The Use of a Multiplanar, Multi-Axis External Fixator to Achieve Knee Arthrodesis in a Worst Case Scenario: A Case Series

Abstract

Background

One of the most catastrophic outcomes following total knee arthroplasty (TKA) is a chronic periprosthetic infection with concomitant failure of the knee extensor mechanism. This study retrospectively reviewed the clinical records of 7 patients who were treated with a 6 axis circular external fixation frame (Taylor Spatial Frame (TSF)) for this condition. Fusion was achieved in 5 of 7 patients (71%) at an average of 8.4 months after surgery. Complications occurred in the treatment of 5 of 7 patients (71%). Infection was controlled in all cases. The TSF presents another valuable tool, which the orthopaedic surgeon should consider when treating these difficult cases.

Purpose

To evaluate the use of the Taylor Spatial Frame (TSF) to achieve knee arthrodesis in patients with chronically infected total knee arthroplasties (TKAs) with concomitant failure of the knee extensor mechanism.

Methods

We retrospectively evaluated the clinical records of 7 patients who were referred to our tertiary care orthopaedic medical center with multiple failed knee arthroplasties, chronic draining infection and complete loss of the extensor mechanism. All patients were treated with a similar protocol including, debridement and bony stabilization with an adjustable, 6 axis circular external fixation frame (TSF). Hospital charts were reviewed for sociodemographic information, surgical details, hospital course and complications. Radiographs were reviewed for healing and alignment. Follow up included clinical examination and radiographs.

Results

The mean age of the patients was 70.9 years (range, 59 – 83 years) at the time of application of the TSF. There were 3 men and 4 women. The average time between TKA and diagnosis of infection was 30.7 months (range, 2.6 – 67.0 months). The 7 patients had undergone an average of 3.3 prior surgical procedures (range, 2-4 procedures) on the ipsilateral extremity. Fusion was achieved in 5 of 7 patients (71%) at an average of 8.4 months after surgery (range, 6 – 10.5 months). Complications occurred in the treatment of 5 of 7 patients (71%) and included infection at the site of the pin tracks (5 patients), antibiotic- induced acute renal failure (1 patient), wound breakdown requiring flap closure (1 patient), and femur fracture secondary to a fall after placement of the antibiotic spacer but before application of the TSF (1 patient). The 2 patients in whom failure of fusion occurred returned to ambulation with an assistive device. Infection was controlled in all cases.

Conclusion

Fusion and complication rates in this cohort are comparable to those reported in previous studies using other techniques to achieve external fixation. The TSF is a versatile external fixator that offers another tool, which the orthopaedic surgeon should consider when treating these difficult cases.

Introduction

The incidence of infection following total knee arthroplasty (TKA) is approximately 0.4-2%^1-5. Despite this relatively low risk, a substantial number of patients will develop periprosthetic infections following this procedure. This is in part due to the sheer number of TKAs performed annually, and in part due to a shift in patient demographics that is projected to increase the demand for TKAs significantly over the next 20 years⁶. It can be expected that the prevalence of patients with periprosthetic infections after TKA will increase, even as the incidence may decrease.

Risk factors for infection after primary TKA have been well characterized⁵. One large retrospective study found patients who were morbidly obese, diabetic, younger, or undergoing unilateral procedures to be at increased risk⁷. Other groups have confirmed these findings, and also found increased risk among patients with rheumatoid arthritis^4,8,9, chronic corticosteroid use¹⁰, use of a hinged prostheses^4,9,11, complications during initial wound healing^12,13, and prior ipsilateral extremity surgeries ⁹.

Standard management of early postoperative or acute hematogenous infections following TKA may begin with surgical debridement, retention of components, and administration of oral antibiotics. In their study of 104 patients, Azzam et al. found this treatment to be effective in eradicating infection in 44% of cases¹⁴. In the case of more chronic infections, explantation with an antibiotic spacer for a period of time followed by reimplantation may be necessary. Definitive care in the worst cases of uncontrollable infection may consist of knee arthrodesis, particularly in patients whose leg extensor mechanisms are compromised, or for whom inadequate soft tissue coverage makes other surgical options untenable¹⁵.

The results of several techniques for knee arthrodesis following infected TKA have been described. The surgical option with the most successful outcomes is intramedullary (IM) nailing, with reported fusion rates of 80-100%^16-18. However, IM nailing is contraindicated by the presence of active infection and the absence of a medullary canal. In such patients, bony fusion may be achieved through compressive external fixation, though this technique is not without its own limitations. Fusion rates are generally lower than those achieved with IM nailing, and complication rates are high^19-22. The purpose of the present study is to present the senior author’s experience with a six axis circular external fixation frame (Taylor Spatial Frame (TSF), Smith & Nephew, Memphis, TN) used for limb salvage in patients with chronically infected TKAs and either catastrophic failure of the leg extensor mechanism, or substantial loss of anterior soft tissue coverage. This construct is a unique tool, which allows bedside adjustment over time to alter the degree of compression and alignment of the knee joint during the fusion process.

Materials and Methods

Our Institutional Review Board approved this study. Patients were included who met the following inclusion criteria: 1) adult patients who presented with a chronic, deep periprosthetic infection of a total knee arthroplasty, 2) failure of the knee extensor mechanism, or substantial anterior soft tissue loss and 3) use of the TSF for arthrodesis. Patients were excluded if they were under the age of 18 years at the time of surgery.

Inpatient and outpatient records were reviewed for all patients. Patient demographics, past medical histories and surgical histories were reviewed. Anteroposterior and lateral radiographs taken at regular follow-up intervals of 1, 3, 6, 12 and 24 months post-operatively were reviewed by a fellowship-trained orthopaedic trauma-tologist to determine time to bony fusion. Intraoperative cultures were reviewed to determine the identities of the infecting organisms.

Surgical Technique

The established midline knee incision was utilized to gain access to the knee. All prosthetic components and cement spacers if present were removed. Deep soft tissue and bone cultures were obtained. Non-viable soft tissues were sharply debrided and the surgical sites were irrigated with 6-9 liters of saline lavage. The medullary canals were irrigated and debrided as well. In some cases the patella was excised. Fresh flat bony cuts were made at the distal end of the femur and proximal tibia using a sagittal saw leaving two flat surfaces for compression. The soft tissue envelope was then closed with monofilament heavy sutures over large drains. Next the frame was applied to help stabilize the soft tissues and eventually to be used to gain fusion.

One full carbon fiber ring was centered over the distal end of the femur and attached using three, 6.0 mm hy- droxyapatite coated half pins drilled bi-cortically through the bone and fixed to the ring with Rancho cubes. Position of the ring perpendicular to the bone was confirmed on anteroposterior and lateral intraoperative radiographs. A second ring was applied to the proximal tibia in a similar manner. Next the two rings were connected with 6 telescoping struts applied in a standard manner (Figure 1a,b). With the frame assembly complete, final radiographs were obtained at the end of surgery to be used for frame management post operatively.

Fig 1. AP (1a) and lateral (1b) view of a saw bone model depicting the standard placement of the TSF.

Following surgery, a computer-based program was used to develop a plan for compression alignment at the arthrodesis site. Pin sites were dressed with sterile bandages initially. Following wound healing they were covered with sponges and cleaned daily with soap and water. Patients were allowed to be full weightbearing on the limb in the frame. Serial radiographs were obtained during the correction phase of treatment and then every 2-3 months following to assess healing.

The treating surgeon and an infectious disease specialist followed all patients. All patients were treated with long-term culture specific, intravenous antibiotics. Complications were defined as sequelae of the arthrodesis that required further medical or surgical treatment.

Results

General Demographics

Between 2002 and 2010, 7 eligible patients were identified and constituted the study population (Table 1) . There were 4 females and 3 males. The average time between TKA and diagnosis of infection was 30.7 months (range, 2.6 – 67.0 months). Average patient age at the time of application of the TSF was 70.9 years (range, 59 – 83 years). Six arthrodeses were performed on right knees and 1 on a left knee. The 7 patients had undergone an average of 3.3 prior surgical procedures (range, 2-4 procedures) on the ipsilateral extremity. The primary TKA was indicated for treatment of osteoarthritis in 6 patients and rheumatoid arthritis in 1 patient. Four of the primary TKAs were performed at the same institution as that in which the arthrodeses were performed; 3 patients underwent primary TKAs at a different institution. All patients had been diagnosed with deep, chronic peri- prosthetic infections on the basis of elevated ESR/CRP levels and clinical findings, including gross purulence. All patients had developed a non-traumatic disruption of the extensor mechanism in the affected limb. Thus in all patients, indications for arthrodesis included persistent infection and an incompetent, un-reconstructable extensor mechanism.

Table 1.

Patient demographics, durations of treatment and outcomes

Patient	Age/Sex	Prior Procedures (no.)	Antibiotic Spacer (wk.)	External Fixator (mo.)	Infection Resolved	Bony Fusion
1	73/M	4	7.86	7.93	Yes	Yes
2	59/F	3	7.29	7.70	Yes	No
3	75/F	3	5.56	10.40	Yes	Yes
4	82/M	4	1.43	6.67	Yes	Yes
5	83/F	3	n/a	5.97	Yes	No
6	62/F	4	1.57	9.50	Yes	Yes
7	62/M	2	6.0	10.5	Yes	Yes
Avg.	70.9 yrs.	3.3 procedures	4.95 wk.	8.4 mo.	7/7 (100%)	5/7 (71%)

The average length of surgery was 222 minutes (range, 91 – 450 minutes). The average length of hospital stay was 25 days (range, 19 – 32 days). Average follow-up was 20.8 months (range, 7.9 – 64.6 months). Of the 7 patients, 6 had been treated with a surgically implanted antibiotic spacer before undergoing arthrodesis for an average time of 4.95 weeks (range, 1.4 – 7.3 weeks). In the 1 remaining patient, the prosthesis used in the primary TKA was removed during the same surgery in which the TSF was applied.

Arthrodesis Results

Fusion was achieved in 5 of 7 patients (71%) at an average of 8.4 months (range, 6 – 10.5 months) after application of the TSF (Figure 2a-h). The remaining 2 patients never achieved bony fusion. One of these two patients was an 83 year-old female who requested removal of the TSF at the 6-month follow-up visit, despite exhibiting no radiographic evidence of bony union. This patient was undergoing intensive treatment for metastatic cancer and refused further elective procedures. At 8 month follow-up, this patient exhibited reduced range of motion at the knee and reported ambulating in her home with the assistance of a brace. Her limb was free of infection clinically and she denied knee pain. The second patient who was considered a treatment failure was a 59-year-old female with a history of chronic osteomyelitis of the ipsilateral tibia and femur. In this patient, the TSF was removed at 7.7 months after surgery, when it was determined that further fusion was unlikely. At the most recent follow-up visit of 64.6 months after surgery, this patient continued to experience almost no knee pain. Although the knee remained unstable, the patient reported being able to ambulate in her home with the use of a brace, and even to dance slowly on occasion.

Fig 2. AP and lateral radiographs of the right knee of a 62 year old female after total knee arthroplasty (2a,b), after removal of the knee prostheses and insertion of an antibiotic spacer (2c,d), after application of the TSF (2e,f), and at 16-month follow-up (2g,h). The significant anterior defect is secondary to a previously failed knee extensor mechanism. At final follow-up, the patient had a stable, pain free extremity.

The cause of infection was identified as a specific organism in 6 patients, and polymicrobial in 1 patient (Table 2). Infections resolved in all of the 7 patients. None of the 7 patients underwent further reconstructive surgeries or salvage procedures of the knee to date.

Table 2.

Results of intraoperative cultures

Patient	Infecting Organism
1	Serratia marcescens
2	Staphylococcus aureus
3	Enterobacter cloacae
4	Polymicrobial
5	Candida albicans
6	Staphylococcus aureus
7	Staphylococcus aureus

Complications

Complications associated with this surgical procedure occurred in the treatment of 5 of the 7 patients (71%). The most common of these was infection at the site of the pin track, which occurred in 4 patients (Table 3). In all cases, pin track infections resolved with local wound care or surgical debridement and pin exchange. These patients were also given supplementary oral antibiotics for coverage of gram positive bacteria if intravenous antibiotics were covering a different organism.

Table 3.

Complications of treatment with the TSF

Patient	Complications
1	None
2	Pin track infection
3	Pin track infection, femur fracture
4	Pin track infection, antibiotic-induced acute renal failure
5	None
6	Pin track infection x2, wound break-down requiring flap closure
7	Wound breakdown
Complication Rate	5/7 (71%)

One of the 4 patients underwent additional procedures at the time of pin exchange surgery. This patient had previously sustained a fracture of the ipsilateral medial femoral condyle after falling in the shower at the rehabilitation facility prior to placement of the spatial frame. A secondary bone graft procedure at both the acute fracture and arthrodesis sites was used to aid in healing and to achieve successful fusion.

One of the 4 patients developed a wound breakdown 2 months following TSF removal. This wound was successfully treated with irrigation and debridement and closure with a local gastrocnemius muscle flap.

Discussion

This series reports the use of a multiplanar circular external fixator for the purpose of gaining fusion following severe chronically infected TKAs with an absent extensor mechanism. These cases represent one of the worst case scenarios following failed TKA. In 1971, Nelson and Evarts first described knee arthrodesis as a treatment for patients with failed TKAs¹⁹. This was a time when the ability to treat infections of primary TKAs was limited, leading the authors to advocate for arthrodesis so broadly. In fact, Nelson and Evarts did not differentiate between TKAs that failed due to infection from those that failed for other reasons; the treatment for either case was arthrodesis. Since then, treatment for infected TKA has evolved, and many techniques have been described to achieve such an arthrodesis. While IM nailing has produced satisfactory results, the procedure is not advisable in patients with ongoing medullary infections^16-18. In these cases, the orthopaedic surgeon must rely on external fixation in an attempt to fuse the knee and salvage the affected lower extremity.

To this end, various groups have presented their results with techniques for external fixation of the knee for arthrodesis. In 1983, Rothacker et al. reported outcomes in 29 patients whose knees were fused utilizing the Hoffmann device. In this retrospective study, 86.2% of patients achieved fusion, while the complication rate was 20.6%(²²). Oostenbroek et al. reported a slightly improved fusion rate of 93% in their study of 15 patients whose knees were fused using the Ilizarov technique. This cohort had a surgical complication rate of 80%, though 100% of patients developed pin track infections during the study period²¹. Most recently, Klinger et al. presented their experience with 18 patients. The fusion rate was 89%, and 100% of patients developed pin track infections²⁰.

The knees of 5 of the 7 patients in the present study successfully fused, yielding a fusion rate of 71%. This rate is consistent with those reported in previous studies, particularly in light of the fact that 1 of the 2 patients who is considered a treatment failure elected to remove the external fixator because of discomfort incurred during oncological treatment. Given that these 7 cases all represented patients with complete loss of the extensor mechanism, several with extensive necrosis and substantial anterior soft tissue loss, this rate may represent an improvement, as these cases would have been high risk for failure using other techniques. Further, we are encouraged that neither of the patients who is considered a treatment failure has undergone subsequent above-knee amputation, a procedure to which both patients objected strongly. Both patients are functioning at a level that is acceptable to them, and neither is completely bed or wheelchair bound. The complication rate of 71% is also consistent with those seen in similar cohorts. While these sequelae are clearly undesirable, they remain an unavoidable aspect of treating patients with difficult periprosthetic infections.

An additional benefit of the TSF is its inherent versatility. The frame has been successfully used to correct complex post-traumatic^23,24 and hereditary²⁵ deformities of the lower extremity. While similar to a traditional external fixator, the TSF has a significant advantage in its ability to be dynamically adjusted as well. Modifying the struts individually may allow the surgeon to produce forces that would require modifying the entire frame of other constructs. Such gradual strut corrections in the TSF have been shown to be effective over time²⁶. This aspect of the implant allowed us to make adjustments and corrections during the healing phase and to add compression to the arthrodesis site as needed post operatively.

This case series has several limitations that are inherent to the retrospective review of existing clinical data. One notable example is the lack of subjective, patient-based outcome measures. While questionnaires that provide such data are an important part of orthopaedic research, it is unclear which questionnaire would be best suited for this patient population. To date, 3 groups have utilized the Short Form 36 (SF-36) questionnaire in studying patients after knee arthrodesis for infected TKAs^20,27,28. On the one hand, these studies have provided a useful basis for outcome comparisons to future studies. On the other hand, Benson et al. suggested that the Arthritis Impact Measurement Scale (AIMS) may be a more appropriate indicator of patient outcomes in this patient population than is the SF-36²⁷. Without further research, it is difficult to know which outcomes tool to use.

The present study presents our group’s experience with the Taylor Spatial Frame, a multi-planar external fixator used for arthrodesis and limb salvage in patients with worst case scenario, chronically infected TKAs. Patient outcomes were generally satisfactory, and all patients avoided above-knee amputation. This technique adds another viable option which orthopaedic surgeons should consider when treating patients in this difficult population.