This retrospective study aims to compare the intraoperative and postoperative parameters of patients operated with open and MIS technique for IDEM tumors. The patients were operated between September 2012 to September 2019. This study was approved by the Institutional Review Board, and written informed consent for participation and publication was obtained from all patients. A minimum follow-up of 3 years postsurgery are done in all patients, though some patients are in a longer follow-up. Patients lost to minimum follow-up of 3 years have been excluded from the study. Demographic date recorded include, name, age, sex, symptom duration, magnetic resonance imaging (MRI) images showing location type and extent of tumor, type of surgery MIS or open, intraoperative parameters like blood loss, duration of surgery, complications like CSF leak, infection or neurological deterioration and completeness of tumor resection from intraoperative observation and immediate postoperative scans. Tumor pathology was also noted. Postoperative duration of stay on hospital, return to normal activities, return to work, and postoperative pain visual analogue scale (VAS) were noted and documented on 3 months, 1 year, and 3 years.

MIS is suitable for IDEM lesions extending up to 2 vertebral body levels using expandable tubular retractors. For tumor spanning more than 3 levels, MIS and expandable retractors are not recommended in view of inadequacy of exposure. Hence open surgery is preferred. MIS can address ventrolateral or dorso-lateral lesions via lateral durotomy and allows for en bloc resection of schwannomas, neurofibromas, or piecemeal removal of meningiomas. If the tumor is placed completely ventral or dorsal then an open surgery was preferred by the surgeon. However, complete ventral or dorsal lesions are very rare. Open surgery was implemented in all lesions including those larger than 3 vertebral body levels or other reasons like cost of surgery, patients wish or factors against MIS approach. Dumbell lesions were treated by either method.

(1) Diagnosis: patients diagnosed with IDEM spinal tumor confirmed by imaging (MRI) and clinical evaluation

(2) Type of tumor: neurofibroma, schwannomas, meningiomas or mimics

(3) Location: cervical, thoracic and lumbar

(4) Levels: 3 vertebral body levels above or below

(5) Type: IDEM and dumbbell lesions extending to the foramina or extra foraminal region

(6) Age range: adults aged 18 years and older

(7) Tumor size: tumors suitable for removal through minimally invasive or traditional open surgical techniques

(9) Symptoms: patients with symptoms warranting surgical intervention, such as progressive pain, paraesthesia, neurological deficits, or spinal cord compression causing myelopathy.

(10) Minimum 3 years with visits at 3 months, 1 year, and 2 years with MRI follow-up immediately after surgery, 1, and 3 years.

(1) Prior spinal surgery: patients with a history of prior spinal surgeries or spinal fusion affecting the tumor site

(2) High-risk comorbidities: patients with severe cardiovascular, pulmonary, or other systemic conditions that contraindicate prolonged surgery

(3) Severe neurological deficits (American Spinal Injury Association Impairment scale - ASIA A or B) open surgery was preferred in this situation.

(4) Those lost to follow up or were not able to do MRI scans at 1 and 3 years.

The procedure begins with the patient in a prone position under general anesthesia and neuromonitoring. The target spinal level is confirmed using intraoperative x-ray imaging. After confirming the site, a vertical incision approximately 1.5–2 cm from the midline is made, based on pre operative MRI findings that determine the tumor’s extent and the appropriate retractor type fixed or expandable. The side of approach is based on the predominance of the tumor and direct approach to the tumor. In case it’s a midline tumor, then the most symptomatic side was chosen.

The subcutaneous tissue is dissected to expose the thoracolumbar fascia. Serial dilators are docked over the hemilamina to gently separate muscles from the underlying lamina, guided by lateral fluoroscopy to ensure precision. The final position is confirmed with anteroposterior and lateral images before placing the tubular retractor over the largest dilator. The retractor is angulated as needed and secured with a table-mounted holder.

An operating microscope which enhances visualization, providing the necessary depth perception for accurate dissection is used in all cases. Hemilaminectomy is performed using a high-speed drill and Kerrison punch, preserving critical structures like the spinous, interspinous, and supraspinous ligaments. The tube is angulated in such a way that the base of the spinous process is drilled to reach the midline. Simultaneously, the table is also tilted away from the surgeons to see the midline on the opposite side. After this step, the tube can be adjusted to the unilateral side or as required. Once the bony decompression is done, the ligamentum flavum is excised, and minimal medial facetectomy is performed to fully expose the lateral edge of the dura. Bayoneted instruments prevent visual obstruction, maintaining a clear surgical field.

The dura mater is opened longitudinally away from the midline in an lateral position with a No. 11 scalpel blade, and 6-0 polypropylene tack sutures hold it open. The arachnoid membrane is carefully incised to reveal the spinal cord, nerve roots, and tumor. The surgical approach, whether en bloc resection or piecemeal removal depends on the tumor’s characteristics, including size, dural attachment, and origin. The full extent of the tumor must be visualized from cranial to caudal before micro resection of tumor is started. An ultrasonic aspirator may be used to fragment and aspirate the tumor while minimizing damage to surrounding tissues. Meticulous dural closure is critical to prevent CSF leaks. Continuous or interrupted sutures are employed using fine bayoneted toothed forceps and needle holders. We prefer continuous sutures for watertight closure. Fibrin sealant is applied to reinforce the repair only if there is a suspicion of CSF leak from an inadequate closure. Valsalva mauver is used at the end of the procedure to check the adequacy of the closure. The tubular retractor is carefully removed, muscles are reapproximated, and fascia is closed followed by subcutaneous tissue with 3-0 vicryl sutures. The skin is closed using continuous nonabsorbable sutures to ensure a watertight seal.

In open surgery the spine is exposed midline to expose the tumor to its full extend cranial and caudally. The muscles are stripped on both sides to expose the spine process and the lamina. The spinous process is removed along with the attachments and a complete laminectomy is done to expose the dura. The facet joints are not disturbed, but sometimes medial facet removal may be required, depending on the size of the tumor. Dumbbell lesions may necessitate unilateral facet removal in order to expose the tumor in the foraminal region. Following this epidural hemostasis is achieved, and the dura is opened in the midline to expose the tumor. The full extent of the tumor is assessed and depending on whether it is a schwannoma, neurofibroma, or meningioma, the tumor is excised accordingly. If it is a meningioma, then the dura is treated with diathermy and the arachnoid is removed completely. Meticulous closure of the dura is achieved with continuous nonabsorbable 6-0 proline. The fascia is closed with 2.0 vicryl and subcutaneous closure is done with 3-0 vicryl. The skin is closed with nonabsorbable continuous 3-0 ethilon.

Minimally invasive spine surgery (MISS) technique can be effectively employed in selected cases of IDEM tumors. MISS is ideally suited for tumors that: (1) involve 1–2 vertebral levels with well-demarcated margins; (2) are located laterally or dorsolaterally (such as schwannomas or meningiomas with lateral dural attachment); (3) are noncalcified, soft-tissue tumors amenable to piecemeal or en bloc resection; (4) have minimal extradural extension; and (5) are in patients with good general condition, suitable for longer operative duration under general anesthesia.

On the other hand, MISS is contraindicated or less favorable in: (1) tumors involving more than 3 vertebral levels; (2) midline ventral tumors with broad dural attachment, limiting safe resection via narrow corridors; (3) calcified or highly adherent meningiomas with complex vascular involvement; (4) very large dumbbell tumors requiring extensive foraminal or extraforaminal dissection and stabilization, and; (5) patients with severe neurological deficits (ASIA A/B) where rapid and wide decompression is required.

Hence, careful preoperative assessment of tumor morphology, location, extension, and patient factors is essential to determine suitability for MISS. The surgeon’s experience and availability of advanced instrumentation also play a key role in decision-making.

Meningiomas arise from arachnoid cap cells and usually present from around the nerve root sleeve and laterally which makes it convenient for MIS approach from unilateral side. Those arising from the ventral or dorsal midline have their origin from dural fibroblast or the pia. These can be a relative contraindication for MIS approach however they are seldom rare. When there is dural tail sign which may be seen in 30%–50 % of cases the exposure must include up to the end of the tail to remove the tumor fully. The most common types of meningiomas are meningothelial and psammomatous variety. Gross total excision may sometimes not be possible because of the strong adherence with the dura, calcification or ventral attachment not accessible to remove completely. In meningiomas, GTR involves excising the tumor and coagulating the dural attachment. If part of the tumor is adherent to nerve roots or the spinal cord, a near-total resection may be performed to avoid neurological deficits. Aggressive dural resection and reconstruction is not recommended as the recurrence rates are no different from just a dural coagulation [16-18] and higher complications. The spinal meningiomas are less aggressive than the cranial and hence long-term follow-up of recurrence rate is reported to be less than 15% [16,17,19,20]. Radiotherapy is reserved only for the recurrent tumors that cannot be resected and malignant tumors which are extremely rare.

These tumors arise from the Schwann cells of the posterior roots and hence they tend to be in the posterolateral region. Rarely they may extend extradural or seen intramedullary [18,21]. Small and medium sized tumors can be removed completely as they are well capsulated and tend to arise from a single fascicle and usually the sensory root [18,22]. If the tumor is large an intracapsular debulking is done and then the fascicle and tumor are disconnected and removed fully. If a single fascicle is not identified then the fleshy part of the tumor is removed completely and capsule if left behind. Aggressive removal of the tumors versus the risk of neurological damage must be assessed in large tumors and in tumors where multiple fascicles are attached. By intracapsular debulking the patient is relieved of his symptoms and the peripheral nerve function are also intact. These incomplete tumors removals remain dormant and asymptomatic for many years [18]. However, in the setting of neurofibromatosis (NF) usually type 2 the recurrence rates are 39% at 5 years and 6%–12% in those less than 5 years [21-23]. The occurrence rate of postoperative neuropathic pain is as high as 7.2 %. In intradural tumor removal, especially in nerve sheath tumor there is loss of Schwann cell guidance leading to random sprouting of neurons causing the neuropathic pain [24]; this is also seen with other intradural tumors like meningioma and ependymoma [25]. There is an underestimation in assessing the long-term disabilities like neuropathic pain in tumor patients who are likely in need of pain medication [9]. This can be managed by very careful microsurgical technique when removing the tumor respecting tumor planes.

Unlike schwannomas the cell origin is not defined for a neurofibroma and may arise from mesenchymal lines. Up to 60% are associate with NF type 1 [18]. Though majority of these are intradural about 30% can be dumbbell lesion with an extradural component. These tumors can become extremely large before they can present with symptoms. Sometimes the extradural component in the thoracic region can be very large entering the pleural cavity. Some of the large intradural tumors cause a fusiform enlargement of the tumor which makes it difficult to spare the nerve [26]. In this setting sacrificing the nerve is not always associated with a neurological deficit [18]. The recurrence rate in these tumors is expected to be about 12% [27]. With MIS technique, the tube is first directed at removing the intradural component and then tilting the tube to get the extradural component [28]. The facet joints have to be sacrificed to remove the tumor which may have scalloped the foramen. It is also possible to remove the tumor without completely removing the facet and working around the tumor from both sides, as we had done in one of our cases; to avoid instrumentation. With MIS in our series, we have not done instrumentation for instability. Dural closure is very difficult and CSF leaks are common in this category of tumors. There is increase chance of CSF leak in the thoracic cavity due to the negative pressure if there is an incomplete removal following dumbbell tumor removal.

In our study, we noticed that the demographic results were similar in patients age, sex, and presentation. This retrospective study has a patient selection bias for the type of operative procedure based on the patient factors and surgeons’ preference. The patient factors were the cost of surgery, as MIS was more expensive, some patients preferred open surgery. Some patients had significant co morbidities and since MIS has a slight longer operating time they were opted for open surgery. The surgeon factors were the choice of approach based on the tumor location, more dorsal placed tumors were preferred for open, between 2–3 vertebral bodies were operated open in the initial phase as the tubular retractor will only expand to 3 vertebral bodies. Meningiomas needed exposure of the dural tail extending beyond the preferred opening [29]. If the radiological diagnosis was certain to be a schwannoma, but extending just beyond 3 vertebral levels or borderline, they were operated by tube as the tumor can be debulked with ultrasonic aspirator and shrunk to deliver it within the space created in the expandable tube. Dumbbell tumors with a large extradural extension were preferred open approach as they needed fixation, and the very large size lesions needed larger exposure. In the later part, these were also operated by MIS with complete GTR of tumor. This emphasizes the fact that the MIS tumor operations have a steady learning curve without compromising on the surgical outcome [27]. Interestingly the commonest location of tumors in our study were cervical spine (47.5%) as compared to most studies where lumbar level tumors were more common. Majority of single vertebral tumors (71%) were operated by MIS technique. The duration of symptoms was seen to be more with the open group (11.2 months). Of the 27 patients who had a single level vertebral tumor were mostly schwannomas and had a duration of presentation of 9.4 months. Those who presented with less duration of symptoms had larger tumors mostly in the lumbar (58%) and cervical (30%) regions.

A subgroup analysis of the MIS group showed that the patient selection in the initial phase of the study was restricted to 1–2 level tumors and then later larger lesions were also approached. The operative time was longer in the MIS group in the initial phase compared to the later cases where the duration became like open. Other factors like blood loss, GTR, and complications were similar. There were no neurological deficits on the MIS group. Residual tumors were seen in 4 open cases and 2 MIS cases in the MRI taken within 24 hours of surgery. The reason for the residual tumors were the large size of the extradural neurofibroma component in 2 cases, dural adhesion in case of a meningioma in 2 cases and adherence to the conus in a myxopapillary ependymoma in 1 case and a neurofibroma that was inadequately removed around the conus. These tumors were followed up and one of the cases showed a slight increase in size radiologically at 3 years in the MIS group but then patient did not want further surgery, and the patient is under follow-up. The author generally does not use wound drains for IDEM tumor excision surgery [23]. In one case where a CSF leak was present despite meticulous attempt to close the dura, a lumbar drain was used for 7 days and wound healed without any complications. Lumbar drains should be placed in the level higher where possible though a separate opening and not through the same opening for the procedure. The reason for this is the possibility of CSF leaking into the wound be due to negative suction pressure in the surgical cavity.

Significant differences (p<0.05) were seen between the groups in intraoperative blood loss, postoperative stay in hospital, return to work, and normal activities with superiority for MIS approach. However, the duration in the surgery group was higher in the MIS group and was attributed to the learning curve in the initial phase. This is variable in the literature also [6,7]. The primary outcomes of the study were GTR and complications (Table 3), which was not different in both groups. Several studies have used MIS technique and achieved 100% GTR rates [30-33], without any serious complications. The literature suggests that GTR in open approaches ranges from 74% to 92% of cases [31-33]. The MIS or mini-open techniques have achieved similar rates of GTR in the literature, ranging from 80% to 94% [30,32]. Similar results were reported by Wong et al. [1,6] with both open and MIS technique the GTR ranges from 80% to 90% in another study of Wong et al. [1,6] attained GTR of 92.6% using true MIS approach. Another advantage of using MIS is reduced CSF leak percentage which is about 0 to 5%. In a study by Thavara et al. [34], GTR was achieved in 67% of cases and NTR in 33% cases. In our experience there was no difference in the GTR rates between groups. The one patient in the MIS group with incomplete resection showed a small increase in size of the lesion and is being managed by observation as patient is asymptomatic.

The main complication that is concerning most surgeons adopting this technique is CSF leak and, in our series, we had no CSF leak in the MIS group and no lumbar drain was used. Though dural closure is technically challenging through tubes it becomes easy with experience. We perform a Valsalva manuver at the end of the procedure and if we see a CSF leak, we either use lumbar drain or dural sealants. Following the MIS technique once the tubes are removed the soft-tissue planes oppose and the muscle planes collapse causing a tamponade effect with no dead space for CSF collection and pseudomeningoceal. Lumbar drain was not needed in any case and in one case a dural sealant was used for CSF leak from the suture entry site but this did not result in a CSF leak post op. In open surgery 0%–9% CSF leak rate was reported [30,31]. In the open group, 1 patient required a lumbar drain but no repair or resurgery was required. No patients in the MIS group had blood transfusion and the average Blood loss was about 65 mL (p<0.01). This is because of the muscle splitting and tamponade by the tube during the procedure. Similar results have been reported in other cases compared to open surgery with 115 mL [34]. One patient in the open group needed a blood transfusion 2nd postoperative day. Surgical site infection (SSI) was seen in 1 patient with open surgery. In the open case, 1 SSI needed a would debridement and readmission after discharge. Neurological deficit was seen equal in both patients with minor sensory and motor deficits. The patients in the open group developed sensory numbness in foot and in another patient was worsening of foot weakness by one grade. This weakness improved to normal in 3 months. In the MIS group, similar mild weakness of thigh was seen following a dumbbell lesion that was excised but no significant disability was seen. This deficit has persisted at 3 years possibly due to nerve injury of L3 during removal. Tumors arising from the cord level have a higher risk of neurological injury compared to those at the lumbar level where the roots have a higher tolerance to manipulation [35]. MIS unilateral approach uses a paramedial dural incision which provides a corridor to see the lateral aspect of the cord and roots avoiding cord manipulation and hence reduced neurological damage [1,5,30]. The neurological deficits are more common in the upper thoracic cord region due to the watershed zone [35]. There was no significant difference in the complication of CSF leak, would infection and neurological complications in both groups (p=0.7).

The secondary outcomes including impact of surgery on symptom relief and tumor recurrence. Increased hospital stay postsurgery is likely due to excessive pain, slow mobility, prolonged antibiotics and lumbar or wound drain. In MIS because of reduced soft-tissue damage the postoperative pain is less and patients mobilized earlier and require less postoperative analgesia and are discharged from hospital earlier. Avoiding a lumbar drain speeds up the discharge process significantly allowing patients to go home by average of 4.4 days (3.5–6.6 days) postoperative (p<0.02). Similar finding is seen reported in the literature for both groups (2.3–4.9 days vs. 4.1–6.3 days) [31,32,36,37]. Similarly return to normal activities including self-care in activities of daily living, toileting, walking inside house and eating in sitting position was seen at 12 days for MIS and >32 days for open (p<0.03). Return to work was seen as another advantage of MIS patients. In our series, all patients returned to work ranging from desk top job to manual labor in 22 days (16–32 days) compared to patients in the open group at average of 45 days (18–94 days).

No patient was reoperated at 3 years (p=0.33). Pre- and postoperative VAS scores were significantly different at 1 month and 3 months but declined at 1 year and 3 years. This highlights that early discharge from hospital, early return to work is an important indicator and is directly related to improved VAS scores up to 3 months, no neurological deficits and complications. Hence MIS patients fare better than the open surgery group upto 3 months. Some of the outcomes not measure is work productivity, financial losses, psychological wellness, improved fitness and long-term prevention of degeneration and back pain postsurgery. It is only logical to conclude that in the immediate phase MIS is superior to open and the long-term results, not measured, are still in favor of MIS in a selective group of patients. Safe GTR of tumors is still the best measure of long-term outcome and we propose to present our results in 5 and 10 years.

One of the major limitations of this study is the section bias in choosing larger (>3 vertebral body levels) and more complex tumors in the cervical and upper thoracic region for open surgery compared to the smaller lesions (<3 vertebral bodies) with more preference in the lumbar level for MIS. However, using a multivariant statistical analysis between the 2 groups showed no significant different in the neurological outcome in both the groups. Absence of a randomized study highlights the complexity of designing such a comparison. If only tumors less than 3 vertebral bodies were compared with the 2 groups still MIS is similar or superior in terms of primary and secondary outcomes.