Sixty patients underwent arthroscopy of the knee as day cases using a local anaesthetic technique. This provided satisfactory operating conditions and high patient acceptability.

Magnetic resonance imaging (MRI) was used to compare the appearance of the spine in 20 adolescents with proven symptomatic intervertebral disc herniations with that in 20 asymptomatic patients who acted as controls. Abnormality in the signal from the nucleus pulposus of one or more discs was present in all patients, while only four of the 20 controls had any abnormal signals. In all the patients the symptomatic disc produced an abnormal signal and in most a herniated fragment of the nucleus pulposus was identified. Fifteen of the 20 patients had multiple-disc abnormality: six had three abnormal discs and nine had two. This suggests there was an underlying diathesis in patients who later developed disc herniation.

Magnetic resonance imaging (MRI) of the spine produces images which reflect the chemical composition of the intervertebral disc. We have conducted a prospective study of the serial changes in the MRI appearance of the intervertebral disc after chemonucleolysis with the enzyme chymopapain. Fourteen patients were studied after single-level chemonucleolysis and the results compared with a control group of 17 discs in six patients who had diagnostic discography without enzyme insertion. A consistent pattern of gradual loss of signal from the nucleus pulposus culminating in complete loss of nuclear signal was seen in all cases after chemonucleolysis. Chymopapain therefore produced MRI changes analogous with premature gross disc degeneration. The rate at which this occurred varied; complete loss of signal took at least six weeks. Transitory minor end-plate changes were present in five patients, probably representing a mild chemical discitis. No similar changes were seen in the discography group.

The lumbar spines of 22 patients were examined for disc degeneration by magnetic resonance imaging (MRI) and by discography. The results from 50 intervertebral discs visualised by both techniques were independently assessed and graded on a five-point scale from normality to gross degeneration and then compared. In 44 discs the results agreed. Of the six discs which gave differing results, four discrepancies were due to observer error and two to incorrect placement of the discographic needle. MRI was shown to be more accurate than discography in the diagnosis of disc degeneration. It has several major advantages, which should make it the investigation of choice.