Fixation Systems of Greater Trochanteric Osteotomies: Biomechanical and Clinical Outcomes

Gregg J. Jarit, MD Abstract Sathappan S. Sathappan, MD The development of cerclage systems for fixation of greater Anand Panchal, DO trochanteric osteotomies has progressed from monofilament wires Eric Strauss, MD to multifilament cables to cable grip and cable plate systems. Paul E. Di Cesare, MD Cerclage wires and cables have various clinical indications, including fixation for fractures and for trochanteric osteotomy in hip arthroplasty. To achieve stable fixation and eventual union of

Dr. Jarit is Chief Resident, Department the trochanteric osteotomy, the implant must counteract the of Orthopaedic Surgery, NYU–Hospital destabilizing forces associated with pull of the peritrochanteric for Joint Diseases, New York, NY. Dr. musculature. The material properties of cables and cable grip Sathappan is Joint Reconstructive systems are superior to those of monofilament wires; however, Fellow, Department of Orthopaedic Surgery, NYU–Hospital for Joint potential complications with the use of cables include debris Diseases. Dr. Panchal is Resident, generation and third-body polyethylene wear. Nevertheless, the Department of Orthopaedic Surgery, cable grip system provides the strongest fixation and results in Grandview Hospital Medical Center, Dayton, OH. Dr. Strauss is Resident, lower rates of nonunion and trochanteric migration. Cable plate Department of Orthopaedic Surgery, constructs show promise but require further clinical studies to NYU–Hospital for Joint Diseases. Dr. Di validate their efficacy and safety. Cesare is Chair, Department of Orthopaedic Surgery, University of California Davis Medical Center, steotomy of the greater tro- Trochanteric osteotomy fixation Sacramento, CA. Ochanter in primary total hip ar- systems include monofilament None of the following authors or a throplasty (THA) was strongly advo- wires, multifilament cables, and ca- member of their immediate families has cated by Charnley,1,2 who observed ble grip systems. The effectiveness received anything of value from or owns that it allows for proper hip abductor of a wire or cable system in counter- stock in a commercial company or muscle tensioning and improved hip acting the destabilizing forces that institution related directly or indirectly to stability. With the advent of modu- affect the reattached trochanter de- the subject of this article: Dr. Jarit, Dr. larity in hip arthroplasty implants, pends on the biomechanical proper- Sathappan, Dr. Panchal, Dr. Strauss, the ability to adjust offset and femo- ties of the system and its compo- and Dr. Di Cesare. ral length with the implant itself al- nents; these components include Reprint requests: Dr. Di Cesare, lowed for adjustment of soft-tissue fatigue resistance, maximum load University of California Davis Medical tension about the hip without the capability, and the compressive forc- Center, Suite 3800, 4860 Y Street, need for trochanteric osteotomy. In es generated by the system. Compli- Sacramento, CA 95817. complex primary3-5 and revision hip cations associated with trochanteric arthroplasty,4-8 however, the need for osteotomy include wire or cable J Am Acad Orthop Surg 2007;15:614- improved hip joint exposure is facil- breakage (Figure 1), debris generation 624 itated by trochanteric osteotomy.9 that can result in third-body poly- Copyright 2007 by the American Following the osteotomy, a stable, ethylene wear, trochanteric frag- Academy of Orthopaedic Surgeons. rigid fixation must be achieved to fa- ment migration, and nonunion. cilitate union of the bony fragment. Despite fixation, proximal migra-

614 Journal of the American Academy of Orthopaedic Surgeons Gregg J. Jarit, MD, et al tion of the osteotomized segment Figure 1 Figure 2 can occur from the contractile activ- ity of the hip abductors (gluteus me- dius and minimus muscles). Charn- ley1,2 showed that the abductors primarily exert shear forces when the hip is flexed and that their proximally directed vector component is a secondary one. These shear forces can be greater than four times body weight and are of paramount impor- tance in the activities of daily living, such as climbing stairs and rising from a chair.9,10

Wire Fixation Systems for reattaching the osteotomized trochanter are conventionally categorized as first, second, or third generation. First-generation monofil- Anteroposterior radiograph ament surgical wires have long been demonstrating broken trochanteric used for osseous fixation,11-13 includ- fixation wires. ing fixation of greater trochanteric osteotomies in THA. However, monofilament wires tend to kink (Figure 3). Loaded at a rate of 50 mm during application, compromising of outward strain per minute, the 16- The Charnley wiring technique. The ends of a looped wire are inserted in a the biomechanical integrity of the gauge wire failed at a mean of ap- hole drilled in the lateral femoral cortex, implant; thus, breakage and loss of proximately 1,300 N for the square distal to the edge of the trochanteric trochanteric osteotomy fixation were knot and knot twist versus approxi- osteotomy and below the abductor common, leading to the development mately half that value for the uni- tubercle. The two ends are passed of new cerclage systems.10,14-24 form symmetric twist method. Re- proximal to the trochanter and through Various wiring techniques have sults for the 18-gauge wire showed the hip abductor musculature and are been described that differ in their failure rates at a mean of approxi- finally locked by passing through the ability to resist displacement forces mately 950 N for the square knot eye of the loop. Two transverse wires following trochanteric osteotomy. and knot twist versus approximate- are placed in the anteroposterior Markolf et al24 examined displace- ly half that value for the uniform direction via drill holes, one in the ment as a function of abductor pull, symmetric twist knot11 (Table 1). proximal femur next to the trochanteric bed and the other in the lesser using a device to simulate the prox- These findings confirm that the ulti- trochanter. The two wires are crossed imal pull of the hip abductors on the mate strength of the wire construct during tightening, which results in greater trochanter and the cerclage was primarily dependent on its di- progressive reduction in slack with system. In their comparison of types ameter and on the type of knot used. each twist. of wiring techniques, they deter- When wire is to be used for fixation, mined that the Charnley25 (Figure 2) 16-gauge wire should be utilized, se- and Harris26 techniques resulted in cured with a square knot or knot teric nonunion. Boardman et al32 greater resistance to motion, exhib- twist. reported that 79.2% of patients iting a displacement of only 0.7 mm Analysis of studies conducted be- with trochanteric nonunion in their after the third loading cycle, where- tween 1978 and 1993 reveals an study had broken wires. Harris and as the Amstutz27 and Coventry28 overall wire breakage rate of 22% in Crothers26 evaluated 136 THAs in techniques had two to three times 2,910 THAs using various wiring which none of the trochanters mi- greater displacement. techniques21,23,26,30-33 (Table 2). Clin- grated and all united; the wire break- Bostrom et al11 examined the ically, wire breakage is most often age rate was only 2%. maximum load capabilities of 16- associated with revision surgery,21 Trochanteric migration as a result and 18-gauge wire using three differ- uniplanar osteotomies, simple wire of wire loosening or breakage is a sig- ent types of knotting techniques configurations, and cases of trochan- nificant complication that can lead

Volume 15, Number 10, October 2007 615 Fixation Systems of Greater Trochanteric Osteotomies

Figure 3 ing the same wiring technique with some modifications; the modifications included osteotomy of the trochanter below the trochanteric ridge, advancement of the trochanter, and abrading of the osteotomy surfaces before reattachment.27 The authors reported a 3.23% rate of trochanteric migration and found an association between this phenomenon and previous osteotomy, osteoporosis, and poor surgical technique. Clarke et al23 reviewed 277 THAs with trochanteric osteotomy fixation using multiple cerclage wires. Trochanteric migration was associ- Knotting techniques. A, Knot twist. B, Square knot. C, Symmetric twist. ated with “severe debility” of the spine or the contralateral hip; the authors attributed fixation failure to to loss of mechanical advantage of ment, poor apposition of bone sur- patient inability to observe “tro- the hip abductors. Amstutz and Ma- faces, too much contact with acryl- chanteric” precautions. Following ki31 reported a 4.9% trochanteric ic or cortical bone instead of fixation of the trochanter, the pa- migration rate in 728 consecutive cancellous bone, failure to tighten tient should avoid abduction exer- THAs using a cruciate two-wire the wire loops adequately). Rates of cises after surgery and adhere to technique. They attributed trochan- trochanteric migration decreased strict postoperative hip precaution teric migration to osteoporosis and with time as the surgeons modified protocols.37 No association with tro- to technical error (eg, small size of their technique. In a subsequent chanteric migration was found with the osteotomized trochanteric frag- study, 712 THAs were evaluated us- osteoporosis, previous surgery, age,

Table 1 Properties of Wires and Wiring Techniques

Fatigue Maximum Compressive Source Wire Gauge Wiring Technique Resistance (N) Strength (N) Force (N) Bostrom et al11 16 Twist 156 679 — Knot twist 356 1,259 — Square knot 356 1,357 — 18 Twist 222 480 — Knot twist 400 938 — Square knot 400 961 — Hersh et al17 16 Charnley — 1,380 ± 402 — Shaw and Daubert29 18 Square knot — 966 ± 7* 17 ± 2* Modiﬁed square knot — 790 ± 26 134 ± 9 Knot twist — 798 ± 13 115 ± 21 Twist knot — 540 ± 24 90 ± 11 Clinical twist knot — 494 ± 7 41 ± 9 20 Square knot — 682 ± 6 19 ± 2 Modiﬁed square knot — 543 ± 28 88 ± 10 Knot twist — 537 ± 13 46 ± 7 Twist knot — 344 ± 7 56 ± 4 Clinical twist knot — 344 ± 14 10 ± 3

* All values in Shaw and Daubert29 were reported in kilogram-force (kgf) and are here converted to N and rounded off.

616 Journal of the American Academy of Orthopaedic Surgeons Gregg J. Jarit, MD, et al

Table 2 Wire Breakage Rates In Vivo

Number of Total Hip Trochanteric Source Arthroplasties Surgical Technique Wire Breakage Rate Nonunion

Clarke et al23 277 Uniplanar osteotomy 33.2% 9% Berry and Müller30 53 primary, Chevron osteotomy 19% of primary 2% in primary and 74 revision hip with single-wire THA, 18% of 3% in revision hip arthroplasties technique revision THA arthroplasties Amstutz and Maki31 728 Cruciate 2-wire 77% 17% technique Harris and Crothers26 136 2 horizontal and 1 2% 2% vertical wire conﬁguration Boardman et al32 1,020 Cruciate: 1 double 9% 79.2% of vertical and 2 trochanteric crossed horizontal nonunion associated with wire breakage Wroblewski and Shelley34 226 Double crossover 4% None reported wire compression spring Jenson and Harris35 804 2 vertical and 1 28% Primary, 1% transverse nonunion; revision, 0% nonunion Bernard and Brooks22 59 revision hip 2 longitudinal and 1 9% 31% arthroplasties circumferential or transverse; 1 longitudinal and transverse Schutzer and Harris21 188 4-wire conﬁguration 27% 79% of the had the lowest trochanteric incidence of wire osteotomies had breakage healed Nercessian et al36 214 Biplanar osteotomy Overall breakage rate, 6.4% in the associated with 14%; 7.3% with biplanar group, decreased wire biplanar osteotomy, 6.2% in the breakage 19% with single-plane single-plane group osteotomy weight, sex, or underlying joint dis- unions, and Charnley hip scores Studies of the association between ease process. were significantly (P < 0.001) lower.39 other factors and trochanteric non- Nonunion after greater trochan- Important technical factors lead- union have not revealed any signifi- teric osteotomy using wires alone in ing to trochanteric nonunion in- cant difference with regard to type of primary THA occurs at rates ranging clude limited surgical experience, osteotomy36 or wiring technique.30 from zero to 7.9%.26,31,32,34,35,38,39 Such small size of the osteotomy frag- The great majority of THAs in nonunions have been correlated with ment, wires wrapped around the which trochanteric osteotomy is male sex, rheumatoid arthritis, and lesser trochanter, poorly tightened used are revision THAs. Schutzer revision surgery,39 but not with os- wires, and reattachment of the and Harris21 examined 188 revision teoporosis.27 Rates of incidence of trochanter to a primary acrylic THAs in which wires were used for pain (P < 0.001), limping (P < 0.001), bed.21,27,32 Boardman et al32 also not- trochanteric reattachment; they not- femoral component loosening (P < ed an increase in the rate of non- ed 3% nonunions and 4% delayed 0.005), and revision of existing com- unions when the primary surgeon unions. Technical errors (eg, reat- ponents (P = 0.0225) were signifi- was a resident instead of an attend- taching the trochanter directly to ce- cantly greater in the group with non- ing surgeon. ment, poor wire or mesh placement)

Volume 15, Number 10, October 2007 617 Fixation Systems of Greater Trochanteric Osteotomies

41-43 Figure 4 al wiring techniques. MacDonald et al42 evaluated 45 extended trochanteric osteotomies performed in revision THAs at a minimum 2-year follow-up and found a mean of 2.8 mm of trochanteric migration with the wire configuration and 0.3 mm with the cable technique, a difference that was statistically significant (P = 0.025). Studies of cable43 and two-plane cable41 constructs have reported no trochanteric migration in the early postoperative peri- od. Multifilament cable system configurations. A, 7×7.B, 19×8×7. Maximal compression perpendic- ular to the plane of the osteotomy minimizes both shear and rotation- were causes for most of the non- that there is bone-on-bone contact al deforming forces. Shaw and Dau- unions. In a review of 59 cases of and only limited contact with the bert29 examined the compressive THA with trochanteric osteotomies cement mantle. The bone edges forces generated by wire and cable that required revision, Bernard and should be abraded before fixation, cerclage systems. They showed that Brooks22 found that all nonunions and the wires should be checked for titanium cables achieved mean com- that underwent additional revision sufficient tightness. pression forces of 260 N, compared with wires failed to unite after the Several of the authors cited above with a range of 11 to 136 N for the second revision; pain was relieved have stated the need for proper sur- 18-gauge wire and 10 to 89 N for the in less than half of these THAs. gical technique and patient selection 20-gauge wire, depending on the Hodgkinson et al40 examined 59 re- when using wire constructs. These type of knot used (Tables 1 and 2). vision THAs performed before and conditions may be more important Compared with wires, multifila- after the implementation of the than the type of fixation used. Wires ment cables provide a stronger con- crossed-wire configuration using a should be used with caution in pa- struct at the osteotomy interface. compression spring; these authors tients who have had prior failure of Cables cannot be used in the reported a significantly (P = 0.001) wire constructs. same fashion as monofilament higher rate of union using the com- wires. Whereas wires can be tied into tight knots and tend not to un- pression spring. Cable Fixation Although the use of wires has de- ravel, cables cannot be tied into a se- creased with the availability of cable Second-generation multifilament ca- cure knot; the cables will unravel, and cable grip constructs, there are bles were first used for trochanteric and fixation will be lost. To solve still situations in which they may be reattachment in 1977 by Dall and this problem, cable sleeves are used. preferred. Some authors are hesitant Miles.10 Whereas monofilament wires The cables are threaded through the to use cables because of the inci- are made only of stainless steel or sleeves and tensioned, and then the dence of cable fraying and generation cobalt-chrome alloy, multifilament sleeve is crimped to hold the fixa- of debris. During revision for cables cables are also available in titanium. tion. In addition, cables should not that have frayed, it may be necessary Cables are constructed in several be crossed. Some micromotion will to use wires to avoid further debris configurations—for example, 8 bun- occur between the crossed cables, generation that could lead to third- dles of 7 monofilament strands sur- which will cause debris generation body polyethylene wear. When wires rounding a central bundle of 19 mono- and potential third-body wear. Final- are to be used for trochanteric fixa- filament strands, or 7 bundles of 7 ly, the cables do not hold fixation as tion, the literature supports the use monofilament strands (Figure 4). The well when threaded through a ce- of multiple wires in at least two material composition of these cables ment mantle. planes, such as with the Charnley and their configuration have com- Even when placed properly, how- and Harris techniques. The wires bined to offer more resistance to de- ever, there is still a significant rate of should be 16-gauge and fixed with a forming forces and provide better com- complications accompanying the knot twist or square knot technique. pression at the osteotomy site.10,14-22 use of cable systems. Kelley and If cement is used, an effort should be Clinical studies have compared Johnston33 evaluated 322 THAs in made to realign the trochanter so cable fixation to the more tradition- the early 1990s; they found a 12%

618 Journal of the American Academy of Orthopaedic Surgeons Gregg J. Jarit, MD, et al wire breakage rate and a 43% cable Table 3 breakage rate. They also reviewed the results of wire fixation and cable Properties of Cables and Cable Grips fixation; nonunion rates were 15% Source Cerclage System Maximum Strength (N) and 8%, respectively. In contrast, 17 Hop et al15 found an opposite trend Hersh et al Cable 1,300 ± 402 in nonunion rate between cable fix- Cable grip 1,900 ± 459 ation (19.7%) and wire fixation McCarthy et al14 1.6-mm cable 2,400 (14%). Despite the conflicting re- 2.0-mm cable 2,800 sults, both studies advised against Shaw and Daubert29 Titanium cable 1,027 ± 15* the use of cable-only constructs be- * Reported in Shaw and Daubert29 in kilogram-force (kgf) and here converted to N cause of potential complications and rounded off. related to metal debris. For these rea- sons, and because of the development of the cable grip systems, the cable through the bridge in the grip. nonunions, 11 (73%) were in THAs cable-only construct is generally re- The cable was either pulled through in which there was no cable break- served for repair of an extended tro- the anterior side of the grip and age. Silverton et al18 reported a 25% chanteric osteotomy. around the medial aspect of the prox- nonunion rate with the use of the ca- imal femur, or it was pulled through ble grip system and reported no dif- Cable Grip Fixation the posterior side and around the lat- ference in four versus two cables. eral aspect of the bone. They found McCarthy et al14 reported a 0.9% Dall and Miles10 introduced the that cables positioned posterolater- nonunion rate using the cable grip third generation of fixation, a tro- ally broke approximately 23 times system and found a statistically sig- chanteric grip system consisting of more often than did those placed an- nificant correlation between non- an H-shaped gripping apparatus that teromedially. union and posterolateral positioning supplements the multifilament ca- Hersh et al17 compared the relative of the cables. The nonunion rate was bles. These grip constructs were re- strengths of orthopaedic wires, ca- significantly less when cable fixa- ported to be better than wires and bles, and the cable grip system under tion was directly to host bone as op- cables at capturing the greater tro- quasistatic loading conditions (Tables posed to cement or allograft bone. chanter and resisting peritrochan- 1 and 3). The cable grip systems Prior history of nonunion was also a teric destabilizing forces, thus pre- withstood 1.5 times the maximum statistically significant (P = 0.046) venting trochanteric escape.10 load of the cable or wire alone. The factor in the failure of the trochan- The Dall-Miles cable grip system loads required to induce 1- or 2-cm teric fragment to heal after a second using 1.6-mm cables was initially re- trochanteric displacements also were reattachment procedure. ported to have a 6.2% breakage rate; examined. The cable grip system re- Silverton et al18 reported a 47% after the authors switched to 2.0-mm quired almost twice the load of the incidence of cable fraying and frag- cables, this rate was reduced to cable or wire alone to cause a 1-cm menting with the Dall-Miles cable 3.1%.10 They also found that cable displacement (1,397 N versus 771 N grip system and a 17% incidence of breakage was not seen after bone and 757 N, respectively) and 2 to 2.5 migrating metallic debris. In another union. The results of numerous times the load of the cable or wire to study of the cable grip system, Mc- follow-up studies using the Dall- cause a 2-cm displacement (1,900 N Carthy et al14 noted an 18% rate of Miles cable system have varied ac- versus 757 N and 1,100 N, respec- unraveling. Third-body wear of the cording to surgical technique. Ritter tively). polyethylene liner has been noted to et al20 reported a 32.5% cable break- The cable grip system also is asso- occur following breakage of cables age rate when the cables were ciated with certain complications. and multifilament cable grip con- threaded through the cement mantle Dall and Miles10 reported a non- structs.16,18,33,44 Compared with cable of the femoral prosthesis, and Silver- union rate of 5.4% with one hori- configurations, monofilament wire ton et al18 experienced a 22% cable zontal 1.6-mm cable, 4.8% non- configurations are associated with breakage rate. McCarthy et al14 re- union with two horizontal 1.6-mm significantly (P = 0.0001) less metal- ported a 10% cable breakage rate cables, and 1.5% nonunion with two lic debris and with reduced migra- with no significant difference be- horizontal 2-mm cables. In the last tion of such debris into the articula- tween cases with stainless steel and group, fibrous union occurred in tion; typically, such debris results in Vitallium cables. When assembling 3.1% with “slight positional loss.” volumetric wear and osteolysis.15,33 the cable grip constructs, these au- Ritter et al20 reported a 38% non- Nevertheless, Hop et al15 noted that thors14 varied the placement of the union rate in 40 THAs. Of the 15 there have been no reports of a sig-

Volume 15, Number 10, October 2007 619 Fixation Systems of Greater Trochanteric Osteotomies

Figure 5 Figure 6 these implants (Figure 6), the possi- bility exists for hardware irritation and hip abductor weakness. In the only clinical study of these systems published to date, Barrack and But- ler45 compared a fourth-generation cable plate to wires, cables, and the cable grip system; they noted significantly lower incidences of cable breakage (P < 0.025) and trochanteric nonunion (P < 0.05) with the cable plate. Although the cable plate group had a lower incidence of limp and demonstrated increased strength, these differences were not significant. Long-term clinical outcomes studies on these new devices may fa- cilitate the decision process for the optimal fixation system for trochanteric osteotomies. At the present time, these devices seem to be most clinically applicable for management of trochanteric nonunions or fractures. There are potential disadvantages Anteroposterior radiograph Anteroposterior radiograph to the cable plate systems. They are demonstrating a Dall-Miles cable grip. demonstrating a cable plate used for difficult to use with small fragments fixation of the osteotomized greater or with osteopenic bone. In addition, trochanter. failure of the construct likely will nificant difference in revision rates require a revision to remove the hard- between wire and cable configura- ware. Also, a more extensive dissec- tion is necessary to use these sys- tions. Because of this fact, and taking Cable Plate Systems into account the higher ultimate tems, and they are more expensive strength of the cable systems as well Fourth-generation cable grip systems than wires or cables with sleeves. as the significant rates of wire break- are now reaching the market; the The advancement of orthopaedic age and other complications with hope is that the cable fraying and implant design likely will provide wire fixation, we favor use of the ca- breakage that has occurred with pre- further options for the fixation of the ble grip construct (Figure 5). vious systems will be reduced. These trochanter. With the increased use of There are few contraindications implants use the 19 × 7 multifila- locking plates, it is likely that an im- to the use of cable grips. If the tro- ment pattern and have some poten- plant using this technology will be chanteric fragment has no soft- tial advantages over the traditional available for use in the future. To our tissue attachments, then it is devoid cable grip system, including the abil- knowledge, there have been no stud- of blood supply and the implant does ity to tighten and then retighten the ies or reports of locking plates being not need to be placed. In addition, an cables as needed. Different-sized used for this application. intact medial cortex just distal to the plates are also available, allowing the lesser trochanter must be present for placement of transverse cables below Extended Trochanteric fixation. When cementless femoral the level of the lesser trochanter, Osteotomy implants are used in conjunction thereby theoretically decreasing the with a greater trochanteric osteoto- incidence of trochanteric migration First described by Wagner46 and lat- my, and if there is medial bone loss or rotation. Placing transverse cables er popularized by Younger et al47 for from the femur, the wires should not below the level of the lesser tro- the removal of well-fixed femoral be in contact with the prosthesis; chanter also allows the surgeon to components in cases of revision such contact can lead to premature use extra cables in situations with THA, the extended trochanteric os- wire failure and the generation of extended trochanteric osteotomies. teotomy has proved to be a useful third-body metallic debris. Because of the larger size of some of tool in providing adequate surgical

620 Journal of the American Academy of Orthopaedic Surgeons Gregg J. Jarit, MD, et al exposure for both revision and com- ly, an osteotomy 12 to 15 cm in chanter or placed just distal to the plex primary cases.3,7,43 The extend- length (measured from the tip of the lesser trochanter, in an attempt to ed trochanteric osteotomy facilitates greater trochanter) satisfies these cri- counteract the forces pulling the os- implant and cement removal, the teria.5,8 teotomized segment proximally. The correction of proximal femoral de- The extended trochanteric osteot- most distal cerclage wire or cable is formity, access to the femoral diaph- omy is most commonly performed typically passed 2 to 3 cm proximal ysis for component placement, and through a posterolateral approach to to the distal end of the osteotomy. improved exposure of the acetabu- the hip, which allows adequate ex- The number of wires or cables lum while providing a relatively posure of the posterolateral cortex of spaced evenly between these two large surface area for healing of the the proximal femur.8,48 Approxi- points depends on the length of the osteotomized segment.3 In addition mately 5 to 10 mm of the posterior extended trochanteric osteotomy to allowing for the efficient removal border of the vastus lateralis is dis- segment. A wire is usually placed of femoral implants, this technique sected away from the linea aspera, distal to the osteotomy site to pre- reduces the likelihood of femoral allowing anterior reflection of the vent propagation of a femoral frac- fracture, femoral perforation, and ce- muscle while preserving its innerva- ture during placement of the femoral ment retention seen with more lim- tion. An oscillating saw is then used stem. This wire may be removed af- ited approaches.6 to make a longitudinal cut, starting ter placement of the stem at the dis- The indications for the extended at the posterior aspect of the greater cretion of the surgeon. Typically, the trochanteric osteotomy include revi- trochanter and extending distally most distal cerclage wire or cable is sion of well-fixed cemented and ce- along the femur to the preoperative- tightened securely, the middle wires mentless femoral components, re- ly planned length (Figure 7, A). Next, or cables tightened slightly less se- moval of a loose femoral component using either the oscillating saw or a curely, and the most proximal wire with a well-bonded cement mantle, high-speed burr, a smooth U-shaped left relatively loose to avoid segment cases of proximal femoral deformity end to the osteotomy should be fash- fracture, especially at the junction of in both revision and primary set- ioned, limiting the stress riser effect the greater trochanter and the lateral tings, and cases in which increased associated with a transverse distal cortex where the bone is thin and acetabular exposure is deemed nec- cut. Finally, the posteromedial cor- weak.5,48 Once the osteotomy is re- essary.5,7 A contraindication for ex- tex can be cut the length of the seg- duced and secure, the hip is taken tended trochanteric osteotomy is ment, completing the extended tro- through a complete range of motion femoral component revision in chanteric osteotomy. Once the to assess component stability and which the revision implant will be lateral and medial cortices have been soft-tissue tension and to ensure a cemented because extrusion of ce- osteotomized, broad osteotomes are lack of bony impingement. ment into the osteotomy site may inserted laterally to ease the osteot- Postoperative precautions after an hamper healing.5 omy open, allowing the greater tro- extended trochanteric osteotomy in- The osteotomy can be performed chanter and the lateral cortical seg- clude toe-touch weight bearing for before hip dislocation, after disloca- ment of the proximal femur to be the first 6 to 8 weeks. This is fol- tion with the femoral component reflected anteriorly, in continuity lowed by progression to weight bear- still in place, or after stem remov- with the attached muscles.5,8,48 ing as tolerated. Active hip abduc- al.8 Osteotomy before hip disloca- On completion of the revision or tion also is prohibited for the first tion is recommended when disloca- complex primary THA, the osteoto- 6 weeks postoperatively. tion is difficult or when there is my is reduced onto its bed on the Complications associated with concern of an intraoperative fracture posterior aspect of the femur. Proper the extended trochanteric osteotomy secondary to extensive heterotopic reapproximation of the osteoto- are similar to those seen with other ossification or significant femoral mized segment is important to avoid greater trochanter osteotomy tech- component subsidence. Otherwise, subsequent impingement. Once the niques, including intraoperative and most authors report performing the fragment is reduced and positioned, postoperative fracture, nonunion, osteotomy after dislocation. The ide- fixation is achieved with the use of malunion, and fragment migration. al length of the osteotomy, tailored two to four evenly spaced cerclage Clinical series evaluating the efficacy to each individual patient, is deter- wires or cables passed submuscular- of the extended trochanteric osteot- mined preoperatively. The osteoto- ly around the osteotomy segment omy in revision THA have demon- mized fragment should be long and proximal femoral diaphysis from strated that the rates of incidence of enough to provide maximal expo- posterior to anterior5,8,48 (Figure 7, B). nonunion and proximal segment mi- sure of the femoral canal and allow a The most proximal cerclage wire gration are markedly lower with this minimum of two cerclage cables or cable is most commonly passed technique than with standard and around it for later fixation. Typical- through a drill hole in the lesser tro- sliding osteotomies.5,48

Volume 15, Number 10, October 2007 621 Fixation Systems of Greater Trochanteric Osteotomies

Figure 7 tive “trochanteric” precautions in addition to standard hip precautions. These additional precautions include no active abduction and no single-leg stance on the affected side for 6 weeks. To our knowledge, there have been no studies documenting the effectiveness of these precautions. Management in the setting of early migration or symptomatic nonunion, with or without cable breakage, is revision of the fixation. Revising constructs in which cables have broken is especially important to prevent polyethylene wear result- ing from third-body debris generation. Revising such constructs could prevent the need for revision of otherwise well-functioning THA com- A, Extended trochanteric osteotomy. A longitudinal cut is made starting at the ponents. In addition, the actual posterior tip of the greater trochanter and extended distally along the femur to a wires or cables can migrate into the preoperatively planned length. The medial and lateral cortices are osteotomized, and joint, causing severe damage to the the trochanter is reflected anteriorly to gain exposure. B, Anteroposterior radiograph components. Asymptomatic wire demonstrating repair of the extended trochanteric osteotomy with a cable plate. breakage can be observed, but if the (Radiograph courtesy of Robert Barrack, MD.) wires begin to migrate, they should be removed. During revision surgery, as much Miner et al7 reviewed 192 consec- 68 healed with no evidence of seg- of the fixation construct as possible utive revision THAs in which an ex- ment migration; the other 5 cases should be removed. The bed of the tended trochanteric osteotomy was healed with <5 mm of migration. trochanter should be prepared by re- used for improved exposure. The au- moving all fibrous tissue from the edges of the osteotomy and ensuring thors reported 2 cases of nonunion Management of the presence of bleeding surfaces. Re- (1.2%) and 1 case of malunion (0.6%) Complications among 166 patients at a minimum peat fixation with a cable grip con- 2-year follow-up. Clinical evaluation Marked functional changes have struct should be performed. Repeat of these patients showed that pain been noted to occur following prox- wire fixation has a high failure rate.22 and walking scores improved from imal trochanteric fragment migra- The implementation of trochanteric a mean of 6.5 preoperatively to tion of >2 to 3 cm, including consid- precautions is important to reduce 9.8 postoperatively on the Merle erable weakness in the hip abductor recurrence of the nonunion. In addi- d’Aubigné and Postel scale. Based on musculature.50 Other clinical stud- tion, restricted initial weight bear- these findings, the authors conclud- ies indicate that trochanteric dis- ing26 with use of crutches for 8 ed that the extended trochanteric os- placements of approximately 5 to 20 weeks postoperatively has been asso- teotomy is a useful tool for the mm lead to an unstable fixation con- ciated with fewer nonunions. removal of well-fixed femoral com- struct with subsequent failure and Asymptomatic nondisplaced fi- ponents, with predictable postopera- result in significant gait alterations, brous nonunion is not an indication tive healing and excellent function- such as a Trendelenburg limp.17,31,50 for surgical revision. This should be al outcomes. It is therefore important to recognize evaluated with serial radiographs to Similar successful outcomes were quickly the development of non- ensure that there is no migration; shown by Mardones et al49 in their union or migration of the osteoto- however, surgery should not be con- review of 74 revision THAs. The au- mized fragment. When early break- sidered unless the patient becomes thors demonstrated that at a mean age of the fixation occurs, migration symptomatic. follow-up of 2 years, 73 of the 74 os- should be suspected. To avoid the Trochanteric bursitis after THA is teotomies had healed uneventfully; development of these complications, more common in the setting of tro- there was 1 nonunion. Of these 73, many surgeons institute postopera- chanteric osteotomy. This is often a

622 Journal of the American Academy of Orthopaedic Surgeons Gregg J. Jarit, MD, et al difficult problem to manage, and ies are needed to validate their effi- 10. Dall DM, Miles AW: Re-attachment treatment success is variable. When cacy and safety. of the greater trochanter: The use of this diagnosis is suspected, an injec- the trochanter cable-grip system. J Bone Joint Surg Br 1983;65:55-59. tion of lidocaine can be given to con- References 11. Bostrom MP, Asnis SE, Ernberg JJ, et firm it. Corticosteroid with lidocaine al: Fatigue testing of cerclage stainless may be injected to treat the inflam- Evidence-based Medicine: There are steel wire fixation. J Orthop Trauma mation. Multiple injections are con- 2 level II study (references 14 and 1994;8:422-428. traindicated because of the increased 34), 6 level III studies (references 15, 12. Guadagni JR, Drummond DS: Strength of surgical wire fixation: A risk of infection with each additional 23, 33, 36, 37, and 45), 26 level IV injection. laboratory study. Clin Orthop Relat studies (references 1-3, 7, 9, 14, 16, Res 1986;209:176-181. If pain does not improve after one 18, 20-22, 25-27, 30-32, 35, 38-43, 49, 13. Schultz RS, Boger JW, Dunn HK: or two injections, removal of the tro- and 50), and 2 level V studies (28 and Strength of stainless steel surgical chanteric hardware may be indicated. 47). The remainder are reviews, bio- wire in various fixation modes. Clin Blood tests should be obtained to rule Orthop Relat Res 1985;198:304-307. mechanical studies, case reports, or out infection. The patient should be 14. McCarthy JC, Bono JV, Turner RH, descriptions of technique. cautioned that removal of the hard- Kremchek T, Lee J: The outcome of trochanteric reattachment in revision ware may not completely eliminate Citation numbers printed in bold 22 total hip arthroplasty with a Cable pain. Bernard and Brooks removed type indicate references published Grip System: Mean 6-year follow-up. hardware from 36 patients; pain was within the past 5 years. J Arthroplasty 1999;14:810-814. relieved in only half of them. When 15. Hop JD, Callaghan JJ, Olejniczak JP, trochanteric nonunion or migration 1. Charnley J: Total hip replacement Pedersen DR, Brown TD, Johnston is present, the fixation should be re- by low-friction arthroplasty. Clin RC: The Frank Stinchfield Award: Orthop Relat Res 1970;72:7-21. Contribution of cable debris genera- vised as described above. 2. Charnley J: The long-term results of tion to accelerated polyethylene wear. low-friction arthroplasty of the hip Clin Orthop Relat Res 1997;344:20- Summary performed as a primary intervention. 32. J Bone Joint Surg Br 1972;54:61-76. 16. Bauer TW, Ming J, D’Antonio JA, Trochanteric osteotomy is a surgi- 3. Della Valle CJ, Berger RA, Rosenberg Morawa LG: Abrasive three-body AG, Jacobs JJ, Sheinkop MB, Paprosky wear of polyethylene caused by bro- cal technique that is used primarily WG: Extended trochanteric osteoto- ken multifilament cables of a total hip in complex primary and revision my in complex primary total hip ar- prosthesis: A report of these cases. THAs. Various devices such as wires, throplasty: A brief note. J Bone Joint J Bone Joint Surg Am 1996;78:1244- cables, and cable grip systems can Surg Am 2003;85:2385-2390. 1247. provide stable and rigid fixation. Al- 4. McGrory BJ, Bal BS, Harris WH: Tro- 17. Hersh CK, Williams RP, Trick LW, chanteric osteotomy for total hip ar- Lanctot D, Athanasiou K: Compari- though the material properties of ca- throplasty: Six variations and indica- son of the mechanical performance of bles and cable grip systems are supe- tions for their use. JAmAcad trochanteric fixation devices. Clin rior to those of monofilament wires, Orthop Surg 1996;4:258-267. Orthop Relat Res 1996;329:317-325. the surgeon must consider such po- 5. Archibeck MJ, Rosenberg AG, Berger 18. Silverton CD, Jacobs JJ, Rosenberg tential complications as debris gen- RA, Silverton CD: Trochanteric os- AG, Kull L, Conley A, Galante JO: teotomy and fixation during total hip Complications of a cable grip system. eration and third-body polyethylene arthroplasty. JAmAcadOrthopSurg J Arthroplasty 1996;11:400-404. wear should the cable break. Al- 2003;11:163-173. 19. Berman AT, Zamarin R: The use of though all systems are adequate in 6. Glassman AH: Exposure for revision: Dall-Miles cables in total hip arthro- achieving trochanteric union and Total hip replacement. Clin Orthop plasty. Orthopedics 1993;16:833-835. positive clinical results, we favor the Relat Res 2004;420:39-47. 20. Ritter MA, Eizember LE, Keating EM, 7. Miner TM, Momberger NG, Chong D, Faris PM: Trochanteric fixation by ca- cable grip system, which can provide Paprosky WL: The extended trochan- ble grip in hip replacement. J Bone the strongest fixation and result in teric osteotomy in revision hip arthro- Joint Surg Br 1991;73:580-581. lower rates of nonunion and trochan- plasty: A critical review of 166 cases at 21. Schutzer SF, Harris WH: Trochanteric teric migration. Positive clinical out- mean 3-year, 9-month follow-up. osteotomy for revision total hip arthro- comes are also dependent on patient J Arthroplasty 2001;16(8 suppl 1)188- plasty: 97% union rate using a compre- 194. hensive approach. Clin Orthop Relat selection because such factors as pre- 8. Meek RM, Greidanus NV, Garbuz DS, Res 1988;227:172-183. vious hip surgery and comorbid con- Masri BA, Duncan CP: Extended tro- 22. Bernard AA, Brooks S: The role of tro- ditions (eg, contralateral hip or spine chanteric osteotomy: Planning, sur- chanteric wire revision after total hip disease) may affect the decision to gical technique, and pitfalls. Instr replacement. J Bone Joint Surg Br use trochanteric osteotomy and the Course Lect 2004;53:119-130. 1987;69:352-354. 9. Pritchett JW: Fracture of the greater 23. Clarke RP, Shea WD, Bierbaum BE: subsequent choice of implant. The trochanter after hip replacement. Trochanteric osteotomy: Analysis of newer cable plate constructs are Clin Orthop Relat Res 2001;390:221- pattern of wire fixation failure and promising, but further clinical stud- 226. complications. Clin Orthop Relat

Volume 15, Number 10, October 2007 623 Fixation Systems of Greater Trochanteric Osteotomies

Res 1979;141:102-110. Clin Orthop Relat Res 1978;132:31- 42. MacDonald SJ, Cole C, Guerin J, Rora- 24. Markolf KL, Hirschowitz DL, Am- 38. beck CH, Bourne RB, McCalden RW: stutz HC: Mechanical stability of the 33. Kelley SS, Johnston RC: Debris from Extended trochanteric osteotomy via greater trochanter following osteoto- cobalt-chrome cable may cause ace- the direct lateral approach in revision my and reattachment by wiring. Clin tabular loosening. Clin Orthop Relat hip arthroplasty. Clin Orthop Relat Orthop Relat Res 1979;141:111-121. Res 1992;285:140-146. Res 2003;417:210-216. 25. Charnley J, Ferreiraade S: Transplan- 34. Wroblewski BM, Shelley P: Reattach- 43. Chen WM, McAuley JP, Engh CA, tation of the greater trochanter in ar- ment of the greater trochanter after Hopper RH, Engh CA: Extended slide hip replacement. J Bone Joint Surg Br throplasty of the hip. J Bone Joint trochanteric osteotomy for revision 1985;67:736-740. Surg Br 1964;46:191-197. total hip arthroplasty. J Bone Joint 35. Jensen NF, Harris WH: A system for 26. Harris WH, Crothers OD: Reattach- Surg Am 2000;82:1215-1219. trochanteric osteotomy and reattachment of the greater trochanter in total 44. Bronson JL: Articular interposition of ment for total hip arthroplasty with a trochanteric wires in a failed total hip hip-replacement arthroplasty: A new ninety-nine percent union rate. Clin technique. J Bone Joint Surg Am Orthop Relat Res 1986;208:174-181. replacement. Clin Orthop Relat Res 1978;60:211-213. 36. Nercessian OA, Newton PM, Joshi 1976;121:50-52. 27. Amstutz HC, Mai LL, Schmidt I: Re- RP, Sheikh B, Eftekhar NS: Trochant- 45. Barrack RL, Butler RA: Current status sults of interlocking wire trochanteric eric osteotomy and wire fixation: A of trochanteric reattachment in com- reattachment and technique refine- comparison of 2 techniques. Clin plex total hip arthroplasty. Clin ments to prevent complications fol- Orthop Relat Res 1996;333:208-216. Orthop Relat Res 2005;441:237-242. lowing total hip arthroplasty. Clin 37. Mauerhan DR, Lonergan RP, Mokris 46. Wagner H: A revision prosthesis for Orthop Relat Res 1984;183:82-89. JG, Kiebzak GM: Relationship be- the hip joint [German]. Orthopade 28. Coventry MB: The surgical technique tween length of stay and dislocation 1989;18:438-453. of total hip arthroplasty, modified rate after total hip arthroplasty. 47. Younger TI, Bradford MS, Paprosky from Charnley, as done at the Mayo J Arthroplasty 2003;18:963-967. WG: Removal of a well-fixed cement- Clinic. Orthop Clin North Am 1973; 38. Peters PC, Head WC, Emerson RH: less femoral component with an ex- 4:473-482. An extended trochanteric osteotomy tended proximal femoral osteotomy. 29. Shaw JA, Daubert HB: Compression ca- for revision total hip replacement. Contemp Orthop 1995;30:375-380. pability of cerclage fixation systems: A J Bone Joint Surg Br 1993;75:158-159. 48. Jando VT, Greidanus NV, Masri BA, biomechanical study. Orthopedics 39. Frankel A, Booth RE, Balderston RA, Garbuz DS, Duncan CP: Trochanteric 1988;11:1169-1174. Cohn J, Rothman RH: Complications osteotomies in revision total hip ar- 30. Berry DJ, Müller ME: Chevron osteot- of trochanteric osteotomy: Long-term throplasty: Contemporary techniques omy and single wire reattachment of implications. Clin Orthop Relat Res and results. Instr Course Lect 2005; the greater trochanter in primary and 1993;288:209-213. 54:143-155. revision total hip arthroplasty. Clin 40. Hodgkinson JP, Shelley P, Wroblew- 49. Mardones R, Gonzalez C, Cabanela Orthop Relat Res 1993;294:155-161. ski BM: Re-attachment of the un- ME, Trousdale RT, Berry DJ: Extended 31. Amstutz HC, Maki S: Complications united trochanter in Charnley low femoral osteotomy for revision of hip of trochanteric osteotomy in total hip friction arthroplasty. J Bone Joint arthroplasty: Results and complica- replacement. J Bone Joint Surg Am Surg Br 1989;71:523-525. tions. J Arthroplasty 2005;20:79-83. 1978;60:214-216. 41. Huffman GR, Ries MD: Combined 50. Bergström B, Lindberg L, Persson BM, 32. Boardman KP, Bocco F, Charnley J: An vertical and horizontal cable fixation Onnerfält R: Complications after total evaluation of a method of trochanter- of an extended trochanteric osteoto- hip arthroplasty according to Charnley ic fixation using three wires in the my site. J Bone Joint Surg Am 2003; in a Swedish series of cases. Clin Charnley low friction arthroplasty. 85:273-277. Orthop Relat Res 1973;95:91-95.

624 Journal of the American Academy of Orthopaedic Surgeons