ORIGINAL ARTICLE

Accuracy of patient-specific template-guided vs. free-handfluoroscopically controlled pedicle screw placement in the thoracicand lumbar spine: a randomized cadaveric study

Mazda Farshad1 • Michael Betz1 • Nadja A. Farshad-Amacker2 •

Manuel Moser1

Received: 5 May 2016 / Revised: 1 August 2016 /Accepted: 1 August 2016 / Published online: 9 August 2016

� Springer-Verlag Berlin Heidelberg 2016

Abstract

Purpose Dorsal spinal instrumentation with pedicle screw

constructs is considered the gold standard for numerous

spinal pathologies. Screw misplacement is biomechanically

disadvantageous and may create severe complications. The

aim of this study was to assess the accuracy of patient-

specific template-guided pedicle screw placement in the

thoracic and lumbar spine compared to the free-hand

technique with fluoroscopy.

Methods Patient-specific targeting guides were used for

pedicle screw placement from Th2–L5 in three cadaveric

specimens by three surgeons with different experience

levels. Instrumentation for each side and level was ran-

domized (template-guided vs. free-hand). Accuracy was

assessed by computed tomography (CT), considering per-

forations of\2 mm as acceptable (safe zone). Time effi-

ciency, radiation exposure and dependencies on surgical

experience were compared between the two techniques.

Results 96 screws were inserted with an equal distribution

of 48 screws (50 %) in each group. 58 % (n = 28) of

template-guided (without fluoroscopy) vs. 44 % (n = 21)

of free-hand screws (with fluoroscopy) were fully con-

tained within the pedicle (p = 0.153). 97.9 % (n = 47) of

template-guided vs. 81.3 % (n = 39) of free-hand screws

were within the 2 mm safe zone (p = 0.008). The mean

time for instrumentation per level was 01:14 ± 00:37 for

the template-guided vs. 01:40 ± 00:59 min for the free-

hand technique (p = 0.013), respectively. Increased radi-

ation exposure was highly associated with lesser experi-

ence of the surgeon with the free-hand technique.

Conclusions In a cadaver model, template-guided pedicle

screw placement is faster considering intraoperative

instrumentation time, has a higher accuracy particularly in

the thoracic spine and creates less intraoperative radiation

exposure compared to the free-hand technique.

Keywords Spine surgery � Pedicle screw � Accuracy �Pedicle perforation � Patient-specific

Introduction

Dorsal spinal instrumentation with pedicle screws and rod

constructs has become a widespread surgical procedure in

the treatment of degenerative spinal disease, spinal defor-

mity, trauma and tumors with continuously increasing

numbers [1, 2]. Numerous assistive techniques have been

elaborated over the years harboring advantages and dis-

advantages in terms of screw placement accuracy, radiation

exposure, surgical time, operating room equipment needs,

learning curves and health care costs. Accurate screw

placement can be technically challenging with a demand

for high level spine care centers and surgical expertise.

Screw misplacement not only carries the risk of neuro-

logical or vascular complications, but also has biome-

chanical disadvantages. Patient-specific templates used as

in situ drill guides are an alternative to free-hand or navi-

gated techniques, aiming to improve screw placement

accuracy while at the same time reducing radiation expo-

sure. Originally reported in the late 1990’s by Radermacher

et al. [3], recent advances in computer and additive

& Mazda Farshad

mazda.farshad@balgrist.ch

1 Division of Spine Surgery, Balgrist University Hospital

Zurich, Forchstrasse 340, 8008 Zurich, Switzerland

2 Institute of Diagnostic and Interventional Radiology,

University Hospital Zurich, Ramistrasse 101, 8091 Zurich,

Switzerland

123

Eur Spine J (2017) 26:738–749

DOI 10.1007/s00586-016-4728-5

manufacturing technology have led to a more sophisticated

and practical application of template-guided procedures in

spine surgery during the last decade with promising results

[4–13]. We hypothesized that template-guided screw

placement is superior to free-hand placement in terms of

accuracy, radiation exposure and time efficiency. Further,

we hypothesized that surgical experience is less pivotal in

template-guided vs. free-hand pedicle screw placement.

Materials and methods

Three fresh frozen adult cadavers without prior surgery or

deformity of the spine were randomly assigned to three

surgeons with different experience levels, namely two

board-certified orthopedic spine surgeons (chief of spine

surgery and attending spine surgeon at a academic spine

division) and one senior-level neurosurgical resident, all

right-handed. The mean diameter of the pedicles on axial

CT scans at the level Th7 was 4.67 ± 0.85 mm (ranges

3.5–5.5 mm). A randomization list (computerized ran-

domization) was prepared for each cadaver, allocating the

template-guided (MySpine�, Medacta SA International,

Switzerland) and the free-hand technique to the left or right

of each vertebral level (intra-vertebral randomization).

Surgical instrumentation time as well as cumulative radi-

ation dose was recorded. No fluoroscopy was used during

template-guided instrumentation. Time measurement for

the template-guided technique started with the positioning

of the guide onto the vertebra or with the osteotomy to

exhibit the entry point for the free-hand technique. Time

measurement stopped at definitive screw placement. The

diameter of the pedicle screws (MUST�-Medacta Uncon-

strained Screw Technology) ranged from 5 to 6 mm

depending on pre-instrumentation planning on CT and the

screw length ranged from 25 to 55 mm. An illustrative

flow-chart comparing both techniques step-by-step is

shown in Fig. 1.

Presurgical planning, vertebra replica and drill

guide fabrication

According to the manufacturer’s protocol a 0.64 mm

spiral CT scan was performed of each specimen’s spine.

After three-dimensional reconstruction with Mimics�

(Materialise, Leuven, Belgium) a digital surgical plan was

developed including entry point, screw length, screw

diameter, as well as screw angulation in the sagittal,

transversal and coronal plane for each level using Solid-

works� (Dassault Systemes, Velizy-Villacoublay,

France). An illustration of the planning report for one of

the thoracic levels (Th5) is shown in Fig. 2. Planned ‘out

screws’ due to narrow pedicles were marked in the pro-

tocol. The participating surgeons reviewed and validated

the planning. Three-dimensional replicas of all vertebrae

and level-specific drill guides were produced using a rapid

prototyping technique with medical grade polyamide

(P2200) and selective laser sintering (SLS) technology

(P395, EOS e-Manufacturing Solutions, Munich, Ger-

many). The dorsal third of the vertebral replicas including

lamina, spinous process, transverse and articulate pro-

cesses (Fig. 3) allowed the surgeons to check for adequate

contact areas and positioning of the guides prior to

definitive in situ placement.

Surgical technique

A dorsal midline incision was performed at the levels Th1–

L5, including dissection and retraction of the paravertebral

musculature to expose the anatomical bony landmarks,

including the transverse processes. Facet joints and their

respective capsules were fully preserved and the supras-

pinous ligament completely removed. The lamina, spinous

process and transverse processes were depicted as being the

main contact areas for the template (Fig. 3). For this rea-

son, instrumentation was first done with the template-gui-

ded system (Fig. 4) according to the randomization list,

which was blinded to the surgeons until the experiments

started. A K-wire was drilled into the vertebra following

the guide trajectory to a depth consistent with the presur-

gical planning of the desired screw length. With the K-wire

in place the guide was removed and a cannulated drill was

used to burr cortical bone. Finally, predefined cannulated

pedicle screws were inserted over the K-wire. For the free-

hand technique intraoperative fluoroscopy was installed

and the surgeon was free to use it or not for proper screw

placement, depending on how safe he felt with the free-

hand instrumentation. First, an osteotomy was done using

bone rongeurs or a chisel to expose the entry point. The

pedicle was prepared with a Lenke bone probe (Fig. 5) and

a small ball tip was used to check bony integrity of the

pedicle wall. No taps were used. Finally, a screw of the

same diameter and length as planed for the template-guided

technique was inserted free-hand. The level Th1 had to be

excluded from the analysis due to anatomical difficulties

with the specimens and repeated failure to prepare the

pedicles for both techniques.

Postoperative evaluation of pedicle wall integrity

Following instrumentation, a 0.64 mm spiral CT scan of all

cadaveric specimens was performed. An independent

Eur Spine J (2017) 26:738–749 739

123

board certified radiologist, who was blinded to the surgical

technique of screw insertion, reviewed all scans in the

sagittal, transversal and coronal plane on two different

occasions. In cases of disagreement between the two

readouts, the worse grade was used for analysis. Pedicle

wall integrity was classified as normal when the

surrounding cortical bone was fully intact in all three

planes on CT. Perforation was documented whenever the

cortex was harmed or interrupted by any part of the screw,

evident in at least one of the three planes on CT. Perfo-

rations were classified as following: (A) \2 mm,

(B) 2–4 mm or (C) [4 mm (see Fig. 6 for illustrative

Fig. 1 Simplified step-by-step procedures of the free-hand and

template-guided pedicle screw placement technique in the presurgical

phase and during surgery. Dotted lines indicate possible returns to an

earlier stage of the procedure or switching to the free-hand technique

if drill-guide application fails or screw misplacement is suspected

740 Eur Spine J (2017) 26:738–749

123

cases). A perforation of \2 mm was considered accept-

able (safe zone). Additionally, the localization of the per-

foration, if present, was categorized into superior, inferior,

lateral or medial. Possible dependencies on the practical

experience of the surgeons were also evaluated, as well as

individual expenditure of time and cumulative radiation

dose.

Statistical analyses

For the comparison of categorical data Pearson’s Chi-

squared test and Fisher’s exact test were used. A two-tailed

Student’s t test was used for the comparison of means and

descriptive statistics was employed to report means, stan-

dard deviations and ranges. Analysis was performed with

Fig. 2 Three-dimensional planning report for template-guided pedi-

cle screw placement on the level Th5 in one of three cadavers used.

SAR/SAL sagittal plane angle right/left with angulation of the screw

shaft in relation to pedicle center line while center of rotation is

located at the minimal cross section of the pedicle (red dot), TAR/TAL

transversal plane angle right/left with angulation of the screw shaft in

relation to the pedicle center line while center of rotation is located at

the minimal cross section of the pedicle (red dot), HDL/HDR

horizontal distance left/right, VDL/VDR vertical distance left/right

Eur Spine J (2017) 26:738–749 741

123

SPSS Statistics version 23.0 (IBM Corp., Armonk NY,

USA). Results at a probability value p\ 0.05 were con-

sidered statistically significant.

Results

Pedicle perforation rates and grading

96 pedicle screws were inserted with an equal distribution

of 48 screws (50 %) in each group (template-guided vs.

free-hand). 4 screws (Th6 right, Th8 left, Th11 right and L1

right) were repositioned in the free-hand group and 1 screw

(L5 left) in the template-guided group because of suspected

misplacement while operating. Accuracies for each tech-

nique are illustrated in Fig. 7 and categorized in Fig. 8

using the 2 mm cutoff (safe zone).

58.33 % (n = 28) of template-guided screws were fully

contained inside the pedicle, 39.58 % (n = 19) were grade

A and 2.08 % (n = 1) grade B perforations. No grade C

perforation was observed in the template-guided screws.

Lateral perforations were the most frequent ones in this

Fig. 3 Replica of the vertebra Th11 and corresponding drill guide, seen from cranial as labeled on the guide. a Bony contact areas for

appropriate placement of the template. b K-wires inserted on both sides illustrating the screw trajectory and entry points

Fig. 4 Left-sided template-guided screw placement in the upper thoracic spine. a K-wire drilling while ensuring adequate bony contact of the

template. b Use of a cannulated drill to burr cortical bone for widening the entry point. c Screw insertion over the K-wire. d Final screw position

742 Eur Spine J (2017) 26:738–749

123

group with 40 % (n = 8) followed by 35 % (n = 7) medial

perforations, 15 % (n = 3) inferior, 5 % (n = 1) lateral-

inferior and 5 % (n = 1) lateral-medial perforations. All

medial, lateral-inferior and lateral-medial perforations were

grade A. The only grade B perforation was inferior.

43.75 % (n = 21) of the free-hand screws were fully

contained inside the pedicle, 37.5 % (n = 18) were grade

A, 10.42 % (n = 5) grade B and 8.33 % (n = 4) grade C

perforations. Lateral perforations were most frequent in

this group with 81.48 % (n = 22). The remaining 18.52 %

(n = 5) were medial perforations. All medial perforations

were grade A. Grade B and C perforations (n = 9) were

laterally.

Comparison of no perforation vs. any perforation

(grades A–C), in terms of a simple ‘in or out’ method

showed no statistically significant difference depending on

the used surgical technique (p = 0.153). Considering per-

forations of \2 mm as acceptable (safe zone), 97.92 %

(n = 47) of template-guided vs. 81.25 % (n = 39) of free-

hand screws were ‘safe’. A C2 mm breach of the pedicle

cortex was observed in only one pedicle (2.08 %) in the

template-guided group vs. in nine pedicles (18.75 %) in the

free-hand group. Therefore, the selected surgical technique

had a statistically important influence on the severity of

perforation (p = 0.008).

In the thoracic spine, 96.97 % (n = 32) of the template-

guided screws were within and 3.03 % (n = 1, grade B,

Th2 right) outside the safe zone compared to 75.76 %

(n = 25) and 24.24 % (n = 8) of free-hand screws

(p = 0.027), respectively. No significant difference was

seen in the lumbar spine as 100 % (n = 15) of template-

guided and 93.33 % (n = 14) of free-hand screws were

within the safe zone. Only one severe perforation was

observed in the free-hand group (grade C, L1 right) in the

lumbar region.

Surgeon’s experience

Given the \2 mm safe zone, the practical experience of

the surgeons (board-certified vs. senior-level resident) had

no influence on the severity of pedicle perforation with

the free-hand technique (p = 0.697). The limited amount

of severe perforations of the template-guided screws did

not allow analysis in this regard. Likewise, the side of

instrumentation (right vs. left) had no influence on the

severity of pedicle perforation in the free-hand group

(p = 0.137). There was practically no difference in the

template-guided group with 24 left-sided and 23 right-

sided screws within the safe zone and one grade B per-

foration (Th2 right).

Radiation dose and fluoroscopy time

Spiral CT scans for surgical planning had a mean dose

length product (DLP) of 1627.5 ± 71.65 mGy*cm (range

1526.3–1682.4 mGy*cm). It should be noted that, because

of the use of cadavers, these CTs were scanned with a

higher dose [fixed X-ray tube current of 399 mAs instead

of automatic exposure control (AEC)] than would have

been used in living individuals with standard protocols.

Mean intraoperative fluoroscopy dose was 889 ± 604.6

mGycm2 (range 225.1–1687.5 mGycm2) for the free-hand

instrumentation of L5–Th1, and the mean radiation time

was 01:14 ± 00:29 min (range 00:37–01:49 min). Fluo-

roscopy time and cumulative dose showed a negative

correlation with the practical experience of the performing

surgeon (ranges corresponding to the most and least

experienced—see Table 1 for details).

Expenditure of time

Time values for each technique and level are listed for each

cadaver in Table 1. Mean time for template-guided

instrumentation (n = 48) per level was 01:14 ± 00:37 min

(range 00:38–03:58 min) and 01:40 ± 00:59 min (range

00:28–05:48 min) for free-hand instrumentation (n = 48),

the difference being statistically significant (p = 0.013).

Mean total time for template-guided instrumentation of

Th2–L5 was 19:47 ± 04:35 min (range 14:42–25:49 min)

per specimen as compared to 26:40 ± 09:38 min (range

13:21–35:46 min) for the free-hand technique.

Fig. 5 Intraoperative lateral fluoroscopy showing free-hand prepara-

tion of the right pedicle Th5 using a Lenke bone probe with cranial

corresponding to the left side of the picture with template-guided

screws already inserted. The same specimen is also depicted in

Fig. 6c

Eur Spine J (2017) 26:738–749 743

123

Discussion

Screw placement accuracy is based on definition and

assessment with no standardized method or consensus. A

review by Aoude et al. [14] showed that the 2 mm incre-

ment grading, originally proposed in 1990 by Gertzbein

and Robbins [15], is the most commonly used method to

assess pedicle screw accuracy with substantial intra-rater

(j = 0.83) [16] and inter-rater agreement (j = 0.65–0.85)

[16, 17]. Nevertheless, it is a pure radiological grading

system not taking into account possible clinical or biome-

chanical consequences of misplaced pedicle screws, which

are not always necessarily negative (see Fig. 6c).

Parker et al. [18] retrospectively reviewed 6816 free-

hand placed pedicle screws and found a thoracic and

lumbar spine screw accuracy rate of 97.5 and 99.1 %,

respectively. Belmont et al. [19] showed that only 57 % of

279 free-hand placed thoracic pedicle screws in 40 patients

were fully inside the pedicle and 28 % within a 2 mm

breach, the accuracy being the highest in the lower thoracic

spine (Th9–Th12) and lateral breaches occurring signifi-

cantly more often than medial ones (68 vs. 32 %,

p\ 0.005). Similarly, Motiei-Langroudi et al. [20] repor-

ted an overall accuracy rate of 97.7 % for the free-hand

technique with lateral fluoroscopy in 770 thoracolumbar

screws, the accuracy being the highest at the L3–S1 levels

(mean 99 %), followed by the thoracolumbar junction area

(T10–L2, mean 96.5 %) and with the lowest accuracy at

the mid-thoracic area (T7–T9, mean 89.5 %). This is

consistent with our findings that screw placement accuracy

was almost perfect in the lumbar spine for both techniques,

but misplacement was much higher in the thoracic spine for

the free-hand technique. This may be related to several

confounding factors: first, a surgeon’s experience may be

higher for lumbar than thoracic pedicle screw placement.

Second, each surgeon may have his preferred method as

Fig. 6 Axial CT scans illustrating the degrees of perforation. a No

perforation, both L5 screws are fully contained within the pedicle.

b Slight medial wall violation of a left L2 screw (free-hand) being

classified as grade A (\2 mm). c Lateral wall violation of a right Th5

screw (free-hand) being classified as grade B (2–4 mm). d Severe

lateral wall violation of a right Th4 screw (free-hand) being classified

as grade C ([4 mm). Although radiologically classified as severe, this

screw is expected to have strong purchase due to its ‘in-out-in’

trajectory with a tricortical anchorage. Sagittal and coronal planes

were also considered for the evaluation of perforation and are not

depicted here for the ease of illustration

744 Eur Spine J (2017) 26:738–749

123

numerous techniques have been proposed for accurate

screw placement in the thoracic spine [21–24] with

potentially different learning curves. Third, there is a large

inter-segmental variability concerning shape, size and ori-

entation of thoracic pedicles and the relation to adjacent

neural structures [25–27]. The relatively small transversal

diameter of the thoracic pedicles in one specimen may

have contributed to an above-average number of perfora-

tions with both techniques although instrumented by an

experienced surgeon. We observed a greater chance to

misplace pedicle screws laterally, a tendency that has been

reported earlier [19, 25, 28, 29]. This might be due to the

intentional salvage ‘in-out-in technique’ used in narrow

thoracic pedicles to increase biomechanical stability (see

Fig. 6d) and a surgeon’s reasonable fear of neurological

complications in cases of medial wall violations.

Anatomical studies have shown that the distance between

the dural sac and the medial pedicle wall can be within

0.0–0.7 mm in the thoracic spine [30, 31]. Increasing

pedicle screw size in the thoracic spine has been shown to

cause pedicle expansion and breaches laterally without

significantly altering transversal spinal canal diameter in a

cadaveric study [29]. For pedicle screw placement in sin-

gle-level degenerative spondylolisthesis lateral perforations

have also been reported to be more common because of

instability at the instrumented level leading to translation

and rotation of the vertebral body while placing pedicle

screws, a problem that could not be prevented using

intraoperative O-arm navigation for screw placement [32].

In contrary to the well-known learning curves for free-hand

screw placement in thoracic deformity surgery [33–35], no

learning curve was observed for template-guided screws.

In a meta-analysis by Kosmopoulos et al. [36] including

37,337 screws, a median screw accuracy of 86.6 % without

navigation compared to 93.7 % with navigation was

reported for in vivo lumbar and/or thoracic spine surgery,

which was higher than for in vitro studies (79 % without

and 87.3 % with navigation). Mason et al. [37] reviewed 30

articles including 1973 patients and 9310 pedicle screws

and found an accuracy of 68.1 % for free-hand conven-

tional fluoroscopy, 84.3 % for two-dimensional and 95.5 %

for three-dimensional fluoroscopic navigation. Shin et al.

[38] compared fluoroscopically controlled vs. navigation-

guidance coupled with O-arm for screw placement in the

thoracic and lumbosacral spine in a randomized prospec-

tive manner. They found 91.9 % of navigated screws to be

fully contained within the pedicle vs. 87.7 % of fluoro-

scopically placed screws. Tian et al. [39] found median

in vivo screw accuracies of 90.76 % for CT-navigated vs.

85.48 % for two-dimensional fluoroscopy compared to a

higher in vitro median accuracy of 94.59 % for CT-navi-

gated vs. 90.12 % for two-dimensional fluoroscopy. The

reported accuracy of patient-specific template-guided

screws in our study showed a considerably high accuracy

with 97.92 %, which is somehow inferior to the previously

published data with MySpine by Lamartina et al. [13] with

100 % accuracy in 42 screws in a cadaver model. This may

be due to the exclusion of 8/12 perforating screws in the

latter study because the pedicle diameter was considered

too small to avoid perforation, whereas every perforation in

our study was considered for analysis. As for template-

based patient-specific techniques, the accuracy rates for

Fig. 7 Screw placement accuracies for the template-guided and free-

hand technique. Accuracy depicted as no perforation or grade of

pedicle perforation expressed by percentage of screws inserted for

each technique

Fig. 8 Screw placement accuracies for the template-guided and free-

hand technique categorized into no perforation or acceptable perfora-

tions (\2 mm) vs. severe perforations (C2 mm) and expressed by

percentage of screws inserted for each technique. Pearson’s Chi-

squared test revealed that the selected surgical technique had a

statistically important influence on the severity of perforation

(p = 0.008)

Eur Spine J (2017) 26:738–749 745

123

thoracolumbar surgeries given the \2 mm safe zone

have been reported to lie between 96.1 and 100 %

[6–13] and are summarized in Table 2. Lamartina et al.

[13] reported satisfactory results with MySpine target-

ing guides calculating the mean deviation of pedicle

screws between planned and actual screw position in

different planes, with an average deviation from the

planned position at the midpoint of the pedicle of

0.7 mm.

The time needed for unilateral single level preparation

of the pedicle and screw placement was significantly

shorter in the template-guided group, in concordance to a

comparison of computer navigation vs. fluoroscopy in a

review by Meng et al. [40] analyzing 14 articles with 1723

patients and 9019 pedicle screws. It must be critically

stated that this possible reduction of instrumentation time

does not include calculation of the time needed for pre-

operative planning, meticulous soft tissue removal and

adequate bone surface preparation.

Radiation exposure to surgeons and patients during

pedicle screw placement may be unacceptably high using

fluoroscopy [41–43] and largely depends on surgical

experience and the technique used for screw insertion.

Computer assisted image guidance has been shown to

allow a significant reduction in intraoperative radiation

exposure [44], especially when compared to the freehand-

technique [43]. Equally, we could demonstrate that tem-

plate-guided pedicle screw placement potentially allows a

complete reduction of intraoperative fluoroscopy without

sacrificing accuracy, but the definitive need for a presur-

gical CT scan may counterbalances this possible reduction

of radiation exposure.

Table 1 Pure instrumentation time (min) for each level and technique for all three cadavers including surgical experience level, diameter of the

pedicle Th7 on axial CT scan and cumulative fluoroscopy dose and time (see text for details)

Level Senior resident Chief of spine surgery Spine consultant

Pedicle Th 7 on axial CT 5.5 mm Pedicle Th 7 on axial CT 3.5 mm Pedicle Th 7 on axial CT 5.0 mm

Free-

hand

(min)

Template-

guided (min)

Left Right Free-

hand

(min)

Template-

guided (min)

Left Right Free-

hand

(min)

Template-

guided (min)

Left Right

Th2 01:11 02:03 0 0a 01:34 01:07 B Ba 02:22 02:26 0 0a

Th3 01:18 01:26 A Aa 01:09 00:57 A Aa 01:39 01:17 Aa A

Th4 01:02 02:35 0a C 00:52 00:54 B Aa 01:57 02:02 A Aa

Th5 01:22 01:40 0a B 00:55 00:43 Aa A 01:41 01:47 0 Aa

Th6 01:04 01:55 Aa 0 00:53 00:41 Aa A 01:27 04:53 A Aa

Th7 01:10 01:32 A 0a 01:09 00:46 Aa A 01:32 02:44 0a 0

Th8 01:00 02:00 A 0a 01:27 00:50 A 0a 01:11 01:16 Aa C

Th9 01:02 02:45 0 0a 00:51 00:48 Aa B 01:50 01:36 Aa B

Th10 00:53 02:06 0a 0 00:45 00:47 Aa C 01:34 01:06 Aa A

Th11 00:52 02:09 0a 0 00:43 01:09 0 0a 01:13 01:50 0 0a

Th12 00:58 02:07 A Aa 00:39 01:09 0 0a 01:16 02:20 A 0a

L1 00:49 01:30 Aa A 00:41 00:28 0a 0 01:07 05:48 0a C

L2 01:09 02:01 A 0a 00:41 00:35 0a 0 00:58 01:29 0 0a

L3 00:55 01:43 0a 0 00:44 00:46 0 0a 01:04 01:29 0a 0

L4 03:15 01:59 0a A 00:38 00:53 0 0a 01:00 01:21 0a 0

L5 00:51 01:23 0 0a 01:01 00:48 Aa 0 03:58 02:22 A 0a

Mean 01:11 01:56 00:55 00:50 01:37 02:14

SD 00:33 00:22 00:16 00:11 00:43 01:16

Min 00:49 01:23 00:38 00:28 00:58 01:06

Max 03:15 02:45 01:34 01:09 03:58 05:48

Fluoroscopy

dose

(mGycm2)

1687.5 225.1 754.3

Fluroscopy time

(min)

01:49 00:37 01:17

Template-guided screws (left or right) and perforations (0, A, B, C) are highlighted with a superscript letter a

746 Eur Spine J (2017) 26:738–749

123

Limitations of our study are the cadaveric design and

the relatively small sample size, making the assumption

of any of these in vitro data into a clinical context

challenging. The measure of clinical significance of

misplaced pedicle screws is the percentage of symp-

tomatic patients or the need for screw replacement sur-

gery [45], however we can only provide radiological

results. Except for the relative small pedicles in one

specimen, the anatomy found was fairly normal and may

not adequately resembles the anatomy found in an

average spine patient requiring dorsal instrumentation.

Potential drawbacks of the template-guided technique

are: (1) the necessity for meticulous soft tissue removal

including the dorsal ligaments while fully preserving the

bone surface and facet joints, (2) inadequate handling,

e.g. placing the drill template too loose on the bone or

having soft tissue interposed between the guide and the

bone, (3) inadequate digital planning, specifically con-

cerning the quality of the presurgical CT scan with a low

dose protocol and (4) time expenditure for the preoper-

ative planning and production of the drill-guide tem-

plates. Completely resigning intraoperative fluoroscopy

for templated-guided screw placement may not be rec-

ommendable and accuracy should be checked at least at

the end of the procedure or during suspected misplace-

ment while operating. To which extent spinal bone

quality influences the accuracy of template-guided screw

placement and if such a system is useful in (severely)

osteoporotic bone needs to be further investigated,

especially when considering the application of pressure

needed to properly place the template and hold it in firm

contact with the dorsal bony structures.

Conclusions

To our knowledge, this is the first randomized cadaveric

study comparing patient-specific template-guided vs. free-

hand fluoroscopically controlled pedicle screw placement

in the thoracic and lumbar spine using a commercially

available targeting system. The template-guided technique

showed a significantly higher pedicle screw placement

accuracy considering perforations of\2 mm as acceptable,

particularly in the thoracic spine.

Compliance with ethical standards

The study presented here was conducted in accordance with Swiss

and international law requirements. Ethical board’s approval was

obtained from the Ethical Committee of Northwestern and Central

Switzerland with the ID number EKNZ BASEC 2016-00204. The

cadaver workshop was carried out on March 12th 2016 at the

Academy of Medical Training and Simulation (AMTS) in Muttenz,

Switzerland.

Conflict of interest None.

Table 2 Overview of articles describing patient-specific drill templates for pedicle screw placement in the thoracic and lumbar spine including

reported screw placement accuracies

Article Study design Thoracic (T),

lumbar (L), sacral

(S)

Number

of screws

No

perforation

(%)

\2 mm

perforation

(%)

2–4 mm

perforation

(%)

[4 mm

perforation

(%)

Other (%)

Lu et al. [6] Cadaveric ? clinical L 58 58 (100) 0 0 0

Ma et al.

[7]

Cadaveric T 240 224 (93.4) 16 (6.6) 0 0

Lu et al. [8] Clinical T 168 157 (93.45) 11 (6.55) 0 0

Merc et al.

[9]

Clinical L ? S 54 48 (88.89) n.a. n.a. n.a. 6 (11.11)

Sugawara

et al. [10]

Clinical T 58 58 (100) 0 0 0

Lamartina

et al. [13]

Cadaveric T ? L 46 42 (91.3) 4 (8.7) 0 0

Takemoto

et al. [11]

Clinical T (scoliosis) 415 408 (98.4) 1 (0.2) 0

T (OPLL) 46 46 (100) 0 0

Hu et al.

[12]

Clinical T 582 559 (96.05) n.a. n.a. n.a. 23 (3.95)

Current

study

Cadaveric T ? L 48 28 (58.33) 19 (39.58) 1 (2.08) 0

n.a. not available, OPLL ossification of the posterior longitudinal ligament

Eur Spine J (2017) 26:738–749 747

123

References

1. Rajaee S, Bae H, Kanim L, Delamarter R (2012) Spinal fusion in

the United States: analysis of trends from 1998 to 2008. Spine

(Philadelphia, Pa 1976) 37(1):67–76

2. Boos N, Webb JK (1997) Pedicle screw fixation in spinal disor-

ders: a European view. Eur Spine J 6(1):2–18. doi:10.1007/

bf01676569

3. Radermacher K, Portheine F, Anton M, Zimolong A, Kaspers G,

Rau G, Staudte H-W (1998) Computer assisted orthopaedic sur-

gery with image based individual templates. Clin Orthop Relat

Res 354:28–38

4. Birnbaum K, Schkommodau E, Decker N, Prescher A, Klapper

U, Radermacher K (2001) Computer-assisted orthopedic sur-

gery with individual templates and comparison to conventional

operation method. Spine (Philadelphia, Pa 1976)

26(4):365–370

5. Berry E, Cuppone M, Porada S, Millner PA, Rao A, Chiverton N,

Seedhom BB (2005) Personalised image-based templates for

intra-operative guidance. Proc Inst Mech Eng 219(2):111–118

6. Lu S, Xu YQ, Zhang YZ, Li YB, Xie LE, Shi JH, Guo H, Chen

GP, Chen YB (2009) A novel computer-assisted drill guide

template for lumbar pedicle screw placement: a cadaveric and

clinical study. Int J Med Robot Comput Assist Surg

5(2):184–191. doi:10.1002/rcs.249

7. Ma T, Xu Y-Q, Cheng Y-B, Jiang M-Y, Xu X-M, Xie L, Lu S

(2012) A novel computer-assisted drill guide template for tho-

racic pedicle screw placement: a cadaveric study. Arch Orthop

Trauma Surg 132(1):65–72

8. Lu S, Zhang YZ, Wang Z, Shi JH, Chen YB, Xu XM, Xu YQ

(2012) Accuracy and efficacy of thoracic pedicle screws in sco-

liosis with patient-specific drill template. Med Biol Eng Comput

50(7):751–758. doi:10.1007/s11517-012-0900-1

9. Merc M, Drstvensek I, Vogrin M, Brajlih T, Recnik G (2013) A

multi-level rapid prototyping drill guide template reduces the

perforation risk of pedicle screw placement in the lumbar and

sacral spine. Arch Orthop Trauma Surg 133(7):893–899. doi:10.

1007/s00402-013-1755-0

10. Sugawara Taku, Higashiyama Naoki, Kaneyama Shuichi, Taka-

batake Masato, Watanabe N, Uchida Fujio, Sumi Masatoshi,

Mizoi Kazuo (2013) Multistep pedicle screw insertion procedure

with patient-specific lamina fit-and-lock templates for the tho-

racic spine. J Neurosurg Spine 19(2):185–190. doi:10.3171/2013.

4.SPINE121059

11. Takemoto M, Fujibayashi S, Ota E, Otsuki B, Kimura H, Saka-

moto T, Kawai T, Futami T, Sasaki K, Matsushita T, Nakamura

T, Neo M, Matsuda S (2016) Additive-manufactured patient-

specific titanium templates for thoracic pedicle screw placement:

novel design with reduced contact area. Eur Spine J

25(6):1698–1705. doi:10.1007/s00586-015-3908-z

12. Hu Y, Yuan ZS, Spiker WR, Dong WX, Sun XY, Yuan JB,

Zhang J, Zhu B (2016) A comparative study on the accuracy of

pedicle screw placement assisted by personalized rapid proto-

typing template between pre- and post-operation in patients with

relatively normal mid-upper thoracic spine. Eur Spine J

25(6):1706–1715. doi:10.1007/s00586-016-4540-2

13. Lamartina C, Cecchinato R, Fekete Z, Lipari A, Fiechter M,

Berjano P (2015) Pedicle screw placement accuracy in thoracic

and lumbar spinal surgery with a patient-matched targeting guide:

a cadaveric study. Eur Spine J 24(7):937–941. doi:10.1007/

s00586-015-4261-y

14. Aoude AA, Fortin M, Figueiredo R, Jarzem P, Ouellet J, Weber

MH (2015) Methods to determine pedicle screw placement

accuracy in spine surgery: a systematic review. Eur Spine J

24(5):990–1004. doi:10.1007/s00586-015-3853-x

15. Gertzbein SD, Robbins SE (1990) Accuracy of pedicular screw

placement in vivo. Spine 15(1):11–14

16. Bai YS, Niu YF, Chen ZQ, Zhu XD, Gabriel L, Wong H, Li M

(2013) Comparison of the pedicle screws placement between

electronic conductivity device and normal pedicle finder in pos-

terior surgery of scoliosis. J Spinal Disord Tech 26(6):316–320

17. Kim M-C, Chung H-T, Cho J-L, Kim D-J, Chung N-S (2011)

Factors affecting the accurate placement of percutaneous pedicle

screws during minimally invasive transforaminal lumbar inter-

body fusion. Eur Spine J 20(10):1635–1643. doi:10.1007/s00586-

011-1892-5

18. Parker S, McGirt M, Farber S, Amin A, Rick AM, Suk I, Bydon

A, Sciubba D, Wolinsky JP, Gokaslan Z, Witham T (2011)

Accuracy of free-hand pedicle screws in the thoracic and lumbar

spine: analysis of 6816 consecutive screws. Neurosurgery

68(1):170–178

19. Belmont PJ, Klemme WR, Dhawan A, Polly DW (2001) In vivo

accuracy of thoracic pedicle screws. Spine (Philadelphia, Pa

1976) 26(21):2340–2346

20. Motiei-Langroudi R, Sadeghian H (2015) Assessment of pedicle

screw placement accuracy in thoracolumbosacral spine using

freehand technique aided by lateral fluoroscopy: results of post-

operative computed tomography in 114 patients. Spine J

15(4):700–704. doi:10.1016/j.spinee.2014.12.012

21. Fennell Vernard S, Palejwala Sheri, Skoch Jesse, Stidd David A,

Baaj Ali A (2014) Freehand thoracic pedicle screw technique

using a uniform entry point and sagittal trajectory for all levels:

preliminary clinical experience. J Neurosurg Spine

21(5):778–784. doi:10.3171/2014.7.SPINE1489

22. Kim Y, Lenke L, Bridwell K, Cho Y, Riew K (2004) Free hand

pedicle screw placement in the thoracic spine: is it safe? Spine

(Philadelphia, Pa 1976) 29(3):333–342

23. Lehman R, Polly D, Kuklo T, Cunningham B, Kirk K, Belmont P

(2003) Straight-forward vs. anatomic trajectory technique of

thoracic pedicle screw fixation: a biomechanical analysis. Spine

(Philadelphia, Pa 1976) 28(18):2058–2065

24. Modi H, Suh S, Song H-R, Yang J-H (2009) Accuracy of thoracic

pedicle screw placement in scoliosis using the ideal pedicle entry

point during the freehand technique. Int Orthop 33(2):469–475

25. Panjabi MM, O’Holleran JD, Crisco JJ, Kothe R (1997) Com-

plexity of the thoracic spine pedicle anatomy. Eur Spine J

6(1):19–24

26. Potter BK, Lehman RAJ, Kuklo TR (2004) Anatomy and

biomechanics of thoracic pedicle screw instrumentation. Curr

Opin Orthop 15(3):133–144

27. Zindrick MR, Wiltse LL, Doornik A, Widell EH, Knight GW,

Patwardhan AG, Thomas JC, Rothman SL, Fields BT (1987)

Analysis of the morphometric characteristics of the thoracic and

lumbar pedicles. Spine (Philadelphia, Pa 1976) 12(2):160–166

28. Carbone J, Tortolani PJ, Quartararo L (2003) Fluoroscopically

assisted pedicle screw fixation for thoracic and thoracolumbar

injuries: technique and short-term complications. Spine

(Philadelphia, Pa 1976) 28(1):91–97

29. Cho SK, Skovrlj B, Lu Y, Caridi JM, Lenke LG (2014) The effect

of increasing pedicle screw size on thoracic spinal canal dimen-

sions: an anatomic study. Spine 39(20):E1195–E1200. doi:10.

1097/brs.0000000000000514

30. Ebraheim NA, Jabaly G, Xu R, Yeasting RA (1997) Anatomic

relations of the thoracic pedicle to the adjacent neural structures.

Spine (Philadelphia, Pa 1976) 22(14):1553–1556 discussion 155731. Ugur H, Attar A, Uz A, Tekdemir I, Egemen N, Genc Y (2001)

Thoracic pedicle: surgical anatomic evaluation and relations.

J Spinal Disord 14(1):39–45

32. Boon Tow BP, Yue WM, Srivastava A, Lai JM, Guo CM, Wearn

Peng BC, Chen JL, Yew AK, Seng C, Tan SB (2015) Does

navigation improve accuracy of placement of pedicle screws in

748 Eur Spine J (2017) 26:738–749

123

single-level lumbar degenerative spondylolisthesis? A compar-

ison between free-hand and three-dimensional O-arm navigation

techniques. J Spinal Disord Tech 28(8):E472–E477. doi:10.1097/

BSD.0b013e3182a9435e

33. Lonner B, Auerbach J, Estreicher M, Kean K (2009) Thoracic

pedicle screw instrumentation: the learning curve and evolution

in technique in the treatment of adolescent idiopathic scoliosis.

Spine (Philadelphia, Pa 1976) 34(20):2158–2164

34. Samdani A, Ranade A, Saldanha V, Yondorf M (2010) Learning

curve for placement of thoracic pedicle screws in the deformed

spine. Neurosurgery 66(2):290–294 (discussion 294)35. Gang C, Haibo L, Fancai L, Weishan C, Qixin C (2011) Learning

curve of thoracic pedicle screw placement using the free-hand

technique in scoliosis: how many screws needed for an appren-

tice? Eur Spine J 21(6):1151–1156. doi:10.1007/s00586-011-

2065-2

36. Kosmopoulos V, Schizas C, Md P (2007) Pedicle screw place-

ment accuracy: a meta-analysis. Spine (Philadelphia, Pa 1976)

32(3):E111–E120

37. Mason A, Paulsen R, Babuska JM, Rajpal S, Burneikiene S,

Nelson EL, Villavicencio AT (2014) The accuracy of pedicle

screw placement using intraoperative image guidance systems.

J Neurosurg Spine 20(2):196–203. doi:10.3171/2013.11.

spine13413

38. Shin M-H, Hur J-W, Ryu K-S, Park C-K (2015) Prospective

comparison study between the fluoroscopy-guided and navigation

coupled with O-arm—guided pedicle screw placement in the

thoracic and lumbosacral spines. Clin Spine Surg 28(6):E347–

E351. doi:10.1097/BSD.0b013e31829047a7

39. Tian N-F, Xu H-Z (2009) Image-guided pedicle screw insertion

accuracy: a meta-analysis. Int Orthop 33(4):895–903. doi:10.

1007/s00264-009-0792-3

40. Meng XT, Guan XF, Zhang HL, He SS (2016) Computer navi-

gation vs. fluoroscopy-guided navigation for thoracic pedicle

screw placement: a meta-analysis. Neurosurg Rev

39(3):385–391. doi:10.1007/s10143-015-0679-2

41. Ul Haque M, Shufflebarger H, O’Brien M, Macagno A (2006)

Radiation exposure during pedicle screw placement in adolescent

idiopathic scoliosis: is fluoroscopy safe? Spine (Philadelphia, Pa

1976) 31(21):2516–2520

42. Srinivasan D, Than K, Wang A, La Marca F, Wang P, Scher-

merhorn T, Park P (2014) Radiation safety and spine surgery:

systematic review of exposure limits and methods to minimize

radiation exposure. World Neurosurg 82(6):1337–1343

43. Villard J, Ryang YM, Demetriades AK, Reinke A, Behr M,

Preuss A, Meyer B, Ringel F (2014) Radiation exposure to the

surgeon and the patient during posterior lumbar spinal instru-

mentation: a prospective randomized comparison of navigated vs.

non-navigated freehand techniques. Spine (Phila Pa 1976)

39(13):1004–1009. doi:10.1097/brs.0000000000000351

44. Smith H, Welsch M, Sasso R, Vaccaro A (2008) Comparison of

radiation exposure in lumbar pedicle screw placement with flu-

oroscopy vs computer-assisted image guidance with intraopera-

tive three-dimensional imaging. J Spinal Cord Med

31(5):532–537

45. Fogelson JL, Krauss WE (2015) Are pedicle screws dangerous?

World Neurosurg 83(5):744–746. doi:10.1016/j.wneu.2014.09.

019

Eur Spine J (2017) 26:738–749 749

123

本文献由“学霸图书馆-文献云下载”收集自网络，仅供学习交流使用。

学霸图书馆（www.xuebalib.com）是一个“整合众多图书馆数据库资源，

提供一站式文献检索和下载服务”的24 小时在线不限IP

图书馆。

图书馆致力于便利、促进学习与科研，提供最强文献下载服务。

图书馆导航：

图书馆首页 文献云下载 图书馆入口 外文数据库大全 疑难文献辅助工具