Biomechanical evaluation of sacroiliac joint fixation with decortication

Accepted Manuscript Title: Biomechanical evaluation of sacroiliac joint fixation with decortication Author: Yushane C. Shih, Brian P. Beaubien, Qingshan Chen, Jonathan N. Sembrano PII: DOI: Reference:

S1529-9430(18)30074-3 https://doi.org/10.1016/j.spinee.2018.02.016 SPINEE 57609

To appear in:

The Spine Journal

Received date: Revised date: Accepted date:

19-10-2017 7-2-2018 13-2-2018

Please cite this article as: Yushane C. Shih, Brian P. Beaubien, Qingshan Chen, Jonathan N. Sembrano, Biomechanical evaluation of sacroiliac joint fixation with decortication, The Spine Journal (2018), https://doi.org/10.1016/j.spinee.2018.02.016. 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.

SIJ Biomechanics with Threaded Implants

1

Biomechanical Evaluation of Sacroiliac Joint Fixation with

2

Decortication

3 4 5

Yushane C. Shih, MD, University of Minnesota, Department of Orthopaedic Surgery, Minneapolis, MN 55414

6

Brian P. Beaubien, MS, Primordial Soup (Psoup), Saint Paul, MN, 55105

7

Qingshan Chen, MS, Excelen Center for Bone and Joint Research, Minneapolis, MN 55105

8 9

Jonathan N. Sembrano, MD, University of Minnesota, Department of Orthopaedic Surgery, Minneapolis, MN 55414

10 11 12 13 14 15 16 17 18 19 20 21

Corresponding Author Address: Brian P. Beaubien [email protected] 287 East 6th Street Suite 160 Saint Paul, MN 55105 Fax: (651) 291-5045 Phone: (651) 291-5045

22

Keywords: Biomechanics

23

Spine

24

Sacroiliac

25

Fixation

26

Screws

27

Fusion

28 29

ABSTRACT Page 1 Page 1 of 17

SIJ Biomechanics with Threaded Implants 1

Background Context: Fusion typically consists of joint preparation, grafting and rigid fixation.

2

Fusion has been successfully used to treat symptomatic disruptions of the SIJ and degenerative

3

sacroiliitis using purpose-specific, threaded implants. The biomechanical performance of these

4

systems is important but has not been studied.

5

Purpose: To compare two techniques for placing primary (12.5mm) and secondary (8.5mm)

6

implants across the SIJ.

7

Study Design: A human cadaveric biomechanical study of sacroiliac joint (SIJ) fixation.

8

Methods: Pure moment testing was performed on fourteen human sacroiliac joints in flexion-

9

extension (FE), lateral bending (LB) and axial rotation (AR) with motion measured across the

10

SIJ. Specimens were tested Intact; after Destabilization (cutting the pubic symphysis); after

11

decortication and Implantation of a primary 12.5mm implant at S1 plus an 8.5mm secondary

12

implant at either S1 (S1/S1, n=8) or S2 (S1/S2, n=8); after Cyclic Loading; after removal of the

13

secondary implant. Ranges of motion (ROM) were calculated for each test. Bone density was

14

assessed on CT and correlated with age and ROM. This study was funded by Zyga Technology

15

but was run at an independent biomechanics laboratory.

16

Results: The mean ± SD intact ROM was 3.0±1.6 degrees in FE, 1.5±1.0 degrees in LB and

17

2.0±1.0 in AR. Destabilization significantly increased the ROM by a mean 60-150%.

18

Implantation in-turn significantly decreased ROM by 65-71%, below the intact ROM. Cyclic

19

loading did not impact ROM. Removing the secondary implant increased ROM by 46-88%

20

(nonsignificant). There was no difference between S1/S1 and S1/S2 constructs. Bone density

21

was inversely correlated to age (R=.69) and ROM (R=.36-.58).

22

Conclusions: Fixation with two threaded rods significantly reduces SIJ motion even in the

23

presence of joint preparation and after initial loading. The location of the secondary, 8.5mm

24

implant does not affect construct performance. Low bone density significantly affects fixation

25

and should be considered when planning fusion constructs. Findings should be interpreted in

26

the context of ongoing clinical studies.

27 28

INTRODUCTION Page 2 Page 2 of 17

SIJ Biomechanics with Threaded Implants 1

Low back pain is a common physical ailment with an annual prevalence of approximately 38%.1

2

The contribution of the sacroiliac joint (SIJ) to chronic low back pain ranges from 13 to 30%.2-8

3

Nonsurgical therapy is preferred in SIJ pain treatment, but is sometimes ineffective, and in

4

these cases fusion can provide a viable long-term solution.9-12 Traditional open SIJ fusion

5

represents a complex and invasive procedure involving exposure of the SIJ, decortication of

6

articular surfaces, graft harvesting and instrumented fixation. These procedures are associated

7

with significant blood loss, risk of neurovascular injury, disruption of ligamentous structures,

8

graft harvest-site morbidity, and protracted recovery.13-15

9

More recently, purpose-specific, minimally invasive sacroiliac joint fixation systems have been

10

introduced. These systems consist of up to three implants and are intended for treatment of

11

degenerative sacroiliitis and other non-traumatic conditions.16-17 Their implants are typically

12

shorter but larger in diameter compared to traditional sacroiliac screws. These systems aim to

13

fixate the SIJ to limit motion and thus facilitate fusion and minimize mechanically-induced pain;

14

therefore, a clear understanding of their contribution to fusion construct biomechanics is

15

important. Published data exist on the in vitro biomechanical performance of systems employing

16

triangular rods.18-19 However, there is a paucity of information on large-diameter (>8mm)

17

threaded implants, which are more common among modern purpose-specific SIJ fusion

18

systems, and no published data exist on systems involving joint decortication in addition to SIJ

19

fixation.

20

The purpose of this study was to determine the biomechanical impact of SIJ decortication and

21

implantation with purpose-specific, threaded implants and to evaluate the impact of secondary

22

implant location. Secondary objectives included evaluation of the relationship between fixation

23

and bone mineral density, and evaluation of the effect secondary implant removal on construct

24

rigidity. The system evaluated in this study was the SImmetry SIJ fusion system (Zyga

25

Technology, Minnetonka, MN), which involves minimally invasive decortication, grafting and

26

fixation with two threaded implants.

27

Page 3 Page 3 of 17

SIJ Biomechanics with Threaded Implants 1

MATERIALS AND METHODS

2

Specimen handling and preparation

3

Eight human lumbo-pelvic cadaveric specimens were procured fresh-frozen and screened using

4

Computed Tomography (CT) for bridging bone across the SIJ. Bone mineral density was also

5

estimated as attenuation on axial CT slices in Hounsfield Units (HU).20-21 Hounsfield Units were

6

evaluated by three authors over four ~0.8cm2 circular regions in the cancellous bone: mid-ilium

7

(bilaterally), lateral ala (bilaterally), mid-ala (bilaterally) and mid S1 vertebral body.

8

Measurements at each location were averaged across the three observers to obtain the

9

presented values.

10

Donor tissue was sharply dissected to yield osteoligamentous test specimens including the

11

pelvis, L4 and L5 vertebrae and intervening discs. Care was taken during dissection to prevent

12

damage to joint tissues. Wood screws were driven through L4 and into the L5 body to facilitate

13

fixation, and the entire L4 body was potted into a rigid resin (Smooth Cast 300Q; Smooth-On,

14

Eaton, PA). The pelvis was then potted at the left and right ischium with wood screws driven

15

into the bone prior to potting. The resultant specimen configuration allowed gripping of the L4

16

body and either the left or right innominate for testing of the respective SIJ. Loading via the

17

acetabulum was considered, but the ischium presented a more secure point of fixation, and

18

since measurements were made directly across the SIJ and the ilium was treated as a rigid

19

body, the exact point of load application was not considered important.

20

Specimens were tested immediately after potting or stored refrigerated at 4˚C overnight prior

21

to testing. Testing was performed at room temperature. Saline-soaked gauze was draped over

22

the specimens during testing to prevent dehydration.

23

Study Design and Simulated Surgery

24

Specimens were tested using a pure-moment loading scheme in five conditions as shown in

25

Figure 1. These states included the following: Intact, Destabilized, Implanted with two implants

26

(either “S1/S1” or “S1/S2” constructs), Cyclically-Loaded, and after removal of the secondary

27

implant (“S1 only”). Testing of the S1/S1 and S1/S2 constructs were performed contralaterally

28

with side (left/right) and order (first/second) randomized. Destabilization involved sharply

29

dissecting the pubic symphysis, and was performed for two reasons: first, cutting the pubic Page 4 Page 4 of 17

SIJ Biomechanics with Threaded Implants 1

symphysis increased implant loading thereby producing a worse-case configuration as may be

2

seen with a lax pelvis. Second, cutting the pubic symphysis allowed the left and right joints to

3

be tested independently, thereby allowing repeated-measures comparison of S1/S1 and S1/S2

4

constructs.

5

Simulated SIJ fusion surgery was performed using the SImmetry SIJ Fusion System (Zyga

6

Technology, Inc., Minneapolis, MN, FDA approved), which involves decortication, grafting and

7

fixation with two threaded implants. Procedures were performed under fluoroscopy per the

8

manufacturer’s recommended surgical technique except for grafting, which was not performed

9

as it was not considered to impact fixation. The SIJ decortication was accomplished using a

10

purpose-specific instrument, which has a flexible blade that is deployed into the joint to remove

11

cartilage and prepare, but not remove, the joint cortices (Figure 2). Access was perpendicular to

12

the joint surface and centered in the articular region of the joint. The primary, 12.5mm implant

13

was placed through this decortication access channel, which orients the implant towards the

14

first sacral segment (“S1”). A secondary, 8.5mm implant was either placed dorsal and superior

15

to the primary implant (“S1/S1”, Figure 2) or ventral and distal to the primary implant (“S1/S2”,

16

Figure 2). Accommodations to these constructs were made where sacral dysmorphism

17

prevented the prescribed placement and were reported.

18

After testing, each construct was loaded to 5 Newton-Meters (Nm) for 1,000 cycles to simulate

19

initial post-operative motion. Pilot testing indicated no difference between 1,000 and 5,000

20

cycles, and since the implantation and testing regime for both sides took approximately 8 hours,

21

longer cyclic loading durations were not pursued to minimize specimen degradation.

22

A final test configuration was performed after removal of the secondary implant to evaluate its

23

incremental contribution to construct rigidity. This configuration was tested last to prevent

24

damage to the primary implant fixation prior to testing the two-implant construct.

25

Loading and Motion Measurement

26

Pure-moment flexibility testing was conducted in the three primary body planes: flexion-

27

extension (nutation/counternutation), lateral bending (coronal plane rotation) and axial rotation

28

(transverse plane rotation). Moments were applied to L4 and in-turn across the L5-S1 disc and

29

ligaments to the SIJ using the previously-described system shown in Figure 3 (MTS 858 Mini

Page 5 Page 5 of 17

SIJ Biomechanics with Threaded Implants 1

Bionix, MTS, Eden Prairie, MN).22 Loading was applied unilaterally with the contralateral ilium

2

anchored vertically, but floating freely in anteroposterior and lateral translation; rotation planes

3

normal to the applied moment were also allowed to float freely. This “pure moment” loading

4

scheme was not intended to replicate in vivo loading, but was selected to evaluate construct

5

rigidity in each anatomic plane and to allow comparison to previously-reported data.18 Moments

6

of ±7.5 Nm were applied based on published data of lower lumbar spine loading23, and were

7

applied in moment-control at a rate of 0.5 Nm/s. Two preconditioning cycles were applied

8

followed by the test cycle, during which the specimen motion was measured.

9

Motions for the third loading cycle were measured directly across the SIJ using a non-contact

10

motion measurement system. One triad of reflective markers was placed into the sacrum at S1,

11

and another into the ilium near the joint. Marker motions were tracked using a non-contact

12

motion measurement system (VICON, Oxford Metrics, UK), and were converted into Euler

13

angles as previously-described.22 The range of motion (ROM) was calculated as the difference

14

between the peak positive and negative angles in plane of applied loading.

15

This study was funded by Zyga Technology, Inc., but was run according to previously published

16

protocols at an independent, ISO-certified laboratory. Neither surgeon authors nor their

17

institution received any funding or consulting fees.

18

Sample Size and Data Analysis

19

The sample size was calculated using the following assumptions: minimum detectable

20

difference of 1.5 degrees (considered by authors to be clinically-relevant), a standard deviation

21

of 50% of the mean, a mean intact ROM of 2 degrees, a power of 0.8 and a significance level

22

of p=.05. A minimum sample size of 6 specimens was identified, but 8 specimens were

23

procured to account for possible attrition due to degradation or excessive rigidity as previously

24

reported.18,19

25

Whereas destabilized segments exhibited a region of laxity, or “Neutral Zone”, both intact and

26

instrumented specimens exhibited linear load-displacement relationships. Thus, the neutral zone

27

was not evaluated or compared, nor was the specimen stiffness, which was not independent,

28

but rather inversely related to the ROM.

29

Statistical comparisons were made on ROM for construct group (S1/S1 and S1/S2) and fixation Page 6 Page 6 of 17

SIJ Biomechanics with Threaded Implants 1

state using repeated measures analysis of variance methods (ANOVA). Pairwise comparisons

2

were made using a Holm–Šidák post-hoc test. Linear regression was used to evaluate the

3

relationship between age and mean bone mineral density, and between mean bone mineral

4

density and post-cyclic loading ROM.

5 6

RESULTS

7

Specimen details are shown in Table 1. One specimen (SI-06) was discarded during testing due

8

to degradation of the bone and potting interface. Decortication and implantation were

9

performed as intended apart from accommodations made for abnormal anatomy (i.e., sacral

10

dysmorphism). Specifically, one dysmorphic specimen had the 12.5 mm, primary implant placed

11

at S2 instead of S1, where there was not sufficient room for the larger implant. A second

12

dysmorphic specimen had sufficient room for the primary implant in S1 for the S1/S2 construct,

13

but there was not sufficient room for both implants in the S1 body.

14

Significant differences were not observed across constructs (S1/S1 vs. S1/S2, p>.1), but were

15

observed across states (p<.05). Compared to the intact state, destabilization significantly

16

increased the ROM in flexion-extension by an overall mean 60%, lateral bending by 151% and

17

axial rotation by 100% (each at p≤.02). Implantation in-turn significantly reduced motion by a

18

mean 71%, 65% and 65% in each direction, respectively (each at p<.001). Cyclic loading had a

19

minimal impact on ROM before vs. after cycling (p>.6). Subsequent removal of the secondary,

20

8.5mm implant increased the ROM by 88%, 73% and 46%, respectively, compared to the post-

21

cyclically loaded state; however, this difference was not statistically significant (p=.1 to .6).

22

ROM results are summarized in Figure 4 and Table 2.

23

Bone density varied along the screw trajectory, and was negatively correlated with age and

24

ROM. Radiodensity data, shown in Table 1, showed the highest density in the Ilium (mean 213

25

HU), followed by the S1 Body (153 HU), the Lateral Ala (100 HU) and the Central Ala (11 HU).

26

Some Central Ala measurements were less than zero indicating a radiodensity less than that of

27

distilled water. Bone mineral density was inversely correlated with age (R = .69) and ROM in

28

flexion-extension (R= .58), lateral bending (R =.36) and axial rotation (R=.41) as shown in

29

Figure 5. Gender-related differences could not be assessed as there was only one male

30

specimen. Page 7 Page 7 of 17

SIJ Biomechanics with Threaded Implants 1

DISCUSSION

2

Fixation construct rigidity was evaluated in fourteen sacroiliac joints before and after

3

decortication and fixation with two threaded implants. The average ± standard deviation intact

4

ROM was 3.0±1.6 degrees in flexion-extension, 1.5±1.0 degrees in lateral bending and 2.0±1.0

5

in axial rotation. Destabilization significantly increased the ROM by a mean 60-150%, and

6

decortication and fixation in-turn significantly decreased the ROM by a mean 65-71%. Cyclic

7

loading had a small (6-13%), non-significant impact on the ROM. Removal of the secondary

8

implant resulted in a mean 46-88% increase in the ROM, but this effect was not statistically

9

significant due to a high degree of variability of the effect. There was no significant difference in

10

means of the S1/S1 vs. S1/S2 constructs.

11

A systematic search was performed to identify comparative published literature. This search

12

generated 24 results of which eight in vitro studies were from the trauma literature: three used

13

Sawbones or composite pelvis models24-27, one was based on a finite element model of unstable

14

pelvic ring injuries28 and two were performed on human cadaveric pelvises.29-30 The results

15

obtained in this study compared favorably with those reported previously. Intact ranges of

16

motion were consistent with previous studies measuring in vivo motions11,31 as well as those

17

obtained in vitro using similar methods.18 Cutting of the pubic symphysis in the current study

18

had a similar effect to that seen previously in vitro19 and a lesser effect vs. disruption of both

19

the symphysis and posterior ligaments.18

20

Compared to the destabilized state, the two-implant fixation constructs evaluated in this study

21

significantly decreased the ROM by 65-71%, even in the presence of decortication. This was a

22

greater reduction compared to that seen with triangular rods, which exhibited a reduction in

23

ROM from the destabilized state of 29-38% in one study18 and 39-41% in another.19 While

24

caution must be exercised when comparing across studies, both studies were performed on a

25

system that has been directly compared to that of the current study and found to produce

26

similar results.22 The impact of removing the secondary implant was similar to that seen in other

27

studies. However, single-implant testing was performed last to favor optimal comparison of the

28

two-implant constructs. For this reason, differences between one- and two-implant constructs

29

should be interpreted with caution as there may be an order-related contribution to the effect.

Page 8 Page 8 of 17

SIJ Biomechanics with Threaded Implants 1

The implant system in this study used larger diameter implants than are typically used in

2

trauma applications, and provided accordingly greater reductions in SIJ motion with greater

3

reductions with two vs. one implant. In a human cadaveric fracture model, van Zwienen et al

4

reported a 28% and 44% reduction in fracture site motion and 170% and 300% increase in

5

rotational stiffness with the addition of a second screw for S2 and S1, respectively.29 Yinger, et

6

al evaluated the effectiveness of multiple fixation constructs for a posterior pelvic ring injury

7

plastic model and concluded that the two transverse iliosacral screw construct provided

8

substantial rigidity, only slightly less than the strongest construct of a screw with multiple

9

anterior plates.27 However, Sagi, et al found no benefit to a supplementary screw in addition to

10

a properly placed S1 iliosacral screw, and there was no statistically significant difference

11

between one or two SI screws under hemipelvis rotational or linear displacement of the SI

12

joint.25

13

Bone density, which was estimated by CT attenuation, averaged 11-213 HU’s in regions

14

spanning a trajectory of the S1 implant and was inversely correlated with age (R=.69) and ROM

15

(R=.36-.58). One prior study from a range of in vivo renal scans of patients aged 13-87 years

16

reported a similar HU range that correlated with age, especially in the lateral sacral ala.20

17

Another study in trauma patients aged 18-49 found lower bone mineral density at the S2 level

18

vs. the S1 level within the body.21 Overall, the relationship between fixation and bone density is

19

important as 45% of sacropelvic instrumentation failures have S1 screw haloing or pullout,32,33

20

and sacral-sided loosening of triangular rod fixation across the sacrum is a known issue.20

21

Adequate bone density appears to be of critical importance in the SIJ given its regions of widely

22

varying density, and especially when the denser bone in the central S1 body is not obtained.

23

Sacral dysmorphism necessitated modified fixation constructs in two specimens. Dysmorphism

24

is a common anatomic variant of the normal SIJ that complicates safe trans-articular fixation.

25

However, in this study all implants placed in dysmorphic specimens were found to reside

26

completely within the bone, and their placement resulted in ROM’s within the range seen for

27

constructs in normal anatomy. The small size of this subgroup did not allow statistical

28

comparison; however, these results demonstrate that biomechanically suitable fixation using

29

modified constructs can be obtained in the presence of sacral dysmorphism.

30

This study had several limitations. First, although the sample size was deemed adequate to

31

perform comparisons to the desired power, it was still limited at a total of 14 joints. Second, Page 9 Page 9 of 17

SIJ Biomechanics with Threaded Implants 1

specimens evaluated in cadaveric studies are often obtained secondary to end-of-life morbidities

2

and can degrade during procurement and handling, which may affect bone quality. Another

3

limitation is that the pure moment flexibility loading scheme may not accurately represent

4

typical in vivo loading. Pure moments are correlated to in vivo motions, such as nutation during

5

push-off, rotation and bending during gait, and positioning during activities of daily living, but

6

do not match the complexity of in vivo loading. It was not the goal of this study to match such

7

loading, but instead to apply moments in each body plane and observe the resulting motion to

8

understand the mechanical integrity of the fixation construct. Finally, the pubic symphysis of

9

each pelvis was cut to isolate left and right joints, and to provide a worst-case, lax pelvis. This

10

allowed independent evaluation of left and right joints, and provide a worst-case fixation

11

environment to challenge the implants; however, this destabilization may have accentuated

12

differences. Although this study was funded by industry, it adhered to state-of-the art protocols

13

and equipment and was run at an independent, ISO-certified laboratory, and findings may apply

14

to other implant systems having similar attributes. The results of this study must be considered

15

in the context of this model, but may be considered relevant as the system used in this study is

16

also indicated in the case of SIJ disruptions.

17

CONCLUSION

18

This study has shown that fixation with two threaded rods significantly reduces SIJ motion even

19

in the presence of minimally invasive decortication, and even after initial cyclic loading. A

20

secondary, 8.5mm implant tends to provide a more rigid construct over a single, 12.5mm

21

implant at S1, but the location of the secondary implant (i.e., in S1 or S2) does not affect the

22

immediate postoperative construct rigidity. Low bone density adversely affects construct rigidity

23

and efforts should be made to improve fixation in this setting. Although not addressed in this

24

study, options may include placing a second supplemental screw, or attempting to place longer

25

screws to gain purchase in the denser sacral vertebral body bone. Overall, these findings may

26

influence surgical decision-making but should be interpreted in the context of ongoing clinical

27

studies.

28

Page 10 Page 10 of 17

SIJ Biomechanics with Threaded Implants 1 2

REFERENCES 1. Hoy, D. et al. A systematic review of the global prevalence of low back pain. Arthritis &

Rheumatism 64, 2028–2037 (2012).

3 4

2. Bernard, T. N. & Kirkaldy-Willis, W. H. Recognizing specific characteristics of nonspecific low back pain. Clin. Orthop. Relat. Res 266–280 (1987).

5 6

3. Schwarzer, A. C., Aprill, C. N. & Bogduk, N. The sacroiliac joint in chronic low back pain.

Spine 20, 31–37 (1995).

7 8

4. Maigne, J. Y., Aivaliklis, A. & Pfefer, F. Results of sacroiliac joint double block and value of sacroiliac pain provocation tests in 54 patients with low back pain. Spine 21, 1889–

9

1892 (1996).

10 11

5. Depalma, M. J., Ketchum, J. M. & Saullo, T. R. Etiology of Chronic Low Back Pain in

12

Patients Having Undergone Lumbar Fusion. Pain Med (2011). doi:10.1111/j.1526-

13

4637.2011.01098.x

14

6. Katz, V., Schofferman, J. & Reynolds, J. The sacroiliac joint: a potential cause of pain after lumbar fusion to the sacrum. J Spinal Disord Tech 16, 96–99 (2003).

15 16

7. Sembrano, J. N. & Polly, D. W. How often is low back pain not coming from the back?

Spine 34, E27-32 (2009).

17 18

8. Dreyfuss, P., Dreyer, S. J., Cole, A. & Mayo, K. Sacroiliac joint pain. J Am Acad Orthop

Surg 12, 255–265 (2004).

19 20

9. Ashman, B., Norvell, D. C. & Hermsmeyer, J. T. Chronic sacroiliac joint pain: fusion

21

versus denervation as treatment options. Evid Based Spine Care J 1, 35–44 (2010).

22

10. Whang P, et al. Sacroiliac Joint Fusion Using Triangular Titanium Implants vs. Non-

23

Surgical Management: Six-Month Outcomes from a Prospective Randomized Controlled

24

Trial. Int J Spine Surg. 2015 Mar 5;9:6.

Page 11 Page 11 of 17

SIJ Biomechanics with Threaded Implants 1

11. Sturesson, B., Selvik, G. & Udén, A. Movements of the sacroiliac joints. A roentgen stereophotogrammetric analysis. Spine 14, 162–165 (1989).

2 3

12. Sturesson, B. et al. Six-month outcomes from a randomized controlled trial of minimally

4

invasive SI joint fusion with triangular titanium implants vs conservative management.

5

Eur Spine J 26, 708–719 (2017).

6

13. Schutz, U. & Grob, D. Poor outcome following bilateral sacroiliac joint fusion for degenerative sacroiliac joint syndrome. Acta Orthop Belg 72, 296–308 (2006).

7 8

14. Simpson, L. A., Waddell, J. P., Leighton, R. K., Kellam, J. F. & Tile, M. Anterior approach

9

and stabilization of the disrupted sacroiliac joint. J Trauma 27, 1332–1339 (1987).

10

15. Waisbrod, H., Krainick, J. U. & Gerbershagen, H. U. Sacroiliac joint arthrodesis for chronic lower back pain. Arch Orthop Trauma Surg 106, 238–240 (1987).

11 12

16. Rudolf, L. Sacroiliac Joint Arthrodesis-MIS Technique with Titanium Implants: Report of

13

the First 50 Patients and Outcomes. Open Orthopaedics Journal 6, 495–502 (2012).

14

17. Block, J. & Miller, L. Minimally invasive arthrodesis for chronic sacroiliac joint dysfunction

15

using the SImmetry SI Joint Fusion system. Medical Devices: Evidence and Research

16

125 (2014). doi:10.2147/MDER.S63575

17

18. Lindsey, D. P. et al. Evaluation of a minimally invasive procedure for sacroiliac joint

18

fusion - an in vitro biomechanical analysis of initial and cycled properties. Med Devices

19

(Auckl) 7, 131–137 (2014).

20

19. Soriano-Baron, H. et al. The Effect of Implant Placement on Sacroiliac Joint Range of Motion: Posterior vs Trans-articular. Spine (2015). doi:10.1097/BRS.0000000000000839

21 22

20. Hoel, R. J., Ledonio, C. G. T., Takahashi, T. & Polly, D. W. Sacral bone mineral density (BMD) assessment using opportunistic CT scans. J. Orthop. Res. 35, 160–166 (2017).

23

Page 12 Page 12 of 17

SIJ Biomechanics with Threaded Implants 1

21. Salazar, D. et al. Investigation of bone quality of the first and second sacral segments

2

amongst trauma patients: concerns about iliosacral screw fixation. J Orthop Traumatol

3

16, 301–308 (2015).

4

22. Wheeler, D. J. et al. Inter-laboratory variability in in vitro spinal segment flexibility testing. J Biomech 44, 2383–2387 (2011).

5 6

23. White, A. & Panjabi, M. M. SI Joint Biomechanics. in Clincal Biomechanics of the Spine Ch. 2, (Lippincott-Raven 1990).

7 8

24. Vigdorchik, J. M. et al. A biomechanical study of standard posterior pelvic ring fixation versus a posterior pedicle screw construct. Injury 46, 1491–1496 (2015).

9 10

25. Sagi, H. C., Ordway, N. R. & DiPasquale, T. Biomechanical analysis of fixation for

11

vertically unstable sacroiliac dislocations with iliosacral screws and symphyseal plating. J

12

Orthop Trauma 18, 138–143 (2004).

13

26. Sar, C. & Kilicoglu, O. S1 pediculoiliac screw fixation in instabilities of the sacroiliac

14

complex: biomechanical study and report of two cases. J Orthop Trauma 17, 262–270

15

(2003).

16

27. Yinger, K., Scalise, J., Olson, S. A., Bay, B. K. & Finkemeier, C. G. Biomechanical comparison of posterior pelvic ring fixation. J Orthop Trauma 17, 481–487 (2003).

17 18

28. Zhang, L., Peng, Y., Du, C. & Tang, P. Biomechanical study of four kinds of

19

percutaneous screw fixation in two types of unilateral sacroiliac joint dislocation: a finite

20

element analysis. Injury 45, 2055–2059 (2014).

21

29. van Zwienen, C. M. A., van den Bosch, E. W., Snijders, C. J., Kleinrensink, G. J. & van

22

Vugt, A. B. Biomechanical comparison of sacroiliac screw techniques for unstable pelvic

23

ring fractures. J Orthop Trauma 18, 589–595 (2004).

Page 13 Page 13 of 17

SIJ Biomechanics with Threaded Implants 1

30. Mears, S. C., Sutter, E. G., Wall, S. J., Rose, D. M. & Belkoff, S. M. Biomechanical

2

comparison of three methods of sacral fracture fixation in osteoporotic bone. Spine 35,

3

E392-395 (2010).

4

31. Jacob, H. a. C. & Kissling, R. O. The mobility of the sacroiliac joints in healthy volunteers between 20 and 50 years of age. Clin Biomech (Bristol, Avon) 10, 352–361 (1995).

5 6

32. Lu, J., Ebraheim, N. A., Yang, H. & Heck, B. E. Anatomic evaluation of the first three

7

sacral vertebrae and dorsal screw placement. Am J. Orthop. 29, 376–379 (2000).

8

33. Harimaya, K. et al. Etiology and revision surgical strategies in failed lumbosacral fixation of adult spinal deformity constructs. Spine 36, 1701–1710 (2011).

9 10 11 12 13

Page 14 Page 14 of 17

SIJ Biomechanics with Threaded Implants 1

FIGURE LEGENDS

2

Figure 1: Study design flowchart. Specimens were tested 1) Intact, 2) “Destabilized” (after

3

cutting the pubic symphysis, 3) “Implanted” with the prescribed construct, 4) “Cyclically

4

Loaded” for 1000 cycles and 5) “S1 Only” after removal of the secondary implant. Left vs. right

5

sides were randomized to receive two implants at the S1 level (“S1/S1”) and one implant at S1

6

and another at S2 (“S1/S2”).

7

Figure 2: Test setup. The upper gimbal rotates to impose the applied moment while the linear

8

slide below allows free translation. Note that the specimen is potted in the anatomic position,

9

with the pelvis tilted such that the alar slopes are oriented vertically.

10

Figure 3: Decortication (A and B) and fixation with threaded implants (C and D) was performed

11

per the manufacturer’s recommended technique. For each specimen, one side had both the

12

12.5mm and 8.5mm diameter implants placed toward the S1 body (“S1/S1”, C), whereas the

13

other side had the 12.5mm implant placed toward S1 and the secondary, 8.5mm implant

14

toward S2 (“S1/S2”, D).

15

Figure 5: Bone mineral density was inversely correlated with donor age (left), and ROM was in-

16

turn inversely-correlated to bone mineral density (right). Only Flexion Extension vs. bone

17

density data are shown (R2=.58); similar trends were observed in Lateral Bending (R2=.36) and

18

Axial rotation (R2=.41).

19

Figure 4: Range of Motion for all directions tested for the S1/S1 and S1/S2 constructs. Bars

20

indicate statistically-significant differences (p<.05). The S1 and S2 constructs did not differ

21

significantly from one another.

22 23 24

Page 15 Page 15 of 17

SIJ Biomechanics with Threaded Implants

1

Table 1: Specimen details and Bone Mineral Density estimates. “*” indicates specimens

2

having bone noted to be soft during implantation. “†” indicates the specimen discarded due to

3

degradation during testing. **Mean bone density estimate reported as CT attenuation.

Cancellous Bone Density (HU)** (left to right) Left

Left

Left

Ilium

Lateral

Central

Ala

Ala

Right

Right

Central

Lateral

Ala

Ala

Right

ID

Mid-S1

Ilium

2

Age (y)

Gender

BMI (kg/m )

SI-01

31

Male

26.3

255

103

-1

186

3

154

249

SI-02

39

Female

23.3

394

88

28

284

69

102

314

SI-03

34

Female

23.4

294

157

55

257

77

114

302

SI-04

62

Female

14.2

107

61

-17

105

-14

15

131

SI-05*

59

Female

37.6

133

74

-7

87

2

78

150

SI-06*†

57

Female

20.5

210

78

-16

144

-1

116

236

SI-07

52

Female

21.8

137

101

34

63

29

118

179

SI-08*

47

Female

32.3

150

78

-33

98

-35

171

167

Mean

48

-

24.9

210

93

6

153

16

108

216

Std. Dev

12

-

7.2

99

30

30

82

39

48

70

4 5

Page 16 Page 16 of 17

SIJ Biomechanics with Threaded Implants 1

Table 2: Pairwise differences represented as degrees and percentage changes in

2

Flexion-Extension (FE), Lateral Bending (LB) and Axial Rotation (AR). Difference in means Comparison

Direction Percent Change

FE

-0.3

-21%

.857

LB

-0.2

-17%

.779

AR

0.3

22%

.122

FE

1.8

60%

.023

LB

2.2

151%

<.001

AR

2.0

100%

<.001

FE

-3.4

-71%

<.001

LB

-2.4

-65%

<.001

AR

-2.6

-65%

<.001

FE

-1.6

-54%

.040

LB

-0.2

-12%

.946

AR

-0.6

-29%

.449

FE

0.2

13%

.757

LB

0.1

6%

.956

AR

0.2

13%

.635

FE

-2.2

-45%

.005

LB

-1.6

-44%

.001

AR

-1.7

-43%

.001

S1/S1 vs. S1/S2 Constructs (implanted)

Destabilization (vs. intact)

Implanted, with decortication (vs. destabilized)

Implanted, with decortication (vs. intact)

Cyclic Loading

p-value Degrees

(vs. implanted)

S1 only (vs. destabilized, always tested last)

3

4 Page 17 Page 17 of 17