The Journal of Arthroplasty xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
The Journal of Arthroplasty journal homepage: www.arthroplastyjournal.org
Intraoperative Proximal Femoral Fracture in Primary Cementless Total Hip Arthroplasty Danielle Y. Ponzio, MD , Alisina Shahi, MD, Andrew G. Park, MD, MBS, James J. Purtill, MD Rothman Institute/Department of Orthopaedic Surgery, Thomas Jefferson University, Philadelphia, Pennsylvania
a r t i c l e
i n f o
Article history: Received 18 December 2014 Accepted 28 February 2015 Available online xxxx Keywords: total hip arthroplasty cementless hip arthroplasty femoral fracture cerclage press-fit calcar fracture
a b s t r a c t Intraoperative proximal femoral fracture is a complication of primary cementless total hip arthroplasty (THA) at rates of 2.95–27.8%. A retrospective review of 2423 consecutive primary cementless THA cases identified 102 hips (96 patients) with fracture. Multivariate analysis compared fracture incidences between implants, Accolade (Stryker Orthopaedics) and Tri-Lock (DePuy Orthopaedics, Inc.), and evaluated potential risk factors using a randomized control group of 1150 cases without fracture. The fracture incidence was 4.4% (102/2423), 3.7% (36/ 1019) using Accolade and 4.9% using Tri-Lock (66/1404) (P = 0.18). Female gender (OR = 1.96; 95% CI 1.19–3.23; P = 0.008) and smaller stem size (OR = 1.64; 95% CI 1.04–2.63; P = 0.03) predicted increased odds of fracture. No revisions of the femoral component were required in the fracture cohort. © 2015 Elsevier Inc. All rights reserved.
The clinical success of cementless total hip arthroplasty (THA) depends upon initial stability at the prosthesis-bone interface. Stability is achieved through press-fitting of a femoral component compatible with the endosteal geometry of the bone [15]. Press-fitting may lead to an increase of intraoperative proximal femoral fractures for uncemented THA compared to cemented designs [9,15]. Fracture occurs during primary cementless THA at rates of 2.95–27.8% [2,3]. If recognized, intraoperative fracture can be addressed with cerclage wire techniques. Cerclage wires reduce the risk of crack propagation [8] and achieve satisfactory initial implant stability without compromising clinical outcome [1,2,7,14,15]. However, intraoperative fracture may lead to increased cost and intraoperative time as well as risk of vascular or nerve injury. Inadequate fixation or unrecognized fracture may lead to fracture displacement or nonunion, persistent thigh pain, poor bone ingrowth, and aseptic loosening of the femoral stem [6] necessitating further surgery. Previously reported risk factors for fracture during femoral component implantation include use of press-fit cementless femoral stems [3,7,12,14], minimally invasive techniques, anterolateral approach [3], previous surgery on the ipsilateral hip, diagnoses other than osteoarthritis [1], revision surgery [15], and female gender [3,4,12]. In the largest single surgeon consecutive series reported, this study aims to (1) present the incidence of proximal femoral fracture during
No author associated with this paper has disclosed any potential or pertinent conflicts which may be perceived to have impending conflict with this work. For full disclosure statements refer to http://dx.doi.org/10.1016/j.arth.2015.02.043. Reprint requests: J. J. Purtill, MD and D. Y. Ponzio, MD, Rothman Institute/Department of Orthopaedic Surgery, Thomas Jefferson University, 1025 Walnut Street, Room 516 College Building, Philadelphia, PA 19107.
primary cementless THA, (2) describe the fracture cohort demographically and by radiographic femoral morphology (Dorr classification), (3) compare the fracture incidence between two femoral stem types and multiple stem sizes, and (4) identify risk factors for fracture. Methods The present study is a retrospective review of 2423 patients who underwent primary cementless THA between January 2000 and December 2013 performed by a single senior adult reconstruction fellowship-trained surgeon (J.J.P.). Over this thirteen year period, the surgeon documented 102 hips (96 patients) with an intraoperative proximal femoral fracture, defined as a nondisplaced or minimally displaced incomplete linear discontinuity along an approximately 90 degree arc of the medial calcar proximal to the lesser trochanter. Of twenty-two patients within the fracture cohort who underwent bilateral THA, six had bilateral fractures. The 102 hips with intraoperative femoral fracture were matched by date of surgery (DOS) to 1150 patients who underwent primary cementless THA without fracture. Matching by DOS was intended to capture a control group of patients with similar demographics and surgical details, including implants utilized, as these variables could change over time. For all patients in each group, a chart review was performed to extract patient demographic data including: age at the time of surgery, gender, body mass index (BMI), preoperative diagnosis, operative side, and length of stay (LOS) (Table 1). Operative notes were reviewed to confirm femoral component type and size. Anteroposterior pelvis radiographs were retrospectively and independently reviewed by two authors for classification of the native proximal femoral morphology according to the Dorr classification [10]. Clinical outcomes were
http://dx.doi.org/10.1016/j.arth.2015.02.043 0883-5403/© 2015 Elsevier Inc. All rights reserved.
Please cite this article as: Ponzio DY, et al, Intraoperative Proximal Femoral Fracture in Primary Cementless Total Hip Arthroplasty, J Arthroplasty (2015), http://dx.doi.org/10.1016/j.arth.2015.02.043
2
D.Y. Ponzio et al. / The Journal of Arthroplasty xxx (2015) xxx–xxx
Table 1 Patient Demographics.
Table 2 There Is No Significant Difference in Fracture Incidence between Two Implant Types.
Demographic
Cohort (Fracture)
Control (No Fracture)
Patients (Hips) Age (Years) Sex Female Male Operative Side Right Left Body Mass Index Length of Stay (Days)
98 (102) 60.5 ± 13.4
1150 63.1 ± 11.8
P-value
68 30
544 520
P b 0.001
57 45 29.1 ± 6.5 3.3 ± 1.3
598 552 29.1 ± 5.1 2.7 ± 1.7
P = 0.47
P = 0.03
P = 0.67 P b 0.001
Annual Usage Total Cases Cohort (Fracture) Control (No Fracture) Fracture Incidence
Accolade
Tri-Lock
Total
2000–2008 1019 36 983 3.7%
2008–2013 1404 66 1338 4.9%
2000–2013 2423 102 2321 4.4%
⁎ No significant difference in fracture incidence between implant types, P = 0.18.
Statistical Analysis
⁎ Represents values that are statistically significant, P b 0.05.
determined by reviewing patient records to identify postoperative complications and need for reoperation. All procedures were performed by one surgeon with consistent technique using a modified anterolateral (Hardinge) approach with the patient in the supine position on a standard radiolucent operating table. Femoral components were wedge-shaped, tapered, proximally porous coated, press-fit stem designs. In both the control and study groups, the Accolade (Stryker Orthopaedics, Kalamazoo, MI) was used from August 2000 until January 2008 and Tri-Lock (DePuy Orthopaedics, Inc, Warsaw, IN) was used from 2008 until the present time. Surgical technique for both stems includes femoral canal reaming followed by sequential broaching until stability of the implant is reached. All proximal femoral fractures were visually recognized in the operating room upon impaction of the final femoral component. No fractures occurred during reaming or broaching of the canal or during trial reductions. There were no periprosthetic fractures in the fracture or control cohorts at any time postoperatively. Fracture fixation involved removal of the prosthesis, cerclage of the proximal femur with a cable above the level of the lesser trochanter, impaction of the femoral component then final tightening of the cable and crimping of the fixation clip (Fig. 1). Visual and tactile implant stability was confirmed for all cases. Postoperatively, the patient was permitted to weight bear as tolerated in accordance with the surgeon's normal established protocol.
Univariate analyses were performed using Student's t-test and Fisher's exact test where appropriate (SigmaStat 4.0, San Jose, CA; IBM SPSS Statistics 21.0, Armonk, NY). Multivariate analyses were performed with logistic regression modeling (R project 3.0.0, Auckland, New Zealand). Significance was established at P ≤ 0.05. Source of Funding No external source of funding was utilized. Results Intraoperative proximal femoral fractures were identified in 102 hips (96 patients) with an overall fracture incidence of 4.4% (Table 2). Of 22 patients undergoing bilateral THA, 6 patients, all female, had bilateral intraoperative proximal femoral fractures (Fig. 1). One patient received one of each implant type on either hip. The fracture incidence was similar using Accolade and Tri-Lock, 3.7% and 4.9%, respectively (P = 0.18). Patient demographics among the fracture group were similar between Accolade and Tri-Lock (Table 3). The majority of patients in the fracture cohort had a diagnosis of osteoarthritis (Table 4) and moderate bone quality, 50% Type B, as determined by the Dorr classification (Fig. 2). Most fractures occurred despite good bone quality, with 86% graded as type A or B. Interobserver reliability was excellent for radiographic assessment of the Dorr classification (Gamma = 0.893 ± 0.048). Univariate analysis of the fracture cohort and control group was employed to identify potential risk factors for proximal femoral fracture during THA. Gender (female, P b 0.001) and younger age (mean of 60.5 years, P = 0.03) were statistically significant risk factors in the fracture cohort, whereas BMI and operative side were not. The fracture incidence increased over time with increasing volume of THA (P = 0.008) (Fig. 3). Multivariate analysis using a logistic regression model of fracture incidence versus stem size was created. Covariates included BMI, gender, and age. Stem size and date of surgery were included as binary variables centered on the median value. Both Tri-Lock and Accolade have 11 gradations of sizing therefore data was compiled on a 0–10 scale for ease of analysis. Females (OR = 1.96, CI = 1.19–3.23), smaller stem sizes (OR = 1.64, CI = 1.04–2.63) (Fig. 4), and age (OR = 0.98, CI = 0.96–0.99) were significant and independent risk factors for proximal femoral fracture (Table 5). Date of surgery, potentially capturing a “learning curve” or “experience factor” effect did not achieve statistical significance in this model.
Table 3 Patient Demographics Are Similar between Two Implant Types.
Fig. 1. Postoperative anteroposterior radiograph of the pelvis demonstrating bilateral primary cementless total hip arthroplasty complicated by intraoperative proximal femoral fracture treated with cerclage of the proximal femur with a cable above the level of the lesser trochanter.
Gender (Female/Male) Side (Right/Left) Age (Years) BMI Length of Stay (Days)
Accolade
Tri-Lock
P-value
23/11 25/11 62.5 ± 13.9 30.5 ± 7.6 3.3 ± 1.0
45/19 32/34 60.2 ± 12.2 28.5 ± 5.7 3.4 ± 1.5
P P P P P
= = = = =
0.82 0.06 0.82 0.18 0.99
Please cite this article as: Ponzio DY, et al, Intraoperative Proximal Femoral Fracture in Primary Cementless Total Hip Arthroplasty, J Arthroplasty (2015), http://dx.doi.org/10.1016/j.arth.2015.02.043
D.Y. Ponzio et al. / The Journal of Arthroplasty xxx (2015) xxx–xxx Table 4 Osteoarthritis Is the Most Common Preoperative Diagnosis in the Fracture Cohort. Diagnosis
Hips (%)
Osteoarthritis Avascular Necrosis SCFE Congenital Dysplasia Rheumatoid Arthritis Post-traumatic Arthritis
83 (81%) 14 (13%) 2 (2%) 2 (2%) 1 (1%) 1 (1%)
There were no femoral component failures, periprosthetic fractures, or dislocations necessitating revision. Two patients underwent reoperation for infection, one of which retained their initial components, and the other underwent a girdlestone procedure. Two patients had residual anterior thigh pain at final follow-up office visits. One of these had findings of femoral component loosening, including radiolucency and subsidence. Their function was, however, satisfactory and in consideration of multiple co-morbidities (EF b 25%), the patient was felt not to be an operative candidate, and use of a cane was recommended. Discussion Intraoperative proximal femoral fracture is a well-described complication of THA in the literature though investigational studies suffer from small sample sizes and poor methodological quality. Fracture may go undetected with subsequent failure of the femoral implant. Fixation is associated with increased cost and operative time, and may risk vascular or nerve injury. The incidence of this complication in primary cementless THA ranges from 2.95% to 27.8% [11]. The variability in fracture incidences is likely attributed to studies which vary in sample size, femoral stem manufacturers, and insertion techniques [5]. In the current study, spanning 13 years, the overall fracture incidence was 4.4% with no significant differences in fracture rate between two implants Accolade (used from 2000–2006) and Tri-Lock (2007–2013). However, there was a change in the overall fracture rate over time, which coincides with increasing THA volume. At its lowest the fracture incidence was 1.28% in 2001, while its highest occurred in 2013 at 6.75%. Minimizing the risk of intraoperative proximal femoral fracture may be accomplished by recognizing predisposing factors. Previously reported risk factors for fracture during femoral component implantation include use of press-fit cementless femoral stems [3,7,12,14], minimally invasive techniques, anterolateral approach [3], previous surgery on the ipsilateral hip, diagnoses other than osteoarthritis [1], revision surgery [15], and female gender [3,4,12]. The present study considers primary cementless THA using an anterolateral approach. We found that
3
age was statistically significant, but given the odds ratio (0.98) and associated confidence interval (0.96–0.99) closely centered around the null hypothesis, there is most likely no clinical significance. Interestingly, when controlling for all other covariates, whether date of surgery was before or later than the median date there was no statistical effect on calcar fracture incidence (OR = 1.18, CI = 0.57–2.43). Future research with more rigorous statistical modeling and additional surgeons and patient cohorts may be necessary to evaluate the temporal discrepancy. We found that the incidence of fracture was significantly increased by two-fold for females, and notably, all of the cases of bilateral THA with bilateral fracture were female. Proposed etiologies of the higher rate of fracture in females focus on estrogen related changes in bone mineral density or proximal femoral geometry mismatch. Additionally, we found smaller stem sizes were associated with approximately a 1.5-fold increase in calcar fracture incidence (OR = 1.64, CI = 1.04–2.63). While smaller stem sizes were significantly associated with female gender, these risk factors are independent predictors of fracture. The effect of stem size on calcar fractures has been investigated previously in a study comparing ten patients with intraoperative fracture versus a matched cohort of twenty control patients undergoing primary cementless THA [15]. Sharkey et al found that the implants for the fracture group were larger on average than for the control group which they attributed to greater hoop stresses present with the application of similar forces in a larger cylinder than in a smaller one. However, this explanation seems counterintuitive as smaller stem sizes are selected to match smaller femurs that have thinner cortices and weaker mechanical properties. Other research has associated under reaming of the femoral cortex, use of a large-diameter femoral stem, and a low ratio between the diameters of the femoral cortex and canal with a greater risk of intraoperative fracture [5]. While a biomechanical explanation, without cross-sectional radiographic analysis, is out of the scope of the present investigation, we found a higher incidence of fracture upon using smaller stems. The interplay between femoral preparation instrumentation and the femoral prosthesis may offer an interesting possible explanation of our study's findings. The femoral instrumentation used to achieve an intimate fit between bone and prosthesis can be significantly undersized compared with the real prosthesis depending on the hip system used [2]. When the oversized femoral prosthesis is implanted into the proximal femur, fracture of the femur can occur easily. In cadaveric femurs, insertion of an optimal-size prosthesis after preparing the femoral canal with instruments the same size as the prosthesis produced moderate assembly strains, up to 1000 microstrain [2]. Half a millimeter press-fit of optimal prostheses produced larger assembly strains, up to 2000 microstrain, and half a millimeter press-fit of a prosthesis that was also one size (1.0 mm) larger than that determined to be optimum
Fig. 2. The Dorr classification serves as a descriptor of femoral morphology and bone quality. Type A femurs have a narrow canal with thick cortical walls (champagne flute canal). Type B femurs have moderate cortical walls. Type C femurs have a wide canal with thin cortical walls (stove-pipe canal). Radiographic analysis of the fracture cohort revealed that most fractures occurred despite good bone quality, with 86% graded as type A and B.
Please cite this article as: Ponzio DY, et al, Intraoperative Proximal Femoral Fracture in Primary Cementless Total Hip Arthroplasty, J Arthroplasty (2015), http://dx.doi.org/10.1016/j.arth.2015.02.043
4
D.Y. Ponzio et al. / The Journal of Arthroplasty xxx (2015) xxx–xxx
Fig. 3. Fracture incidence increased over time with increasing volume of THA.
produced even larger assembly strains (2000–6000 microstrain) and longitudinal linear fractures in the femoral cortex [9]. The magnitude of peak strains produced by press-fitting the femoral component and the small amount of disparity between the size of the recess and the prosthesis necessary to produce these strains show the narrow range of tolerances available to the surgeon. Given that smaller broaches are used more frequently, it is possible that instrumentation sets exhibit wear that may result in a relative oversizing of the femoral prosthesis resulting in calcar fracture most prevalent in smaller stem sizes. Femoral broach wear also would explain the rise in fracture incidence with increasing THA volume and over time. To our knowledge, there is no literature examining femoral preparation instrumentation wear which could have important industrial implications in quality control testing and timing of set turnover.
The present study, to our knowledge, is the largest single surgeon series that reports proximal femoral fracture incidence and risk factors during THA. However, our findings are specific to two implant manufacturers and insertion techniques utilized by one surgeon which may not be generalizable. Additionally, the study included limited data in terms of radiographic analysis and future studies may include micro-CT, biomechanical testing, and clinical follow-up data. In summary, we reviewed one surgeon's experience with intraoperative proximal femoral fractures. These fractures should be anticipated with the use of cementless THA designs. The surgeon must have a high index of suspicion for iatrogenic fractures particularly in females and with use of smaller femoral stems. During insertion of the femoral stem, a sudden change in resistance is highly suggestive of a femoral fracture. Intraoperative assessment of the stability of the component is
Fig. 4. Fracture incidence is highest with the use of smaller stem sizes.
Please cite this article as: Ponzio DY, et al, Intraoperative Proximal Femoral Fracture in Primary Cementless Total Hip Arthroplasty, J Arthroplasty (2015), http://dx.doi.org/10.1016/j.arth.2015.02.043
D.Y. Ponzio et al. / The Journal of Arthroplasty xxx (2015) xxx–xxx Table 5 Multivariate Analysis Reveals Increased Odds of Fracture with Female Gender and Small Stem Size. Female OR CI P-value
1.96 1.19–3.23 0.008
Age 0.98 0.96–0.99 0.02
BMI 0.99 0.96–1.03 0.67
DOSa b
1.18 0.57–2.43 0.66
Stem Sizea 1.64c 1.04–2.63 0.03
OR = odds ratio, CI = confidence interval, BMI = body mass index, DOS = date of surgery. ⁎ Represents values that are statistically significant, P b 0.05. a Represents binary variables centered on the median value. b Corresponds to surgery later than the median date. c Corresponds to stem size smaller than median size.
crucial even if there is no overt sign of fracture[5,13]. This complication, if managed properly, did not compromise clinical results.
References 1. Berend KR, Lombardi AV, Mallory TH, et al. Cerclage wires or cables for the management of intraoperative fracture associated with a cementless, tapered femoral prosthesis: results at 2 to 16 years. J Arthroplasty 2004;19:17.
5
2. Berend KR, Lombardi AV. Intraoperative femur fracture is associated with stem and instrument design in primary total hip arthroplasty. Clin Orthop 2010;468:2377. 3. Berend ME, Smith A, Meding JB, et al. Long-term outcome and risk factors of proximal femoral fracture in uncemented and cemented total hip arthroplasty in 2551 hips. J Arthroplasty 2006;21:53. 4. Berry DJ. Management of periprosthetic fractures: the hip. J Arthroplasty 2002;17:11. 5. Davidson D, Pike J, Garbuz D, et al. Intraoperative periprosthetic fractures during total hip arthroplasty. Evaluation and management. J Bone Joint Surg Am 2008;90:2000. 6. Fishkin Z, Han S-M, Ziv I. Cerclage wiring technique after proximal femoral fracture in total hip arthroplasty. J Arthroplasty 1999;14:98. 7. Fitzgerald RH, Brindley GW, Kavanagh BF. The uncemented total hip arthroplasty. Intraoperative femoral fractures. Clin Orthop 1988:61. 8. Incavo SJ, DiFazio F, Wilder D, et al. Longitudinal crack propagation in bone around femoral prosthesis. Clin Orthop 1991:175. 9. Jasty M, Bragdon CR, Rubash H, et al. Unrecognized femoral fractures during cementless total hip arthroplasty in the dog and their effect on bone ingrowth. J Arthroplasty 1992;7:501. 10. Khanuja HS, Vakil JJ, Goddard MS, et al. Cementless femoral fixation in total hip arthroplasty. J Bone Joint Surg Am 2011;93:500. 11. Mayle RE, Della Valle CJ. Intra-operative fractures during THA: see it before it sees us. J Bone Joint Surg (Br) 2012;94:26. 12. Moroni A, Faldini C, Piras F, et al. Risk factors for intraoperative femoral fractures during total hip replacement. Ann Chir Gynaecol 2000;89:113. 13. Schutzer SF, Grady-Benson J, Jasty M, et al. Influence of intraoperative femoral fractures and cerclage wiring on bone ingrowth into canine porous-coated femoral components. J Arthroplasty 1995;10:823. 14. Schwartz JT, Mayer JG, Engh CA. Femoral fracture during non-cemented total hip arthroplasty. J Bone Joint Surg Am 1989;71:1135. 15. Sharkey PF, Hozack WJ, Booth RE, et al. Intraoperative femoral fractures in cementless total hip arthroplasty. Orthop Rev 1992;21:337.
Please cite this article as: Ponzio DY, et al, Intraoperative Proximal Femoral Fracture in Primary Cementless Total Hip Arthroplasty, J Arthroplasty (2015), http://dx.doi.org/10.1016/j.arth.2015.02.043
Contents lists available at ScienceDirect
The Journal of Arthroplasty journal homepage: www.arthroplastyjournal.org
Intraoperative Proximal Femoral Fracture in Primary Cementless Total Hip Arthroplasty Danielle Y. Ponzio, MD , Alisina Shahi, MD, Andrew G. Park, MD, MBS, James J. Purtill, MD Rothman Institute/Department of Orthopaedic Surgery, Thomas Jefferson University, Philadelphia, Pennsylvania
a r t i c l e
i n f o
Article history: Received 18 December 2014 Accepted 28 February 2015 Available online xxxx Keywords: total hip arthroplasty cementless hip arthroplasty femoral fracture cerclage press-fit calcar fracture
a b s t r a c t Intraoperative proximal femoral fracture is a complication of primary cementless total hip arthroplasty (THA) at rates of 2.95–27.8%. A retrospective review of 2423 consecutive primary cementless THA cases identified 102 hips (96 patients) with fracture. Multivariate analysis compared fracture incidences between implants, Accolade (Stryker Orthopaedics) and Tri-Lock (DePuy Orthopaedics, Inc.), and evaluated potential risk factors using a randomized control group of 1150 cases without fracture. The fracture incidence was 4.4% (102/2423), 3.7% (36/ 1019) using Accolade and 4.9% using Tri-Lock (66/1404) (P = 0.18). Female gender (OR = 1.96; 95% CI 1.19–3.23; P = 0.008) and smaller stem size (OR = 1.64; 95% CI 1.04–2.63; P = 0.03) predicted increased odds of fracture. No revisions of the femoral component were required in the fracture cohort. © 2015 Elsevier Inc. All rights reserved.
The clinical success of cementless total hip arthroplasty (THA) depends upon initial stability at the prosthesis-bone interface. Stability is achieved through press-fitting of a femoral component compatible with the endosteal geometry of the bone [15]. Press-fitting may lead to an increase of intraoperative proximal femoral fractures for uncemented THA compared to cemented designs [9,15]. Fracture occurs during primary cementless THA at rates of 2.95–27.8% [2,3]. If recognized, intraoperative fracture can be addressed with cerclage wire techniques. Cerclage wires reduce the risk of crack propagation [8] and achieve satisfactory initial implant stability without compromising clinical outcome [1,2,7,14,15]. However, intraoperative fracture may lead to increased cost and intraoperative time as well as risk of vascular or nerve injury. Inadequate fixation or unrecognized fracture may lead to fracture displacement or nonunion, persistent thigh pain, poor bone ingrowth, and aseptic loosening of the femoral stem [6] necessitating further surgery. Previously reported risk factors for fracture during femoral component implantation include use of press-fit cementless femoral stems [3,7,12,14], minimally invasive techniques, anterolateral approach [3], previous surgery on the ipsilateral hip, diagnoses other than osteoarthritis [1], revision surgery [15], and female gender [3,4,12]. In the largest single surgeon consecutive series reported, this study aims to (1) present the incidence of proximal femoral fracture during
No author associated with this paper has disclosed any potential or pertinent conflicts which may be perceived to have impending conflict with this work. For full disclosure statements refer to http://dx.doi.org/10.1016/j.arth.2015.02.043. Reprint requests: J. J. Purtill, MD and D. Y. Ponzio, MD, Rothman Institute/Department of Orthopaedic Surgery, Thomas Jefferson University, 1025 Walnut Street, Room 516 College Building, Philadelphia, PA 19107.
primary cementless THA, (2) describe the fracture cohort demographically and by radiographic femoral morphology (Dorr classification), (3) compare the fracture incidence between two femoral stem types and multiple stem sizes, and (4) identify risk factors for fracture. Methods The present study is a retrospective review of 2423 patients who underwent primary cementless THA between January 2000 and December 2013 performed by a single senior adult reconstruction fellowship-trained surgeon (J.J.P.). Over this thirteen year period, the surgeon documented 102 hips (96 patients) with an intraoperative proximal femoral fracture, defined as a nondisplaced or minimally displaced incomplete linear discontinuity along an approximately 90 degree arc of the medial calcar proximal to the lesser trochanter. Of twenty-two patients within the fracture cohort who underwent bilateral THA, six had bilateral fractures. The 102 hips with intraoperative femoral fracture were matched by date of surgery (DOS) to 1150 patients who underwent primary cementless THA without fracture. Matching by DOS was intended to capture a control group of patients with similar demographics and surgical details, including implants utilized, as these variables could change over time. For all patients in each group, a chart review was performed to extract patient demographic data including: age at the time of surgery, gender, body mass index (BMI), preoperative diagnosis, operative side, and length of stay (LOS) (Table 1). Operative notes were reviewed to confirm femoral component type and size. Anteroposterior pelvis radiographs were retrospectively and independently reviewed by two authors for classification of the native proximal femoral morphology according to the Dorr classification [10]. Clinical outcomes were
http://dx.doi.org/10.1016/j.arth.2015.02.043 0883-5403/© 2015 Elsevier Inc. All rights reserved.
Please cite this article as: Ponzio DY, et al, Intraoperative Proximal Femoral Fracture in Primary Cementless Total Hip Arthroplasty, J Arthroplasty (2015), http://dx.doi.org/10.1016/j.arth.2015.02.043
2
D.Y. Ponzio et al. / The Journal of Arthroplasty xxx (2015) xxx–xxx
Table 1 Patient Demographics.
Table 2 There Is No Significant Difference in Fracture Incidence between Two Implant Types.
Demographic
Cohort (Fracture)
Control (No Fracture)
Patients (Hips) Age (Years) Sex Female Male Operative Side Right Left Body Mass Index Length of Stay (Days)
98 (102) 60.5 ± 13.4
1150 63.1 ± 11.8
P-value
68 30
544 520
P b 0.001
57 45 29.1 ± 6.5 3.3 ± 1.3
598 552 29.1 ± 5.1 2.7 ± 1.7
P = 0.47
P = 0.03
P = 0.67 P b 0.001
Annual Usage Total Cases Cohort (Fracture) Control (No Fracture) Fracture Incidence
Accolade
Tri-Lock
Total
2000–2008 1019 36 983 3.7%
2008–2013 1404 66 1338 4.9%
2000–2013 2423 102 2321 4.4%
⁎ No significant difference in fracture incidence between implant types, P = 0.18.
Statistical Analysis
⁎ Represents values that are statistically significant, P b 0.05.
determined by reviewing patient records to identify postoperative complications and need for reoperation. All procedures were performed by one surgeon with consistent technique using a modified anterolateral (Hardinge) approach with the patient in the supine position on a standard radiolucent operating table. Femoral components were wedge-shaped, tapered, proximally porous coated, press-fit stem designs. In both the control and study groups, the Accolade (Stryker Orthopaedics, Kalamazoo, MI) was used from August 2000 until January 2008 and Tri-Lock (DePuy Orthopaedics, Inc, Warsaw, IN) was used from 2008 until the present time. Surgical technique for both stems includes femoral canal reaming followed by sequential broaching until stability of the implant is reached. All proximal femoral fractures were visually recognized in the operating room upon impaction of the final femoral component. No fractures occurred during reaming or broaching of the canal or during trial reductions. There were no periprosthetic fractures in the fracture or control cohorts at any time postoperatively. Fracture fixation involved removal of the prosthesis, cerclage of the proximal femur with a cable above the level of the lesser trochanter, impaction of the femoral component then final tightening of the cable and crimping of the fixation clip (Fig. 1). Visual and tactile implant stability was confirmed for all cases. Postoperatively, the patient was permitted to weight bear as tolerated in accordance with the surgeon's normal established protocol.
Univariate analyses were performed using Student's t-test and Fisher's exact test where appropriate (SigmaStat 4.0, San Jose, CA; IBM SPSS Statistics 21.0, Armonk, NY). Multivariate analyses were performed with logistic regression modeling (R project 3.0.0, Auckland, New Zealand). Significance was established at P ≤ 0.05. Source of Funding No external source of funding was utilized. Results Intraoperative proximal femoral fractures were identified in 102 hips (96 patients) with an overall fracture incidence of 4.4% (Table 2). Of 22 patients undergoing bilateral THA, 6 patients, all female, had bilateral intraoperative proximal femoral fractures (Fig. 1). One patient received one of each implant type on either hip. The fracture incidence was similar using Accolade and Tri-Lock, 3.7% and 4.9%, respectively (P = 0.18). Patient demographics among the fracture group were similar between Accolade and Tri-Lock (Table 3). The majority of patients in the fracture cohort had a diagnosis of osteoarthritis (Table 4) and moderate bone quality, 50% Type B, as determined by the Dorr classification (Fig. 2). Most fractures occurred despite good bone quality, with 86% graded as type A or B. Interobserver reliability was excellent for radiographic assessment of the Dorr classification (Gamma = 0.893 ± 0.048). Univariate analysis of the fracture cohort and control group was employed to identify potential risk factors for proximal femoral fracture during THA. Gender (female, P b 0.001) and younger age (mean of 60.5 years, P = 0.03) were statistically significant risk factors in the fracture cohort, whereas BMI and operative side were not. The fracture incidence increased over time with increasing volume of THA (P = 0.008) (Fig. 3). Multivariate analysis using a logistic regression model of fracture incidence versus stem size was created. Covariates included BMI, gender, and age. Stem size and date of surgery were included as binary variables centered on the median value. Both Tri-Lock and Accolade have 11 gradations of sizing therefore data was compiled on a 0–10 scale for ease of analysis. Females (OR = 1.96, CI = 1.19–3.23), smaller stem sizes (OR = 1.64, CI = 1.04–2.63) (Fig. 4), and age (OR = 0.98, CI = 0.96–0.99) were significant and independent risk factors for proximal femoral fracture (Table 5). Date of surgery, potentially capturing a “learning curve” or “experience factor” effect did not achieve statistical significance in this model.
Table 3 Patient Demographics Are Similar between Two Implant Types.
Fig. 1. Postoperative anteroposterior radiograph of the pelvis demonstrating bilateral primary cementless total hip arthroplasty complicated by intraoperative proximal femoral fracture treated with cerclage of the proximal femur with a cable above the level of the lesser trochanter.
Gender (Female/Male) Side (Right/Left) Age (Years) BMI Length of Stay (Days)
Accolade
Tri-Lock
P-value
23/11 25/11 62.5 ± 13.9 30.5 ± 7.6 3.3 ± 1.0
45/19 32/34 60.2 ± 12.2 28.5 ± 5.7 3.4 ± 1.5
P P P P P
= = = = =
0.82 0.06 0.82 0.18 0.99
Please cite this article as: Ponzio DY, et al, Intraoperative Proximal Femoral Fracture in Primary Cementless Total Hip Arthroplasty, J Arthroplasty (2015), http://dx.doi.org/10.1016/j.arth.2015.02.043
D.Y. Ponzio et al. / The Journal of Arthroplasty xxx (2015) xxx–xxx Table 4 Osteoarthritis Is the Most Common Preoperative Diagnosis in the Fracture Cohort. Diagnosis
Hips (%)
Osteoarthritis Avascular Necrosis SCFE Congenital Dysplasia Rheumatoid Arthritis Post-traumatic Arthritis
83 (81%) 14 (13%) 2 (2%) 2 (2%) 1 (1%) 1 (1%)
There were no femoral component failures, periprosthetic fractures, or dislocations necessitating revision. Two patients underwent reoperation for infection, one of which retained their initial components, and the other underwent a girdlestone procedure. Two patients had residual anterior thigh pain at final follow-up office visits. One of these had findings of femoral component loosening, including radiolucency and subsidence. Their function was, however, satisfactory and in consideration of multiple co-morbidities (EF b 25%), the patient was felt not to be an operative candidate, and use of a cane was recommended. Discussion Intraoperative proximal femoral fracture is a well-described complication of THA in the literature though investigational studies suffer from small sample sizes and poor methodological quality. Fracture may go undetected with subsequent failure of the femoral implant. Fixation is associated with increased cost and operative time, and may risk vascular or nerve injury. The incidence of this complication in primary cementless THA ranges from 2.95% to 27.8% [11]. The variability in fracture incidences is likely attributed to studies which vary in sample size, femoral stem manufacturers, and insertion techniques [5]. In the current study, spanning 13 years, the overall fracture incidence was 4.4% with no significant differences in fracture rate between two implants Accolade (used from 2000–2006) and Tri-Lock (2007–2013). However, there was a change in the overall fracture rate over time, which coincides with increasing THA volume. At its lowest the fracture incidence was 1.28% in 2001, while its highest occurred in 2013 at 6.75%. Minimizing the risk of intraoperative proximal femoral fracture may be accomplished by recognizing predisposing factors. Previously reported risk factors for fracture during femoral component implantation include use of press-fit cementless femoral stems [3,7,12,14], minimally invasive techniques, anterolateral approach [3], previous surgery on the ipsilateral hip, diagnoses other than osteoarthritis [1], revision surgery [15], and female gender [3,4,12]. The present study considers primary cementless THA using an anterolateral approach. We found that
3
age was statistically significant, but given the odds ratio (0.98) and associated confidence interval (0.96–0.99) closely centered around the null hypothesis, there is most likely no clinical significance. Interestingly, when controlling for all other covariates, whether date of surgery was before or later than the median date there was no statistical effect on calcar fracture incidence (OR = 1.18, CI = 0.57–2.43). Future research with more rigorous statistical modeling and additional surgeons and patient cohorts may be necessary to evaluate the temporal discrepancy. We found that the incidence of fracture was significantly increased by two-fold for females, and notably, all of the cases of bilateral THA with bilateral fracture were female. Proposed etiologies of the higher rate of fracture in females focus on estrogen related changes in bone mineral density or proximal femoral geometry mismatch. Additionally, we found smaller stem sizes were associated with approximately a 1.5-fold increase in calcar fracture incidence (OR = 1.64, CI = 1.04–2.63). While smaller stem sizes were significantly associated with female gender, these risk factors are independent predictors of fracture. The effect of stem size on calcar fractures has been investigated previously in a study comparing ten patients with intraoperative fracture versus a matched cohort of twenty control patients undergoing primary cementless THA [15]. Sharkey et al found that the implants for the fracture group were larger on average than for the control group which they attributed to greater hoop stresses present with the application of similar forces in a larger cylinder than in a smaller one. However, this explanation seems counterintuitive as smaller stem sizes are selected to match smaller femurs that have thinner cortices and weaker mechanical properties. Other research has associated under reaming of the femoral cortex, use of a large-diameter femoral stem, and a low ratio between the diameters of the femoral cortex and canal with a greater risk of intraoperative fracture [5]. While a biomechanical explanation, without cross-sectional radiographic analysis, is out of the scope of the present investigation, we found a higher incidence of fracture upon using smaller stems. The interplay between femoral preparation instrumentation and the femoral prosthesis may offer an interesting possible explanation of our study's findings. The femoral instrumentation used to achieve an intimate fit between bone and prosthesis can be significantly undersized compared with the real prosthesis depending on the hip system used [2]. When the oversized femoral prosthesis is implanted into the proximal femur, fracture of the femur can occur easily. In cadaveric femurs, insertion of an optimal-size prosthesis after preparing the femoral canal with instruments the same size as the prosthesis produced moderate assembly strains, up to 1000 microstrain [2]. Half a millimeter press-fit of optimal prostheses produced larger assembly strains, up to 2000 microstrain, and half a millimeter press-fit of a prosthesis that was also one size (1.0 mm) larger than that determined to be optimum
Fig. 2. The Dorr classification serves as a descriptor of femoral morphology and bone quality. Type A femurs have a narrow canal with thick cortical walls (champagne flute canal). Type B femurs have moderate cortical walls. Type C femurs have a wide canal with thin cortical walls (stove-pipe canal). Radiographic analysis of the fracture cohort revealed that most fractures occurred despite good bone quality, with 86% graded as type A and B.
Please cite this article as: Ponzio DY, et al, Intraoperative Proximal Femoral Fracture in Primary Cementless Total Hip Arthroplasty, J Arthroplasty (2015), http://dx.doi.org/10.1016/j.arth.2015.02.043
4
D.Y. Ponzio et al. / The Journal of Arthroplasty xxx (2015) xxx–xxx
Fig. 3. Fracture incidence increased over time with increasing volume of THA.
produced even larger assembly strains (2000–6000 microstrain) and longitudinal linear fractures in the femoral cortex [9]. The magnitude of peak strains produced by press-fitting the femoral component and the small amount of disparity between the size of the recess and the prosthesis necessary to produce these strains show the narrow range of tolerances available to the surgeon. Given that smaller broaches are used more frequently, it is possible that instrumentation sets exhibit wear that may result in a relative oversizing of the femoral prosthesis resulting in calcar fracture most prevalent in smaller stem sizes. Femoral broach wear also would explain the rise in fracture incidence with increasing THA volume and over time. To our knowledge, there is no literature examining femoral preparation instrumentation wear which could have important industrial implications in quality control testing and timing of set turnover.
The present study, to our knowledge, is the largest single surgeon series that reports proximal femoral fracture incidence and risk factors during THA. However, our findings are specific to two implant manufacturers and insertion techniques utilized by one surgeon which may not be generalizable. Additionally, the study included limited data in terms of radiographic analysis and future studies may include micro-CT, biomechanical testing, and clinical follow-up data. In summary, we reviewed one surgeon's experience with intraoperative proximal femoral fractures. These fractures should be anticipated with the use of cementless THA designs. The surgeon must have a high index of suspicion for iatrogenic fractures particularly in females and with use of smaller femoral stems. During insertion of the femoral stem, a sudden change in resistance is highly suggestive of a femoral fracture. Intraoperative assessment of the stability of the component is
Fig. 4. Fracture incidence is highest with the use of smaller stem sizes.
Please cite this article as: Ponzio DY, et al, Intraoperative Proximal Femoral Fracture in Primary Cementless Total Hip Arthroplasty, J Arthroplasty (2015), http://dx.doi.org/10.1016/j.arth.2015.02.043
D.Y. Ponzio et al. / The Journal of Arthroplasty xxx (2015) xxx–xxx Table 5 Multivariate Analysis Reveals Increased Odds of Fracture with Female Gender and Small Stem Size. Female OR CI P-value
1.96 1.19–3.23 0.008
Age 0.98 0.96–0.99 0.02
BMI 0.99 0.96–1.03 0.67
DOSa b
1.18 0.57–2.43 0.66
Stem Sizea 1.64c 1.04–2.63 0.03
OR = odds ratio, CI = confidence interval, BMI = body mass index, DOS = date of surgery. ⁎ Represents values that are statistically significant, P b 0.05. a Represents binary variables centered on the median value. b Corresponds to surgery later than the median date. c Corresponds to stem size smaller than median size.
crucial even if there is no overt sign of fracture[5,13]. This complication, if managed properly, did not compromise clinical results.
References 1. Berend KR, Lombardi AV, Mallory TH, et al. Cerclage wires or cables for the management of intraoperative fracture associated with a cementless, tapered femoral prosthesis: results at 2 to 16 years. J Arthroplasty 2004;19:17.
5
2. Berend KR, Lombardi AV. Intraoperative femur fracture is associated with stem and instrument design in primary total hip arthroplasty. Clin Orthop 2010;468:2377. 3. Berend ME, Smith A, Meding JB, et al. Long-term outcome and risk factors of proximal femoral fracture in uncemented and cemented total hip arthroplasty in 2551 hips. J Arthroplasty 2006;21:53. 4. Berry DJ. Management of periprosthetic fractures: the hip. J Arthroplasty 2002;17:11. 5. Davidson D, Pike J, Garbuz D, et al. Intraoperative periprosthetic fractures during total hip arthroplasty. Evaluation and management. J Bone Joint Surg Am 2008;90:2000. 6. Fishkin Z, Han S-M, Ziv I. Cerclage wiring technique after proximal femoral fracture in total hip arthroplasty. J Arthroplasty 1999;14:98. 7. Fitzgerald RH, Brindley GW, Kavanagh BF. The uncemented total hip arthroplasty. Intraoperative femoral fractures. Clin Orthop 1988:61. 8. Incavo SJ, DiFazio F, Wilder D, et al. Longitudinal crack propagation in bone around femoral prosthesis. Clin Orthop 1991:175. 9. Jasty M, Bragdon CR, Rubash H, et al. Unrecognized femoral fractures during cementless total hip arthroplasty in the dog and their effect on bone ingrowth. J Arthroplasty 1992;7:501. 10. Khanuja HS, Vakil JJ, Goddard MS, et al. Cementless femoral fixation in total hip arthroplasty. J Bone Joint Surg Am 2011;93:500. 11. Mayle RE, Della Valle CJ. Intra-operative fractures during THA: see it before it sees us. J Bone Joint Surg (Br) 2012;94:26. 12. Moroni A, Faldini C, Piras F, et al. Risk factors for intraoperative femoral fractures during total hip replacement. Ann Chir Gynaecol 2000;89:113. 13. Schutzer SF, Grady-Benson J, Jasty M, et al. Influence of intraoperative femoral fractures and cerclage wiring on bone ingrowth into canine porous-coated femoral components. J Arthroplasty 1995;10:823. 14. Schwartz JT, Mayer JG, Engh CA. Femoral fracture during non-cemented total hip arthroplasty. J Bone Joint Surg Am 1989;71:1135. 15. Sharkey PF, Hozack WJ, Booth RE, et al. Intraoperative femoral fractures in cementless total hip arthroplasty. Orthop Rev 1992;21:337.
Please cite this article as: Ponzio DY, et al, Intraoperative Proximal Femoral Fracture in Primary Cementless Total Hip Arthroplasty, J Arthroplasty (2015), http://dx.doi.org/10.1016/j.arth.2015.02.043
