Emerging evidence has linked the long-term use of alendronate (fosamax) with subtrochanteric insufficiency fractures. However, findings to date have been anecdotal. The aims of this study were to determine the incidence of subtrochanteric insufficiency fractures and identify whether they were more prevalent following the introduction of alendronate in Australia.

All patients that presented between January 2007 and February 2009 with low- energy subtrochanteric fracture were identified. Similar data were collected between January 1995 and February 1997 as this was immediately prior to introduction of alendronate in Australia. The radiographs were examined for failure due to pre- existing insufficiency fracture. Characteristic findings were a transverse fracture line on the tension side of the femur with lateral cortical thickening immediately adjacent to the fracture. Relevant details from the history were recorded. We also separately identified all patients that presented between 2007 and 2009 with a proximal femoral fracture and determined the proportion taking alendronate.

One hundred and seventeen patients with low-energy subtrochanteric fracture were included. Seventy-nine patients presented between 2007 and 2009 and 38 presented between 1995 and 1997. Forty-one of the 79 (52%) patients were identified as having radiograph findings suggestive of underlying insufficiency fracture, whilst none were identified prior to the introduction of alendronate. Of the 41 patients with subtrochanteric insufficiency fracture, 40 (98%) had been taking alendronate and one had been taking risedronate. Twenty-nine of the 41 (71%) complained of prodromal pain in the affected femur. Eighteen of the 41 (44%) demonstrated subtrochanteric insufficiency changes on the contralateral side and 9 of 41 (22%) sustained spontaneous non-traumatic fracture during activities of daily living. Of the 38 patients without insufficiency changes, 12 (32%) had been taking alendronate. Alendronate use was therefore strongly suggestive of insufficiency fracture (sensitivity = 98%, specificity = 84%, PPV = 77%, NPV = 99%, LR+ = 6). The mean duration of alendronate use in those with insufficiency fracture was 7.1 years (95% CI, 6.6-7.6 years). The mean duration in those without was 3.2 years (95% CI, 2.6-3.8 years, P<0.0001). Three hundred and ninety eight patients presented with a low-energy proximal femur fracture between 2007 and 2009. Of these, only 52 (13%, P<0.0001) were taking alendronate.

This is the largest study in the literature on subtrochanteric insufficiency fractures and alendronate therapy. Confirming recent reports, alendronate use was strongly suggestive of subtrochanteric insufficiency fracture. Our findings provide the most compelling evidence to date of the potential long-term sequelae of alendronate but more research is needed before definitive conclusions can be made.

A finite element study was carried out to compare the performance of a three-hole locking plate with angled screws to the ‘gold-standard’ four-hole hip plate. Two cases of the three-hole hip plate were examined; (a) three screws and (b) two screws (most proximal and most distal).

A 3D model of the proximal femur was constructed from CT scans. A 3D CAD model of the four-hole hip plate was also created. The three-hole hip plate was then created from the four-hole implant in a way that it was possible to switch between all three models by activating/deactivating sections and/or switching material properties. A single common finite element model was generated, and a static analysis of each model variation was then performed in two steps using ABAQUS/standard. In the first, screws were pre-tensioned up to 150N. In the second, loads corresponding to stair climbing were applied.

Forces in the screws, permitted to change in the second step, were examined and compared. Maximum principal stresses in the bone were also examined, with a focus on the stresses in the bone at the end of the plate in each model. The highest tensile force was in the proximal screw of the three-hole plate with three screws, followed by the most distal screw in the standard four-hole plate. This suggests that the risk of screw pull-out is highest at the proximal screw of the three-hole hip plate with three screws.

A comparison of the forces in the distal screws for all cases shows that the highest tensile force was in the four-hole plate, followed by the three-hole plate with two screws. The lowest was the three-hole plate with three screws, which was in compression at full load. The maximum tensile stresses in the bone at the end of the plate were greatest for the standard four-hole hip plate, followed by the three-hole plate with two screws and then the three-hole plate with three screws. This indicates that the risk of bone fracture at the end of the plate is lowest for the three-hole hip plate with three screws.

The risk of bone fracture is significantly lower for the three-hole hip plate, with either two or three screws, compared to the ‘gold-standard’ four-hole hip plate. This is partially offset by a small increase in the risk of screw pull out (in the proximal rather than the distal screw).