Operative management of a proximal tibial nonunion is challenging, compromised by limited bone stock, pre-existing hardware, stiffness, poor soft tissue conditions, and infection. The goals of treatment include bone union, re-establishment of both joint stability and lower extremity alignment, restoration of an anatomic articular surface, and recovery of function.1 Currently, various treatment options such as plate fixation, bone grafting, intramedullary nailing, external fixation, functional bracing, or a combination of these are available.1-8 Rigid internal fixation is the gold standard for most nonunions. However, sometimes local soft tissues or bone quality preclude standard internal fixation. Bolhofner9 described the combination of a single plate and an external fixator on the contralateral side for the management of extra-articular proximal tibial fractures with compromised soft tissues, and the technique known as composite fixation was coined. The external fixator on 1 side and the plate on the other, generate a balanced, stable environment while limiting the use of foreign hardware, thereby avoiding both additional soft-tissue damage and periosteal stripping.9-11 In this technical article, we describe the indication, technique, and outcomes of 3 patients with proximal tibial nonunions, who were successfully treated with composite fixation.
MATERIALS AND METHODS
PATIENTS
Between January 2014 and July 2016, 3 patients each with a proximal tibial nonunion that developed after a bicondylar tibial plateau fracture (Schatzker type VI) were treated with composite fixation (Table). The 3 patients were female with an average age of 61 years (range, 60-62 years), and a body mass index of 23.7 kg/m2 (range, 19.0-31.9 kg/m2). All 3 patients had sustained a tibial plateau fracture that was primarily treated with open reduction and internal fixation. Two of them had a diagnosis of rheumatoid arthritis and were being treated with methotrexate and Humira (adalimumab) (case 1), and with methotrexate, prednisolone, and etanercept (case 3). The etanercept was discontinued after discussion with the treating rheumatologist when a deep infection developed. Two patients (cases 1 and 2) were referred to us because of their nonunions. All 3 patients developed extra-articular nonunions with compromised bone stock. Two patients had developed deep infections during treatment of their plateau fractures; 1 of these patients underwent a medial gastrocnemius flap for wound coverage (case 1). The second patient (case 3) with a deep infection underwent partial hardware removal, a Masquelet salvage procedure, and revision plate fixation. However, the infection recurred. The hardware was removed, and 2 débridements with conversion to a hybrid external fixator with thin wire fixation were done. Due to her longstanding rheumatoid arthritis, the patient had bilateral valgus knee malalignment causing the ring fixator to strike her contralateral knee when she walked. The period from the initial tibial plateau fracture to our composite fixation averaged 11.3 months (range, 11-12 months). Indications for the use of the composite fixation comprised previously infected soft tissue on the lateral side and inability to walk with a hybrid thin wire fixator because of valgus knees (case 3), a medial gastrocnemius flap (case 2), and poor bone quality (case 1). Follow-up consisted of clinical examination, Timed Up and Go (TUG) test that is a standardized test for mobility, and radiographic evaluation at routine appointments up to 1 year or until healed.12 At the last follow-up visit, patients filled out the International Knee Documentation Committee (IKDC) subjective knee form.13
SURGICAL TECHNIQUE
A fellowship-trained orthopedic trauma surgeon treated all patients. Patients were placed on a radiolucent operating table after general or regional anesthesia. Previous incisions were used. Two patients had a midline incision; the third had both a posteromedial and an anterolateral incision. Five deep tissue cultures were taken after which antibiotics were given intravenously. All unstable or failed hardware was removed. Aggressive débridement of the nonunion was performed. After débridement, multiple holes were drilled with a 2.0 mm drill bit until blood was seen to egress from both sides of the medullary canal. Malalignment of the proximal tibia was corrected and checked fluoroscopically. Fixation was done with an anatomic locking plate (LCP Proximal Tibia Plate 3.5; DePuy Synthes) with a mixture of locking and non-locking screws. In 2 patients, a tricortical graft from the posterior iliac crest was positioned in the defect. Additional autologous bone graft and demineralized bone matrix was added around the nonunion. Although locking screws were used, the fixation did not appear to be strong enough to resist the varus (cases 1 and 2), or the valgus (case 3) deforming forces. Additional fixation was thus needed. However, the contralateral soft tissues were compromised in case 2 (medial gastrocnemius flap), and case 3 (a previously infected area with very tenuous skin laterally), whereas the bone was considered to be of insufficient quality in case 1. The opposite side of the nonunion was stabilized using composite fixation with a 2-pin external fixator to circumvent the need for additional plate fixation. In 2 patients, the plate was placed laterally, and the external fixator medially. In the third patient, the plate was positioned medially, and the external fixator laterally. The plate was always placed first. The external fixator was placed last. Using fluoroscopy, we ensured that the fixator pins would not interfere with the screws. The pins were predrilled and positioned perpendicular to the tibia through small stab incisions. We prefer hydroxyapatite-coated pins (6-mm diameter, XCaliber Bone Screws; Pro-Motion Medical) to increase their holding power in the often osteopenic bone. Postoperative management consisted of toe-touch weight-bearing for 6 weeks and progressed to full weight-bearing at 3 months. Radiographs were taken on postoperative day 1, at 6 weeks, and at 12 weeks until healed. No continuous passive motion was used postoperatively. Antibiotics were continued until cultures were negative. No specific pin care was used. We advised patients to shower daily with the external fixator in place, once the wounds have healed.
Continue to: RESULTS...