Total en bloc spondylectomy for spinal tumours: Technical aspects and surgical details

ARTICLE IN PRESS n e u r o c i r u g i a . 2 0 1 7;x x x(x x):xxx–xxx

NEUROCIRUGÍA www.elsevier.es/neurocirugia

Clinical research

Total en bloc spondylectomy for spinal tumours: Technical aspects and surgical details夽 Pedro David Delgado-López a,∗ , Antonio Rodríguez-Salazar a , Vicente Martín-Velasco a , José Manuel Castilla-Díez a , Javier Martín-Alonso a , Ana Galacho-Harriero a , Cecilia Gil-Polo b , Elena Araus-Galdós c a b c

Servicio de Neurocirugía, Hospital Universitario de Burgos, Burgos, Spain Servicio de Neurología, Hospital Universitario de Burgos, Burgos, Spain Servicio de Neurofisiología Clínica, Hospital Universitario de Burgos, Burgos, Spain

a r t i c l e

i n f o

a b s t r a c t

Article history:

Objective: To describe the specific surgical details and report the lessons learned with a series

Received 2 May 2016

of patients suffering from spinal tumours that underwent total en bloc spondylectomy (TES).

Accepted 24 July 2016

Methods: A retrospective case series review is presented, together with an analysis of the

Available online xxx

clinical and technical variables, as well as the outcomes.

Keywords:

chondrosarcoma, fibrosarcoma and chordoma) and secondary spinal tumours (lung,

Results: A total of 10 patients underwent TES (2000–2016) for primary (osteosarcoma, En bloc spondylectomy

breast, thyroid, oesophagus, and meningioma metastases). According to the Tomita

Total vertebrectomy

classification, 2 patients had intra-compartmental tumours, and the rest presented as extra-

Tomita

compartmental. All patients experienced an improvement in their pain level after surgery.

Spinal tumour

Nine patients preserved ambulation post-operatively and one patient developed paraplegia.

Surgical technique

Six patients needed subsequent operations for wound debridement, tumour recurrence, or

Radical resection

revision of the fixation. Other complications included pneumothorax, pleural effusion and venous thrombosis. Four patients remain alive (4 months to 15 years follow-up). The rest died due to primary tumour progression (6.5 months to 12 years). A detailed description of the surgical steps, tips, and pitfalls is provided. Modifications of the technique and adjuncts to resection are commented on. Observation of some considerations (selection of candidates, careful blunt vertebral dissection, strict blood loss control, careful handling of the spinal cord, and maintenance of the radical resection concept at all stages) is key for a successful operative performance. Conclusion: TES is a paradigmatic operation, in which the concept of radical resection provides functional effectiveness and improves survival in selected patients suffering from

夽 Please cite this article as: Delgado-López PD, Rodríguez-Salazar A, Martín-Velasco V, Castilla-Díez JM, Martín-Alonso J, Galacho-Harriero A, et al. Vertebrectomía total en bloque en tumores vertebrales: consideraciones técnicas y detalles quirúrgicos. Neurocirugía. 2016. http://dx.doi.org/10.1016/j.neucir.2016.07.002 ∗ Corresponding author. E-mail address: [email protected] (P.D. Delgado-López). ˜ ˜ S.L.U. All rights reserved. 2529-8496/© 2016 Sociedad Espanola de Neurocirug´ıa. Published by Elsevier Espana,

NEUCIE-253; No. of Pages 16

ARTICLE IN PRESS 2

n e u r o c i r u g i a . 2 0 1 7;x x x(x x):xxx–xxx

spinal tumours. Our preliminary experience allows us to highlight some specific and relevant features, especially those favouring a simpler and safer operation. ˜ ˜ S.L.U. All rights © 2016 Sociedad Espanola de Neurocirug´ıa. Published by Elsevier Espana, reserved.

Vertebrectomía total en bloque en tumores vertebrales: consideraciones técnicas y detalles quirúrgicos r e s u m e n Palabras clave:

Objetivo: Describir detalles quirúrgicos específicos y comentar algunas lecciones aprendidas,

Vertebrectomía en bloque

a partir de una serie de pacientes con tumores vertebrales a los que se les practicó una

Vertebrectomía total

vertebrectomía total en bloque (VTB).

Tomita

Métodos: Presentamos una serie retrospectiva de casos. Se analizan variables clínicas, téc-

Tumor vertebral

nicas y de resultado.

Técnica quirúrgica

Resultados: Se intervinieron 10 pacientes (2000-2016) mediante VTB por tumor primario

Extirpación radical

vertebral (osteosarcoma, condrosarcoma, fibrosarcoma y cordoma) o secundario (pulmón, mama, tiroides, esófago y meningioma). Según la clasificación de Tomita, 2 pacientes presentaban lesiones intracompartimentales y el resto extracompartimentales. Todos los pacientes mejoraron del dolor tras la cirugía. Nueve pacientes conservaron la capacidad de caminar en el postoperatorio y uno desarrolló paraplejía. Seis pacientes precisaron reoperaciones por desbridamiento de la herida quirúrgica, recidiva o revisión de la fijación. Otras complicaciones fueron neumotórax, derrame pleural y trombosis venosa. Cuatro pacientes ˜ sobreviven (tras 4 meses y hasta 15 anos). El resto fallecieron por progresión del tumor ˜ primario (de 6,5 meses a 12 anos). Se realiza una descripción detallada de los pasos quirúrgicos, consejos y dificultades de la técnica. Se comentan ciertas modificaciones de la técnica y otras cuestiones relativas a la resección. El respeto a ciertas consideraciones (selección de los candidatos, disección vertebral cuidadosa, control estricto del sangrado, manejo cuidadoso de la médula y mantenimiento del concepto de resección radical en todo momento) es clave para realizar con éxito esta intervención. Conclusión: La VTB es una intervención paradigmática en la que el concepto de resección radical implica efectividad funcional y mejora la supervivencia en pacientes seleccionados portadores de tumores vertebrales. Esta experiencia preliminar nos permite destacar algunas de sus características relevantes, especialmente aquellas dirigidas a simplificar la técnica y hacerla más segura. ˜ ˜ S.L.U. Todos © 2016 Sociedad Espanola de Neurocirug´ıa. Publicado por Elsevier Espana, los derechos reservados.

Introduction Total en bloc spondylectomy (TES) is a surgical technique designed to conduct radical resection of primary and secondary spinal tumours. It is generally called the Tomita technique when it is done using a single posterior approach. This author and his collaborators described the procedure for the first time in the nineties.1–3 Since then, various studies have been published that provide clinical results, as well as improvements and modifications of the original technique.4–7 Performing a TES involves the complete resection of two fragments of the vertebra. Once the pedicles are selected, the posterior arch of the vertebra is separated from the rest of the vertebral body. With a series of surgical manoeuvres, it is possible to extract both fragments en bloc while maintaining the integrity of the tumour capsule. Compared with curettage or intralesional resection (piecemeal), this technique

reduces the risk of local recurrence4,8,9 and is presented as an alternative to total spondylectomy using combined methods as previously described by Boriani et al.,10,11 Sundaresan et al.,12 and others.13,14 In solitary metastasis of the thoracolumbar spine, TES has recently been compared to the mini-open anterior corpectomy approach. Although complication rates were similar in both, the mini-open technique caused less bleeding, there was less need for instrumentation material and it involved less surgery time. However, patients that underwent TES developed fewer local recurrences.15 TES is currently considered to be an effective option for selected patients, those harbouring non-metastasised spinal tumours, in which en bloc resection is considered to improve functionality and increase survival.4,5,9 However, many consider this procedure to be high-risk, due to the inherent technical difficulty and the potential risk of severe neurovascular morbidity.4–6,15–18 A systematic review by the

ARTICLE IN PRESS 3

n e u r o c i r u g i a . 2 0 1 7;x x x(x x):xxx–xxx

Spine Oncology Study Group concluded that the degree of recommendation for conducting a TES is strong, keeping in mind the high adverse events profile even at experienced centres.19 In this study, we discuss our preliminary experience with TES, various technical aspects and lessons learned from the first 10 cases.

Patients and methods Selecting patients, indications and strategies We present a retrospective case series. The potential TES candidates were referred to our unit from the Neurology and Medical Oncology departments, or they were directly identified in the Neurosurgery outpatient clinic. Only a small percentage of patients with spinal metastasis were considered possible candidates. In general, the procedure is recommended as long as a needle biopsy of the spinal tumour is

Extra-compartmental

Intra-compartmental

Type 1

Lesion limited to vertebral body

Type 2

Lesion extends to the pedicle

Type 5

Type 4

Lesion presents epidural extension

possible before planning a TES, given that certain primary spinal tumours only benefit from radiotherapy or chemotherapy treatment. In addition, a computed tomography (CT) of the entire body is needed to rule out systemic metastasis. The surgical stage is determined according to the Weinstein–Boriani–Biagini staging system.20 In general, those patients in which long-term local control is considered feasible are potential candidates for TES. According to the prognostic scoring system developed by Tomita et al.,5 patients with slow-growing spinal tumours and no single isolated visceral metastasis or spinal metastasis with a life expectancy, a priori, of at least two years are preferred for this group. TES candidates are selected and classified according to the degree of spinal impairment, according to the Tomita classification2,3 developed for spinal metastasis (Fig. 1). A TES is recommended both in intra- and extra-compartmental tumours: types 3–5, and possibly in types 1, 2 and 6. However, these criteria are not strict and the suitability of candidates must be discussed in an individualised manner addressing

Lesion presents paravertebral extension

Type 3

Lesion extends to the pedicle and blade

Type 6

Lesion extends to adjacent vertebrae

Multiple

Type 7

Lesions at multiple non-contiguous levels

Fig. 1 – Tomita classification based on the degree of spinal involvement. For type 1 the following is recommended: no treatment (in primary benign latent tumours), intralesional or marginal resection (in primary benign active tumours) or TES (in primary benign aggressive tumours). For type 2, an intralesional or marginal resection or a TES (in primary aggressive tumours) is recommended. For types 3–5, the recommended technique is TES. For type 6, TES (when there is no large vessel involvement) or a combined anterior-posterior approach (for invasive tumours) or palliative treatment is recommended. Lastly, for type 7 only palliative treatment is administered.

ARTICLE IN PRESS 4

n e u r o c i r u g i a . 2 0 1 7;x x x(x x):xxx–xxx

the particular clinical and radiological characteristics of each case.

Surgical protocol TES is a complex technique previously described in detail.2,6,21 Various key steps exist that need to be performed in consecutive order. The generic TES procedure includes: exposing the posterior arch at the level of the affected vertebra, including the proximal segment of the ribs bilaterally; horizontally cutting the pedicles (using a tool called a threadwire saw or Tsaw22 ) and en bloc resection of the posterior elements; blunt dissection of the lateral and anterior sides of the vertebral body (separating the vertebrae from the pleura and from the large vessels); rigid unilateral stabilising before conducting the proximal and distal cuts of the intervertebral discs; en bloc extraction of the vertebral body by rotating it around the dural sac, and reconstructing the spine by placing a substitute on the vertebral body and finally fixing the internal posterior pedicle. Postoperative patients are allowed to walk after the second or third day, with or without a brace. Drainage is left at the surgical site in all patients, with aspiration at all times to minimise the appearance of pneumothorax. Broadspectrum antibiotics are administered (second generation cephalosporins or vancomycin) at the start of the procedure. A second dose is administered 4–5 h after surgery and every 8 h from then onwards for the next 3–4 days. Patients previously treated with corticosteroids due to cord compression symptoms will have their dose progressively decreased until the drugs are discontinued. Common analgesics are administered (paracetamol and non-steroidal anti-inflammatory drugs) for the initial postoperative period. Opiates are occasionally needed. Patients are evaluated at the Neurosurgery outpatient clinic through clinical and radiological evaluations. Patients are followed until the disease stabilises, progresses or causes death.

(three were type 4 and two were type 5), including another three with tumours spreading to adjacent vertebrae or type 6. All patients underwent a TES procedure as has been described above. A three-level spondylectomy was required in one case and two two-level spondylectomies in another two cases. Six patients needed to undergo surgery again for debridement due to a wound infection (3), tumour recurrence (1) or revision of the fixation (2), as described in Table 1. The duration of the surgery varied between 345 and 535 min (average of 410 min). Blood loss was not recorded quantitatively during the procedure, but all patients eventually needed an intra or postoperative blood transfusion (from 2 to 7 units, average of 4 units). All patients experienced improvement in pain and were able to reduce or eliminate analgesic medication (three patients completely free of pain). Eight patients preserved their neurological state after the procedure. One patient developed a mild lower limb proprioceptive deficit that was resolved in three months. Intraoperative neurological damage occurred in one patient who developed a complete and permanent spinal cord syndrome with paraplegia, sensory deficit and sphincter dysfunction. No visceral or large vessel lesions occurred. The surgical wound was evaluated for infection in three patients; all were satisfactory. Two cases of postoperative pneumothorax occurred, one of which required pleural drainage for 48 h. Other complications included pleural effusion, venous thrombosis and mood alteration related to corticotherapy in two patients. All patients, except the one that developed paraplegia, maintained their ability to walk during the entire follow-up period and until the late stages of the disease. One patient needed the left screws at T9T10 to be replaced 8 days after surgery and another patient had a broken L5 screw (with symptoms of local pain) that was replaced 11 years after the spondylectomy. Only four patients remain alive at this time after 15 and 13 years, and 8 and 4 months of follow-up. The rest died due to progression of the primary tumour after a minimum of 6.5 months up to 12 years (Table 1).

Results Between July 2000 and January 2016, 10 TES procedures were conducted on patients with primary or metastatic spine tumours. Clinical, anatomopathological, surgical and progression characteristics of the cohort are described in Table 1. The primary tumours included one osteosarcoma, two chondrosarcomas, one fibrosarcoma and one chordoma. Secondary lesions corresponded to metastasis of the lung, breast, thyroid and oesophagus. A rare case of spinal metastatic meningioma was also included that spread from an intracranial meningioma.23 A needle biopsy was only conducted in lesions that were thought to be primary (five cases). The histopathological study confirmed the clinical suspicion in three cases. The lesions mostly affected the thoracic segment. One of the patients presented preoperative paresis of the lower limbs. All patients presented local pain in the affected area with no significant radicular pain. According to the Tomita classification, two patients harboured type 3 intra-compartmental lesions, and the rest harboured extra-compartmental lesions

Discussion Total en bloc spondylectomy rationale and contribution Following the idea of anatomical barriers described by Enneking et al. for soft tissue spinal tumours,24 vertebrae can be considered a single oncological compartment.4 The anterior longitudinal ligament, vertebral discs and annulus fibrosus seem to be the most important barriers against tumour progression, while the posterior longitudinal ligament and the periosteum on the lateral sides of the vertebral body are weaker barriers, according to studies by Fujita et al.25 This anatomical distinction has implications on the way tumours tend to spread to surrounding tissues and it influences the chances of a radical resection.4 Tomita et al. devised a surgical classification of spinal metastasis based on the degree of spinal progression with implications for the excision strategy2,3 (Fig. 1). In primary malignant tumours and tumours that are benign but

Table 1 – Clinical characteristics and progression of the patients that underwent a total en bloc spondylectomy. Case

Age at diagnosis/sex

Primary tumour

Location

Preoperative state and Tomita classification

Surgical technique

Postoperative evolution

Morbidity and complications

Survival

Thyroid (follicular)

L1

ASIA E Type: 4

TES + fixation T11-L3 Wound debridement 40 days postop.

ASIA E Improvement in pain

2

11/female

Vertebral osteosarcoma

L4

ASIA E Type: 3

TES + fixation L3-L5 Revision of the fixation (L5 screw breakage) at 11 years

ASIA E Improvement in pain

3

37/female

Spinal chondrosarcoma

T4

ASIA D. Paraparesis 4/5. Hypoesthesia T6 level Type: 4

TES + fixation T2-T6 Reoperations due to tumour recurrence at 6, 11 and 12 years.

ASIA D Same paraparesis Improvement in pain

4

65/female

Spinal chordoma

T11

ASIA E Type: 5

TES + fixation T9-L1

ASIA E Improvement in pain

None

13 years and 3 months (alive)

5

56/female

Breast carcinoma

T4-T5

ASIA E. Hypoesthesia T7 Type: 6

TES + fixation T3-T7

ASIA E Improvement in pain

6.5 months (deceased)

6

37/male

Intracranial meningioma (extracranial metastasis)

T11

ASIA E Type: 5

TES + fixation T9-L1 Replacement of the right T9T10 screws 8 days postop.

ASIA E Improvement in pain

Mild apical pneumothorax (no pleural drainage) None

7

63/female

Spinal myxofibrosarcoma

T9T10

ASIA E Type: 6

TES + fixation T6-L1 Revision of the wound due to infection (Proteus mirabilis and Escherichia coli) 20 days postop.

ASIA A Paraplegia, anaesthesia and complete sphincter dysfunction No pain

Cord damage (presumably due to excessive traction during surgery) Wound infection

10 months (deceased)

Pneumothorax (pleural drainage 48 h) Wound dehiscence and infection (Pseudomonas aeruginosa) None Delayed malfunction of the fixation system needing surgery None

15 months (deceased)

Thyroidectomy + iodine131, 7 years before TES

15 years and 2 months (alive)

Lumbar lordosis deformity due to impaction (subsidence) of the interbody cage

12 years and one month (deceased)

Death due to abdominal surgery complications. Unable to walk after the 12th year of progress Postoperative radiotherapy. No recurrence at the moment Able to walk until her death

3 years and 9 months (deceased)

Able to walk until meningioma metastasis to the bone. Death due to multi-organ metastasis Cause of death unknown

ARTICLE IN PRESS

66/female

n e u r o c i r u g i a . 2 0 1 7;x x x(x x):xxx–xxx

1

Comments

5

6

Table 1 – (Continued) Case

Age at diagnosis/sex

Primary tumour

Location

Preoperative state and Tomita classification

Oesophageal cancer

T9

ASIA E Type: 3

9

40/male

Spinal chondrosarcoma (low grade G1)

T11

ASIA E Type: 4

54/female

Adenocarcinoma T7T8 of the lung

10

ASIA E Type: 6

TES + fixation T7-T11 Revision of the wound due to infection (Staphylococcus epidermidis) on day 19 postop. Neurophysiological monitoring Navigation guided by CAT TESm + fixation T9-L1 Neurophysiological monitoring Navigation guided by CAT TESm + fixation T5-T10 Neurophysiological monitoring Navigation guided by CAT

ASIA E Improvement in pain

ASIA E Mild proprioceptive deficit in lower limbs No pain ASIA E No pain

Morbidity and complications

Survival

Comments

Bilateral pleural effusion (no pleural drainage) Thrombosis of the superior vena cava and of the jugular vein Wound infection that required debridement surgery None

3 years and 2 months (deceased)

Oesophagus carcinoma that caused severe dysphagia Postoperative radio and chemotherapy Death due to pulmonary metastasis

8 months (alive)

Able to walk despite sensory deficit Needs analgesia occasionally due to dorsal pain

None

4 months (alive)

Received radiotherapy three days before TES due to pain Mood alteration induced by steroids which required sedatives

ASIA: American Spinal Injury Association; TES: total en bloc spondylectomy; TESm: modified total en bloc spondylectomy (modified according to the technique by Guo et al.6 ).

ARTICLE IN PRESS

53/male

Postoperative evolution

n e u r o c i r u g i a . 2 0 1 7;x x x(x x):xxx–xxx

8

Surgical technique

ARTICLE IN PRESS 7

n e u r o c i r u g i a . 2 0 1 7;x x x(x x):xxx–xxx

A

B Type 4 Type 3

D

C

Type 5

Type 6

Fig. 2 – (A) Type 3 spinal involvement. The lamina and articular process (slanting arrow) have been resected en bloc but the spinous process (arrow pointing down, only partially visible in the image) was resected in pieces (not affected by the tumour). (B) In type 4, epidural extension is occurring (white arrow) that protrudes towards the channel. C) A voluminous paravertebral extension is shown in this type 5 resection (black arrow). D) Type 6 implies spreading to the adjacent vertebrae; the unilateral spreading of one pedicle is visible here (black arrow).

aggressive, such as the giant-cell tumour, first time en bloc excision is preferred rather than an intralesional resection (or piecemeal resection), due to the possibility of relapse and the difficulty of salvage reoperations, due to fibrosis and adherence of the vertebrae to the vital surrounding tissues such as the aorta, vena cava and dura mater.2,4 As a general rule, Tomita et al. recommend the TES procedure in type 3–5 lesions. It is also relatively indicated in type 1, 2 and 6 lesions. Type 7 lesions (multiple-skip lesions) are not candidates for TES and should only receive palliative care.4 All of our patients were classified as types 3–5, or 6 (Table 1). Our type 6 cases required double or triple spondylectomies (Fig. 2). In general, TES is limited to non-spreading spinal malignant tumours, primary aggressive benign tumours and isolated spinal metastasis with long life expectancy.4,6 All of our cases belonged to these subgroups. We frequently find that many patients who were initially potential candidates for TES based on vertebral involvement, had to eventually be ruled out because the primary tumour had already spread at the time of diagnosing the vertebral lesion. A needle biopsy of the pedicle is generally recommended before planning a TES.4,6 This biopsy helps to differentiate between non-aggressive primary tumours and metastasis. In cases where a biopsy is performed, it is necessary to resect all tissues from the trajectory of the needle at the beginning of the TES, in order to minimise the risk of malignant cell dissemination.4 The decision to conduct

a biopsy is made on a discretionary basis and based on preoperative radiological findings. The biopsy provided diagnostic confirmation in only three of the five cases biopsied in the series. There is currently no doubt that en bloc resection of primary and secondary spinal tumours provides an increase in survival rates when compared to curettage or piecemeal resection.4,6 This fact is especially true for certain pathologies, such as thyroid metastasis, where expected local recurrence rates after debulking surgery and after TES are 57 and 10% respectively, with an average follow-up time of 55 months.8 Long-term clinical results of TES are also favourable in terms of survival and quality of life.26,27 Patients that underwent a TES preserved or improved their neurological state in most cases.4,6,28 The ability to walk was postoperatively preserved in at least 90% of cases,18 as we have also observed in our small series. Kato et al. have recently reported a 90% satisfaction rate ten years after the procedure, according to the SF-36 questionnaire among patients that underwent a TES.26 They also indicate a survival rate for renal and thyroid metastasis of 33 and 25%, respectively, 10 years after a TES.26 A recent study on quality of life referred by the patient after TES concluded that mental health status and ratings in social roles were equivalent to healthy individuals, although this was at the expense of a significant physical dysfunction that generally returned to normal three years after surgery.27

ARTICLE IN PRESS 8

n e u r o c i r u g i a . 2 0 1 7;x x x(x x):xxx–xxx

Table 2 – TES procedure: surgical steps and technical considerations (see text for a detailed explanation). TES surgical steps 1. Exposing the posterior arch and ribs.

2. Cutting the pedicle and extracting the posterior arch

3. Anterolateral dissection and cutting the discs

4. Unilateral fixation and en bloc resection of the vertebral body

5. Anterior reconstruction and fixation

Technical considerations The longitudinal cutaneous incision needs to be large enough to expose the affected vertebrae and at least two levels above and below them. The 3 cm next to the ribs in two or three levels of the affected vertebra must be dissected and removed, including the rib heads. Carefully dissect the pleura and the lateral sides of the vertebral body. The segmental arteries must be coagulated and ligated so that they are not damaged or pulled out of the aorta during the blunt dissection. A special guide is needed to pass the T-saw around the pedicle. On occasions, this manoeuvre is not possible since it can compress the spinal cord. The pedicle should be cut as horizontal as possible in order to not affect the lamina. The resection of the posterior arch can be done with the piecemeal method as long as there is no tumour involvement (in type 3 lesions) The dissection of the anterior side of the vertebral body of the main arteries is a dangerous manoeuvre that must be done very carefully with the fingertips of both hands. Massive bleeding could occur at this stage due to a lesion of the segmental arteries or veins that would need to be controlled. Once the spatulas are placed around the vertebral body, a safe space for cutting the discs is obtained. Cutting the discs in an oblique way from medial to lateral (using a thick lumbar puncture needle to introduce the T-saw) as proposed by Guo et al.6 is a safer alternative than the classic disc cut described by Tomita et al.1–4 One side must be fastened with screws and rods before completing the disc cutting. Sacrificing nerve roots (bilaterally?) is necessary to obtain sufficient direct access to conduct a blunt dissection and resection of the vertebral body, rotating around the spinal cord. Slight distraction may be applied during this manoeuvre The substitute graft of the anterior vertebral body must be placed with no distraction at all. The interbody cages can be filled with autogenous bone not affected by the tumour. Posterior fixation is completed. Rigid metallic pedicles may not be recommended from a biomechanical point of view. However, fixation with moderate compression is preferable to improve stability. The spinal cord is more likely to become irreversibly damaged after stretching than due to moderate shortening (around one or two thirds of the height of the body). In fact, moderate spinal shortening appears to increase spinal cord blood flow and hastens neurological recovery.

TES: total en bloc spondylectomy.

TES is primarily indicated instead of other techniques because it avoids leaving residual tumour tissue and, therefore, minimises the possibility of intraoperative contamination by tumour cells. Before the Tomita technique was invented, corpectomies and total spondylectomies were conducted using an anteroposterior or posteroanterior double approach. The tumours that presumably affect major vessels (types 5 and 6) can be removed using the TES method, although an anterior resection followed by posterior stabilisation in a posteroanterior combined approach is also a reasonable indication. On the other hand, lesions that affect levels L4–L5 could benefit from decompression and posterior fixation, prior to an anterior corpectomy in a posteroranterior combined approach.4 The main contribution of the TES is that the entire vertebra or vertebrae can be removed with a single posterior approach, thus reducing morbidity rates compared to combined approaches. This technique is usually used for one, two or three levels, although cases have been published of

up to four levels.29 This procedure is feasible for all thoracic and lumbar levels, keeping in mind that, below L1, the psoas muscle insertions in the vertebral bodies complicate the anterior dissection manoeuvres. For levels L4 and L5, a posteroanterior combined approach is probably the best option, as has been stated previously. This technique preserves lumbar nerves and eliminates local recurrence in two steps, as reported by Kawahara et al. in a study on ten patients who underwent combined approaches for malignant lesions at L4–L5, with an average follow-up period of 52 months.30 TES has also been described for cervical vertebrae, where preserving the nerve roots and the vertebral arteries is mandatory. Therefore, more than two osteotomies or discectomies are required.31 According to a study by Yokogawa et al. on 16 patients who underwent a TES of the upper thoracic region and of the T2 nerve roots, when extensive dissection of nerve roots T1 and above is conducted, 56% deterioration in upper limb motor function is expected, while the section

ARTICLE IN PRESS n e u r o c i r u g i a . 2 0 1 7;x x x(x x):xxx–xxx

9

Fig. 3 – (A) Pedicle cut with the T-saw and pulley system. (B) Disc cut once the dissection of the vertebral body is performed. (C) Vertebral body free of adherences once the discs have been cut and the dural sac has been dissected from the common posterior longitudinal ligament. (D) Careful rotation of the piece around the spinal cord. This manoeuvre is done using unilateral fixation and very slight distraction. (E) Coronal view of the piece dissected en bloc, both the posterior arch (left) and the body (right). (F) Axial view of the pieces dissected en bloc. (G) Reconstruction of the anterior spine using stackable PEEK cages filled with autogenous bone.

comprising only T2 nerve roots is a safe and acceptable manoeuvre.32 In a TES, both reconstruction of the vertebral body and stabilisation of the spinal column can be carried out with the same approach. The main advantage of this technique is that the spinal cord is directly visible and it can be monitored throughout the procedure, especially when manoeuvres are conducted that entail certain risk such as anterior discectomy, corpectomy and reconstruction of the spinal column. A notable feature of TES is the need to conduct a blunt dissection of the anterior part of the vertebral body using a posterior approach. This is done blindly with the tips of the fingers, using a bi-manual, bilateral and extrapleural manoeuvre.2–4

It is very important that this be done slowly and carefully in order to not damage the pleura, segmental arteries or main arteries. A certain degree of ability and experience is needed since any vascular complications that could arise at this point would be very worrying.

Technical aspects and surgical steps in total en bloc spondylectomy Table 2 summarises the primary technical considerations to keep in mind in a TES. It is especially important to control blood loss at every stage, since TES is always a prolonged surgery (more than 5–6 h). Blood must be available for an

ARTICLE IN PRESS 10

n e u r o c i r u g i a . 2 0 1 7;x x x(x x):xxx–xxx

eventual transfusion if it is needed during the procedure. The use of a T-saw is preferable to the conventional Gigli saw, due to its small diameter (0.5 mm).22 The T-saw is considered minimally invasive due to its smooth cutting surface and multifilament twisted stainless steel structure. The slice thickness is extremely thin, which reduces bleeding and possibly the spread of tumour cells.22 For a detailed description of the main risks of TES and possible solutions, the reader is referred to a thorough review of Tomita et al.4 To summarise, these authors consider it key to have five important questions in mind during the procedure: excessive bleeding (consider a preoperative embolisation, hypotensive anaesthesia and tamponade of the epidural space with fibrin glue); injury to the major vessels (take special care to avoid pulling out the segmental arteries from the aorta); spinal cord injury (by atraumatic handling of the spinal cord; avoid stretching it or using excessive nerve root traction); contamination by tumour cells (which is reduced by using the T-saw and, perhaps, by rinsing with distilled water and anticancer substances), and spinal stability (by reconstructing the anterior pillar, circumferential stabilisation and slight spinal shortening). The possibility of infection in the surgical site is always something that requires attention. Hayashi et al. have recently reported an infection rate of 6.4% in a retrospective study of 125 patients that underwent TES.33 The rate reached 12.5% after a conventional TES but was only 1.4% when implants soaked in iodine were used. In this study, the combined anteroposterior approach and not using said implants with iodine were identified as independent risk factors for infection after the TES.33 Prostaglandin E1 has also been recommend to prevent postoperative infection.34 In our series, three patients needed debridement due to an infected wound, which was satisfactorily resolved in all cases. For didactic purposes, we can subdivide the TES into five steps. Some of the technical characteristics and details of TES are described below in some detail. The details of a typical TES are summarised in Table 2 and Fig. 3 shows intraoperative images of the surgical sequence. In order to be able to conduct a total radical spondylectomy and minimise morbidity, each step needs to be performed consecutively and thoroughly as follows: 1. Exposing the posterior arch and ribs. The patient is put in prone position under general anaesthesia. Some authors recommend continuous intravenous infusion of methylprednisolone (5 mg/kg),6 although the use of corticosteroids in preventing spinal cord damage is still debated.35 We do not routinely use continuous infusion of glucocorticoids during surgery. We only administer a bolus of 8 mg of dexamethasone at the beginning of the procedure. Once the posterior arches are exposed, the ribs are divided bilaterally a few distal 3–4 cm from the costotransverse joints and they are removed, including the rib heads. Guo et al. recommend removing one rib at the lower thoracic levels and two at the upper thoracic levels.6 In all of our cases, we found it very difficult to conduct blunt dissection of the vertebral bodies unless two or three ribs were resected independently of the affected level, since

the free space for resection with the fingers is very narrow (Figs. 3D and 5A). We did not find any functional respiratory dysfunction or significant pain attributable to the resection of three levels of ribs. 2. Cutting the pedicles and extracting the posterior arch. To facilitate the introduction of the T-saw through the intervertebral foramen and around the pedicle, the inferior articular facet of the upper vertebra needs to be resected. Tomita et al. recommend introducing the saw in a craniocaudal direction.2–4 In our experience, the caudocranial view is also possible (Fig. 4A). In fact, in some cases, it is even easier to reach the saw, depending on how it emerges from the channel, when introduced caudocranially. The pedicles are cut to keep the saw as horizontal as possible and helped by an ingenious pulley system (Fig. 5B). In our experience and in contrast to previous papers,6 we did not find significant bleeding coming from the surface of the cut pedicles, even when they were affected by a tumour (Fig. 2D). At this moment of the procedure, a biopsy can be taken by curettage of the pedicle for a preliminary histopathological study. It is important to remove the entire vertebral posterior segment en bloc in order to maintain the radical resection concept. In cases where the preoperative MRI does not show tumour invasion of the posterior elements, such as in type 3, piecemeal resection is an acceptable option (Fig. 2A). 3. Anterolateral vertebra dissection and cutting the discs. The segmental vessels (artery and vein) are identified bilaterally and ligated or coagulated. These vessels are anatomical markers, since they emerge directly from the aorta. Preoperative embolisation of these vessels has been recommended, especially for hypervascular lesions such as kidney, thyroid or hepatocellular carcinoma metastasis.9,28,36 Given that interruption of the artery of Adamkiewicz during TES does not appear to affect neurological function, Tomita et al. advocate sacrificing up to three segmental artery sets, even if one these is the artery of Adamkiewicz.36 In fact, the study by Sugita et al. has demonstrated that in patients who later underwent another TES, due to metastasis at different levels, embolisation of more than three levels of segmental arteries is feasible and safe.37 The intercostal nerve root should also be ligated and cut. In the series by Yokogawa et al. on 72 TES, they reported a 23.6% rate of postoperative cerebrospinal fluid fistulas, especially if radiation had been given previously (fivefold higher risk).38 In our series, no fistulas occurred even though we only coagulated and cut the root. Tomita et al.2–4 and others6 recommend ligating only one root on the side on which the vertebral body is going to be rotated for its extraction. Again, we found that blunt dissection is quite complicated unless the roots are cut bilaterally. In our experience, we did not find evidence of radicular pain or significant anaesthesia in the postoperative period when two or even three roots were selected in the thoracic region. A blunt dissection lengthwise between the pleura and the vertebral body, from both sides, in a special bi-manual manoeuvre is conducted below (Fig. 5A), until the fingertips are in front of the anterior side of the vertebral body and the surgeon is able to place a pair of curved spatulas,

ARTICLE IN PRESS n e u r o c i r u g i a . 2 0 1 7;x x x(x x):xxx–xxx

11

Fig. 4 – (A) Insertion of the pedicle guide of the saw through the intervertebral foramen (broken arrow) and cranial exit under the lamina (black arrow). (B) The surgeon’s assistant holds the pulleys firmly which allows for the pedicle cuts to be made horizontally.

which overlap in front of the anterior side of the vertebral body, protecting surrounding vital structures. This manoeuvre must be done slowly in order to not damage the large vessels, especially the azygous vein or the vena cava. At this point, the disc is identified and through it, in an oblique manner and under the thecal sac (Fig. 5B), a lumbar puncture needle of sufficient thickness to be able to introduce a T-saw is inserted. Said needle is inserted until it emerges contralaterally and bumps against the spatula, according to the modification of the technique introduced by Guo et al.6 The saw is inserted, the needle is withdrawn and the end of the saw meets the side it entered. The disc is then cut with a rocking movement, far from the thecal sac. This is done bilaterally (both proximally and distally)

in order for only one triangular portion of the disc to be left uncut under the sac. 4. Unilateral fixation and en bloc resection of the vertebral body. Just before conducting the contralateral disc cuts, one side must be fastened with a rod; the side opposite to the one chosen in order to rotate the vertebral body. The side does not matter, unless a voluminous paravertebral affectation exists on one of the sides (Fig. 2C). In these cases, it was easier to rotate the vertebra towards the more protruding or voluminous side. The rest of the disc is cut at this point under the dural sac. The sac must be dissected from the common posterior longitudinal ligament to free any adhesion to it. The use of fibrin glue has been recommended since the epidural space may bleed profusely.4,6 In our experience, this

Fig. 5 – (A) Bi-lateral and bi-manual dissection manoeuvre. The fingertips of both hands must be on the anterior side of the vertebral body and they must create enough room to introduce the protective spatulas. (B) The lumbar puncture needle is inserted in an oblique manner, just under the dural sac, towards the opposite side. The saw is inserted through the needle and is extracted from the same side in order to partially cut the disc. This is done bilaterally, on both the proximal and distal discs. The thecal sac is always protected from direct trauma since the saw is never directed towards the spinal cord.6

ARTICLE IN PRESS 12

n e u r o c i r u g i a . 2 0 1 7;x x x(x x):xxx–xxx

Fig. 6 – (A) Anterior reconstruction with stackable PEEK cages filled with autogenous bone obtained from unaffected ribs. The size of the cage must not be greater than the space available in order to not stretch the spinal cord. (B) The artificial pedicles (black star) join the anterior cages to the posterior fixation system. From a biomechanical point of view, this construction does not appear to offer any benefits compared to fixation with moderate compression.

bleeding can be controlled effectively with bipolar coagulation. A minimal distraction of the fixation system may be applied to help the already freed vertebral body to rotate around the thecal sac and be extracted en bloc (Fig. 3D). 5. Anterior reconstruction and fixation The distraction is stopped quickly and a cage or mash cylinder is placed as a vertebral body substitute. The size must be the exact height so the spinal cord is not stretched at all. It can be filled with autogenous bone obtained from unaffected ribs (Fig. 6A). Allografts can also be used as substitutes, although cases of contamination and morbidity due to their use have been reported, raising suspicions regarding their safety.17 The contralateral rod is inserted and moderate compression is applied (some 10 mm). Tomita et al.1–4 recommend shortening the spine around one or two thirds of the height of the vertebral body. This provides better stability and increases spinal cord blood flow, as proven in animal studies39,40 and clinical studies.28,41 A study on biomechanics by Kato et al. demonstrated that reconstruction with cages that are 10 mm shorter than the space available provides a more rigid construction compared to when larger cages are used, independent of the cross-link number that is used (one or two).42 The study by Murakami et al. on 79 patients with TES concluded that the increase in spinal cord blood flow, due to shortening the spine and added to decompression, was responsible for the resolution or improvement of neurological symptoms after TES.28 The study by Yoshioka et al. on spinal reconstructions after TES at three or more levels, in 26 patients showed a graft impaction rate (subsidence) of 50%, visible one month after surgery.41 They also concluded that shortening the spine (an average of 10.4 mm through axial compression of the system) increased stability and reduced thee subsequent failure rate to 3.8%.41

(four cases) and carbon and stackable polyether ether ketone (PEEK) cages were used to provide the same stability. Although rigid artificial metallic pedicles (pieces of titanium that join the vertebral body with the posterior instrumentation) were used in two patients (Fig. 6B), they have subsequently been discontinued since they do not seem to provide any benefit from a biomechanical standpoint compared with fixation with moderate compression. There is also the theoretical risk that these pedicles could transmit tension or instability to the vertebral body graft in the event of subsequent instrumentation failure. Matsumoto et al. reported a late subsequent failure rate of 40% after TES.43 Predisposing factors they found were the impact of the cage (>5 mm), preoperative irradiation and the number of instrumented vertebrae (more than four).43 Reconstructing the posterior elements (laminas) with titanium mesh has recently been proposed, with the goal of enhancing stability and protecting neural elements.44 It is generally recommended to leave a suction drain at the surgical site. We use fibrin glue in a spray to facilitate haemostasis, as recommended.6 It was also useful to protect the pleura with absorbent patches or haemostatic agents such as Surgicel (Johnson & Johnson, USA) since, due to manipulation, it is not uncommon to find pleural fissures and tears at the end of the procedure. In the postoperative period, pleural thorax tubes were only required in one patient. In our opinion, if the vacuum is maintained constantly during drainage, the onset of pneumothorax is minimised. Using the radioscopy arch or preferably the O-arm (Medtronic, Minneapolis, USA), the final placement of the fixation system and the anterior graft is verified before leaving the operating theatre.

In most cases, pedicle screws were inserted two levels above and below the lesion. In the case where a three-level TES was conducted, the screws were placed three levels above and below (Fig. 3), and only on one level in the girl with L4 osteosarcoma. This may have been the cause why the patient needed subsequent revision surgery. Titanium mesh cylinders

Our preliminary experience with TES allows us to highlight a series of important issues to keep in mind:

Lessons learned

1. Patient selection and prior irradiation. Selecting a good candidate for TES requires adequate preoperative imaging tests and a certain consensus between the oncologist and

ARTICLE IN PRESS n e u r o c i r u g i a . 2 0 1 7;x x x(x x):xxx–xxx

13

Fig. 7 – (A) D-Wave cable (black arrow) inserted in the epidural space in caudal direction to the affected vertebra. It continuously monitors the function of the spinal cord during the entire procedure. (B) Normal D-Waves visible on the neurophysiologist’s monitor during the procedure.

the spinal surgeon regarding the presumably long lifeexpectancy of the patient. Prior irradiation of the vertebra is a relative contraindication for TES, since radiotherapy induces the formation of scar tissue and adherences that impede a safe dissection of the vertebra, viscera and large vessels. In addition, preoperative radiotherapy increases the risk of dural lesion, liquid fistula, wound dehiscence and pleural effusion by a factor of 3.34 The safest time to conduct the TES appears to be between three and six weeks after radiotherapy.34 However, in our experience, radiotherapy administered immediately before surgery does not appear to make the blunt dissection of the vertebra more difficult, as observed in case 10. 2. Bleeding control. The bi-manual dissection of the anterior side of the vertebra must be done smoothly and carefully. Accelerating this manoeuvre can lead to uncontrollable bleeding due to avulsion of the intercostal arteries or, more frequently, due to vein rupture. The long duration of TES means that attention must be paid to haemostasis at all times. Oxygenated water was useful in controlling mild diffuse bleeding. 3. Spinal cord function. The best method to prevent permanent spinal cord damage is to carefully manage the instrumentation around the spinal cord and to conduct the procedure according to the specified steps. Neurophysiological monitoring is also useful; continuous monitoring of the D-Wave ensures the functioning of the spinal cord during the entire procedure (Fig. 7). Somatosensory and motor evoked potentials are also helpful. We believe that neurophysiological monitoring is currently a basic requirement when conducting a TES, especially when unintentional trauma of the spinal cord occurs while being manipulated. 4. Spinal navigation. Intraoperative navigation helps to identify the affected level, especially in dorsal spondylectomies

where the disc space may be very narrow or collapsed. It is also useful for placing screws. Navigation was guided by the O-arm (Medtronic, Minneapolis, USA) in the last three cases (Fig. 8). In our opinion, spinal navigation increases the precision of the screw placement and does not increase the duration of the procedure. 5. Modified Tomita technique. In our view, the modification introduced by Guo et al.,6 by cutting the discs obliquely far from the spinal cord, is a significant improvement on the classic Tomita technique in terms of neurological safety. This modification was adopted in the last two cases. The original technique by Tomita et al.1–4 involves cutting the discs in the direction of the spinal cord, protecting it with a special spatula under the sac. The surgeon’s assistant must keep the spatula steady while the saw moves closer to the common posterior ligament to prevent a direct trauma with the saw against the spinal cord. This is completely omitted with the referenced modification. Another modification proposed by Huang et al. also aims to increase safety during disc cutting.7 6. Length of the spinal cord. A presumably excessive traction on the spinal cord (although not seen during the operation) may have been the cause of full and permanent spinal cord deficit that occurred in one of our patients (case 7). Special care must be taken when rotating and extracting the anterior bloc, with as few distractions as possible. A circumferential fixation with moderate compression is always recommended (with a certain shortening of the spine) to improve spinal stability and enhance spinal cord blood flow. 7. Surgeon’s experience. Given that there is an obvious learning curve for TES, the surgeon must have a certain amount of previous experience in complex spinal surgery4 in order to face a TES with confidence. In our opinion, the personal experience of the surgeon can be a key factor in how

ARTICLE IN PRESS 14

n e u r o c i r u g i a . 2 0 1 7;x x x(x x):xxx–xxx

Fig. 8 – (A) Navigation-assisted placement of screws. The CT intraoperative images provide spatial information in real time for spinal navigation. (B) Intraoperative verification of the fixation by CT once the circumferential stabilisation is completed.

it is conducted and the consequences of the procedure. Technological aids such as neurophysiological monitoring and CT guided navigation are important tools for the surgeon. However, they cannot substitute the necessary surgical skills. Lastly, this study presents the same limitations inherent to any retrospective study of a series of cases. The most important is the modest number of patients. Although conducting TES on 10 patients has given us a certain degree of expertise (enabling tricks and failures to be identified), it is not enough to statistically evaluate its efficiency in terms of progression-free survival or overall survival. In addition, some technical aids such as intraoperative neuromonitoring and CT guided spinal navigation were only available in the last cases, impeding any comparative analysis. It is believed that these technical aids, together with the experience gained, has changed our view of TES from a risky and difficult procedure to one that is more standardised and safe.

Conclusion In this series of cases various technical aspects related to the Tomita technique were described. It is believed that certain considerations of interest must be observed for the procedure to be satisfactorily conducted. They include: a careful selection of candidates, thorough vertebral dissection, strict bleeding control, careful handling of the spinal cord and maintaining the radical resection concept at all stages. Our preliminary experience allows us to rule out some relevant and specific characteristics of TES, especially those that

favour a safer and simpler procedure. Keeping in mind that the first radical oncological surgery is the best option for healing,4 TES is a paradigmatic operation where some advanced surgical abilities provide functional efficacy and improve the survival of patients harbouring spinal tumours.

Conflicts of interest The authors declare that there are no conflicts of interest.

Acknowledgements A M. Rodríguez Miguélez, MD, and Gwyn Roberts for their careful review of the English version of this study.

references

1. Tomita K, Toribatake Y, Kawahara N, Ohnari H, Kose H. Total en bloc spondylectomy and circumspinal decompression for solitary spinal metastasis. Paraplegia. 1994;32:36–46. 2. Tomita K, Kawahara N, Baba H, Tsuchiya H, Fujita T, Toribatake Y. Total en bloc spondylectomy. A new surgical technique for primary malignant vertebral tumors. Spine. 1997;22:324–33. 3. Tomita K, Kawahara N, Baba H, Tsuchiya H, Nagata S, Toribatake Y. Total en bloc spondylectomy for solitary spinal metastases. Int Orthop. 1994;18:291–8. 4. Tomita K, Kawahara N, Murakami H, Demura S. Total en bloc spondylectomy for spinal tumors: improvement of the technique and its associated basic background. J Orthop Sci. 2006;11:3–12.

ARTICLE IN PRESS n e u r o c i r u g i a . 2 0 1 7;x x x(x x):xxx–xxx

5. Tomita K, Kawahara N, Kobayashi T, Yoshida A, Murakami H, Akamaru T. Surgical strategy for spinal metastases. Spine. 2001;26:298–306. 6. Guo C, Yan Z, Zhang J, Jiang C, Dong J, Jiang X, et al. Modified total en bloc spondylectomy in thoracic vertebra tumour. Eur Spine J. 2011;20:655–60. 7. Huang L, Chen K, Ye JC, Tang Y, Yang R, Wang P, et al. Modified total en bloc spondylectomy for thoracolumbar spinal tumors via a single posterior approach. Eur Spine J. 2013;22:556–64. 8. Demura S, Kawahara N, Murakami H, Abdel-Wanis ME, Kato S, Yoshioka K, et al. Total en bloc spondylectomy for spinal metastases in thyroid carcinoma. J Neurosurg Spine. 2011;14:172–6. 9. Salame K, Regev G, Keynan O, Lidar Z. Total en bloc spondylectomy for vertebral tumors. Isr Med Assoc J. 2015;17:37–41. 10. Boriani S, Chevalley F, Weinstein JN, Biagini R, Campanacci L, de Iure F, et al. Chordoma of the spine above the sacrum. Treatment and outcome in 21 cases. Spine. 1996;21:1569–77. 11. Boriani S, Biagini R, de Iure F, di Fiore M, Gamberini G, Zanoni A. Lumbar vertebrectomy for the treatment of bone tumors: surgical technique. Chir Organ Mov. 1994;79:163–73. 12. Sundaresan N, Rosen G, Huvos AG, Krol G. Combined treatment of osteosarcoma of the spine. Neurosurgery. 1988;23:714–9. 13. Stener B. Complete removal of vertebrae for extirpation of tumors. A 20-year experience. Clin Orthop Relat Res. 1989;245:72–82. 14. Roy-Camille R, Saillant G, Bisserié M, Judet T, Hautefort E, Mamoudy P. Total excision of thoracic vertebrae. Rev Chir Orthop Reparatrice Appar Mot. 1981;67:421–30. 15. Fang T, Dong J, Zhou X, McGuire RA Jr, Li X. Comparison of mini-open anterior corpectomy and posterior total en bloc spondylectomy for solitary metastases of the thoracolumbar spine. J Neurosurg Spine. 2012;17:271–9. 16. Mesfin A, El Dafrawy MH, Jain A, Hassanzadeh H, Kebaish KM. Total en bloc spondylectomy for primary and metastatic spine tumors. Orthopedics. 2015;38:e995–1000. 17. Neves RP, Oliveira VC, Costa LM, Soares DF, Cardoso PF, Costa PG, et al. Major complications following total en bloc spondylectomy for giant-cell tumor. J Surg Case Rep. 2014;2014, http://dx.doi.org/10.1093/jscr/rjt131. 18. Luzzati AD, Shah S, Gagliano F, Perrucchini G, Scotto G, Alloisio M. Multilevel en bloc spondylectomy for tumors of the thoracic and lumbar spine is challenging but rewarding. Clin Orthop Relat Res. 2015;473:858–67. 19. Yamazaki T, McLoughlin GS, Patel S, Rhines LD, Fourney DR. Feasibility and safety of en bloc resection for primary spine tumors: a systematic review by the Spine Oncology Study Group. Spine. 2009;34 Suppl.:S31–8. 20. Boriani S, Weinstein JN, Biagini R. Primary bone tumors of the spine. Terminology and surgical staging. Spine. 1997;22:1036–44. 21. Hsieh PC, Li KW, Sciubba DM, Suk I, Wolinsky JP, Gokaslan ZL. Posterior-only approach for total en bloc spondylectomy for malignant primary spinal neoplasms: anatomic considerations and operative nuances. Neurosurgery. 2009;65 Suppl.:173–81, discussion 181. 22. Tomita K, Kawahara N. The threadwire saw: a new device for cutting bone. J Bone Joint Surg Am. 1996;78:1915–7. 23. Delgado-López PD, Martín-Velasco V, Castilla-Díez JM, Fernández-Arconada O, Corrales-García EM, Galacho-Harnero A, et al. Metastatic meningioma to the eleventh dorsal vertebral body: total en bloc spondylectomy. Case report and review of the literature. Neurocirugia (Astur). 2006;17:240–9. 24. Enneking WF, Spanier SS, Goodman MA. A system for the surgical staging of musculoskeletal sarcoma. Clin Orthop Relat Res. 1980;153:106–20.

15

25. Fujita T, Ueda Y, Kawahara N, Baba H, Tomita K. Local spread of metastatic vertebral tumors. A histologic study. Spine. 1997;22:1905–12. 26. Kato S, Murakami H, Demura S, Yoshioka K, Kawahara N, Tomita K, et al. More than 10-year follow-up after total en bloc spondylectomy for spinal tumors. Ann Surg Oncol. 2014;21:1330–6. 27. Kato S, Murakami H, Demura S, Yoshioka K, Kawahara N, Tomita K, et al. Patient-reported outcome and quality of life after total en bloc spondylectomy for a primary spinal tumour. Bone Joint J. 2014;96-B:1693–8. 28. Murakami H, Kawahara N, Demura S, Kato S, Yoshioka K, Tomita K. Neurological function after total en bloc spondylectomy for thoracic spinal tumors. J Neurosurg Spine. 2010;12:253–6. 29. Druschel C, Disch AC, Melcher I, Engelhardt T, Luzzati A, Haas NP, et al. Surgical management of recurrent thoracolumbar spinal sarcoma with 4-level total en bloc spondylectomy: description of technique and report of two cases. Eur Spine J. 2012;21:1–9. 30. Kawahara N, Tomita K, Murakami H, Demura S, Yoshioka K, Kato S. Total en bloc spondylectomy of the lower lumbar spine: a surgical techniques of combined posterior–anterior approach. Spine. 2011;36:74–82. ˇ ˇ T, Nesnídal P, Kryl J, Klézl Z. Total 31. Stulík J, Barna M, Vyskocil en bloc spondylectomy of C3: a new surgical technique and literature review. Acta Chir Orthop Traumatol Cech. 2015;82:261–7. 32. Yokogawa N, Murakami H, Demura S, Kato S, Yoshioka K, Hayashi H, et al. Motor function of the upper-extremity after transection of the second thoracic nerve root during total en bloc spondylectomy. PLoS ONE. 2014;9:e109838. 33. Hayashi H, Murakami H, Demura S, Kato S, Yoshioka K, Shinmura K, et al. Surgical site infection after total en bloc spondylectomy: risk factors and the preventive new technology. Spine J. 2015;15:132–7. 34. Yokogawa N, Murakami H, Demura S, Kato S, Yoshioka K, Hayashi H, et al. Perioperative complications of total en bloc spondylectomy: adverse effects of preoperative irradiation. PLoS ONE. 2014;9:e98797. 35. Ahn H, Fehlings MG. Prevention, identification, and treatment of perioperative spinal cord injury. Neurosurg Focus. 2008;25:E15. 36. Murakami H, Kawahara N, Tomita K, Demura S, Kato S, Yoshioka K. Does interruption of the artery of Adamkiewicz during total en bloc spondylectomy affect neurologic function? Spine. 2010;35:E1187–92. 37. Sugita S, Murakami H, Demura S, Kato S, Yoshioka K, Yokogawa N, et al. Repeated total en bloc spondylectomy for spinal metastases at different sites in one patient. Eur Spine J. 2015;24:2196–200. 38. Yokogawa N, Murakami H, Demura S, Kato S, Yoshioka K, Hayashi H, et al. Postoperative cerebrospinal fluid leakage associated with total en bloc spondylectomy. Orthopedics. 2015;38:e561–6. 39. Kawahara N, Tomita K, Kobayashi T, Abdel-Wanis ME, Murakami H, Akamaru T. Influence of acute shortening on the spinal cord: an experimental study. Spine. 2005;30:613–20. 40. Carlson GD, Warden KE, Barbeau JM, Bahniuk E, Kutina-Nelson KL, Biro CL, et al. Viscoelastic relaxation and regional blood flow response to spinal cord compression and decompression. Spine. 1997;22:1285–91. 41. Yoshioka K, Murakami H, Demura S, Kato S, Kawahara N, Tomita K, et al. Clinical outcome of spinal reconstruction after total en bloc spondylectomy at 3 or more levels. Spine. 2013;38:E1511–6. 42. Kato S, Murakami H, Higashino K, Okada M, Ito Z, Demura S, et al. The effect of spinal shortening after total en bloc

ARTICLE IN PRESS 16

n e u r o c i r u g i a . 2 0 1 7;x x x(x x):xxx–xxx

spondylectomy: a biomechanical study in the thoracic spine. J Spinal Disord Tech. 2012;25:E183–90. 43. Matsumoto M, Watanabe K, Tsuji T, Ishii K, Nakamura M, Chiba K, et al. Late instrumentation failure after total en bloc spondylectomy. J Neurosurg Spine. 2011;15:320–7.

44. Chung JY, Kim SK, Jung ST, Lee KB. New posterior column reconstruction using titanium lamina mesh after total en bloc spondylectomy of spinal tumour. Int Orthop. 2013;37: 469–76.