We have carried out a prospective study of 17 patients (14 women, 3 men) of mean age 48 years (21 to 76) with transcervical fractures of the femur using MRI to detect early evidence of avascular necrosis of the head.

Two fractures were Garden stage I, 12 stage II, and three stage III. We performed internal fixation under radiological control at a mean of five days (2 to 15) after injury using a titanium cannulated cancellous screw or a titanium compression hip screw. MRI was performed at one, six and 12 months and then yearly after operation. T1- and T2-weighted images were obtained by a spin-echo technique. The duration of follow-up of patients who did not subsequently require replacement of the head of the femur was from 2 to 5 years (mean 3.2).

One month after operation eight of the 17 hips showed a band of low signal intensity on T1-weighted images and high signal intensity on T2-weighted images indicating lesions in the femoral head away from the fracture line. These were of three types: type I was a small infarct at the superolateral region of the femoral head and was seen in three hips; type II was a shallow lesion from the superolateral region to the fovea of the femoral head (three hips); and type III was a large lesion occupying most of the femoral head (two hips). No further changes were seen in the MRI after six months from operation. Collapse of the femoral head did not occur in the three hips with type-I lesions, but two of the three type-II hips and both type-III hips subsequently collapsed. At the final follow-up the three hips with a type-I lesion and one with a type-II were still asymptomatic but radiography showed sclerosis in the femoral head corresponding to the MRI lesions. The nine hips which showed no changes on MRI at one month had no abnormal findings on physical examination, radiography or MRI at final follow-up.

Superior labral tears of the shoulder involve the biceps tendon and labrum complex which may be detached, displaced inferiorly, and interposed between the glenoid and the humeral head. We have treated ten young athletes with painful shoulders due to this lesion by arthroscopic stapling. Arthroscopy at the time of staple removal, after three to six months, showed that all the lesions had been stabilised. Clinical review at over 24 months showed an excellent or good result in 80%. The two relative failures were due in one to residual subacromial bursitis, and the other to multidirectional shoulder instability. Arthroscopic stapling can restore the shoulder anatomy, and it is recommended for active adolescent athletes with this lesion.

In an attempt to repair articular cartilage, allograft articular chondrocytes embedded in collagen gel, were transplanted into full-thickness defects in rabbit articular cartilage. Twenty-four weeks after the transplantation, the defects were filled with hyaline cartilage, specifically synthesising Type II collagen. These chondrocytes were autoradiographically proven to have originated from the transplanted grafts. Assessed histologically the success rate was about 80%, a marked improvement over the results reported in previous studies on chondrocyte transplantation without collagen gel. By contrast, the defects without chondrocyte transplantation healed with fibrocartilage. Immunological enhancement induced by transplanted allogenic chondrocytes or collagen was not significant at eight weeks after treatment, so far as shown by both direct and indirect blastformation reactions. Thus, allogenic transplantation of isolated chondrocytes embedded in collagen gel appears to be one of the most promising methods for the restoration of articular cartilage.