Eur Spine J (1999) 8 : 210–217
`© Springer-Verlag 1999
`
`O R I G I N A L A RT I C L E
`
`W. S. Zeegers
`L. M. L. J. Bohnen
`M. Laaper
`M. J. A. Verhaegen
`
`Artificial disc replacement
`with the modular type SB Charité III:
`2-year results
`in 50 prospectively studied patients
`
`Received: 15 November 1996
`Revised: 17 December 1998
`Accepted: 11 January 1999
`
`W. S. Zeegers · L. M. L. J. Bohnen ·
`M. Laaper · M. J. A. Verhaegen
`Department of Orthopaedics,
`Maasland Hospital, Sittard,
`The Netherlands
`
`L. M. L. J. Bohnen (쾷)
`Oude Rijksweg Noord 30,
`NL-6114 JE Susteren, The Netherlands
`
`Abstract The Modular Type SB
`Charité disc prosthesis has been de-
`veloped as a device for artificial disc
`replacement (ADR) in patients with
`symptomatic discopathies. Here, we
`report on our first series of 50 (out
`of 350) patients, who had a satisfac-
`tory clinical result in 70% of cases
`(2 years’ follow-up). Subgroup ana-
`lysis revealed that patients with an
`isolated discopathy without previous
`spinal operations or other pathology
`at the same or other spinal level ben-
`efitted more from the surgery. How-
`ever, this technique was associated
`with some problems: a 13% rate of
`permanent side-effects and/or com-
`plications was observed caused by
`the anterior approach. Four percent
`were related to poor implantation
`
`technique. There were no problems
`related to the material of the prosthe-
`sis. Twelve patients needed re-opera-
`tion, but this was beneficial in only
`three of them. In one patient we had
`to convert to an interbody fusion. We
`conclude that in patients with severe
`isolated symptomatic discopathies
`that are resistant to conservative
`treatment, a mobile disc prosthesis is
`worth considering as a real alterna-
`tive to a spondylodesis. However, ac-
`curate patient selection is imperative.
`With these criteria we were encour-
`aged by our results to continue the
`implantation of this artificial disc.
`
`Key words Lumbar spine ·
`Discopathy · Disc prosthesis ·
`Outcome · Complications
`
`Introduction
`
`Spinal fusion has been used as a treatment of chronic dis-
`abling low back pain not responding to conservative treat-
`ment. Results of spinal fusion, however, are often unpre-
`dictable [13,14]. Complications of spinal fusion, such as
`bone graft donor site pain, prolonged period of postop-
`erative recuperation, pseudarthrosis, spinal stenosis, and
`spondylolysis acquisita have been reviewed by Lee et al.
`[13, 14]. In addition, Lee lists several complications that
`can affect adjacent unfused segments, such as disc hernia-
`tion, accelerated degeneration, spinal stenosis, spondylo-
`lysis, facet joint arthritis, and instability. In this context,
`preservation of segmental mobility is perceived favourably
`[11]. Artificial disc replacement (ADR) has been devel-
`
`oped as a technique to preserve segmental motion after sur-
`gical treatment of a symptomatic discopathy [11]. ADR
`has the advantage of reduced force on the adjacent seg-
`ments, and facilitates postoperative mobilisation. Büttner-
`Janz et al. [2, 3, 5] evaluated several designs of ADR: The
`prosthesis with a metal ball of Fernström showed penetra-
`tion of the metal ball into the vertebral body in most cases,
`which is due to the high local mechanical load at the con-
`tact areas. However, Fernström [9] found better clinical
`results in patients with this spherical endoprosthesis after
`evacuation of the disc than in those without this implant,
`at a follow-up of 0.5 –2.5 years. MacKenzie [16] also
`used this technique in patients with disabling sympto-
`matic discopathy and discal prolapse. He too reported en-
`couraging clinical results at that time (1972), which also en-
`couraged the consideration of artificial disc replacement.
`GLOBUS MEDICAL, INC.
`EXHIBIT 1020
`IPR2015-to be assigned
`(Globus v. Flexuspine)
`1 of 8
`
`

`
`211
`
`Fig. 1 Anteroposterior view of
`Modular Type SB Charité III
`disc prosthesis at L4-L5, con-
`sisting of two cover plates
`(cobalt-chromium alloy),
`which are fixed without ce-
`ment, as well as a moveable
`polyethylene sliding core be-
`tween the plates
`
`Fig. 2 Lateral view of Modu-
`lar Type SB Charité III disc
`prosthesis at L4-L5. The
`choice of endplates is deter-
`mined by the disc that is being
`replaced. Oblique models are
`used at L5-S1 and the plane-
`parallel are used at levels
`above L5-S1
`
`1
`
`2
`
`Correction of decreased disc space height has been thought
`to be important, because diminished disc space height has
`been linked to several problems, including narrowing of
`the intervertebral foramen, change of kinetic centre, over-
`load of facet joints, disturbance of the structures with hy-
`permobility, and overload at insertions of ligaments [22,
`23]. The design of the Link SB Charité disc prosthesis
`was based on the concepts of ADR developed by Büttner-
`Janz and Schellnack [2–5] and it has been tested clini-
`cally. Its development and composition have been exten-
`sively described by Büttner-Janz and Schellnack [2] and
`Griffith et al. [11]. The prosthesis underwent several ma-
`jor modifications in design (structure and manufacturing)
`from type I to type II and finally to type III (Figs. 1, 2), be-
`cause of prosthetic failure (model and material problems)
`and subsidence in clinical use [2, 5]. There is paucity of
`published literature on clinical results with ADR. There-
`fore, we want to present our first experiences and prelim-
`inary results of our learning curve with a 2-year postop-
`erative follow-up of our prospective study.
`
`verted into an interbody fusion because of malposition. Twenty-
`nine patients had a prosthesis inserted at one level, 18 at two lev-
`els, and 3 patients had three prostheses inserted (Table 1). The mean
`age at the operation was 43 years (24–59 years). Thirty women and
`20 men were operated. Mean duration of history of low back com-
`plaints was 10 years (range 1–35 years).
`Patients with predominant symptoms or deficits in the legs that
`could be related to involvement of the nerve roots were excluded.
`The radiographic features for each operated level were diverse:
`40 discopathies (DP), 29 post-discectomy discopathies, and 6 dis-
`copathies with possible signs of disc protrusion at CT imaging but
`without any clinical radicular symptoms. All patients had failed
`conservative management. Twenty of 50 patients were involved in
`heavy physical labour. Fifty-four percent (27/50) of all patients had
`undergone previous surgery (Table 2). Special attention was paid
`to possible confounding factors, such as discitis, spondylolysis/
`spondylolisthesis, transitional vertebra, or conservative treatment
`of former disc herniation. We used standard anteroposterior (50/50),
`lateral (50/50) and flexion/extension radiographic views (40/50) of-
`ten supported by CT (37/50), discography (36/50) and MRI (2/50).
`Special attention was paid to preoperative characteristics of degen-
`eration at the same or other lumbar levels. According to our own
`criteria, we divided these into ‘no’, ‘minor’ or ‘definite’ character-
`istics of other lumbar degeneration (Table 3). We used criteria de-
`rived from those of Stauffer and Coventry [20] to assess the clini-
`cal response to surgery. The patients were classified as having a
`‘good’, ‘fair’ or ‘poor’ clinical result, based on relief of pain, re-
`
`Materials and methods
`
`Fifty patients with medically refractory lumbar discopathies un-
`derwent placement of a Modular Type SB Charité III disc prosthe-
`sis at the Maasland Hospital, Sittard (the Netherlands), between
`June 1989 and June 1991. All patients were operated by the main
`author. All patients were studied prospectively, and were evaluated
`after a postoperative period of 2 years. Four patients were lost to
`follow-up. The patients were interviewed and examined before op-
`eration and at regular periods afterwards. Other data were ab-
`stracted from the patients’ case files and radiographs. Seventy-five
`disc prostheses were inserted in 50 patients. Of these, seven pros-
`theses were placed at another level in patients already having re-
`ceived one or two disc prostheses. One prosthesis insertion was con-
`
`Table 1 Operated levels of
`50 patientsa
`
`a One level with a prosthesis
`(in malposition) is not men-
`tioned in this Table, because it
`had already been converted
`into an interbody fusion at the
`time of this follow-up study
`
`Level
`
`L5-S1
`L4-L5
`L4-S1
`L3-S1
`L3-L4
`L3-L5
`
`L5-S1 + L3-L4
`
`n
`
`8
`16
`12
`3
`5
`5
`
`1
`
`2 of 8
`
`

`
`212
`
`Table 2 Previous lumbar surgery including the levels of a disc
`prosthesis (n = 27/50 patients, 54%)
`
`Table 4 Criteria for classification of technical results
`
`Surgical procedure
`
`No. of patients
`
`Percutaneous microdiscectomy
`
`Laminectomy and/or discectomy:
`+ percutaneous microdiscectomy
`+ chemonucleolysis
`+ chemonucleolysis + percutaneous
`microdiscectomy
`
`Spinal fusion:
`+ percutaneous microdiscectomy
`+ denervation of facet joint
`
`11
`
`8
`1
`2
`
`1
`
`2
`1
`1
`
`Table 3 Criteria used for classification of preoperative lumbar de-
`generation apart from discopathies to be operated
`
`Group A: No other characteristics of (lumbar) degeneration
`
`Group B: Minor characteristics of other degeneration:
`(with a maximum of 3 of these)
`B1 – Spondylosis: osteophytes, slight sclerosis or cysts
`B2 – Segmental hyperflexion, hyperextension, or pinching
`B3 – Shift > 3 mm during flexion and extension
`B4 – Minor arthritis of facet joints
`B5 – Herniation on CT (without radicular symptoms)
`B6 – Minor narrowing of intervertebral foramen or disc
`space height of other level
`
`Group C: Definite signs of other lumbar degeneration:
`C1 – More than 3 group B characteristics
`C2 – Evident degeneration of another disc
`C3 – Evident degeneration of facet joints
`C4 – Vacuum phenomenon
`
`turn to employment, physical activities and consumption of anal-
`gesics. A good or fair result was defined as a positive clinical re-
`sult. The assessment of the technical results was based on position
`and angulation of the prosthesis. The position of the prosthesis was
`assessed by normal AP and lateral radiographs. Attention was paid
`to the position of the prosthesis in the transversal intervertebral disc
`space (central, asymmetric, or malposition) and to the eventual an-
`gulation of the prosthesis to the neighbouring borders of the verte-
`bral bodies (acceptable, borderline, or wrong). Technical results were
`classified into ‘good’, ‘fair’ or ‘poor’ according to our own criteria
`for position and angulation (Table 4). We used the method of Far-
`fan to measure the disc space height [8].
`
`Good technical result
`Central position: the centre of the prosthesis is located at
`a distance of less than 1/8 of the (anteroposterior and lateral)
`diameters from the centre of the disc space.
`Angulation of endplates of the prosthesis is less than 15°
`to the transversal vertebral plane.
`
`Fair technical results
`Asymmetric position: position is not central, but the
`prosthesis is inside the side-lines of the disc space.
`Angulation between plates of the prosthesis and the trans-
`versal vertebral plane is between 15° and 30°.
`
`Poor technical result
`The prosthesis is partially outside the side-lines of the disc space.
`Angulation is more than 30°
`
`Oversized prosthesis
`The maximum diameter of the prosthesis exceeds the minimal
`diameter of the endplates of the vertebra.
`
`Undersized prosthesis
`In the sagittal or frontal plane, the diameter of the pros-
`thesis is less than 2/3 of the vertebral body diameter.
`
`the intervertebral space. The selected metal endplates are intro-
`duced by gently tapping on the introducer only. The polyethylene
`sliding core is placed between the plates after distraction of the bo-
`dies has been performed. The height of the core is dependent on
`the degree of distraction. Teeth in the coverplates and release of dis-
`traction prevent sliding during removal of the introducer.
`
`Peri-operative and postoperative period
`
`The patients receive peri-operative antibiotics and no urinary
`catheter. Oral fluid in the first days are dependent on abdominal re-
`covery. Early mobilisation is encouraged without restriction di-
`rectly after the operation. A supporting belt is optional.
`
`Results
`
`Clinical results
`
`Seventy percent of the patients had a positive clinical re-
`sult (32/46). Four patients were lost to follow-up, despite
`repeated summones.
`
`Technique of operation
`
`Griffith et al. [11] have described the technical surgical details of
`this disc prothesis. The three-piece disc prosthesis consists of two
`cover plates of cobalt-chromium alloy, implanted without cement,
`as well as a moveable polyethylene sliding core between the plates
`(Figs. 1, 2). The size of the artificial disc is determined by the di-
`ameter of the endplates of the intervertebral body. Typically, the
`oblique models are used at the L5-S1 level and the plane-parallel
`models are used at levels above L5-S1. Each geometric configura-
`tion of artificial endplate is manufactured in three sizes. The ante-
`rior retroperitoneal approach was used as described for interbody
`fusion [17]. The size of the prosthesis is measured after cleaning of
`
`Factors in clinical outcome
`
`There were no significant differences in clinical outcome
`for the following factors: Central versus asymmetric posi-
`tion of the prosthesis; one versus two prostheses in the same
`patient; pure discopathy versus post-discectomy discopa-
`thy; duration of history, less versus more than 5 years; and
`good versus fair technical result. Previous surgery no
`longer showed a difference in clinical outcome by 2 years
`after the operation, in contrast to 1 year postoperatively.
`
`3 of 8
`
`

`
`213
`
`The following factors showed a non-significant ten-
`dency toward better outcome: Patients without previous
`surgery and without other lumbar degenerative character-
`istics or other possible (roentgenographic) pathology (our
`‘ideal candidate’) demonstrated better clinical results
`(81% vs the others 66%). This difference was statistically
`significant at 1 year after surgery. Patients without com-
`plications did better than those with (82% vs 62%), and fi-
`nally, patients without re-operations after a disc prosthesis
`did better than those with (79% vs 53%).
`The following factor was associated with a statistically
`significant better outcome: Age under 45 years showed
`better clinical results. All factor significance was tested by
`Chi-square at the P < 0.05 level.
`
`Reduction of pain
`
`Sixty-five percent (30/46) showed improvement of low
`back pain. Four patients did not have preoperative leg
`pain. Sixty-four percent (27/42) reported improvement of
`leg pain at 2 years after surgery.
`
`Return to work
`
`All working activities at home and during leisure and em-
`ployed labour time are included. Three patients did not
`work at all before surgery. In four patients no information is
`available about their work at the 2-year follow-up. Eighty-
`one percent returned to some work (35/43), and 43% re-
`turned to their original work.
`
`Analgesics
`
`to our own criteria (Table 4), 74% (28/38) showed a good
`technical result, and 24% (9/38) showed a fair position-
`ing, although clinically well accepted. Two out of 50 pa-
`tients showed a malpositioning by the surgeon. One of
`these 2 patients (who had a double prosthesis) needed a
`spondylodesis after removal of the malpositioned prosthe-
`sis. This finally resulted in a positive clinical result. The
`other patient with a malposition and some slip was lost to
`follow-up despite repeated summones.
`Eighteen percent of all prostheses showed an asym-
`metric position, that clinically could be accepted. The ma-
`jority of implants were located fairly centrally. Two per-
`cent of all prostheses were positioned rather steeply, but
`all were less than 30° to the transversal plane. No end-
`plates were oversized, but 23% were undersized.
`According to the disc space height quotient of Farfan
`[8], 14% of all levels with a prosthesis showed a decrease
`in height (> 25%), 2 years after surgery. There was no
`significant migration (> 2 mm) of the prostheses in our
`patients who attended the 2-year evaluation. Unfortunately,
`the above-mentioned patient with a malposition and some
`slip of the prosthesis did not attend follow-up The range
`of motion of the prosthesis (ROM) between flexion and
`extension (on lateral radiographs) averaged 9° (range 2°–
`17°) 2 years postoperatively, which equalled the preoper-
`ative ROM, namely 9° (range 2°–20°).
`
`Re-operations
`
`Twelve patients (24%) out of the 50 patients needed re-op-
`erations. Seven re-operations were related to complications,
`
`Fifteen out of 34 patients were able to decrease their anal-
`gesic intake.
`
`Table 5 Types of re-operation (n = 24) in 12 patients
`
`Patient satisfaction
`
`Here, 83% (38/46) did not regret their surgery at all. The
`most frequent argument for regret was a disappointing clin-
`ical result (6/8).
`
`Peri-operative and postoperative period
`
`Generally, all patients underwent (supine) bedrest averag-
`ing 4 days (0–45 days) followed by a mean time to mobil-
`isation of another 7 days (1–45 days). The total hospital
`stay averaged 10 days (3–90 days), according to the pro-
`tocol of the period between June 1989 and June 1991. A
`supporting belt was provided to half of the patients for an
`average period of 12 weeks (0.5 week–1 year).
`
`Technical results
`
`Thirty-eight out of 50 patients had complete roentgeno-
`graphic documentation 2 years postoperatively. According
`
`Procedure
`
`A. Re-operations at the segment with a prosthesis
`Dorsal release:
`+ ligamentoplasty (Graf)
`+ arthrodesis of facet joints
`+ lateral decompression + block of facet joints
`+ facetectomy
`Dorsal fusion after partial removal of facet joint,
`with prosthesis in situ
`
`B. Re-operations at other levels
`Percutaneous nucleotomy
`Disc prosthesis
`Discectomy
`Dorsal release
`
`C. Re-operations related to complications (3 patients)
`Release of haematoma
`Unsuccessful effort to reposition prosthesis
`Conversion of prosthesis (malposition) into fusion
`Vascular surgery (aorta)
`
`a Same patient
`
`n
`
`6
`4
`(1)
`(1)
`(1)
`(1)
`
`2
`
`11
`2
`7
`1
`1
`
`7
`2
`1a
`1a
`3a
`
`4 of 8
`
`

`
`214
`
`11 concerned other segments, while 6 reoperations were
`performed at the level of the disc prosthesis (Table 5).
`All operations at the level of the prosthesis, apart from
`salvage of complications, were performed to improve pain
`control (Table 5 A). Pain control was also the indication for
`additional operative treatment at other levels (Table 5 B).
`Re-operation was beneficial in only three patients.
`Re-operation for complications (seven, in only three
`patients) included release of haematoma (two times), with
`the others being in our first and only patient to undergo re-
`moval of the prosthesis, which resulted in severe difficul-
`ties (Table 5 C).
`
`Complications and side-effects
`
`We also paid attention to subjective complaints (pain, dis-
`turbances in sensation, cramps, etc. (Table 6). During this
`postoperative period, side-effects or complications at or
`after the first implantation operation were reported 52 times
`by 30 patients. Only 16% of these were permanent. Only
`2/52 were clearly related to the implantation technique,
`
`Table 6 Complications and side-effects 2 years after surgery, in-
`cluding re-operations
`
`Temporary Permanent
`(n = 44)
`(n = 8)
`
`Neurological (n = 10)
`Dysaesthesia of legs
`Paresis/muscle weakness
`Cramps in legs
`
`Wound, haematoma (n = 17)
`Painful/numb scar
`Haematoma
`
`Abdominal problems (n = 3)
`Retroperitoneal haematoma
`Visceral dysfunction
`Abdominal pain
`
`New or progression of old pain (n = 5)
`Low back or leg pain
`
`Vegetative dysfunctions (n = 8)
`Sympathectomy effect
`Disturbance of miction
`
`Vascular problems (n = 1)
`Aortal lesion (at removal
`of prosthesis)
`
`General complications (n = 5)
`Infection of urinary tract
`Impotence, retrograde ejaculation
`Deep venous thrombosis
`
`Prosthesis (n = 2)
`Malposition of prosthesis
`
`4
`1
`2
`
`5
`12
`
`1
`1
`1
`
`5
`
`3
`1
`
`1
`
`4
`1
`1
`
`1
`
`3
`–
`–
`
`–
`–
`
`–
`–
`–
`
`–
`
`4
`–
`
`–
`
`–
`–
`–
`
`1
`
`while there were no problems related to the material of the
`prosthesis itself.
`
`Discussion
`
`Clinical results
`
`Enker et al. [7] tried to compare their clinical results using
`the Acroflex ADR with the series described by Büttner-
`Janz et al. [5], who used the SB Charité ADR and found
`83% satisfactory clinical results at an average period of
`15 months (max 3 years) after surgery. Enker et al. [7]
`pointed out that differences in age and surgical indications
`and a failure to differentiate the results of patients under-
`going primary from those undergoing revision surgery did
`not allow a valid comparison between the two studies. A
`comparison of our results with the SB Charité type III
`ADR and the experiences of the multicentre study with
`the SB Charité types I, II and III ADR [11], shows similar
`improvements in back pain. Wittig et al. [23] described
`their experiences with the SB Charité type III ADR and
`found good or satisfactory clinical results in 77% of their
`patients at a follow-up of 3–18 months. In contrast to
`Cinotti et al. [6], we did not find a significant difference in
`clinical outcome depending on whether patients had re-
`ceived an ADR at one or at more segments. We showed
`that patients without previous surgery had better clinical
`results 1 year postoperatively, but that this difference faded
`away 1 year later. Similarly, no difference according to pre-
`vious surgery was reported by Büttner-Janz et al. [5]. Our
`patients who had no previous surgery and no other degen-
`eration or spinal disorder (apart from the symptomatic
`discopathy) – our ‘ideal candidates’ – showed a positive
`clinical result in 80% at 2-year follow-up. Total return to
`original working activities was rather limited (43% 2 years
`after the operation). Similar disappointing results were re-
`ported by Griffith et al. [11]. Fourteen patients had a poor
`clinical outcome 2 years after the operation. Other degen-
`erative pathology at the same or other lumbar level was
`found in six of those 14 patients. Another patient became
`pain free after psychiatric treatment. So far, there is no
`satisfactory explanation for the poor outcome of the other
`patients. Cinotti et al. [6] reported better clinical outcome
`after ADR for patients with no previous back surgery. We
`can confirm this finding only at 1 year after surgery, and
`we should also note that in our series of 50 operated pa-
`tients, 54% had undergone previous spinal surgery. A crit-
`ical review of our good and poor clinical results makes
`clear how difficult it is to find the real origin of low back
`pain. We identified discopathy as a possible cause of their
`pain
`syndrome when a preoperative provocative
`discogram caused identical pain. Despite this selection,
`there remain some patients whose poor clinical results
`cannot be explained by complications or other pathology.
`This illustrates that reasons for low back pain can be ob-
`
`5 of 8
`
`

`
`215
`
`scure. It needs to be emphasized that chronic pain behav-
`iour is also related to psychosocial factors. Psychosocial
`screening in this field will be absolutely necessary.
`What, then, makes a patient a candidate for a disc
`prosthesis? Logically, this would be a patient with proven
`discogenic low back pain, without any other (organic or
`psychological) cause for their complaints. Therefore, we
`looked at our patients without any previous spinal surgery
`or other organic disease in the lower back, and we found
`good 2-year results in 80% of this selected subgroup of
`patients (called our ‘ideal candidates’).
`
`Indications and contra-indications
`
`Several indications and contra-indications for ADR have
`been previously reported, but are not unanimously ac-
`cepted. An ADR is primarily designed for severe disco-
`pathies and directly related morbidity: disc degeneration
`[7, 11], postnucleotomy situations [5, 7, 11], isolated disc
`resorption [7], and lateral recess stenosis because of di-
`minished disc height [7]. For disc degeneration at one or
`more levels, some authors [7, 22] point to the importance
`of ADR in case of juxtafusion degeneration. Wittig et al.
`[22] report that adjacent to a spondylodesis, increased
`damage can be expected because of loss of function of the
`fused spine. A mobile prosthesis at the adjacent level, in-
`stead of an extra fused level of an adjacent discopathy, can
`be a solution for additional detrimental stresses on other
`levels. In contrast, others [10, 12] do not confirm these
`ideas in their studies on spinal fusion. For painful back
`syndromes after discectomy we agree with Enker et al.
`[7], that these theoretically can be good indications, pro-
`vided that there is no recurrent disc herniation, no insta-
`bility and no damaged posterior elements. Lateral recess
`stenosis due to disc degeneration and loss of disc space
`height might respond to disc space height restoration by
`ADR but should be considered a relative indication [7].
`Enker et al. [7] point to the potential importance of the
`condition of the posterior elements in assessing results of
`a disc prosthesis, because the posterior elements play a
`role not only in a (partly) load-carrying capacity and in re-
`sisting shear forces, but the facet joints also play a role in
`the kinematic functioning of the motion segment. Enker et
`al. therefore, suggest that ADR could be indicated in the
`face of intact posterior elements, to allow the ADR to serve
`as a load-sharing device in concert with the posterior ele-
`ments [7].
`There are more definite contra-indications, including
`spondylolisthesis, spinal stenosis, altered posterior ele-
`ments, infection, metabolic bone diseases like osteoporo-
`sis and osteomalacia, severe scarring after previous surgery
`and insufficient motivation of the patient. Spondylolisthe-
`sis should be a contra-indication because of the risk of
`dislocation [11]. In spinal stenosis, the encroachment will
`not be released by ADR. Altered posterior elements, like
`
`prior decompressive laminectomy, prior posterior fusion ef-
`forts [22] or degenerative facet joints [3, 22] are also re-
`ported as contra-indicative. We think that here, a fusion in
`the long run will provide more stability than an ADR with
`the possibilities of slip or wear. Previous and latent infec-
`tions have also been mentioned as contra-indications [3,
`22]. Several authors [3, 11, 22] have concluded that ADR
`should be avoided in metabolic bone diseases such as os-
`teoporosis and osteomalacia, because of subsidence. Wit-
`tig et al. [23] add severe scarring (at the same or other
`level) after spinal surgery to the list of contra-indications,
`because in their experience this problem is not solved by
`the distraction provided by the prosthesis. Some authors
`also report insufficient compliance or motivation of the
`patient [3, 22].
`
`Segmental mobility by ADR
`
`Our series shows maintenance of segmental mobility in
`patients with ADR. This was also shown in other clinical
`studies [6, 15], as well as in a cadaver study with this pros-
`thesis [1].
`
`Size and position of the prosthesis
`
`From a mechanical point of view, the prosthesis as a load-
`bearing device should be placed centrally in the interver-
`tebral disc space. Wittig et al. [23] describe that the centre
`of the implanted prosthesis did not exactly coincide with
`the natural axes for sagittal movements. However, like
`Büttner-Janz et al. [4], they did not find a relationship be-
`tween the horizontal position and postoperative conse-
`quences. Similarly, we did not find significant differences
`in clinical results between centrally and asymmetrically
`located prostheses. Wittig et al. [23] focus on the prob-
`lematic cranio-caudad tilt of the prosthesis, because this
`not only can diminish the range of motion but also, unfor-
`tunately, can relocate more compression forces from the
`weight-bearing disc prosthesis to the facet joints. Further
`kinematic studies will be needed to address this anticipated
`detrimental effect on motion and load sharing.
`
`Complications and re-operations
`
`We reported several complications. Some of them have
`been reported before. We had complications both due to
`the operation and to poor implantation technique. Büttner-
`Janz et al. [5] mentioned a low number of general compli-
`cations, as did Griffith et al. [11] in 1994, related to the
`anterior surgical approach.
`Two of our patients required evacuation of an anterior
`abdominal wall haematoma. Frymoyer [10] mentions a
`1–2% chance of sexual impotence in male patients by us-
`
`6 of 8
`
`

`
`216
`
`ing the anterior approach for the two lower levels, partic-
`ularly the L5-S1 level. We had one patient with temporary
`sexual problems (retrograde ejaculation). We did not en-
`counter any infections of the prosthesis in our series. We
`observed complications due to poor implantation tech-
`nique in only two patients. Büttner-Janz et al. [5] men-
`tioned 10 times intracorporal migration and 2 times ven-
`tral dislocation in 67 endoprostheses. They state [2, 4, 5]
`that migration or slipping can probably be reduced by us-
`ing the largest area of contact between the endplates and
`the prosthesis. Apart from a doubtful difference in posi-
`tion (< 2 mm) in the transversal plane, no significant mi-
`gration in this plane or slip could be seen in our patients
`who attended the 2-year follow-up. Some of the complica-
`tion-related re-operations will be due to a learning curve
`that every physician will go through when learning a new
`technique [11]. The rate of complications caused directly by
`the implantation of the disc prosthesis was, according to
`Griffith [11], 6.5% of the patients (4.3% of all prostheses
`implanted). This compares to 4% in our series (due to mal-
`positioning, but not due to the material). From our first
`and only patient with revision of the prosthesis in this series
`we learned about its difficulties and dangers (Table 5 C).
`Comparing complications of a disc prosthesis and those
`of an anterior interbody fusion is very difficult. Despite
`the same surgical approach, specific complications and re-
`operations are associated with each type of implant, like
`dislocation of the prosthesis or complications of spinal fu-
`sion (outlined in the Introduction). Studies differ not only
`in indications, patient selection and follow-up, but also in
`the assessment of side-effects or complications. Some stud-
`ies mention postoperative transient hypotonic ileus, but
`do not count it as a significant complication [20, 21]. This
`also can be the case for bladder dysfunction [21]. We did
`pay attention to these minor problems, including altered
`
`sensation or muscle strength, scarring, haematomas, veg-
`etative changes and abdominal problems. In this way, we
`found that 15% of our patients had side-effects or compli-
`cations of a more permanent character. Interbody fusion
`studies have reported complication rates of 6–25% of their
`patients [17, 19–21], while permanent problems are re-
`ported here as 0–25% [17, 19, 20]. Twenty-four percent
`of our patients were reoperated: 29% (7/24) for complica-
`tions, and the rest for improving pain control. Anterior in-
`terbody studies have shown reoperation rates of 6–20%,
`with 12–100% of these being performed for complica-
`tions or non-unions [19–21]. The only effort for revision in
`our present study resulted in damage to the large vessels.
`Vascular damage to the large vessels is also reported in
`1–4% in some studies of anterior interbody spondylodesis
`[17, 21], while other studies did not report this problem
`[19, 20]. Larger studies with a much longer follow-up are
`needed to learn the true indications for ADR. In this lim-
`ited group of 50 patients with severe low back pain, of
`whom 54% had undergone previous failed spinal surgery,
`we report satisfactory clinical results (70%). From our to-
`tal population of more than 350 operated patients with a
`disc prosthesis, this study concerns only our first 50 suc-
`cessive patients. It is important to realise that this study
`includes our learning curve for technique, criteria and pa-
`tient selection. Therefore, we need a long-term follow-up
`of a larger group of patients with optimal (radiographic
`and psychological) selection criteria. Finally, important
`questions like migration, collapse, kinematics and wear of
`the prosthesis can only be studied satisfactorily in a long-
`term follow-up study.
`
`Acknowledgements We thank Dr. Nico Bohnen, neurologist,
`Pittsburgh, USA, for correction of the English in the manuscript.
`
`References
`
`1. Ahrens JE, Shelokow AP, Carver JL
`(1997) Normal joint mobility is main-
`tained with an artificial disc prosthesis.
`Paper presented at the North American
`Spine Society meeting, New York, Oc-
`tober 1997
`2. Büttner-Janz K, Schellnack K (1990)
`Bandscheibenendoprothetik, Entwick-
`lungsweg und gegenwärtiger Stand.
`Beitr Orthop Traumatol 37 : 137–147
`3. Büttner-Janz K, Schellnack K, Zippel
`H (1987) Eine alternative Behand-
`lungsmethode beim lumbalen Band-
`scheiben Schaden mit der Band-
`scheiben-endoprothese Modulartyp SB
`Charité. Z Orthop 125 : 1–6
`4. Büttner-Janz K, Schellnack K, Zippel
`H (1989) Biomechanics or the SB
`Charit