The unstable iliac fracture: a biomechanical evaluation of internal fixation

Injury Vol. 28, No. 7, pp. 469-475, 1997 0 1997 Elsevier Science Ltd. All rights reserved Printed in Great Britain 0020.1383/97 $17.00 + 0.00

ELSEVIER

PII: SOO20-1383(97)00112-5

The unstable iliac fracture: a biomechanical evaluation of internal fixation Peter T. Simonian,

M. L. Chip Routt Jr, Richard

M. Harrington

and Allan F. Tencer

Harborview Medical Center, Biomechanics Laboratory, Department of Orthopaedic Surgery, University of Washington, Seattle, WA, USA

Neither plating nor lag screw fixation of a displaced iliac wing fracture as part of an unstable pelvic ring disruption has been studied biomechanically. The purpose of this study was to compare the stability of various combinations of fixation, specifically contrasting lag screws placed between the tables of the ilium with plating in different locations across the fracture line. Various combinations of these fixation implants were evaluated for an unstable iliac fracture. A longitudinal iliac fracture was created in each of six hemi-pelzGc specimens prior to testing. Compressiz)e force, up to 500 N or to the magnitude necessay to displace the fracture 2 mm, was applied to the fracture line through the hemi-pelvis for each of the plate and lag screw combinations tested. There was no statistical difference between any of the implants or combinations tested. A single 3.5-mm reconstruction (cephalad) plate placed along the cephalad internal aspect of the iliac crest prozlided the least stability allowing 2 mm of displacement with a mean load of 80 N. The two combinations of fixation that required the greatest loads for 2 mm of displacement were a single 3.5-mm lag (cephalad) screw inserted into the iliac crest between the tables of the ilium coupled with either a 3.5-mm reconstruction (brim) plate placed along the internal aspect of the inferior iliac fossa at the pelvic brim (239 N) or a 4.5-mm lag (brim) screw between the inner and outer tables at the inferior aspect of the fracture just above the greater sciatic notch (225 N). 0 1997 Elsevier Science Ltd.

Injury, Vol. 28, No. 7,469-475, 1997

Introduction Unstable pelvic ring injuries can be present in a multitude of patterns. Attempts to classify these injuries have been helpful 1,2.The common pelvic ring fracture patterns have been studied both clinically and biomechanically>“. A fracture, first described by Malgaigne in 185912,can present with multiplanar or vertical instability courses longitudinally through the ilium where the bone is attenuated along with an anterior ring disruption through either the pubic rami or symphysis. Internal fixation of the displaced iliac fracture as a part of an unstable pelvic ring

disruption has not been studied. This is an important fracture to study because of the inherent osseous limitations for stabilization. The iliac fracture often exits through the greater sciatic notch where the superior gluteal artery is vulnerable to injury13-‘*. Because the fracture passes through the ilium itself, external fixation is not always a good option since the optimal position for the fixation pins are within the fracture line. When significant displacement and instability exist, accurate reduction along with rigid stabilization using internal fixation is rational. Both plates and lag screws offer advantages and liabilities when used for fixation. One way of differentiating between them would be if one type of fixation offered significantly greater mechanical stability. The stability offered by these different forms of fixation has not been described. Therefore, the purpose of this study was to compare the stability of plates and/or intramedullary lag screws in different locations and in different combinations along the fracture line for the displaced iliac fracture as a part of an unstable pelvic ring disruption.

Materials

and methods

Specimens Six fresh-frozen, non-embalmed cadaveric hemipelvic specimens, along with the proximal two-thirds of the intact femora were obtained. These were sectioned sagittally through the midline of the sacrum and the symphysis. The mean age of the donors was 77years (range 71-87years). The specimens originated from patients without skeletal disease. No signs of pelvic bone or soft tissue disease were found at the time of procurement. The sacroiliac, sacrotuberous, and sacrospinous ligaments as well as the hip capsular tissues were preserved. A longitudinal iliac fracture was created in each hemi-pelvis prior to testing. This was done by drilling partial thickness (through the outer table cortex) holes with a 1.5-mm drill approximately 3 mm apart. A thin osteotome was used to connect the

Injury:

470

International

Journal

of the Care of the Injured

drilled holes by breaking through only the outer table cortex. The fracture was then completed by manually loading the hemi-pelvis. This created fracture planes with bony interdigitation allowing for a more realistic anatomic reduction. The fracture line extended from the cephalad aspect of the ilium beginning just posterior to the gluteus medius bone pillar at the iliac crest along the attenuated bone longitudinally, posterior to the pillar, through the pelvic brim, and exiting through the upper aspect of the greater sciatic notch (Figure 1). Loading arrangement In preparation for mechanical testing, the sacrum was potted in a fixture to facilitate loading of the hemi-pelvis in a materials testing machine (Model 858 Bionix, MTS Systems Corporation, Minneapolis, MN) which allowed axial compression (Figure2). The femoral shaft of the specimen was then mounted in a fixture on the base of the materials testing system. This fixture consisted of a l-cm diameter ball bearing placed between the distal end of the femoral shaft and a plate mounted to the base of the materials testing machine (Figure 2). This allowed angular motion of the femur with minimal constraint but not displacement of the end of the femur. Cyclic compressive force was applied to the potted sacrum which was mounted to the actuator of the materials Fracture throughiliac crest

Vol. 28, No. 7,1997

MTS LOAD CELL (CYCLIC: AXIAL LOADING)

I

’POTTED SACRUM

I

-/

FEMUR ALLOWED TO ROTATE ON INTRAMEDUALIARY BALL BEARING

I

Figure2 The sacrum was potted in a ‘fixture to facilitate loading of the hemi-pelvis in a materials testing machine for compressive loading. The femoral shaft of the specimen was then mounted in a fixture on the base of the materials testing system. This fixture consisted of a ball bearing at the base of the femoral shaft placed in a hemi-spherical slot allowing motion with minimal constraint.

tester, into the hemi-pelvis, and across the fracture line. This created combined compression and bending across the fracture plane. The specimen was loaded to the point where either the fracture displaced 4 mm or to a maximum force of 500 N, then unloaded. Loading and unloading were repeated over four cycles to ensure reproducibility of the measurements. All specimens were displaced a minimum of 2 mm through all cycles, which always occurred prior to reaching a force of 500 N. The force was limited to 500 N or 4 mm of fracture displacement

to prevent

damage

to the

strain

gauges

and

bone since each specimen was used for a number of sequential tests. Experimental sequence The loading protocol was applied to each hemi-pelvis after placing implants in a randomized fashion, SO that no implant would always be tested first or last. The following constructs were tested:

(1) Iliac fracture fixed with a single 100 x 3.5 mm lag

Figure 1. The fracture line was not cornminuted and extended from the cephalad aspect of the ilium at the crest longitudinally exiting through the superior aspect of the greater sciatic notch.

screw (Synthes, Paoli, PA) inserted along the iliac crest at the aspect of the fracture between the inner and outer tables (Figure 3). (2) Iliac fracture fixed with two lag screws (Synthes, Paoli, PA) between the inner and outer tables: one located along the iliac crest at the cephalad aspect of the fracture (100 x 3.5 mm; as in no. l), and the other medullary lag screw directed from the anterior superior iliac spine to the posterior iliac spine along the pelvic brim (140 x 4.5 mm) just above the greater sciatic notch (Figure 4).

Simonian

et al.: The unstable

iliac fracture

(3) Iliac fracture fixed with a single 100 x 3.5 mm lag screw (Synthes, Paoli, PA) between the inner and outer tables in the cephalad aspect of the fracture (as in no. l), and a single 4-hole 3.5-mm reconstruction plate (Synthes, Paoli, PA) contoured to fit along the pelvic brim at the caudad aspect of the fracture along the inner cortical surface (Figure 5). (4) Iliac fracture fixed with two 4-hole 3.5-mm reconstruction plates (Synthes, Paoli, PA): one plate contoured to fit the iliac crest along the inner cortical surface at the cephalad aspect of the fracture, and the other plate contoured to fit the pelvic brim within the internal iliac fossa at the caudad aspect of the iliac fracture (as in no. 3) (FiglLre6). (5) Iliac fracture fixed with a single 4-hole 3.5-mm reconstruction plate (Synthes, Paoli, PA) alone contoured to fit the iliac crest along its inner cortical surface at the cephalad aspect of the fracture (as in no. 3) (Fip~rc 7). Iliac fracture fixed with a single 4-hole 3.5-mm (6) reconstruction plate (Synthes, Paoli, PA) contoured to fit the iliac crest along its inner cortical surface at the cephalad aspect of the fracture (as in no. 3), and a single pelvic brim 140 x 4.5 mm lag screw (Synthes, Paoli, PA) between the inner and outer tables at the inferior

471

mm

Figure 4. Iliac fracture fixed with two lag screws(Synthes,

Paoli, PA) between the inner and outer tables; one in the cephalad (100mm lengtW3.5mm) and one in the pelvic brim (140lengtW4.5mm) aspectsof the fracture.

3.5 mmcortical screw

aspect of the fracture just above the greater sciatic notch (as in no. 2) (Figure 8). A 3.5-mm intramedullary lag screw was used in the cephalad, iliac crest region of the fracture because it was the largest diameter that could consistently be placed between the inner and outer tables of the superior ilium. A 4.5-mm intramedullary lag screw passed with ease between the inner and outer tables at the brim region of the fracture without hip or sacroiliac joint or greater sciatic notch violation.

Figure 3. Iliac fracture stabilized with a single 100x 3.5 mm lag screw (Synthes, Paoli, PA) in the cephalad aspectof the fracture along the iliac crest between the inner and outer tablesof the iliac crest.

Motion measurement Fracture displacement (gap opening) during loading was measured with two liquid mercury strain gauges, each 25 mm long (Parks Medical Electronics, Aloha, OR). Each was placed transversely across the fracture on the outer table: one along the cephalad aspect and one lower along the pelvic brim region. Each strain gauge was calibrated prior to application. The strain gauge output with displacement is linear from approximately lo-50 per cent strain”. Calibration using a micrometer was performed to establish this interval of linearity for each strain gauge. Both strain gauges produced similar curves with slopes that were within 1 per cent of each other. The plots contained a non-linear region below 10 per cent strain. Therefore each gauge was preloaded to approximately 10 per cent strain when affixed to the specimen.

472

Injury: International Journal of the Care of the Injured Vol. 28, No. 7, 1997

Statistical analysis of data Outputs from the strain gauges and the material testing machine load cell were recorded on a personal computer with an analog-to-digital converter (ZSOZA, Data Translation Inc., Marlboro, MA). The data were analyzed by averaging displacements from each strain gauge for each of the four cycles. The forces to displace either strain gauge 2 mm were compared for each cycle. Means and standard deviations were calculated for each condition, and differences tested using a repeated measures analysis of variance (ANOVA) (Statview 4.02, Abacus Concepts Inc., Berkeley, CA) on a personal computer.

Results The mean forces to displace either the cephalad or caudad strain gauge 2 mm for all cycles are reported in Table 1 and Figure 9. None of the differences in motion between implants tested were significant at p < 0.05. The cephalad reconstruction plate alone allowed 2 mm of displacement with the least amount of load (mean load = 80.3 N). The cephalad intramedullary screw alone was the same as the superior cephalad reconstruction plate combined with a pelvic brim intramedullary screw; both requiring essentially the same load to produce 2 mm of fracture gap displacement, 134 N and 128 N, respectively. The

Figure 6. Iliac fracture fixed with two 4-hole 3.5-mm reconstruction plates (Synthes, Paoli, PA): one contoured along the inner cortical surface of the iliac crest at the cephalad aspect of the fracture, and one contoured along the pelvic brim region of the fracture. cephalad plate combined with a pelvic brim plate required a mean of 162 N. The two combinations of fixation that required the great&t load to produce 2 mm of displacement were the cephalad intramedullary screw in combination with either a pelvic brim reconstruction plate (239 N) or a pelvic brim intramedullary screw (225 N). Cephalad intramedullary screw fixation alone required 68 per cent greater load compared with the cephalad reconstruction plate alone prior to 2 mm of fracture gap displacement. When pelvic brim reconstruction plate or pelvic brim intramedullary screw fixation was combined with cephalad intramedullary screw fixation the load capacity increased by 78 and 68 per cent, respectively, compared with the cephalad intramedullary screw alone. The ,KJvalues from ANOVA statistical analysis comparing each specific form of fixation to one another are reported in Table 11;no ~7value was less than 0.05. The modes of fixation could not be compared with the disrupted pelvis without fixation because the model could not stand upright or withstand any load without fixation.

Figure 5. Iliac fracture fixed with a single 100 x 3.5 mm lag screw (Synthes, Paoli, PA) between the inner and outer tables in the cephalad aspect of the fracture, and a single 4-hole 3.5-mm reconstruction plate (Synthes, Paoli, PA) contoured along the inner cortical surface of the pelvic brim region of the fracture.

Discussion Iliac fractures which are isolated, stable, and/or closed do not mandate operative stabilization. Displaced and/or unstable iliac fractures occur longitudinally from the superior ilium exiting through the greater

Simonian

et al.: The unstable

473

iliac fracture

mean values for the loads supported by different constructs for the same fracture displacement. Statistical significance could not be established because of the spread in measured values for each construct. This spread most likely results from differences in bone density and therefore purchase strength and interfragmentary compression achieved in different specimens. These bone density variations tend to mask the structural differences in the constructs that we sought to measure, in terms of developing statistical significance. Internal fixation at the cephalad aspect of the iliac wing would require less dissection and screw place-

Figure 7. Iliac fracture fixed with a single 4-hole 3.5-mm

reconstruction plate (Synthes, Paoli, PA) contoured along the inner cortical surface of the iliac crest at the cephalad aspectof the fracture. sciatic notch as part of a multiplanar or vertically unstable injury. Such unstable injuries often have some form of anterior ring disruption’“. Comminuted iliac fractures may also occur in these injuries because of direct high energy loads to the subcutaneous iliac crest. Superior gluteal vascular injury due to the fracture extension or displacement into the greater sciatic notch may result in haemodynamic instability”~‘“. Accurate reduction and rigid stabilization of displaced and/or unstable iliac fractures provides comfort and allows patient mobilization from the recumbent position. Because of the fracture location and the attenuated cortex in the iliac crest, options for fixation are limited. Spica cast immobilization and traction are treatment alternatives, but they increase the risk of complications associated with immobilization. External fixator placement is usually not possible because of the location of the fracture. When placement is possible, external fixators may provide some form of control to the anterior ring disruptior?” but will likely provide minimal stability at the site of the unstable iliac fracture unless the fixation pins are placed across the iliac fracture itself. Similarly, some forms of complex anterior ring internal fixation alone may provide limited, indirect stability to the unstable iliac fracture component, as they do to posterior ring disruptions through the sacroiliac joint or sacrumxmlU. No implant or combination of implants were significantly different at reducing measured motion, although there were numerical differences in the

Figure 8. Iliac fracture fixed with a single 4-hole 3.5-mm reconstruction plate (Synthes, Paoli, PA) contoured along the inner cortical surface of the iliac crest at the cephalad aspect of the fracture, and a single 140 x 4.5 mm lag screw (Synthes, Paoli, PA) between the inner and outer tables at the pelvic brim region of the fracture.

Table I. The mean force (N) to displace either the cephalad or pelvic brim strain gauge 2 mm for all cycles with standard error and standard deviation. Fixation CP CP+BP CP+BS cs CS+BP CS+BS

Mean

(N)

80.3 162.2 127.5 133.7 239.3 225.2

CP-cephalad reconstruction tion plate; CS-cephalad medullary screw.

Std. Err.

Std. Dev.

55.2 64.8 58.2 55.1 61.0 65.9

135.2 142.6 134.9 158.8 161.5 149.5

plate; BP-pelvic medullary screw;

brim reconstrucBS-pelvic brim

474

Injury:

CP

CP+BS

CS

CP+BP

CS+BS

International

CS+BP

fixation

Figure 9. Mean maximal load (N) with standard deviation to displace either the cephalad or pelvic brim strain gauge 2 mm (CP-cephalad reconstruction plate; BP-pelvic brim reconstruction plate; CS-cephalad medullary screw; BSpelvic brim medullary screw).

ment should be safer than at the pelvic brim region of the fracture, but fracture fixation is limited secondary to the thin cortex and narrow intramedullary space available along the iliac crest. Plate fixation at the cephalad aspect of the fracture is another alternative, but in this study the cephalad reconstruction plate alone provided the least stability. Cephalad intramedullary screw fixation alone was stronger, allowing 68 per cent greater load compared to the cephalad reconstruction plate. Adding some form of pelvic brim fracture fixation to a cephalad lag screw almost doubled the strength. The two combinations of fixation that allowed maximal loading prior to displacement included a cephalad intramedullary screw combined with either a pelvic brim reconstruction plate or a pelvic brim intramedullary screw. In all casesthe motion occurred at the bone-implant interface; no implant failed structurally. An advantage of intramedullary lag screw fixation is the possibility for percutaneous placement. This would provide mechanical stability with minimal

Table II. The p values from ANOVA statistical analysis comparing each specific form of fixation to one another

p Value

Fixation CP/CP+BP CP/CP+BS CPKS cP/cs+BP cP/cs+Bs CP+BP/SP+BS cp+ BPKS CP+BP/CS+BP CP+BP/CS+BS CP+BS/CS CP+BS/CS+BP CP+BS/CS+BS CS/CS+BP CS/CS+BS CS+BP/CS+BS CP-cephalad tion plate;

0.344 0.584 0.536 0.072 0.099 0.687 0.740 0.372 0.465 0.943 0.199 0.260 0.224 0.291 0.869 reconstruction

CS-cephalad medullary screw.

plate;

BP-pelvic screw;

medullary

brim reconstrucBS-pelvic brim

Journal

of the Care of the Injured

Vol. 28, No. 7,1997

surgical exposure. However, reduction is not always possible using percutaneous techniques and there has been a report of bowel herniation through the fracture site2’ which likely would not be appreciated unless the fracture was opened. Therefore, open reduction is recommended if bowel interposition is suspected or closed fracture reduction is not possible. The importance of an anatomic reduction must be underscored when interpreting the results of this study. Without an anatomic reduction using boney interdigitations to provide added stability, the effectiveness of all of these implants would likely be diminished. In the clinical case, anatomic reduction may be compromised because of iliac comminution. An added factor in the complexity of treatment is the bone quality of the pelvis. Our model incorporates the use of elderly specimens with increased osteopenia which may validate the results in the most difficult clinical situation. This study is best interpreted on a comparative basis. In summary, fixation of an unstable iliac fracture can be accomplished using a variety of techniques. A medullary lag screw or reconstruction plate applied to stabilize the upper iliac crest portion of the fracture are similar. The stability of either upper implant was augmented by some form of internal fixation applied to stabilize the lower aspect of the fracture ai the same time. In our study, both the pelvic brim reconstruction plate and medullary lag screw directed from the anterior superior iliac spine to the posterior superior iliac spine performed well in this capacity.

Acknowledgements This work was supported by a grant from the American Fracture Association.

References Burgess A., Eastridge B. and Young J. Pelvic ring disruptions: effective classification system and treatment protocols. ] Trauma 1990; 30: 8484356. Tile M. and Pennal G. Pelvic disruptions: principles of management. CIin Orthop 1980; 151: 56-64. Lange R. and Hansen S. Pelvic ring disruptions with symphysis pubis diastasis: indications, technique, and limitations of anterior internal fixation. Cli?~ Orthop 1985; 201: 130-137. Matta J. and Saucedo T. Internal fixation of pelvic ring fractures. Clin Ovthop 1989; 242: 83-87. Mears D. and Rubash H. External and internal fixation of the pelvic ring. AAOS instructional Course Lectures 1984; 37: 144-158. Peltier L. Fractures of the pelvis. A report of eighty cases treated at university hospitals. Minn Med 1955; 38: 563-564. Simonian P., Routt M., Harrington R. and Tencer A. Biomechanical comparison of different forms of posterior internal fixation for the unstable transforaminal sacral fracture. Trans Orfhopaedic Research Society 1995; 41: 575.

Simonian

et al.: The unstable

475

iliac fracture

8 Simonian I’., Routt M., Harrington R. and Tencer A. Box plate fixation of the disrupted pelvic ring: biomechanical evaluation of a new technique. J Orthop Trauma 1994; 8: 483-489. 9 Simonian I’., Routt M., Harrington R. and Tencer A. Internal fixation of the unstable anterior pelvic ring: a plating biomechanical comparison of standard techniques and the retrograde medullary superior pubic ramus screw, J Orfhop Trauma 1994; 8: 476-482. 10 Simonian I’. T., Routt M. L. R., Harrington R. M., Mayo K. A. and Tencer A. F. Biomechanical simulation of the anteroposterior compression injury of the pelvis: an understanding of instability and fixation. C/iii Ortliop Rel. Res. 1994; 309: 245-256. 11 Tile M. Pelvic fractures: operative versus nonoperative treatment. Orthop Clin North America 1980; 11: 423-464. 12 Malgaigne J. Traite des fmctztres et des luxatiorzs. J. B. Lippincott Co., Paris, Philadelphia, 1859. 13 Ben-Menachem Y., Coldwell D., Young J. and Burgess A. Hemorrhage associated with pelvic fractures: causes, diagnosis, and emergent management. AJR Am 1 Raentgcrzol 1991; 157: 1005-1014. 14 Brown J., Greene F. and McMillin R. Vascular injuries associated with pelvic fractures. Anz Surg 1984; 50: 150-154. 15 Kahn B. Superior gluteal artery laceration: a complication of iliac bone graft surgery. C/in Orfhop 1979; 140: 204-207.

16 Rothenberger D., Fischer R. and Perry J. Jr. Major vascular injuries secondary to pelvic fractures: an unsolved clinical problem. Am J Surg 1978; 136: 660-662. 17 Smith K., Ben-Menachem Y., Duke J. J. and GL H. The superior gluteal: an artery at risk in blunt pelvic trauma. 1 Trauma 1976; 16: 273-279. 18 Sundaram M., Pate1 B. and Woverson M. Superior gluteal artery hemorrhage following pelvic fractures controlled by embolisation. Clirl Radio1 1981; 32: 187-190. 19 Brown T. D., Sigal L., Njus G. O., Njus N. M., Singerman R. J. and Brand R. A. Dynamic performance of the liquid metal strain gauge. 1 Biomeclz 1986; 19(2): 165-173. 20 Stocks G., Gabel G. and Nobel I’. Anterior and posterior internal fixation of vertical shear fractures of the pelvis. J Orfhop Res 1991; 9: 237-245. 21 Charnley G. and Dorrell J. Small bowel entrapment in an iliac wing fracture. Injury 1993; 24: 627-628.

Paper

accepted

25 June 1997.

Requests for reprints should be addressed to: Peter T. Simonian, University of Washington Medical Center, Department of Orthopaedic Surgery, Box 356500, Seattle, WA 98195, USA.