Does It Matter: Total Hip Arthroplasty or Lumbar Spinal Fusion First? Preoperative Sagittal Spinopelvic Measurements Guide Patient-Specific Surgical Strategies in Patients Requiring Both

The Journal of Arthroplasty xxx (2019) 1e11

Contents lists available at ScienceDirect

The Journal of Arthroplasty journal homepage: www.arthroplastyjournal.org

Does It Matter: Total Hip Arthroplasty or Lumbar Spinal Fusion First? Preoperative Sagittal Spinopelvic Measurements Guide Patient-Specific Surgical Strategies in Patients Requiring Both Frank W. Parilla, MS a, Ritesh R. Shah, MD a, b, c, d, *, Alexander C. Gordon, MD a, b, c, d, Steven M. Mardjetko, MD a, b, c, d, Nancy E. Cipparrone, MA a, Wayne M. Goldstein, MD a, b, c, d, Jeffrey M. Goldstein, MD a, b, c, d a

Department of Orthopedic Surgery, Illinois Bone & Joint Institute, Morton Grove, IL Department of Orthopedic Surgery, University of Illinois at Chicago, Chicago, IL c Department of Orthopedic Surgery, Advocate Lutheran General Hospital, Park Ridge, IL d Department of Orthopedic Surgery, NorthShore University HealthSystem - Skokie Hospital, Skokie, IL b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 28 January 2019 Received in revised form 24 May 2019 Accepted 29 May 2019 Available online xxx

Background: In patients requiring both total hip arthroplasty (THA) and lumbar spinal fusion (LSF), consideration of preoperative sagittal spinopelvic measurements can aid in the prediction of postfusion compensatory changes in pelvic tilt (PT) and inform adjustments to traditional THA cup anteversion. This study aims to identify relationships between spinopelvic measurements and post-THA hip instability and to determine if procedure order reveals a difference in hip dislocation rate. Methods: Patients at a single practice site who received both THA and LSF between 2005 and 2015 (292: 158 ¼ LSF prior to THA, 134 ¼ THA prior to LSF) were retrospectively reviewed for incidents of THA instability. Those with complete radiograph series (89) had their sagittal (standing) spinopelvic profiles measured preoperatively, immediately postoperatively, and 3 months, 6 months, 1 year, 1.5 years, and 2 years postoperatively. Measured parameters included lumbar lordosis (LL), pelvic incidence (PI), PT, and sacral slope (SS). Results: No significant differences in dislocation rates between operative order groups were elicited (7/73 LSF first, 4/62 THA first; Z ¼ 0.664, P ¼ .509). Compared to nondislocators, dislocators had lower LL (10.9) and SS (7.8), and higher PT (þ4.3) and PI-LL (þ7.3). Additional risk factors for dislocation included sacral fusion (relative risk [RR] ¼ 3.0) and revision fusion (RR ¼ 2.7) . Predictive power of the model generated through multiple regression to characterize individual profiles of post-LSF PT compensation based on perioperative measurements was most significant at 1 year (R2 ¼ 0.565, F ¼ 0.000456, P ¼ .028) and 2 years (R2 ¼ 0.741, F ¼ 0.031, P ¼ .001) postoperatively. Conclusion: In performing THA after LSF, it is theoretically ideal to proceed with THA at a postfusion interval of at least 1 year, beyond which further compensatory PT change is minimal. However, the order of surgical procedure revealed no statistical difference in hip instability rates. In cases characterized by large PI-LL mismatch (larger or less predictable compensation profiles) or large SS or LL loss (considerably atypical muscle recruitment), consideration of full functional anteversion range between sitting and standing positions to account for abnormalities not appreciated with standing radiographic assessment alone may be warranted. © 2019 Elsevier Inc. All rights reserved.

Keywords: lumbar spinal fusion total hip arthroplasty radiographic outcomes complications sequence

Funding: This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. One or more of the authors of this paper have disclosed potential or pertinent conflicts of interest, which may include receipt of payment, either direct or indirect, institutional support, or association with an entity in the biomedical field which may be perceived to have potential conflict of interest with this work. For full disclosure statements refer to https://doi.org/10.1016/j.arth.2019.05.053. * Reprint requests: Ritesh R. Shah, MD, Department of Orthopedic Surgery, Illinois Bone & Joint Institute, 9000 Waukegan Road, Suite 200, Morton Grove, IL 60053. https://doi.org/10.1016/j.arth.2019.05.053 0883-5403/© 2019 Elsevier Inc. All rights reserved.

Degenerative diseases of the hip and spine rank among the most common sources of pain and disability in the United States, with a reported 51.8 and 59 million Americans suffering from osteoarthritis and chronic back pain, respectively [1]. As the population ages and the average retirement age increases, healthcare costs and wages lost to disability due to hip and spine degeneration continues to similarly increase. Complicating the treatment of these diseases and further exacerbating the economic burdens they

2

F.W. Parilla et al. / The Journal of Arthroplasty xxx (2019) 1e11

represent is the fact that a considerable deal of overlap exists between degenerative hip and spine disease populations, with low back pain prevalence among patients with severe hip osteoarthritis estimated to be between 21.2% and 49.4% [2,3]. Among patients electing to undergo total hip arthroplasty (THA), between 18%-25% have previously seen a spine surgeon [4]. With proper surgical indications and patient selection, THA and lumbar spinal fusion (LSF) are both effective surgical interventions. However, management of concurrent presentation of hip and spine degeneration in a single patient poses a unique challenge. Differences in sagittal spinal balance between healthy individuals and those with spinal degeneration, particularly with regard to pelvic tilt (PT), complicate the optimization of THA component positioning in spinal-degenerative patients. Techniques and physiologic markers normally used by the hip surgeon to access relative safe zone cup placement may not be ideal for degenerative spine patients. Compromised sagittal balance and compensatory reliance on atypical muscle group recruitment and movement patterns to achieve motion ranges required for day-today life can lead to extremes of pelvic version that render a typically safe cup orientation problematic. The implications of this reality are well-documented in the literature. Numerous studies have demonstrated that THA patients with concomitant spinal deformity experience episodes of instability and dislocation at disproportionately high rates despite traditional, safe zone cup placement [5]. Furthermore, patients with an initial dislocation are at significantly higher risk for future THA instability episodes, inferior outcomes, and further revision surgery. In addition to affecting quality of life, revision surgery represents a significant financial burden to both patients and healthcare systems, with the mean total cost for revision surgery for instability exceeding $11,000 [6,7]. Presently, there exist neither clear clinical guidelines for determination of the sequence of operative treatment in patients with both hip and spine degeneration, nor any clear guidelines to inform how the targets of biomechanical restoration, standard to either intervention, may be modified to minimize deleterious effects on the outcome goals of the other. An improved understanding of sagittal spinal correction and resultant changes in spinopelvic balance following LSF may decrease the disproportionately high

rates of THA instability in this population. In that context, this study aims to identify relationships between spinopelvic measurements and post-THA hip instability and to determine if procedure order reveals a difference in hip instability rate. Methods An electronic health record database for a single practice study site was cross-queried for patient billing histories containing Current Procedural Terminology codes for both THA and LSF between 2005 and 2015. Cross-reference of codes yielded an initial cohort of 292 patientsd158 of whom received LSF prior to THA and 134 of whom received THA prior to LSF. This cohort was then narrowed through exclusion of patients who received lumbar fusions that extended superior to L1; received nontotal arthroplasty, hemiarthroplasty, or resurfacing arthroplasty; had revision THA/LSF prior to primary LSF/THA; or were lost to follow-up subsequent to their second operative procedure. These exclusions yielded a 135patient cohort. Average time between procedures, time to complication from surgery, and complication rate were reviewed. Of these patients, 89 had sufficient preoperative and postoperative radiographic data for a complete series of measurements to be made. This standard was defined as an at-minimum availability of preoperative, immediate postoperative, 3-month, 6month, and 1-year postoperative sagittal spinal radiographs for which measurements could be reproduced on 3 separate occasions within a 3% tolerance-for-variability margin. Due to the imposition of this standard for radiographic measurements, all correlative values between radiographic measurements and outcomes took into account only 66% (89/135) of the cases from which the overall complication rates and relationship between outcomes and surgical timing were assessed. Data recorded for all 135 patients included age, gender, weight, height, hip diagnosis and laterality, spine diagnosis and levels, fusion levels, and all instances of complication including dislocation, fracture, loosening, nonunion, and revision. All radiographic measurements were performed by a single, trained observer using the spinal imaging analytic software Surgimap (Fig. 1). The parameters assessed were lumbar lordosis (LL), PT, pelvic incidence (PI), sacral slope (SS), and PI-LL. These were

Fig. 1. Sample measurement performed with Surgimap spinal analytic software.

F.W. Parilla et al. / The Journal of Arthroplasty xxx (2019) 1e11

3

Table 1 Cohort Demographics and Complication Time Courses.

Combined cohorts All patients Dislocation Fracture Loosening (hip) Spine complication Fusion / THA cohort All patients Dislocation Fracture Loosening (hip) Spine complication THA / fusion cohort All patients Dislocation Fracture Loosening (hip) Spine complication

#M

#F

Weight

Height

BMI

Age

Total

Between Surgeries (mo)

Time to Complication (mo)

56 6 2 2 12

79 5 7 3 15

190.99 187.43 173.25 225.00 204.95

66.34 66.29 65.5 68.3 66.2

0.31 0.31 0.29 0.35 0.34

67.6 62.6 73.1 61.7 65.9

135 11 9 5 27

29.40 44.45 25.89 18.00 33.65

± ± ± ± ±

27.21 34.77 25.11 20.10 28.80

e 12.80 42.25 33.00 35.53

± ± ± ±

15.24 36.37 30.77 33.51

73 7 4 2 13

31.71 48.71 17.25 28.00 36.33

± ± ± ± ±

30.50 40.33 13.94 33.94 27.64

e 11.71 39.66 33.50 44.58

± ± ± ±

18.23 33.61 19.09 29.15

62 4 5 3 15

26.16 37.00 32.80 11.33 31.50

± ± ± ± ±

23.94 25.70 31.32 8.08 31.20

e 15.33 43.80 41.00 27.40

± ± ± ±

5.86 41.73 40.95 35.81

M, male; F, female; BMI, body mass index; THA, total hip arthroplasty.

accessed preoperatively, immediately postoperatively, 3 months postoperatively, 6 months postoperatively, 1 year postoperatively, and, when possible, 1.5 years postoperatively and 2 years postoperatively. Measurements were performed in blocks comprised of full radiograph series for each patient and presented to the observer in temporal order (preoperative, immediate postoperative, 3 months, etc.). Radiographs were sorted as such prior to each measurement session and presented to the observer with no identifying demographic, surgical, or complication data so as to maintain blindness throughout the measuring process. Measurements were repeated 3 times for each patient’s full series of radiographs for intraobserver reliability. Following confirmation that the 3 data points for each measured variable fell within the tolerated range of measurement variance, the 3 values were averaged to arrive at the final value used in the study.

dislocators was greater in all measured spinopelvic parameters, most notably in total (2 year) change in PT (2.80) and PI-LL (3.20). Two patient subgroups which shared characteristics of this average dislocator profile independently arose as populations at increased risk of dislocation. Patients with fusions extending to the sacrum, though representing only 9.0% (18 patients) of the patients included in the study, made up 27.3% (3 patients) of the population of patients who experienced a dislocation (relative risk [RR] ¼ 3.0) (Table 4). Similarly, spinal revision patients, though comprising only 13.6% of the studied fusion cases, made up 36.4% (4) of the dislocation case group (RR ¼ 2.7) (Table 5).

Discussion Standardized Correction for PT

Results Of the 135 individuals who met inclusion criteria, 73 underwent LSF followed by THA (average time between surgeries ¼ 31.7 ± 30.5 months) and 62 underwent THA followed by LSF (26.2 ± 23.9 months). Among the cohort there were 11 patients (8.1%) with at least 1 recorded dislocation incident. No significant differences in dislocation rate between operative order groups were elicited (7/73 LSF first, 4/62 THA first; Z ¼ 0.664, P ¼ .509). Similarly, there were no significant differences in time between surgeries or order of surgeries between noncomplication cases and cases with a recorded incident of fracture, component loosening, or mechanical spinal complication (Table 1). There were also no significant differences in pre-LSF spinopelvic parameter values between operative order groups (Table 2). All 11 patients who experienced a dislocation event fell within the cohort of 89 patients for which full radiographic data were collected and measured. Consistent with the literature, when compared to nondislocators, these patients had significantly lower LL (11.6), higher PT (þ3.5), lower SS (7.5), and higher PI-LL (þ7.4) following LSF (Table 3). Notable, however, was that these dislocating patients experienced greater operative alteration to spinopelvic parameters during LSF. This difference was particularly pronounced with regard to PT, with dislocators experiencing a 284% (þ8.0 vs þ2.82) greater on average intraoperative increase in PT compared to nondislocators. Additionally, the magnitude of average directional post-LSF compensatory movement among

The disturbance of PT in degenerative spine patients is of particular importance in considerations of hip-spine interplay, given the implications of the parameter’s range of movement at the hip joint. In a normal flexible pelvis with intact functional range of motion, there is a change in PT across functional positions, with PT typically increasing as much as 20 -35 between standing and sitting positions [8,9]. These changes in PT produce concomitant changes in the orientation of the acetabulum relative to the femur. An increase in PT is accompanied by a functional increase in the anteversion of the acetabulum. Conversely, a decrease in PT results in a functional decrease in acetabular anteversion (Fig. 2). For every degree of additional PT, the anatomically typical acetabulum will gain 0.7 of anteversion (Fig. 3) [10e12]. This increase in acetabular anteversion which necessarily follows PT increase between standing and sitting positions helps clear the anterior lip of the acetabulum from impingement by the femoral neck, thereby preventing posterior instability as the upper range of flexion is approached. Conversely, in transitioning from a flexed position toward the upper range of functional extension experienced while standing, there is a decrease in PT and subsequent functional retroversion of the acetabulum, clearing the posterior lip of the acetabulum from impingement by the femoral neck, thereby preventing anterior instability. Inability to safely achieve this full range of functional version change can result in femoral neck impingement at the extremes of flexion or extension and lead to either anterior or posterior instability at the hip joint.

Preop

4

Table 2 Operative Order Groups. Preop DFH

Operative Change

jOperative Changej

12.44 9.96 9.00 8.00 13.82

1.13 3.24 3.35 0.29 4.48

6.68 5.37 4.16 4.06 2.52

13.48 10.48 9.23 9.27 13.30

2.43 3.70 3.30 0.35 5.74

11.81 9.64 8.86 7.23 14.13

0.36 2.97 3.38 0.67 3.74

Postop DFH

Change From Preceding Measurement

2 y Change Sum

jChange From Preceding Measurementsj 3 mo

6 mo

1y

1.5 y

2y

2 y Change Sum

3 mo

6 mo

1y

1.5 y

2y

9.58 11.56 6.66 4.38 16.24

0.41 0.80 0.66 0.19 1.07

0.49 0.11 0.47 0.30 0.02

0.58 0.03 0.25 0.75 0.33

0.58 0.03 0.25 0.75 0.33

0.11 0.20 0.15 0.28 0.26

0.69 1.43 1.03 0.15 1.72

3.16 2.56 3.04 2.46 1.07

2.13 2.09 2.15 1.82 0.02

1.78 1.93 1.95 1.85 0.33

1.57 1.26 1.39 1.17 0.70

0.84 0.95 1.21 0.74 0.26

9.47 8.79 9.73 8.04 2.37

5.57 5.52 3.83 2.87 1.74

11.17 13.27 6.92 5.87 18.09

0.14 0.55 0.82 0.27 0.95

0.79 0.74 0.16 0.74 0.63

0.82 0.82 0.73 1.09 1.55

0.60 0.00 0.60 0.60 0.00

0.67 0.33 0.67 0.67 0.00

1.37 0.13 1.33 0.64 0.04

2.59 2.00 2.91 2.45 0.32

1.95 1.79 2.37 2.00 0.42

1.91 1.73 2.91 3.09 1.00

2.60 0.80 1.00 1.40 1.60

0.67 0.33 0.67 0.67 0.00

9.71 6.65 9.85 9.61 0.14

7.33 5.28 4.36 4.77 2.97

8.64 10.53 6.50 3.48 15.14

0.57 0.95 0.57 0.46 1.14

0.33 0.58 0.64 0.06 0.31

0.48 0.38 0.62 0.62 0.14

0.17 0.72 0.72 0.00 0.19

0.00 0.19 0.19 0.13 0.19

0.58 1.55 1.46 0.02 2.04

3.49 2.89 3.11 2.46 0.38

2.22 2.25 2.03 1.72 0.19

1.72 2.00 1.59 1.38 0.14

1.28 1.39 1.50 1.11 0.22

0.88 1.06 1.31 0.75 0.44

9.59 9.59 9.53 7.42 0.05

LL, lumbar lordosis; PT, pelvic tilt; PI, pelvic incidence; SS, sacral slope; LSF, lumbar spinal fusion; THA, total hip arthroplasty; DFH, distance of variable from healthy average in matched nondegenerative control groups; jVariablej, absolute value of a given variable.

Table 3 Dislocators. Preop

Dislocators LL 37.60 PT 19.40 PI 49.80 SS 30.60 PI12.20 LL Nondislocators LL 46.48 PT 20.25 PI 53.73 SS 33.52 PI7.25 LL

Preop DFH

jPreop DFHj

Operative Change

jOperative Changej

Postop DFH

Change From Preceding Measurement 3 mo

6 mo

1y

1.5 y

2y

17.00 6.80 0.80 7.10 16.20

17.00 11.28 9.92 7.10 17.40

3.20 8.00 3.60 4.20 6.80

5.00 7.33 3.00 6.17 2.00

20.20 14.80 2.80 11.30 23.00

0.00 0.80 1.40 2.00 1.40

1.00 0.02 1.20 1.20 0.20

0.40 0.60 1.20 0.60 1.60

1.20 1.00 1.00 0.00 2.20

1.40 0.20 0.80 1.00 0.60

7.69 8.45 3.67 4.45 11.36

12.03 9.84 8.92 8.08 13.50

0.95 2.82 3.33 0.68 4.28

6.82 5.20 4.26 3.88 2.56

8.64 11.28 7.00 3.77 15.64

0.45 0.80 0.85 0.02 1.30

0.65 0.14 0.64 0.46 0.01

0.72 0.12 0.11 0.77 0.62

0.03 0.43 0.26 0.17 0.23

0.45 0.21 0.55 0.03 0.11

2 y Change Sum

jChange From Preceding Measurementsj

2y Change Sum

3 mo

6 mo

1y

1.5 y

2y

2.00 2.80 1.20 1.20 3.20

3.20 4.00 2.60 2.00 1.40

1.80 1.40 1.60 1.60 0.20

1.60 1.00 2.00 1.80 1.60

2.00 1.00 2.20 1.60 2.20

1.80 1.00 2.00 1.00 0.60

10.40 8.40 10.40 8.00 6.00

0.10 1.17 1.13 0.15 1.03

3.15 2.43 3.08 2.50 1.30

2.16 2.16 2.20 1.84 0.01

1.80 2.07 1.94 1.86 0.62

1.43 1.34 1.14 1.04 0.23

0.44 0.93 0.88 0.63 0.11

8.99 8.93 9.24 7.87 2.26

Bolded values indicates significant deviation from control value in the positive direction, Italics values indicates significant deviation from control value in the negative direction. LL, lumbar lordosis; PT, pelvic tilt; PI, pelvic incidence; SS, sacral slope; DFH, distance of variable from healthy average in matched nondegenerative control groups; jVariablej, absolute value of a given variable.

F.W. Parilla et al. / The Journal of Arthroplasty xxx (2019) 1e11

Combined operative order group LL 45.76 8.45 PT 20.18 8.32 PI 53.42 3.30 SS 33.29 4.67 PI7.65 11.75 LL THA / LSF LL 45.87 8.73 PT 22.17 9.57 PI 54.22 3.62 SS 32.17 5.53 PI8.35 12.35 LL LSF / THA LL 45.70 8.28 PT 19.00 7.56 PI 52.95 3.11 SS 33.95 4.15 PI7.24 11.39 LL

jPreop DFHj

Table 4 Sacral Fusions. Preop

jPreop DFHj

Operative Change

jOperative Changej

Postop DFH

15.60 7.40 0.77 6.26 16.38

18.28 11.40 8.45 9.50 19.75

1.65 5.41 3.06 1.76 1.41

5.53 6.59 3.29 4.24 2.24

5.91 8.64 4.20 4.10 10.11

10.36 9.44 9.20 7.47 11.71

1.68 2.35 3.22 1.04 4.90

6.60 4.71 4.24 3.76 2.36

Change From Preceding Measurement 3 mo

6 mo

1y

1.5 y

2y

13.95 12.81 3.83 8.03 17.79

1.12 0.18 0.71 0.35 1.82

0.36 0.21 0.50 0.36 0.86

1.00 0.64 1.55 1.18 0.55

0.89 1.22 0.67 0.89 1.56

0.38 0.00 0.25 0.13 0.13

7.59 10.99 7.42 3.06 15.01

1.06 1.06 1.01 0.16 2.07

0.90 0.19 0.65 0.42 0.25

0.60 0.35 0.29 0.62 0.89

0.13 0.55 0.99 0.56 0.86

0.00 0.22 0.08 0.23 0.08

2y Change Sum

jChange From Preceding Measurementsj

2y Change Sum

3 mo

6 mo

1y

1.5 y

2y

1.96 1.47 1.26 0.29 0.70

3.59 2.29 2.82 2.00 1.82

1.64 1.36 1.36 1.36 0.86

1.73 1.55 1.73 1.73 0.55

2.00 1.44 2.00 1.33 1.56

1.63 0.75 1.25 0.88 0.13

10.58 7.39 9.16 7.29 4.91

1.23 1.29 1.13 0.37 2.36

4.12 3.41 3.91 3.35 2.07

2.96 3.05 3.08 2.55 0.25

1.81 2.13 1.90 1.88 0.89

1.37 1.42 1.24 0.93 0.86

0.76 0.99 0.98 0.53 0.08

11.03 10.99 11.11 9.24 4.15

Bolded values indicates significant deviation from control value in the positive direction, Italics values indicates significant deviation from control value in the negative direction. LL, lumbar lordosis; PT, pelvic tilt; PI, pelvic incidence; SS, sacral slope; DFH, distance of variable from healthy average in matched nondegenerative control groups; jVariablej, absolute value of a given variable.

Table 5 Revision Fusions. Preop

Preop DFH

Revision fusions (primary op) LL 45.10 9.50 PT 26.50 13.90 PI 56.60 6.00 SS 29.60 8.10 PI11.50 17.22 LL Combined operative group LL 45.76 8.45 PT 20.18 8.32 PI 53.42 3.30 SS 33.29 4.67 PI7.65 11.75 LL

jPreop DFHj

Operative Change

jOperative Changej

Postop DFH

Change From Preceding Measurement 3 mo

6 mo

1y

1.5 y

2y

2y Change Sum

jChange From Preceding Measurementsj 3 mo

6 mo

1y

1.5 y

2y

2y Change Sum

14.86 15.42 9.48 11.42 21.22

2.70 0.00 2.70 3.00 0.00

6.10 7.00 4.90 7.00 1.20

6.80 13.90 8.70 5.10 17.22

1.63 0.50 0.25 0.88 1.38

0.14 0.43 0.57 0.43 0.71

0.17 0.83 1.33 0.33 1.17

0.67 0.67 0.17 0.83 0.83

1.00 0.67 1.00 0.33 2.00

1.60 0.10 1.85 1.95 0.24

4.38 3.25 2.25 4.13 1.38

1.57 2.43 0.86 2.43 0.71

2.50 3.83 2.67 2.33 1.17

4.33 2.00 2.50 1.50 0.83

1.00 1.33 1.00 0.33 2.00

13.78 12.85 9.27 10.72 6.09

12.44 9.96 9.00 8.00 13.82

1.13 3.24 3.35 0.29 4.48

6.68 5.37 4.16 4.06 2.52

9.58 11.56 6.66 4.38 16.24

0.41 0.80 0.66 0.19 1.07

0.49 0.11 0.47 0.30 0.02

0.58 0.03 0.25 0.75 0.33

0.58 0.03 0.25 0.75 0.33

0.11 0.20 0.15 0.28 0.26

0.69 1.43 1.03 0.15 1.72

3.16 2.56 3.04 2.46 1.07

2.13 2.09 2.15 1.82 0.02

1.78 1.93 1.95 1.85 0.33

1.57 1.26 1.39 1.17 0.70

0.84 0.95 1.21 0.74 0.26

9.47 8.79 9.73 8.04 2.37

F.W. Parilla et al. / The Journal of Arthroplasty xxx (2019) 1e11

Sacral fusions LL 39.00 PT 20.00 PI 51.38 SS 31.44 PI12.38 LL Nonsacral fusions LL 48.12 PT 20.24 PI 54.13 SS 33.93 PI6.01 LL

Preop DFH

Bolded values indicates significant deviation from control value in the positive direction, Italics values indicates significant deviation from control value in the negative direction. LL, lumbar lordosis; PT, pelvic tilt; PI, pelvic incidence; SS, sacral slope; DFH, distance of variable from healthy average in matched nondegenerative control groups; jVariablej, absolute value of a given variable.

5

6

F.W. Parilla et al. / The Journal of Arthroplasty xxx (2019) 1e11

Fig. 2. Cup orientation with pelvic tilt of 14 (left) and þ14 (right). In order to investigate the influence of pelvic tilt, synthetic X-rays of a pelvis with an acetabular component were generated. A 3-dimensional model of a pelvis orientated with anterior and posterior superior iliac spines at the same height in the frontal plane had a 56-mm diameter Birmingham hip resurfacing implanted at 40 inclination and 15 anteversion. Pelvic tilt was applied to this 3-dimensionl model by rotating its vertices about an axis through the hip joint centers. An orthographic projection matrix was used to produce a 2-dimensional image of the pelvis and acetabular component. The resulting image was analyzed using Ein-Bild-Roentgen-Analyse cup to determine inclination and anteversion (http://www.ors.org/Transactions/57/1008.pdf).

To avoid impingement and instability issues, hip surgeons aim for acetabular cup placement which approximates the baseline biological anteversion of the physiologically normal acetabulum. Traditionally, this is attempted through adherence to Lewinnek’s suggested “safe zone” for acetabular cup placement in THAs, specifically, 40 ± 10 of acetabular inclination and 15 ± 10 of anteversion [13]. However, the patient groups studied in the establishment of these guidelines came from general THA populations, and the guidelines were thus developed in a way not necessarily considerate of the potential, unique needs of THA populations with concomitant abnormalities in sagittal spinal balance, either from spinal degeneration or treatment with lumbar fusion. Thus, anteversion ranges of acetabular cup placement considered safe in patients with uncompromised sagittal balance may still represent stability risks in patients with atypical sagittal balance parameters. A recent study by DelSole et al found that, among a studied cohort of primary THA patients with concomitant spinal deformity (either untreated lumbar degeneration or lumbar degeneration treated with fusion) who experienced a dislocation incident, 80% had acetabular cup anteversion values which fell into the range defined as safe by Lewinnek’s guidelines. Despite this degree of adherence to the aims of traditional safe zone cup placement, dislocations occurred at a rate of 8% with a 5.8% revision rate for instability (compared to the general primary THA dislocation rate of 1%-3% commonly cited in the literature) [14]. These findings are part of a growing body of literature which supports the existence of a relationship between THAs performed on patients with concurrent spinal deformity and increased rates of THA instability. Bedard et al demonstrated a similarly disproportionate rate of dislocation in patients with concurrent THA and spinopelvic fusion within both a home institution (16.7%) and within a large national database (8.3%). THA patients in the same national database with no history of spinopelvic fusion dislocated at a rate of 2.9%. The relative risk of dislocation among patients with both THA and spinal fusion was 2.96 (1.16-7.57) compared to patients with only THA (P ¼ .02) [15]. A similar study by Perfetti et al

[5] found that, at 12 months, THA patients with prior lumbar fusion were 7.19 times more likely to dislocate (control: 0.4%, fusion: 3.0%, P < .001) and 4.64 times more likely to undergo revision (control: 0.9%, fusion: 3.9%, P < .001). These disparities in instability rates have been widely attributed to significant average differences between the spinopelvic profiles of patients with concomitant hip-spine degeneration and those without such degeneration. The most consistently and significantly disparate parameter cited among these differences has been PT (24 ± 12 in degenerative populations vs 13 ± 10 in healthy populations) [16e18]. Among dislocators within such degenerative populations,

Fig. 3. Effect of pelvic tilt on cup anteversion. All cup measurements were made using Ein-Bild-Roentgen-Analyse by 2 observers with excellent interclass correlation coefficients. Acetabular component orientation was measured in 13 patients (7:6 males:females) with adequate quality supine and standing radiographs (patient cohort). The changes in anteversion between supine and standing were calculated as follows: DAnteversion ¼ Standing AnteversioneSupine Anteversion (http://www.ors.org/ Transactions/57/1008.pdf).

F.W. Parilla et al. / The Journal of Arthroplasty xxx (2019) 1e11

this directional abnormality in PT is further magnified. One large and recent study reported that, compared to spinal deformity patients who received THA and did not dislocate, spinal deformity patients who did sustain a dislocation had significantly higher PT (29.282 ± 14.188 vs 22.141 ± 10.389, P ¼ .50) [14]. For such a population, the simplest and highest yield spinopelvic consideration in planning a THA cup orientation least likely to produce impingement and instability is PT, for which, as previously described, every additional degree of deviation will produce approximately 0.7 of anteversion change. Deviation in PT from that of a profile of a typical balanced spine should be adjusted for with a 0.7 modification to the cup anteversion target for every degree in variance of PT, taking into consideration both standing PT at the time of THA, and the approximated magnitude and direction of any further PT changes that may be expected to occur as a result of compensatory adjustment to prior LSF that may have yet to complete its course. When possible, the time between procedures should be planned so that THA proceeds at a post-LSF interval beyond which further LSFinduced compensatory changes are both minimal and most reliably predictable. In situations in which THA must or has already preceded LSF, care should be taken by the spine surgeon to note the patient’s implanted cup orientation and standing PT, with an understanding of how particular parameter alteration targets might be chosen to minimize deleterious directional changes to PT, or improve presently apparent impingement propensities by driving a directional change in PT toward a value better suited for the orientation of the previously implanted cup. This consideration would take preoperative spinopelvic parameters into consideration, as well as the post-LSF compensatory changes in PT expected to occur in response to a given profile of operative modification to these parameters.

Operative Order Significant differences in spinopelvic parameter values and complication rates between operative order groups were not elicited in this study. The sequence of intervention at the hip and spine may not itself be of significance in outcome optimization. An important consideration in this result’s interpretation, however, is the lack of direct coordination between the hip and spine surgeons operating on the patients included in this study. To our knowledge, all hip prostheses were placed without comprehensive consideration of patients’ spinopelvic balance profile and all fusions implemented without explicit reference to the orientation of prior THA patients’ prosthetic cup positioning. Due to the comparative complexity and relatively numerous degrees of freedom present within the spine and spinopelvic joint systems as compared to the hip joint, it is both more difficult to controllably manipulate the spine via fusion surgery into a desired profile and more difficult to reliably predict how the postoperative spinal profile will compensate to surgical alteration in the long term. The constraint of the hip joint and the clear, static relationship of the implanted cup relative to the acetabulum allow for prosthesis orientation goals to be met with greater precision and reliability. Although this may implicate THA as a preferable candidate for secondary interventiondto follow spinal fusiondwhen coordinating surgical interventions in concomitant hip-spine degeneration patients, other considerations are relevant in the decision-making process of operative order. Specifically, patients may want to address the more painful, disabling condition first or may want to undergo THA first due to its more consistent and reliable postoperative rehabilitation experience.

7

Time Between Surgeries To ensure that functional anteversion remains centered within a target range in the long-term, hip surgeons must determine the following: (1) the post-LSF duration beyond which further compensatory changes to PT are minimal or most reliably predictable and (2) the relative magnitude and direction of these changes. To help inform these determinations, a multiple regression analysis was conducted for variables affecting postfusion PT change at 6 months, 1 year, and 2 years postoperatively. The input variables included operative change to LL, PT, PI, and SS, as well as the postoperative distance of each of these 4 variables from their respective values in a spinopelvic profile representative of an average healthy, balanced spine [19] (Table 6). The predictive model generated for the 6-month postfusion duration accounted for only 33% of the compensatory alteration of PT with indications that predictive power is poor within the first 6 postoperative months (significance F ¼ 0.013, P ¼ .655). The 1-year predictive model accounted for 57% of measured compensatory alteration of PT with evidence of significant predictive power at this time interval (significance F ¼ 0.000456, P ¼ .028). Of the model’s inputs, significance was greatest among the operative change values, which indicated that for every degree of operative increase in LL a loss of 0.25 PT (P ¼ .006), for every degree in PT a loss of 0.393 PT (P ¼ .012), and for every increase in SS a gain of 0.384 PT (P ¼ .006) was observed. The 2-year predictive model accounted for 74% of measured compensatory variation of PT with evidence of significant predictive power at this time interval (significance F ¼ 0.031, P ¼ .001). Similar to the 1-year predictive model, significance was greatest among operative change values, which indicated that for every degree of operative increase in PT a loss of 0.46 PT (P ¼ .019) and for every degree of PI a loss of 1.26 PT (P ¼ .013) was observed. Although, as in the 1-year model, there was a less significant relationship between postoperative spinopelvic parameters’ deviation from healthy spinopelvic parameter profile values and compensatory PT changes, the relationship coefficients and their respective levels of significance indicated that deviation from healthy values postoperatively had a nearly 2-fold increase in both predictive significance and attributable impact on PT changes (per degree of deviation) at 2 years as compared to 1 year, suggesting that operative change may bear its greatest predictive value at shorter postoperative intervals, and that the postoperative profile and the magnitude and direction of its values’ deviations from those of a healthy, balanced profile becomes most clinically useful in predicting compensatory changes at greater postoperative intervals. Also evident was the fact that alteration to parameters besides PT were of greatest predictive value at shorter postoperative durations, with alterations to PT itself gaining increasing predictive value as time went on, relative to the other parameters. Although the 2-year model was able to attribute a greater proportion of overall changes to immediate postoperative inputs, the influence of individual value changes on PT was most clear at the 1-year point, suggesting that PT changes beyond this point may not be as simply or linearly attributable to specific LL, PI, or SS values. Among all cases included in the study, 53% of the PT changes that cumulatively occurred over the 2-year measurement had occurred by 6 months and 75% by 1 year. The smaller fraction of total compensatory change undergone at the 6-month interval combined with the relatively unpredictable course of changes during this time suggest that a postfusion wait period of at least 1 year would be preferable both for the significantly decreased likelihood of further compensatory alteration to PT and for the enhanced ability to reliably predict such changes after that

8

F.W. Parilla et al. / The Journal of Arthroplasty xxx (2019) 1e11

Table 6 Postoperative Pelvic Tilt Change vs Parameter Value at the Time of Surgery. 6 mo

1y

Change in Pelvic Tilt vs Measured Parameter Value

Change in Pelvic Tilt vs Measured Parameter Value

Regression Statistics

Regression Statistics

Multiple R R2 Adjusted R2

Multiple R R2 Adjusted R2

0.576 0.332 0.213 df

Regression

8.000 Coefficients

SS

MS

F

Sig F

214.048

26.756

2.792

Standard Error

t Stat

P-Value

df

0.013 Low 95%

0.752 0.565 0.453

Regression Up 95%

Low 95%

Up 95%

8.000 Coefficients

SS

MS

530.366 Standard Error

66.296 t Stat

0.316

0.704

0.450

.655

1.101

1.733

1.101

1.733

2.468

1.067

2.312

LL PT PI SS

0.131 0.130 0.294 0.238

0.062 0.109 0.144 0.105

2.111 1.198 2.049 2.274

.040 .237 .046 .028

0.256 0.349 0.583 0.027

0.006 0.089 0.005 0.449

0.256 0.349 0.583 0.027

0.006 0.089 0.005 0.449

LL PT PI SS

¡0.247 ¡0.393 0.302 0.384

0.083 0.147 0.190 0.131

¡2.973 ¡2.682 1.592 2.919

LL PT PI SS

0.013 0.022 0.038 0.009

0.053 0.239 0.231 0.217

0.255 0.090 0.164 0.040

.800 .929 .870 .968

0.120 0.504 0.503 0.428

0.093 0.461 0.427 0.446

0.120 0.504 0.503 0.428

0.093 0.461 0.427 0.446

LL PT PI SS

0.004 0.630 0.496 0.383

0.078 0.479 0.459 0.414

0.049 1.315 1.079 0.924

Intercept

pointdespecially in cases involving particular risk factors for the generation of unsafe functional anteversion ranges. In patients with immediately debilitating pain or disability emanating from both spine and hip sources, the benefits of such a planned extension of interoperative duration would have to be weighed against potentially prolonging of pain or disability and lengthening of time until return to uncompromised daily living. Special Considerations Modifications to these general recommendations may be necessary in cases in which there are indications that (1) there exists a compromise to the maintenance of a safe functional range of anteversion that may not be adequately appreciated and adjusted for with static lateral radiographic assessment of PT alone or (2) the profile of postfusion compensatory changes in sagittal balance may deviate considerably from the change profile predicted by measurements of preoperative/postoperative spinopelvic parameters and the magnitude and direction of the operational adjustments made to them. Directional Biasing of Functional Anteversion Range Through Atypical Muscle Recruitment Aside from PT abnormalities, a defining characteristic of the spinopelvic profiles of the patients who experienced a dislocation in this study was exaggerated loss of LL and SS when compared to nondislocators. This was seen preoperatively (17.00 LL, 7.10 SS vs 7.69 LL, 4.45 SS), as an operative change (3.20 LL, 4.20 SS vs 0.95 LL, 0.68 SS), and as a postoperative difference (20.20 LL, 11.30 SS vs 8.64 LL, 3.77 SS) from that of a healthy, balanced spine. Such exaggerated loss of lumbar sagittal curvature may prompt a proportionally exaggerated reliance on intrinsic pelvic flexion to achieve movement normally accomplished in part through extrinsic pelvic (lumbar) flexion, resulting in compensatory pelvic hyperextension, SS loss, and an increase in PT. Such compensatory hyperextension may translate to a routine surpassing of the positive range of anteversion safely allowed by traditional cup anteversion targets (even when modified in accordance with the above adjustment coefficients to account for standing PT abnormalities and projected post-LSF change profiles) and necessitate further adjustment to the target range of cup anteversion to avoid

Intercept

increased risk of posterior impingement and anterior instability when the hip is extended. Conversely, while less evident in this particular cohort, increased fixed LL secondary to fusion could drive an increase in SS, produce a concomitant decrease in PT, and necessitate an increase in the target range of cup anteversion to avoid anterior impingement and posterior instability when the hip is flexed. Although not explicitly considered in this study, an increase in the number of fused lumber vertebrae may correlate with an increasing degree of spinal rigidity capable of compromising functional PT range and thus safe version range at the hip joint at either or both extremes of articulation. A similar mechanism may also partially underlie this study’s observation of disproportionately high rates of instability in patients with fusions extending to the sacrum, when instability cannot be attributed to large changes in PT alone. Patients with spinopelvic profiles characterized by considerable LL and SS loss, particularly if projected to incur further LL/SS loss or changes to PT that would exacerbate the functional implications of such losses, would benefit from an analysis of functional PT range between sitting and standing positions. Similar caution should be exercised in patients with lumbar fusions extending to the sacrum. In cases where compensation for achievement of safe movement range necessitates significant alteration to cup anteversion targets, or if compensation for a narrowing of safe movement range secondary to fixed spine rigidity may not be adequately achieved at both articulation extremes with directional adjustment of anteversion, consideration of dual-mobility constructs may be merited. Larger-Than-Predicted Postoperative Compensatory Change in (Static) Standing PT Although comparatively low PI was characteristic among this study’s population of dislocators (2.80 above that of the average healthy spine vs 7.00), large PI-LL mismatch was additionally a trait of this population (23.00 above healthy vs 15.64), demonstrating that even a PI value large enough to otherwise suggest some retained ability to safely accommodate imbalance without producing instability-related alterations to functional anteversion range, could still exist in an instability-prone patient, should there exist a large enough mismatch between PI and LL values (Table 7). This reality is demonstrated clearly among the population of

F.W. Parilla et al. / The Journal of Arthroplasty xxx (2019) 1e11

1y

2y

Change in Pelvic Tilt vs Measured Parameter Value

Change in Pelvic Tilt vs Measured Parameter Value

Regression Statistics

Regression Statistics

0.752 0.565 0.453 F

Multiple R R2 Adjusted R2 Sig F

5.042

0.861 0.741 0.535 df

0.00046

Regression

P-Value

Low 95%

.028

0.291

Up 95% 4.644

Low 95% 0.291

Up 95% 4.644

.006 .012 .122 .006

¡0.417 ¡0.692 0.690 0.116

¡0.078 ¡0.094 0.085 0.652

¡0.417 ¡0.692 0.690 0.116

¡0.078 ¡0.094 0.085 0.652

.961 .198 .289 .362

0.155 1.607 0.441 1.228

0.163 0.347 1.433 0.462

0.155 1.607 0.441 1.228

0.163 0.347 1.433 0.462

9

8.000 Coefficients

SS

MS

274.867

t Stat

F

Sig F

3.584

0.031

P-Value

Low 95%

Up 95%

Low 95%

Up 95%

9.719

2.244

4.330

.001

4.718

14.719

4.718

14.719

LL PT PI SS

0.166 ¡0.456 ¡1.259 0.107

0.093 0.163 0.420 0.150

1.783 ¡2.799 ¡3.000 0.710

.105 .019 .013 .494

0.374 ¡0.818 ¡2.194 0.228

0.042 ¡0.093 ¡0.324 0.441

0.374 ¡0.818 ¡2.194 0.228

0.042 ¡0.093 ¡0.324 0.441

LL PT PI SS

0.072 1.162 0.802 0.750

0.103 1.351 1.343 1.348

0.696 0.860 0.597 0.556

.502 .410 .564 .590

0.158 4.172 2.190 3.755

0.302 1.848 3.793 2.254

0.158 4.172 2.190 3.755

0.302 1.848 3.793 2.254

Intercept

Standard Error

34.358

Gray cells indicate the value of operative modification and white cells indicate the distance of postoperative value from healthy; discussed significant values boldened. df, degrees of freedom; MS, mean squared; LL, lumbar lordosis; PT, pelvic tilt; PI, pelvic incidence; SS, sacral slope.

studied patients who experienced “catastrophic” changes to PT, defined as postoperative compensatory PT movement in excess of 7 within the first year following fusion surgerydan extent of change itself capable of destabilizing a THA placed at a postfusion interval which the aforementioned models would otherwise suggest to be sufficient to assume that further compensatory PT change would be minimal and relatively predictable (Table 8). Despite the group’s large PI compared to the average of the full study cohort (60.30 vs 53.42), these patients experienced significantly more dramatic and deleterious postoperative changes to PT. This event appears attributable to the only other significant difference between these groupsdPI-LL valuesdwhich were far greater in the “catastrophic” change group than in the full study comparison cohort (18.50 vs 12.13). Compensatory alterations to spinopelvic balance parameters capable of producing greater-than-expected alterations to PT and, consequently, functional anteversion ranges, occur at increased rates not only in individuals with low PI, but also, and perhaps more predictably or significantly, in individuals with large PI-LL mismatch. Evidently, large PI alone should not itself be considered a profile characteristic that indicates reduced likelihood of large postfusion compensatory changes in spinopelvic balance. When planning cup orientations most likely to produce functional anteversion ranges that remain safe in the long-term, special attention should thus be paid to cases involving spinopelvic profiles characterized by either trait, as well as to cases involving revision fusions, which appear to correlate with greater and less predictable compensatory change profiles (Table 5). In addition to larger directional adjustments to anteversion targets in these populations, dual-mobility constructs may also be helpful in the assurance that safe functional anteversion ranges are maintained in the long-term.

dislocation. It should also be noted that implant position and pelvic position are not the only factor in hip stability. These data does not acknowledge the possibility of patients with other relevant factors for hip instability such as patients with previous nonarthroplasty hip surgery or patients with underlying diagnoses of neuromuscular dysfunction or muscle imbalance. Additionally, functional measurements were not reliably available across patients and measurement intervals, which limited this study to consideration of standing radiographic data. Furthermore, the inclusion of some patients for whom there was lack of radiographic data beyond the 1-year measurement interval (1.5 and 2 years) may have blunted the magnitude of measured spinopelvic change averages within the first postoperative year, and magnified those in the second postoperative year. This possibility exists as a result of the failure to adjust for the fact that there may have existed a relationship between follow-up loss beyond this 1-year time point and outcomes. The predictive model for postoperative compensatory change in PT generated in this study took into account only preoperative and immediately postoperative radiographic data. Although useful in preoperative screening for the potential need for modification of cup anteversion targets, dual-mobility cup use, or adjustment to planned intervals between surgeries, a more dynamic model might progressively incorporate postoperative deviations from predicted change patterns using additional, postoperative radiographs, and allow for an ongoing enhancement of initial preoperative projections to better inform further clinical decision-making. Furthermore, use of both standing and sitting radiographs in the future would give a better sense of the rigidity of spine deformity as well as whether a postfusion pelvic position mirrors a position more commonly seen in a healthy standing or sitting pelvis.

Limitations

Conclusion

Rigorous exclusion criteria required by the longitudinal radiographic nature of this study yielded a limited study cohort size. The small cohort size prohibited meaningful comparisons to be made between a number of secondary factors including bone density, soft tissue variation, and various prosthetic configurations. Although all of the THAs in the cohort were performed with a posterior approach and posterior capsular repair, the data presented does not specifically acknowledge femoral head size or direction of

Recognition of and adjustment for a number of readily identifiable spinopelvic profile characteristics may decrease the disproportionately high rates of instability currently reported among patients who require dual THA-LSF intervention for concomitant hip-spine degeneration. In performing THA after LSF, this study suggests that it may be ideal to wait at least 1 year after LSF prior to performing THA, beyond which further compensatory PT change is minimal and more reliably predictable. However, the order of

10

Table 7 PI-LL Mismatch. Preop

jPreop DFHj

Operative Change

jOperative Changej

Postop DFH

Change From Preceding Measurement 3 mo

6 mo

1y

1.5 y

2y

15.71 11.87 6.93 4.54 22.63

15.96 13.68 10.55 6.49 22.63

4.05 6.42 4.21 0.11 8.47

8.37 6.95 4.32 4.63 10.89

19.76 18.29 11.14 4.44 31.11

0.37 1.58 1.32 0.21 0.95

1.42 0.05 0.53 0.68 0.84

0.42 1.21 0.05 0.42 0.26

0.33 0.67 0.42 0.58 0.08

0.45 0.09 0.55 0.27 0.09

0.09 2.64 2.36 0.18 0.45

2.16 3.58 3.74 1.89 1.58

2.16 2.68 2.32 1.63 0.21

4.24 4.13 1.22 2.16 5.45

11.09 7.15 8.95 8.31 10.09

2.05 0.09 2.27 2.41 0.55

6.05 4.09 3.45 3.86 5.18

2.19 4.04 3.49 0.25 6.00

1.00 0.24 0.05 0.24 0.95

0.30 0.10 0.00 0.15 0.30

0.79 1.26 0.16 1.42 0.95

0.50 0.70 1.10 0.80 0.60

0.29 0.43 0.14 0.43 0.43

2.50 2.38 0.13 1.88 1.38

3.48 1.67 2.90 2.24 0.38

1.80 2.30 2.20 1.65 0.50

2y Change Sum

3 mo

6 mo

1y

2y Change Sum

1.5 y

2y

1.58 1.84 2.37 1.37 0.89

1.17 1.00 0.75 0.92 0.42

0.82 1.00 1.09 0.64 0.27

7.88 10.11 10.26 6.45 2.54

1.84 2.11 1.11 1.84 0.74

1.10 1.70 1.50 0.80 0.40

0.86 1.00 1.00 0.43 0.14

9.08 8.77 8.71 6.96 0.07

Bolded values indicates significant deviation from control value in the positive direction, Italics values indicates significant deviation from control value in the negative direction. LL, lumbar lordosis; PT, pelvic tilt; PI, pelvic incidence; SS, sacral slope; DFH, distance of variable from healthy average in matched nondegenerative control groups; jVariablej, absolute value of a given variable.

Table 8 “Catastrophic” PT Changers. Preop

Preop DFH

Catastrophic PT changers LL 48.10 6.50 PT 24.10 11.50 PI 60.30 9.70 SS 36.10 1.60 PI12.10 16.10 LL Combined operative order group LL 45.76 8.45 PT 20.18 8.32 PI 53.42 3.30 SS 33.29 4.67 PI7.65 11.75 LL

jPreop DFHj

Operative Change

14.10 11.50 14.30 8.60 16.10

1.90 3.40 4.60 1.30 6.40

12.44 9.96 9.00 8.00 13.82

1.13 3.24 3.35 0.29 4.48

jOperative Changej

Postop DFH

Total Directional Change 6 mo

1y

2y

Total Absolute Change 6 mo

1y

2y

8.14 5.71 5.71 3.28 9.29

8.30 15.00 14.30 0.30 22.60

2.60 0.60 0.00 0.10 2.60

0.70 0.60 0.70 1.60 1.40

3.30 3.00 7.70 1.00 11.00

6.29 5.43 3.43 3.29 3.71

5.57 10.00 5.29 5.57 6.00

9.33 8.33 9.00 5.00 11.00

6.68 5.37 4.16 4.06 2.52

9.58 11.56 6.66 4.38 16.24

0.90 0.69 0.19 0.49 1.09

0.32 0.66 0.44 0.26 0.76

0.69 1.43 1.02 0.15 1.72

5.29 4.65 5.19 4.28 1.09

7.07 6.58 7.14 6.13 1.42

9.48 8.79 9.74 8.04 2.38

Bolded values indicates significant deviation from control value in the positive direction, Italics values indicates significant deviation from control value in the negative direction. LL, lumbar lordosis; PT, pelvic tilt; PI, pelvic incidence; SS, sacral slope; DFH, distance of variable from healthy average in matched nondegenerative control groups; jVariablej, absolute value of a given variable.

F.W. Parilla et al. / The Journal of Arthroplasty xxx (2019) 1e11

Mismatched (PI-LL) LL 38.90 PT 24.47 PI 57.53 SS 33.16 PI18.63 LL Matched (PI-LL) LL 50.36 PT 16.73 PI 51.82 SS 35.55 PI1.46 LL

jChange From Preceding Measurementj

Preop DFH

F.W. Parilla et al. / The Journal of Arthroplasty xxx (2019) 1e11

surgical procedure itself revealed no statistical difference in hip instability rates. In cases characterized by large PI-LL mismatch or large SS or LL loss, a greater consideration of full functional anteversion range between sitting and standing positions is needed to account for abnormalities not appreciated with standing radiographs alone. Patients with LSF to the sacrum warrant particular attention due to their higher risk of dislocation. Further exploration of different surgical approaches to the hip with purported lower risk of dislocation (anterior, direct lateral, etc.) as well as dualmobility constructs and other emerging technologies may be warranted to assess if instability rates are similarly affected in the higher risk, dual THA-LSF patients. Finally, hip instability is one of many complications that hip and spine surgeons must consider and needs to be weighed in the context of patient pain, disability, and loss of quality of life when making surgical planning decisions. References [1] Goode AP, Carey TS, Jordan JM. Low back pain and lumbar spine osteoarthritis: how are they related? Curr Rheumatol Rep 2013;15:305. https://doi.org/ 10.1007/s11926-012-0305-z. [2] Staibano P, Winemaker M, Petruccelli D, de Beer J. Total joint arthroplasty and preoperative low back pain. J Arthroplasty 2014;29:867e71. [3] Parvizi J, Pour AE, Hillibrand A, Goldberg G, Sharkey PF, Rothman RH. Back pain and total hip arthroplasty: a prospective natural history study. Clin Orthop Relat Res 2010;468:1325e30. [4] Prather H, Van Dillen L, Kymes S. Impact of coexistent lumbar spine disorders on clinical outcomes and physician charges associated with total hip arthroplasty. Spine J 2012;12:363e9. [5] Perfetti DC, Schwarzkopf R, Buckland AJ, Paulino CB, Vigdorchik JM. Prosthetic dislocation and revision after primary total hip arthroplasty in lumbar fusion patients: a propensity score matched-pair analysis. J Arthroplasty 2017;32: 1635e40.

11

[6] Jones S. The prevention and treatment of dislocation following total hip arthroplasty: efforts to date and future strategies. Hip Int 2015;25:388e92. [7] Hailer N, Weiss R, Stark A, K€ arrholm J. The risk of revision due to dislocation after total hip arthroplasty depends on surgical approach, femoral head size, sex, and primary diagnosis. Acta Orthop 2012;83:442e8. [8] Lazennec JY, Brusson A, Rousseau MA. Lumbar-pelvic femoral balance on sitting and standing lateral radiographs. Orthop Traumatol Surg Res 2013;99: S87e103. [9] Philippot R, Wegrzyn J, Farizon F, Fessy MH. Pelvic balance in sagittal and Lewinnek reference planes in the standing, supine and sitting positions. Orthop Traumatol Surg Res 2009;95:70e6. [10] Maratt JD, Esposito CI, McLawhorn AS. Pelvic tilt in patients undergoing total hip arthroplasty: when does it matter? J Arthroplasty 2015;30:387e91. [11] Lembeck B, Mueller O, Reize P. Pelvic tilt makes acetabular cup navigation inaccurate. Acta Orthop 2005;76:517e23. [12] Wan Z, Malik A, Jaramaz B. Imaging and navigation measurements of acetabular component position in THA. Clin Orthop Relat Res 2009;467:32e42. [13] Lewinnek GE, Lewis JL, Tarr R, Compere CL, Zimmerman JR. Dislocations after total hip-replacement arthroplasties. J Bone Joint Surg Am 1978;60:217e20. [14] DelSole EM, Vigdorchik JM, Schwarzkopf R, Errico TJ, Buckland AJ. Total hip arthroplasty in the spinal deformity population: does degree of sagittal deformity affect rates of safe zones placement, instability, or revision? J Arthroplasty 2017;32:1910e7. [15] Bedard NA, Martin CT, Slaven SE, Pugely AJ, Mendoza-Lattes SA, Callaghan JJ. Abnormally high dislocation rates of total hip arthroplasty after spinal deformity surgery. J Arthroplasty 2016;31:2884e5. [16] Jackson RP, McManus AC. Radiographic analysis of sagittal plane alignment and balance in standing volunteers and patients with low back pain matched for age, sex, and size: a prospective controlled clinical study. Spine J 1994;19:1611e8. [17] Charosky S, Guigui P, Blamoutier A, Roussouly P, Chopin D. Complications and risk factors of primary adult scoliosis surgery: a multicenter study of 306 patients. Spine J 2012;37:693e700. [18] Lafage V, Schwab F, Patel A, Hawkinson N, Farcy JP. Pelvic tilt and truncal inclination: two key radiographic parameters in the setting of adults with spinal deformity. Spine J 2009;34:E599e606. [19] Guigui P, Levassor N, Rillardon L, Wodecki P, Cardinne L. Physiological value of pelvic and spinal parameters of sagittal balance: analysis of 250 healthy volunteers. Rev Chir Orthop Reparatrice Appar Mot 2003;89:496e506.