- Email: [email protected]
Rigid Patient Positioning is Unreliable in Total Hip Arthroplasty
Accepted Manuscript Rigid Patient Positioning is Unreliable in Total Hip Arthroplasty Michael T. Milone, MD, Ran Schwarzkopf, MD, MSc, Patrick A. Meere, MD, Kaitlin Carroll, BS, Seth Jerabek, MD, Jonathan Vigdorchik, MD PII:
S0883-5403(16)30923-8
DOI:
10.1016/j.arth.2016.12.038
Reference:
YARTH 55569
To appear in:
The Journal of Arthroplasty
Received Date: 27 August 2016 Revised Date:
10 December 2016
Accepted Date: 19 December 2016
Please cite this article as: Milone MT, Schwarzkopf R, Meere PA, Carroll K, Jerabek S, Vigdorchik J, Rigid Patient Positioning is Unreliable in Total Hip Arthroplasty, The Journal of Arthroplasty (2017), doi: 10.1016/j.arth.2016.12.038. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Rigid Patient Positioning is Unreliable in Total Hip Arthroplasty Michael T. Milone MDA*, Ran Schwarzkopf MD MScA, Patrick A. Meere MDA, Kaitlin Carroll BSB, Seth Jerabek MDB, Jonathan Vigdorchik MDA A
NYU Langone Orthopaedics, Hospital for Joint Diseases, 301 East 17th Street, New York, 10003 [email protected] [email protected] *[email protected]
Hospital for Special Surgery, 535 East 70th Street, New York, 10021 [email protected] [email protected]
M AN U
*Corresponding Author – Jonathan Vigdorchik, MD
SC
B
RI PT
[email protected]
AC C
EP
TE D
Investigation performed at the NYU Langone Hospital for Joint Diseases, New York, NY, USA
1
ACCEPTED MANUSCRIPT 1
Rigid Patient Positioning is Unreliable in Total Hip Arthroplasty
2 ABSTRACT
4
Background
5
To our knowledge, no study has assessed the ability of rigid patient positioning devices to
6
afford arthroplasty surgeons with ideal acetabular orientation throughout surgery.
7
purpose of this study is to utilize robotic-arm assisted computer navigation to assess the
8
reliability of pelvic position in total hip arthroplasty performed on patients positioned with
9
rigid positioning devices.
The
M AN U
SC
RI PT
3
10 Methods
12
A prospective cohort of 100 hips (94 patients) underwent robotic guided total hip arthroplasty
13
in the lateral decubitus position from the posterior approach; 77 stabilized by universal lateral
14
positioner, and 23 by peg board.
15
generated true values of pelvic anteversion and inclination based on the position of the robot
16
arm registered to the patient’s preoperative pelvic CT.
TE D
11
EP
17
Prior to reaming, CT-templated computer software
Results
19
Mean alteration in anteversion and inclination values were 1.7º (absolute value 5.3º, range -
20
20 - 20º) and 1.6º (absolute value 2.6º, range -8 - 10º) respectively. 22% of anteversion
21
values were altered by >10º; 41% by > 5º. There was no difference between hip positioners
22
used (p=0.36). Anteversion variability was correlated with BMI (p=0.02).
AC C
18
23 24
Conclusion
2
ACCEPTED MANUSCRIPT 25
Despite the use of rigid patient positioning devices - a lateral hip positioner or peg board –
26
this study reveals clinically important malposition of the pelvis in many cases, especially with
27
regards to anteversion.
28
anatomic landmarks or computer assisted techniques to assure accurate acetabular cup
29
positioning. Patient positioning should not be solely trusted.
30
RI PT
These results show a clear need to pay particular attention to
Keywords
32
primary hip arthroplasty; patient positioning; component positioning; acetabular orientation;
33
computer navigation
M AN U
SC
31
34 35
MANUSCRIPT
36
Introduction
Acetabular component position is integral to successful total hip arthroplasty. Hip
38
instability is the leading cause for revision total hip arthroplasty, accounting for 23% of all
39
revision procedures.[1] Malpositioned acetabular components also accelerate polyethylene
40
wear and subsequent bearing surface complications.[2]
EP
TE D
37
In 1978, Lewinnek described an acetabular component safe zone of 40±10° of
42
inclination and 15±10º of anteversion.[3] This landmark paper has guided surgeons for
43
decades; however, in 2011, Callanan et al. revealed that, in a review of 1823 consecutive
44
primary total hip arthroplasties performed at a prominent tertiary institution, only 59.3% of
45
cups were positioned within the target safe zone.[4]
AC C
41
46
Jolles et al. showed that when placing 150 cups by 10 surgeons in 10 identical plastic
47
pelvis models, freehand placement led to 10° of anteversion and 3.5° of inclination error
48
despite fixing the model directly to the table with a vice.[5]
3
ACCEPTED MANUSCRIPT 49
In addition to human error, variations in acetabular orientation at time of impaction
50
further predispose surgeons to acetabular component malpositioning.
The ideal lateral
51
decubitus position places the patient both parallel and perpendicular to the table/floor without
52
pelvic tilt, rotation, or obliquity.
53
acetabular orientation. Pelvic tilt defines the relationship between the anterior pelvic plane,
54
defined as the plane connecting the anterior superior iliac spines and the pubic symphysis,
55
and the coronal plane of the patient.[6] Each degree of anterior pelvic tilt towards an inlet
56
view results in approximately 0.7-0.8 degrees of decreased acetabular anteversion and has
57
variable effects on inclination.[7-9] Axial pelvic rotation also impacts both anteversion and
58
inclination,[10] while pelvic obliquity, as defined as elevation or depression of a hemipelvis,
59
directly alters acetabular inclination.
M AN U
SC
RI PT
Alteration in any of these parameters can influence
Despite the advent of accurate computer-assisted techniques[11], most surgeons rely
61
on anatomic landmarks or mechanical guides to avoid acetabular component malposition.
62
Archbold et al described utilization of the transverse acetabular ligament,[12] but others have
63
only been able to identify the ligament in 47% of patients intra-operatively, and cup
64
positioning was not improved in this subset of patients.[13] Referencing the transverse
65
acetabular notch and anterior acetabular notch has also been described;[14] however, this
66
more complicated technique requires pre-operative CT images, a functional pelvic plane, and
67
intra-operative mathematical formulas. For these reasons, the method has not been shown to
68
be reproducible in less-experienced hands nor is it widely implemented.
EP
AC C
69
TE D
60
Mechanical guides, although more practically implemented, fail to consider variations
70
in pelvic orientation. DiGioia et al. showed that reliance on mechanical guides results in a
71
mean of 18° of anteversion variation between actual and desired orientation primarily due to
72
variation in patient positioning at time of impaction.[15] This study, however, used a self-
73
stated suboptimal bean bag positioning device, and more rigid patient positioning devices
4
ACCEPTED MANUSCRIPT 74
may allow for more accurate free hand cup placement by more firmly fixing the patient’s
75
pelvis to the operating table. To our knowledge, no study has assessed the ability of rigid
76
devices to afford arthroplasty surgeons with ideal pelvic positioning throughout surgery. The purpose of this study is to utilize robotic-arm assisted computer navigation to
78
assess the reliability of pelvic position in total hip arthroplasty performed on patients
79
positioned with a rigid lateral hip positioner or peg-board device.
80
pelvic and acetabular orientation will be inconsistent, yielding another source of error
81
contributing to cup malpositioning.
RI PT
77
SC
82 Materials and Methods
M AN U
83
We hypothesize that
One hundred consecutive hips (94 patients) were prospectively enrolled after being
85
indicated for CT-guided robotic arm assisted total hip arthroplasty with Mako robotic arm
86
navigation system (Robotic Arm Interactive Orthopaedic System, Stryker Orthopaedics,
87
Mahwah, NJ, USA).[16,17] All procedures were performed via a mini posterior approach
88
with the patient in the lateral decubitus position.
89
universal lateral positioner (Innovative Medical Products, Plainville, CT), and last 23 by rigid
90
peg board fixation (Innomed, Savannah, GA). As routinely performed, free-hand level was
91
used to optimize patient positioning prior to prepping and draping. The goal of this technique
92
was to minimize variations in sagittal tilt, axial rotation, and coronal obliquity relative to the
93
neutral operating table. To achieve this, the level was aligned with the lateral aspect of the
94
hip, gluteal folds, and spinopelvic junction.
The first 77 hips were stabilized by
AC C
EP
TE D
84
95
After dislocation and robotic registration, but prior to reaming, one fellowship trained
96
arthroplasty surgeon manually placed the robotic arm parallel to both the longitudinal axis of
97
the patient and the horizontal surface of the operating table, which, if the pelvis were oriented
98
perfectly, would represent 0º of anteversion and 0º of inclination (Figure 1). Although there
5
ACCEPTED MANUSCRIPT
is potential for error with malplacement of the robotic arm, we believe this error is small
100
because the arm’s default position is parallel to both the ground (horizontal aspect of table)
101
and parallel to the operating table, as the Mako base is parked perpendicular to the table
102
(longitudinal axis of the patient). The CT-templated computer software then generated true
103
values of this perceived zero° of anteversion and inclination based on the position of the
104
robotic arm, which was registered to the patient’s preoperative pelvic CT.
105
variations in acetabular orientation due to pelvic malpositioning are represented by these
106
robotic navigation generated values. The robotic arm was then placed in the robotic arm
107
determination of 0º of anteversion and 0º of inclination, and measurements were taken with a
108
sterile protractor compared to the floor and longitudinal axis of the patient.
RI PT
99
M AN U
SC
Therefore,
To test the assumption that variations in acetabular orientation were due to patient
110
positioning and not errors in registration of the pelvis to the CT scan, and to assure the
111
accuracy of robotic measurements, cup anteversion and inclination at times of impaction
112
were recorded and compared to those calculated via Lewinnek’s trigonometric ellipse method
113
on standardized 3-months postoperative supine AP pelvis x-rays.[3] Supine x-rays were
114
chosen to ensure consistency of measurements because the Mako robotic values are recorded
115
using the supine coronal plane of the body in the CT scanner.
116
Statistical analysis
EP
AC C
117
TE D
109
Statistical analyses were performed utilizing Stata Statistical Software: Release 11
118
(StataCorp LP, College Station, TX, USA).
119
assess the significance (threshold p<0.05) of recorded anteversion and inclination errors as
120
well as differences across rigid positioner types and body mass index subgroups. Similarly,
121
paired t-tests confirmed consistency between robotic navigation acetabular cup position at
122
impaction and post-operative X-ray position.
123
More specifically, t-tests were employed to
6
ACCEPTED MANUSCRIPT 124
Results
125 126
48% of enrolled patients were male. The mean age was 52.7 ± 13 years (range 22-78 years) and mean BMI was 29.4 ± 6.5 kg/m2 (range 19.2-51.4). The mean variations in anteversion and inclination values were 1.7º (absolute value
128
5.3º, range -200 - 20º) and 1.6º (absolute value 2.6º, range -8° - 10º) respectively (Table 1).
129
Hypothesis testing revealed that these anteversion and inclination errors were different from
130
zero (p=0.02 and p=0.0001, respectively). 41% of anteversion values were altered by 5 or
131
more°, and 22% were off by at least 10° (Figure 2). 18% and 2% of inclination values were
132
altered by >5 and >10º, respectively (Figure 3).
M AN U
SC
RI PT
127
Paired t-tests showed no difference in the absolute value of acetabular anteversion
134
(p=0.36) or inclination (p=0.59) for the two types of positioners used in this study.
135
Regression analysis revealed that anteversion differences were correlated with BMI (p=0.02),
136
but inclination differences were not (p=0.09). The mean absolute anteversion error of 6.4º
137
for the top 50th percentile of patients sorted by BMI, those with a BMI > 27.5 mg/kg2, was
138
higher (p=0.02) than the mean anteversion error of 4.1º for the bottom 50th percentile (Table
139
1).
TE D
133
Robotic navigation acetabular cup anteversion at impaction (mean 21.8º) was not
141
different from postoperative X-ray anteversion (mean 21.9º) (p=0.50), nor was robotic
142
navigation acetabular cup inclination (mean 40.6º) different from postoperative X-ray
143
inclination (mean 40.5º)(p=0.34) (Table 2)
145
AC C
144
EP
140
Discussion
146
Accurate acetabular component placement may be altered by variations in patient
147
positioning at the time of cup impaction. Despite this, no study has assessed the ability of
148
rigid patient positioners, which attempt to stabilize the pelvis, to afford surgeons with ideal
7
ACCEPTED MANUSCRIPT 149
pelvic positioning intra-operatively. The purpose of this study is to utilize robotic-arm
150
assisted computer navigation to assess the reliability of acetabular orientation in patients
151
undergoing total hip arthroplasty in the lateral decubitus position stabilized by rigid lateral
152
positioner or peg board devices. This study is not without limitations. First, a single surgeon manually placed the
154
robotic arm at what he perceived to be parallel to the longitudinal axis of the patient and the
155
horizontal surface of the operating table/floor. Because this was performed free hand, this
156
may account for some error in perceived values of inclination and anteversion. However, the
157
robotic arm provides a long reference frame to facilitate accurate placement, and divergence
158
from parallelism is more easily appreciated than variation from non-right angles.
159
Additionally, the robotic arm defaults to a neutral position as described in the methods
160
section of this paper. For these reasons, this error is likely small and wound not account for
161
the magnitude of this study’s findings. Second, no intra-operative measurements of pelvic
162
orientation were recorded in this study.
163
presumed to be secondary to discrepancies in pelvic positioning on the operative table. Still,
164
we are not able to identify other variables that would alter the acetabular measurements,
165
given that the navigation system was accurate as evidenced by the postoperative x-ray
166
evaluation. Thirdly, we used two different types of patient positioners in unequal amounts.
167
This was limited by clinical availability. However, both firmly position the patient by
168
providing fixed mechanical blocks to both the anterior and posterior pelvis.
SC
M AN U
EP
TE D
All alterations in acetabular orientation were
AC C
169
RI PT
153
Despite these limitations, this study provides important information about the
170
shortcomings of rigid patient positioning in total hip arthroplasty. The devices used in this
171
study failed to assure consistent acetabular orientation throughout surgery, especially with
172
regards to anteversion. Although the mean version difference of 5.3° is superior to the 18°
173
mean variation reported by DiGioia’s bean bag positioning,[15] reliance on rigid positioners
8
ACCEPTED MANUSCRIPT 174
still would have resulted in unacceptable acetabular alignment in 22% of hips.
This
175
percentage of patients were positioned in such a way that anteversion was altered by more
176
than 10°, the acceptable variability in described safe zones.[3,4]
177
variable, which is intuitive, as intra-operative hip manipulation relies on flexion and
178
extension but to a lesser extent abduction and adduction. Moreover, changes in pelvic tilt
179
have not been shown to have a reproducible effect on acetabular inclination.[10]
180
somewhat expected, there was a relationship between anteversion differences and BMI.
181
Excess body habitus not only obscures anatomic landmarks utilized in initial patient
182
positioning but also inhibits firm juxtaposition to bony prominences.
183
make an attempt to identify the pelvic parameters responsible for changes in acetabular
184
orientation, and future investigations could seek to delineate common errors with hip
185
positioners.
As
SC
RI PT
Inclination was less
M AN U
This study does not
Consistent with the literature,[11,18] the MAKO robotic arm navigation system
187
accurately navigate acetabular cup positioning in total hip arthroplasty, as evidenced by the
188
robot-measured impaction and 3-month post-operative X-ray anteversion and inclination
189
values. The largest single difference for both cup anteversion and inclination was 4°. The
190
CT based registration of patient anatomy by the MAKO robotic arm navigation system
191
accommodates for known pelvic tilt and permits adjustment of acetabular anteversion
192
appropriately.[16] Importantly, this ensures that we utilized accurate “true” navigated values
193
for zero° of cup anteversion and inclination for the purpose of comparison to the surgeon’s
194
determined zero values.
AC C
EP
TE D
186
195
The results of this study help explain acetabular component malpositioning. In 1978,
196
Lewinnek described the first safe zone centred at 40° of inclination and 15° of
197
anteversion, with 20 degree ranges for both parameters[3]. Over 20 years later,
198
Callaan et al. revealed that arthroplasty surgeons are missing the acetabular target
9
ACCEPTED MANUSCRIPT
over 40% of the time[4]. Human error has been shown to be partly responsible,
200
accounting for 10° of variability in cup positioning even when the pelvic model is
201
fixed, and directly visualized[5]. Now, this study shows that rigid pelvic positioners
202
complicate an already troublesome situation by presenting up to 20° of variability in
203
acetabular orientation.
204
Nonetheless, this paper highlights a clear need to pay particular attention to anatomic
205
landmarks or use computer assisted techniques, and not solely rely on positioning guides. All
206
of these strategies account for pelvic orientation to assure accurate acetabular component
207
placement.
SC
RI PT
199
Patient positioning should not be trusted alone.
M AN U
208 209 210 211
Figures and Tables
Table 1. Summary statistics of absolute variations in acetabular orientation
Anteversion
Mean ± S.D.
Range
N
5.3 ± 5.3º
0 - 20
100
5.0 ± 5.0º
0 - 20
77
EP
Variable
AC C
213
TE D
212
Lateral Positioner
p-value*
0.36
Peg Board
6.2 ± 6.1º
0 - 20
23
BMI > 27.5
6.4 ± 5.9º
0 - 20
50 0.02
BMI < 27.5 Inclination Lateral Positioner
4.1 ± 4.3º
0 - 15
50
2.6 ± 2.3º
0 - 10
100
2.6 ± 2.4º
0 - 10
77
0.59
10
ACCEPTED MANUSCRIPT Peg Board
2.9 ± 2.1º
0-8
23
BMI > 27.5
2.9 ± 2.5º
0 - 10
48
BMI < 27.5
2.4 ± 2.1º
0-8
52
0.16
*p-values determined by t-tests, significance set at <0.05
RI PT
214 215
SC
216 217
M AN U
218 219 220
TE D
221
Table 2. Final cup position at time of impaction and 3 months post-operatively (N=100)
Anteversion
AC C
Impaction*
Mean ± S.D.
EP
Variable
Post-operative X-Ray
p-value
21.8 ± 1.8º 0.5
21.9 ± 1.9º
Inclination
Impaction*
40.6 ± 1.2º 0.34
Post-operative X-Ray
40.5 ± 1.3º
*values recorded by CT-guided robotic navigation system intra-operatively
11
ACCEPTED MANUSCRIPT 222 223 224
RI PT
225 226
228
EP
TE D
M AN U
SC
227
Figure 1: Robotic arm positioned parallel to longitudinal axis of patient and horizontal
230
surface of operating table (left). In this case, the patient is malpositioned, and actual values
231
of arm inclination and anteversion are nonzero.
232
inclination and anteversion, confirming malpositioning (right).
233
AC C
229
Robotic arm positioned at zero° of
12
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
234 235
Figure 2: Histogram of variation in acetabular anteversion due to patient positioning. 40%
236
of patients altered by ± 5º; 22% by ± 10º.
EP AC C
238
TE D
237
13
239
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
Figure 3: Histogram of variation in acetabular inclination due to patient positioning. 18% of
241
patients altered by ± 5º; 2% by ± 10º.
242 243
References
244
1.
TE D
240
EP
Bozic KJ, Kurtz SM, Lau E, Ong K, Vail TP, Berry DJ. The epidemiology of revision total hip arthroplasty in the United States. The Journal of bone and joint surgery.
246
American volume. Jan 2009;91(1):128-133.
247
2.
248
Patil S, Bergula A, Chen PC, Colwell CW, Jr., D'Lima DD. Polyethylene wear and
acetabular component orientation. The Journal of bone and joint surgery. American
249 250
AC C
245
volume. 2003;85-A Suppl 4:56-63.
3.
Lewinnek GE, Lewis JL, Tarr R, Compere CL, Zimmerman JR. Dislocations after
251
total hip-replacement arthroplasties. The Journal of bone and joint surgery. American
252
volume. Mar 1978;60(2):217-220.
14
ACCEPTED MANUSCRIPT 253
4.
Callanan MC, Jarrett B, Bragdon CR, et al. The John Charnley Award: risk factors for
254
cup malpositioning: quality improvement through a joint registry at a tertiary hospital.
255
Clinical orthopaedics and related research. Feb 2011;469(2):319-329.
256
5.
Jolles BM, Genoud P, Hoffmeyer P. Computer-assisted cup placement techniques in total hip arthroplasty improve accuracy of placement. Clinical orthopaedics and
258
related research. Sep 2004(426):174-179.
259
6.
RI PT
257
Babisch JW, Layher F, Amiot LP. The rationale for tilt-adjusted acetabular cup navigation. The Journal of bone and joint surgery. American volume. Feb
261
2008;90(2):357-365. 7.
263 264
Lembeck B, Mueller O, Reize P, Wuelker N. Pelvic tilt makes acetabular cup
M AN U
262
SC
260
navigation inaccurate. Acta orthopaedica. Aug 2005;76(4):517-523. 8.
Dandachli W, Ul Islam S, Richards R, Hall-Craggs M, Witt J. The influence of pelvic tilt on acetabular orientation and cover: a three-dimensional computerised
266
tomography analysis. Hip international : the journal of clinical and experimental
267
research on hip pathology and therapy. Jan-Feb 2013;23(1):87-92.
268
9.
TE D
265
Maratt JD, Esposito CI, McLawhorn AS, Jerabek SA, Padgett DE, Mayman DJ. Pelvic tilt in patients undergoing total hip arthroplasty: when does it matter? The
270
Journal of arthroplasty. Mar 2015;30(3):387-391. 10.
272
C RM, Spitzer AI. How does pelvic rotation or tilt affect radiographic measurement of acetabular component inclination angle during THA? Journal of orthopaedics. Dec
273 274
AC C
271
EP
269
2015;12(4):222-227.
11.
Domb BG, El Bitar YF, Sadik AY, Stake CE, Botser IB. Comparison of robotic-
275
assisted and conventional acetabular cup placement in THA: a matched-pair
276
controlled study. Clinical orthopaedics and related research. Jan 2014;472(1):329-
277
336.
15
ACCEPTED MANUSCRIPT 278
12.
Archbold HA, Mockford B, Molloy D, McConway J, Ogonda L, Beverland D. The transverse acetabular ligament: an aid to orientation of the acetabular component
280
during primary total hip replacement: a preliminary study of 1000 cases investigating
281
postoperative stability. The Journal of bone and joint surgery. British volume. Jul
282
2006;88(7):883-886.
283
13.
RI PT
279
Epstein NJ, Woolson ST, Giori NJ. Acetabular component positioning using the transverse acetabular ligament: can you find it and does it help? Clinical orthopaedics
285
and related research. Feb 2011;469(2):412-416.
286
14.
SC
284
Ha YC, Yoo JJ, Lee YK, Kim JY, Koo KH. Acetabular component positioning using anatomic landmarks of the acetabulum. Clinical orthopaedics and related research.
288
Dec 2012;470(12):3515-3523.
289
15.
M AN U
287
Digioia AM, 3rd, Jaramaz B, Plakseychuk AY, et al. Comparison of a mechanical acetabular alignment guide with computer placement of the socket. The Journal of
291
arthroplasty. Apr 2002;17(3):359-364. 16.
294
platform. Current reviews in musculoskeletal medicine. Sep 2011;4(3):151-156. 17.
295
robotic-assisted arthroplasty. Surgical technology international. Mar 2014;24:302-
296 297 298 299 300 301
Werner SD, Stonestreet M, Jacofsky DJ. Makoplasty and the accuracy and efficacy of
EP
293
Tarwala R, Dorr LD. Robotic assisted total hip arthroplasty using the MAKO
AC C
292
TE D
290
306.
18.
Elson L, Dounchis J, Illgen R, et al. Precision of acetabular cup placement in robotic integrated total hip arthroplasty. Hip international : the journal of clinical and experimental research on hip pathology and therapy. Nov 25 2015;25(6):531-536.
S0883-5403(16)30923-8
DOI:
10.1016/j.arth.2016.12.038
Reference:
YARTH 55569
To appear in:
The Journal of Arthroplasty
Received Date: 27 August 2016 Revised Date:
10 December 2016
Accepted Date: 19 December 2016
Please cite this article as: Milone MT, Schwarzkopf R, Meere PA, Carroll K, Jerabek S, Vigdorchik J, Rigid Patient Positioning is Unreliable in Total Hip Arthroplasty, The Journal of Arthroplasty (2017), doi: 10.1016/j.arth.2016.12.038. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Rigid Patient Positioning is Unreliable in Total Hip Arthroplasty Michael T. Milone MDA*, Ran Schwarzkopf MD MScA, Patrick A. Meere MDA, Kaitlin Carroll BSB, Seth Jerabek MDB, Jonathan Vigdorchik MDA A
NYU Langone Orthopaedics, Hospital for Joint Diseases, 301 East 17th Street, New York, 10003 [email protected] [email protected] *[email protected]
Hospital for Special Surgery, 535 East 70th Street, New York, 10021 [email protected] [email protected]
M AN U
*Corresponding Author – Jonathan Vigdorchik, MD
SC
B
RI PT
[email protected]
AC C
EP
TE D
Investigation performed at the NYU Langone Hospital for Joint Diseases, New York, NY, USA
1
ACCEPTED MANUSCRIPT 1
Rigid Patient Positioning is Unreliable in Total Hip Arthroplasty
2 ABSTRACT
4
Background
5
To our knowledge, no study has assessed the ability of rigid patient positioning devices to
6
afford arthroplasty surgeons with ideal acetabular orientation throughout surgery.
7
purpose of this study is to utilize robotic-arm assisted computer navigation to assess the
8
reliability of pelvic position in total hip arthroplasty performed on patients positioned with
9
rigid positioning devices.
The
M AN U
SC
RI PT
3
10 Methods
12
A prospective cohort of 100 hips (94 patients) underwent robotic guided total hip arthroplasty
13
in the lateral decubitus position from the posterior approach; 77 stabilized by universal lateral
14
positioner, and 23 by peg board.
15
generated true values of pelvic anteversion and inclination based on the position of the robot
16
arm registered to the patient’s preoperative pelvic CT.
TE D
11
EP
17
Prior to reaming, CT-templated computer software
Results
19
Mean alteration in anteversion and inclination values were 1.7º (absolute value 5.3º, range -
20
20 - 20º) and 1.6º (absolute value 2.6º, range -8 - 10º) respectively. 22% of anteversion
21
values were altered by >10º; 41% by > 5º. There was no difference between hip positioners
22
used (p=0.36). Anteversion variability was correlated with BMI (p=0.02).
AC C
18
23 24
Conclusion
2
ACCEPTED MANUSCRIPT 25
Despite the use of rigid patient positioning devices - a lateral hip positioner or peg board –
26
this study reveals clinically important malposition of the pelvis in many cases, especially with
27
regards to anteversion.
28
anatomic landmarks or computer assisted techniques to assure accurate acetabular cup
29
positioning. Patient positioning should not be solely trusted.
30
RI PT
These results show a clear need to pay particular attention to
Keywords
32
primary hip arthroplasty; patient positioning; component positioning; acetabular orientation;
33
computer navigation
M AN U
SC
31
34 35
MANUSCRIPT
36
Introduction
Acetabular component position is integral to successful total hip arthroplasty. Hip
38
instability is the leading cause for revision total hip arthroplasty, accounting for 23% of all
39
revision procedures.[1] Malpositioned acetabular components also accelerate polyethylene
40
wear and subsequent bearing surface complications.[2]
EP
TE D
37
In 1978, Lewinnek described an acetabular component safe zone of 40±10° of
42
inclination and 15±10º of anteversion.[3] This landmark paper has guided surgeons for
43
decades; however, in 2011, Callanan et al. revealed that, in a review of 1823 consecutive
44
primary total hip arthroplasties performed at a prominent tertiary institution, only 59.3% of
45
cups were positioned within the target safe zone.[4]
AC C
41
46
Jolles et al. showed that when placing 150 cups by 10 surgeons in 10 identical plastic
47
pelvis models, freehand placement led to 10° of anteversion and 3.5° of inclination error
48
despite fixing the model directly to the table with a vice.[5]
3
ACCEPTED MANUSCRIPT 49
In addition to human error, variations in acetabular orientation at time of impaction
50
further predispose surgeons to acetabular component malpositioning.
The ideal lateral
51
decubitus position places the patient both parallel and perpendicular to the table/floor without
52
pelvic tilt, rotation, or obliquity.
53
acetabular orientation. Pelvic tilt defines the relationship between the anterior pelvic plane,
54
defined as the plane connecting the anterior superior iliac spines and the pubic symphysis,
55
and the coronal plane of the patient.[6] Each degree of anterior pelvic tilt towards an inlet
56
view results in approximately 0.7-0.8 degrees of decreased acetabular anteversion and has
57
variable effects on inclination.[7-9] Axial pelvic rotation also impacts both anteversion and
58
inclination,[10] while pelvic obliquity, as defined as elevation or depression of a hemipelvis,
59
directly alters acetabular inclination.
M AN U
SC
RI PT
Alteration in any of these parameters can influence
Despite the advent of accurate computer-assisted techniques[11], most surgeons rely
61
on anatomic landmarks or mechanical guides to avoid acetabular component malposition.
62
Archbold et al described utilization of the transverse acetabular ligament,[12] but others have
63
only been able to identify the ligament in 47% of patients intra-operatively, and cup
64
positioning was not improved in this subset of patients.[13] Referencing the transverse
65
acetabular notch and anterior acetabular notch has also been described;[14] however, this
66
more complicated technique requires pre-operative CT images, a functional pelvic plane, and
67
intra-operative mathematical formulas. For these reasons, the method has not been shown to
68
be reproducible in less-experienced hands nor is it widely implemented.
EP
AC C
69
TE D
60
Mechanical guides, although more practically implemented, fail to consider variations
70
in pelvic orientation. DiGioia et al. showed that reliance on mechanical guides results in a
71
mean of 18° of anteversion variation between actual and desired orientation primarily due to
72
variation in patient positioning at time of impaction.[15] This study, however, used a self-
73
stated suboptimal bean bag positioning device, and more rigid patient positioning devices
4
ACCEPTED MANUSCRIPT 74
may allow for more accurate free hand cup placement by more firmly fixing the patient’s
75
pelvis to the operating table. To our knowledge, no study has assessed the ability of rigid
76
devices to afford arthroplasty surgeons with ideal pelvic positioning throughout surgery. The purpose of this study is to utilize robotic-arm assisted computer navigation to
78
assess the reliability of pelvic position in total hip arthroplasty performed on patients
79
positioned with a rigid lateral hip positioner or peg-board device.
80
pelvic and acetabular orientation will be inconsistent, yielding another source of error
81
contributing to cup malpositioning.
RI PT
77
SC
82 Materials and Methods
M AN U
83
We hypothesize that
One hundred consecutive hips (94 patients) were prospectively enrolled after being
85
indicated for CT-guided robotic arm assisted total hip arthroplasty with Mako robotic arm
86
navigation system (Robotic Arm Interactive Orthopaedic System, Stryker Orthopaedics,
87
Mahwah, NJ, USA).[16,17] All procedures were performed via a mini posterior approach
88
with the patient in the lateral decubitus position.
89
universal lateral positioner (Innovative Medical Products, Plainville, CT), and last 23 by rigid
90
peg board fixation (Innomed, Savannah, GA). As routinely performed, free-hand level was
91
used to optimize patient positioning prior to prepping and draping. The goal of this technique
92
was to minimize variations in sagittal tilt, axial rotation, and coronal obliquity relative to the
93
neutral operating table. To achieve this, the level was aligned with the lateral aspect of the
94
hip, gluteal folds, and spinopelvic junction.
The first 77 hips were stabilized by
AC C
EP
TE D
84
95
After dislocation and robotic registration, but prior to reaming, one fellowship trained
96
arthroplasty surgeon manually placed the robotic arm parallel to both the longitudinal axis of
97
the patient and the horizontal surface of the operating table, which, if the pelvis were oriented
98
perfectly, would represent 0º of anteversion and 0º of inclination (Figure 1). Although there
5
ACCEPTED MANUSCRIPT
is potential for error with malplacement of the robotic arm, we believe this error is small
100
because the arm’s default position is parallel to both the ground (horizontal aspect of table)
101
and parallel to the operating table, as the Mako base is parked perpendicular to the table
102
(longitudinal axis of the patient). The CT-templated computer software then generated true
103
values of this perceived zero° of anteversion and inclination based on the position of the
104
robotic arm, which was registered to the patient’s preoperative pelvic CT.
105
variations in acetabular orientation due to pelvic malpositioning are represented by these
106
robotic navigation generated values. The robotic arm was then placed in the robotic arm
107
determination of 0º of anteversion and 0º of inclination, and measurements were taken with a
108
sterile protractor compared to the floor and longitudinal axis of the patient.
RI PT
99
M AN U
SC
Therefore,
To test the assumption that variations in acetabular orientation were due to patient
110
positioning and not errors in registration of the pelvis to the CT scan, and to assure the
111
accuracy of robotic measurements, cup anteversion and inclination at times of impaction
112
were recorded and compared to those calculated via Lewinnek’s trigonometric ellipse method
113
on standardized 3-months postoperative supine AP pelvis x-rays.[3] Supine x-rays were
114
chosen to ensure consistency of measurements because the Mako robotic values are recorded
115
using the supine coronal plane of the body in the CT scanner.
116
Statistical analysis
EP
AC C
117
TE D
109
Statistical analyses were performed utilizing Stata Statistical Software: Release 11
118
(StataCorp LP, College Station, TX, USA).
119
assess the significance (threshold p<0.05) of recorded anteversion and inclination errors as
120
well as differences across rigid positioner types and body mass index subgroups. Similarly,
121
paired t-tests confirmed consistency between robotic navigation acetabular cup position at
122
impaction and post-operative X-ray position.
123
More specifically, t-tests were employed to
6
ACCEPTED MANUSCRIPT 124
Results
125 126
48% of enrolled patients were male. The mean age was 52.7 ± 13 years (range 22-78 years) and mean BMI was 29.4 ± 6.5 kg/m2 (range 19.2-51.4). The mean variations in anteversion and inclination values were 1.7º (absolute value
128
5.3º, range -200 - 20º) and 1.6º (absolute value 2.6º, range -8° - 10º) respectively (Table 1).
129
Hypothesis testing revealed that these anteversion and inclination errors were different from
130
zero (p=0.02 and p=0.0001, respectively). 41% of anteversion values were altered by 5 or
131
more°, and 22% were off by at least 10° (Figure 2). 18% and 2% of inclination values were
132
altered by >5 and >10º, respectively (Figure 3).
M AN U
SC
RI PT
127
Paired t-tests showed no difference in the absolute value of acetabular anteversion
134
(p=0.36) or inclination (p=0.59) for the two types of positioners used in this study.
135
Regression analysis revealed that anteversion differences were correlated with BMI (p=0.02),
136
but inclination differences were not (p=0.09). The mean absolute anteversion error of 6.4º
137
for the top 50th percentile of patients sorted by BMI, those with a BMI > 27.5 mg/kg2, was
138
higher (p=0.02) than the mean anteversion error of 4.1º for the bottom 50th percentile (Table
139
1).
TE D
133
Robotic navigation acetabular cup anteversion at impaction (mean 21.8º) was not
141
different from postoperative X-ray anteversion (mean 21.9º) (p=0.50), nor was robotic
142
navigation acetabular cup inclination (mean 40.6º) different from postoperative X-ray
143
inclination (mean 40.5º)(p=0.34) (Table 2)
145
AC C
144
EP
140
Discussion
146
Accurate acetabular component placement may be altered by variations in patient
147
positioning at the time of cup impaction. Despite this, no study has assessed the ability of
148
rigid patient positioners, which attempt to stabilize the pelvis, to afford surgeons with ideal
7
ACCEPTED MANUSCRIPT 149
pelvic positioning intra-operatively. The purpose of this study is to utilize robotic-arm
150
assisted computer navigation to assess the reliability of acetabular orientation in patients
151
undergoing total hip arthroplasty in the lateral decubitus position stabilized by rigid lateral
152
positioner or peg board devices. This study is not without limitations. First, a single surgeon manually placed the
154
robotic arm at what he perceived to be parallel to the longitudinal axis of the patient and the
155
horizontal surface of the operating table/floor. Because this was performed free hand, this
156
may account for some error in perceived values of inclination and anteversion. However, the
157
robotic arm provides a long reference frame to facilitate accurate placement, and divergence
158
from parallelism is more easily appreciated than variation from non-right angles.
159
Additionally, the robotic arm defaults to a neutral position as described in the methods
160
section of this paper. For these reasons, this error is likely small and wound not account for
161
the magnitude of this study’s findings. Second, no intra-operative measurements of pelvic
162
orientation were recorded in this study.
163
presumed to be secondary to discrepancies in pelvic positioning on the operative table. Still,
164
we are not able to identify other variables that would alter the acetabular measurements,
165
given that the navigation system was accurate as evidenced by the postoperative x-ray
166
evaluation. Thirdly, we used two different types of patient positioners in unequal amounts.
167
This was limited by clinical availability. However, both firmly position the patient by
168
providing fixed mechanical blocks to both the anterior and posterior pelvis.
SC
M AN U
EP
TE D
All alterations in acetabular orientation were
AC C
169
RI PT
153
Despite these limitations, this study provides important information about the
170
shortcomings of rigid patient positioning in total hip arthroplasty. The devices used in this
171
study failed to assure consistent acetabular orientation throughout surgery, especially with
172
regards to anteversion. Although the mean version difference of 5.3° is superior to the 18°
173
mean variation reported by DiGioia’s bean bag positioning,[15] reliance on rigid positioners
8
ACCEPTED MANUSCRIPT 174
still would have resulted in unacceptable acetabular alignment in 22% of hips.
This
175
percentage of patients were positioned in such a way that anteversion was altered by more
176
than 10°, the acceptable variability in described safe zones.[3,4]
177
variable, which is intuitive, as intra-operative hip manipulation relies on flexion and
178
extension but to a lesser extent abduction and adduction. Moreover, changes in pelvic tilt
179
have not been shown to have a reproducible effect on acetabular inclination.[10]
180
somewhat expected, there was a relationship between anteversion differences and BMI.
181
Excess body habitus not only obscures anatomic landmarks utilized in initial patient
182
positioning but also inhibits firm juxtaposition to bony prominences.
183
make an attempt to identify the pelvic parameters responsible for changes in acetabular
184
orientation, and future investigations could seek to delineate common errors with hip
185
positioners.
As
SC
RI PT
Inclination was less
M AN U
This study does not
Consistent with the literature,[11,18] the MAKO robotic arm navigation system
187
accurately navigate acetabular cup positioning in total hip arthroplasty, as evidenced by the
188
robot-measured impaction and 3-month post-operative X-ray anteversion and inclination
189
values. The largest single difference for both cup anteversion and inclination was 4°. The
190
CT based registration of patient anatomy by the MAKO robotic arm navigation system
191
accommodates for known pelvic tilt and permits adjustment of acetabular anteversion
192
appropriately.[16] Importantly, this ensures that we utilized accurate “true” navigated values
193
for zero° of cup anteversion and inclination for the purpose of comparison to the surgeon’s
194
determined zero values.
AC C
EP
TE D
186
195
The results of this study help explain acetabular component malpositioning. In 1978,
196
Lewinnek described the first safe zone centred at 40° of inclination and 15° of
197
anteversion, with 20 degree ranges for both parameters[3]. Over 20 years later,
198
Callaan et al. revealed that arthroplasty surgeons are missing the acetabular target
9
ACCEPTED MANUSCRIPT
over 40% of the time[4]. Human error has been shown to be partly responsible,
200
accounting for 10° of variability in cup positioning even when the pelvic model is
201
fixed, and directly visualized[5]. Now, this study shows that rigid pelvic positioners
202
complicate an already troublesome situation by presenting up to 20° of variability in
203
acetabular orientation.
204
Nonetheless, this paper highlights a clear need to pay particular attention to anatomic
205
landmarks or use computer assisted techniques, and not solely rely on positioning guides. All
206
of these strategies account for pelvic orientation to assure accurate acetabular component
207
placement.
SC
RI PT
199
Patient positioning should not be trusted alone.
M AN U
208 209 210 211
Figures and Tables
Table 1. Summary statistics of absolute variations in acetabular orientation
Anteversion
Mean ± S.D.
Range
N
5.3 ± 5.3º
0 - 20
100
5.0 ± 5.0º
0 - 20
77
EP
Variable
AC C
213
TE D
212
Lateral Positioner
p-value*
0.36
Peg Board
6.2 ± 6.1º
0 - 20
23
BMI > 27.5
6.4 ± 5.9º
0 - 20
50 0.02
BMI < 27.5 Inclination Lateral Positioner
4.1 ± 4.3º
0 - 15
50
2.6 ± 2.3º
0 - 10
100
2.6 ± 2.4º
0 - 10
77
0.59
10
ACCEPTED MANUSCRIPT Peg Board
2.9 ± 2.1º
0-8
23
BMI > 27.5
2.9 ± 2.5º
0 - 10
48
BMI < 27.5
2.4 ± 2.1º
0-8
52
0.16
*p-values determined by t-tests, significance set at <0.05
RI PT
214 215
SC
216 217
M AN U
218 219 220
TE D
221
Table 2. Final cup position at time of impaction and 3 months post-operatively (N=100)
Anteversion
AC C
Impaction*
Mean ± S.D.
EP
Variable
Post-operative X-Ray
p-value
21.8 ± 1.8º 0.5
21.9 ± 1.9º
Inclination
Impaction*
40.6 ± 1.2º 0.34
Post-operative X-Ray
40.5 ± 1.3º
*values recorded by CT-guided robotic navigation system intra-operatively
11
ACCEPTED MANUSCRIPT 222 223 224
RI PT
225 226
228
EP
TE D
M AN U
SC
227
Figure 1: Robotic arm positioned parallel to longitudinal axis of patient and horizontal
230
surface of operating table (left). In this case, the patient is malpositioned, and actual values
231
of arm inclination and anteversion are nonzero.
232
inclination and anteversion, confirming malpositioning (right).
233
AC C
229
Robotic arm positioned at zero° of
12
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
234 235
Figure 2: Histogram of variation in acetabular anteversion due to patient positioning. 40%
236
of patients altered by ± 5º; 22% by ± 10º.
EP AC C
238
TE D
237
13
239
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
Figure 3: Histogram of variation in acetabular inclination due to patient positioning. 18% of
241
patients altered by ± 5º; 2% by ± 10º.
242 243
References
244
1.
TE D
240
EP
Bozic KJ, Kurtz SM, Lau E, Ong K, Vail TP, Berry DJ. The epidemiology of revision total hip arthroplasty in the United States. The Journal of bone and joint surgery.
246
American volume. Jan 2009;91(1):128-133.
247
2.
248
Patil S, Bergula A, Chen PC, Colwell CW, Jr., D'Lima DD. Polyethylene wear and
acetabular component orientation. The Journal of bone and joint surgery. American
249 250
AC C
245
volume. 2003;85-A Suppl 4:56-63.
3.
Lewinnek GE, Lewis JL, Tarr R, Compere CL, Zimmerman JR. Dislocations after
251
total hip-replacement arthroplasties. The Journal of bone and joint surgery. American
252
volume. Mar 1978;60(2):217-220.
14
ACCEPTED MANUSCRIPT 253
4.
Callanan MC, Jarrett B, Bragdon CR, et al. The John Charnley Award: risk factors for
254
cup malpositioning: quality improvement through a joint registry at a tertiary hospital.
255
Clinical orthopaedics and related research. Feb 2011;469(2):319-329.
256
5.
Jolles BM, Genoud P, Hoffmeyer P. Computer-assisted cup placement techniques in total hip arthroplasty improve accuracy of placement. Clinical orthopaedics and
258
related research. Sep 2004(426):174-179.
259
6.
RI PT
257
Babisch JW, Layher F, Amiot LP. The rationale for tilt-adjusted acetabular cup navigation. The Journal of bone and joint surgery. American volume. Feb
261
2008;90(2):357-365. 7.
263 264
Lembeck B, Mueller O, Reize P, Wuelker N. Pelvic tilt makes acetabular cup
M AN U
262
SC
260
navigation inaccurate. Acta orthopaedica. Aug 2005;76(4):517-523. 8.
Dandachli W, Ul Islam S, Richards R, Hall-Craggs M, Witt J. The influence of pelvic tilt on acetabular orientation and cover: a three-dimensional computerised
266
tomography analysis. Hip international : the journal of clinical and experimental
267
research on hip pathology and therapy. Jan-Feb 2013;23(1):87-92.
268
9.
TE D
265
Maratt JD, Esposito CI, McLawhorn AS, Jerabek SA, Padgett DE, Mayman DJ. Pelvic tilt in patients undergoing total hip arthroplasty: when does it matter? The
270
Journal of arthroplasty. Mar 2015;30(3):387-391. 10.
272
C RM, Spitzer AI. How does pelvic rotation or tilt affect radiographic measurement of acetabular component inclination angle during THA? Journal of orthopaedics. Dec
273 274
AC C
271
EP
269
2015;12(4):222-227.
11.
Domb BG, El Bitar YF, Sadik AY, Stake CE, Botser IB. Comparison of robotic-
275
assisted and conventional acetabular cup placement in THA: a matched-pair
276
controlled study. Clinical orthopaedics and related research. Jan 2014;472(1):329-
277
336.
15
ACCEPTED MANUSCRIPT 278
12.
Archbold HA, Mockford B, Molloy D, McConway J, Ogonda L, Beverland D. The transverse acetabular ligament: an aid to orientation of the acetabular component
280
during primary total hip replacement: a preliminary study of 1000 cases investigating
281
postoperative stability. The Journal of bone and joint surgery. British volume. Jul
282
2006;88(7):883-886.
283
13.
RI PT
279
Epstein NJ, Woolson ST, Giori NJ. Acetabular component positioning using the transverse acetabular ligament: can you find it and does it help? Clinical orthopaedics
285
and related research. Feb 2011;469(2):412-416.
286
14.
SC
284
Ha YC, Yoo JJ, Lee YK, Kim JY, Koo KH. Acetabular component positioning using anatomic landmarks of the acetabulum. Clinical orthopaedics and related research.
288
Dec 2012;470(12):3515-3523.
289
15.
M AN U
287
Digioia AM, 3rd, Jaramaz B, Plakseychuk AY, et al. Comparison of a mechanical acetabular alignment guide with computer placement of the socket. The Journal of
291
arthroplasty. Apr 2002;17(3):359-364. 16.
294
platform. Current reviews in musculoskeletal medicine. Sep 2011;4(3):151-156. 17.
295
robotic-assisted arthroplasty. Surgical technology international. Mar 2014;24:302-
296 297 298 299 300 301
Werner SD, Stonestreet M, Jacofsky DJ. Makoplasty and the accuracy and efficacy of
EP
293
Tarwala R, Dorr LD. Robotic assisted total hip arthroplasty using the MAKO
AC C
292
TE D
290
306.
18.
Elson L, Dounchis J, Illgen R, et al. Precision of acetabular cup placement in robotic integrated total hip arthroplasty. Hip international : the journal of clinical and experimental research on hip pathology and therapy. Nov 25 2015;25(6):531-536.