J Robotic Surg (2015) 9:331–338 DOI 10.1007/s11701-015-0536-x

ORIGINAL ARTICLE

Accuracy of thoracolumbar transpedicular and vertebral body percutaneous screw placement: coupling the RosaÒ Spine robot with intraoperative flat-panel CT guidance—a cadaver study M. Lefranc1,2,3 • J. Peltier1,2

Received: 9 August 2015 / Accepted: 27 September 2015 / Published online: 22 October 2015 Ó Springer-Verlag London 2015

Abstract The primary objective of the present study was to evaluate the accuracy of a new robotic device when coupled with intraoperative flat-panel CT guidance. Screws (D8-S1) were implanted during two separate cadaver sessions by coupling the RosaÒ Spine robot with the flat-panel CT device. Of 38 implanted screws, 37 (97.4 %) were fully contained within the pedicle. One screw breached the lateral cortical of one pedicle by \1 mm. The mean ± SD accuracy (relative to pre-operative planning) was 2.05 ± 1.2 mm for the screw head, 1.65 ± 1.11 for the middle of the pedicle and 1.57 ± 1.01 for the screw tip. When coupled with intraoperative flat-panel CT guidance, the RosaÒ Spine robot appears to be accurate in placing pedicle screws within both pedicles and the vertebral body. Large clinical studies are mandatory to confirm this preliminary cadaveric report. Keywords

Robot
Pedicle screw placement
Accuracy

is now a widely used and accepted technique in thoracic and lumbar spine procedures [1–5]. The advantages of percutaneous dorsal pedicle screw instrumentation are clear: significantly less blood loss, a lower postoperative infection rate and faster recovery (due to minimal injury to the tissues around the spine). Accurate pedicle screw placement is mandatory during these procedures [6]. A high level of accuracy can be obtained with intraoperative double fluoroscopic guidance (albeit at the cost of high exposure to ionizing radiation, in most cases) [6– 8]. In order to provide high levels of accuracy and limit exposure to radiation, robotic devices and intraoperative image-guided navigation have been developed [9, 10]. In the present cadaver study of percutaneous transpedicular thoracolumbar screw placement, we evaluated the accuracy of a new robotic device (the RosaÒ Spine from Medtech Surgical, Montpellier, France) coupled with intraoperative flat-panel CT guidance (the O’ArmÒ from Medtronic, Minneapolis, MN, USA).

Introduction

Materials and methods

Over the last two decades, there has been a notable surgical trend towards minimizing injury to healthy tissues while nevertheless obtaining an acceptable or even excellent technical outcome. Percutaneous transpedicular instrumentation

Two cadaver specimens were used in two separate sessions. All procedures were performed by the same spine surgeon.

& M. Lefranc [email protected] 1

Department of Neurosurgery, Amiens University Hospital, Avenue Laennec, 80054 Amiens, France

2

Lae¨nnec Saloue¨l, 80054 Amiens Cedex 1, France

3

Neurosurgery Department, CHU Amiens Picardie, Salouel, France

Operating technique The procedures were performed in order to mimic the percutaneous procedure in patients as closely as possible. In the operating theatre, the cadaver was placed in the prone position on a radiotransparent spinal operating table. All pressure points were padded as appropriate. The thoracolumbar spine was prepared and draped under sterile conditions. A percutaneous reference pin was placed in the right iliac wing or attached on spinous process. The

123

332

O’armÒ device was placed to the left of the surgeon, and the robot faced the surgeon. Three-dimensional (3D) CT images from the O-armÒ were transferred to the RosaÒ Spine adjunct surgical workstation (MedtechÒ, Montpellier, France). Registration was performed by the automatic recognition of a ‘‘fiducial box’’ held by the robotic arm. The three-dimensional planning of bilateral transpedicular screw placement ranged from T8 to S1 in one specimen and from T10 to S1 in the other was performed. All sizes of screw and all kinds of trajectory can be taken into account in this procedure, which enables the surgeon to plan the screw placement exactly as wished. Specifically, a transpedicular hole was drilled (using a 3 mm bit) with real-time robotized navigation guidance (i.e. the robot is able to monitor and follow the movements of the patient’s body in real time). A guide-tube needle was placed through the skin and through the pedicle into the posterior part of the vertebral body. After a guide wire had been placed through the guide-tube needle, the latter was removed. The guide-tube needle, guide wire and all instruments were monitored by the robot (i.e. via real-time computer-aided navigation), which provided the instruments’ exact spatial positions throughout the surgery. The procedure used the LongitudeÒ percutaneous ancillary system (MedtronicÒ). Dilators were placed via the guide wire, and the screws were then inserted via the guide wire with real-time robotic guidance. The size of each screw was selected on the basis of pedicle size measurements from the initial 3D planning. Following placement of the pedicle screws, another CT scan was acquired with the O-armÒ. The pre-operative and post-operative CT datasets were co-registered, in order to measure the difference (in mm) between the planned and actual screw positions. Two surgeons assessed the accuracy of pedicle screw placement on the final CT images. Figure 1 illustrates the operating technique. Our evaluation of the accuracy of pedicle screw placement was based on (1) postoperative O’ArmÒ scans; (2) the scoring system described by Zdichavsky and used by Heintel et al. and (3) the scoring system described by Ravi et al. [11, 12]. On Zdichavsky’s scale, the best positioning options are those rated as 1a and 1b. Positions rated as 2a or 2b must be evaluated for stability, but do not usually require revision. Positions rated as 3a and 3b are bad (poor stability or possible nerve irritation) and so revision must be considered. The Ravi scale is based on the presence of pedicle breach (I: no breach, II: breach \2 mm, III: breach of 2–4 mm), IV: breach [4 mm). The direction of pedicle breach (lateral–medial, superior or inferior) and any breaches of the vertebral body were also recorded. The accuracy of positioning (in mm) of three different parts of the screw (the head, the segment in the middle of the pedicle, and the tip) was also evaluated and compared with the pre-operative plan (Fig. 2).

123

J Robotic Surg (2015) 9:331–338

Statistics All statistical analyses were performed with R software. A Mann–Whitney test (two-sided, when n \ 20) and a T test (two-sided, when n [ 20 in each group and the data were normally distributed) were used to compare the two cadaver sessions and the accuracy of positioning for the different parts of the screw. The threshold for statistical significance was set to p \ 0.05.

Results Accuracy of pedicle screw placement A total of 38 screws were inserted, and 37 (97.2 %) did not breach the pedicle. During the second session, one screw caused a lateral cortical pedicle breach of less than 1 mm (grade II on Ravi’s scale and grade 1b on Zdichavsky’s scale). However, the screw’s position appeared to be biomechanically effective and would probably not have induced nerve irritation in a patient. There were no vertebral body breaches with any of the 38 screws. All screws were placed in relevant locations and did not require revision, as judged by Zdichavsky’s classification. Figure 3 shows a typical screw position within the vertebral body (a) and the pedicle (b), and the position that led to a breach (c). The positional data are summarized in Fig. 4. Agreement between planned and actual screw positions The mean ± SD (range) accuracy (relative to the pre-operative planning) was 2.05 ± 1.2 mm (0.12–5.12) for the screw head, 1.65 ± 1.11 mm (0.16–3.94) for the middle of the pedicle and 1.57 ± 1.01 mm (0.18–3.8) for the screw tip. Figure 5 summarizes the accuracy of screw placement. There were no significant differences in accuracy when comparing the screw head and the middle of the pedicle (p = 0.2), the screw head and the screw tip (p = 0.127) or the middle of the pedicle and the screw tip (p = 0.87). Likewise, there were no significant differences in the accuracy of screw placement when comparing thoracic and lumbar sites [1.88 ± 0.8 (0.79–3.47) and 2.41 ± 1.24 (0.35–5.12), respectively; p = 0.015]. These data are summarized in Fig. 5.

Discussion In the present study, we reported on the accuracy of percutaneous pedicle screw placement by a new robotic device designed for coupling with intraoperative CT.

J Robotic Surg (2015) 9:331–338

333

Fig. 1 The operating technique: the flat-panel CT is placed to the surgeon’s left, with the robot facing the surgeon and the camera on the right. The percutaneous ancillary was not modified and was introduced via a dedicated reducer held by the robotic arm

Minimally invasive surgery (MIS) has proven to be a safe, effective alternative to traditional open spinal surgery [13–16]. Percutaneous transpedicular osteosynthesis is now a widely accepted technique in MIS in general and in degenerative spine surgery [1, 14] and traumatic spine surgery [4, 17, 18, 11] in particular. Accurate pedicle screw placement is mandatory during these procedures. Pedicle breaches can potentially cause haemorrhage, nerve root injury, spinal cord injury and low biomechanical strength of the operated region [6]. Furthermore, optimal biomechanical performance requires the tip of the screw to be located in the midportion of the vertebral body (near the anterior cortex). However, excessively long pedicle screws are associated with a life-threatening risk of damage to adjacent structures such as the aorta and the pleural cavity [11].

Here, we demonstrated a high level of accuracy for pedicle screw placement throughout the lumbar and thoracolumbar regions of the spine. The intraoperative use of image-guided devices may provide surgeons with a safer, reliable, accurate method for placing thoracic pedicle screws while limiting exposure to ionizing radiation. Our present data compare favourably with all the other percutaneous methods currently in use. Based on Gelalis et al.’s review, the proportion of screws that perforate the cortical margins of the pedicle ranges from 6 to 31 % with freehand techniques, from 81 to 92 % with fluoroscopic navigation and from 89 to 100 % with flat-panel CT navigation [6]. In another review of the accuracy of pedicle screw placement, Mason et al. determined an accuracy rate of 68.1 % for conventional fluoroscopy, 84.3 % with two-dimensional fluoroscopic navigation, and

123

334

J Robotic Surg (2015) 9:331–338

Fig. 2 The method used to evaluate the accuracy of screw positioning: the distance between screw’s position and the planned trajectory was measured for the screw head, the middle of the pedicle and the screw tip

95.5 % with 3D fluoroscopic navigation (i.e. coupled with flat-panel CT). With a strict percutaneous approach, a very high accuracy rate can be obtained with intraoperative double fluoroscopic guidance (98 % in Heintel et al.’s study [11]), albeit usually at the cost of high radiation exposure [8] and after a non-negligible learning period. In order to provide high level of accuracy and to limit the extent to which the patient, surgeons and operating theatre staff are exposed to X-rays during percutaneous osteosynthesis, a number of appropriate robotic devices and intraoperative image-guided navigation techniques have been developed [19, 20]. The high level of accuracy and easy workflow associated with intraoperative CT image-guided navigation for pedicle screw instrumentation has increased the popularity of this approach [16, 19, 21, 22]. However, intraoperative CT image-guided navigation lacks real-time follow-up of body movement induced by surgery. This limitation makes it hard to place the screw in accordance with pre-operative planning and (in our opinion) is the prime cause of screw misplacement and facet joint violation (affecting up to 18 % of cases in Yson et al. series [23]). However, the value of robotic devices in spine surgery has not been extensively determined [24–26]. Most literature reports concern the use of the Spine-assistÒ robotic device (Mazor Robotics, Israel) for pedicle screw

123

placement [2, 20, 27–30]. The use of a robot does appear to reduce radiation exposure and to improve accuracy of screw positioning [29–33]. However, accuracy of the screws placement using Spine-assistÒ varies in relatively high range in literature and the proportion of pedicle breaches can reach 6 to 20 % in some reports [31, 32, 34– 36]. Here, we present a new device which has been designed to couple the advantages of CT image-guided navigation with robotic guidance. The device is a development of the RosaÒ Brain robot used in brain surgery [37, 38]. The robot’s ability to track movement of the vertebrae in real time—as the Cyberknife does in radiosurgery [39]—is to improve accuracy of screw placement not only in regard of the pedicle, but also all along the trajectory planned. The accuracy levels (for all parts of the screw) are similar to those obtained in stereotactic brain surgery [40]. The data presented here compare favourably with the results of other cadaver studies of robot-assisted surgery [41]. One screw breached the lateral cortical of one pedicle by less than 1 mm, but all the other screws were classified as grade 1. In our opinion, even if there are a relatively low number of screws implanted in this study, our ability to compare the planning and the actual position of the screw clearly demonstrates the ability of this robotic device to

J Robotic Surg (2015) 9:331–338

335

Fig. 3 Typical placement of the screws within the vertebral body and the pedicle (a). Only one screw (b) breached the pedicle (classified as grade II on Ravi’s scale and grade 1b on Zdichavsky’s scale)

accurately place screws within the vertebra with a 2 mm accuracy range. The accuracy obtained in the present study might result in a lower risk of neurotoxicity after screw placement. Improved accuracy for all parts of the screw might also increase the mechanical advantages of

pedicle screw fixation and decrease the risk of facet joint violation via adjacent segment degeneration. Finally, our use of only 2 cadavers and 48 screws constitutes a clear study limitation and our present data must be confirmed in the clinic with large series.

123

336

J Robotic Surg (2015) 9:331–338

Fig. 4 The accuracy of pedicle screw placement with the RosaÒ Spine robot (on Ravi’s scale). Only one breach was noted (grade II)

Conclusion When coupled with intraoperative flat-panel CT, the RosaÒ Spine robot appears to be highly accurate in pedicle screw placement with regard to both the pedicle and the vertebral body. This level of accuracy should lead to greater safety and biomechanical strength when the technique is applied in patients. These preliminary data must be confirmed by large clinical studies. Compliance with ethical standards Conflict of interest The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper. Dr Lefranc has received a speaker honorarium from Medtech (Montpellier, France). Ethical approval Not available (cadaver study—patients consent to give their bodies to science after death to the ‘‘Laboratoire d’anatomie et d’organogene`se’’—UPJV UFR medecine Amiens).

Fig. 5 The accuracy of pedicle screw placement was 2.05 ± 1.2 mm (0.12–5.12) for the screw head, 1.65 ± 1.11 mm (0.16–3.94) for the middle of the pedicle and 1.57 ± 1.01 mm (0.18–3.8) for the screw tip

123

References 1. Foley KT, Holly LT, Schwender JD (2003) Minimally invasive lumbar fusion. Spine 28(15S):S26–S35

J Robotic Surg (2015) 9:331–338 2. Ringel F, Stoffel M, Stu¨er C, Meyer B (2006) Minimally invasive transmuscular pedicle screw fixation of the thoracic and lumbar spine. Neurosurgery 59(4 Suppl 2):ONS361–366 (discussion ONS366–367, Oct. 2006) 3. Kim D-Y, Lee S-H, Chung SK, Lee H-Y (2005) Comparison of multifidus muscle atrophy and trunk extension muscle strength: percutaneous versus open pedicle screw fixation. Spine 30(1):123–129 4. Zairi F, Aboukais R, Marinho P, Allaoui M, Assaker R (2014) Minimally invasive percutaneous stabilization plus balloon kyphoplasty for the treatment of type A thoraco lumbar spine fractures: minimum 4 year’s follow-up. J Neurosurg Sci 58(3):169–175 5. Zairi F, Arikat A, Allaoui M, Marinho P, Assaker R (2012) Minimally invasive decompression and stabilization for the management of thoracolumbar spine metastasis. J Neurosurg Spine 17(1):19–23 6. Gelalis ID, Paschos NK, Pakos EE, Politis AN, Arnaoutoglou CM, Karageorgos AC, Ploumis A, Xenakis TA (2012) Accuracy of pedicle screw placement: a systematic review of prospective in vivo studies comparing free hand, fluoroscopy guidance and navigation techniques. Eur Spine J 21(2):247–255 7. Fuentes S, Metellus P, Fondop J, Pech-Gourg G, Dufour H, Grisoli F (2007) Traitement des fractures de type burst de la charnie`re thoracolombaire par kyphoplastie et oste´osynthe`se percutane´e. Neurochirurgie 53(4):272–276 8. Rampersaud YR, Foley KT, Shen AC, Williams S, Solomito M (2000) Radiation exposure to the spine surgeon during fluoroscopically assisted pedicle screw insertion. Spine 25(20):2637–2645 9. Waschke A, Walter J, Duenisch P, Reichart R, Kalff R, Ewald C (2013) CT-navigation versus fluoroscopy-guided placement of pedicle screws at the thoracolumbar spine: single center experience of 4,500 screws. Eur Spine J 22(3):654–660 10. Beck M, Mittlmeier T, Gierer P, Harms C, Gradl G (2009) Benefit and accuracy of intraoperative 3D-imaging after pedicle screw placement: a prospective study in stabilizing thoracolumbar fractures. Eur Spine J 18(10):1469–1477 11. Heintel TM, Berglehner A, Meffert R (2013) Accuracy of percutaneous pedicle screws for thoracic and lumbar spine fractures: a prospective trial. Eur Spine J 22(3):495–502 12. Ravi B, Zahrai A, Rampersaud R (2011) Clinical accuracy of computer-assisted two-dimensional fluoroscopy for the percutaneous placement of lumbosacral pedicle screws. Spine 36(1):84–91 13. Bydon M, Xu R, Amin AG, Macki M, Kaloostian P, Sciubba DM, Wolinsky J-P, Bydon A, Gokaslan ZL, Witham TF (2014) Safety and efficacy of pedicle screw placement using intraoperative computed tomography: consecutive series of 1148 pedicle screws. J Neurosurg Spine 21(3):320–328 14. Terman SW, Yee TJ, Lau D, Khan AA, La Marca F, Park P (2014) Minimally invasive versus open transforaminal lumbar interbody fusion: comparison of clinical outcomes among obese patients. J Neurosurg Spine 20(6):644–652 15. Sidhu GS, Henkelman E, Vaccaro AR, Albert TJ, Hilibrand A, Anderson DG, Rihn JA (2014) Minimally invasive versus open posterior lumbar interbody fusion: a systematic review. Clin Orthop Relat Res 472(6):1792–1799 16. Bourgeois AC, Faulkner AR, Bradley YC, Pasciak AM, Barlow PB, Gash JR, Reid WSJ (2014) Improved accuracy of minimally invasive transpedicular screw placement in the lumbar spine with three-dimensional stereotactic image guidance: a comparative meta-analysis. J Spinal Disord Tech [Epub ahead of print] 17. Blondel B, Fuentes S, Pech-Gourg G, Adetchessi T, Tropiano P, Dufour H (2011) Percutaneous management of thoracolumbar burst fractures: evolution of techniques and strategy. Orthop Traumatol Surg Res 97(5):527–532

337 18. Zairi F, Court C, Tropiano P, Charles YP, Tonetti J, Fuentes S, Litrico S, Deramond H, Beaurain J, Orcel P, Delecrin J, Aebi M, Assaker R (2012) Minimally invasive management of thoracolumbar fractures: Combined percutaneous fixation and balloon kyphoplasty. Orthop Traumatol Surg Res 98(6 Supplement):S105–S111 19. Houten JK, Nasser R, Baxi N (2012) Clinical assessment of percutaneous lumbar pedicle screw placement using the O-arm multidimensional surgical imaging system. Neurosurgery 70(4):990–995 20. Schatlo B, Molliqaj G, Cuvinciuc V, Kotowski M, Schaller K, Tessitore E (2014) Safety and accuracy of robot-assisted versus fluoroscopy-guided pedicle screw insertion for degenerative diseases of the lumbar spine: a matched cohort comparison: clinical article. J Neurosurg Spine, 1–8 21. Eck J, Lange J, Street J, Lapinsky A, DiPaola C (2013) Accuracy of intraoperative computed tomography-based navigation for placement of percutaneous pedicle screws. Glob Spine J 03(02):103–108 22. Van de Kelft E, Costa F, Van der Planken D, Schils F (2012) A prospective multicenter registry on the accuracy of pedicle screw placement in the thoracic, lumbar, and sacral levels with the use of the O-arm imaging system and stealth station navigation. Spine 37(25):E1580–E1587 23. Yson SC, Sembrano JN, Sanders PC, Santos ERG, Ledonio CGT, Polly DW (2013) Comparison of cranial facet joint violation rates between open and percutaneous pedicle screw placement using intraoperative 3-D CT (O-arm) computer navigation. Spine 38(4):E251–E258 24. Shweikeh F, Amadio JP, Arnell M, Barnard ZR, Kim TT, Johnson JP, Drazin D (2014) Robotics and the spine: a review of current and ongoing applications. Neurosurg. Focus 36(3):E10 25. Roser F, Tatagiba M, Maier G (2013) Spinal robotics: current applications and future perspectives. Neurosurgery 72(Suppl 1):12–18 26. Mattei TA, Rodriguez AH, Sambhara D, Mendel E (2014) Current state-of-the-art and future perspectives of robotic technology in neurosurgery. Neurosurg Rev 37(3):357–366 (discussion 366) 27. Onen MR, Simsek M, Naderi S (2014) Robotic spine surgery: a preliminary report. Turk Neurosurg 24(4):512–518 28. Onen MR, Naderi S (2014) Robotic systems in spine surgery. Turk Neurosurg 24(3):305–311 29. Devito DP, Kaplan L, Dietl R, Pfeiffer M, Horne D, Silberstein B, Hardenbrook M, Kiriyanthan G, Barzilay Y, Bruskin A, Sackerer D, Alexandrovsky V, Stuer C, Burger R, Maeurer E, Gordon DG, Schoenmayr R, Friedlander A, Knoller N, Schmieder K, Pechlivanis I, Kim I-S, Meyer B, Ds Shoham B (2010) Clinical acceptance and accuracy assessment of spinal implants guided with spine assist surgical robot: retrospective study. [Miscellaneous Article]. Spine 35(24):2109–2115 30. Pechlivanis I, Kiriyanthan G, Engelhardt M, Scholz M, Lu¨cke S, Harders A, Schmieder K (2009) Percutaneous placement of pedicle screws in the lumbar spine using a bone mounted miniature robotic system: first experiences and accuracy of screw placement. Spine 34(4):392–398 31. Ringel F, Stuer C, Reinke A, Preuss A, Behr M, Auer F, Stoffel M, Meyer B (2012) Accuracy of robot-assisted placement of lumbar and sacral pedicle screws: a prospective randomized comparison to conventional freehand screw implantation. Spine 37(8) 32. Schizas C, Thein E, Kwiatkowski B, Kulik G (2012) Pedicle screw insertion: robotic assistance versus conventional C-arm fluoroscopy. Acta Orthop. Belg. 78(2):240–245 33. Hu X, Lieberman IH What is the learning curve for roboticassisted pedicle screw placement in spine surgery? Clin Orthop Relat Res 1–6

123

338 34. Lieberman IH, Togawa D, Kayanja MM, Reinhardt MK, Friedlander A, Knoller N, Benzel EC (2006) Bone-mounted miniature robotic guidance for pedicle screw and translaminar facet screw placement: part I-Technical development and a test case result. Neurosurgery 59(3):641–650 (discussion 641–650) 35. Hu X, Lieberman IH (2014) What is the learning curve for robotic-assisted pedicle screw placement in spine surgery? Clin Orthop Relat Res. 472(6):1839–1844 36. Hu X, Ohnmeiss DD, Lieberman IH (2013) Robotic-assisted pedicle screw placement: lessons learned from the first 102 patients. Eur Spine J 22(3):661–666 37. Lefranc M, Capel C, Pruvot AS, Fichten A, Desenclos C, Toussaint P, Le Gars D, Peltier J (2014) The impact of the reference imaging modality, registration method and intraoperative flatpanel computed tomography on the accuracy of the ROSAÒ stereotactic robot. Stereotact Funct Neurosurg 92(4):242–250

123

J Robotic Surg (2015) 9:331–338 38. Lefranc M, Capel C, Pruvot-Occean A-S, Fichten A, Desenclos, P. Toussaint C, Le Gars D, Peltier J (2014) Frameless robotic stereotactic biopsies: a consecutive series of 100 cases. J Neurosurg 1–11 39. Gerszten PC, Ozhasoglu C, Burton SA, Vogel WJ, Atkins BA, Kalnicki S, Welch WC (2003) CyberKnife frameless singlefraction stereotactic radiosurgery for benign tumors of the spine. Neurosurg. FOCUS 14(5):e16 40. Maciunas RJ, Galloway RL Jr, Latimer JW (1994) The application accuracy of stereotactic frames. Neurosurgery 35(4):682–694 (discussion 694–695) 41. Lieberman IH, Hardenbrook MA, Wang JC, Guyer RD (2012) Assessment of pedicle screw placement accuracy, procedure time, and radiation exposure using a miniature robotic guidance system. J Spinal Disord 25(5):241–248