Clinical characteristics and surgical outcomes of split cord malformations: A retrospective analysis

Abstract

Background:

Split cord malformation (SCM) is a rare congenital anomaly characterized by a longitudinal division of the spinal cord into two hemicords. Clinical manifestations are diverse, often including neurological, orthopedic, and cutaneous abnormalities. Surgical management remains the mainstay of treatment, though outcomes and complication profiles require further study. This study aimed to analyze the clinical spectrum, surgical management, and outcomes of pediatric and young adult patients with SCM.

Methods:

A retrospective review was conducted of patients with SCM who underwent surgery at our tertiary care institute between 2018 and 2024. From 168 identified cases, 35 patients with complete clinical, radiological, and surgical data were included. Demographics, clinical features, radiological findings, operative details, and postoperative outcomes were recorded. Descriptive statistics were generated using GNU PSPP 2.0.0.

Results:

Among 35 patients (18 males, 17 females; mean age 5.4 years, range 2 months–20 years), SCM Type II was more common (62.9%) than Type I (37.1%). Motor deficits (77.1%) were the predominant presentation, followed by bowel/bladder involvement (51.4%). Cutaneous markers were observed in 45.7% and skeletal anomalies in 65.7%. Radiological findings included tethered cord (60%), low-lying cord (22.9%), and lipomyelomeningocele (17.1%). All patients underwent laminectomy with septum excision and dural repair; intraoperative neuromonitoring was used in 40%. Postoperative complications were minimal (cerebrospinal fluid leaks 8.6%, wound dehiscence 2.9%, and re-exploration 5.7%). Neurological outcomes were favorable, with 59.3% improving in motor deficits and no patient experiencing deterioration.

Conclusion:

SCM is a complex congenital anomaly often associated with cutaneous and skeletal markers. Surgical treatment is safe and effective, with most patients achieving neurological improvement or stabilization and a low complication rate. Early recognition and comprehensive surgical management are crucial to preventing long-term morbidity.

INTRODUCTION

Split cord malformation (SCM) is a rare congenital anomaly featuring a longitudinal cleft in the spinal cord, resulting in two distinct hemicords.[1] The condition can occur as an isolated entity or in conjunction with other congenital defects.[17] SCM diagnosis and surgical outcomes continue to be debated, as symptoms are nonspecific and often linked to other anomalies. SCM is a rare congenital anomaly characterized by a longitudinal cleft in the spinal cord, resulting in two distinct hemicords. There are two recognized types of SCM: Type I, where a bony spur separates the two hemicords, each encased in its own dural sac; and Type II, where the hemicords are not separated by a cartilaginous spur but are contained within a single dural sac. Clinical features of SCM can vary widely depending on the location and severity of the malformation, as well as the presence of associated anomalies. Common presentations include cutaneous markers such as hypertrichosis, dimples, or hemangiomas over the affected area, as well as neurological deficits such as weakness, sensory changes, or bowel and bladder dysfunction.[10,19] Scoliosis and leg length discrepancies may also be observed. We aim to present a retrospective analysis of patients with SCM, focusing on clinical presentations, management strategies, and surgical outcomes.

MATERIALS AND METHODS

Patient eligibility and study setting

We retrospectively searched our hospital database for patients with SCM operated between 2018 and 2024. The following keywords were used to retrieve patient data: congenital spinal malformations, diastematomyelia, diplomyelia, and SCM.

Patients were included if they met all of the following conditions:

1. Confirmed diagnosis of SCM Type I or Type II based on magnetic resonance imaging (MRI) and/or computed tomography (CT) findings showing a longitudinal division of the spinal cord

2. Underwent surgical management for SCM at King George’s Medical University between 2018 and 2024

3. Had complete clinical, radiological, and intraoperative data available, including demographic details, clinical presentation, surgical technique, and postoperative follow-up information.

Patients were excluded if they:

1. Had inaccessible or poor-quality radiological studies that precluded confirmation of SCM type or level

2. Had insufficient clinical data, defined as missing key information on neurological status, surgical details, or follow-up outcomes

3. Were diagnosed intraoperatively or radiologically with other spinal dysraphisms or anomalies (e.g., myelomeningocele, lipomyelomeningocele without SCM, tethered cord without split cord, dermoid cysts, or isolated syringomyelia) without evidence of a true split cord

4. Did not undergo surgery for SCM during the study period.

The final diagnosis of SCM was confirmed based on radiological imaging (MRI and CT) reviewed independently by two neurosurgeons. Type I and Type II distinctions were made according to the Pang classification, depending on the presence or absence of a bony/cartilaginous spur and the number of dural sacs.

Out of the 168 cases that the search produced, we examined and eliminated 133 cases that did not meet the inclusion criteria. Patient demographics, comorbidities, clinical aspects (neurologic, orthopedic, and cutaneous symptoms), radiologic and surgical findings, and immediate and long-term outcomes were thus examined retrospectively for 35 patients with complete data sets.

Statistical analysis

Data were analyzed using GNU PSPP version 2.0.0. Continuous variables (e.g., age, postoperative McCormick scores) were expressed as mean ± standard deviation, and categorical variables as frequencies and percentages. Comparative analyses between SCM Type I and II, and between folic-acid intake and non-intake groups, were performed using the independent-samples t-test for continuous variables and the Chi-square test or Fisher’s exact test for categorical variables. p < 0.05 was considered statistically significant.

Surgery

All patients underwent preoperative CT and MRI scans to visualize the bony anatomy and neural elements. Postoperative MRI scans were obtained in some patients to assess the adequacy of the surgical decompression and identify any potential complications. All patients underwent laminectomy for septum removal and dura repair, which are the standard surgical procedures for SCM. After the dura was stripped from the bony spur, the spur was removed using surgical instruments such as drills or nibblers. The medial side of the dura of the hemichords was opened and cut to release the tethering. Connective formations such as fibrous band and dentate ligament were sectioned to further untether the spinal cord. Associated intraspinal pathologies such as dermoid and lipoma were also removed as appropriate to prevent future complications. The filum was cut distally in cases with low lying cord to release the tethering. Other associated anomalies such as myelomeningocele and lipomyelomeningocele were addressed as appropriate to correct all the congenital defects. A waterproof dura closure was performed to prevent cerebrospinal fluid (CSF) leakage.

RESULTS

A total of 35 patients with SCM were analyzed, with ages ranging from 2 months to 20 years. The cohort included 18 males (51.4%) and 17 females (48.6%). Only 11 patients (31.4%) had a history of folic acid intake during pregnancy, while 24 (68.6%) did not, highlighting a potential preventive gap. The difference in postoperative McCormick scores between folic-acid groups (2.9 ± 0.4 vs. 2.6 ± 0.3) was not statistically significant (p = 0.21). Similarly, no significant differences were found between SCM Type I and II in postoperative outcomes (p > 0.05). SCM Type I was identified in 13 patients (37.1%), predominantly in the lumbar region (12), while SCM Type II was seen in 22 patients (62.9%), with lumbar (14), dorsolumbar (6), dorsal (1), and cervical (1) involvement. Clinically, motor deficits were the most common presenting symptom, seen in 77.1%, followed by bowel and bladder involvement in 51.4%, back or limb pain in 45.7%. Associated skin anomalies included dermal sinus in 6 (17.1%), hypertrichosis in 4 (11.4%), skin dimples in 3 (8.6%), hyperpigmented patches in 4 (11.4%), and subcutaneous lipomas in 9 (25.7%). Skeletal anomalies were also common: Congenital talipes equinovarus (CTEV) in 17.1%, scoliosis in 14.3%, kyphosis in 5.7%, congenital dislocation of the hip (CDH) in 5.7%, along with flat foot (2.9%), limb thinning (11.4%), limb length discrepancy (8.6%), and spina bifida aperta (14.3%). Radiologically, tethered cord was observed in 21 patients (60%), low-lying cord in 8 (22.9%), lipomyelomeningocele in 6 (17.1%), meningomyelocele in 4 (11.4%), and syrinx/Chiari in 2 (5.7%) [Tables 1 and 2].

Table 1:

Demography and clinical presentation.

Table 2:

Associated anomalies/features.

Intra operatively, a drill was used in 18 patients (51.4%), nibbler in 17 (48.6%), and intraoperative neuromonitoring (IONM) in 14 (40%). Postoperative complications were limited: CSF leaks occurred in 3 patients (8.6%), wound dehiscence in 1 (2.9%), and re-exploration was needed in 2 (5.7%) [Table 3].

Table 3:

Intra operative adjunct and complications.

Outcomes were encouraging: among those with motor deficits (n = 27), 16 (59.3%) improved and 11 (40.7%) remained static. For sensory deficits (n = 9), 6 (66.7%) improved and 3 (33.3%) were static. Among those with autonomic symptoms (n = 8), 2 (25%) improved while 6 (75%) remained static. Of the 18 patients (51.4%) with urinary symptoms, 8 (44.4%) improved and 10 (55.6%) remained static. Notably, no patient showed worsening in any of the outcome domains [Table 4].

Table 4:

Clinical Outcome.

Follow-up duration and long-term outcomes

The mean duration of follow-up in our cohort was 18.6 months (range: 6–48 months). Long-term follow-up data were available for 27 of the 35 patients (77.1%). During this period, no cases of retethering or progressive spinal deformity were documented. Two patients (5.7%) required re-exploration in the early postoperative period due to CSF leakage, but no delayed reoperations were needed. Mild residual motor weakness persisted in some patients with severe preoperative deficits, but no late neurological deterioration was observed in any case.

DISCUSSION

Our retrospective analysis of 35 SCM patients provides valuable insights into the presentation, management, and outcomes of this rare congenital anomaly. The age range of our cohort, from 2 months to 20 years, with a mean of 5.43 years, consistent with the understanding that SCM often presents early in life. Other studies have also reported a wide age range at presentation, reflecting the variability in symptom severity and the timing of diagnosis. The mean age of patients at presentation in another study was 7.3 years.[15]

The slightly higher incidence of SCM Type II (22 patients) compared to Type I (13 patients) in our series differs from some reports in the literature, where Type I SCM is more frequently observed. However, it is important to note that the relative prevalence of each type can vary depending on the referral patterns and diagnostic criteria used in different studies. However, similar demography was observed with Type I SCM in 156 patients, while Type II was found in 98 in a paper by Mahapatra and Gupta.[15] Several factors may explain this discrepancy. First, our institution functions as a tertiary neurosurgical referral center catering to a large pediatric population in Northern India. As such, patients referred to our unit typically represent symptomatic and surgically indicated cases, particularly those presenting early in life. Type II SCMs, characterized by a single dural sac and fibrous septum, are known to cause earlier neurological symptoms due to tethering, prompting earlier referral and intervention. In contrast, Type I lesions, often associated with bony spurs and dual dural sacs, may remain asymptomatic for longer or may have been surgically managed at other centers before referral. This referral and selection bias may partly account for the higher prevalence of Type II SCM observed in our cohort. Second, the improved availability of high-resolution MRI after 2020 may have enhanced the detection of subtle Type II lesions that previously went unrecognized, especially those without an osseous spur. Third, variability in classification criteria and interobserver interpretation between mixed or transitional forms of SCM could influence categorization; some borderline lesions might have been designated as Type II based on single dural sac anatomy rather than septal composition. Finally, regional embryologic or environmental influences, including differences in maternal nutrition and folic acid intake, may subtly affect the prevalence or expression of dysraphic anomalies, although this remains speculative. Taken together, these findings suggest that the epidemiologic spectrum of SCM may vary regionally and temporally, reflecting evolving diagnostic capabilities and referral biases. By highlighting a predominance of Type II SCM in our series, this study underscores the need for standardized diagnostic imaging criteria and multicenter data to more accurately define the true distribution of SCM subtypes across populations.

The lumbar region was the most common location for both SCM types in our cohort, which is in line with other published series. In a study by Borkar et al., the thoracic region was the most common site of split, followed by the lumbar region, with some patients exhibiting splits at multiple levels[5] [Figure 1].

Figure 1:

(a) Axial computed tomography (CT) scan demonstrating a bony spur characteristic of split cord malformation Type I. (b) Postoperative CT images following laminectomy and excision of the bony spur, showing adequate decompression of the spinal canal.

Asymmetric weakness of lower limbs is a common neurological deficit, and sphincter disturbances have also been described. Occult urological abnormalities can be detected through urodynamic studies, even in patients without apparent urological symptoms, and these abnormalities may stabilize or improve after surgical intervention.[16,18] In a study by Erşahin, unilateral leg paresis and atrophy, and bladder and/or bowel incontinence were observed in a subset of patients, with the proportion of patients with neurological or orthopedic deficits not significantly varying with the type and level of SCM.[10] In our study, there was prominence of motor deficits (27 patients), followed by bowel and bladder involvement, and back or limb pain.

Our study found numerous skin anomalies which included dermal sinus in 17.1%, hypertrichosis in 11.4%, followed by skin dimples, hyperpigmented patches and subcutaneous lipomas. Skeletal anomalies were also common which included CTEV, scoliosis, kyphosis, CDH, flat foot, limb thinning, and discrepancy. Hypertrichosis, naevi, dimples, dermal sinus, and lipomatous swellings are commonly observed in SCM patients. Often, spinal deformity correction surgery uncovers a cord anomaly in an otherwise asymptomatic patient. Associated anomalies, such as syringomyelia, lipomeningomyeloceles, and congenital tumors, can significantly influence neurological outcomes. The presence of cutaneous markers such as dermal sinus, hypertrichosis, skin dimples, hyperpigmented patches, and subcutaneous lipomas further supports the established association between SCM and these external signs. Orthopedic manifestations such as scoliosis, kyphosis, congenital talipes equinovarus, congenital hip dislocation, rib anomalies, limb thinning/shortening, and flat foot have also been observed in SCM patients in numerous published literatures.[14,15] These orthopedic manifestations can be attributed to the altered biomechanics and neurological imbalances caused by the spinal cord malformation. The observation of tethered cord, low-lying cord, lipomyelomeningocele, meningomyelocele, and syrinx/Chiari on radiological imaging highlights the complex nature of SCM and its frequent association with other spinal cord abnormalities. These findings emphasize the need for careful preoperative imaging to delineate the anatomy and identify any coexisting conditions that may impact surgical planning.

Intraoperatively, the use of a drill, nibbler, and IONM reflects the surgical techniques employed for septum removal and dura repair. IONM was utilized in 14 patients (40%) in this cohort. The selective use of IONM was primarily due to its recent availability at our institute after 2022. Consequently, all patients operated on after 2023 underwent surgery with IONM, while those operated before this period did not have access to it. IONM modalities included motor-evoked potentials and somatosensory-evoked potentials to monitor functional integrity of the spinal cord during spur excision and cord detethering. Although not statistically analyzed due to the small sample size, patients operated with IONM showed a trend toward fewer postoperative neurological deficits and complications, with no cases of neurological worsening observed. The introduction of IONM thus provided an added layer of intraoperative safety, particularly in complex cases with severe adhesions or distorted anatomy, and has since become a standard adjunct in all SCM surgeries at our center [Figure 2].

Figure 2:

(a) Axial magnetic resonance imaging demonstrating the classic u201cowlu2019s eyeu201d appearance consistent with split cord malformation. (b) Axial computed tomography scan of the same patient showing a bony spur characteristic of split cord malformation type I.

The limited postoperative complications, including CSF leaks, wound dehiscence, and re-exploration, suggest that surgical intervention for SCM can be performed with an acceptable safety profile.

Postoperative complications in our series were uncommon and generally manageable, but a thorough description of their management and clinical impact is important for reproducibility and patient counseling. In total, 3 patients (8.6%) developed postoperative CSF leak and 1 patient (2.9%) experienced wound dehiscence. All CSF leaks were identified within the first 7 days postoperatively by persistent drainage from the wound or the development of a small subcutaneous collection confirmed on inspection. Initial management prioritized conservative measures: Reinforced multilayer wound closure at bedside when appropriate, application of pressure dressings, short periods of prone position bed rest with head-end elevation avoided to reduce intracranial hypotension, and close wound monitoring. One case resolved with conservative treatment and was discharged on acetazolamide for 14 days. In two patients, the leak responded to bedside wound reinforcement and a short course of controlled lumbar subarachnoid drainage (continuous lumbar drain at 5–10 mL/h for 5 days) together with prophylactic broad-spectrum antibiotics while cultures were pending; both healed without further surgery. No patient developed meningitis, pneumocephalus, or subdural hematoma, and there were no long-term neurological sequelae attributable to the CSF leaks. These management steps are consistent with established algorithms emphasizing initial conservative management with lumbar drainage and re-exploration only for persistent or large leaks or when infection is suspected.[2,11]

The single case of wound dehiscence was managed with prompt local debridement, resuturing with layered closure, and adjunctive wound care; after which the wound healed. Clinically, these complications had limited impact on long-term outcomes in our cohort. Two patients required early re-exploration (one for persistent CSF leak and one for residual spur remnant), both recovered well without further neurological deterioration. Median additional hospital stay attributable to these complications was 3–7 days (range 2–12 days), with the longest stay occurring in the patient who underwent re-exploration; no patient required intensive care solely for complication management. These short extensions in length of stay and the absence of major infective sequelae suggest that, with early recognition and institution of standardized management pathways, most post-SCM complications can be resolved with limited morbidity, a finding echoed in other SCM series reporting low permanent morbidity despite modest early complication rates.[3]

Preventive measures that likely contributed to our relatively low and non-catastrophic complication profile included meticulous multilayer dural closure under the operating microscope, layered subcutaneous closure to obliterate dead space, and early postoperative wound checks with low thresholds for drain placement or radiological assessment. Recent systematic reviews and institutional guidelines emphasize that meticulous dural repair, selective use of lumbar drainage, early bedside wound revision for minor leaks, and prompt surgical re-exploration for persistent leaks or infected wounds yield the best balance between morbidity and resource utilization.[13]

Finally, we acknowledge that the management of CSF leaks and wound complications carries risks (e.g., lumbar drain-related headaches, nerve root irritation, or drain-associated infections) and must be individualized; in our series, the benefits of short controlled drainage and targeted re-exploration outweighed these risks. Prospective standardization of postoperative protocols (timing of mobilization, use of drains, and wound surveillance schedules) and longer follow-up would help determine whether these interventions also reduce late complications such as retethering or late wound breakdown.

Complications can significantly affect the patient’s quality of life, length of stay, and the likelihood of hospital readmissions.[6] Awareness of potential complications can improve surgical planning and patient counseling. The encouraging outcomes observed in our study, with improvement or stabilization of motor, sensory, and autonomic symptoms, are consistent with the goals of surgical management in SCM. The fact that no patient experienced worsening in any of the outcome domains is particularly noteworthy. While the natural history of untreated SCM can involve progressive neurological deterioration, surgical intervention aims to halt or reverse this process by releasing the tethering effect of the septum and associated abnormalities.

We observed that only 11 of 35 mothers (31.4%) reported periconceptional folic acid intake. This low proportion is notable because folic acid supplementation during the periconceptional period is a well-established preventive measure against neural tube defects (NTDs), including myelomeningocele and spina bifida, with multiple studies demonstrating up to a 70% reduction in NTD risk following adequate supplementation or food fortification programs.[4,8,9] Although direct evidence linking folic acid deficiency to SCM development is limited, some experimental and epidemiological studies have suggested that folate deficiency during early embryogenesis can impair neural tube closure and secondary canalization, processes implicated in SCM pathogenesis.[7,12] In our cohort, patients with maternal folic acid intake had slightly better postoperative McCormick scores (2.9 vs. 2.6), but this difference did not reach statistical significance. Given the small sample size and potential recall bias regarding maternal supplementation, this finding should be interpreted cautiously. Nonetheless, the observed trend highlights the importance of strengthening public health efforts toward routine folic acid supplementation and prenatal counseling. Further multicenter, prospective studies assessing the timing, dosage, and biochemical confirmation of folate status are warranted to clarify whether folic acid deficiency contributes to SCM incidence or postoperative outcomes.

Long-term follow-up considerations

Long-term follow-up is crucial in SCM patients, as retethering, progressive scoliosis, or late neurological decline may occur even years after apparently successful surgery.[10,16] In our series, the mean follow-up period was 18.6 months, during which no delayed neurological deterioration or retethering was observed. However, the relatively short duration and incomplete follow-up limit our ability to draw conclusions regarding late outcomes. Previous studies have reported retethering rates of 5–15% and late deformity progression in a subset of patients, particularly those with complex Type I anomalies or extensive adhesions.[15,18] Hence, lifelong surveillance with clinical and radiological assessment remains advisable for all operated patients.

CONCLUSION

SCM is a complex congenital anomaly that requires a multidisciplinary approach for diagnosis and management, and surgical intervention can lead to favorable outcomes, particularly when combined with meticulous surgical techniques and comprehensive pre- and postoperative care. Cutaneous markers can aid in the diagnosis of occult spinal dysraphism, and the entire neuraxis should be evaluated to treat any tethering lesions during the initial surgery. Further research is needed to explore the potential role of folic acid in preventing SCM and to identify other factors that may influence its development. Consideration should be given to the benefits of early diagnosis and intervention to prevent irreversible neurological damage. Once diagnosed, all SCMs should be treated surgically, even if patients are asymptomatic and neurologically intact, to prevent future neurological deterioration. Postoperative complications are usually minimal, and overall outcomes are often favorable, including stabilization, symptomatic improvement, and resolution of hydronephrosis. Long-term follow-up is essential to monitor for potential complications such as retethering and spinal deformities, which may necessitate additional interventions. Continued research into risk factors and a better understanding of correction goals may aid in prevention and management at all stages of care.

Data availability:

The data used and analyzed in this study are available from the corresponding author on reasonable request.

Ethical approval:

The Institutional Review Board approval is not required. This was a retrospective analysis of previously collected clinical and radiological records, with no prospective intervention or alteration in patient management. In accordance with institutional policy, formal Institutional Ethics Committee approval was not required for this retrospective analysis. Patient confidentiality was maintained, and all data were anonymized before analysis.

Declaration of patient consent:

The authors certify that they have obtained all appropriate patient consent forms. In the forms, the patients/parents/legal guardians have given consent for clinical information and images to be reported in the journal. The patients understand that their names and initials will not be published and due efforts will be made to conceal their identity.

Financial support and sponsorship:

Nil.

Conflicts of interest:

There are no conflicts of interest.

Use of artificial intelligence (AI)-assisted technology for manuscript preparation:

The authors confirm that there was no use of artificial intelligence (AI)-assisted technology for assisting in the writing or editing of the manuscript and no images were manipulated using AI.

Disclaimer

The views and opinions expressed in this article are those of the authors and do not necessarily reflect the official policy or position of the Journal or its management. The information contained in this article should not be considered to be medical advice; patients should consult their own physicians for advice as to their specific medical needs.