This study was carried out in order to document the indications, methods, complications and results of revision lumbar disc replacement surgery.

Out of a total series of 29 disc replacement operations carried out by the senior author 8 patients have required revision (10 operations). All of these patients have been followed prospectively since their initial operation and data was extracted from the case notes, theatre logs and radiographs.

The primary indications for revision were recurrence of, or failure to relieve back pain. Of the 8 patients requiring revision for their symptoms, 3 had prostheses that had clearly failed on CT. Four patients had the pros-thesis removed and were revised to a circumferential fusion using iliac crest autologous bone graft anteriorly and pedicle screws posteriorly. Three patients had a posterolateral fusion only using pedicle screw fixation with retention of the prosthesis. Two patients had a posterior procedure carried out which failed to resolve the situation satisfactorily, and subsequently had the prosthesis removed and interbody cage fusion. Out of the 6 anterior cases there were 2 injuries to the great veins requiring vascular repair. There were few complications with posterior surgery. Although numbers are small, the patients who had anterior surgery seemed to do better than those retaining their prosthesis with posterior surgery alone.

We conclude that a posterolateral fusion is a safe but symptomatically unsatisfactory revision strategy for failed disc replacement, and that anterior revision with conversion to inter-body fusion is a better option but has a higher rate of surgical morbidity.

This prospective study was carried out to document the efficacy of lumbar disc replacement using a viscoelastic prosthesis in the medium term.

Eleven patients were included in the pilot study of this prosthesis, with an average age at surgery of 41, and these have now reached 5 years follow-up. They have been followed up prospectively with clinical and radiological evaluation, and outcome assessment using the Oswestry Disability Index (ODI), Visual Analogue pain scales (VAS), and the Low Back Outcome Score (LBOS).

There have been 3 revision operations in patients who failed to get lasting symptomatic relief. Two patients have had their prosthesis removed with revision to an inter-body fusion supplemented with pedicle screw fixation, and one has been revised to a posterolateral fusion while retaining an apparently well functioning prosthesis.

In the 8 ‘survivors’ the improvement in the ODI, VAS and LBOS seen at the 2 year follow-up has been maintained in the longer term. Fine cut helical CT scan has shown that the prosthesis has undergone mechanical failure in 3 patients, without significant clinical problems. One patient (with a virtually perfect ODI and LBOS) appears to have undergone ‘autofusion’ with significant heterotopic ossification anterior to the disc prosthesis.

In conclusion, despite some successes, the rate of early failure of this prosthesis, both radiologically and clinically, is unacceptable, and its use has therefore been abandoned.