We investigated the clinical and dosimetric predictors for radiation-associated femoral fractures in patients with proximal lower extremity soft tissue sarcomas (STS).

We examined 131 patients with proximal lower extremity STS who received limb-sparing surgery and external-beam radiation therapy between 1985 and 2006. Five (4%) patients sustained pathologic femoral fractures. Dosimetric analysis was limited to 4 fracture patients with full three-dimensional dose information, who were compared with 59 nonfracture patients. The mean doses and volumes of bone (V(d)) receiving specified doses (≥30 Gy, 45 Gy, 60 Gy) at the femoral body, femoral neck, intertrochanteric region, and subtrochanteric region were compared. Clinical predictive factors were also evaluated.

Of 4 fracture patients in our dosimetric series, there were three femoral neck fractures with a mean dose of 57.6 ± 8.9 Gy, V30 of 14.5 ± 2.3 cc, V45 of 11.8 ± 1.1 cc, and V60 of 7.2 ± 2.2 cc at the femoral neck compared with 22.9 ± 20.8 Gy, 4.8 ± 5.6 cc, 2.5 ± 3.9 cc, and 0.8 ± 2.7 cc, respectively, for nonfracture patients (p < 0.03 for all). The femoral neck fracture rate was higher than at the subtrochanteric region despite lower mean doses at these subregions. All fracture sites received mean doses greater than 40 Gy. Also, with our policy of prophylactic femoral intramedullary nailing for high-risk patients, there was no significant difference in fracture rates between patients with and without periosteal excision. There were no significant differences in age, sex, tumor size, timing of radiation therapy, and use of chemotherapy between fracture and nonfracture patients.

These dose-volume toxicity relationships provide RT optimization goals to guide future efforts for reducing pathologic fracture rates. Prophylactic femoral intramedullary nailing may also reduce fracture risk for susceptible patients.