 9 and 12 cases respectively. Out of 12 children with lower limb weakness, 3 had paraplegia with complete anesthesia. Autoamputation of little toe was seen in two children and sphincter disturbance was noted in 6. Three children showed non-healing trophic ulcers on their feet. Neuro-orthopedic syndrome (NOS) was observed in the form of scoliosis/hemivertebrae and congenital talipes equino varus in five each, congenital dislocation of hip in two, rib cage anomaly in one and leg length discrepancy in two cases. Two children with MMC also had a patch of hair near the sac. Children having MMC and SCM simultaneously were labeled as harboring the "complex spina bifida".[5]
Four of 19 children were operated for MMC at peripheral hospital at birth by general surgeon without adequate investigation, where incomplete surgery was performed. These children had progressive deterioration in neurological function within 6 years of their primary repair, the oldest child being 6-years-old-male with non-healing trophic ulcer over the right great toe and 50% hypoaesthesia on dorsum of foot. Another 4 years old boy, whose meningomyelocele was operated, following the shunt insertion for hydrocephalus at one month of life, presented with rapid onset of paraplegia with sphincteric dysfunction and anesthesia below L-1 dermatome. One patient had progressive bilateral congenital talipes equino varus and another had history of frequent falls and weakness of both lower limbs, after two years of surgery. MR scan of these children revealed underlying split cord malformation.Figure3, Figure4, Figure5, Figure6, Figure7
GROUP II (SCM WITHOUT MMC)
Group II, consisting of 27 children, presented with cutaneous stigmata, progressive neurological deterioration or both, leading to clinical suspicion of tethered cord syndrome. Of cutaneous markers, hypertrichosis was present in thirteen children, cutaneous capillary hemangioma in four and dermal sinus with skin dimple in three. A 14-year-old boy had multiple neurofibromas all over the body with cervical cord splitting and anamolous vertebral bodies. Two patients in this group presented with low backache paraesthesia and progressive weakness of lower limbs. Eleven of the twenty seven children manifested mainly with progressive and asymmetric motor deficit, one child had left leg weakness only. Ten patients had graded or patchy hypoaesthesia below the level of lesion. One child presented with non-healing trophic ulcer on the foot at the pressure points. Seven patients had sphincteric involvement in the form of urinary incontinence and constipation of varying severity.
Amongst the neuro-orthopedic syndrome (NOS), congenital talipes equino varus (one bilateral) was seen in five children and nine had progressive scoliosis. Rib cage anomaly in 1, limb length discrepancy in three, high arch foot in 1 and valgus deformity of foot in 1 were other types of NOS seen in this group of patients.
Radiological Findings
Plain X-ray spine of two patients of group I had bony spur visible at the lumbar and dorsal spine, two had hemivertebrae in dorsal spine and 4 patients had deficient posterior arch.
Out of total 46 cases of SCM, 38 patients had type I and 8 had type II SCM. Fifteen children of group I had type I SCM, while four had type II SCM. In group II patients SCM type I was present in 23 of 27 and SCM type II in four cases.
Bar diagrams Figure6 and Figure7 show the incidence of various associated spinal and cranial lesions in both group I and II cases of SCM. Only nine patients of group II, who had split cord in dorsal region, had normal position of the conus. The other eighteen cases had their conus ending between L2 and S2. Of 19 patients of group I, nine had low-lying cord ending below L2, whereas in ten patients conus was lying at the level of L2. Hydrosyrinx was observed in seven patients of group I and in 4 of group II children. Hydrosyrinx in both the groups remained rostral to the defects and extent was 2-4 vertebral level except in one of group I patient, who had holocord syrinx. None of these patients warranted surgery for the syrinx in isolation. Neural placode was seen in all patients of group I, whereas intraspinal lipoma in 4, dermoid in 1, arachnoid cyst in 2 and thick/fatty filum seen in 5 cases. In group II other associated tethering lesions were intraspinal lipoma in 7, intraspinal dermoid in 3 patients and neurenteric cyst in only 1 patient. Neural placode was not found in any of the children of group II.
Five patients of group I had hydrocephalus, of which two were symptomatic to have enlargement of head and bulging fontanelle. These two required shunt CSF diversion before definitive surgery. Two children became symptomatic on long-term follow-up with progressive enlargement of head, proved on serial scans and required ventriculoperitoneal shunting while one other child remained asymtomatic. Three children of group II had hydrocephalus, two of them had symptomatic hydrocephalus and required shunt before definitive surgery. One patient, whose preoperative scan showed minimal ventricular dilatation, required shunt in follow-up, as he developed clinically significant increase in head size.
In group I, two children had dysgensis of corpus-callosum and another two showed Arnold Chiari malformation More Details. Aqueductal stenosis was noticed in 2 patients. In group II, none of the patients had Arnold Chiari Malformation (ACM), but aqueductal stenosis was noted in two children.
Surgical procedure and findings
Surgical Procedure
Bony spur in type I SCM was extradural, to lie between two dural tube. After laminectomy, bony spurs were identified following separation of two dural sleeves. Complete excision was done with small rongeur or micro drill. Both dural tubes were opened in their midline; usually there are fibrous adhesions with hemicords and dura on medial side. After opening dura, a thorough adhesiolysis was done and dural closure performed posteriorly in such a manner to form a common tube for both the cords. In type II SCM, the two cords were exposed and identified till the two become single rostrally and caudally. Fibrous septum between two was excised, arachnoids adhesions and anomalous nerve roots were also divided to allow free mobility of both the cords. A tight filum terminale, if diagnosed was also divided and a watertight dural closure done. Excision and repair of MMC sac were performed simultaneously in cases having meningomyelocele. At times, fascial graft or synthetic dura was used for lax dural closure. Associated tethering lesions like dermoid, lipoma were also removed simultaneously.
Surgical Findings
In group I, SCM was found one to two vertebral levels rostral in 14 cases, while it was at the level of meningomyelocele in five cases. In none of the case, SCM was caudal to MMC. While in group II, SCM was found either beneath the cutaneous marker or in close proximity, coinciding with bony anomalies of spina bifida.
Various Image findings (as described under radiological findings) were confirmed during surgical procedure. However, four of group I and one of group II patients macroscopically showed that some of the nerve roots were seen very thin and atrophic as compared to others. In five of group I and three of group II children, few nerve roots were seen significantly adherent to dura with arachnoid bands. Considering these as additional tethering lesions, these were also divided. One case of group I had intramedullary abscess without the prior evidence of infection, dermal sinus or previous surgery.
Surgical Outcome
Pre-operative sphincter dysfunction was present in 13 patients (group I = 6, group II = 7) Five patients belonging to group II showed satisfactory improvement subjectively, whereas only four children of group I had improvement in the bladder function with increase in their dry periods. Neuro-orthopedic syndrome (NOS) became stabilized in all patients of group I and group II, no further improvement or deterioration was noticed in follow-up. However, we feel that the physiotherapy and management of neuroorthopedic deformities were not adequate in most of these children.
Post operative complications
Post-operative CSF leak occurred in eight patients of group I and three of group II. Five of group I and one of group II required re-exploration and repair of dural defect with facia lata patch or synthetic graft (Gore tex), whereas 3 of group I and 2 of group II were treated conservatively with pressure dressing and acetazolamide therapy and they responded well. Pseudomeningocele was noticed in five of group I and three of group II patients. One patient of group I required thecoperitoneal shunt. Two patients required reexploration and repair, whereas in two patients only aspiration and pressure dressing were needed. In group II, one child required reexploration and repair, whereas in two patients only aspiration and pressure dressing were done. Four patients in group I and two of group II developed meningitis. All of them responded to antibiotic except one patient of group I who died due to fulminant meningitis. One child of cervical complex spina bifida expired while recovering in post-operative period due to sudden respiratory arrest. Instability of cervical spine was blamed for such catastrophe.
Discussion
SCM is not an uncommon finding in spinal dysraphism.[1],[8],[9],[10],[11] On review of our total 138 patients of spinal dysraphism SCM constituted 46 of all, reflecting an incidence of SCM to be 33 % (46/138), slightly more than reported by others, which may be on account of referral bias.[3] Nineteen children of MMC had SCM, which represent 14% of total spinal dysraphism cases. Previous pathological studies of children with MMC revealed a significant incidence of associated SCM. Campbell et al[8] found 36 cases of SCM in 100 infants of open spinal dysraphism and Emery & Lendon[9] reported 78 cases amongst 100 spinal dysraphic patients. In the present study nineteen children of MMC with unsuspected SCM could not have been diagnosed without an awareness of such coexistence. Hence, it is apparent from the above observations that the association of SCM in MMC is not an infrequent finding, which needs a special attention while screening and planning to treat these children for favourable results.
SCM is more common in females according to western literature[2],[3],[11] but male predominance is seen in Indian subcontinent, including the present study.[5],[6],[12]
The skin manifestations in occult dysraphism represent minor aberrations in the development of the surface ectoderm, brought about by the adverse influences of a dorsal endomesenchymal tract, but these changes are overshadowed by changes in the surface ectoderm in case of an associated meningomyelocele.[4],[5],[13],[14],[15] The skin markers are quite frequent in occult spinal dysraphism i.e. in SCM. In the present series, the incidence of cutaneous marker is 90% in group II cases as compared to only 10.4% in group I,[3],[4],[10],[16] signifying a low incidence of cutaneus markers in SCM presenting with meningomyelocele. It is an important observation because these cases of SCM have a higher chance of missed diagnosis without awareness. Amongst these markers the hypertrichosis was the commonest, being present in 50% of the cases of group II. The other markers like capillary hamangiomas and dermal sinus were seen in group II only and not in group I. This agrees to the consensus of incidence of cutaneous markers in occult dysraphism varying between 20 to 75%.[4],[7],[17],[18],[19],[20],[21]
Neurological deficits, as such are common in meningomyelocele, these become further dense in presence of other tethering lesions like SCM in same patients.[13],[22] Considering the high possibility of progressive and often irreversible neurological deterioration, it is logical to advocate the screening of entire spine with MRI in every spina bifida child. If SCM is found in association with MMC, it should be treated simultaneously to circumvent its deleterious effect on the developing cord.[4] While comparing both the groups, it was observed that the incidence of neurological deficit was more significant in group I as compared to group II: motor weakness (63.1% vs 40.7%), sensory loss (47.3% vs 37%) and sphincter involvement (31.5% vs 25.9%) table1. Following surgery the improvement in neurological functions were more marked in group II cases compared to group I: motor weakness (54.4% vs 33.3%), sensory functions (70% vs 66.6%) and bladder control (71.4 vs 66.6%) table2. While the incidence of neuro-orthopedic syndrome was 66.6% in group I and 96% in group II patients, this is in accordance with other reported series.[1],[4]
Children with MMC or previously operated for open defect tend to have more wound complication and CSF leakage, because of the thin skin covering the lesion.[4] Similar observations were made in the present study (42.7%).
Amongst associated craniospinal anomalies, hydrosyrinx was seen in 7 of 19 (36.8 %) cases of group I and 4 of 27 (14.8 %) of group II, which is within the reported range of 25 to 40%.[11],[15],[20] but relative incidence is higher in group I cases in this series. Amongst the other anomalies Chiari malformation (ACM) was present in 13% of cases of group I and none in group II suggesting an association, more with MMC and least with SCM. Moreover the incidence of clinically significant ACM in cases of spinal dysraphism ranges from 30 to 50% in existing literature,[14] but no comparative data is available for two groups of SCM reported by us. The overall incidence of hydrocephalus in meningomyelocele is reported as 86% in western literature;[15] in the present study, it was encountered in 5/19 (26.3%) of group I patients and 3/27 (11.11%) of group II patients, which is 17.39% of total number of cases. Our previous studies[5],[6] and other studies[12] from Indian subcontinent revealed almost similar low incidence of hydrocephalus in cases of SCM. Though the incidence of hydrocephalus is relatively more in cases of SCM with MMC, a larger study would be required to establish it.
Incidence of low lying cord in SCM with MMC is reported 30% by Iskander whereas in pure SCM it was found in 40.5% of cases as reported by Ersahin et al.[10] In our series, it was noticed in 47.3%, and 66.6% of group I and group II patients respectively. We encountered relatively higher incidence of low lying cords in both groups of cases. Intraspinal lipoma was the significant finding in the present study, both in group I, 4/19 (21.0%), and group II 7/27 (25.9%). Barring two patients in group I, in whom lipoma was rostral to the split cord, all the patients had distal lipomas in both the groups. This coincides with the reported incidence of terminal lipomas by Pang[4] who encountered 6 lipomas in total 37 cases of SCM. Other associated lesions with SCM are dermoid, epidermoid, arachnoid cyst and neurenteric cyst, as reported in literature.[6],[12],[23]
Type I SCM was encountered in 38 of 46 (82%) of our cases, while type II was seen in rest of 8 children. As an additional observation, 15 of 19 (79%) children of group I (SCM with MMC) had type I SCM signifying a higher incidence of bony spur in cases of SCM associated with meningomyelocele. It was interesting to note that 14 of 19 split cord malformations were located one or two vertebral levels rostral to myelomeningoceles and remaining 5 were found at the level of meningomyeloceles itself. It is obvious that no SCM was found below the level of MMC. The type I SCM is reported to have higher incidence of 75% in comparison 25 of type II in literature.[10],[11],[12] A higher incidence of type I SCM is reported in cases associated with MMC by Pang et al[10] and Ersahin et al.[10]
Occasionally, the shortened nerve roots, conjoined nerve roots, angulated nerve roots and arachnoid bands are seen at the level of dysraphic state, causing tethering of cord. [24],[25] Lassman and James have devised the term 'meningocele manque' for the cases in whom adhesion of the conus, filum, cauda equina roots to the inner aspect of the posterior dura occur.[26] The present series also witnessed the thin and aberrant nerve roots in 4 of goup I and1 of group II patients, whereas nerve roots were adherent to dura with arachnoid bands in 5 of group I and 3 of group II cases. The presence of such structures, most probably signify the additional tethering elements, need to be catered during surgery.
Conclusion
The incidence of SCM with meningomyelocele being 41% of total SCM cases is obviously significance. The prognosis of SCM with MMC is not as good as noted in pure SCM cases. Progressive sensorimotor deficit also has higher incidence in cases of SCM with MMC, whereas the incidence of NOS is relatively more in pure SCM cases. A screening of entire neural axis is seemingly indispensable, in order to cater to the occult associations simultaneously (in cases of overt spina bifida) for better outcome. Because of significant occurrence of MMC in SCM and concurrent hydrocephalus, it seems appropriate to investigate these children thoroughly by cranio-spinal MRI. In case of equivocal radiological findings, SCM should be searched one to two level higher than MMC, if not present at same level in view of its higher incidence of rostral location.
Acknowledgement
The authors are extremely grateful to Mr. A.P. Dhar Dwivedi for preparation of this manuscript.
References
1. Andar UB, Harkness WFJ and Hayward RD: Split cord malformation of the Lumbar region. Pediatr Neurosurg 1997;26: 17-24.
2. Anderson FM: Occult spinal dysraphism: A series of 73 cases. Pediatrics 1975; 55: 826 - 835.
3. Pang. D, Dias MS & Ahab-barmada M: split cord malformation: Part I; unified theory of embryogenesis for double spinal cord malformation, Neurosurg 1992 ; 31: 3, 451-480.
4. Pang. D; split cord malformation: Part II: clinical syndrome, Neurosurg 1992; 31: 3, 481-500.
5. Kumar R, Bansal KK, and Chhabra DK: Occurrence of Split Cord Malformations in Meningomyelocele: Complex Spina Bifida. Pediatr Neurosurg 2002 ; 36: 119 - 127.
6. Kumar R, Bansal KK, and Chhabra DK: Split cord malformation in pediatric patients: out come of 19 cases. Neurology India 2001; 49: 128-133.
7. Eid K, Hochberg J, Saunders DE. Skin abnormalities of the back in diastematomyelia. Plast Reconstr Surg 1979; 63: 534- 549.
8. Campbell LR, Dayton D and Sohal GS. Neural tube defects: A review of human and animal studies on the etiology of neural tube defects. Teratology 1986; 34: 171-187.
9. Emery JL and Lendon RG. The local cord lesion in neurospinal dysraphism (Meningomyelocele). J Path 1973; 110: 83-96.
10. Ersahin Y, Mutluer S, Kocaman S, and Demirtas E: Split cord malformations in children. J Neurosurg 1998 ; 88: 57 - 65.
11. Iskander BJ, Mclaughlin C and Oakes WJ. Split cord malformation in myelomeningocele patients. Br J Neursurg 2000 ; 14(3): 200-203.
12. Jindal A, Mahapatra AK. Split cord malformation - a clinical study of 48 cases. Indian Pediatrics 2000 ; 37: 603 - 607.
13. Miller A, Guille JT, Brown Jr: Evaluation and treatment of diastematomyelia. J Bone Joint Surg (Am) 1993; 75: 1308-1317.
14. Reigel DH, Rotenstien D. Spina bifida. In pediatric neurosurgery by section of pediatric neurosurgery of American association of neurological surgeon, 3rd edition, USA; WB Saunders, 1994; 51-76.
15. Rokos J. Pathogenesis of diastematomyelia and spina bifida. J Path 1975; 117: 155-161.
16. Ersahin Y, Demirtas E, Mutluer S, Tosun A & Saydam: Split cord malformation: Report of three unusual cases. Pediatr Neurosurg 1996; 24: 155-159.
17. Dale AJD. Diastematomyelia. Arch Neurol 1969; 20: 309 - 317.
18. Guthkelch AN. Diastematomyelia with median septum. Brain 1974; 97: 729-742.
19. Hood RW, Riseborough EJ, Nehme AM, Micheli LJ, Strand RD, Neuhauser EB: Diastematomyelia and structural spinal deformities. J Bone Joint Surg 1980; 62 A: 520-528.
20. Humphreys RP. Spinal dysraphism. In Rengachary SS, Wilkins RH, eds. Neurosurgery, 2nd edn. Mc Graw - Hill, 1996.
21. James CCM, Lassman LP. Spinal dysraphism. The diagnosis and treatment of progressive lesions in spina bifida occulta. J Bone Joint Surg 1962; 44B: 828-840.
22. Meacham WF. Surgical treatment of diastematomyelia. J Neurosurg 1967; 27: 78-85.
23. Kumar R, Jain R, Rao KM, Nuzhat H. Intraspinal neurenteric cysts. Child Nerv Syst 2001; 17: 584-588.
24. McLone DG. Technique for closure of meningomyelocele. Childs Brain 1980; 6: 65-73.
25. Guthkelch AN, Pang D, Varies JK. Influence of closure technique on results in myelomeningocele. Childs Brain 1981;8: 350-355.
26. Lassman LP, James CCM. Meningocele manque. Child Nerv Syst 1977; 3: 1.(Kumar Raj, Singh SN, Bans)