Background: Fibrous dysplasia of the proximal femur is a progressive, often recurrent condition of bone that can cause skeletal deformity, fractures, and pain [corrected]. Allogeneic cortical strut grafting to minimize the risk of fracture or as part of fracture treatment is a promising treatment option, but evidence is scarce on the intermediate- to long-term results of this procedure and there are no data on factors associated with graft failure.
Questions/purposes: The purposes of this study were (1) to evaluate the revision-free survivorship; (2) radiographic findings; (3) factors associated with failure; and (4) complications associated with cortical strut allograft to prevent or treat fractures of the proximal femur in patients with fibrous dysplasia.
Methods: Between 1980 and 2013 we performed cortical strut allografting in 30 patients for impending or actual fractures of the proximal femur, of whom 28 (93%) were available for followup at a minimum of 2 years (mean, 13 years; range, 4-37 years) and of whom 22 (73%) had also been evaluated within the last 5 years. During that time, the indications for cortical strut allografting were an impending fracture of the proximal femur, persistent pain, or an actual nondisplaced femoral fracture. In patients who presented with a diaphyseal fracture, a fracture with severe dislocation of severe varus deformity, which required an osteotomy, placement of a blade plate was instead performed and these patients are not included here. During that time, for patients with diaphyseal fractures, and in patients with a displaced femoral fracture of the proximal femur, placement of a blade plate without strut grafting was instead performed; these patients are not included here. The primary outcome was the success rate of allogeneic cortical strut grafting surgery as assessed by the absence of revision surgery for a newly sustained fracture, resorption of the graft, or progressive deformity of the proximal femur. The association of possible contributing factors to graft failure such as gender, age at surgery, preoperative fracture, and anchoring distances of the graft in healthy bone was also evaluated using Cox regression analysis.
Results: Revision surgery was performed in 13 patients, resulting in a mean survival time of 13 years (Kaplan-Meier 95% confidence interval [CI], 10-16). Radiological resorption of the graft was observed in 15 of 28 patients (54%). However, revision surgery was not performed in all patients who developed graft resorption, because of the absence of a risk for fracture on the basis of the anatomical site of resorption. Identified risk factors for graft failure included preoperative fractures (hazard ratio [HR], 4.5; 95% CI, 1.2-17.2; p = 0.028) and insufficient proximal anchoring of the graft in healthy bone (HR, 6.02; 95% CI, 1.3-27; p = 0.02). One patient sustained a refracture after surgery resulting from an in-hospital fall. The fracture was treated without further surgery, and it healed.
Conclusions: Our findings from this study suggest that cortical strut allografting may be a viable option for treatment of fibrous dysplasia of the proximal femur a without previous pathological fracture. Surgeons should pay particular attention to the proximal fixation point of the allograft to decrease the risk of failure. Patients with a fracture have an increased risk of failure and reoperation and so should be treated with an osteosynthesis.
Level of evidence: Level IV, therapeutic study.