Failed fusions are difficult to treat. This is an analysis of a subgroup of 19 patients who received lumbar disc prostheses in a segment adjacent to previous posterolateral instrumented fusions.

Out of 142 patients who received 181 lumbar disc replacements of two different models, 19 patients (11 men and eight women) had previously undergone posterolateral fusions of one or two lumbar segments. Seven of these patients had junctional replacement using a Charité disc prosthesis and 12 using a Centurion disc prosthesis. The mean age was 46 years (33 to 63).

All patients had a significant flat back deformity of their lumbar spine before the index operation. All presented with sacro-iliac joint pain and 17 had spinal stenosis symptoms. The mean postoperative hospital stay was 3.4 days (2 to 8). All patients returned to their previous occupations at a mean of 29 days (21 to 42). Follow up ranges from 6 to 32 months. At the latest follow-up, 10 patients were very satisfied and nine were satisfied. The Oswestry score in the last 12 (Centurion) patients decreased from 49 ± 12 preoperatively to 16 ±22 at the latest follow-up.

There are no published data on this particular subgroup of patients. They are difficult to treat and would otherwise require combined front and back procedures with extension of the fusion in order to correct the sagittal imbalance. Disc replacement seems to be a good salvage procedure for junctional failure after previous fusions.

Lumbar disc replacement with a modern articulating disc prosthesis was first done in 1984, using the Charité SB I prosthesis. Since then other discs, easier to implant, have been developed (Prodisc, Maverick, Centurion).

We present short-term (6 to 18 month) clinical results of implantation of 121 Centurion disc prostheses in 93 consecutive patients. This modular three-component titanium/polyethylene disc is inserted as one unit and suitable for minimal interventional approaches.

The mean age of the patients was 43 years (23 to 63). The primary diagnosis was degenerative disc disease, with 16 patients having symptoms of spinal stenosis. Forty-four patients underwent single-level disc replacements and 28 double-level disc replacements. Nine patients had a fusion of another level at the time of the index procedure and 12 patients had disc replacement at a junctional level after previous fusion procedures. At 6-month follow-up of 50 patients, clinical outcome was excellent in 27 patients, good in 16, fair in four and poor in three. The Oswestry score improved from 51.2 ±16.0 preoperatively to 13.2 ± 17.8 (p < 0.01) at 6 months. One of the first 29 patients was lost to follow-up. In the other 28 there were 16 excellent, seven good, three fair and two poor outcomes. The Oswestry score improved from 49.7 ± 17.5 preoperatively to 13.0 ± 14.9 (p < 0.01) at 1 year.

Good short-term clinical results were achieved in a heterogeneous patient group, largely with advanced disc degeneration and severe disc space narrowing. Using a new insertion mechanism, posterior placement within the disc space and disc space distraction could be reliably achieved even in preoperatively collapsed and immobile disc spaces.

Sixty-one lumbar disc replacements in 50 consecutive patients were performed between June 2001 and October 2002. The mean age of the patients was 42.42 years (24 to 61). All presented with mechanical back and/or leg pain. The primary diagnosis was degenerative disc disease. Three patients underwent fusions of another level during the same operation. Seven patients presented with adjacent level disc disease after previous instrumented posterolateral fusion.

The mean postoperative hospital stay was 3.6 days (2 to 8). One patient with a double-level disc replacement and misplacement of the proximal prosthesis underwent revision surgery three days later. Six patients had subsidence on one of the endplates, all within six weeks postoperatively. All but one patient went back to their previous occupation at a mean time of 26 days (7 to 91).

Patients completed questionnaires at 3, 6, 12 and 24 months postoperatively. In the latest follow-up questionnaire, 47 patients said they were satisfied or very satisfied. Forty-five patients would undergo the same operation again or recommend it to friends, two patients were uncertain and two indicated they would not undergo the procedure again. One patient from another country was lost to follow-up, but a telephonic enquiry six months after surgery suggested that he was doing well.

Although our follow-up period is short, the clinical results compare favourably to those in the literature. The rate of immediate subsidence, which seems to be linked to the anchoring teeth, is of concern.

A subgroup of nine patients, five women and four men, who received lumbar disc prostheses in a segment adjacent to previous posterolateral instrumented fusions (one or two levels) is reviewed. Two patients underwent double-level disc replacement. The mean age was 46.1 years (33 to 62). All patients had marked flattening of the lumbar spine before the operation. One patient developed subsidence within two weeks of the procedure but went on to a satisfactory result. One patient with a double-level disc replacement needed another operation because of a missed far out lateral disc. On day five after the procedure, the top-level disc was removed, the level properly decompressed and the disc reinserted, with good results.

The mean postoperative hospital stay was 3.9 days (3 to 8). Within a mean time of 32 days (21 to 42), all patients returned to their previous occupations. Follow-up ranged from 3 to 19 months. At the latest follow-up, four of the patients were satisfied and five were very satisfied.

Disc replacement seems an ideal salvage procedure for junctional failure after previous fusions. The short-term clinical results are good.