Purpose of the study: External fixation has not been widely used for femoral fractures and few series are reported in the literature. External fixation is generally reserved for severe open fractures, for vessel injury or multiple trauma with life threatening. We present a retrospective analysis of a serie treated in a single center in order to detail the indications of this fixation technique. Material and methods: From 1984 to Jun 2002, 49 patients with femoral fractures were treated by external fixation. The series included 36 men and 13 women, mean age 31 years. All were victims of high-energy trauma: traffic accident (n = 40), fall from high level (n = 4), firearm wound (n = 5). Multiple fractures were present in all patients except seven and 24 patients had multiple injuries. Forty fractures were open fractures: two type 1, ten type 2, four type 3a, 23 type 3b and five type 3c in the Gustilo classification. Twenty-seven were shaft fractures and 26 involved the distal metaphyseoepiphyseal portion of the femur. Loss of cortical stock was noted in five cases and total loss of a segment in four. Surgery was deferred in 19 patients, mean six days. A single-plane external fixation was used (Orthofix) with a femorofemoral frontolatateral assembly. Transepiphyseal screw fixation was also used to stabilize the distal fracture in eleven cases. Results: One patient with a bifocal fracture of the femur died from head trauma. Three patients required above knee amputation after failure of a vessel bypass or due to septic necrosis of the reconstruction flap. Five patients required a second reduction within days of external fixation. On the AP view, femoral alignment was successfully reestablished at +/- 51 in 45 cases, ranged from 50 to 10 degrees in seven and was greater than 10 degrees in one. On the lateral view, alignment was between 50 and 10 degrees in 42 cases and greater than 10 degrees in one. Femur length was equal to the healthy side in 23 cases, and was shortened 1-2 cm in 26. Four metaphyseal fractures resulted in a 3 cm shortening. Bone healing time was available for 42 patients (1 death, 3 amputations, 3 lost to follow-up). Elective conversion to internal fixation was performed in ten patients (five lateral cortical plates and five centromedullary nailings). These patients all achieved first-intention bone healing with a mean time of 7.4 months. Exclusive external fixation was planned for 34 fractures. First-intention healing was achieved in 25 (17 shaft and 8 distal) without bone graft with an average time of 7.3 months. Ten patients had one or more osteitis foci on pin tracts. Two patients in this group developed recurrent fracture after removal of the external fixator. Nine fractures did not heal and required revision with centromedullary nailing (n = 5) or plate fixation with autograft (n = 4). Nailings for nonunion were successful but plate fixation was compromised by infection in one patient and recurrent fracture after plate removal in another. Fourteen patients underwent joint mobilization under general anesthesia and 14 had open arthrolysis. Mean follow-up was 2.8 years. Mean active flexion was 90 degrees (30-130 degrees). Ten patients exhibited flexion between 30 degrees and 60 degrees and 19 between 70 degrees and 100 degrees. Knee flexion was greater than 110 degrees in 15 patients. Residual 10 degrees flexion was noted in six knees. Mean leg length discrepancy was 0.4 +/- 0.6 after distal fracture and 0.8 +/- 1.3 after diaphyseal fracture. Discussion: The indications and results of external fixation in this series are in line with reports in the literature. For diaphyseal fractures, healing is long and difficult, partly because of the insufficient mechanical properties of external fixation. The rate of infection and stiff knee is high, particularly for distal fractures of the femur.
