Minimally Invasive Pedicle Screws Fixation and Percutaneous Vertebroplasty for the Surgical Treatment of Thoracic Metastatic Tumors With Neurologic Compression

Abstract

Study Design. 

A retrospective study.

Objective. 

To describe minimally invasive pedicle screw fixation (MIPS) combined with percutaneous vertebroplasty (PVP), minimally invasive decompression and partial tumor resection for the treatment of thoracic metastasis with symptoms of neurologic compression and evaluate the feasibility, efficacy, and safety.

Summary of Background Data. 

Neurologic decompression, spinal tumor resection, and stabilization with instrumentation should be performed from an anterior, a posterior, or a combined approach for spinal metastatic tumors with symptoms of neurologic compression. These operations, however, have significant morbidity related to the surgical approach, potential blood loss, extensive dissection, or biomechanical instability.

Methods. 

Eighteen patients who sustained single-level thoracic vertebral metastasis and neurologic compression underwent MIPS (The minimal-access in a paraspinal sacrospinalis muscle-splitting approach was performed to insert the pedicle screws into the vertebrae under direct vision and two rods of appropriate size were placed over the pedicle screws through subcutaneous soft tissues and muscles) combined with PVP, minimally invasive neurologic decompression, and partial tumor resection. The patients were evaluated preoperatively according to the Tomita, revised Tokuhashi, Bilsky grading system, and Spinal Instability Neoplastic Score. Pre- and postoperative VAS score, American Spinal Injury Association grade, ambulatory status, and urinary continence were also recorded. The Cobb angles, central, and anterior vertebral body height were measured on the lateral radiographs before surgery and during the follow-up.

Results. 

Clinical follow-up was available for 17 patients in this study ranging from 12 to 16 months (mean time, 14.2 months), and 1 patient died 8 months after surgery. The Visual Analog Scale was significantly decreased after surgery. Improvement of paraplegia was observed after surgery in all of these patients. Spine stability was observed in all of the surviving patients during the follow-up.

Conclusion. 

MIPS combined with PVP, minimally invasive decompression, and partial tumor resection is a good choice of surgical treatment of thoracic metastatic tumors with neurologic compression.

Level of Evidence: 2

Percutaneous vertebroplasty (PVP) is a minimally invasive, radiologically guided therapeutic procedure that is performed to reduce pain and stabilize structurally weakened vertebrae through the injection of cement into the vertebral body. Analgesic efficacy was achieved in most patients with spinal malignancies treated with PVP.^1–6 In addition to increasing the strength and stiffness of the vertebral body, PVP is important for the prevention of further vertebral collapse and spinal cord compression.^7,8

Only PVP without decompression is not regarded as optimal for spinal metastatic tumor with symptoms of neurologic compression because the procedure cannot improve neurologic function.⁹ Neurologic decompression, spinal tumor resection, and stabilization with instrumentation should be performed from an anterior, a posterior, or a combined approach.^10,11 The surgical options of spinal tumor resection include total en bloc spondylectomy (TES), vertebrectomy, sagittal resection, resection of the posterior arch, piecemeal excision, and eggshell curettage.^10,11 These operations, however, have significant morbidity related to the surgical approach, potential blood loss, extensive dissection, or biomechanical instability due to the collapse of the involved vertebrae.^12–14 In the present study, minimally invasive pedicle screw fixation (MIPS)^15,16 combined with PVP, minimally invasive neurologic decompression, and partial tumor resection through a mini-posterior midline approach were performed to treat thoracic metastasis with neurologic compression to decrease the aggressiveness and blood loss, and to prevent anterior column collapse and hardware failure. This surgical technique was described and the feasibility, efficacy, and safety of minimally invasive surgery were evaluated.

MATERIALS AND METHODS

Population Data

The radiologic and clinical records of patients with single-level thoracic vertebral metastasis and symptoms of neurologic compression who underwent MIPS combined with PVP, minimally invasive decompression and partial tumor resection in our hospital between December 2013 and June 2014 were retrospectively reviewed. Before the procedure, all patients provided informed consent, and our hospital review board did not require further approval for the use of patients’ records and images.

Neurologic deficit with one thoracic vertebral body lesion associated with neurologic compression related to metastatic spinal disease was an indication for MIPS combined with PVP, minimally invasive decompression, and partial tumor resection. Patients who underwent PVP for a malignant vertebral lesion without epidural involvement or for trauma, osteoporosis, or angioma were also excluded from the present study.

Pre- and Postprocedural Imaging

All patients were evaluated before the procedure by computed tomography (CT) and magnetic resonance imaging (MRI) to determine whether there was cortical breakthrough, extravertebral extension, or to determine the type of lesion (osteolytic, osteoblastic, or mixed) and quantify the degree of vertebral collapse and neurologic compression. Epidural spinal cord compression (ESCC) was evaluated using Bilsky grading system¹⁷ with the T2-weighted MRI images: grade 0 (bone only disease), grade 1a (epidural impingement), 1b (thecal sac deformation), 1c (cord impingement), grade 2 (cord compression with cerebrospinal fluid visible) and grade 3 (cord compression with no cerebrospinal fluid visible). After treatment, CT images were obtained to assess neurologic decompression, the position of the pedicle screws, and the extravertebral leakage of cement.

Spine Instability Assessment

The Spinal Instability Neoplastic Score (SINS) system^18,19 (Table 1), developed by the Spinal Oncology Study Group, evaluates spinal stability by adding together six radiographic (location, bone lesion, radiographic spinal alignment, vertebral body collapse, and posterolateral involvement of spinal elements) and clinical (pain) components, with a score ranging from 0 to 18. The total score is divided in three categories of stability: stable (0–6 points), potentially unstable (7–12 points), and unstable (13–18 points).

TABLE 1: Spinal Instability Neoplastic Scoring System

Surgical Procedure

All surgeries were undertaken by the senior spine surgeon (Y.-T.G). Patients were positioned prone on a radiolucent table. MIPS (Viper; Depuy Spine, Raynham, MA) combined with PVP (Ruibang, Shanghai, China), minimally invasive decompression and partial tumor resection were performed on all cases under general anesthesia and antibiotic prophylaxis (2 g of cefazolin during surgery and 2 g twice in the following 24 hours). The pedicles of the involved vertebrae were identified, and the skin was marked under posteroanterior fluoroscopic control before beginning the surgical procedure.

The first surgical stage consisted of performing MIPS and PVP. During MIPS, the cannulated pedicle screws were placed into the adjacent two-level vertebrae above and below the involved vertebrae through the pedicles using the minimally invasive technique. The minimal access in a paraspinal sacrospinalis muscle-splitting (Wiltse) approach²⁰ was performed to expose the superior articular facet and root of the transverse process. The entry site to the pedicle was located at the junction between the lateral border of the superior articular facet and superior 1/3 line of the transverse process. Once the pedicle has been identified, either a pedicle probe or a hand-held curette was used to enter the pedicle (Figure 1A). Preoperative anteroposterior and lateral x-rays and CT scans through the pedicles of the vertebral body to be instrumented were studied to determine the correct angle of entry in both the coronal and sagittal planes. The pedicle integrity was verified in all four quadrants to ensure that a solid tube of bone exists and that violation into the spinal canal or inferiorly into the neuroforamen has not occurred. Eight pedicle screws of the appropriate length were then introduced into the vertebral body via the pedicle to engage at least 75% of the vertebral body's anterior-posterior width. Posteroanterior and lateral x-rays were taken to confirm their position, and two 13-gauge puncture needles were then passed into the anterior central aspect of the metastatic vertebral body through the pedicles under fluoroscopic guidance. For PVP, bone cement (PMMA, polymethylmethacrylate) was injected under constant fluoroscopy into the target vertebral body (Figure 1B) until the cement approached the posterior aspect of the vertebral body or leaked into an extraosseous space, such as the intervertebral disc or an epidural or paravertebral vein. Two rods of the appropriate size were contoured to maintain a normal spine curve and placed over the pedicle screws through subcutaneous soft tissues and muscles (Figure 1C).

Figure 1: Procedure of minimally invasive pedicle screw fixation (MIPS) and percutaneous vertebroplasty (PVP). A, Photography showing a pedicle probe used to enter the pedicle after exposure of the superior articular facet and root of transverse process through the minimal-access in a paraspinal sacrospinalis muscle-splitting (Wiltse) approach before the introduction of the annulated pedicle screw into the vertebrae above the involved vertebrae under direct vision with the minimally invasive technique. B, Injection of cement through 13-gauge needles into the target vertebral body under fluoroscopic guidance after the placement of eight pedicle screws. C, Photography showing two contoured rods of the appropriate size placed over the pedicle screws through subcutaneous soft tissues and muscles.

In the second surgical phase, minimally invasive neurologic decompression and partial tumor resection were performed. The small midline incision was used to cut the skin, superficial facial and deep facial. The paraspinal sacrospinalis muscle was elevated subperiosteally to expose the spinal process and lamina through the mini-midline approach. The spinal process and lamina were removed to expose the dural sac involved in neurologic decompression. Facetectomies and pedicle resection were performed to get the posterior part of vertebral body with the rongeur and Kerrison Rongeur or drill. Bipolar cautery and curettes or pituitary rongeurs were used for piecemeal resection of tumor to create a cavity in the vertebral body. Then the curved dura dissector was carefully inserted into the interface between the tumor and dura to push forward the posterior tumor compressing neurologic elements into the cavity of vertebral body and make “separation” of spinal cord and tumor for complete decompression, a procedure termed partial tumor resection (Figure 2). Hemostasis by compression was performed by packing absorbable hemostatic gauze into the cavity of vertebral body. Attention should be paid to protect the spinal cord and nerve root during the surgery.

Figure 2: After minimally invasive pedicle screw fixation (MIPS) and percutaneous vertebroplasty (PVP), the spinal process and lamina were removed to expose the dural sac involved using the mini-midline approach for minimally invasive neurologic decompression. In addition, the pedicle involved and posterior parts of the vertebral body compressing neurologic elements were removed as much as possible for partial tumor resection.

No external braces were used after the operation. The patients were mobilized as soon as feasible after surgery. After leaving the hospital, the patients were encouraged to resume their daily routine and were followed up as outpatients at the hospital ward. Postoperative chemotherapy was offered to patients with good medical conditions and radiation therapy to 12 patients to prevent local recurrence of thoracic metastasis. Intensity-modulated radiotherapy was given to the operated level 2 weeks after surgery when the wound had healed. The total radiation dosage was usually 4000 cGy and was administered during a 20-day course in 4 weeks, with 200 cGy doses per day delivered in the first 5 days and then a break of 2 days during 1 week.

Clinical Follow-Up

All of the patients had clinical and x-ray reevaluation immediately, 1 month, 2 months, 3 months, 6 months, 1 year, and final follow-up after surgery. Postoperative complications including wound dehiscence and infection were recorded.

The pain intensity in the previously symptomatic region was graded using the visual analog scale (VAS) pain score. A neurologic examination was performed before and after treatment in all of the patients. The severity of the neurologic deficit was assessed using the ASIA (American Spinal Injury Association) impairment scale as follows²¹: A (complete impairment), no motor or sensory function is preserved in segments S4 through S5; B (incomplete impairment), sensory but not motor function is preserved below the neurologic level and includes segments S4 through S5; C (incomplete impairment), motor function is preserved below the neurologic level and more than half of key muscles below the neurologic level have a muscle grade of less than 3; D (incomplete impairment), motor function is preserved below the neurologic level and at least one-half of key muscles below the neurologic level have a muscle grade of 3 or more; and E (normal impairment), motor and sensory function are normal. Ambulatory ability and urinary continence were also evaluated before and after surgery.

The spine was considered stable when no modification in the curvature of the spine or height of the vertebral body was observed during the follow-up, as observed on spine x-rays. The Cobb angles and the central and anterior vertebral body heights were measured on the lateral radiographs by the same physician. The vertebral body height was calculated as a percentage of the estimated, intact vertebral body height by averaging the respective central and anterior heights from the adjacent levels.²²

Statistical Analysis

Comparison of preoperative and postoperative VAS, Cobb angles, and the central or anterior vertebral body height was performed using a linear mixed-effects model for multiple comparison procedures. Statistically significant differences were defined at a 95% confidence level. SPSS software (SPSS Inc., Chicago, IL) was used to perform statistical evaluations.

RESULTS

Eighteen patients, ten women and eight men with a mean age of 57.2 ± 10.7 years, were included in the present study (Table 2). The mean prognostic score was 7 (range, 6–7 points) according to the Tomita scoring system¹⁰ (Table 3), 11 (range, 9–12 points) of modified Tokuhashi score^23,24 (Table 4) at baseline. There were 7 patients with ESCC grade 2 and 11 patients with ESCC grade 3 on the MRI image (Figure 3 A and B). The mean SINS was 11 (range, 6–14) for spinal instability assessment.

TABLE 2: Summary of the Clinical Data of the Patients

TABLE 3: Tomita Scoring System

TABLE 4: Modified Tokuhashi Scoring System

Figure 3: A, Sagittal and (B) axial magnetic resonance (MR) images show metastasis of T6 from liver cancer with neurologic compression in a 75-year-old man. C, The lateral x-ray picture after MIPS combined with PVP, minimally invasive decompression, and partial tumor resection shows good position of the pedicle screw construct. There was no leakage of cement into the spinal canal, and neurologic decompression is complete on (D) axial CT and (E) sagittal CT reconstruction. (F) Photography shows minimally invasive results 14 months after surgery (Case 1).

The mean interval between the onset of neurologic deficit and surgery was 3.5 days (range, 1–7 days). MIPS and PVP, minimally invasive decompression, and partial tumor resection were successfully performed for all of the cases in the present study. The mean duration of the operation was 204.7 ± 10.5 minutes. The mean length of the midline incision was 3.1 ± 0.3 cm (Figure 3F). There was a mean blood loss of 150 mL (range, 70–600 mL). The mean amount of cement injected was 5.8 ± 1.3 mL. The mean stay at the hospital was 6 days (range, 4–7 days).

The postoperative x-rays and CT scan images showed that the position of the pedicle screw construct was good (Figure 3C), and the neurologic decompression was complete (Figure 3D and E). There is no leakage of PMMA into the spinal canal, and three cases of leakage into the intervertebral disc, paraspinal soft tissues, or paravertebral vein without clinical consequences were observed.

Clinical Outcome

Clinical follow-up was available for 17 patients in the present study ranging from 12 to 16 months (mean time, 14.2 mo), and 1 patient died 8 months after surgery. There were no perioperative complications such as wound dehiscence, infection, and no death due to complications of the procedure itself.

The VAS significantly dropped from 9 (range, 7–10) preoperatively to 3 (range, 2–4) (P < 0.001) immediately after surgery and to 1 (range, 0–1) (P < 0.001) at the 1-year follow-up. In the present study, 4 patients presented with complete motor paralysis (no motor function, ASIA scale B), and 14 presented with incomplete motor paralysis (ASIA scale C or D). Improvement in paraplegia was observed after surgery in all of these patients. At the 3-month follow-up, 2 of 4 patients with complete motor paralysis improved from ASIA scale B to D, 12 of 14 patients with incomplete motor paralysis improved from ASIA scale C or D to E. Thirteen of 17 surviving patients reached ASIA scale E at the 1-year follow-up (Table 2). Five of 7 (71%) patients who were unable to walk before surgery regained ambulatory ability. Of those patients who were ambulant preoperatively, 100% maintained mobility. Six of 8 (75%) patients who were incontinent before surgery recovered urinary sphincter function.

Radiological Examination

Spine stability was observed in all of the surviving patients at the 1-year follow-up, and there was no significant difference in the postoperative Cobb angle, and central or anterior vertebral body height on spine x-rays during the follow-up (Table 5).

TABLE 5: Variations of Each Measured Parameters From the Initial Evaluation to the Last Follow-up

DISCUSSION

The strategy for spinal metastases was decided along with the treatment goal set according to the Tomita prognostic scoring system: a prognostic score of 2 to 3 points suggested a wide or marginal excision such as TES, vertebrectomy, sagittal resection, or resection of the posterior arch for long-term local control; 4 to 5 points indicated marginal or intralesional excision such as piecemeal excision, eggshell curettage for middle-term local control; 6 to 7 points justified palliative surgery for short-term palliation; and 8 to 10 points indicated nonoperative supportive care.^10,11,26 TES for complete resection of spinal tumors is part of our armamentarium to potentially cure neoplastic disease in selected patients.²⁷ Although TES of primary malignant tumors or aggressive benign tumors in medically fit patients is accepted, controversy exists in cases of spinal metastatic disease that is distant from the primary neoplasm,²⁸ and there are no data supporting improved quality of life and prolonged survival following TES over alternative, less aggressive, surgical and medical treatment paradigms. Although surgical techniques have been remarkably improved, TES, vertebrectomy, and sagittal resection may remain too invasive surgical procedures in such cases, and the morbidity and mortality rates related to the surgical approach, potential blood loss, extensive dissection, and long operative time remain relatively high.^12–14 In addition, piecemeal excision, eggshell curettage, and palliative surgery still carry the risk of massive blood loss, and it is not easy to achieve biomechanical stability due to the collapse of the involved vertebral body, although screws and rods are used.

Minimally invasive approaches, including MIPS and PVP, dramatically decrease paraspinal musculature iatrogenic injury.^15,16 Unlike traditional midline incision, the Wiltse approach protected the attachment of the muscle to bone, avoided disruption of the supraspinous and interspinous ligaments, provided a more direct approach to the transverse processes and pedicles, and decreased bleeding and postoperative pain.^29–31 Compared with percutaneous pedicle screws, MIPS uses incisions of similar size but with easier manipulation and less fluoroscopic monitoring during the operation. In the present study, we designed MIPS and PVP, minimally invasive decompression, and partial tumor resection to treat 18 cases of thoracic vertebral metastasis with symptoms of neurologic compression. Two rods were placed over the pedicle screws through subcutaneous soft tissues and muscles following the insertion of cannulated pedicle screws and PVP before neurologic decompression and partial tumor resection to decrease the blood loss. The procedure can be completed as soon as possible after decompression, if there is massive blood loss during neurologic decompression and partial tumor resection. In addition, minimally invasive neurologic decompression and partial tumor resection were performed using the mini midline approach to decrease aggressiveness and blood loss. In addition, PVP, the injection of cement into the vertebral body, could embolize the rich vessels of the tumor and may help reduce hemorrhage when the posterior part of the vertebra compressing neurologic elements was removed. The morbidity and mortality rates of surgery were minimized in our study, although there was the absence of any data supporting prolonged survival following the procedure.

In the present study, pain improvement in the previously symptomatic region was achieved in all patients of the present study after surgery. The postoperative x-rays and CT scan images showed that neurologic decompression using the mini-midline approach was complete (Figure 3D and E), and proper angulation could be achieved for neurologic decompression medially, although the small incision made it difficult to reconstruct the anterior spine column with cage or other instrument. Improvement in paraplegia was observed after surgery in all of these patients. Thirteen of the 17 surviving patients reached ASIA scale E at the 1-year follow-up. The neurologic function prognosis had likely relevance to the interval between the onset of the neurologic deficit and surgery, particularly in patients with complete motor paralysis. Earlier, decompression was performed, better recovery of neurologic function was observed, and the surgery should be undertaken during 2 days after the onset of a neurologic deficit for patients with complete motor paralysis. Spine stability was observed in all of the surviving patients at the 1-year follow-up. In the cadaveric biomechanical study by Mermelstein et al³², it was found that the injection of cement in a burst fracture reduced the load on the pedicle screw construct that was inserted for fracture stabilization, and vertebroplasty with cement after posterior instrumentation might reduce hardware failure and anterior column collapse. Our study confirmed that there was no hardware failure in any patient following MIPS and PVP without fusion during the follow-up.

In the study by Bilsky et al,³³ open circumferential decompression and fusion using a single-stage posterolateral transpedicular approach were performed for patients with metastatic spinal cord compression and the mean intraoperative blood loss was 1700 mL (200–4000 mL) as assessed by the anesthesiologist. The mean intraoperative transfusion requirement was 3.5 U. The mean operating time was 7 hours. The mean hospital stay was 11 days. Three deaths occurred within 30 days after surgery, and wound dehiscence occurred in 4% of patients in this series. Compared with open circumferential decompression and fusion, MIPS combined with PVP, minimally invasive decompression and partial tumor resection had advantages in blood loss, duration of operation, stay at the hospital, and postoperative complications. In addition, Bilsky et al³³ reported that 5 out of 6 patients with ASIA D improved a grade to become neurologically intact (ASIA E) after open circumferential decompression and fusion, and patients with significant neurologic deficit before surgery (ASIA C) did not fare as well as patients who were neurologically intact. The two patients with radiographic high-grade spinal cord compression in this group worsened neurologically. Four of 6 (67%) patients who were not able to walk before surgery improved to ambulatory status. Hatrick et al,³⁴ Fourney et al,³⁵ and Sundaresan et al³⁶ found that the proportion of nonambulatory patients with metastasis who regained ambulatory function after circumferential spinal cord decompression was 57%, 46%, and 94%, respectively. The study of Patchell et al³⁷ showed that 10 of 16 (62%) patients who were unable to walk regained the ability after direct circumferential decompression of spinal cord. Jonsson et al³⁸ described that 19 of 25 (76%) preoperatively nonambulatory patients regained walking ability after direct or indirect decompression and stabilization surgery for spine metastases. Quan et al³⁹ reported that 17 out of 25 (68%) patients who were unable to walk preoperatively regained mobility after spinal surgery for symptomatic vertebral metastases. Thirteen out of 22 (59%) patients who were incontinent preoperatively recovered urinary sphincter function. Three patients, however, became incontinent. Our study showed that 5 of 7 (71%) nonambulatory patients before surgery regained walking ability and 6 of 8 (75%) patients who were incontinent before surgery recovered continent function after minimally invasive surgery. Compared with most published benchmarks, the outcomes using this minimally invasive technique were noninferior in the improvement of neurologic deficit, ambulatory ability, and urinary continence.

The feasibility and relative safety of MIPS combined with PVP were supported by the postoperative x-rays and CT scan images showing that the screws and cement were all properly positioned in the present series of patients. None of the patients was found to have any postoperative neurologic complications. Like all surgical interventions, pedicle screw stabilization is not devoid of risks because it can cause nerve injuries. The pedicle must be carefully probed in all four quadrants to ensure that a solid tube of bone exists and that violation into the spinal canal or inferiorly into the neuroforamen has not occurred before the cannulated pedicle screws are implanted into the vertebrae with the minimally invasive technique under direct vision in our study. Cement instillation involves the risks of complications, including cement leakage into the spinal canal. During the PVP procedure, we injected bone cement into the target vertebral body under constant fluoroscopy, a step that must be stopped if the cement approached the posterior aspect of the vertebral body or leaked into an extraosseous space. These measures were taken to avoid the aggravation of neurologic deficits and guarantee the safety of the operation. Although the fluoroscopic monitor was needed during the minimally invasive surgery, the amount of radiation the patient received was as limited as that in the only PVP procedure, because MIPS technique described in the present study was performed under direct vision, which did not depend on the fluoroscopy. The study of Saliou et al⁶ showed that no deterioration of neurologic symptoms was observed after only PVP among the patients with malignant vertebral fractures with spinal cord or cauda equina compression and irreversible neurologic deficit. PVP before neurologic decompression is feasible and safe for patients with spine vertebral metastasis and neurologic symptoms and would not increase the risk of aggravating the symptoms, finding that were also supported in our study.

In our study, inclusion criteria were thoracic metastasis with neurologic compression (ESCC grade 2 or 3) of one vertebral body lesion, and Tomita prognostic score of 6 to 7, which was treated with our minimally invasive strategy. The ESCC grading system has been shown by some authors to be an informative instrument that may guide treatment decisions.¹⁷ In the absence of mechanical instability, Grades 0, 1a, 1b, and 1c may be considered for radiation or PVP as initial treatment. Grades 2 and 3, described high-grade ESCC, need neurologic decompression surgery. The SINS can also be analyzed as a binary indicator of surgical referral status: “stable” (0–6) or “current or possible instability” (7–18).⁴⁰ Surgical consultation is recommended for those patients with a score of ≥7, PVP or internal fixation for possible instability (7–12), and internal fixation for current instability (13–18). The patients of the present study had 11 (range, 6–14) SINS. Although SINS of some patients were less than 7, the neurologic decompression and internal fixation surgery was needed for their ESCC grade 2 or 3.

Although the patients had a Tomita score of 6 to 7 which corresponds to a life expectancy of 6 to 12 months, only one patient died before 12 months in the present study. The prognostic results were simultaneously evaluated using modified Tokuhashi score system^23,24 (Table 4), which consists of the sum of six parameters that are used to measure the severity of the disorder: the general condition of the patient, the number of extraspinal bone metastases, the number of vertebral metastases, the number of metastases to the major internal organs, the primary site of the cancer, and spinal cord palsy. The final score divides patients into three prognostic categories, 12 or more with a good prognosis (predicted survival > 12 months), 9 to 11 a moderate prognosis (predicted survival > 6 months), and 8 or less a poor prognosis (predicted survival < 6 months). The mean prognostic score of patients was 11 (range, 9–12 points) according to modified Tokuhashi system, which indicates a moderate or good prognosis in the present study. It seems that modified Tokuhashi score is slightly better than Tomita in the ability to accurately predict individual survival, although it remains challenging to identify patients with spinal metastasis who will die early or, conversely, who will be long-term survivors.

MIPS combined with PVP, minimally invasive decompression, and partial tumor resection is a good choice of surgical treatment for thoracic metastatic tumors with neurologic compression.

Key Points

• The feasibility and safety of MIPS combined with PVP, minimally invasive decompression, and partial tumor resection were supported by postoperative x-rays and CT scan images showing that the screws and cement were all properly positioned in the present series of patients.
• The improvement of pain and ASIA score were achieved, and spine stability was observed, in all of the surviving patients of thoracic vertebral metastasis with neurologic compression at the 1-year follow-up after MIPS combined with PVP, minimally invasive decompression, and partial tumor resection.
• The minimally invasive surgery decreased the aggressiveness and blood loss, findings that were supported by the mean blood loss of 150 mL (range, 70–600 mL), length of midline incision of 3.1 ± 0.3 cm, and the ability of patients to return home quickly after a mean hospital stay of only 6 days (range, 4–7 days).
• More studies should be performed to compare this minimally invasive surgery with open surgery and compare MIPS with percutaneous pedicle screw fixation to treat thoracic metastatic tumors with neurologic compression.