There have been recent reports linking alendronate and a specific pattern of subtrochanteric insufficiency fracture. We performed a retrospective review of all subtrochanteric fractures admitted to our institution between 2001 and 2007. There were 20 patients who met the inclusion criteria, 12 of whom were on long-term alendronate. Alendronate-associated fractures tend to be bilateral (Fisher’s exact test, p = 0.018), have unique radiological features (p < 0.0005), be associated radiologically with a pre-existing ellipsoid thickening of the lateral femoral cortex and are likely to be preceded by prodromal pain. Biomechanical investigations did not suggest overt metabolic bone disease. Only one patient on alendronate had osteoporosis prior to the start of therapy. We used these findings to develop a management protocol to optimise fracture healing. We also advocate careful surveillance in individuals at-risk, and present our experience with screening and prophylactic fixation in selected patients.

Recent reports suggest that long-term alendronate therapy may result in an unusual pattern of femoral subtrochanteric fracture. We aimed to determine if the presence of a specific radiographic feature in patients on alendronate could be used to predict an impending insufficiency fracture and thereby prevent its occurrence through further investigations and prophylactic fixation in high-risk patients.

Sixty-two subtrochanteric fractures treated surgically from 2001 to 2007 were reviewed and radiographs of 25 low-energy fractures were independently evaluated. Incidence of alendronate therapy, clinical data, and other investigations like bone mineral density (BMD) scans were recorded.

Seventeen fractures (68%) were associated with alendronate therapy. Hypertrophy of the lateral cortex of the femur with splaying of the fracture ends was noted in 70.1% of patients on alendronate; initial radiographs were not available in 17.6% and 11.8% had stress fractures identified by bone scan. None of the fractures in the non-alendronate group had this pattern. The fracture configuration in the alendronate group suggested that an ellipsoid thickening in the lateral cortex had been present prior to fracture. Indeed, 6 patients on alendronate (35.3%) had pre-existing radiographs as early as 3 years prior to fracture and all had this feature. Four of them had bone scans, which confirmed a stress fracture. Hip pain was often associated with this radiographic sign but may not be specific as patients were already on follow-up for other musculoskeletal conditions. BMD scans were not predictive of an impending fracture as they were mostly in the osteopaenic range. Only 50% with proven stress fractures had prophylactic fixation, while the remainder sustained overt fractures.

Alendronate-related subtrochanteric fractures are associated with a specific pre-existing radiographic abnormality. We recommend that all patients on long-term alendronate - particularly those with hip pain or a previous subtrochanteric fracture - be routinely followed-up with plain radiographs of the pelvis. If an ellipsoid feature is noted in the subtrochanteric region, further investigations like bone scan or MRI should be sought. Patients with evidence of stress fracture should be strongly considered for prophylactic operative fixation. We believe this is a cost-effective strategy to prevent subtrochanteric insufficiency fractures in patients on alendronate.