Lumbar Total Disc Replacement: Seven to Eleven-Year Follow-Up

Tropiano, Patrick MD; Huang, Russel C. MD; Girardi, Federico P. MD; Cammisa, Frank P. Jr. MD; Marnay, Thierry MD

Journal of Bone & Joint Surgery - American Volume:
doi: 10.2106/JBJS.C.01345
Scientific Articles
Supplementary Content

Background: Symptomatic lumbar degenerative disc disease is a challenging entity to treat. The results of arthrodesis may be compromised in the short term by pseudarthrosis and in the long term by pain at the iliac-crest donor site and by junctional degeneration. Total disc replacement has the potential to provide long-lasting relief to these patients. The purpose of this study was to present the clinical and radiographic results assessed seven to eleven years following a Prodisc total lumbar disc replacement.

Methods: Sixty-four patients had single or multiple-level implantation of a total lumbar disc replacement between 1990 and 1993. The mean duration of follow-up was 8.7 years. Clinical results were evaluated by assessing preoperative and postoperative lumbar pain, radiculopathy, disability, and modified Stauffer-Coventry scores. Preoperative and postoperative radiographs were evaluated as well. Subgroup analysis was performed to determine if gender, an age of less than forty-five years, previous surgery, or multilevel surgery had an effect on outcome.

Results: At an average of 8.7 years postoperatively, there were significant improvements in the back-pain, radiculopathy, disability, and modified Stauffer-Coventry scores. Thirty-three of the fifty-five patients with sufficient follow-up had an excellent result, eight had a good result, and fourteen had a poor result. Neither gender nor multilevel surgery affected outcome. An age of less than forty-five years and prior lumbar surgery had small but significant negative effects on outcome. Radiographs did not demonstrate loosening, migration, or mechanical failure in any patient. Five patients had approach-related complications.

Conclusions: The Prodisc lumbar total disc replacement appears to be effective and safe for the treatment of symptomatic degenerative disc disease. Gender and multilevel surgery did not affect the outcomes, whereas prior lumbar surgery or an age of less than forty-five years was associated with slightly worse outcomes. Longer follow-up of this cohort of patients and randomized trials comparing disc replacement with arthrodesis are needed.

Level of Evidence: Therapeutic Level IV. See Instructions to Authors for a complete description of levels of evidence.

Author Information

1 Department of Orthopaedic Surgery, Hôpital CHU Nord, Chemin des Bourrelly, 13915 Marseille CEDEX 20, France

2 The Hospital for Special Surgery, 535 East 70th Street, New York, NY 10021. E-mail address for R.C. Huang:

3 Department of Orthopaedic Surgery, Clinique du Parc, 50 Rue Emile-Combes, Boite Postal 20, 34171 Castelnau-le-Lez, France

Article Outline

Intervertebral disc degeneration is a major cause of pain and disability in adults and is associated with large economic and social costs1. The initial treatment for lumbar degenerative disc disease is nonoperative, but that often fails. The most common surgical treatment is arthrodesis, which can relieve pain, eliminate segmental instability, maintain disc space height, preserve sagittal balance, and halt further degeneration at the operatively treated level. A recent randomized clinical trial demonstrated that, at the time of short-term follow-up, lumbar arthrodesis was superior to nonoperative management for the treatment of chronic low-back pain due to degenerative disc disease2.

However, lumbar arthrodesis is not a perfect operation. In a meta-analysis of forty-seven publications on lumbar arthrodesis, the rates of pseudarthrosis and chronic pain at the iliac crest donor site were estimated to be 14% and 9%, respectively3. Many postoperative protocols include cumbersome bracing. At the time of long-term follow-up after intersegmental fusion, stress concentrations have been found cephalad and caudad to the fused segments. Hypertrophic facet arthropathy, dynamic instability, spinal stenosis, disc degeneration, and osteophyte formation have all been reported to occur at levels adjacent to a fusion. These pathological processes may result in pain and require revision surgery in many patients4-6.

Many investigators have attempted to develop intervertebral disc prostheses7. Although multiple studies have demonstrated good short-term clinical results after lumbar total disc replacement8-10, to our knowledge none have included clinical follow-up of greater than 4.3 years11 after a contemporary lumbar total disc replacement.

The purpose of this study was to report the clinical and radiographic outcomes seven to eleven years following treatment with a first-generation Prodisc lumbar total disc replacement (Aesculap, Tuttlingen, Germany). We also sought to determine the patient factors that might affect the clinical outcome.

Back to Top | Article Outline

Materials and Methods

Patient Characteristics

Sixty-four patients underwent single or multiple-level implantation of a Prodisc lumbar total disc replacement between 1990 and 1993. Fifty-five (86%) of the sixty-four patients in the original cohort had sufficient clinical and radiographic follow-up for analysis and gave informed consent to participate in this study. Of the patients who were not included in the follow-up analysis, three had died from causes unrelated to the implantation surgery, two could not be located for the follow-up evaluation, two refused to return for the follow-up examination, and two did not complete all items of the questionnaire. The ethical compliance of this study was certified by the Committee on Medical Ethics of the Clinique du Parc.

Each patient had radiographic evidence of degenerative disc disease and had had severe lumbar pain despite nonoperative treatment for a minimum of six months prior to the index procedure. Exclusion criteria included facet arthrosis, central or lateral recess stenosis, osteoporosis, sagittal or coronal plane deformity, postoperative absence of posterior elements, and a sequestrated herniated nucleus pulposus that could not be decompressed from an anterior approach. There were thirty men and twenty-five women with an average age of forty-six years (range, twenty-five to sixty-five years) at the time of surgery. Twenty-seven patients (49%) had had no prior lumbar surgery, ten patients (18%) had had one prior operation, and eighteen patients (33%) had had two, three, or four prior operations. The prior operations included discectomy, laminotomy, percutaneous nucleotomy, chemonucleolysis, and thermocoagulation. A total of seventy-eight prostheses were implanted in the fifty-five patients. Thirty-five patients had a one-level implantation; seventeen, a two-level implantation; and three, a three-level implantation (Table I). Seven patients underwent additional surgical procedures during the same surgical session; these operations included an L5-S1 arthrodesis with an L4-L5 lumbar total disc replacement in six patients and an L5-S1 arthrodesis adjacent to L3-L4 and L4-L5 lumbar total disc replacements in one patient.

Preoperatively, forty-six (84%) of the fifty-five patients were employed: twelve were sedentary workers, and thirty-four were manual laborers. Of the forty-six employed patients, twenty-two were on medical leave from work because of back pain. Of the nine unemployed patients, two were receiving disability benefits and one was retired.

Back to Top | Article Outline
Implant Design

The first-generation Prodisc lumbar total disc replacement consists of two titanium-alloy end plates and a polyethylene core (Fig. 1). The implants are manufactured in four sizes, three heights, and two lordosis angles. Each end plate is coated with a titanium-plasma ingrowth surface and has two keels to provide primary stability. The convex polyethylene core is securely mated to the inferior titanium end plate and provides a congruent ball-and-socket articulation.

Back to Top | Article Outline
Surgical Protocol

Forty-five patients (82%) were treated through a retroperitoneal approach and ten, through a transperitoneal approach. The transperitoneal approach was preferred for obese patients or patients who had had previous anterior surgery. Anterior anulotomy and nucleotomy back to the posterior anular fibers was performed. The end-plate cartilage was removed, but the mechanical integrity of the osseous end plates was carefully preserved. The patients were mobilized out of bed on the first postoperative day. No patient wore a brace postoperatively. Running, jumping, heavy lifting, twisting, and bending at the waist were restricted until the third postoperative month.

Back to Top | Article Outline
Clinical Evaluation

The patients were evaluated preoperatively by the surgeon who treated them (T.M.) and postoperatively by another one of the authors (P.T.) and by a research assistant, neither of whom were involved in the patient selection, surgery, or postoperative care. The average follow-up period was 104 months (8.7 years), with a range of eighty-five to 128 months.

Four scoring systems were used to evaluate the patients preoperatively and postoperatively. A 20-point modified Stauffer-Coventry11 score, ranging from 0 (worst) to 20 (best), was used to assess pain, neurologic deficits, need for medication, disability, and psychiatric status (see Appendix). Absolute scores were converted into relative improvement scores by dividing the actual improvement by the maximum possible improvement. A result was rated as excellent when the relative improvement score was >70%, good when it was between 60% and 70%, and poor when it was <60%. Three separate subjective areas were scored on 3-point scales preoperatively and postoperatively by the patients (see Appendix): low-back pain was graded as severe (3 points), moderate (2 points), mild (1 point), or absent (0 points); lower-limb (radicular) pain was graded as severe (3 points), moderate (2 points), mild (1 point), or absent (0 points); and the ability to perform work-related activities and activities of daily living was scored as normal (0 points), slightly limited (1 point), substantially impaired (2 points), or severely limited or impossible (3 points). Finally, it was noted whether the patient was entirely satisfied, satisfied, or not satisfied with the result of the surgical procedure.

Several patient factors, including younger age (less than forty-five years), gender, previous lumbar surgery, and multilevel disc implantation, were evaluated to determine their effect on outcome. Student t testing (parametric), Wilcoxon matched-pairs signed-ranks testing (ordinal, paired), and Mann-Whitney testing (ordinal, unpaired) were performed with significance defined as p < 0.05. All data analysis was performed with Microsoft Excel 97 software (Microsoft, Redmond, Washington) and InStat 3 software (GraphPad Software, San Diego, California).

Back to Top | Article Outline
Radiographic Evaluation

Standing anteroposterior and lateral radiographs, lateral flexion-extension radiographs, magnetic resonance images, and computed tomography scans were assessed for periprosthetic radiolucent lines, implant migration, mechanical failure (dissociation or fracture), and wear of the polyethylene bearing. The height of the polyethylene core was measured on standing lateral radiographs made immediately after the surgery and at the final evaluation; the width of the subjacent vertebral end plate was used to correct for differences in magnification between radiographs.

Back to Top | Article Outline


Clinical Results

Thirty-three (60%) of the fifty-five patients had an excellent result (relative improvement score, >70%), eight (15%) had a good result (relative improvement score, 60% to 70%), and fourteen (25%) had a poor result (relative improvement score, <60%). There were significant postoperative improvements (p < 0.0001) in the modified Stauffer-Coventry, low-back-pain, lower-limb-pain, and disability scores (Table II). The mean modified Stauffer-Coventry score was 7.0 of 20 points preoperatively and 16.1 of 20 points postoperatively. Forty-three of the fifty-five patients rated the low-back pain as severe preoperatively, and fifty patients rated it as moderate, mild, or absent at the time of final follow-up (Fig. 2). Thirty-eight patients had severe lower-limb pain preoperatively compared with two at the time of the last-follow-up. Of the forty-eight patients who had had preoperative lower-limb pain, forty-three had no or mild pain at the time of the last follow-up (Fig. 3). Preoperatively, forty-six of the fifty-five patients were substantially or severely impaired with regard to their ability to perform work-related activities or activities of daily living, whereas, at the time of the last follow-up, forty-four patients had a normal or slightly impaired ability and eleven had a moderately or severely impaired ability (Fig. 4).

Of the twenty-four patients who had been employed full-time before the surgery, thirteen (54%) were employed, six were retired, four were receiving disability benefits, and one was on sick leave from employment at the time of final follow-up. Of the eight patients who had been unemployed or receiving disability benefits before the surgery, one was still receiving disability benefits, two were still unemployed, and five were retired at the time of final follow-up. Finally, thirty-five patients (64%) reported that they were entirely satisfied with the result of the surgery, fifteen reported that they were satisfied, and five were not satisfied.

The results of the analysis of the effects of age, gender, previous lumbar surgery, and multilevel disc implantation are shown in the Appendix. Briefly, patients who were forty-five years of age or older had better postoperative modified Stauffer-Coventry scores (p = 0.006) and higher relative improvement scores (p = 0.005). Although, preoperatively, women considered themselves to be more disabled than did men (p = 0.005), there were no significant differences in postoperative outcome based on gender. Patients who had had previous lumbar surgery had inferior postoperative modified Stauffer-Coventry scores (p = 0.042) and relative improvement scores (p = 0.045). Finally, with the numbers available, there were no significant differences in outcome between single and multiple-level implantations.

Of the seven patients who had had both disc replacement and an adjacent arthrodesis during the same operation, five had a relative improvement score of >70% and two had a relative improvement score of <60%. Three of the five patients who were not satisfied with the result of the index surgery subsequently underwent an arthrodesis with posterior instrumentation and retention of the prosthesis. These patients all had had multiple previous operations. One patient underwent the arthrodesis at five years after the disc replacement, and the other two patients, at three years.

Back to Top | Article Outline
Radiographic Results

No periprosthetic radiolucencies, migration, or mechanical failures were noted at the time of the last follow-up (Fig. 5). There was no substantial decrease in the mean height of the polyethylene bearing between the immediate postoperative and final radiographs. Seventeen patients (31%) had intraoperative penetration of the implant through the posterior part of the superior and/or inferior end plates and into the vertebral body. The end-plate penetration was ≤2 mm in fifteen patients and >2 mm in two. End-plate penetration had no significant effect on clinical outcome, with the numbers available. A spontaneous interbody fusion developed at the level of the implantation in one patient with end-plate penetration.

Back to Top | Article Outline

Five patients (9%) had surgical complications, consisting of one deep venous thrombosis, one iliac vein laceration that was repaired primarily, one transient retrograde ejaculation, and two incisional hernias. No complication was related to the prosthesis itself. Five patients had increased radicular pain postoperatively, without radiographically apparent neural compression. Their pain was attributed to nerve root traction resulting from intervertebral distraction in the setting of postoperative epidural fibrosis. These patients, who all had had previous discectomies, were treated with medication. The radicular pain had resolved by three months postoperatively in all patients.

Back to Top | Article Outline


Lumbar arthrodesis is the most commonly performed procedure for patients with painful degenerative disc disease for which nonoperative management has failed12. Despite excellent short-term clinical success rates approaching 80% in series ranging in size from fifty-one to sixty-nine patients13-15, lumbar arthrodesis is an imperfect procedure. One long-term study of patients treated with lumbar arthrodesis showed that 44% (twenty-seven of sixty-one) had back pain, 53% (thirty-three of sixty-two) were using medications for back pain, and 15% (nine of sixty-two) had required additional lumbar surgery at a mean of thirty-three years postoperatively16.

Total disc replacement has numerous potential advantages, including the avoidance of pseudarthrosis, postoperative bracing, and the development of junctional degeneration. Because candidates for lumbar total disc replacement are often young, a successful disc-replacement prosthesis must endure decades of high loading without mechanical failure. To our knowledge, we are the first to report the seven to eleven-year results of a contemporary lumbar total disc replacement. We consider this to be medium-term follow-up, and it is essential that follow-up of this cohort be continued in the decades to come.

The longest previous follow-up study of a contemporary lumbar total disc replacement, of which we are aware, was reported by Lemaire et al.11. In their study, 105 patients were evaluated with use of the modified Stauffer-Coventry score fifty-one months following treatment with the SB Charité implant (Waldemar Link, Hamburg, Germany). Using the same criteria that we used in the current study, they found that 79% of their patients had a relative improvement score of >70% (excellent), 6% had a score of 60% to 70% (good), and 15% had a score of <60% (poor). Complications occurred in 10% (eleven) of the 105 patients.

It is difficult to directly compare the results of lumbar total disc replacement with those of arthrodesis for the treatment of degenerative disc disease. Penta and Fraser17 reported on 103 patients who had been followed for ten to twelve years following anterior lumbar interbody arthrodesis. Seventy-eight percent (ninety-eight) of the original cohort of 125 patients had the surgery for treatment of discogenic back pain. Not including those who had nonunion, 10% of the 103 patients who were followed had complications. Seventy-eight percent of the 103 patients reported that they had “complete relief” or “a good deal of relief.” According to the Low-Back Outcome Score18, 68% (seventy) of the 103 patients had a “fair” result or better. Twenty-two percent (twenty-nine) of the 129 treated levels failed to fuse. Given the differences in patient-evaluation techniques between our study and the one by Penta and Fraser, it is impossible to determine whether fusion or lumbar total disc replacement provides superior results. In fact, since it may take more than a decade for symptomatic junctional degeneration to develop after a fusion19, longer follow-up periods may be necessary to distinguish between the two techniques even in randomized prospective studies.

We believe that the Prodisc device can be used successfully at two or three contiguous levels. Some authors have reported that they had not observed problems with multilevel implantation of SB Charité devices11; however, they did not compare single and multiple-level subgroups as we did in the current study. Cinotti et al.9 reported that the SB Charité prosthesis is not suitable for implantation at two contiguous levels. When it is implanted at two levels, adequate distraction may be difficult to achieve at the second level, resulting in an undersized or anteriorly located prosthesis. Because of the monoconvex configuration of the Prodisc polyethylene core, the intersegmental distraction required for insertion of the Prodisc replacement is probably less than that needed for insertion of the biconvex core of the SB Charité replacement. Twenty of our fifty-five patients had multilevel implantation, and their clinical results were similar to those of the single-level implantations.

Five patients with previous, failed back surgery experienced notable radicular pain after implantation of the Prodisc replacement, possibly because of epidural fibrosis resulting in nerve root traction after intervertebral distraction. The pain resolved spontaneously in all five patients after three months of observation. This complication occurred in five of the twenty-eight patients who had had previous lumbar surgery.

After an average of 8.7 years, polyethylene wear was not evident on plain radiographs. Although compressive loads in the lumbar spine are high, the previously reported9-11,20 arcs of flexion-extension have been small, averaging 4° to 10°. The accuracy of our technique for wear assessment is unknown, but it is likely that radiostereometric analysis is the most sensitive technique. Postmortem recovery of prostheses could also provide useful information regarding wear and bone ingrowth, but this information has not been reported, to our knowledge. Currently it is unknown whether osteolysis will occur after lumbar total disc replacement, but we are encouraged by the absence of radiographically apparent osteolysis in our patients at seven to eleven years postoperatively. Although metal and polyethylene lumbar total disc replacements were first implanted in the early 1980s, we are aware of only one reported case of presumed osteolysis after lumbar total disc replacement; this occurred thirteen years after placement of an SB Charité lumbar total disc replacement21.

The current study had all of the limitations inherent to nonrandomized case series. We cannot make any definitive statements regarding the outcomes of lumbar total disc replacement compared with those of nonoperative treatment or arthrodesis. The data reflect the results associated with a single implant type and may be different for other designs of lumbar total disc replacement22. Furthermore, the study would be more useful if validated outcomes instruments had been used. Nevertheless, our data indicate that the prosthesis is safe, with an acceptably low rate of complications at seven to eleven years postoperatively, and that its efficacy appears promising and should be tested in a randomized prospective trial.

Back to Top | Article Outline


Tables showing the modified Stauffer-Conventry scores; scores for low-back pain, lower-limb pain, and disability; and subgroup analysis by age, gender, prior lumbar surgery, and multilevel disc implantation are available with the electronic versions of this article, on our web site at (go to the article citation and click on “Supplementary Material”) and on our quarterly CD-ROM (call our subscription department, at 781-449-9780, to order the CD-ROM). ▪

A video supplement to this article is available from the Video Journal of Orthopaedics. A video clip is available at the JBJS web site, . The Video Journal of Orthopaedics can be contacted at (805) 962-3410, web site: .

In support of their research or preparation of this manuscript, one or more of the authors received grants or outside funding from Spine Solutions. In addition, one or more of the authors received payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity (Spine Solutions). No commercial entity paid or directed, or agreed to pay or direct, any benefits to any research fund, foundation, educational institution, or other charitable or nonprofit organization with which the authors are affiliated or associated.

Investigation performed at the Department of Orthopaedic Surgery, Clinique du Parc, Castelnau-le-Lez, France

1. , Simeone FA, Bernini PM. Lumbar disc disease. In: Rothman RH, Simeone FA, editors. The spine. 2nd ed. Philadelphia: Saunders; 1982. p 508-645.
2. , Hagg O, Wessberg P, Nordwall A; Swedish Lumbar Spine Study Group. Lumbar fusion versus nonsurgical treatment for chronic low back pain: a multicenter randomized controlled trial from the Swedish Lumbar Spine Study Group. Spine. 2001;26: 2521-34.
3. , Ersek M, Herron L, Haselkorn J, Kent D, Ciol MA, Deyo R. Patient outcomes after lumbar spinal fusions. JAMA. 1992;268: 907-11.
4. . Accelerated degeneration of the segment adjacent to a lumbar fusion. Spine. 1988;13: 375-7.
5. , Hanley EN Jr, Howe J, Kuhlmann D, Matteri RE. A comparison of radiographic findings in fusion and nonfusion patients ten or more years following lumbar disc surgery. Spine. 1979;4: 435-40.
6. , Jacquot F, Hall H. Long-term follow-up of functional outcomes and radiographic changes at adjacent levels following lumbar spine fusion for degenerative disc disease. Eur Spine J. 2001;10: 309-13.
7. , Gunzburg R, Mayer M. Spine arthroplasty: a historical review. Eur Spine J. 2002;11 Suppl 2: S65-84.
8. , Shelokov AP, Buttner-Janz K, LeMaire JP, Zeegers WS. A multicenter retrospective study of the clinical results of the LINK SB Charite intervertebral prosthesis. The initial European experience. Spine. 1994;19: 1842-9.
9. , David T, Postacchini F. Results of disc prosthesis after a minimum follow-up period of 2 years. Spine. 1996;21: 995-1000.
10. , Bohnen LM, Laaper M, Verhaegen MJ. Artificial disc replacement with the modular type SB Charite III: 2-year results in 50 prospectively studied patients. Eur Spine J. 1999;8: 210-7.
11. , Skalli W, Lavaste F, Templier A, Mendes F, Diop A, Sauty V, Laloux E. Intervertebral disc prosthesis. Results and prospects for the year 2000. Clin Orthop. 1997;337: 64-76.
12. , Montgomery DM. Diagnosis and treatment of discogenic low back pain. Orthop Rev. 1993;22: 311-8.
13. , Williams H. Anterior and combined anteroposterior fusion for lumbar disc pain. A preliminary study. Spine. 1991;16: 143-5.
14. , O'Brien JP. Simultaneous combined anterior and posterior fusion. An independent analysis of a treatment for the disabled low-back pain patient. Spine. 1990;15: 322-8.
15. , Pinto MR, Butler LM. Degenerative disc disease treated with combined anterior and posterior arthrodesis and posterior instrumentation. Spine. 2002;27: 1680-6.
16. , Spratt KF, Tozzi JE, Weinstein JN, Reinarz SJ, el-Khoury GY, Colby H. Long-term follow-up of lower lumbar fusion patients. Spine. 1987; 12: 97-104.
17. , Fraser RD. Anterior lumbar interbody fusion. A minimum 10-year follow-up. Spine. 1997;22: 2429-34.
18. , Fraser RD. Assessment of outcome in patients with low-back pain. Spine. 1992;17: 36-41.
19. , Smith JA, Schleusener RL. Lumbar motion segment pathology adjacent to thoracolumbar, lumbar, and lumbosacral fusions. Spine. 1996; 21: 970-81.
20. , Girardi FP, Cammisa FP Jr, Tropiano P, Marnay T. Long-term flexion-extension range of motion of the prodisc total disc replacement. J Spinal Disord Tech. 2003;16: 435-40.
21. , Oner FC, Verbout AJ. Complications of artificial disc replacement: a report of 27 patients with the SB Charite disc. J Spinal Disord Tech. 2003;16: 369-83.
22. , Girardi FP, Cammisa FP Jr, Wright TM. The implications of constraint in lumbar total disc replacement. J Spinal Disord Tech. 2003; 16: 412-7.
Copyright 2005 by The Journal of Bone and Joint Surgery, Incorporated