ORIGINAL ARTICLE

Biomechanical Evaluation of Peri- and Interprosthetic Fractures of the Femur

Wolfgang Lehmann, MD, Martin Rupprecht, MD, Nils Hellmers, MD, Kai Sellenschloh, Daniel Briem, MD, Klaus Puschel, MD, Michael Amling, MD, Michael Morlock, PhD, and Johannes Maria Rueger, MD

Background: Because of an increasing life expectancy of patients and the rising number of joint replacements, peri- and interprosthetic femoral fractures are a common occurrence in most trauma centers. This study was designed to answer two primary questions. First, whether the fracture risk increases with two intramedullary implants in one femur; and second, whether a compression plate osteosynthesis is sufcient for stabilizing an interprosthetic fracture. Methods: Twenty-four human cadaveric femurs were harvested and four groups were matched based on the basis of bone density using a peripher quantitative computer tomography (pQCT). All groups(I) hip prosthesis with a cemented femoral stem; (II) hip prosthesis and retrograde femoral nail; (III) hip prosthesis, retrograde femoral nail, and lateral compression plate; (IV) all three implants with an additional simulated interprosthetic fracturewere biomechanically tested in a four-point bending in lateralmedial direction. Results: The second group with two intramedullary implants exhibited 20% lower fracture strength in comparison with group 1 with proximal femoral stem only. The stabilization of an interprosthetic fracture with a lateral compression plate (group IV) resulted in a fracture strength similar to femur with prosthesis only. Conclusion: Two intramedullary implants reduce the fracture strength signicantly. If an interprosthetic fracture occurs, sufcient stability can be achieved by a lateral compression plate. Because two intramedullary implants in the femur may decrease the fracture strength, the treatment of supracondylar femoral fractures with a retrograde nail in cases with preexisting ipsilateral hip prosthesis should be reconsidered. Key Words: Interprosthetic fracture, Periprosthetic fracture, Femoral fracture, Locking plate. (J Trauma. 2010;68: 1459 1463)

urrently, the prevalence of periprosthetic femoral fractures has been estimated from 0.1% to 3.2% for primary total hip arthroplasties without cement and from 3% to 12%
Submitted for publication February 3, 2009. Accepted for publication July 23, 2009. Copyright 2010 by Lippincott Williams & Wilkins From the Departments of Trauma, Hand, and Reconstructive Surgery (W.L., M.R., D.B., M.A., J.M.R.) and Legal Medicine (K.P.); Center for Biomechanics and Skeletal Biology (W.L., M.R., M.A.), University Medical Center Hamburg-Eppendorf, Hamburg, Germany; and Biomechanics Section (N.H., K.S., M.M.), Hamburg University of Technology, Hamburg, Germany. First two authors contributed equally and therefore share the rst authorship. Address for reprints: Wolfgang Lehmann, MD, Department of Trauma, Hand, and Reconstructive Surgery, University Medical Center Hamburg-Eppendorf, Martinistrasse 52, Hamburg 20246, Germany; email: wlehmann@uke. uni-hamburg.de. DOI: 10.1097/TA.0b013e3181bb8d89

for revisions performed with cement.1 With the increasing life expectancy of most patients, the physical activity of the elderly,2,3 and the increasing number of patients undergoing total hip or knee arthroplasty, the prevalence of periprosthetic fractures is expected to rise.3,4 As a consequence, orthopedic surgeons increasingly face complications that are associated with arthroplasties. The most common periprosthetic fracture of the femur is located immediately distal to the stem tip (Vancouver type B). Fractures of the femoral shaft between ipsilateral hip and knee arthroplasties or between an intramedullary nail and an endoprosthesis appear infrequently but challenge every trauma surgeon. The incidence of interprosthetic fractures is unknown. Kenny et al.5 found four cases among a total of 320 limbs with ipsilateral hip and knee arthroplasties, making it an incidence of 1.25%. Fink et al.6 have seen 11 interprosthetic fractures in 5 years in their institutions. There are a number of reports on the biomechanical testing of the various surgical treatment options for periprosthetic fractures.712 Two procedures are generally considered in these studies: (i) a revision arthroplasty with a long stem, which crosses the fracture distally, especially in cases where the hip prosthesis is already loosened and (ii) plate osteosynthesis that uses a broad variety of plates from different manufacturers. In addition, retrograde nails represent a further utility to stabilize supracondylar femoral fractures. Particularly, elder patients may benet by such a therapy.13,14 Reports on a successful treatment of femoral fractures between two intramedullary implants are very limited in medical literature. Dave et al.15 successfully treated an interprosthetic femoral shaft fracture with a Mennen plate, which is a nonrigid internal xation device. Jensen et al.16 used the Partridge Cerclage system, but loosening of the nylon band and a less stability were common problems. In contrast to this, Kenny et al.5 reported a 100% failure (4 of 4) of osteosynthesis with a dynamic compression plate. To the best of our knowledge, there is neither a biomechanical investigation of the risk getting an interprosthetic femoral fracture nor a biomechanical study analyzing the stability of these fractures after plate osteosynthesis. In recent years, we used a retrograde intramedullary nail for treatment of patients with a supracondylar femoral fracture in the presence of an ipsilateral hip prosthesis. This procedure allows early mobilization with timely rehabilitation and, depending on the fracture type, faster functional recovery and rapid weight bearing. However, in some cases, interprosthetic fractures occurred after minimal trauma (4 in
1459

The Journal of TRAUMA Injury, Infection, and Critical Care Volume 68, Number 6, June 2010

Lehmann et al.

The Journal of TRAUMA Injury, Infection, and Critical Care Volume 68, Number 6, June 2010

6 years). Therefore, it is conceivable that two intramedullary implants may alter the stiffness in the longitudinal direction of the femur, which may then lead to decreased stability in the region between the implant tips. We sought to determine whether two intramedullary implants, i.e., so-called kissing implants, may alter the biomechanical stability of the femur and may therefore be associated with a higher fracture risk. Furthermore, it is unclear whether lateral plating provides sufcient biomechanical stability in cases of interprosthetic fractures. The results of this study have important clinical implications not only for the management of ipsilateral hip prosthesis but also for conditions in which hip prosthesis and long-stem knee prosthesis or retrograde nails are present on the same side.

MATERIALS AND METHODS Specimens

A matched-pair experimental design using human femurs was used. Twenty-four fresh femurs were frozen at 20C directly after dissection and defrosted before the biomechanical testing. The specimens were from 12 men and 12 women, all aged 65 years and older (average age, 79.4 years; SD, 11.7 years). An osteoporotic bone model was chosen because patients with osteoporosis represent the group with the highest incidence of periprosthetic fractures. The femurs were randomized on the basis of pQCT data. The specimens of the rst group were match paired with the femurs of the second one and femurs of the third with them of the fourth group.

pQCT Measurements
Matched femur pairs were selected based on bone density using a pQCT (XCT-2000; Stratec Medizintechnik, Pforzheim, Germany). The diaphysis and the supracondylar region of the femur were scanned (Fig. 1, A). The pQCT device was calibrated using a standard phantom and a cone phantom provided by the manufacturer. A 2-mm thick single tomographic slice with a pixel size of 0.59 mm 0.59 mm is taken from the transverse plane of the condylus and the middle of the diaphysis. Femur length was dened as the distance between the distal end of the condylus and tip of the greater trochanter. Image processing and calculation were performed using the manufacturers software package (version 2.01). The total cross-sectional area was dened as the area enclosed within the outer bone border. The total volumetric bone mineral density (vBMD) was dened as the quotient of total bone mineral content on the volume. Additional parameters were used to assess the cortical structure of the femoral shaft. The total femoral cross-sectional area was separated into three parts (i.e., cortex, subcortex, and marrow cavity) based on two thresholds: 710 mg/cm3 distinguishes the cortex from the subcortex and 100 mg/cm3 separates the subcortex from the marrow cavity. Mean value for the diaphysis was 785 mg/cm3 (SD, 38) and for the condylar region 186 mg/cm3 (SD, 8). Based on own preliminary examinations on healthy human femurs and because of the ndings from other groups for the BMD of the femur, the threshold for the specimens to be considered osteoporotic was dened lower than 1,000 mg/cm3.17
1460

Figure 1. pQCT measurements. (A) Cross-sections of the condylar region and the femoral diaphysis. The cortical thickness and the bone mineral density were analyzed by pQCT. Colors illustrate the corresponding bone mineral density (BMD). (B) The femurs were randomized on the basis of pQCT data in four groups. There was no statistical significance between the BMD and cortical thickness of all groups.

On the basis of these data, the femurs were divided into the different groups to obtain maximum comparability (Fig. 1, B). Group I: The rst group (n 6) consisted of femurs with a hip prosthesis (Fig. 2). Group II: Six femurs with an implanted hip prosthesis and a distal femur nail (length, 220 mm; diameter, 10 mm; T2 Femoral Nailing System; Stryker GmbH & Co. KG, Duisburg, Germany). The nail was always locked on with two locking screws on the proximal and one locking screw (diameter, 5.0 mm) on the distal side. The screw holes were drilled with a 4.0 mm diameter (Fig. 2). Group III: In the third group (n 6), the femur is additionally stabilized on its lateral side with a locked compression plate (width, 4.5 mm; Locking Compression Plate; Stryker, Howmedica Osteonics, Allendale, NJ) crossing the tips of both implants. The plate design allows choosing between locking and compression screw insertion separately for each screw. In this study, all screws were locked. At least three unicortical screws were placed on each side of the prosthesis of the nail tips. The tips of the screws were attened to allow a maximum depth of insertion onto the prosthesis and respectively onto the nail. In the space between the two femoral components, two bicortically anchored screws were inserted (Fig. 2).
2010 Lippincott Williams & Wilkins

The Journal of TRAUMA Injury, Infection, and Critical Care Volume 68, Number 6, June 2010

Peri- and Interprosthetic Fractures

middle two supports (length l2 in Fig. 3) with a constant bending moment according to Eq. 1 M F l1 l2 l 2 2 l1 2 l2 8 cm (1)

with: l1: distance between the outer supports; l2: distance between the inner supports; F: applied force; l: lever arm length. This setup omitted transverse force across the femur (Fig. 3). Before testing, bones were embedded in polyurethane (Ureol FC53, Gol&Pfaff, Karlskron, Germany) in steel pots. The lever arm was adjusted to 8 cm. Load was applied at a constant speed of 0.1 mm/s until fracture.

Statistics
Statistical analyses were performed using SPSS for windows. The probability of a type I error was set to 5%. Students t test was used for normally distributed independent variables. Data are presented as mean and SD.
Figure 2. Group arrangement. Group I: hip prosthesis with a cemented femoral stem; Group II: hip prosthesis retrograde femoral nail; Group III: hip prosthesis retrograde femoral nail lateral compression plate; Group IV: all three implants additional osteotomy simulating an interprosthetic fracture.

RESULTS
All femurs with hip prosthesis only (group I) exhibited a periprosthetic fracture around the tip of the stem. Mean fracture strength is 4,692 N (SD, 183 N), which is signicantly higher than in femurs instrumented with hip prosthesis and intramedullary nail (group II: F 3,875 N; SD, 229 N; p 0.01; Fig. 4, A and B). In group II, the fracture occurred between the two kissing implants in all cases examined. The group with lateral plating construct without an osteotomy (group III) exhibited a higher fracture strength than group II and showed the highest strength of all test congurations (group III: F 5,020 N; SD, 639 N; p 0.04). The difference to group I failed to reach statistical signicance (p 0.37). Group IV had a similar fracture strength as group I (F 4,591 N; SD, 690 N; p 0.79; Fig. 4, A and B). In femurs with lateral plating (groups III and IV), the fracture always occurred at the distal end of the plate at the level of the most inferior screw hole. As predicted, the femurs with osteotomy (group IV) needed a smaller force to failure (4,591 N; SD, 690 N; p 0.01), similar to those of group I (4,692 N; SD, 183 N; Fig. 4, A and B).

Group IV: The femurs in the fourth group (n 6) were additionally osteotomized in an oblique plane at the level of the tip of the prosthetic stem, starting medially and extending laterally at 45 degrees, to simulate an interprosthetic fracture (Fig. 2). All femurs were stripped of soft tissue and the medullary canals were prepared for prosthesis implantation. Each femur was implanted with an Exeter V40 stem (Stryker, Howmedica Osteonics) with Simplex P Bone Cement (Stryker, Howmedica Osteonics). The cement stopper was implanted to a depth that allowed inserting the nail from the distal side. This further minimizes the variation in the distance between cement and end of the retrograde nail. To ensure an adequate cement mantle, position, length, and orientation, radiographs are taken after specimen preparation and implantation of the prosthesis. Noncemented stems were not tested because it is impossible to simulate osseous ingrowth in a laboratory setting.8

Four-Point Bending Test

In preliminary studies, we tested different loading directions. If the force was directed from anterior to posterior, the fractures occurred in the supracondylar region. Pure torsional loading alone was insufcient to bring about a fracture between the implants. The hip prosthesis cracked the cement and caused a pertrochanteric fracture around the proximal part of the prosthesis. In lateral-medial direction, we were able to create fractures between the two intramedullary implants. Therefore, we used this loading condition in all specimens. To determine the fracture strength in the boneimplant constructs, a four-point bending test was conducted. A servohydraulic testing machine was used (MTS 858.2, Eden Prairie, MN). The femur was loaded between the
2010 Lippincott Williams & Wilkins

Figure 3. Biomechanical testing. A servohydraulic testing machine was used (MTS 858.2, Eden Prairie, MN). The femur was loaded between the middle two supports (length l2) with a constant bending moment according to. 1461

Lehmann et al.

The Journal of TRAUMA Injury, Infection, and Critical Care Volume 68, Number 6, June 2010

Figure 4. Fracture strength. (A) Average load to failure in groups I to IV. The femurs with a lateral plating construct (group III) exhibited the highest fracture strength. The maximum load to failure of group II (femurs with two intramedullary implants) was significantly reduced in comparison with femurs with prosthesis only (group I). (B) Presentable forcedisplacement diagrams for matched pairs for the four groups. Note the steepened incline with two intramedullary components (group II).

DISCUSSION
As the incidence of prosthetic joint replacement is continually rising, surgeons will be more frequently confronted with peri- and interprosthetic fractures of the femur. An aggravating factor is that those patients often suffer from reduced bone quality and severe osteoporosis. However, there is only limited information on the biomechanics of periprosthetic fractures and until now, there has been no biomechanical investigation of interprosthetic femoral fractures or the treatment of those fractures. Furthermore, to date, it is unknown whether fracture strength is altered by the implantation of two intramedullary carriers into a single femur. To address these questions, we performed a series of experiments where we analyzed the fracture strength and biomechanical stability in four different paradigms. To enhance the clinical signicance, we used postmortem femurs,
1462

which had considerable osteoporosis, as determined by reduced cortical thickness and bone density. Our data indicate that two intramedullary implants decrease the fracture strength signicantly. Furthermore, in preliminary studies, we observed that the loading force, which causes peri- and interprosthetic fractures, is the load application from the lateral side. Simple load application in anteroposterior direction led to supracondylar fractures, regardless of the presence of an implanted retrograde nail in the distal femur. Because the dynamic force from the lateral side is considered to be the main trauma direction, we conducted the further experiments in a four-point bending test in this direction. We used a four-point bending with a constant moment that acts between the middle of the two supports allowing failure at the less resistant region. Because the load is split between the two middle supports, a possible bias by local material tiredness at the contact point is minimized. Previously, various stabilization methods have been tested for the xation of periprosthetic fractures around the tip of a single implanted stem for hip replacement surgery. These studies simulated a periprosthetic fracture by osteotomy similar to a Vancouver C fracture and compared the stability by a variety of parameters including torsional testing.7,8,11 We did not include torsional testing in our study because of technical restrictions. In preliminary experiments, we observed that torsional testing without an osteotomy burst the cement mantle, resulting in a pertrochanteric fracture. Therefore, we chose the four-point bending that produces a constant moment between the two middle support points. Moreover, torsion and shear are difcult to apply in an ex vivo model because the load application often causes failure at undesired sites of the bone. We further evaluated plates in tension banding because plates represent currently the gold standard for fracture treatment of osteoporotic bones (i.e., on the lateral cortex). From a clinical perspective, our results suggest that patients with two ipsilateral intramedullary femoral implants may benet by lateral femoral plating, to stress shield the area between the tips of the two implants. Although such a procedure is not totally consistent with the concept to develop treatment strategies that minimize invasive surgery, it should be seriously considered to protect patients from a higher fracture risk in the endangered area between the implant tips. After open reduction of the mostly comminuted fractures, the screws on the proximal part of the plate can be applicated throughout a targeter by minimal invasive approaches. If additional plating cannot be considered as a treatment option, an extramedullary stabilization or an in situ lengthening of the indwelling prosthesis may be used to avoid the problem of two intramedullary components. The locking condylar plates can be usually applied through a small incision along the lateral condyle and locked percutaneously at the proximal end. Meyer et al. used a lengthening of the indwelling prosthesis by a custom-made slotted hollow intramedullary nail for treating periprosthetic femoral fractures. In all cases, they observed fracture healing, and in only one case, a subsequent prosthesis loosening.18 A further important nding of our study demonstrates that the locking compression plate between the tip of the stems leads to a remarkable stability regardless of the presence of an inter 2010 Lippincott Williams & Wilkins

The Journal of TRAUMA Injury, Infection, and Critical Care Volume 68, Number 6, June 2010

Peri- and Interprosthetic Fractures

prosthetic fracture. Although methodological differences make it often difcult to compare the results of biomechanical studies, our ndings are very similar to the observations from other studies. Schmotzer et al.19 performed a biomechanical study that included six different xation techniques in seven cadaver femurs using cementless implants with a transverse osteotomy. They concluded that a well-xed implant, i.e., a plate with proximal unicortical screws, provided the best xation until pull out of the proximal screws occurred. Dennis et al.7 observed that plate constructs with proximal unicortical screws and distal bicortical screws or with proximal unicortical screws, proximal cables, and distal bicortical screws did yield the most benet in terms of biomechanical stability. Fulkerson et al.8 compared a locked plating construct with the Ogden construct and found that a locking plate is stiffer than the Ogden construct in axial loading and torsion. Light of these ndings, plate-screw techniques appear to provide the best stability in periprosthetic fractures. In these experiments, we simulated the most unstable situation in group IV by application of an additional osteotomy. The two implants only allowed monocortical screw xation of the lateral plate with three orat mostfour screws on either side proximal and distal to the respective implant tips. This construct resulted in a very high degree of stability comparable with those of the femur without osteotomy and retrograde nailing. No pull out of the screws was observed, which further implicates a strong connection between screw and plate as well as plate and bone. Therefore, our data are not only in line with previous biomechanical studies, suggesting that treatment with a locking plate is probably the best method to stabilize periprosthetic femoral fractures. The present data also demonstrate that this osteosynthesis is a sufcient procedure for stabilizing a fracture between two intramedullary implants. This study is limited by the use of nonliving tissue and failure to account for soft tissue contributions to stability. The loading scheme does not replicate adduction forces or tension provided by the iliotibial band. In clinical routine, comminuted fractures are most common. To facilitate the comparability of different studies, we used an established fracture model without a comminution defect. A technical limitation is the variability of the distance between the tips of the ipsilateral femoral component and the retrograde nail (mean, 46 mm; SD, 18 mm). To minimize this potential bias, we used nails with the same length and femurs of a comparable size. To take this issue into further account, we implemented the cement mantle always as deep as possible, so that the nail almost directly touched the cement. Furthermore, a study by Iesaka et al.20 suggests that a variable gap size may only have a minor inuence on stability and resistance of femoral stems in humans. In a series of experiments, they could demonstrate that the size of the gap between two ipsilateral femoral stems does not affect peak tensile stress on the femur. According to this study, the stress riser effect of an intramedullary metal stem inside a femur is clearly dependent on the femurs cortical density instead of gap distance. Taken together, the results from this study clearly indicate that two intramedullary implants in the femur are
2010 Lippincott Williams & Wilkins

associated with a decreased fracture strength between these implants. In clinical practice, an extramedullary xation technique for supracondylar femoral fractures should be preferred instead of an additional intramedullary implant. In cases of an oblique, noncomminuted interprosthetic femoral fracture, the lateral compression plate offers a sufcient treatment option and provides a similar stability in comparison with femurs with a single implanted hip prosthesis. REFERENCES
1. Springer BD, Berry DJ, Lewallen DG. Treatment of periprosthetic femoral fractures following total hip arthroplasty with femoral component revision. J Bone Joint Surg Am. 2003;85:2156 2162. 2. Beals RK, Tower SS. Periprosthetic fractures of the femur. An analysis of 93 fractures. Clin Orthop Relat Res. 1996;327:238 246. 3. Berry DJ. Epidemiology: hip and knee. Orthop Clin North Am. 1993; 30:183190. 4. Dixon T, Shaw M, Ebrahim S, Dieppe P. Trends in hip and knee joint replacement: socioeconomic inequalities and projections of need. Ann Rheum Dis. 2004;63:825 830. 5. Kenny P, Rice J, Quinlan W. Interprosthetic fracture of the femoral shaft. J Arthroplasty. 1998;13:361364. 6. Fink B, Fuerst M, Singer J. Periprosthetic fractures of the femur associated with hip arthroplasty. Arch Orthop Trauma Surg. 2005;125: 433 442. 7. Dennis MG, Simon JA, Kummer FJ, Koval KJ, Di Cesare PE. Fixation of periprosthetic femoral shaft fractures: a biomechanical comparison of two techniques. J Orthop Trauma. 2001;15:177180. 8. Fulkerson E, Koval K, Preston CF. Fixation of periprosthetic femoral shaft fractures associated with cemented femoral stems: a biomechanical comparison of locked plating and conventional cable plates. J Orthop Trauma. 2006;20:89 93. 9. Kuptniratsaikul S, Brohmwitak C, Techapongvarachai T, Itiravivong P. Plate-screw-wiring technique for the treatment of periprosthetic fracture around the hip: a biomechanical study. J Med Assoc Thai. 2001;84(suppl 1): 415 422. 10. Peters CL, Bachus KN, Davitt JS. Fixation of periprosthetic femur fractures: a biomechanical analysis comparing cortical strut allograft plates and conventional metal plates. Orthopedics. 2003;26:695 699. 11. Shawen SB, Belmont PJ Jr, Klemme WR. Osteoporosis and anterior femoral notching in periprosthetic supracondylar femoral fractures: a biomechanical analysis. J Bone Joint Surg Am. 2003;85:115121. 12. Wilson D, Frei H, Masri BA, Oxland TR, Duncan CP. A biomechanical study comparing cortical onlay allograft struts and plates in the treatment of periprosthetic femoral fractures. Clin Biomech. 2005;20:70 76. 13. El-Kawy S, Ansara S, Moftah A, Shalaby H, Varughese V. Retrograde femoral nailing in elderly patients with supracondylar fracture femur; is it the answer for a clinical problem? Int Orthop. 2007;31:83 86. 14. Papadokostakis G, Papakostidis C, Dimitriou R, Giannoudis PV. The role and efcacy of retrograding nailing for the treatment of diaphyseal and distal femoral fractures: a systematic review of the literature. Injury. 2005;36:813 822. 15. Dave DJ, Koka SR, James SE. Mennen plate xation for fracture of the femoral shaft with ipsilateral total hip and knee arthroplasties. J Arthroplasty. 1995;10:113115. 16. Jensen TT, Overgaard S, Mossing NB. Partridge Cerclene system for femoral fractures in osteoporotic bones with ipsilateral hemi/total arthroplasty. J Arthroplasty. 1990;5:123126. 17. Sievanen H, Koskue V, Rauhio A, Kannus P, Heinonen A, Vuori I. Peripheral quantitative computed tomography in human long bones: evaluation of in vitro and in vivo precision. J Bone Miner Res. 1998;13:871 882. 18. Meyer C, Alt V, Schroeder L, Heiss C, Schnettler R. Treatment of periprosthetic femoral fractures by effective lengthening of the prosthesis. Clin Orthop Relat Res. 2007;463:120 127. 19. Schmotzer H, Tchejeyan GH, Dall DM. Surgical management of intraand postoperative fractures of the femur about the tip of the stem in total hip arthroplasty. J Arthroplasty. 1996;11:709 717. 20. Iesaka K, Kummer FJ, Di Cesare PE. Stress risers between two ipsilateral intramedullary stems: a nite-element and biomechanical analysis. J Arthroplasty. 2005;20:386 391.

1463