We identified five (2.3%) fractures of the stem in a series of 219 revision procedures using a cementless, cylindrical, extensively porous-coated, distally-fixed femoral stem. Factors relating to the patients, the implant and the operations were compared with those with intact stems. Finite-element analysis was performed on two of the fractured implants.

Factors associated with fracture of the stem were poor proximal bone support (type III–type IV; p = 0.001), a body mass index > 30; (p = 0.014), a smaller diameter of stem (< 13.5 mm; p = 0.007) and the use of an extended trochanteric osteotomy (ETO 4/5: p = 0.028). Finite-element analysis showed that the highest stresses on the stem occurred adjacent to the site of the fracture. The use of a strut graft wired over an extended trochanteric osteotomy in patients lacking proximal femoral cortical support decreased the stresses on the stem by 48%.

We recommend the use of a strut allograft in conjunction with an extended trochanteric osteotomy in patients with poor proximal femoral bone stock.

We report the five- to-ten year results of Anderson Orthopaedic Research Institute type-2 bone defects treated with modular metal augments in revision knee surgery.

A total of 102 revision knee arthroplasties in patients with type-2 defects treated with augments and stems were prospectively studied. Seven patients (seven knees) had incomplete follow-up and 15 patients (16 knees) died with the arthroplasty in situ. The mean follow-up of the 79 remaining knees was 7 ± 2 years (5 to 11). The presence of non-progressive radiolucent lines around the augment in 14% of knees was not associated with poorer knee scores, the range of movement, survival of the component or the type of insert which was used (p > 0.05). The survival of the components was 92 ± 0.03% at 11 years (95% CI, 10.3 to 11.2).

We recommend the use of modular augmentation devices to treat type-2 defects in revision knee surgery.

1. Thirty-seven cases of fracture of the dens have been studied.

2. The incidence of non-union was high: 64 per cent after apparently adequate closed treatment.

3. Possible causes of the high incidence of non-union have been studied : attention is drawn to the effect of displacement and to that of posterior displacement in particular.

4. Non-union of the dens with potential instability at the atlanto-axial joint is not acceptable in a patient who expects to lead a normal active life.

5. Atlanto-axial fusion is the method of choice in the treatment of instability ; once that has been secured, pseudarthrosis of the dens is no longer significant.