Arch Orthop Trauma Surg (1994) 113:222-227 © Springer-Verlag 1994

R Haentjens • R R Casteleyn • R Opdecam

Hip arthroplasty for failed internal fixation of intertrochanteric and subtrochanteric fractures in the elderly patient

Received: 11 October 1993

A b s t r a c t Nine elderly patients were treated by salvage hip arthroplasty fol lowing failed internal fixation o f an in- tertrochanteric or subtrochanteric hip fracture. The mean age at time of fracture was 79 years (range 6 7 - 9 4 years). The mean interval f rom initial fracture fixation to con- version arthroplasty was 7 months (range 5 days to 19 months). One 84-year-old patient died 6 weeks postopera- tively f rom a pulmonary infection. The remaining eight patients were assessed clinically and roentgenographi- cally after a mean fol low-up period of 41 months (range 4 - 8 5 months). The functional results, rated according to the Merle d 'Aubign6 hip rating scale, were excellent in one patient, very good in four patients, good in two pa- tients, and fair in one patient. Serial roentgenographic analysis showed new bone formation around the ex- tramedullary part of the femoral component in all these patients and bone remodell ing o f the diaphyseal part o f the femur in all patients. The preservation o f the func- tional continuity o f the abduction apparatus during sur- gery and the early w a k i n g with full unrestricted weight- bearing made possible by the arthroplasty are considered to be the major contributing factors to these results.

introduction

At the intertrochanteric or subtrochanteric level several treatment options may be considered for failed internal fixation necessitating reoperation. For a young patient all authors agree that a second attempt at internal fixation with or without bone grafting or os teotomy is the favoured procedure [1, 3, 10-12, 16]. For an elderly pa- tient, hip arthroplasty might also be considered [10-13] . Salvage by a prosthesis at the intertrochanteric and sub- trochanteric level has indeed been advocated in carefully selected cases [10, 12, 13, 16]. Salvage prosthetic hip

P. Haentjens ([~). P. P. Casteleyn - P. Opdecam Department of Orthopaedics and Traumatology, Academisch Ziekenhuis V.U.B., Vrije Universiteit Brussel, Laarbeeklaan 101, Bq090 Brussels, Belgium

surgery has been advocated in elderly patients with ace- tabular destruction or osteoporosis and in elderly patients who are unable to participate in the postoperative rehabil- itation program with protected weight-bearing on the op- erated limb [10, 12, 16].

Separate reports on salvage hip arthroplasty for the treatment of failed internal fixation of intertrochanteric and subtrochanteric fractures are rare [12, 13]. The pur- pose of this paper is to review our experience with failed internal fixation of intertrochanteric and subtrochanteric hip fractures with conversion to hip arthroplasty. The sur- gical approach and the femoral reconstruction technique peculiar to this procedure are discussed.

Patients and methods

During the period from March 1983 to April 1991, nine elderly pa- tients were treated with hip arthroplasty following failure of inter- nal fixation of an intertrochanteric or subtrochanteric fracture of the proximal femur (Table 1). The patients were made up of five women and four men. The mean age at the time of fracture was 79 years (range 67-94 years). Three patients had their original fixa- tion performed elsewhere. Six patients had originally been treated at our University Hospital.

The initial fractures were classified according to the system de- scribed by Evans [4] and modified by Jensen [8] for the inter- trochanteric fractures, and according to Seinsheimer's system. [20] for the subtrochanteric fractures. There were two stable and three unstable intertrochanteric fractures, and one Seinsheimer type II and three Seinsheimer type V subtrochanteric fractures.

At initial operative reduction and internal fixation a variety of internal fixation devices had been used. An articulated nail-plate device was used in one patient (Fig. 1A), flexible Ender nails were used in one, a dynamic hip screw was used in two, and a one-piece AO/ASIF condylar blade plate with lag screws was used in five pa- tients. In one patient (V.J.) a second attempt at union with internal fixation was performed. This ultimately also resulted in a failure.

The mean interval from initial fracture fixation to conversion arthroplasty was 7 months (range 5 days to 19 months). The inter- nal fixation device had broken in six patients; implant cut-out was noted in four patients (Fig. 2A), while acetabular damage was noted in two patients.

In all nine patients the proximal femur was reconstructed with a cemented large femoral component with diaphyseal support [19]. If destruction of the acetabulum was present, an acetabular recon- struction was performed (total hip arthroplasty: Fig. 1). If on the

Table 1 Clinical details of nine patients who underwent Pa- Sex Age a Type of hip arthroplasty for failed in- tient (years) fracture ternal fixation of intertrochan- [4, 20] teric and subtrochanteric frac- tures of the femur D.L. F 94 Evans 2

P.R. M 71 Evans 5

V.E. F 82 Evans 5

V.H. M 84 Seinsheimer V

V.M. F 67 Seinsheimer V

V.A. M 74 Evans 5

W.J. F 83 Seinsheimer II

V.V. F 78 Evans 2

V.J. M 75 Seinsheimer V a At time of initial fracture

223

Initial treat- Time to Implant Implant ment of failure failure cut-out fracture

Dynamic hip screw 19 months + -

AO/ASIF condylar plate 9 months + -

Ender nails 4 weeks - +

AO/ASIF condylar plate 5 days - +

AO/ASIF condylar plate 6 months + -

McLaughlin blade plate 5 months + +

AO/ASIF condylar plate 13 months + -

Dynamic hip screw 10 days - +

AO/ASIF condylar plate 9 months + -

preoperative roentgenographs the acetabulum appeared uninvolved and if during operation the acetabular cartilage appeared normal, a mobile self-centring cup was used (bipolar arthroplasty: Fig. 2).

Surgical approach [6, 14]

The patient is placed on the uninvolved side on a standard operat- ing table, with the trunk supported by chest-rests. The surgical ap- proach is through a straight lateral incision, which begins three to four fingerwidths proximal to the greater trochanter and extends toward the lateral femoral condyle for a distance of about 20 cm.

The subcutaneous layer is dissected and the fascia lata is split longitudinally, exposing the greater trochanter. The previous fixa- tion devices are removed. The greater trochanter is osteotomized vertically along a plane that will leave the insertions of the gluteus medius proximally and the vastus lateralis distally untouched, to maintain the continuity of the abduction apparatus. The abduction apparatus is then reflected anteriorly, exposing the ununited frac- ture and the proximal femoral diaphysis. The diaphysis is sec- tioned transversely, just distal to the level of non-union. This sec- tion must be accurate in order to obtain a large contact-bearing area between the femoral prosthesis and the diaphyseal cortex.

The capsule is opened in line with the femoral neck. Whenever possible the hip capsule is carefully preserved. Only the femoral head and the femoral neck are removed. All the remaining vascu- larized bone fragments of the proximal femur are carefully pre- served.

Acetabular reconstruction [6, 18]

In two patients destruction of the acetabulum necessitated recon- struction by total hip arthroplasty (Fig. 1). In one of these patients there was only minor acetabular involvement, so a cementless self- tapping hemispherical threaded ring was used (Mecring, Mecron, Berlin). This hemispherical shaped titanium alloy ring is available in 14 different sizes with an outer diameter ranging from 42 to 71 ram. The corresponding inner polyethylene cup with an internal di- ameter of 32 mm was used. In the other patient there was more ex- tensive destruction of the acetabulum; in this patient reconstruction was performed using three fully threaded cancellous bone screws functioning as pile screws, and a cemented (Palacos with gentam- icin, Schering Corporation USA, Hemilworth, New Jersey), con- ventional polyethylene cup with internal diameter of 32 mm (Pro- tek AG, Bern, Switzerland; Fig. 1D, F). In seven patients there was no acetabular involvement and a bipolar component (De Puy, War- saw, Indiana) was used (Fig. 2). This self-centring cup is locked onto the femoral component after the component has been ce- mented in and the polymethylmethacrylate cement in the femo- ral diaphysis has set. This self-centring mobile cup is available in 19 different sizes and its external diameter ranges from 43 to 65 mm.

Femoral reconstruction

In all patients a Mtiller stainless steel large femoral component (type 316 L; Protek AG) was used (Figs. 1C, E, 2B, C). This large femoral component has a 32-ram head and is available in 11 extra- medullary lengths ranging from 60 to 180 mm. The greater trochan- ter can be fixed to the lateral part of the prosthetic component by a polyacetal washer and two screws. The lesser trochanter can be fixed to the medial part with a cerclage wire that is passed through two holes located in the medial part of the femoral component. The straight, intramedullary part of the femoral component is always 140 mm in length and 13 mm in diameter. This femoral component has a conical enlargement to avoid an abrupt transition between the extramedullary and intramedullary parts (Figs. 1C, E, 2B, C).

The femoral shaft is prepared using first a straight intra- medullary reamer. Next, the proximal part of the shaft is prepared further with a reamer that has a conical enlargement, so that a cor- rect fit is obtained between it and the larger proximal shaft of the prosthesis. The appropriate length of the extramedullary portion of the femoral component can be chosen by using the adjustable trial stem provided by the manufacturer.

The trial stem is assembled with a trial cup, and test reductions are performed to determine the exact length that will provide the desired tension of the abductor muscles. The hip is dislocated again, and all trial components are removed and a final trial reduc- tion is performed with the prosthetic components that will be in- serted. When the medullary canal is large, AO/ASIF semitubular plates are inserted in the canal to help fill the space between the stem and the cortex. The degree of anteversion is determined dur- ing this final reduction. A notch on the inferior part of the ex- tramedullary portion of the stem is used as a guide for this purpose, the exact position of the notch in relation to the femoral cortex be- ing marked by electric cautery on the cortical bone.

After all componens have been removed, a Seidel plug (How- medica, Rutherford, New Jersey) is inserted and the medullary canal is rinsed with saline solution. One or two units of poly- methylmethacrylate cement (Palacos with gentamicin, Schering Corporation) is injected under pressure, and the femoral compo- nent and, i f necessary, the AO/ASIF semitubular plates are in- serted. In the bipolar arthroplasty group after the polymethyl- methacrylate has set the self-centring cup is locked onto the pros- thetic head. The prosthesis is then reduced into the acetahulum.

Soft tissue reconstruction [14, 16]

The capsular incision is carefully closed. The remaiping part of the lesser trochanter is fixed to the femoral component by a cerclage wire of 1.0 mm which is passed through the two holes on the medial side of the femoral component. The g!uteus medius, greater trochanter, and vastus lateralis are returned to their anatomical locations, and 4.5-mm holes are drilled through the greater trochanter. The greater trochanter is then fixed through these holes to the lateral part of the prosthetic stem by one or two screws and

224

Fig. 1A-F Roentgenograms of a 74-year-old man (V.A.) showing the indication for salvage total hip arthroplasty and the long-term follow-up roentgenographic changes following salvage total hip arthroplasty. A Nail penetration into the acetabulum 5 months fol- lowing operative reduction and internal fixation. B Tomogram after removal of the fixation device clearly shows the collapse of the proximal femur and its poor mechanical properties. C On the femoral side a reconstruction is performed using a cemented fem- oral component with diaphyseal support. During the surgical ap- proach the functional continuity of the abduction apparatus is pre- served. The greater trochanter is fixed to the femoral component by a washer and one screw. D On the acetabular side reconstruc- tion is performed using pile screws and a cemented polyethylene cup. E At 7 years and 1 month the overall hip score according to Merle d'Aubign6 [15] is rated as good. The patient has occasional pain (score five points), a normal range of motion (score six points) and uses a cane only after long distances (score five points). The femoral roentgenogram shows bone formation around the extra- medullary part of the femoral component. The remaining part of the femur shows bone remodelling with cancellization. F The ace- tabular roentgenogram shows no migration of the prosthetic com- ponents after 7 years and 1 month

Fig. 2A-C Roentgenograms of a 78-year-old woman (V.V.) show- ing the indication for salvage bipolar arthroplasty and the long- term roentgenographic changes following salvage bipolar arthro- plasty. A Initially the stable intertrochanteric fracture (Evans type 2) had been treated by sliding hip-screw fixation. On the 10th day after surgery the patient suffered acute onset of pain. This roent- genogram reveals implant cut-out and further extension of the frac- ture into the diaphyseal area. B Salvage bipolar arthroplasty allows ambulation with full unrestricted weight-bearing. C Five years and 8 months after salvage bipolar arthroplasty the overall hip score according to Merle d'Aubign6 [15] is rated as very good: the pa- tient has no pain (score six points), a normal range of motion (score six points) and uses a cane only for long distances (score five points). This roentgenogram shows the bony envelope sur- rounding the extramedullary part of the femoral component and cancellization of the femur

a polyacetal washer (Figs. 1C, E, 2B, C). The fascia lata is sutured and the skin is closed. Suction drainage is placed in all patients for 48-72 h. The mean duration of anaesthesia was 150 min; mean du- ration of surgery was 98 min; mean operative blood loss was 1286 ml. No intraoperative fractures occurred.

Postoperative care

An antibiotic is given just before operation and is continued for only 24 h postoperatively (cefamandole, 2 g i.v. every 8 h). Post- operatively the patients receive thromboembolic prophylactic treat- ment (calcium heparinate, 5000 units s.c. every 8 h) for 5 days. Un- less contraindicated, anti-inflammatory medication is administered postoperatively for 5 days (indomethacin, 50 mg p.o. every 12 h). Active and passive mobilization of both limbs is started as soon as possible, taking care to avoid forced adduction or rotation of the hip that was operated on. Moderate flexion of both the hip and knee joints, with a large pillow between the ankles during bed rest, is recommended. Walking with full weight-bearing is allowed as soon as the postoperative roentgenograms confirm that the pros- thetic components are properly positioned. Postoperative ambula- tion with full weight-bearing was possible on the 4th day for all patients but one. This one patient unable to walk was an 84-year old man who contracted a fatal pulmonary infection.

Clinical and roentgenographic assessment

Each patient was reviewed for postoperative complications. The functional results were judged according to the hip rating scale of Merle d'Aubign6 [15]. In this scale pain around the hip joint, mo- tion and walking ability are assessed separately on a six-point scale. Summation of the three scores gives a total score with a cor- responding overall final functional result.

For all patients, roentgenograms made at final follow-up exami- nation were compared with the immediate postoperative roentgeno- grams. On the acetabular side these roentgenograms were analysed for signs of either loosening or migration of a cemented acetabular component, or protrusion of a bipolar component. On the femoral side they were analysed for signs of bone remodelling of the femo- ral cortex, new bone formation, loosening of the stem, and subsidence of the stem. As most of the patients were quite elderly, the duration of follow-up was determined largely by how long the patients lived.

Results

Funct iona l results

The funct ional results were assessed according to the Mer le d ' A u b i g n 6 scale [15] in all pat ients but the one suf-

225

fer ing the fatal pu lmonary infection, who d ied 6 weeks af- ter the convers ion procedure (Table 2). The mean fo l low- up per iod was 41 months (range 4 - 8 5 months) . A t the fi- nal fo l low-up examina t ion all pat ients were sat isf ied with the result of the conversion procedure. No patient reported be ing worse fo l lowing the convers ion to hip replacement . Pr ior to convers ion ar throplas ty all pat ients exper ienced cons iderable pain and d isabi l i ty associa ted with the fail- ure o f the internal f ixation.

At f inal fo l low-up evaluat ion according to the Mer le d ' A u b i g n 6 scale the mean pain score was 5.75 points , the mean score for mobi l i ty was 5.87 points, and the mean score for walk ing abi l i ty 4.75 points.

Serial roen tgenographic analysis showed no cases of acetabular prot rus ion fo l lowing b ipo la r ar throplasty (Fig. 2B, C). In all cases the ace tabu la r ca r t i l age he igh t ap- peared preserved. Serial roen tgenographic analysis o f the distal femur showed signs o f bone remode l l ing in all pa- tients. The cortex decreased in densi ty and was rep laced by a t rabecular bone densi ty with a very thin shell o f cor- t ical bone at the per iphery (Figs. 1C, E, 2B,C) . In all pa- t ients new bone format ion could be observed around the ex t ramedul la ry part o f the femora l componen t (Figs. 1E, 2C). It started very ear ly pos topera t ive ly and the vascu- lar ized bone f ragments were incorpora ted in this bony en- velope. This bone format ion did not affect the c l in ical re- sults in terms o f pers is tent pain or loss o f mobil i ty.

There were no cases of loosening or subs idence o f the femoral component .

Compl ica t ions

One pat ient suffered a fatal pu lmonary infection. This 84- year -o ld pat ient had been ly ing on the f loor for more than 24 h after the ini t ial fracture. His Se inshe imer type V frac- ture was t reated with a 95 ° A O / A S I F b lade plate and three lag screws; implan t cut-out occurred on the 5th day post- operatively. Therefore a reoperat ion using a bipolar arthro- p las ty was performed. Af te r this opera t ion a pu lmonary infect ion developed. The pat ient d ied 6 weeks later.

One pat ient showed p rob lems with wound heal ing, probably due to subcutaneous fat necrosis. This was treated

Table 2 Results of hip ar- throplasty in the nine patients Pa- Type of Length

tient arthro- of fol- (THA total hip a~aroplasty, ROM range of movement) plasty low-up

(months)

D.L. Bipolar 35 P.R. Bipolar 40 V.E. Bipolar 4 V.H. Bipolar - V.M. THA 78 V.A. THA 85 W.J. Bipolar 9 V.V. Bipolar 68 V.J. Bipolar 12

Functional results [15]

Pain ROM Gait Overall result

Comments

6 6 5 Very good 6 6 5 Very good 5 5 5 Good

6 6 6 Excellent 5 6 5 Good 6 6 2 Fair 6 6 5 Very good 6 6 5 Very good

Died 6 weeks postop. (pneumonia) Dislocation 5 weeks postop.

Wound healing problem

226

by local debridement and closure of the defect. There were no cases of deep infection.

One patient suffered posterior dislocation of the total hip arthroplasty 5 weeks postoperatively. This was treated by closed reduction followed by 3 weeks of immobilization. The same patient fell at home 4.5 months postoperatively and sustained a fracture of the distal part of the femur, which was treated by internal fixation and bone grafting. More than 5 years after this procedure, the fracture of the distal femur is healed and the overall hip score is rated as excellent.

One patient complained of pain over the trochanteric area 5 months after insertion of the prosthesis. This pain was due to progressive loosening and migration of the screws. The screws and the polyacetal washer were re- moved. At final follow-up examination the hip score is rated as very good.

No patients developed pressure sores. No neurological or vascular complications were observed.

Discussion

The reoperation rate due to failure of union or material failure following open reduction and internal fixation of intertrochanteric and subtrochanteric fractures has been reported at 0 .5%-2% [12].

In a young patient with a normal bone structure one should always consider operative reduction and internal fixation to treat failed internal fixation of an in- tertrochanteric or subtrochanteric fracture. This often re- quires osteotomy and augmentation with autogenous bone grafts [1, 3, 11, 16]. In a young patient weight-bearing is usually avoided until there is some evidence of osseous bridging, which usually takes up to 3 months [5, 17].

In an elderly patient there are several treatment options for failed internal fixation of an intertrochanteric or a sub- trochanteric fracture [11, 12, 16]. In an elderly patient, as in a young patient, one must always consider a second at- tempt at operative reduction and internal fixation. Here, too, the principles for salvage of a failed first procedure are stable bone apposition, preferable in valgus position, secure internal fixation and bone grafting [1, 3, 10, 12, 16]. Only in stable configurations can partial weight-bear- ing of 10-15 kg be allowed immediately [5, 16, 17]. An elderly patient is often unable to adhere to this postopera- tive rehabilitation program with protected weight-bearing or non-weight-bearing on the operated limb [16]. Further- more, elderly patients often have marked osteoporosis, jeopardizing fixation with a screw-plate device [16]. Fre- quently, the purchase of the threads in the head-neck frag- ment and screws in the shaft fragment is less than ideal due to osteoporosis [16]. Screw purchase in osteoporotic bone is poor and the threads may cut out [16, 21]. This failure does not seem to be related to any particular type of device but merely to local and general conditions [1, 3, 10, 12, 16]. Therefore, in the elderly patient, salvage by arthroplasty has been proposed [10, 16].

Salvage of failed internal fixation of a femoral neck fracture by hip arthroplasty has been described as a suc-

cessful procedure [7, 9, 11, 13]. The functional results are comparable to those seen in primary total hip arthroplasty, where the same conventional femoral prosthetic compo- nents are used [7, 13]. Therefore, in the elderly patient with failed fixation of an intracapsular fracture, replace- ment hip surgery is considered the treatment of choice [7, 9, 13, 16]. At this level the procedure can usually be performed using a conventional femoral component [13, 16].

Salvage of failed internal fixation of a fracture at the intertrochanteric or subtrochanteric level by hip arthro- plasty has occasionally been considered on the basis of carefully selected indications [12, 13, 16]. In patients un- able to restrict the weight borne on the operated limb [16] and in those with acetabular destruction [12, 13], replace- ment hip surgery may be preferred to a second attempt at operative reduction and internal fixation. Salvage arthro- plasty following failure of fixation of an intertrochanteric or subtrochanteric fracture is usually considerably more difficult than salvage arthroplasty after failed fixation of a femoral neck fracture [13].

Failed fixation of an intertrochanteric or subtrochan- teric fracture converted to arthroplasty does not usually allow the use of a conventional femoral component. Moreover, loss of attachment for the hip abductor muscles might contribute to poorer fuctional results [12]. There- fore, special prosthetic components and surgical recon- strnction techniques must be used to obtain satisfactory results in failed fixations at this level [13]. Some surgeons prefer to use a standard conventional femoral prosthetic component and to reconstruct the proximal femur using bone grafts [13]. In an elderly patient population, how- ever, the present authors would not recommend the use of bone grafts and a conventional prosthesis because this makes the procedure more complex and does not allow weight to be put on the operated limb. Weight-bearing should indeed be delayed until there is some evidence of osseous bridging and healing at the host-graft junction [13]. The present authors prefer to use a large femoral prosthesis with diaphyseal support as a technically simpler solution in these elderly patients. This large prosthetic component allows the poor mechanical quality of the proximal femur to be bypassed. This cemented prosthesis allows immediate ambulation with full unrestricted weight-bearing. We prefer a cemented device in these el- derly patients because we believe that cement has still a place in prosthetic hip surgery in the older osteoporotic patient [16]. The femoral component in our patients was always fixed with polymethylmethacrylate cement. A sin- gle unit of cement was usually sufficient to fill the medullary canal, but when the canal was large, AO/ASW semitubular plates were used to augment the filling [19]. According to Schneider [19], these semitubular plates, or even multiple 2.0- to 2.5-mm Kirschner wires, provide a press-fit of the femoral stem, which enhances prosthetic stability. In the present series cement containing gentamicin was used in all cases. Comparison of the deep infection rates using regular or antibiotic-loaded cement suggests that the use of gentamicin-loaded cement offers significant benefit in cases of secondary operative procedures [21].

The surgical approach is based on maintaining an in- tact muscular-osseous-muscular sleeve constituted of the gluteus medius, part of the greater trochanter and the vas- tus lateralis [6, 14]. This approach thereby respects the functional continuity of the abduction apparatus. In con- trast to other approaches employing osteotomy of the greater trochanter, loosening or removal of the screws from the femoral component does not result in loss of continuity of the abduction apparatus.

In the present series the hip joint capsule was pre- served and carefully sutured in all bipolar arthroplasties. Resection of the capsule was performed only when such resection was essential for good visualization of the ac- etabulum for reconstruction. Preservation and systematic closure of the capsule has indeed been recommended in order to prevent dislocation [6].

Bone formation around the extramedullary part of the femoral component has also been described by several au- thors who have used the same type of femoral prosthetic component. This bone formation might be attributed to the presence of viable periosteum and to the preservation of all vascularized bone fragments around the extra- medullary part of the stem [6, 19]. Only the femoral head, the femoral neck and the loose non-vascularized bone fragments should be removed when performing this surgi- cal procedure. Several studies have indeed demonstrated that the fracture fragments from the patient's trochanteric and proximal diaphyseal region that are left behind par- ticipate in the callus formation. This bone formation some- times results in a bony envelope completely surrounding the extramedullary part of the stem [6, 19]. Cortical re- modelling of the remaining femoral diaphyseal cortex has also been described when using the same cemented large femoral components [6, 19]. The reason for this might be a change in loading pattern or change in vascularity of the remaining bone [2]. Despite this remodelling that occurs around intramedullary steins of both large and standard femoral stems, cement fixation has proved to be a reliable and effective method of fixation for many years.

Conclusion

On the basis of these results we consider that prosthetic replacement offers a viable treatment solution for the dif- ficult problem of failed internal fixation at the inter- trochanteric or subtrochanteric level in the elderly patient.

227

References

1.Baker HR (1960) Ununited intertrochanteric fractures of the femur. Clin Orthop 18 : 209-220

2.Blunn GW, Wait ME (1990) A comparison of the fixation of custom-made prosthesis with the fixation of standard joint re- placements. In: Oldner J (ed) Implant bone interface. Springer, Berlin Heidelberg New York, pp 151-161

3.Boyd HB, Lipinski SW (1957) Nonunion of trochanteric and subtrochanteric fractures. Surg Gynecol Obstet 104:463-470

4. Evans EM (1949) The treatment of trochanteric fractures of the femur. J Bone Joint Surg [Br] 31-B : 190-203

5. Ganz R, Thomas RJ, Hammerle CP (1979) Trochanteric frac- tures of the femur. Treatment and results. Clin Orthop 138: 30-40

6. Haentjens P, Casteleyn PP, De Boeck H, Handelberg F, Op- decam P (1989) Treatment of unstable intertrochanteric and subtrochanteric fractures in elderly patients. J Bone Joint Surg [Am] 71-A : 1214-1225

7.H~igglund G, Nordstrtm G, Lindgren L (1984) Total hip re- placement after nailing failure in femoral neck fracture. Arch Orthop Trauma Surg 103 : 125-127

8. Jensen JS (1980) Classifications of trochanteric fractures. Acta Orthop Scand 51 : 803-810

9. Johnson J (1984) Comparison between hemiarthroplasty and total hip replacement following failure of nailed femoral neck fractures focused on dislocation. Arch Orthop Trauma Surg 102:187-190

10. Knight WM, De Lee JC (1982) Nonunion of intertrochanteric fractures of the hip: a case study and review. Orthop Trans 16 : 438

ll.Lifeso R, Younge D (1990) The neglected hip fracture. J Or- thop Trauma 4 : 287-292

12.Mariani EM, Rand JA (1987) Nonunion of intertrochanteric fracture of the femur following open reduction and internal fix- ation. Clin Orthop 218 : 81-89

13.Mehlhoff T, Landon GC, Tullos HS (1991) Total hip ar- throplasty following failed internal fixation of hip fractures. Clin Orthop 269 : 32-37

14.Mercati E, Guary A, Miquel L, Bourgeon A (1972) Une voie d'abord posttro-exteme de la hanche. Int4r4t de la rtalisation d'un muscle "digastrique". J Chir (Paris) 103:499-504

15. Merle d'Aubign6 R (1970) Cotation chiffrte de la fonction de la hauche. Rev Chir Orthop 56:481-486

16. Meyers MH (1985) Delayed union, nonunion and malunion of hip fractures. In: Meyers MH (ed) Fractures of the hip. Year Book Medical, Chicago, pp 343-353

17. Schatzker J, Tile M (1987) The rationale of operative fracture care. Springer, Berlin Heidelberg, New York, pp 217-234

18. Schneider R (1980) Die Armierung der Pfanne bei der Total- endoprothese der Htifte. Unfallheillcunde 83 : 482-488

19. Schneider R (1982) Die Krtickstockprothese. In: Schneider R (ed) Die Total-Prothese der Htifte. Ein biomechanisches Kon- zept und seine Konsequenzen. Huber, Bern, pp 237-246

20. Seinsheimer F (1978) HI: Subtrochanteric fractures of the fe- mur. J Bone Joint Surg [Am] 60-A: 300-306

21.Wroblewski BM (1990) Infection. In: Wroblewski BM (ed) Revision surgery in total hip arthroplasty. Springer, New York Berlin Heidelberg, pp 47-61

本文献由“学霸图书馆-文献云下载”收集自网络，仅供学习交流使用。

学霸图书馆（www.xuebalib.com）是一个“整合众多图书馆数据库资源，

提供一站式文献检索和下载服务”的24 小时在线不限IP

图书馆。

图书馆致力于便利、促进学习与科研，提供最强文献下载服务。

图书馆导航：

图书馆首页 文献云下载 图书馆入口 外文数据库大全 疑难文献辅助工具