BIOMECHANICS OF SPINAL INSTRUMENTATION

SPINAL INSTRUMENTATION

Goal:– To maintain anatomic alignment of injured spinal segments

by sharing the loads acting on the spine until a solid biologic fusion takes place

Prerequisite Understandings:– Pathology– Spine Biomechanics– Biomechanics of implant constructs

SPINAL INSTRUMENTATION

ADVANTAGES - Promote fusion - Enhance early mobilization

DISADVANTAGES - Adverse effects on the adjacent segments - Stress-shielding effects on the stabilized segment - Hardware Failure

HARDWARE FAILURE

Breakage of Rods or Screws Hook Dislodgement Loosening Screw Pull-out

in cases of extreme Osteoporosis

Biomechanical Strength (Static and Dynamic)– Impact strength– Surgical construct strength– Metal-metal interface strength– Bone-metal interface strength

Stability– Segmental stiffness/flexibility

Load Sharing– Implant survival– Stress shielding– Fusion rate and quality

Biomechanical Consideration Factors

Implant Design Factors

Mechanical strength:Profile:

– Low profile preferred particularly in anterior instrumentation

User Friendliness:– Top loading– Poly-axial screw insertion– Lateral-medial ajustment– Easy-to-use surgical insruments

Versatility:

Biomechanical Consideration Factors

Fretting Corrosion:– Any damage to the material that takes place at the edge of

contacting parts or within the local contact area• Associated with wear, surface damage and accumulated debris

• The rate of fretting corrosion is mostly governed by micromotion between component surfaces within an interconnection

– Cause of late infection and loosening of interconnections– Greater tendency for:

• designs with lower strength interconnections

• Improper assembly such as misalignment and incomplete seating

• Higher applied loads

PULL-OUT AND CYCLIC FAILURES OF THE ANTERIOR VERTEBRAL SCREW

FIXATION IN RELATION TO BONE MINERAL DENSITY

Howard S. An, M.D.

Tae-Hong Lim, Ph.D.

Christopher Evanich, M.D.

Toru Hasegawa, M.D.

Kaya Y. Hasanoglu, M.D.

Linda McGrady, B.S.

Anterior Spinal Instrumentation

Fractures Tumors Correction of Deformities

Hardware Failure

Breakage of rods or screws Screw loosening Screw pullout

in cases of extreme Osteoporosis

PREVIOUS STUDIES

Screw Pullout Strength in relation to Design Variables and Insertion Methods of Pedicle Screw

- Zindrick et al., 1986; Krag et al., 1986; Skinner -et al., 1990; Daftari et al., 1992

Screw Pullout Strength in relation to Bone Mineral Density (BMD)

- Coe et al., 1990; Soshi et al., 1991; Yamagata et al., 1992; An et al., 1993

PREVIOUS STUDIES

Loosening of the Pedicles Screw in relation to

Design Variables and Insertion Methods of Pedicle Screw

- Zindrick et al., 1986 No BMD Measurement

Correlation between

Anterior Vertebral Screw Fixation Failure

And

Bone Mineral Density (BMD)

Of the Verebral Body

PURPOSE OF THE STUDY

Relationships between BMD and:

- Pull-out Strength

- Cyclic Screw Loosening

PART I: Pull-out Test (6 fresh-frozen lumbar spines)

PART II: Cyclic Loosening Test (5 fresh-frozen lumbar spines)

MEASUREMENT OFBMD (g/cm2)

Dual Energy X-ray Absorptiometry (DEXA) - Scan Speed: 60 mm/sec - Resolution: 1 x 1 mm

Lateral view of each lumbar spine

Kaneda Screw

(6.5 mm diameter; 55 mm long)

Measurement of Torque (Nm) and Vertebral Width (mm) for screw insertion

Pull-out Strength Test Screw was pulled out

along its long axis.

Loading Rate: 10 mm/min

(Displacement Control)

DATA ANALYSIS(Pull-out Strength Test)

Pull-out Strength– Maximum pull-out force in the load

displacement curve

Regression Analyses

Cyclic Loading Test to Induce Screw Loosening

Cyclic loading was applied to the screw in the cephalad-caudal direction using an MTS

machine.

Loading Frequnecy: 0.5 Hz Loading Amplitude: 200 N

100 N

DATA ANALYSIS(Cyclic Loosening Test)

Number of Loading Cycles (NLC) to Induce the Screw Loosening:

– when the displacement > the displacement at the first peak load + 1 mm.

Regression analysis NLCs for

– specimens with BMD < 0.45 vs. BMD > 0.45.

RESULTS

Means (SD) ofMeasured Parameters

(Pull-out Test)

BMD: 0.58 g/cm2 (STD 0.14)

Torque: 0.86 Nm (STD 0.2)

Depth: 41.0 mm (STD 3.28)

Pull-out Strength: 211 N (STD 124)

CORRELATION COEFFICIENTS(Pull-out Test)

Pull-outStrength BMD Torque (T) Width (W)

BMD

Torque (T)

Width (W)

Pull-outStrength 1.0 0.85 0.47 No relation

1.0 0.58 No relation

1.0 No relation

1.0

-

-

---

-

CORRELATION

Pull-out Strength vs. BMDR = 0.83 (p < 0.0002)

Pull-out Strength = -226 + 774 x BMD

Pull-out Strength vs. BMD

BMD (g/cm2)

Pull-out Force (N)

0

100

200

300

400

500

0.3 0.4 0.5 0.6 0.7 0.8

Means (SD) ofMeasured Parameters(Cyclic Loosening Test)

BMD: 0.32 g/cm2 (STD 0.10)

Torque: 6.9 Kg-cm (STD 2.8)

NLC: 149 (STD 234)

(5 to 960)

CORRELATION(Cyclic Loosening Test)

NLC vs. BMDSecond Order Polynomial Relationship

NLC = -1190 BMD + 3168 BMD2

R = 0.80 (p < 0.01)

NLC vs. BMD

BMD (g/cm/cm)

NLC

0

200

400

600

800

1000

0.3 0.4 0.5 0.6 0.7

CORRELATIONS(Cyclic Loosening Test)

NLC vs. Screw Insertion Torque:

Second Order Polynomial

R = 0.68 (p < 0.01)

BMD vs. Screw Insertion Torque: Linear

R = 0.51 (p < 0.02)

MEAN NLCs (SD)

12 Specimens with BMD<0.45:

18.0 (20.1)

11 Specimens with BMD>0.45:

270.8 (278.8)These were significantly different.

(p=0.003)

BMD: – correlation with pull-out strength as well as cyclic loosening failure of

the anterior vertebral screw– a useful means to evaluate the severity of osteoporosis

DEXA: measurement of BMD - Low-radiation dose - Short scanning time - Increased image resolution - Improved precision

CONCLUSIONQuantitative assessment of BMD using a

DEXA unit may be a good predictor of anterior vertebral screw fixation failure.

BMD < 0.45 g/cm2 may be the critical value for osteoporosis in screw loosening.

PREDICTION OF FATIGUE LOOSENING FAILURE OF THE

PEDICLE SCREW FIXATION

Tae-Hong Lim, Ph.D.

Lee H. Riley, III, M.D.

Howard S. An, M.D.

Linda M. McGrady, B.S.

John Klein, M.S.

Pullout Strength of the Pedicle Screw

In relation to Pedicle Screw Design Variables- Zindrick et al., 1986; Krag et al., 1986; Skinner -et al.,

1990; Daftari et al., 1992

Effects of screw insertion torque, screw hole preparation method, and screw angulation

- Skinner et al. 1991; Daftari et al. , 1992; Zdeblick et al. , 1993

In relation to Bone Mineral Density (BMD)

- Coe et al., 1990; Soshi et al., 1991; Yamagata et al., 1992

Loosening Failure of the Screw Fixation

Pedicles Screw Loosening in relation to Design Variables and Insertion Methods of Pedicle Screw

- Zindrick et al., 1986

Anterior Vertebral Screw Loosening in relation to the Bone Mineral Density of the Vertebral Body

- Lim et al., 1994

PURPOSEFatigue Loosening

of the Pedicle Screw

in relation to

Bone Mineral Density

of the Vertebral Body

PURPOSEFatigue Loosening

of the Pedicle Screw

in relation to

Pedicle Size and Screw Insertion Torque

MATERIALSThree fresh frozen human lumbar

spines (L1 - L5) were used in this study.

Anterior-posterior and lateral radiographs were taken to exclude spines with gross pathology.

MEASUREMENT OFBMD (g/cm2)

Dual Energy X-ray Absorptiometry (DEXA) - Scan Speed: 60 mm/sec - Resolution: 1 x 1 mm

Lateral view of each lumbar spine

Pedicle Size Measurement– Pedicle Height (PH): Long Axis (cephalad-caudal

direction)– Pedicle Width (PW): Short Axis (medial-lateral

direction) Pedicle Screw Placement

– A 3.5 mm drill hole and 6.25 mm tapper– A 6.25 mm Steffee Screw (40 mm long)

Screw Insertion Torque (TQ) – Measured using a Torque Wrench

Cyclic Loading Test to Induce Screw Loosening

Cyclic loading was applied to the screw in the cephalad-caudal direction using an MTS machine.

Loading Frequency: 0.5 Hz Loading Amplitude: 200 N

100 N

Cyclic Test Set-up

MMA

MTSLoad Cell

Fixture

Loading Direction

RAM

DATA ANALYSISNumber of Loading Cycles (NLC) to

Induce the Screw Loosening:– when the displacement > the displacement at the first

peak load + 1 mm.

Regression analysis – Relationships between NLC, BMD, PH, PW, and TQ

Means (SDs) of the Measured Parameters

BMD: 0.465 g/cm2 (0.187)

PH: 15.57 mm (4.97)

PW: 12.08 mm (2.44)

TQ: 9.25 Kg-cm (3.14)

NLC: 300 (377)

Correlation between the Measured Parameters

BMD PH PW TQ NLC

BMD 1.0 - - - -

PH 0.54 1.0 - - -

PW *0.07 0.51 1.0 - -

TQ *0.44 0.59 *0.11 1.0 -

NLC 0.58 0.72 *0.42 0.50 1.0

* indicates relationship with no statistical significance (p > 0.05).

Linear Relationship between NLC and BMD

0

200

400

600

800

1000

1200

0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

NLC

BMD (g/cm/cm)

Linear Relationship between NLC and PH

0

200

400

600

800

1000

1200

5 10 15 20 25

PH (mm)

NLC

Multiple regression analysis demonstrated a significant correlation between NLC and other measured parameters.

(R = 0.79, p = 0.02)

NLC = 775.8 x BMD + 22.4 x PH -

44.3 PW + 15.5 TQ - 17.2

The stepwise regression analysis

revealed that PH was the most significant

predictor of pedicle screw loosening.

DISCUSSION

Screw loosening: a significant complication in spinal fixation.

Pedicle screw loosening was experimentally induced under cyclic loading in this study.

Application of the cyclic load in the cephalad-caudal direction on the screw

– in order to simulate the load transferred to the screw by the connecting rod or plate

NLC was significantly correlated with BMD, PH, and TQ, while PH was the most significant predictor of screw loosening failure.

– Honl et al. also found that the amount of loosening correlated more with pedicle geometry than BMD. (Proceedings, 2nd World Congress of Biomechanics, 1994)

Limitations:– Small number of Specimens;– No studies on the optimal ratios of the screw diameter to PH for the best

fixation strength

DISCUSSION

A significant relationship between the fatigue loosening of the pedicle screw and PH and BMD was found.

The assessment of the pedicle geometry as well as the vertebral body BMD may provide valuable preoperative information to predict the early loosening failure of the pedicle screw fixation.

BIOMECHANICAL EVALUATION OF ANTERIOR SPINAL

INSTRUMENTATIONJae Won You, M.D., Ph.D.

Tae-Hong Lim, Ph.D.

Howard S. An, M.D.

Jung Hwa Hong, M.S.

Jason M. Eck, B.S.

Linda McGrady, B.S.

Anterior Spinal Instrumentation

Plate Type– Syracuse I-Plate, CASP, Z-Plate, University Plate, etc.

Rod Type– Zielke, Kaneda, TSRH instrumentation, etc.

Purpose of the Study

Compare the biomechanical Stability of anterior fixation constructs,

particularly comparing the use of plate vs. rod type anterior fixators.

METHODS

Specimen Preparation

20 Fresh Calf Lumbar Spines (L2-L5)– The lumbar spine dissected free of muscles, leaving ligaments, capsules,

and disc intact

L5 fixed to the loading rig, and L2 mounted in an unconstrained loading setup.

3 markers reflecting infra-red light attached to L3 and L4 vertebral body.

Loading Setup

Pure Moments:– using dead weights, unconstrained

6 Directions:– FLX, EXT, RLB, LLB, RAR, and LAR

Incremental Loading:– to a maximum of 6.4 Nm in 6 steps

Motion Measurements

3-D Motion Analysis System:– 3 Vicon Cameras (Oxford, England)– Micro-Vax Workstation (DEC, Maynard, MA)

Marker position data were transformed to rotational angles in FLX/EXT, LBs, and ARs.

Testing Constructs

Intact Spine Anterior fixation with an interbody graft

following total discectomy and endplate excision of L3-L4

– PMMA block was used as a bone graft

Anterior fixation only

Tested Anterior Fixators

Plate Type:– University Anterior Plating System (UNIV): AcroMed Corp.– Z-Plate (ZP): Danek, Inc.

Rod Type:– Kaneda device with transfixator (KAN): AcroMed, Corp– TSRH vertebral body screw construct: Danek, Inc.

Data AnalysesStabilizing Effect:

– % change of motion as compared with the intact motion

Role of Graft:– % change of motion between with and without graft cases

ANOVA with Tukey’s multiple mean comparison

Results

In Axial Rotation:– Similar to the Intact Motion in UNIV, ZP, TSRH Fixation– Significantly stabilized by KAN (p < 0.05)

In Lateral Bending and Flexion:– Singnificantly stabilized by all tested devices (p < 0.05)– No significant difference between devices

Stabilizing Effect compared with Intact Motion (Anterior Fixation with Graft)

Stabilizing Effect compared with Intact Motion (Anterior Fixation with Graft)

*

In Extension:– Singnificantly stabilized by all tested devices (p < 0.05)– Stabilizing effect of KAN > ZP (p = 0.05), but no

difference between any other devices

Stabilizing Effect compared with Intact Motion (Anterior Fixation with Graft)

Stabilizing Effect compared with Intact Motion (Anterior Fixation with Graft)

*

In Axial Rotation:– Similar to the intact motion in UNIV, ZP, TSRH Fixation– Significantly stabilized by KAN (p < 0.05)

In Lateral Bending:– Singnificantly stabilized by all tested devices (p < 0.05)– No difference between devices (p > 0.05)

Stabilizing Effect compared with Intact Motion (Anterior Fixation only)

In Flexion:– Significantly stabilized by UNIV, KAN, TSRH(p < 0.05), but

not in ZP fixation– Stabilizing Effect of ZP < KAN (p = 0.01) and TSRH (p =

0.05)

In Extension:– Singnificantly stabilized by KAN (p < 0.05)– Stabilizing Effect of KAN > TSRH (p = 0.005)

Stabilizing Effect compared with Intact Motion (Anterior Fixation only)

Stabilizing Role of Interbody Graft

In Axial Rotation:– No significant stabilizing effect

In Lateral Bending:– Significantly increase the stabilizing effect in all tested cases

In Flexion and Extension:– Significantly increase the stabilizing effect in ZP fixation only

Stabilizing Role of Interbody Graft

DISCUSSION

Biomechanical Spinal Implant Testing Protocols

Measuring Construct Stiffness in Axial Load, Torsion, Flexion, and Extension in nondestructive manner

– Ashmann et al., 1989

Measuring flexibility in terms of uncostrained motion under pure moment application

– Panjabi, 1988

Previous Studies

Abumi et al. 1989:– Kneda device provides good stability in FLX and EXT, but

not in AR.

Gaines et al., 1991:– Kaneda device provided good stabilization and resistance to

torsional and lateral bending loads.

Zdeblick et al., 1993:– Kaneda and TSRH devices are effective in restoring acute

stability to the lumbar spine after corpectomy.

In this study

Measure the unconstrained motion under pure moments application.

Findings:– Kaneda device provides effective stabilization in all directions,

particularly with the use of interbody graft.– UNIV, ZP, TSRH are also effective in stabilizing LB, FLX,

and EXT with an interbody graft, although they restore to the intact motion in AR.

CONCLUSIONModern plate and rod type anterior

fixation devices are effective in restoring stability of anterior and middle column defects in flexion, extension, and lateral bending beyond the intact specimen.

CONCLUSIONInterbody grafting may be important in

anterior instrumentation for effective stabilization immediately after surgery.

Anterior fixation restores torsional stability relatively less, although Kaneda device was the best among the tested devices.

Percentage AR Motion Changes compared to the Intact Case

Percentage LB Motion Changes compared to the Intact Case

Percentage FLX Motion Changes compared to the Intact Case

Percentage EXT Motion Changes compared to the Intact Case

Effects of Crosslinking Devices in Pedicle Screw Instrumentation:

A Biomechanical and Finite Element Modeling Study

Tae-Hong Lim, Ph.D.

Howard S. An, M.D.

Jason C. Eck, B.S.

Jae Y. Ahn, M.D.

PURPOSETo evaluate the stabilizing effect of

transfixation in flexion, extension, lateral bending, and axial rotation modes

To determine the optimal position of transfixation to achieve greater stabilizing effect

Flexibility tests Unstable Calf Spine Model

Finite element studies

FLEXIBILITY TESTS10 Ligamentous calf spines

- 5 spines for L3-L4 stabilization - 5 spines for L2-L4 stabilization

Maximum pure moment of 6.4 Nm in FLX, EXT, LB, and AR

Vicon 3-D motion analysis system was used to measure the resultant segmental motions.

Tested Constructs

One Segment (L3-L4) Stabilization- Intact

- Instrumentation without Transfixation after total discectomy

- Instrumentation with Single Transfixation

Tested ConstructsTwo Segment (L2-L4) Stabilization

- Intact

- Instrumentation without Transfixation after L3 corpectomy

- Instrumentation with Single Transfixation at the Middle Points

- Instrumentation with Double Transfixation at the proximal and distal 1/3 points

Finite Element Studies To determine the optimal position of

transfixation

Boundary and Loading Conditions:- Nodes in lower vertebra were held fixed.- Axial compression (445 N), FLX, EXT, LB, and AR Moments (5

Nm) at the middle point of the vertebra element

MENTAT II Finite Element Analysis Package (Marc Analysis Research Corp.)

Finite Element Models

One Segment Stabilization Two Segment Stabilization

ISOLA System

Vertebrae

Transfixators

Fixed Nodes

Axial Compression forceand Moment

RESULTS

Rotational Motions (deg) responding to Applied Moments of 6.4 Nm in One Segment

Stabilization

p<.01

Rotational Motions (deg) responding to Applied Moments of 6.4 Nm in Two Segment

Stabilization

p = .01

p = .06

Single Transfixation(FE Model Predictions)

Compared with no TF case:No additional stabilizing effect in axial

compression, FLX, and EXT In one segment stabilization:

- Up to 10.8% LB motion decrease

- Up to 25.0% AR motion decrease

In two segment stabilization:- Up to 13.0% LB motion decrease

- Up to 18.2% AR Motion decrease

Stabilizing Effect of 1 TF in AR Mode with respect to the implanting Positions

2 Segment Stabilization

1 segment stabilization

Double Transfixation(FE Model Predictions)

Compared with no TF case:

No additional stabilizing effect in axial compression, FLX, and EXT

Up to 14.6% LB motion decrease

Up to 30.3% AR motion decrease

Stabilizing Effect of 2 TF in AR Mode with respect to the Implanting Positions

CONCLUSIONFlexibility Tests:

Significant stabilizing effect of TF in AR only as shown in previous studies

2 TFs can provide more AR stability than one TF.

With 2 TFs, the restored AR stability is similar to the intact specimen.

CONCLUSION

Finite Element Model Predictions:

Model prediction correlated well with the experimental results.

The greatest AR stability can be obtained by implanting two TFs, one at the proximal 1/8 points and the other at the midpoints of the longitudinal rods in two level stabilization.

A BIOMECHANICAL COMPARISON BETWEEN MODERN ANTERIOR

VERSUS POSTERIOR PLATE FIXATION OF UNSTABLE CERVICAL

SPINE INJURIESTae-Hong Lim, Ph.D.Howard S. An, M.D.Young Do Koh, M.D.

Linda M. McGrady, B.S.*

Unstable Cervical Spine Injuries

Flexion-distraction Injury– Three column injury– Unilateral or bilateral facet dislocation

Burst Fracture of the Vertebral Body– produce severe instability by destroying the anterior and middle column

of the cervical spine

Surgical TreatmentsClosed or open reduction and fusion with

Anterior Fixation– anterior plate

Posterior Fixation– Wiring– Posterior lateral mass screw system

Combined Fixation

Surgical Treatment Goals Reduction

Maintenance of alignment

Early rehabilitation

Enhancement of fusion

Decreased use of external orthoses

Biomechanical stability provided by various fixation methods is an important factor to achieve these goals.

Purpose of the StudyTo determine and compare the

biomechanical stability provided by modern posterior, anterior, and combined screw-plate fixation in flexion-distraction and corpectomy models simulated in human cadaveric cervical spines.

METHODS

Flexibility Tests

10 cadaveric cervical spines (C2-T1)– Group I (n = 5): One-level 3-column injury at c4-5– Group II (n =5): Corpectomy of C5 vertebral body

Pure moment of 2.45 Nm was achieved in five steps using dead weights in FLX, EXT, LB, and AR directions.

Vicon 3-D motion analysis system was used to measure the resultant segmental motions.

Stability was quantified in terms of the percent changes in rotational motions of the stabilized segment compared with those in the intact segment.

Tested Constructs Group I (Flexion Distraction Model)

– Intact– Posterior fixation without interbody grafting after 3-column injury– Posterior fixation + an interbody graft (PMMA block)– Combined anterior and posterior fixation + PMMA block– Anterior fixation + PMMA block

3-Column injury was made by cutting supraspinous, interspinous, capsular and posterior longitudinal ligaments, ligamentum flavum, and the posterior half of the intervertebral disc.

Tested Constructs Group II (Corpectomy Model)

– Intact– Posterior fixation without interbody grafting after corpectomy– Posterior fixation + an interbody graft (PMMA block)– Combined anterior and posterior fixation + PMMA block– Anterior fixation + PMMA block

C5 vertebral body was removed (corpectomy) to simulate a very unstable burst fracture.

Axis plate and Orion plate systems (Danek Inc., Memphis, TN) were used for posterior and anterior fixation, respectively.

Axial Rotation(Flexion-Distraction Model)

The posterior fixation (cases 1, 2, and 3) decreased the motion significantly.

Anterior fixation alone provided the rigidity similar to the intact specimen.

1 2 3 4

-100

0

100

200

300%

Ch

ang

e o

f M

oti

onAxial Rotation

Instrumentation

(fracture-dislocation)

Lateral Bending(Flexion-Distraction Model)

Posterior fixation with or without PMMA block reduced motion as compared with the intact case, although the differences were not statistically significant.

When anterior fixation was used alone, the lateral bending became significantly greater than the intact motion (p < 0.05).

1 2 3 4

-100

0

100

200

300

% C

hang

e of

Mot

ion

Lateral Bending

Instrumentation

(fracture-dislocation)

Flexion(Flexion-Distraction Model)

The posterior fixation (cases 1, 2, and 3) significantly reduced the flexion motion from the intact case.

No significant difference was found between the posterior fixation constructs.

In case of anterior fixation (case 4), the flexion motion was significantly larger than the intact motion (128%, p < 0.01).

1 2 3 4

-100

0

100

200

300 %

Ch

ang

e o

f M

oti

on

Flexion

Instrumentation

(fracture-dislocation)

Extension(Flexion-Distraction Model)

Extension motion was significantly less than the intact motion in the combined anterior posterior fixation only (case 3).

There were no significant differences between the posterior constructs with and without PMMA block.

Extension motion was also not significantly different between the posterior and anterior fixation constructs.

1 2 3 4

-100

0

100

200

300%

Ch

ang

e o

f M

oti

on

Extension

Instrumentation

(fracture-dislocation)

Axial Rotation(Corpectomy Model)

Axial rotational motion was significantly reduced from the intact motion only when the combined anterior and posterior fixation was used (case 3).

Axial rotational motions in other constructs were not significantly different from the intact motion.

1 2 3 4

Instrumentation

-130

-110

-90

-70

-50

-30

-10

10

30

50

70

90

% M

otio

n C

hang

e

Axial Rotation(Corpectomy)

Lateral Bending(Corpectomy Model)

Lateral bending motions in posterior fixation cases 1, 2, and 3 were significantly less than the intact motion.

Anterior fixation alone (case 4) did not provide the rigidity beyond the intact specimen even with the anterior grafting material.

1 2 3 4

-130

-110

-90

-70

-50

-30

-10

10

30

50

70

90%

Ch

ang

e o

f M

oti

onLateral Bending

Instrumentation

(corpectomy)

Flexion(Corpectomy Model)

Flexion motion was significantly reduced from the intact motion in all tested constructs (p < 0.05).

The combined anterior and posterior fixation provided most rigid fixation than the other constructs.

1 2 3 4

-130

-110

-90

-70

-50

-30

-10

10

30

50

70

90 %

Ch

ang

e o

f M

oti

on

Flexion

Instrumentation

(corpectomy)

Extension(Corpectomy Model)

Extension motion was significantly less than the intact motion in all tested constructs except for the posterior fixation with PMMA block (case 2).

1 2 3 4

-130

-110

-90

-70

-50

-30

-10

10

30

50

70

90%

Ch

ang

e o

f M

oti

on

Extension

Instrumentation

(corpectomy)

DISCUSSION

Human cadaveric cervical spines were used to minimize the anatomical difference between the in vitro and in vivo cases.

OrionTM and AxisTM plate system were chosen for testing because they represent relatively rigid modern anterior posterior plates constructs in cervical spine fixation.

A PMMA block was used to simulate the interbody grafting technique while eliminating potential variation in inherent material properties of bone graft.

Limitations– old age specimens and Bone quality variations– Exclusion of supporting structures such as muscle, fascia, and ligamentum nuche

SUMMARY

Anterior fixation system is biomechanically inferior to the posterior lateral mass screw-plate fixation, particularly in flexion-distraction injury in which posterior ligamentous structures are disrupted.

The anterior fixation seems to be more suitable for anterior middle column injury where the posterior ligamentous elements are intact.

Postoperative use of external orthoses should be considered when the anterior plate is used alone for the treatment of unstable cervical spine injuries with disruption of posterior stabilizing elements.

SUMMARY Anterior fixation was particularly ineffective to prevent the

flexion motion in the flexion-distraction injury model. This seems to occur mostly due to no rigid connection between the plate and screws. Rigid connection at this junction may significantly improve the fixation.

Combined fixation seemed to improve the stability as compared with anterior or posterior fixation alone. However, the difference in stability was not always significant and the combined fixation requires additional incision. These suggest that a combined anterior and posterior fixation should be carefully indicated.

Findings of Finite Element Studies

Stress-shielding effects exist in the surgical construct due to the presence of spinal fixation devices and healed bone mass.

The fixation device may transmit 9 to 40% of the applied compression load depending upon the stabilized technique used.

Semi-rigid fixation may reduce the stress-shielding effect and incidence of hardware failure.

Finite Element Models

Intact L3-L4 (INT) Bilateral Fixation(STVSP & PVSP)

Unilateral Fixation(UVSP)

Important Factors in Finite Element Modeling

Adequate AssumptionsUse of Accurate Input Data

– Geometry; material properties; Element types

Proper Solution Procedures– Linear or nonlinear analysis; viscoelasticity; poroelasticity,

etc.

Complete Understanding of Results

EFFECT OF INSTRUMENTED FUSION ON THE BIOMECHANICS

OF ADJACENT SEGMENT: AN IN VIVO CANINE STUDY

Tae-Hong Lim, PhD, Avinash G. Patwardhan, PhD,* Jung Hwa Hong, MS, Howard S. An, MD,

Lee H. Riley III, MD, Scott Hodge, MD,* Jason Eck, BS, Michael M. Zindrick, MD*

Complications in the Segments Adjacent to Fusion

Degenerated Disc Stenosis Segmental Instability Osteoarthritis

Previous StudiesEx-vivo and In-vivo Studies of Post-fusion

Mechanics:– Increased motion and loads at the adjacent segment.

In-vivo Animal Studies:– Hypoactive metabolism in the adjacent discs;– Significant biochemical changes indicating a degeneration

process.– Changes in the biomechanical response of the adjacent segment

resulting from theses alterations in the disc were not investigated.

Previous Studies

It is believed that a concentration of load can cause degeneration at the adjacent segment.

There are little data on the long-term changes in the biomechanical properties of the adjacent segment.

OBJECTIVEQuantify the long-term changes in the

flexibility and viscoelastic properties of the intervertebral disc at the adjacent segment due to the instrumented lumbar spinal fusion in a canine model.

In-vivo Canine Model

Flexibility Tests

Relaxation and Cyclic Loading Tests

In-vivo Canine Model Development

Control Group:– 5 adult mongrel dogs (age: 2 yr and weight: 25-30 kg) – Euthanized at the beginning of the study

Experimental Group:– 5 adult mongrel dogs (age: 1.5 yr and weight: 25-30 kg) – Posterior instrumented fusion surgery across L5-S1 levels using

ISOLA system (AcroMed, Cleveland, OH)– Follow-up Period: 30 weeks

Flexibility Tests Ligamentous lumbar spines (L2-S1)

Maximum pure moment of 2.0 Nm applied in FLX, EXT, LB, and AR

Vicon 3-D motion analysis system was used to measure the resultant segmental motions at L4-5, L5-L7, and L7-S1 levels.

Flexibility Tests Control Group

– Control-Intact– Control-Implant

Isola instrumentation following a partial dissection

of the L5-6, L6-7, and L7-S1 facets

Experimental Group– Fusion + Instrumentation– Fusion Alone

After removing Isola instrumentation

Viscoelastic Property MeasurementMTS Load

Cell

L4-5 V B - Disc - VB

MTS Ram

Relaxation Test–A constant strain of 8% were

applied using a ramp function

in 1 minute.–Relaxation Period: 1 hour

Viscoelastic Property Measurement Cyclic Deformation Tests in Axial Compression

Test Parameters:Test Step Control

TypePre-Strain

MaximumStrain

LoadingFrequency

# ofCycles

1 StrainControl

8% 12%0.1 Hz

502 1.0 Hz

Str

ain

(%

)

Time

Measured Parameters

Disc Height (mm)– using AP and Lateral radiographs before relaxation and cyclic deformation

tests

Gross Anatomic or Degenerative Changes in the L4-5 Discs– dissection of L4-5 VB-Disc-VB unit after mechancial tests

Disc Cross-sectional Area (mm2)– using an image processor

3-D angular displacements of L4 relative to L5, L5 relative to L7, and L7 relative to S1 in response to a 2.0 Nm

Measured Parameters E1 (equilibrium modulus); E2 (instantaneous modulus);

and (damping coefficient) using a three-parameter standard linear solid (SLS) model

Relaxation Time Constant ( = /(E1 +E2)) Dynamic Stiffness (MN/m): peak-to-peak load/peak-to-

peak deformation– from the last three load-unload cycles

Hysteresis– from the last three load-unload cycles

RESULTS

General ObservationsNo complications were observed in the

experimental dogs during the follow-up period.

At sacrifice, the loosening of the rod-screw connection was observed at the sacrum level in four out of five experimental dogs.

Flexibility Testing Results Significant acute stabilization in all loading modes in

the control-implant group. FLX/EXT motion of the experimental groups was

statistically smaller than the intact control group, but it was nearly 12 degrees of motion.

No significant difference in LB and AR motion between the experimental groups and the intact control group.

No solid fusion was achieved across the L7-sacrum level.

L7-S1 Motions

Control-Intact

Control-Implant

Fusion alone

Fusion+Implant

FLX EXT LB AR

0

7

14

21

28

Ro

tati

on

An

gle

s (d

eg)

Flexibility Testing Results

Significant acute stabilization in all loading modes in the control-implant group.

As compared to the intact control group, motions of the experimental groups were significantly reduced in all loading modes.

Solid fusion across the L5-L7 segments was achieved at the end of 30 weeks of follow-up.

L5-L7 Motions

Control-Intact

Control-Implant

Fusion alone

Fusion+Implant

FLX EXT LB AR

0

4

8

12

16

Ro

tati

on

An

gle

s (d

eg)

Flexibility Testing Results

No significant L4-L5 motion changes were found among the tested groups in all loading modes.

The flexibility of the L4-5 segment (adjacent to instrumentation and fusion) was affected either immediately following instrumentation across L5-sacrum or after the 30 weeks follow-up.

L4-L5 Motions(Adjacent to Fusion)

FLX EXT LB AR

0

3

6

9

12 Control-Intact

Control-Implant

Fusion alone

Fusion+Implant

Ro

tati

on

An

gle

s (d

eg)

SLS Model Parameters(mean ± SD)

No significant differences in SLS model parameters between the control and experimental groups

Parameters Control Group(n=5)

Experimental Group(n=5)

E1 (MPa)

E2 (MPa)

(MPa-min)

/(E1+E2)

1.69 ± 0.38

3.40 ± 1.59

47.5 ± 22.3

9.14 ± 2.55

1.46 ± 0.82

3.21 ± 1.99

39.7 ± 22.1

8.99 ± 1.25

Relaxation of the L4-5 Disc

Cyclic Load-Displacement Testing Results(mean ± SD)

No significant changes in measured quantities between the control and experimental groups.

MeasuredQuantity Control Experimental Control Experimental

1.46 ± 0.89 0.98 ± 0.54 1.65 ± 1.33 1.46 ± 0.54

13.4 ± 4.22 18.0 ± 8.75 14.4 ± 5.29 14.3 ± 5.34

1 Hz

DynamicStiffness(MN/mm)

Hysteresis (%)

0.1 Hz

Steady State Load-Unload Curves

Observations of L4-5 Discs No morphological changes as a result of

instrumentation and fusion across L5-sacrum. Similar Disc Heights:

– 3.21 (SD 0.4) mm for the control group– 3.17 (SD 0.5) mm for the experimental group

Similar cross-sectional areas– 303.5 (SD 53.2) mm2 for the control group– 335.1 (SD 24.3) mm2 for the experimental group

Observations of L4-5 Discs

No visual signs of disc degeneration were observed in both the control and experimental L4-5 intervertebral discs.

An in-vivo canine model:– It has been frequently used for the studies of spinal instrumentation.

In the dog spine, the L7-sacrum segment is most mobile in FLX-EXT as compared to the other lumbar segments. This may have contributed to the rod loosening at the sacrum level and the subsequent development of nonunion at the lumbosacral junction.

Discussion

Results of this study indicate that the solid fusion across the L5-L7 segments did not alter the biomechanical properties of the adjacent segment at 30 weeks postoperatively in the canine spine.

Fusion may induce changes in the biochemical and nutritional environments, that may indicate the degeneration process. However, no biochemical analysis was performed in this study.

Discussion

What causes degeneration at the adjacent segment?

Increased motion and loads at the the adjacent segment due to:

– Solid fusion at the lumbosacral junction (Elimination of the most mobile segment) vs. floating fusion

– Loss of lordotic curves due to instrumentation

Metabolic changes due to the presence of screws in the spine

Genetic factors? Other factors?

CONCLUSION Solid fusion across the L5-L7 segments but not solid

fusion across the lumbosacral junction could be achieved in the dog model even using a rigid Isola pedicle screw instrumentation across the L5-sacrum levels.

The flexibility of the L4-5 segment was not changed due to the instrumentation across L5-sacrum in the control as well as the experimental dogs.

Fusion of the L5-L7 segments did not significantly affect the viscoelastic properties of the adjacent disc at the end of 30 weeks.

Current Findings of Spinal Instrumentation

Rigid Spinal Instrumentation can – enhance the solid fusion rate– strong enough to allow early mobilization without serious hardware problems

Construct stability (Anterior vs. Posterior Fixation)– Posterior fixation is superior to anterior fixation in general.– Both fixation systems can not provide the axial rotational stability beyond the intact

AR stability in most cases.

Need to determine the optimum stability of the surgical construct

– Too much rigid fixation may cause various complications particularly at the adjacent level.

Current Concepts Minimal Invasive Surgery

– Laprascopic surgery techniques

Solid Fusion with minimum use of spinal implants

– BAK screw system, – Metal cages– Artificial bone graft– use of BMP

Ongoing Debates

Use of rigid or semirigid fixationFixation in more or less lordosisCombination of posterior and anterior

fusionOthers

Design Factors for Improvement

User friendliness– fixing all components posteriorly

Rigid Fixation between Components– rigid connection between screw and rod (or plate), particularly

important in anterior cervical fixation

Adjustable Connection between Components– Poly-axial screw insertion– Allow adjustment in medial-lateral direction as well as in vertical

direction

Devices needs to be developed

Anterior graft device (Cage, Ceramic, etc.)Biological enhancement such as BMPCervical spine fixation deviceInstrument for larprascopic surgeryComputer-aided surgery techniquesScoliosis reduction systemArtificial intervertebral joint

EFFECT OF INTERVERTEBRAL JOINT STIFFNESS CHANGES ON THE LOAD SHARING CHARACTERISTICS IN THE

STABILIZED LUMBAR SEGMENT

Tae-Hong Lim, Ph.D.

Department of Orthopaedic Surgery

Rush-Presbyerian-St. Luke’s Medical Center

Chicago, Illinois

Vijay K. Goel, Ph.D.

Department of Biomedical Engineering

The University of Iowa

Iowa City, Iowa

Load Sharing

Ability of the instrumented segment to resist a fraction of the externally applied load

Clinical Relevance of Load Sharing:– Implant survival– Stress shielding of the instrumented segment– Fusion/healing rate and quality

Load sharing characteristics varies as a function of:– Spinal column stiffness– Implant stiffness

LOAD SHARING MECHANISM

F

Fd FP

Spinal Segment

Spinal

Instrumentation

Plate Stiffness FPFd

When stiffnesses of the spinal column andthe screws are not changed;

Previous Studies

20% of the axial load through the VSP plate in case of PLIF:

– Goel et al., 1988

10%, 20% and 23% of the axial load through the 4.76mm rods, 6.35mm rods, and VSP plates

– Duffield et al., 1993

Spinal Column Stiffness Changes

In STVSP, Edisc of the elements in dashed area was changed;

– Edisc= 0.0 MPa (Total Nucleotomy)

– Edisc= 4.2 MPa

– Edisc= 8.4 MPa

– Edisc= 1,000 MPa (PLIF)

– Edisc= 2,000 MPa (PLIF)

– Edisc= 3,000 MPa (PLIF)

Stabilization with stainless steel VSP system was maintained.

Purpose of This Study

To investigate the effects of variations in the intervertebral joint stiffness as well

as the implant stiffness on the load sharing characteristics across the

stabilized motion segment

METHODS

Finite Element Analysis

Finite Element Models

Intact L3-L4 (INT) Bilateral Fixation(STVSP & PVSP)

Unilateral Fixation(UVSP)

Implant Stiffness Changes

While keeping a degenerated intact disc:

STVSP: – L3-L4 motion segment stabilized by VSP system bilaterally

PVSP: – L3-L4 motion segment stabilized by non-metal plates and metal screws bilaterally

Unilateral: – L3-L4 motion segment stabilized by VSP system unilaterally

Boundary & Loading Conditions

Boundary Conditions:– All nodes in the midsagittal plane were not allowed to move in lateral

direction because of assumed midsagittal symmetry, while these constraints were not imposed on UVSP.

– Nodes in the inferior most surface of the VB were constrained not to move in any direction.

Loading Conditions:– Axial compressive loads of 200, 413, and 700N– Simulated by a uniform distribution of loads on the superior most surface

of VB and superior facets of L3

Data Analysis

FE models were executed using ANSYS, and solutions were obtained in an iterative manner.

Output was processed to obtain:– Stresses in various components (axial component of facet contact force)– Axial forces transmitted through facets and VSP plate

– Fp = Plate x-are x sum of axial stresses in the middle of the plate

Axial Forces across the VSP Platesin STVSP Model

0

50

100

150

200

250

300

0 200 413 700Applied Axial Compression Load (N)

Axi

al F

orc

es (

N)

Axial Forces (N) across the VSP Plates in Case of Varying Spinal Column Stiffness

0

50

100

150

200

250

300

350

400

450

0 (SD) 4.2 8.4 1000(PLIF)

2000(PLIF)

3500(PLIF)

Axi

al F

orc

e (N

)

100%

43%38%

10.9% 10.1% 9.1%

Edisc Variations

020

4060

80100

120140

160180

Intact VSP PVSP UVSP

17 %24 %

Axial Forces (N) Across the VSP Plates in Case of Varying Implant Stiffness

8% on facets

38 %

Axi

al F

orc

e (N

)

INT

1.83 (1.98)

0.11 (0.19)

0.21 (0.29)

Cortical Bone

Cancellous Bone

Disc Annulus

Average von-Mises Stresses (MPa) in Spinal Segment

VSP

1.51 (2.04)

0.07 (0.15)

0.12 (0.18)

PVSP

1.91 (2.21)

0.10 (0.18)

0.16 (0.23)

UVSP

1.50 (2.03)

0.08 (0.18)

0.15 (0.25)

*Maximum stresses are listed in the parentheses.

Maximum von-Mises Stresses in the VSP System

05

101520253035404550

Superior Screw Inferior Screw Plate

VSP PVSP UVSP

S

tre

ss

(M

Pa

)

A parametric study was conducted to investigate the effect of variations in the intervertebral joint stiffness as well as the implat stiffness on load sharing characteristics across the stabilized motion segment using finite element method.

Limitations:– Rigid connection was assumed at bone-screw interface and metal-to-metal

interfaces.– Only a few limited range of variations in the spinal column and implant

stiffnesses was simulated without modeling the property changes over the healing process.

Implications of Model Predictions

Model Predictions for Spinal Column Stiffness Variations:

– Importance of preserving the IVD to keep the load on the implants low for reducing the incidence of hardware failure, particularly in case of severe discectomy

– Benefit of using an interbody graft from the perspective of load sharing

– Minimal adverse effect of a slight decrease in the graft stiffness on the load sharing when using posterior fixation

– High potential for subsidence in case of interbody fusion with a too stiff graft even with rigid fixation

Implications of Model Predictions Model Predictions for Implant Stiffness Variations:

– Greater load on the spinal column in case of using a less rigid fixation

– Significant stress reduction in the implant components and less stress-shielding effect with decreasing implant stiffness

– Further studies are required to address how to maintain the surgical construct with the use of less rigid fixation.