Most intertrochanteric hip fractures that are treated with contemporary methods of internal fixation heal. 2,10 However, nonunion or early loss of fracture fixation occurs occasionally, most often in patients with unfavorable fracture patterns, poor bone quality, or suboptimal position of internal fixation devices. 1–3,7,9–11 Most patients with hip fracture nonunion or fracture fixation failure have marked pain and functional disability. Hip arthroplasty may be considered a salvage method for older patients, patients with severely compromised bone inadequate for fixation, or patients with a severely damaged hip articular surface. 5,8,12,14 However, for physiologically younger patients with good bone quality, preservation of the patient’s femoral head probably is preferable to prosthetic replacement. The results of revision internal fixation for failure of fixation of intertrochanteric fractures have received little attention in the orthopaedic literature. Several small series 7,11,13,16 have had encouraging success rates with revision internal fixation and bone grafting using multiple techniques. The purpose of the current study was to review a series of patients treated with revision internal fixation and bone grafting for failure of internal fixation of an intertrochanteric hip fracture to determine the results, success rate, and complications associated with the procedure.
MATERIALS AND METHODS
Between 1981 and 2000, 20 patients with 20 intertrochanteric fractures that failed initial internal fixation attempts were treated at the authors’ institution with revision internal fixation and bone grafting. Patients were located using an institutional computerized diagnosis-based database. Institutional review board approval and consent were obtained for a retrospective review. Pathologic fractures were excluded. During a similar time, approximately 60 patients with a mean age of 77 years were treated with hip arthroplasty for failed internal fixation of an intertrochanteric hip fracture. 6 Most patients in the current series lived in the geographic region where they were treated, however, because of the complexity of the reconstructions, many patients were referred from other regions.
There were 11 women and nine men with an average age of 58 years (range, 21–86 years). Only three patients were older than 70 years. The fracture was of the right hip in 14 patients and of the left in six. All patients previously had the intertrochanteric fracture treated with internal fixation. Information on the original internal fixation device was available in all patients: in 14 patients the original implant was a sliding hip screw, in five a blade plate, and in one a cephalomedullary nail. Nineteen of 20 patients had an established nonunion with a mean interval of 12 months (range, 4–27 months) after the original internal fixation, and one of 20 had acute fixation failure (associated with infection) at 6 weeks. One patient had posterior cut-out of a blade plate from the femoral head but had well-preserved acetabular articular cartilage. The remaining 19 patients had fracture nonunion without implant cut-out through the femoral head, and all had well-preserved acetabular articular cartilage. Two of 20 patients had infections of their original internal fixation devices and were treated with hardware removal, debridement, organism-specific intravenous antibiotics, and delayed internal fixation at 6 weeks and 12 weeks after hardware removal, respectively.
The choice of implant used and the type of bone graft used was at the discretion of the treating surgeon. Implants used for revision internal fixation were as follows: 11 hips were revised using an angled blade plate (seven, 95°; two, 90°; one, 110° [Synthes, Paoli, PA], and one Harris blade plate [Zimmer, Warsaw, IN]), five using a Dynamic hip screw (Synthes), three using a Dynamic condylar screw (Synthes), and one using a Zickel nail (Stryker-Howmedica-Osteonics, Allendale, NJ). Cancellous autograft bone was used in 17 patients and particulate allograft bone was used in three patients. The mean operative time was 260 minutes (range, 115–590 minutes). The mean estimated blood loss was 1300 mL (range, 200–3000 mL). In three patients fractures were osteotomized in the intertrochanteric region along the original fracture line to allow valgus positioning of the proximal fragment and good bony apposition. In one patient a valgus producing subtrochanteric osteotomy was done. No cement augmentation was used in any patient.
Clinical and radiographic results were reviewed retrospectively and results and complications were analyzed. Patients were followed up until union, or failure of fixation, and had revision of the original fixation to an arthroplasty. The mean clinical followup for the group was 27 months (range, 3–120 months) and mean radiographic followup was 22 months (range, 3–120 months). In 12 of 20 patients (60%) the postoperative radiographs were available for review with a mean radiographic followup of 30 months (range, 3–120 months). In eight patients, the final postoperative radiographs had been discarded after 10 years as part of institutional policy, but clinical notes of the treating orthopaedic surgeon and the radiologist clearly documented radiographic fracture union.
Neck-shaft angle was measured with a goniometer. The degree of hip degenerative change, if present, was graded according to the criteria of Tönnis. 15 Radiographic union was defined as bridging callus on three of four cortices on anteroposterior and lateral views. Clinical union was defined as painless hip range of motion and painless full weightbearing. Functional assessment was done evaluating pain, use of gait aids, and ambulatory capacity at followup.
Nineteen of 20 nonunions healed (95%). One nonunion treated with a 95° angled blade plate and autograft did not heal and was revised to a cemented bipolar hemiarthroplasty 12 months postoperatively.
Before the salvage operation 13 of 20 patients had severe pain, and seven had moderate pain. Sixteen of 20 required two-arm support, two used a cane full-time, and two were nonambulatory. At followup, 16 of 19 (84%) patients had no pain, and three had mild pain over the trochanter which appeared to be related to the retained hardware. Seven ambulated independently with no support, eight required a cane, and four required a walker full-time. All of the patients who needed upper extremity support were older than 60 years at last followup. Twelve of 19 lived alone, four required assisted living, and three resided in nursing homes.
Of the 12 patients with available postoperative radiographs at the time of the retrospective review, none showed radiographic evidence of degenerative change at last followup (Tönnis Grade 0). 15 The mean neck-shaft angle at union was 130° (range, 102°–145°). None had failure of any internal fixation device. None had osteonecrosis of the femoral head.
There was one intraoperative complication in 20 patients (5%). A patients with preexisting familial platelet dysfunction had severe intraoperative hypotension develop necessitating rapid completion of the procedure. There were two postoperative complications in 20 patients (10%). The patient with the platelet disorder (mentioned above) subsequently had an infected hematoma and wound dehiscence that required multiple debridements and intravenous antibiotics but eventually healed. One patient had severe postoperative hyponatremia develop and was treated medically. There were no intraoperative deaths. At last followup 15 patients were alive and five were dead.
The treatment of failed internal fixation of intertrochanteric hip fractures is challenging because of fracture deformity, bone loss, bone comminution, and retained, often broken, hardware from previous operations. Patients are almost always markedly disabled and except for infrequent patients with prohibitively severe medical problems, reoperation usually is considered. Treatment options include prosthetic replacement and revision internal fixation. In younger patients, and active older patients with good remaining bone stock and a well-preserved hip, revision internal fixation and bone grafting have been advocated, but to date there has been limited information on the results, success rate, and complications of this procedure. 7,11,13,16 The current study showed that a high rate of success can be achieved with internal fixation and bone grafting for properly selected patients. The patients in the current series were a selected cohort of patients with this problem and almost all were physiologically younger patients with good remaining bone. At the authors’ institution during this same time, older patients with low demand activities and poor bone quality, or a damaged hip articular surface were (and are) treated preferentially with hip arthroplasty.
The strengths of the current study included the larger series and excellent followup. The limitations of the current study included the retrospective methodology, the selection bias discussed above, and the different fixation implants used. In eight patients the radiographs had been discarded after 10 years as a part of institutional policy and were not available for review. In all patients, the clinical notes and radiographic recordings clearly documented union but the patients were not available for radiographic review by the authors. In each patient clear evidence of fracture union was documented in the chart by the treating surgeon and the radiologist. These limitations do not undermine the conclusion that for the majority of well-selected patients treated with revision open reduction and internal fixation and bone grafting, successful salvage was obtained with few complications.
Sarathy et al 13 reported on seven patients with established nonunion of an intertrochanteric fracture treated with greater trochanteric osteotomy, valgus alignment of the head and neck fragment, medial displacement of the distal fragment, and bone grafting. A fixed angle 130° blade plate was used for fixation. Six of seven nonunions healed. Wu et al 16 reported on 14 intertrochanteric fractures with cut-out of a lag screw of a dynamic hip screw. Patients were treated with reinsertion of a lag screw inferiorly in the femoral head, cement augmentation, and valgus producing subtrochanteric osteotomy. All nonunions healed at a mean of 5 months. In their series, union of the subtrochanteric osteotomy took longer than union of the original fracture. In the current series bone in the inferior femoral head commonly was used to gain fixation, however, cement augmentation was not used. Additionally, subtrochanteric osteotomy was used in only one case. By freshening the nonunion site the proximal fragment usually can be realigned in either anatomic or slight valgus position without the need for subtrochanteric osteotomy which deforms the proximal femur, requires another site to heal, and potentially can make subsequent arthroplasty, if required, more difficult. 4 For the patients with available postoperative radiographs, the mean neck-shaft angle at union for the current series was 130°.
Mariani and Rand 11 reported on 11 patients with intertrochanteric nonunion with a mean age of 53 years treated with repeat open reduction and internal fixation at a time earlier than the current study. Nine of 11 (82%) patients achieved union at an average of 6 months. One other patient eventually achieved union with repeat bone grafting, and one subsequently required a total hip arthroplasty. Functional results were improved for all patients. Various implants were used successfully based on location of remaining bone stock in the femoral head. The current series confirms that the surgeon can choose from different fixation devices, based on the fracture pattern and the location of remaining bone stock in the femoral head, and if the principles of stable fixation, good proximal purchase, and bone grafting are adhered to, union with a good result can be achieved in the majority of cases.
Implants for revision internal fixation typically are selected based on the quality and location of remaining bone stock in the femoral head. Bone stock in the inferior femoral head often has not been violated by prior hardware used for routine internal fixation of intertrochanteric fractures. Fixed angle devices such as angled blade plates are useful in this setting (Figs 1, 2). Multiple neck-shaft angles are available that provide many options for fixation of the proximal bone fragment based on preoperative templating. In certain cases, sliding hip screws can be useful for revision fixation despite the failure of a previous sliding hip screw. Super-lag screws (Synthes) with larger thread diameters may be useful if good bone stock remains along the tract of a previously well centered lag screw. Alternatively, if the prior lag screw was misplaced being short or high in the femoral head, a new lag screw can be advanced deeper in the same tract or placed lower in the femoral head to afford proximal fragment purchase. All of the above techniques were used with successful results in the current series.
Nonunions of the intertrochanteric region of the femur typically are well vascularized and have excellent surrounding soft tissue coverage, unlike nonunions in other anatomic locations. It is difficult to determine whether bone grafting is necessary to achieve predictable union. Extensive dissection may be necessary to mobilize and align these nonunions before internal fixation, and this may, to some extent, compromise the vascularity at the nonunion site. The authors, therefore, prefer to do some form of bone grafting when salvaging these difficult nonunions.
Revision internal fixation and bone grafting of failed internal fixation of intertrochanteric hip fractures in selected, mostly younger patients with good remaining bone stock and a well-preserved hip provided high union rates and functional improvement in the majority of patients with a low rate of associated complications.
1. Baker HR: Ununited intertrochanteric fractures of the femur. Clin Orthop 18:209–220, 1960.
2. Baumgartner MR, Solberg BD: Awareness of tip-apex distance reduces failure of fixation of trochanteric fractures of the hip. J Bone Joint Surg 79B:969–971, 1997.
3. Boyd HB, Lipinski SW: Nonunion of trochanteric and subtrochanteric fractures. Surg Gynecol Obstet 104:463–470, 1957.
4. Ferguson GM, Cabanela ME, Ilstrup DM: Total hip arthroplasty after failed intertrochanteric osteotomy. J Bone Joint Surg 76B:252–257, 1994.
5. Haentjens P, Casteleyn PP, Opdecan P: Hip arthroplasty for failed internal fixation of intertrochanteric and subtrochanteric fractures in the elderly patient. Arch Orthop Trauma Surg 113:222–227, 1994.
6. Haidukewych GJ, Berry DJ: Hip arthroplasty for salvage of failed treatment of intertrochanteric hip fractures. In Press, J Bone Joint Surg (A).
7. Haidukewych GJ, Israel TA, Berry DJ: Reverse obliquity of fractures of the intertrochanteric region of the femur. J Bone Joint Surg 83A:643–650, 2001.
8. Kim Y-H, Oh J-H, Koh Y-G: Salvage of neglected unstable intertrochanteric fractures with cementless porous-coated hemiarthroplasty. Clin Orthop 277:182–187, 1992.
9. Knight WM, DeLee JC: Nonunion of intertrochanteric fractures of the hip: A case study and review. Orthop Trans 16:438, 1982.
10. Kyle RF, Gustilo RB, Premer RF: Analysis of 622 intertrochanteric hip fractures. J Bone Joint Surg 61A:216–221, 1979.
11. Mariani EM, Rand JA: Nonunion of intertrochanteric fractures of the femur following open reduction and internal fixation: Results of second attempts to gain union. Clin Orthop 218:81–89, 1987.
12. Mehlhoff T, Landon GC, Tullos HS: Total hip arthroplasty following failed internal fixation of hip fractures. Clin Orthop 269:32–37, 1991.
13. Sarathy MP, Madhavan P, Ravichandran KM: Nonunion of intertrochanteric fractures of the femur. J Bone Joint Surg 77B:90–92, 1994.
14. Stoffelen D, Haentjens P, Reynders P, et al: Hip arthroplasty for failed internal fixation of intertrochanteric and subtrochanteric fractures in the elderly patient. Acta Orthop Belg 60:135–139, 1994.
15. Tönnis D: Congenital Dysplasia and Dislocation of the Hip in Children and Adults. New York, Springer 113–130, 167, 370–380, 1987.
© 2003 Lippincott Williams & Wilkins, Inc.
16. Wu CC, Shih CH, Chen WJ, Tai CL: Treatment of cutout of a lag screw of a dynamic hip screw in an intertrochanteric fracture. Arch Orthop Trauma Surg 117:193–196, 1998.