Posttraumatic arthritis of the knee can occur after distal femoral fracture and may be related to residual malalignment or direct intra-articular injury. Although the incidence of arthritis is considerably less than that after tibial plateau fractures, it is not uncommon for orthopedic surgeons to encounter patients with prior femoral fractures who have developed end-stage arthritis requiring total knee arthroplasty (TKA).1
The presence of retained hardware, often combined with femoral sclerosis, can make the use of intramedullary guides difficult during the replacement procedure.2 Staged surgical procedures to remove the hardware combined with femoral osteotomy to correct posttraumatic deformity may be required before TKA in these cases. However, multiple surgical procedures prior to TKA have been linked to an increased risk of arthrofibrosis and infection.3 Simultaneous TKA and hardware removal may require 2 incisions or the use of a skin flap for adequate exposure, both of which increase the risks of soft tissue necrosis and septic contamination. In addition, the screw holes left after hardware removal are potential stress risers, which may limit postoperative weight bearing or require additional support such as stems, intramedullary rods, and strut allografts.4,5 Likewise, removal of older and sometimes bulky hardware can be particularly challenging for the surgeon, leading to a longer and more complex surgical procedure.
Extramedullary jigs have been proposed as an alternative on the tibial side. However, femoral extramedullary instrumentation requires radiographic identification of the center of the femoral head with freehand pinning of the cutting block, affecting the accuracy of implant placement.6,7 In 1998, Ries8 suggested prophylactic femoral nailing during TKA with simultaneous femoral plate removal to reduce the risk of fracture through the screw holes, particularly in osteoporotic bone. The literature suggests that the outcome of TKA following femoral fractures using traditional techniques may be somewhat inferior to those seen after routine TKA.1
Computer-assisted TKA has been shown to improve implant alignment and to correct limb deformity.7,9 Computer navigation does not require intramedullary instrumentation and can therefore be performed without femoral hardware removal. As a result, staged surgery or simultaneous hardware removal are not required, avoiding the risks associated with these procedures and providing economic savings.10–15
The purpose of this study was to assess the effectiveness of computer-assisted TKA in managing posttraumatic arthritic joints after prior distal femoral fractures without removal of retained hardware. The study group was compared with a matched group of patients with atraumatic arthritis who underwent routine TKA, and the results were compared.
Materials and Methods
The study was approved by the ethics committee, and all participants gave their written consent to be included in the study. Between March 2001 and March 2009, seven hundred eighty-nine computer-assisted primary TKAs were performed in the authors’ institution. Of these, 16 patients with posttraumatic knee arthritis after prior distal femoral fracture underwent computer-assisted TKA (group A). At a minimum follow-up of approximately 2 years, each patient in group A was matched with a patient with atraumatic knee arthritis who had undergone a computer-assisted TKA (group B) during the same time period as those in group A. All surgeries were performed by 2 seniors authors (A.M., N.C.), both trained in navigated TKA.
For patients in group A, the average interval between fracture and subsequent TKA was 5.8 years (range, 1.9–12 years). Eight patients were men and 8 were women. In all cases, the hardware used for the original fracture fixation was retained. Retained hardware included a distal lateral plate and screws in 7 patients, an intramedullary nail in 6 patients, and screws alone in 3 patients.
In 10 patients, an e.motion knee prosthesis (B. Braun Aesculap, Tuttlingen, Germany) was implanted using dedicated computer navigation hardware and software (OrthoPilot; B. Braun Aesculap). In the remaining 6 patients, a computer-assisted Genesis II TKA (Smith & Nephew, Memphis, Tennessee) was performed using implant-specific hardware and software (VectorVision; BrainLAB, Munich, Germany). Patient age determined which implant was used. E.motion, which is a high-congruence mobile-bearing prosthesis, was implanted in patients younger than 65 years, whereas the Genesis II was implanted in all other patients.
Preoperatively, extra-articular femoral deformities were seen in 4 patients as a result of malunion (mean, 7.2°). The deformity was situated in the distal third of the femur in 2 patients and the middle third of the bone in 2 patients. In all patients, the correction of the limb deformity was planned using the prostheses, and no extra- or intra-articular corrective osteotomy was required. In 3 patients, implantation of the prosthesis required removal of the most distal screws in the femoral plate, which was achieved through small incisions without extension of the standardized TKA approach using a maximum 14-cm skin incision.
For patients in group B, each e.motion/Genesis II implant was strictly matched to the same implant with the similar features (cruciate retaining or sacrificing) using the same software. Patients were matched in terms of age, sex, preoperative range of motion, preoperative arthritis severity according to the Ahlbäck16 classification, type and grade of deformity, and implant type. Patients were matched with a maximum difference of 3 years with respect to age, 10° with respect to preoperative flexion, and 5° with respect to varus–valgus deformity (valgus to valgus and varus to varus).
In all patients, an anteromedial longitudinal knee incision with a medial parapatellar arthrotomy was used, and in no patient was the patella resurfaced, only denervated with electrocautery. All knee implants were cemented, and the same pre- and postoperative rehabilitation protocols were used for both groups. However, passive knee flexion was delayed in the 2 patients who required a tibial osteotomy. Early weight bearing as tolerated was encouraged in all patients. Operative time was documented in all cases. Surgery was performed using a standard navigation technique with anatomical trackers rigidly fixed in the distal part of the femur and the proximal part of the tibia, carefully avoiding the retained hardware in the distal femur in group A. Preoperatively, all knees were evaluated using the Knee Society Score (KSS) and the KSS functional score.17
At last follow-up, an author (C.C.) not involved in the original surgery evaluated all patients using the KSS and the Western Ontario and McMaster Universities Arthritis Index (WOMAC) self-administered questionnaire.18 Operative time, duration of hospital stay, and intra- and postoperative complications were recorded and compared. A radiological assessment was performed using a standardized protocol and magnification. Standing radiographs were obtained with the knee in maximum extension, the patella pointing forward, and both hips and ankles visible on the radiograph. Lateral radiographs were taken with the knee in 30° of flexion on a 20×40-cm radiographic film. The radiographs were repeated if malrotation was detected. These standard radiographs were used to determine the hip–knee–ankle (HKA) angle, frontal femoral component (FFC) angle, frontal tibial component (FTC) angle, and sagittal orientation (slope) of both femoral and tibial components. Two independent surgeons (A.M., C.C.) measured all angles on 2 separate occasions, and the final value for each angle was derived from the mean of these measurements. The FFC was determined as the angle between the mechanical axis of the femur and the transverse axis of the femoral component. The FTC was determined as the angle between the mechanical axis of the tibia and the transverse axis of the tibial component. The slopes of the femoral and tibial component were evaluated by measuring the angle formed between a line drawn tangential to the base plate (surface in contact with bone) of the respective components and the anterior femoral cortex or mechanical tibial axis.
The ideal alignment for each parameter was determined prior to the study as an FFC angle of 90°, FTC angle of 90°, HKA angle of 180°, femoral slope of 90°, and tibial slope of 87°. The total number of outliers for each parameter was determined. Outliers were defined as prostheses with any alignment parameter beyond 3° of the ideal value according to the literature because of a potential higher failure rate.19,20 Statistical analysis of the results was performed with a non-parametric test (Mann-Whitney U test) using Statistica 7.0 software (StatSoft Inc, Tulsa, Oklahoma). A P value of .05 or less was considered statistically significant.
Each group comprised 8 men and 8 women. Mean patient age was 69.9 years (range, 54–82 years) in group A and 71.3 years (range, 56–84 years) in group B. Mean preoperative flexion was 105.6° (range, 85°–125°) in group A and 108.7° (range, 90°–120°) in group B. Mean number of previous surgeries was 1.6 (range, 1–4) in group A and 0.3 (range, 0–2) in group B. Mean preoperative HKA angle was 175.8° (range, 170°–186°) in group A and 174.9° (range, 171°–183°) in group B. Three patients had a valgus malalignment and 13 had a varus malalignment in each group. In both groups, the arthritis grade according to Ahlbäck exceeded grade III in all patients. Preoperatively, mean KSS was 43.7 (range, 39–51) in group A and 45.1 (range, 40–49) in group B. Preoperative KSS functional score was 46.9 (range, 42–52) in group A and 48 (range, 42–52) in group B. No statistically significant differences existed in preoperative data between the 2 groups (Table 1).
Table 1: Preoperative Data
No complication specifically related to the computer navigation was observed in either group. No statistically significant differences existed in operative time or hospital stay (Table 2). Eleven patients in each group required postoperative blood transfusions. Two patients in group A required a tibial tubercle osteotomy to obtain adequate exposure and to avoid excessive traction on the patellar tendon. In both of these patients, the tubercle was reattached with screws with no adverse effect on the final outcome.
Table 2: Postoperative Results
At last follow-up (55.1 and 53.5 months for groups A and B, respectively), no implant had been revised and no major signs of radiological loosening were seen in either group. Mean KSS was 84.7 (range, 76–94) in group A and 86 (range, 76–94) in group B. Mean KSS functional score was 86.6 (range, 75–95) in group A and 88.4 (range, 80–95) in group B. No statistically significant differences existed in KSS and KSS functional scores between the 2 groups. The WOMAC score showed no statistically significant difference between the 2 groups for pain, function, and stiffness indices (Figure 1). Mean HKA angle was 179.1° (range, 177°–182°) in group A and 178.6° (range, 176°–182°) in group B. In group A, mean FFC and FTC angles were 89.6° (range, 86°–92°) and 89.6° (range, 87°–92°), respectively. In group B, mean FFC and FTC angles were 88.5° (range, 86°–95°) and 88.6° (range, 86°–91°), respectively. Mean slope of the femoral component was 89.9° (range, 88°–94°) in group A and 89.6° (range, 88°–94°) in group B. Mean slope of the tibial component was 87.1° (range, 85°–91°) in group A and 87° (range, 85°–91°) in group B. No statistically significant differences existed for any radiological parameters. Only 7 outliers in 6 patients in group A and 6 outliers in 6 patients in group B were observed across all radiological parameters (80 angles in total). No statistically significant differences existed between outliers in the 2 groups (Figure 2).
Figure 1: Comparison of postoperative Western Ontario and McMaster Universities Arthritis Index (WOMAC) pain, stiffness (Stif), and function (Func) scores (A) and postoperative (Post-op) Knee Society Scores (KSS) and KSS functional (Func) scores (B). Group A includes patients with posttraumatic knee arthritis after prior distal femoral fracture who underwent computer-assisted total knee arthroplasty. Group B includes patients with atraumatic knee arthritis who underwent a computer-assisted total knee arthroplasty.
Figure 2: Postoperative femoral (Fem) slope, tibial slope, frontal femoral component (FFC), and frontal tibial component (FTC) angles (A) and postoperative hip–knee–ankle (HKA) angle (B) alignments as measured on standardized radiographs. Group A includes patients with posttraumatic knee arthritis after prior distal femoral fracture who underwent computer-assisted total knee arthroplasty. Group B includes patients with atraumatic knee arthritis who underwent a computer-assisted total knee arthroplasty.
The number of patients with posttraumatic skeletal deformities undergoing TKA has increased steadily over recent decades.1–3 Total knee arthroplasty in patients with posttraumatic arthritis and retained hardware can be a particularly challenging problem. In these cases, accurate implant positioning is difficult because fracture fixation hardware can prevent the use of intramedullary guides.1,2 The presence of extra-articular deformities and intramedullary sclerosis can further jeopardize accurate prosthesis implantation using traditional alignment systems.1–3 Hardware removal at the time of prosthesis implantation results in stress risers at the site of each screw hole and therefore may lead to concern about potential periprosthetic fracture. This can result in the need to use more complex TKA implants. In older patients with osteoporotic bone, these concerns can lead to a slower postoperative rehabilitation program. In particular, restrictions on weight bearing and continuous passive motion in these patients may delay recovery.4,5 An alternative solution in these cases has been staged surgery. These techniques involve initial hardware removal, on occasion combined with prophylactic femoral nailing, followed by delayed TKA. However, the benefits of these 2-stage procedures must be weighed against the increased risk of complications, such as infections and arthrofibrosis.2,3 Unfortunately, the results of TKA combined with hardware removal performed at the time of prosthesis implantation or in a delayed fashion have been reported as inferior to routine TKA.1,8
Computer-assisted TKA using a surgical navigation system has been shown in the literature to offer results similar to traditional alignment guides in terms of accuracy.7,9,21 It enables the surgeon to make accurate bony cuts, orient the implants correctly, and provide a good qualitative intraoperative assessment of ligament balance and kinematics. In addition, the intramedullary canal is left intact when a computer-assisted technique is used. The ability to implant a TKA accurately without the use of femoral intramedullary guides is particularly useful in patients with posttraumatic arthritis after a prior femoral fracture and retained hardware. Despite the potential advantages of computer-assisted TKA in these patients, Kim et al22 and Tigani et al23 are the only studies in the literature documenting navigation exclusively used to keep the hardware in well-aligned femurs. They reported 2 and 5 cases of navigated TKA with retained hardware, respectively.22,23 In the study by Kim et al,22 the cases were successfully treated using a navigated TKA with modified self-tapping femoral anchoring pins, whereas in the study by Tigani et al,23 the cases were successfully managed using traditional bicortical pins.
The purpose of the current study was to determine the results of a 1-stage computer-assisted TKA with retention of fracture fixation hardware undertaken on patients with posttraumatic knee arthritis after prior femoral fracture. In this study, each patient with posttraumatic knee arthritis and retained hardware (group A) was matched with a similar patient with atraumatic arthritis (group B). Patients were matched for age, sex, preoperative range of motion, preoperative arthritis severity according to the Ahlbäck classification, type and grade of deformity, and type of implant. In addition, patients were also matched for computer navigation software used. All surgeries were performed with standard navigation tools with bicortical tools and a routine TKA approach using a maximum 14-cm skin incision without patella resurfacing. For patients in group A, contact with the retained hardware was avoided by insertion of the femoral pins directly through the surgical incision rather than the routine percutaneous insertion.
At the final assessment, no significant differences existed between the 2 groups in terms of operative time, duration of hospital stay, clinical results, and radiological outcomes, despite the existence of more complex cases considering tibial tubercle osteotomy. An explanation for this may be the extensive experience of the surgeons (more than 12 years) in managing these complex cases with navigation and the similar preoperative range of motion (maximum preoperative flexion difference, 10°) in the 2 groups. All patients in group A were managed successfully with a 1-stage procedure retaining the hardware. No computer navigation–specific complications were observed in either group, despite the use of 2 different types of implants and software. Retention of fracture fixation hardware caused no difficulties with the navigation or implantation processes.
This study is the largest cases series in the literature describing posttraumatic computer-assisted TKAs without hardware removal, despite no power analysis being performed. Other limitations include the use of different implants and software, the variation in surgical technique in 2 patients in group A (tibial tuberosity osteotomy), and the use of 2 different operating surgeons. In addition, the study would have been improved if the data had been collected prospectively. However, to the authors’ knowledge, this study is the largest published series of patients with posttraumatic knee arthritis after prior femoral fracture undergoing computer-assisted TKA with retention of fracture fixation hardware. The authors do not advocate retaining hardware in all knees just because of navigation. In cases of painful and bulky hardware or when quadriceps release is required, they recommend hardware removal, usually with a 2-stage procedure. Nevertheless, based on their own experience, they believe that computer-assisted TKA with retention of hardware could be helpful in managing patients with posttraumatic knee arthritis after prior femoral fracture. Computer navigation allows the surgeon to obtain good, reproducible results similar to a routine TKA with a 1-stage surgery. This has advantages in terms of clinical results and costs.
- Papadopoulos EC, Parvizi J, Lai CH, Lewallen DG. Total knee arthroplasty following prior distal femoral fracture. Knee. 2002; 9(4):267–274. doi:10.1016/S0968-0160(02)00046-7 [CrossRef]
- Weiss NG, Parvizi J, Hanssen AD, Trousdale RT, Lewallen DG. Total knee arthroplasty in post-traumatic arthrosis of the knee. J Arthroplasty. 2003; 18(3 suppl 1):23–26. doi:10.1054/arth.2003.50068 [CrossRef]
- Wu LD, Xiong Y, Yan SG, Yang QS. Total knee replacement for posttraumatic degenerative arthritis of the knee. Chin J Traumatol. 2005; 8(4):195–199.
- Brooks DB, Burstein AH, Frankel VH. The biomechanics of torsional fractures. The stress concentration effect of a drill hole. J Bone Joint Surg Am. 1970; 52(3):507–514.
- Johnson BA, Fallat LM. The effect of screw holes on bone strength. J Foot Ankle Surg. 1997; 36(6):446–451. doi:10.1016/S1067-2516(97)80097-X [CrossRef]
- Baldini A, Adravanti P. Less invasive TKA: extramedullary femoral reference without navigation. Clin Orthop Relat Res. 2008; 466(11):2694–2700. doi:10.1007/s11999-008-0435-9 [CrossRef]
- Confalonieri N, Manzotti A, Pullen C, Rangone V. Computer-assisted technique versus intramedullary and extramedullary alignment systems in total knee replacement: a radiological comparison. Acta Orthop Belg. 2005; 71(6):703–709.
- Ries MD. Prophylactic intramedullary femoral rodding during total knee arthroplasty with simultaneous femoral plate removal. J Arthroplasty. 1998; 13(6):718–721. doi:10.1016/S0883-5403(98)80019-3 [CrossRef]
- Pang CH, Chan WL, Yen CH, et al. Comparison of total knee arthroplasty using computer-assisted navigation versus conventional guiding system: a prospective study. J Orthop Surg (Hong Kong). 2009; 17(2):170–173.
- Bottros J, Klika AK, Lee HH, Polousky J, Barsoum WK. The use of navigation in total knee arthroplasty for patients with extra-articular deformity. J Arthroplasty. 2008; 23(1):74–78. doi:10.1016/j.arth.2007.01.021 [CrossRef]
- Mullaji A, Shetty GM. Computer-assisted total knee arthroplasty for arthritis with extra-articular deformity. J Arthroplasty. 2009; 24(8):1164–1169. doi:10.1016/j.arth.2009.05.005 [CrossRef]
- Kuo CC, Bosque J, Meehan JP, Jamali AA. Computer-assisted navigation of total knee arthroplasty for osteoarthritis in a patient with severe posttraumatic femoral deformity. J Arthroplasty. 2011; 26(6):976.e17–20. doi:10.1016/j.arth.2010.07.017 [CrossRef]
- Martin A, Wohlgenannt O, Prenn M, von Strempel A. Post-traumatic osteoarthritis of the knee: special indication for navigation TKA. Unfallchirurg. 2008; 111(9):754–758. doi:10.1007/s00113-007-1391-7 [CrossRef]
- Klein GR, Austin MS, Smith EB, Hozack WJ. Total knee arthroplasty using computer-assisted navigation in patients with deformities of the femur and tibia. J Arthroplasty. 2006; 21(2):284–288. doi:10.1016/j.arth.2005.07.013 [CrossRef]
- Patai J, Janositz G, Mécs L, Tóth K. Navigated total knee arthroplasty in a patient with severe diaphyseal deformities. Acta Orthop Belg. 2007; 73(4):536–544.
- Ahlbäck S. Osteoarthrosis of the knee. A radiographic investigation. Acta Radiol Diagn (Stockh). 1968; (suppl 277):7–72.
- Insall JN, Dorr LD, Scott RD, Scott WN. Rationale of the Knee Society Clinical Rating System. Clin Orthop Relat Res. 1998; (248):13–14.
- Bellamy N, Buchanan WW, Goldsmith CH, Campbell J, Stitt LW. Validation study of WOMAC: a health status instrument for measuring clinically important patient relevant outcomes to antirheumatic drug therapy in patients with osteoarthritis of the hip and knee. J Rheumatol. 1988; 15(12):1833–1840.
- Lotke PA, Ecker ML. Influence of positioning of prosthesis in total knee replacement. J Bone Joint Surg Am. 1977; 59(1):77–79.
- Ritter MA, Faris PM, Keating EM, Meding JB. Postoperative alignment of total knee replacement. Its effect on survival. Clin Orthop Relat Res. 1994; (299):153–156.
- Manzotti A, Cerveri P, De Momi E, Pullen C, Confalonieri N. Relationship between cutting errors and learning curve in computer-assisted total knee replacement. Int Orthop. 2010; 34(5):655–662. doi:10.1007/s00264-009-0816-z [CrossRef]
- Kim KK, Heo YM, Won YY, Lee WS. Navigation-assisted total knee arthroplasty for the knee retaining femoral intramedullary nail, and distal femoral plate and screws. Clin Orthop Surg. 2011; 3(1):77–80. doi:10.4055/cios.2011.3.1.77 [CrossRef]
- Tigani D, Masetti G, Sabbioni G, Ben Ayad R, Filanti M, Fosco M. Computer-assisted surgery as indication of choice: total knee arthroplasty in case of retained hardware or extra-articular deformity. Int Orthop. 2012; 36(7):1379–1385. doi:10.1007/s00264-011-1476-3 [CrossRef]
|Group Aa||Group Bb|
|Age, y||69.9±8 (54–82)||71.3±7.8 (56–84)||.06|
|No. of M/F||8/8||8/8|
|No. of preop surgeries||1.7±0.9 (1–4)||0.3±0.6 (0–2)||.0005|
|Flexion, deg||105.6±12.7 (85–125)||108.8±9.7 (90–120)||.13|
|KSS||44±3 (39–51)||45.1±2.6 (40–49)||.09|
|KSS functional score||46.9±3.4 (42–52)||48±2.5 (42–52)||.15|
|HKA angle, deg||175.8±4.1 (170–186)||174.9±3.8 (171–183)||.07|
|No. of valgus/varus deformities||3/13||3/13|
|Group Aa||Group Bb|
|Operative time, min||82.6±9.5 (69–101)||78.5±9.9 (67–105)||.1|
|Duration of hospital stay, d||7.1±1.2 (5–9)||7.3±0.6 (6–8)||.8|
|KSS||84.7±6.3 (76–94)||86±5.7 (76–94)||.3|
|KSS functional score||86.6±7.4 (75–95)||88.4±5.9 (80–95)||.3|
|HKA angle, deg||179.1±1.7 (177–182)||178.6±1.5 (176–182)||.2|
|FFC angle, deg||89.6±1.6 (86–92)||88.5±1.5 (86–95)||.08|
|FTC angle, deg||89.6±1.6 (87–92)||88.6±1.2 (86–91)||.08|
|Femoral slope, deg||89.9±1.5 (88–94)||89.6±1.5 (88–94)||.6|
|Tibial slope, deg||87.1±1.9 (85–91)||87±1.4 (85–91)||.9|
|No. of outliers||7||6||1.0|