Session V - Femur

Focal Osteolysis at the Junctions of a Three-Piece, Modular, Femoral Intramedullary Nail

Darron M. Jones, MD; John L. Marsh, MD; James V. Nepola, MD, University of Iowa Hospitals & Clinics, Iowa City, IA; J. J. Jacobs; R.M. Urban; A.K. Skipor, Rush Medical College, Chicago, IL

Purpose: To determine the incidence, etiology, and significance of radiographic osteolysis and periosteal reaction around the junctions of a three-piece stainless steel modular femoral nail.

Methods: Between September 1995 and December 1996, 42 patients with 44 femoral fractures were treated with an FDA-approved TriMax stainless steel modular femoral nail (Smith + Nephew Richards, Memphis, TN). Based on the finding of radiographic osteolysis at the junctions of some of these nails, the following study was begun. Eighteen nails were excluded (8 tumor, 1 infection, 1 broken nail, 4 removed prior to starting our protocol, and 4 in patients who refused to return) leaving 26 patients with 26 intramedullary nails for this study. These 26 patients were invited to return for radiographs, a physical examination, functional outcomes assessment, a pain analog scale for thigh pain, serum chromium levels, and hardware removal if desired. In nine of these patients, nails were removed according to the study protocol. The nails were analyzed for corrosion products, and histopathological analysis of curettings from the femoral canal was performed. A control group of patients treated with a one-piece stainless steel IM nail were evaluated using the same outcome measures and compared to the modular femoral nail group for incidence of thigh pain and serum chromium levels.

Results: The 26 patients were seen at a mean of 22 months post-fracture. Radiographs revealed that 25 fractures were healed, and in 22 femurs there was at least one of three types of abnormalities localized to one or both junctions: osteolysis, periosteal reaction or cortical thickening. Nine patients complained of thigh pain greater than four on a 1-10 visual analog scale. Serum chromium levels averaged 1.21 ng/ml (range = 0.15 to 3.12 ng/ml). In the control group with one piece nails the mean was 0.48 ng/ml (range = 0.015 to 1.25 ng/ml) (p<0.03) and in a group of patients without implants the mean was 0.05 ng/ml(range = 0.015 to.0.25 ng/ml) (p<0.01). There was a trend for patients with higher serum chromium levels to have more osteolysis and more thigh pain. Corrosion products were seen at the junction where the osteolysis occurred and metal debris was present in the tissue samples.

Discussion/Conclusion: This device was designed for intraoperative assembly to decrease inventory and to allow antegrade or retrograde insertion. Each method of insertion had a unique proximal and distal end that assembled to a common shaft by a taper mechanism. These tapers were then secured with a locking bolt through the center of the nail.

The etiology of the radiographic findings of osteolysis and periosteal reaction is not specifically known, but the presence of corrosion products at the taper junctions suggests particulate debris as a factor. There was an association between the radiographic abnormalities (osteolysis and periosteal reaction) and thigh pain and elevated serum chromium levels, indicating that these findings have clinical significance. Serial follow-up of this cohort of patients is indicated and is currently underway.