- Email: [email protected]
Surgical treatment of periprosthetic femoral fractures following hip arthroplasty: Our institutional experience
Injury, Int. J. Care Injured 46 (2015) 1945–1950
Contents lists available at ScienceDirect
Injury journal homepage: www.elsevier.com/locate/injury
Surgical treatment of periprosthetic femoral fractures following hip arthroplasty: Our institutional experience Plamen Kinov a, Gershon Volpin b,c,*, Roger Sevi d, Panayot P. Tanchev a, Boris Antonov a, Geries Hakim b,c a
Department of Orthopaedics and Traumatology, University Hospital Queen Giovanna - ISUL, Medical University of Sofia, ul. Bialo more 8, 1527 Sofia, Bulgaria Department of Orthopaedic Surgery and Traumatology, EMMS Medical Center, Nazaret, Israel c The Galilee Medical Faculty Sfad, Bar Ilan University, Israel d Department of Orthopaedic Surgery and Traumatology, Hillel Yafe Medical Center, Hadera, Israel b
A R T I C L E I N F O
A B S T R A C T
Article history: Accepted 9 June 2015
Purpose: Periprosthetic fractures are one of the most serious complications after hip replacement. The aim of this retrospective study was to evaluate the clinical outcome of surgical treatment of periprosthetic femoral fractures following total hip arthroplasty using treatment algorithm of the Vancouver classification. Materials and methods: Fifty six periprosthetic femoral fractures operated on during the period December 2004–September 2013 were followed-up retrospectively. There were 40 women and 16 men with mean age at the time of surgery 64.7 years (41–88 years). The mean follow-up for the group was 5 years (range, 1–10 years). Periprosthetic fractures were classified according to the Vancouver classification. The clinical evaluation was performed with the Harris hip score, the Western Ontario and McMaster University Osteoarthritis Index (WOMAC) and Short Form 8 (SF-8). Bone healing, implant survival, pain, function and complications were recorded. Bone healing and implant stability were evaluated clinically and on plain radiographs. Results: Uneventful bone healing was achieved in 52 cases. In two fractures (one type B1, one type C) nonunion and plate failure occurred. Two cemented stems were revised for aseptic loosening 6.5 and 7 years after fracture fixation. Uncontrollable prosthesis infection and sepsis in a rheumatoid (immunocompromised) patient required disarticulation of the involved extremity. Discussion and conclusions: Periprosthetic femoral fractures are difficult to treat and require complex treatment approach according to risk assessment, fracture type, implant stability, bone stock and medical status of the patient. Using a treatment protocol of the Vancouver classification we obtained satisfactory outcome. ß 2015 Elsevier Ltd. All rights reserved.
Keywords: Periprosthetic fracture Hip arthroplasty Osteosynthesis Revision Outcome Vancouver classification
Introduction Periprosthetic fractures of the femur following total hip replacement (THR) are one of the most serious complications and represent a difficult treatment challenge. The incidence of these fractures varied between 0.5% and 2% but up to 4% following revision total hip arthroplasty [1,2]. Such fractures may occur intraoperatively or postoperatively. The treatment is based on the
* Corresponding author at: Department of Orthopaedic Surgery and Traumatology, EMMS Hospital, Nazaret, Israel. Tel.: +972 509887561; fax: +972 48370901. E-mail address: [email protected] (G. Volpin). http://dx.doi.org/10.1016/j.injury.2015.06.017 0020–1383/ß 2015 Elsevier Ltd. All rights reserved.
site of fracture, implant stability, quality of bone stock and the medical condition of the patient. Conservative treatment has shown poor results and should be reserved for elderly patients with poor medical status but with stable prostheses [3]. Surgical treatment of periprosthetic fractures of the femur following THR has higher morbidity and mortality compared to revision hip arthroplasty for aseptic loosening [4,5]. Difficult surgical treatment, high rate of complications and usually less favourable clinical outcome compared to aseptic revisions necessitate proper evaluation of the periprosthetic fracture, precise planning and implementation of the treatment plan. Various classifications for periprosthetic femoral fractures around hip arthroplasty have been proposed [6–10]. The Vancouver classification [11] is widely used as it is based on fracture
P. Kinov et al. / Injury, Int. J. Care Injured 46 (2015) 1945–1950
1946
location, implant stability and bone quality. Moreover, the classification has been validated and includes treatment algorithm [11,12]. However, decision for the type of internal fixation depends on preoperative plan, intra-operative findings, and possibility to achieve stable fixation and surgeon’s preferences as well as logistics. Successful treatment of such periprosthetic femoral fractures varies from nonoperative procedures to extensive revision surgeries [4]. The aim of this retrospective study was to (1) evaluate the clinical outcome; (2) to assess the impact of femoral periprosthetic fracture on patient reported outcome measurements; and (3) to analyse the effect on fracture healing and implant stability of when applying a modified treatment algorithm of the Vancouver classification.
Table 2 Distribution of the patients in the series according to ASA (American Society of Anesthesiologists) Physical Status Classification System.
Materials and methods
Surgical data
Fifty six consecutive periprosthetic femoral fractures operated on during the period December 2004–September 2013 were followed-up retrospectively at mean 5 years (range, 1–10 years). During the study period we treated conservatively four other periprosthetic fractures of the greater trochanter that were not included in this study. The cohort included 40 women and 16 men with mean age at the time of surgery 64.7 years (41–88 years). Fifty three of the patients were after primary arthroplasties and three following revision arthroplasties. In one case the prosthesis was bipolar. The primary diagnosis was primary osteoarthritis in 33 hips (58.9%), hip fracture in 14 hips (25%), and in the others – secondary osteoarthritis following Developmental hip dysplasia in 5 (8.9%), osteonecrosis in three (5.4%), and rheumatoid arthritis in one hip (1.8%). In 32 patients the primary THR femoral stem was cemented and in 24 uncemented. Time from insertion of the primary THR prosthesis to periprosthetic fracture was on average 6.2 years (range, 0.5– 22.2 years). In 52 cases the periprosthetic fractures were result of a fall, one fracture was sustained in a car accident, one case occurred intraoperatively during reduction of the implant at the end of the operation, and two fractures occurred intraoperatively during revision surgery. The right hip was fractured in 34 patients and the left hip in 22 patients. The periprosthetic femoral fractures in the series were classified according to the Vancouver classification [11]. Based on this classification, as presented in Table 1, Forty four patients had type B fracture – around or just distal to the stem, of them 16 with type B1 (a fracture of the femur and a stable prosthesis), 14 patients with type B2 (a fracture of the femur with a unstable prosthesis), 14 patients with type B3 (fracture of the femur with an unstable prosthesis and with poor quality of bone). Twelve patients had type C fracture (well below the distal tip of a stable prosthesis). We had four patients with type Ag fracture (greater trochanter) treated conservatively and not included in this series. Six patients were ASA grade I, 22 patients were ASA grade II, and 28 patients were grade III (Table 2). This reflected the prevalence of ischaemic heart disease (22/56 – 39.3%), diabetes (8/56 – 14.3%) and alcohol abuse (2/56 – 3.6%) in some of the patients.
Time from admission to surgery was on average 5 days (range, 1–12 days). In all cases lateral decubitus position was used. In two cases extended trochanteric osteotomy was used for implantation of a longer uncemented stem [14]. To a certain extent we deviated from the treatment algorithm of the Vancouver classification as some oblique Vancouver type B and C fractures were treated by plates or by cerclages. Periprosthetic fractures were operated on as follows: in 16 patients with type B1 – 12 were treated by locking plate (Fig. 1), and 4 by cerclage wires. All 14 patients with Type B2 were treated by revision of the stem with a longer revision femoral stem (uncemented in12 and cemented in 2) and 3 by cerclage wires. All 14 patients with Type B3 were treated by a longer revision femoral stem (uncemented in 12 and cemented in 2). In 12 patients with Type C - 9 were treated by locking plates and 3 by cerclage wires. Additional cortical strut grafts were used in two patients (one with Type B1 and one with Type B3) for reconstruction of the proximal femur together with plate reinforcement. In 30 cases wound drainage was used for 24 h. Third generation cephalosporin was used for antibiotic prophylaxis for five days and for thromboembolic prophylaxis we used low molecular weight heparin for 45 days. In the postoperative period partial weight-bearing (tip-toeing) was allowed at discretion of the surgeon and progressive partial weight-bearing was allowed after clinical visit on the second postoperative month provided radiographical signs of callous formation were present. Full weight-bearing was allowed after radiographical signs for fracture healing. Bone healing, subsidence and implant stability were assessed on plain radiographs. Bone union was diagnosed clinically when patient was able to full weight-bear without pain. The bone implant interface and bone remodeling were assessed by radiographs. Implant survival, pain, function, early and late complications, rehabilitation after discharge, and length of hospital stay were recorded. White blood count, CRP levels and ESR were evaluated in suspicion of infection. The clinical evaluation was performed using the Harris Hip Score [15]. General and disease-specific health-related quality of life (HRQOL) was measured by the patient using two types of instruments. The disease-specific HRQOL was assessed with the Western Ontario and McMaster University Osteoarthritis Index (WOMAC) [16], a 24-item questionnaire that quantifies hip disease in three domains. The general HRQOL was assessed with the Medical Outcome Study 8-Items Short Form Health Survey (SF-8) [17]. The SF-8 generates two summary measures, a physical component summary (PCS) and a mental component summary (MCS). The health-status questionnaires scores were matched with scores found after primary and revision THR (16). The radiological outcome was graded as excellent, good or poor as previously described [18]. An excellent clinical outcome was registered when
Table 1 Periprosthetic fractures in the series. Vancouver type
Number of patients
Ag B1 B2 B3 C
4 (treated conservatively) 16 14 14 12
ASA grade I (a normal healthy patient) II (a patient with mild systemic disease) III (a patient with severe systemic disease)
Number of patients 6 22 28
Preoperative ASA health status of the patients did not correlate with the type of fixation or the type of fracture.
P. Kinov et al. / Injury, Int. J. Care Injured 46 (2015) 1945–1950
1947
Fig. 1. M.S., 80-year-old female. (A) Revision hip replacement for aseptic loosening performed elsewhere. (B) A complex case of a periprosthetic femoral fracture with a stable cemented stem (Vancouver Type B1) and acetabular loosening 7 years after revision. (C) First postoperative radiograph. Exchange of the cemented stem, stabilisation of the fracture with a locking plate, cortical strut graft and cerclage wires. (D) Control radiograph 9 months after surgery. Good clinical outcome, patient walks with a cane.
the hip was pain free and the stem was stable, with minimal deformity and no shortening. Stable subsidence with minimal deformity or shortening was considered a good result. Loosening, nonunion, sepsis, severe deformity or shortening was considered as a poor result. Results In 52 patients (92.8%) bone healing occurred at mean 4.1 months (range, 3–6 months) after surgery. In the seven patients treated with cerclage wires the periprosthetic fractures (four type B1 and three type C) united at mean 4.4 months (range, 4–5 months) after fixation, whereas in the remaining 45 patients fracture united at mean 3.9 months (range, 3–6 months) (p < 0.05). There were two patients with nonunion of the periprosthetic fractures of the femur – in a patient with type B1 fracture the
cemented stem subsided as the fracture did not unite and the plate broke (Fig. 2). The fracture healed after revision with a long uncemented stem. In the second patient with type C fracture the plate broke and malunion with shortening and angulation of the extremity resulted but the patient refused further treatment. Two stems in a type B1 fractures were revised for aseptic loosening 7 and 6.5 years after uneventful fracture union. In a patient with a type B1 fracture 5 mm stem subsidence was observed at 6 years after fracture surgery. Wound haematoma was evacuated on the eighth postoperative day in one of the patients and the surgical wound healed uneventfully. Superficial draining sinus occurred in an 80-year-old female treated with stem exchange, plating and massive allograft. After administration for one month of ciprofloxacin and tubocin it resolved. Another superficial wound infection in a morbidly obese female patient (BMI > 40) caused by MLSE positive Staph. epidermidis was debrided and treated successfully
Fig. 2. V.V., 60-year-old female. (A) Periprosthetic fracture with a stable cemented stem (Vancouver type B1) in a paretic extremity. (B) Stabilisation of the fracture with a locking plate and cerclage wires. Inadequate fracture reduction. (C) Broken plate and nonunion 7 months after surgery. (D) Fracture union 3 months after revision with long uncemented stem and cerclage wires.
1948
P. Kinov et al. / Injury, Int. J. Care Injured 46 (2015) 1945–1950
with prolonged use of antibiotics (2 months) according to microorganism susceptibility. One male with alcohol abuse who had an uneventful bone healing of periprosthetic type B2 fracture treated with long cemented stem and plate, had an additional new distal periprosthetic fracture type B1 after a fall 20 weeks postoperatively (Fig. 3). The plate was exchanged by a longer distal one and the new fracture united uneventfully. Six months later he sustained a patellar fracture that healed uneventfully after surgical treatment. However, after eight months another fracture of the lateral tibial plateau occurred and was treated operatively. One year after the last incident the patient was pain-free and clinical outcome was excellent. Uncontrollable periprosthetic infection and sepsis caused by MRSA positive Staph. aureus in one immunocompromised rheumatoid female patient occurred after uneventful bone healing 19 months postoperatively and required disarticulation of the involved extremity.
Forty four (78.6%) of our patients had excellent or good clinical outcome: the average HHS was 81.3 points (range, 52–91 points), and the average WOMAC score was 73.7 points (range, 50–86 points). Both components of the generic HRQOL measurement instrument SF-8 showed satisfactory scores of the patients in the cohort: physical component summary score was 41.3 (range, 33.1– 56.7 points), mental component summary score was 44.1 (range, 30.4–62.2 points). The preoperative ASA grade, type of primary stem fixation, type of fracture, type of fracture fixation and the clinical outcome measurements did not correlate significantly. However, advanced age correlated significantly with lower physical scores measured by SF-8 (p < 0.05). We observed infections only in 2 female patients (p < 0.05). Discussion Periprosthetic femoral fractures are a serious complication with increasing incidence. Open reduction and internal fixation is a
Fig. 3. G.N., 61-year-old male. (A) Periprosthetic fracture with an unstable cemented stem (Vancouver type B2). (B) Fracture reduction and stabilisation with plate and cerclage wires, long cemented stem exchange. (C) Four months later patients sustained a second traumatic fracture below the tip of the stem (Vancouver type B1). (D) The second fracture was stabilised with a long plate. (E) The patient sustained a third patellar fracture after six months stabilized with cerclage wires. (F) Eight months later the patient was treated for lateral tibial plateau fracture.
P. Kinov et al. / Injury, Int. J. Care Injured 46 (2015) 1945–1950
treatment of choice. Various implants such as plates, cables, cerclage wires, long cemented and uncemented femoral stems, as well as cortical strut grafts have been proposed. Nevertheless, these fractures are difficult to treat and the results are not always predictable. The principles for treatment of periprosthetic fractures are outlined in the Vancouver classification. According to the algorithm proposed by Masri et al. [13] the treatment of type B1 fractures is by internal fixation with cerclage wires, plates or cortical strut grafts. Recently, the advantages of cables in treatment of periprosthetic fractures have been demonstrated [19]. In the cohort under study, using modified approach to periprosthetic fractures, we obtained healing in all the type B1 fractures. We support the conclusion of Park et al. [20] that predictable bone union in type B1 fractures could be achieved with cerclage wires. Careful operative technique with preservation of periosteal blood supply will favour better results. However, two stems were revised for subsidence 7 and 6.5 years after the operation and another stem showed signs of loosening (5 mm subsidence, slight intermittent thigh pain) 6 years after the surgery. In the two femoral revisions uncemented distal fixation was used with successful outcome and the third patient was regularly followed-up for signs of progressive loosening. The stability of the femoral stem could not always be assessed on the preoperative radiographs. In some cases CT scans might also be used. Finally, exploration of the joint during surgery and testing stability of the stem could differentiate between type B1 and type B2 fractures. Distinguishing of type B2 fractures from type B1 fractures is a major challenge in treating periprosthetic fractures. Absence of basis for validation of the stability of the stem is considered a significant drawback of the Vancouver classification [20]. Proximal femoral fractures are more likely to involve loosening of the stem and require stem exchange [21]. Although considered stable, the stem might be exchanged to long stem in unfavourable biomechanical situation. Anatomic reduction of the fracture is mandatory in order to obtain long lasting result. In one of our cases with implant subsidence suboptimal fracture reduction might have added to questionable implant stability. According to the algorithm proposed by Masri et al. [13] type B2 and type B3 periprosthetic femoral fractures require revision with a longer femoral stem. One of the two type B2 fractures in this series treated with longer cemented stem needed plate reinforcement as the stem could not provide sufficient distal stability (Fig. 3). Uncemented tapered femoral stems are considered as a reliable treatment method for Vancouver B2 and B3 periprosthetic femoral fractures [22]. 14 patients with type B3 fractures in the series were revised successfully with medullary stem (12 with uncemented stem and 2 with cemented stem) with distal fixation and the deficient proximal femur was reconstructed with cerclage wires or plates. We suggest that selected type B3 fractures could be treated with uncemented longer stems and without cortical strut grafts. Three out of five type C fractures in our series were treated with cerclage wires and healed at about 18 weeks after surgery. However, one patient with type C fracture treated with locking plate developed malunion after plate failure. This noncompliant patient was lost to follow-up and presented broken plate and malunated fracture 2.5 years later and refused further treatment. Compliance to postoperative regiment and rehabilitation protocol is mandatory for obtaining favourable outcome in these serious injuries. Cortical strut allograft may have added to the construct stability [23,24]. Provided the endosteal blood supply is damaged by the intramedullary device and by the fracture itself cortical strut allografts may be routinely used as proposed by Haddad et al. [23]. However, we suggest that precise operative technique with preservation of periosteal blood is of paramount importance. Fracture fixation with a stiff locking plate provides rigid fixation
1949
and reduces fragment mobility. This inhibits callus formation and results in nonunion as observed in two of our cases with stable implants and relatively good bone quality. This suggests that locking plate should be applied as a bridging plate to allow interfragmentary motion and to promote fracture healing. The slightly longer time (two weeks) to fracture union in cases treated with cerclage wires was explained by the concerns of the surgeon when allowing unprotected weight-bearing in these cases. The goal of treatment of periprosthetic fractures is fracture union and stability to allow early mobilisation. The use of cerclage wires for fracture stabilisation may restrict early mobilisation and may cause delayed bone healing. Instead of cerclage wires, cables in combination with plates or cortical strut grafts were recommended [19]. In our study, we obtained fracture union in all cases treated with cerclage wires. However, this approach led to slower rehabilitation and delayed fracture union by two weeks. In this study, we assessed treatment of periprosthetic femoral fractures not only regarding radiological and clinical outcome, but also in respect to the patient-reported HRQOL measurements. We used quality of life measurement to assess the overall impact of periprosthetic fracture surgery on health [25]. The assessment with the disease specific instrument WOMAC showed satisfactory scores that compared favourably with results from a previous study (16). Patient general health measurement with SF-8 showed slightly lower scores in the physical component summary compared to the mental component summary. Due to advanced age and coexisting comorbidities of some of the patients in the series the clinical outcome was lower compared to the outcome after primary THR or revision for aseptic loosening [16]. We feel that HRQOL measurement added important information allowing direct comparison with other studies. In conclusion, we presented our results after treatment of periprosthetic femoral fractures according to fracture pattern, stability of the prosthesis, quality of bone stock and patient’s medical condition. Next to appropriate surgical technique, regular follow-ups and compliance to postoperative regiment and rehabilitation protocol are a prerequisite for obtaining favourable outcome. Periprosthetic femoral fractures remain a major complication following THR but are not often presented even to a busy arthroplasty centre. Long-term follow-up of a larger cohort will be necessary to establish whether the use of modification to the Vancouver treatment algorithm can favourably affect implant survival and clinical outcome in the long-term. Conflict of interest All authors declare that they do not have any financial or personal relationships with other people or organisations that could inappropriately influence (bias) their work. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. References [1] Lewallen DG, Berry DJ. Periprosthetic fracture of the femur after total hip arthroplasty. J Bone Joint Surg Am 1997;79:1881–90. [2] Sarvilinna R, Huhtala HS, Puolakka TJ, Nevalainen JK, Pajama¨ki KJ. Periprosthetic fractures in total hip arthroplasty: an epidemiologic study. Int Orthop 2003;27(6):359–61. [3] Van der Wal BC, Vischjager M, Grimm B, Heyligers IC, Tonino AJ. Periprosthetic fractures around cementless hydroxyapatite-coated femoral stems. Int Orthop 2005;29(4):235–40. [4] Lindahl H, Oden A, Garellick G, Malchau H. The excess mortality due to periprosthetic femur fracture. A study from the Swedish national hip arthroplasty register. Bone 2007;40:1294–8. [5] Young SW, Walker CG, Pitto RP. Functional outcome of femoral periprosthetic fracture and revision hip arthroplasty: a matched pair study from the New Zealand registry. Acta Orthop 2008;79:483–8. [6] Johansson JE, McBroom R, Barrington TW. Fracture of the ipsilateral femur in patients with total hip replacement. J Bone Joint Surg Am 1981;63:1435–42.
1950
P. Kinov et al. / Injury, Int. J. Care Injured 46 (2015) 1945–1950
[7] Kelley SS. Periprosthetic femoral fractures. J Am Acad Orthop Surg 1994;2(3):164–72. [8] Mont MA, Maar DC. Fractures of the ipsilateral femur after hip arthroplasty: a statistical analysis of outcome based on 487 patients. J Arthroplasty 1994;9:511–9. [9] Morrey BF, Kavanagh BF. Complications with revision of the femoral component of total hip arthroplasty: comparison between cemented and uncemented techniques. J Arthroplasty 1992;7:71–9. [10] Whittaker RP, Sotos LN, Ralston EL. Fractures of the femur about femoral endoprostheses. J Trauma 1974;14:675–94. [11] Brady OH, Garbuz DS, Masri BA, Duncan CP. Classification of the hip. Orthop Clin North Am 1999;30(2):215–20. [12] Brady OH, Garbuz DS, Masri BA, Duncan CP. The reliability and validity of the Vancouver classification of femoral fractures after hip replacement. J Arthroplasty 2000;15(1):59–62. [13] Masri BA, Meek RM, Duncan CP. Periprosthetic fractures evaluation and treatment. Clin Orthop Relat Res 2004;420:80–95. [14] Lakstein D1, Kosashvili Y, Backstein D, Safir O, Lee P, Gross AE. The long modified extended sliding trochanteric osteotomy. Int Orthop 2011;35(1):13–7. [15] Harris WH. Traumatic arthritis of the hip after dislocation and acetabular fractures: treatment by mold arthroplasty. An end-result study using a new method of result evaluation. J Bone Joint Surg Am 1969;51:737–55. [16] Kinov P, Tivchev P. Evaluation of Bulgarian version of WOMAC in patients with hip pathology. In: Proceedings of the 10th National Congress of Bulgarian Orthopedic and Traumatologic Association, Borovets; 2007. p. 221–7.
[17] Ware JE, Kosinski M, Dewey JE, Gandek B. How to score and interpret singleitem health status measures: a manual for users of the SF-8 health survey. Boston: QualityMetric Incorporated; 2001. [18] Mukundan C, Rayan F, Kheir E, Macdonald D. Management of late periprosthetic femur fractures: a retrospective cohort of 72 patients. Int Orthop 2010;34(4):485–9. [19] Giannoudis PV, Kanakaris NK, Tsiridis E. Principles of internal fixation and selection of implants for periprosthetic femoral fractures. Injury 2007;38(6): 669–87. [20] Park SK, Kim YG, Kim SY. Treatment of periprosthetic femoral fractures in hip arthroplasty. Clin Orthop Surg 2011;3(2):101–6. [21] Harris B, Owen JR, Wayne JS, Jiranek WA. Does femoral component loosening predispose to femoral fracture?: an in vitro comparison of cemented hips. Clin Orthop Relat Res 2010;468:497–503. [22] Abdel MP, Lewallen DG, Berry DJ. Periprosthetic femur fractures treated with modular fluted, tapered stems. Clin Orthop Relat Res 2014;472(2):599–603. [23] Haddad FS, Duncan CP, Berry DJ, Lewallen DG, Gross AE, Chandler HP. Periprosthetic femoral fractures around well-fixed implants: use of cortical onlay allografts with or without a plate. J Bone Joint Surg Am 2002;84:945–50. [24] Barden B, von Knoch m, Fitzek JG, Loer F. Periprosthetic fractures with extensive bone loss treated with onlay strut allografts. Int Orhtop 2003;27(3):164–7. [25] Singh J, Sloan JA, Johanson NA. Challenges with health-related quality of life assessment in arthroplasty patients: problems and solutions. J Am Acad Orthop Surg 2010;18(2):72–82.
Contents lists available at ScienceDirect
Injury journal homepage: www.elsevier.com/locate/injury
Surgical treatment of periprosthetic femoral fractures following hip arthroplasty: Our institutional experience Plamen Kinov a, Gershon Volpin b,c,*, Roger Sevi d, Panayot P. Tanchev a, Boris Antonov a, Geries Hakim b,c a
Department of Orthopaedics and Traumatology, University Hospital Queen Giovanna - ISUL, Medical University of Sofia, ul. Bialo more 8, 1527 Sofia, Bulgaria Department of Orthopaedic Surgery and Traumatology, EMMS Medical Center, Nazaret, Israel c The Galilee Medical Faculty Sfad, Bar Ilan University, Israel d Department of Orthopaedic Surgery and Traumatology, Hillel Yafe Medical Center, Hadera, Israel b
A R T I C L E I N F O
A B S T R A C T
Article history: Accepted 9 June 2015
Purpose: Periprosthetic fractures are one of the most serious complications after hip replacement. The aim of this retrospective study was to evaluate the clinical outcome of surgical treatment of periprosthetic femoral fractures following total hip arthroplasty using treatment algorithm of the Vancouver classification. Materials and methods: Fifty six periprosthetic femoral fractures operated on during the period December 2004–September 2013 were followed-up retrospectively. There were 40 women and 16 men with mean age at the time of surgery 64.7 years (41–88 years). The mean follow-up for the group was 5 years (range, 1–10 years). Periprosthetic fractures were classified according to the Vancouver classification. The clinical evaluation was performed with the Harris hip score, the Western Ontario and McMaster University Osteoarthritis Index (WOMAC) and Short Form 8 (SF-8). Bone healing, implant survival, pain, function and complications were recorded. Bone healing and implant stability were evaluated clinically and on plain radiographs. Results: Uneventful bone healing was achieved in 52 cases. In two fractures (one type B1, one type C) nonunion and plate failure occurred. Two cemented stems were revised for aseptic loosening 6.5 and 7 years after fracture fixation. Uncontrollable prosthesis infection and sepsis in a rheumatoid (immunocompromised) patient required disarticulation of the involved extremity. Discussion and conclusions: Periprosthetic femoral fractures are difficult to treat and require complex treatment approach according to risk assessment, fracture type, implant stability, bone stock and medical status of the patient. Using a treatment protocol of the Vancouver classification we obtained satisfactory outcome. ß 2015 Elsevier Ltd. All rights reserved.
Keywords: Periprosthetic fracture Hip arthroplasty Osteosynthesis Revision Outcome Vancouver classification
Introduction Periprosthetic fractures of the femur following total hip replacement (THR) are one of the most serious complications and represent a difficult treatment challenge. The incidence of these fractures varied between 0.5% and 2% but up to 4% following revision total hip arthroplasty [1,2]. Such fractures may occur intraoperatively or postoperatively. The treatment is based on the
* Corresponding author at: Department of Orthopaedic Surgery and Traumatology, EMMS Hospital, Nazaret, Israel. Tel.: +972 509887561; fax: +972 48370901. E-mail address: [email protected] (G. Volpin). http://dx.doi.org/10.1016/j.injury.2015.06.017 0020–1383/ß 2015 Elsevier Ltd. All rights reserved.
site of fracture, implant stability, quality of bone stock and the medical condition of the patient. Conservative treatment has shown poor results and should be reserved for elderly patients with poor medical status but with stable prostheses [3]. Surgical treatment of periprosthetic fractures of the femur following THR has higher morbidity and mortality compared to revision hip arthroplasty for aseptic loosening [4,5]. Difficult surgical treatment, high rate of complications and usually less favourable clinical outcome compared to aseptic revisions necessitate proper evaluation of the periprosthetic fracture, precise planning and implementation of the treatment plan. Various classifications for periprosthetic femoral fractures around hip arthroplasty have been proposed [6–10]. The Vancouver classification [11] is widely used as it is based on fracture
P. Kinov et al. / Injury, Int. J. Care Injured 46 (2015) 1945–1950
1946
location, implant stability and bone quality. Moreover, the classification has been validated and includes treatment algorithm [11,12]. However, decision for the type of internal fixation depends on preoperative plan, intra-operative findings, and possibility to achieve stable fixation and surgeon’s preferences as well as logistics. Successful treatment of such periprosthetic femoral fractures varies from nonoperative procedures to extensive revision surgeries [4]. The aim of this retrospective study was to (1) evaluate the clinical outcome; (2) to assess the impact of femoral periprosthetic fracture on patient reported outcome measurements; and (3) to analyse the effect on fracture healing and implant stability of when applying a modified treatment algorithm of the Vancouver classification.
Table 2 Distribution of the patients in the series according to ASA (American Society of Anesthesiologists) Physical Status Classification System.
Materials and methods
Surgical data
Fifty six consecutive periprosthetic femoral fractures operated on during the period December 2004–September 2013 were followed-up retrospectively at mean 5 years (range, 1–10 years). During the study period we treated conservatively four other periprosthetic fractures of the greater trochanter that were not included in this study. The cohort included 40 women and 16 men with mean age at the time of surgery 64.7 years (41–88 years). Fifty three of the patients were after primary arthroplasties and three following revision arthroplasties. In one case the prosthesis was bipolar. The primary diagnosis was primary osteoarthritis in 33 hips (58.9%), hip fracture in 14 hips (25%), and in the others – secondary osteoarthritis following Developmental hip dysplasia in 5 (8.9%), osteonecrosis in three (5.4%), and rheumatoid arthritis in one hip (1.8%). In 32 patients the primary THR femoral stem was cemented and in 24 uncemented. Time from insertion of the primary THR prosthesis to periprosthetic fracture was on average 6.2 years (range, 0.5– 22.2 years). In 52 cases the periprosthetic fractures were result of a fall, one fracture was sustained in a car accident, one case occurred intraoperatively during reduction of the implant at the end of the operation, and two fractures occurred intraoperatively during revision surgery. The right hip was fractured in 34 patients and the left hip in 22 patients. The periprosthetic femoral fractures in the series were classified according to the Vancouver classification [11]. Based on this classification, as presented in Table 1, Forty four patients had type B fracture – around or just distal to the stem, of them 16 with type B1 (a fracture of the femur and a stable prosthesis), 14 patients with type B2 (a fracture of the femur with a unstable prosthesis), 14 patients with type B3 (fracture of the femur with an unstable prosthesis and with poor quality of bone). Twelve patients had type C fracture (well below the distal tip of a stable prosthesis). We had four patients with type Ag fracture (greater trochanter) treated conservatively and not included in this series. Six patients were ASA grade I, 22 patients were ASA grade II, and 28 patients were grade III (Table 2). This reflected the prevalence of ischaemic heart disease (22/56 – 39.3%), diabetes (8/56 – 14.3%) and alcohol abuse (2/56 – 3.6%) in some of the patients.
Time from admission to surgery was on average 5 days (range, 1–12 days). In all cases lateral decubitus position was used. In two cases extended trochanteric osteotomy was used for implantation of a longer uncemented stem [14]. To a certain extent we deviated from the treatment algorithm of the Vancouver classification as some oblique Vancouver type B and C fractures were treated by plates or by cerclages. Periprosthetic fractures were operated on as follows: in 16 patients with type B1 – 12 were treated by locking plate (Fig. 1), and 4 by cerclage wires. All 14 patients with Type B2 were treated by revision of the stem with a longer revision femoral stem (uncemented in12 and cemented in 2) and 3 by cerclage wires. All 14 patients with Type B3 were treated by a longer revision femoral stem (uncemented in 12 and cemented in 2). In 12 patients with Type C - 9 were treated by locking plates and 3 by cerclage wires. Additional cortical strut grafts were used in two patients (one with Type B1 and one with Type B3) for reconstruction of the proximal femur together with plate reinforcement. In 30 cases wound drainage was used for 24 h. Third generation cephalosporin was used for antibiotic prophylaxis for five days and for thromboembolic prophylaxis we used low molecular weight heparin for 45 days. In the postoperative period partial weight-bearing (tip-toeing) was allowed at discretion of the surgeon and progressive partial weight-bearing was allowed after clinical visit on the second postoperative month provided radiographical signs of callous formation were present. Full weight-bearing was allowed after radiographical signs for fracture healing. Bone healing, subsidence and implant stability were assessed on plain radiographs. Bone union was diagnosed clinically when patient was able to full weight-bear without pain. The bone implant interface and bone remodeling were assessed by radiographs. Implant survival, pain, function, early and late complications, rehabilitation after discharge, and length of hospital stay were recorded. White blood count, CRP levels and ESR were evaluated in suspicion of infection. The clinical evaluation was performed using the Harris Hip Score [15]. General and disease-specific health-related quality of life (HRQOL) was measured by the patient using two types of instruments. The disease-specific HRQOL was assessed with the Western Ontario and McMaster University Osteoarthritis Index (WOMAC) [16], a 24-item questionnaire that quantifies hip disease in three domains. The general HRQOL was assessed with the Medical Outcome Study 8-Items Short Form Health Survey (SF-8) [17]. The SF-8 generates two summary measures, a physical component summary (PCS) and a mental component summary (MCS). The health-status questionnaires scores were matched with scores found after primary and revision THR (16). The radiological outcome was graded as excellent, good or poor as previously described [18]. An excellent clinical outcome was registered when
Table 1 Periprosthetic fractures in the series. Vancouver type
Number of patients
Ag B1 B2 B3 C
4 (treated conservatively) 16 14 14 12
ASA grade I (a normal healthy patient) II (a patient with mild systemic disease) III (a patient with severe systemic disease)
Number of patients 6 22 28
Preoperative ASA health status of the patients did not correlate with the type of fixation or the type of fracture.
P. Kinov et al. / Injury, Int. J. Care Injured 46 (2015) 1945–1950
1947
Fig. 1. M.S., 80-year-old female. (A) Revision hip replacement for aseptic loosening performed elsewhere. (B) A complex case of a periprosthetic femoral fracture with a stable cemented stem (Vancouver Type B1) and acetabular loosening 7 years after revision. (C) First postoperative radiograph. Exchange of the cemented stem, stabilisation of the fracture with a locking plate, cortical strut graft and cerclage wires. (D) Control radiograph 9 months after surgery. Good clinical outcome, patient walks with a cane.
the hip was pain free and the stem was stable, with minimal deformity and no shortening. Stable subsidence with minimal deformity or shortening was considered a good result. Loosening, nonunion, sepsis, severe deformity or shortening was considered as a poor result. Results In 52 patients (92.8%) bone healing occurred at mean 4.1 months (range, 3–6 months) after surgery. In the seven patients treated with cerclage wires the periprosthetic fractures (four type B1 and three type C) united at mean 4.4 months (range, 4–5 months) after fixation, whereas in the remaining 45 patients fracture united at mean 3.9 months (range, 3–6 months) (p < 0.05). There were two patients with nonunion of the periprosthetic fractures of the femur – in a patient with type B1 fracture the
cemented stem subsided as the fracture did not unite and the plate broke (Fig. 2). The fracture healed after revision with a long uncemented stem. In the second patient with type C fracture the plate broke and malunion with shortening and angulation of the extremity resulted but the patient refused further treatment. Two stems in a type B1 fractures were revised for aseptic loosening 7 and 6.5 years after uneventful fracture union. In a patient with a type B1 fracture 5 mm stem subsidence was observed at 6 years after fracture surgery. Wound haematoma was evacuated on the eighth postoperative day in one of the patients and the surgical wound healed uneventfully. Superficial draining sinus occurred in an 80-year-old female treated with stem exchange, plating and massive allograft. After administration for one month of ciprofloxacin and tubocin it resolved. Another superficial wound infection in a morbidly obese female patient (BMI > 40) caused by MLSE positive Staph. epidermidis was debrided and treated successfully
Fig. 2. V.V., 60-year-old female. (A) Periprosthetic fracture with a stable cemented stem (Vancouver type B1) in a paretic extremity. (B) Stabilisation of the fracture with a locking plate and cerclage wires. Inadequate fracture reduction. (C) Broken plate and nonunion 7 months after surgery. (D) Fracture union 3 months after revision with long uncemented stem and cerclage wires.
1948
P. Kinov et al. / Injury, Int. J. Care Injured 46 (2015) 1945–1950
with prolonged use of antibiotics (2 months) according to microorganism susceptibility. One male with alcohol abuse who had an uneventful bone healing of periprosthetic type B2 fracture treated with long cemented stem and plate, had an additional new distal periprosthetic fracture type B1 after a fall 20 weeks postoperatively (Fig. 3). The plate was exchanged by a longer distal one and the new fracture united uneventfully. Six months later he sustained a patellar fracture that healed uneventfully after surgical treatment. However, after eight months another fracture of the lateral tibial plateau occurred and was treated operatively. One year after the last incident the patient was pain-free and clinical outcome was excellent. Uncontrollable periprosthetic infection and sepsis caused by MRSA positive Staph. aureus in one immunocompromised rheumatoid female patient occurred after uneventful bone healing 19 months postoperatively and required disarticulation of the involved extremity.
Forty four (78.6%) of our patients had excellent or good clinical outcome: the average HHS was 81.3 points (range, 52–91 points), and the average WOMAC score was 73.7 points (range, 50–86 points). Both components of the generic HRQOL measurement instrument SF-8 showed satisfactory scores of the patients in the cohort: physical component summary score was 41.3 (range, 33.1– 56.7 points), mental component summary score was 44.1 (range, 30.4–62.2 points). The preoperative ASA grade, type of primary stem fixation, type of fracture, type of fracture fixation and the clinical outcome measurements did not correlate significantly. However, advanced age correlated significantly with lower physical scores measured by SF-8 (p < 0.05). We observed infections only in 2 female patients (p < 0.05). Discussion Periprosthetic femoral fractures are a serious complication with increasing incidence. Open reduction and internal fixation is a
Fig. 3. G.N., 61-year-old male. (A) Periprosthetic fracture with an unstable cemented stem (Vancouver type B2). (B) Fracture reduction and stabilisation with plate and cerclage wires, long cemented stem exchange. (C) Four months later patients sustained a second traumatic fracture below the tip of the stem (Vancouver type B1). (D) The second fracture was stabilised with a long plate. (E) The patient sustained a third patellar fracture after six months stabilized with cerclage wires. (F) Eight months later the patient was treated for lateral tibial plateau fracture.
P. Kinov et al. / Injury, Int. J. Care Injured 46 (2015) 1945–1950
treatment of choice. Various implants such as plates, cables, cerclage wires, long cemented and uncemented femoral stems, as well as cortical strut grafts have been proposed. Nevertheless, these fractures are difficult to treat and the results are not always predictable. The principles for treatment of periprosthetic fractures are outlined in the Vancouver classification. According to the algorithm proposed by Masri et al. [13] the treatment of type B1 fractures is by internal fixation with cerclage wires, plates or cortical strut grafts. Recently, the advantages of cables in treatment of periprosthetic fractures have been demonstrated [19]. In the cohort under study, using modified approach to periprosthetic fractures, we obtained healing in all the type B1 fractures. We support the conclusion of Park et al. [20] that predictable bone union in type B1 fractures could be achieved with cerclage wires. Careful operative technique with preservation of periosteal blood supply will favour better results. However, two stems were revised for subsidence 7 and 6.5 years after the operation and another stem showed signs of loosening (5 mm subsidence, slight intermittent thigh pain) 6 years after the surgery. In the two femoral revisions uncemented distal fixation was used with successful outcome and the third patient was regularly followed-up for signs of progressive loosening. The stability of the femoral stem could not always be assessed on the preoperative radiographs. In some cases CT scans might also be used. Finally, exploration of the joint during surgery and testing stability of the stem could differentiate between type B1 and type B2 fractures. Distinguishing of type B2 fractures from type B1 fractures is a major challenge in treating periprosthetic fractures. Absence of basis for validation of the stability of the stem is considered a significant drawback of the Vancouver classification [20]. Proximal femoral fractures are more likely to involve loosening of the stem and require stem exchange [21]. Although considered stable, the stem might be exchanged to long stem in unfavourable biomechanical situation. Anatomic reduction of the fracture is mandatory in order to obtain long lasting result. In one of our cases with implant subsidence suboptimal fracture reduction might have added to questionable implant stability. According to the algorithm proposed by Masri et al. [13] type B2 and type B3 periprosthetic femoral fractures require revision with a longer femoral stem. One of the two type B2 fractures in this series treated with longer cemented stem needed plate reinforcement as the stem could not provide sufficient distal stability (Fig. 3). Uncemented tapered femoral stems are considered as a reliable treatment method for Vancouver B2 and B3 periprosthetic femoral fractures [22]. 14 patients with type B3 fractures in the series were revised successfully with medullary stem (12 with uncemented stem and 2 with cemented stem) with distal fixation and the deficient proximal femur was reconstructed with cerclage wires or plates. We suggest that selected type B3 fractures could be treated with uncemented longer stems and without cortical strut grafts. Three out of five type C fractures in our series were treated with cerclage wires and healed at about 18 weeks after surgery. However, one patient with type C fracture treated with locking plate developed malunion after plate failure. This noncompliant patient was lost to follow-up and presented broken plate and malunated fracture 2.5 years later and refused further treatment. Compliance to postoperative regiment and rehabilitation protocol is mandatory for obtaining favourable outcome in these serious injuries. Cortical strut allograft may have added to the construct stability [23,24]. Provided the endosteal blood supply is damaged by the intramedullary device and by the fracture itself cortical strut allografts may be routinely used as proposed by Haddad et al. [23]. However, we suggest that precise operative technique with preservation of periosteal blood is of paramount importance. Fracture fixation with a stiff locking plate provides rigid fixation
1949
and reduces fragment mobility. This inhibits callus formation and results in nonunion as observed in two of our cases with stable implants and relatively good bone quality. This suggests that locking plate should be applied as a bridging plate to allow interfragmentary motion and to promote fracture healing. The slightly longer time (two weeks) to fracture union in cases treated with cerclage wires was explained by the concerns of the surgeon when allowing unprotected weight-bearing in these cases. The goal of treatment of periprosthetic fractures is fracture union and stability to allow early mobilisation. The use of cerclage wires for fracture stabilisation may restrict early mobilisation and may cause delayed bone healing. Instead of cerclage wires, cables in combination with plates or cortical strut grafts were recommended [19]. In our study, we obtained fracture union in all cases treated with cerclage wires. However, this approach led to slower rehabilitation and delayed fracture union by two weeks. In this study, we assessed treatment of periprosthetic femoral fractures not only regarding radiological and clinical outcome, but also in respect to the patient-reported HRQOL measurements. We used quality of life measurement to assess the overall impact of periprosthetic fracture surgery on health [25]. The assessment with the disease specific instrument WOMAC showed satisfactory scores that compared favourably with results from a previous study (16). Patient general health measurement with SF-8 showed slightly lower scores in the physical component summary compared to the mental component summary. Due to advanced age and coexisting comorbidities of some of the patients in the series the clinical outcome was lower compared to the outcome after primary THR or revision for aseptic loosening [16]. We feel that HRQOL measurement added important information allowing direct comparison with other studies. In conclusion, we presented our results after treatment of periprosthetic femoral fractures according to fracture pattern, stability of the prosthesis, quality of bone stock and patient’s medical condition. Next to appropriate surgical technique, regular follow-ups and compliance to postoperative regiment and rehabilitation protocol are a prerequisite for obtaining favourable outcome. Periprosthetic femoral fractures remain a major complication following THR but are not often presented even to a busy arthroplasty centre. Long-term follow-up of a larger cohort will be necessary to establish whether the use of modification to the Vancouver treatment algorithm can favourably affect implant survival and clinical outcome in the long-term. Conflict of interest All authors declare that they do not have any financial or personal relationships with other people or organisations that could inappropriately influence (bias) their work. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. References [1] Lewallen DG, Berry DJ. Periprosthetic fracture of the femur after total hip arthroplasty. J Bone Joint Surg Am 1997;79:1881–90. [2] Sarvilinna R, Huhtala HS, Puolakka TJ, Nevalainen JK, Pajama¨ki KJ. Periprosthetic fractures in total hip arthroplasty: an epidemiologic study. Int Orthop 2003;27(6):359–61. [3] Van der Wal BC, Vischjager M, Grimm B, Heyligers IC, Tonino AJ. Periprosthetic fractures around cementless hydroxyapatite-coated femoral stems. Int Orthop 2005;29(4):235–40. [4] Lindahl H, Oden A, Garellick G, Malchau H. The excess mortality due to periprosthetic femur fracture. A study from the Swedish national hip arthroplasty register. Bone 2007;40:1294–8. [5] Young SW, Walker CG, Pitto RP. Functional outcome of femoral periprosthetic fracture and revision hip arthroplasty: a matched pair study from the New Zealand registry. Acta Orthop 2008;79:483–8. [6] Johansson JE, McBroom R, Barrington TW. Fracture of the ipsilateral femur in patients with total hip replacement. J Bone Joint Surg Am 1981;63:1435–42.
1950
P. Kinov et al. / Injury, Int. J. Care Injured 46 (2015) 1945–1950
[7] Kelley SS. Periprosthetic femoral fractures. J Am Acad Orthop Surg 1994;2(3):164–72. [8] Mont MA, Maar DC. Fractures of the ipsilateral femur after hip arthroplasty: a statistical analysis of outcome based on 487 patients. J Arthroplasty 1994;9:511–9. [9] Morrey BF, Kavanagh BF. Complications with revision of the femoral component of total hip arthroplasty: comparison between cemented and uncemented techniques. J Arthroplasty 1992;7:71–9. [10] Whittaker RP, Sotos LN, Ralston EL. Fractures of the femur about femoral endoprostheses. J Trauma 1974;14:675–94. [11] Brady OH, Garbuz DS, Masri BA, Duncan CP. Classification of the hip. Orthop Clin North Am 1999;30(2):215–20. [12] Brady OH, Garbuz DS, Masri BA, Duncan CP. The reliability and validity of the Vancouver classification of femoral fractures after hip replacement. J Arthroplasty 2000;15(1):59–62. [13] Masri BA, Meek RM, Duncan CP. Periprosthetic fractures evaluation and treatment. Clin Orthop Relat Res 2004;420:80–95. [14] Lakstein D1, Kosashvili Y, Backstein D, Safir O, Lee P, Gross AE. The long modified extended sliding trochanteric osteotomy. Int Orthop 2011;35(1):13–7. [15] Harris WH. Traumatic arthritis of the hip after dislocation and acetabular fractures: treatment by mold arthroplasty. An end-result study using a new method of result evaluation. J Bone Joint Surg Am 1969;51:737–55. [16] Kinov P, Tivchev P. Evaluation of Bulgarian version of WOMAC in patients with hip pathology. In: Proceedings of the 10th National Congress of Bulgarian Orthopedic and Traumatologic Association, Borovets; 2007. p. 221–7.
[17] Ware JE, Kosinski M, Dewey JE, Gandek B. How to score and interpret singleitem health status measures: a manual for users of the SF-8 health survey. Boston: QualityMetric Incorporated; 2001. [18] Mukundan C, Rayan F, Kheir E, Macdonald D. Management of late periprosthetic femur fractures: a retrospective cohort of 72 patients. Int Orthop 2010;34(4):485–9. [19] Giannoudis PV, Kanakaris NK, Tsiridis E. Principles of internal fixation and selection of implants for periprosthetic femoral fractures. Injury 2007;38(6): 669–87. [20] Park SK, Kim YG, Kim SY. Treatment of periprosthetic femoral fractures in hip arthroplasty. Clin Orthop Surg 2011;3(2):101–6. [21] Harris B, Owen JR, Wayne JS, Jiranek WA. Does femoral component loosening predispose to femoral fracture?: an in vitro comparison of cemented hips. Clin Orthop Relat Res 2010;468:497–503. [22] Abdel MP, Lewallen DG, Berry DJ. Periprosthetic femur fractures treated with modular fluted, tapered stems. Clin Orthop Relat Res 2014;472(2):599–603. [23] Haddad FS, Duncan CP, Berry DJ, Lewallen DG, Gross AE, Chandler HP. Periprosthetic femoral fractures around well-fixed implants: use of cortical onlay allografts with or without a plate. J Bone Joint Surg Am 2002;84:945–50. [24] Barden B, von Knoch m, Fitzek JG, Loer F. Periprosthetic fractures with extensive bone loss treated with onlay strut allografts. Int Orhtop 2003;27(3):164–7. [25] Singh J, Sloan JA, Johanson NA. Challenges with health-related quality of life assessment in arthroplasty patients: problems and solutions. J Am Acad Orthop Surg 2010;18(2):72–82.