Eur Spine J (2002) 11 (Suppl. 2) : S124–S130
`DOI 10.1007/s00586-002-0446-2
`
`O R I G I N A L A RT I C L E
`
`H. M. Mayer
`K. Wiechert
`A. Korge
`I. Qose
`
`Minimally invasive total disc replacement:
`surgical technique
`and preliminary clinical results
`
`Received: 30 April 2002
`Accepted: 10 May 2002
`Accepted: 10 May 2002
`Published online: 9 August 2002
`Published online: 9 August 2002
`© Springer-Verlag 2002
`
`H.M. Mayer (✉) · K. Wiechert · A. Korge
`Spine Center,
`Orthopedic Clinic Munich-Harlaching,
`Harlachinger Strasse 51,
`81547 Munich, Germany
`e-mail: MMayer@schoen-kliniken.de,
`Tel.: +49-89-62112011,
`Fax: +49-89-62112012
`
`I. Qose
`Ancona University, Italy
`
`Abstract Total disc replacement has
`become an option for the treatment
`of degenerative disc disease of the
`lumbar spine. A new generation of
`implants has been developed that can
`be implanted through minimally in-
`vasive anterior approaches to the
`lumbar levels L2/3, L3/4, L4/5 and
`L5/S1. However mid- and long-term
`data are still lacking. This paper de-
`scribes the minimally invasive surgi-
`cal approach – techniques as well as
`the preliminary results of our first
`34 consecutive patients. The inter-
`vertebral spaces L5/S1, L4/5, L3/4
`and L2/3 were each approached
`through slightly different, but stan-
`dardized, mini-laparotomies either
`through a retroperitoneal or a trans-
`peritoneal route. The clinical results
`
`with a follow-up of up to 1 year
`show satisfactory outcomes in about
`80% of the patients. Oswestry score
`as well as VAS values show signifi-
`cant changes during the postopera-
`tive course. There have been three
`complications (8.8%), two of which
`were specific to the implantation
`process, but were resolved with a
`good clinical outcome in both pa-
`tients. The preliminary results sug-
`gest that total disc replacement may
`become a reasonable alternative to
`spinal fusion under the selection cri-
`teria used in this study.
`
`Keywords Artificial disc ·
`Degenerative disc disease · Lumbar
`spine · Total disc replacement ·
`Low-back pain
`
`Introduction
`
`Degenerative disc disease remains a therapeutic chal-
`lenge. The therapeutic gap that existed between conven-
`tional non-surgical treatment and spinal fusion surgery
`has been closed by a variety of so-called ‘semi-invasive’
`techniques. Epidural catheter treatments, intradiscal elec-
`trothermal therapy (IDET), and partial (nucleus pulposus)
`or total disc replacement techniques have been developed
`and are being used more and more frequently in clinical
`studies, although safety and efficacy are not yet evidence
`based [1, 2, 3, 4, 5, 10, 11, 12]. This is also true for total
`disc replacement in the therapeutic regime of chronic
`discogenic low back pain. The clinical and radiological
`results thus have to be monitored closely. A new genera-
`tion of implants for total disc replacement has been devel-
`oped, designed especially for application through a mini-
`
`mally invasive anterior approach (Prodisc, Spine Solu-
`tions, Tuttlingen). It has been in clinical use since 1999.
`We report on the results of our own series of patients,
`which formed part of an international prospective clinical
`multicenter trial.
`
`Implant and minimally invasive surgical technique
`
`The clinical study was performed with a new generation
`implant (ProDisc). The modular implant technology al-
`lows a stepwise implantation, which fulfils all necessary
`criteria for a minimally invasive surgical access (Fig. 1).
`Preparation as well as application instruments play a key
`role in minimally invasive approaches for total disc re-
`placement. Thus, all instruments for preparation of the
`implantation (probe implant, distractor, chisels) as well as
`GLOBUS MEDICAL, INC.
`EXHIBIT 1014
`IPR2015-to be assigned
`(Globus v. Flexuspine)
`1 of 7
`
`

`
`S125
`
`the patient influences the surgical technique and, in rare
`cases, might lead to a contraindication for disc replace-
`ment (e.g. venous bifurcation covering completely the an-
`terior circumference of the target disc space). All other
`preoperative planning criteria correspond to the ones that
`have been described for minimally invasive anterior inter-
`body fusion (MiniALIF) [6, 7, 8, 9].
`All implantations can be performed through a midline
`mini-laparotomy. The patients are placed in a neutral
`mini-ALIF position (cave: hyperextension of the lumbar
`spine increases segmental lordosis) (Fig. 3). The target
`level is localized under antero-posterior and lateral fluoro-
`scopic control and marked on the skin. All implantations
`are performed through small 4- to 5-cm transverse skin in-
`cisions (Fig. 4). Because of anatomical and topographical
`details, each level has very specific technical demands.
`
`Fig. 1 ProDisc implant: modular design for minimally invasive
`step-by-step implantation
`
`Technique for the L5/S1 level
`
`instruments for performing the implantation (applicator,
`distractor, inlay insertion, etc.) have been designed fol-
`lowing microsurgical criteria (Table 1).
`
`Surgical approaches
`
`General remarks
`
`This is the easiest segment to approach. After exposure of
`the rectus fascial sheet, the linea alba is split in the mid-
`line, and the peritoneum is exposed. There are three op-
`tions for exposing the L5/S1 disc space from anterior:
`retroperitoneal from the right side, retroperitoneal from
`the left side or transperitoneal. The ‘approach decision’
`should follow the following guidelines.
`
`Total disc replacement requires an anterior midline ap-
`proach. Due to the designs of the implant, insertion into
`the intervertebral space must be performed strictly in the
`midline. This requires meticulous preoperative planning
`as well as a modification of the surgical technique, espe-
`cially at L4/5 and higher levels.
`Preoperative planning includes magnetic resonance
`imaging (MRI) investigation of the lumbar spine, as well
`as three-dimensional computed tomography (3D CT) an-
`giography, to evaluate the size, shape and topography of
`the retroperitoneal blood vessels (Fig. 2). This technique
`makes it possible to clearly visualize the venous and arte-
`rial bifurcation and also shows the topographical relation
`between the arterial and venous branches. With these pre-
`operative data, surgical planning can be performed in de-
`tail. The knowledge of the individual vascular situation of
`
`Retroperitoneal approach from the right side
`
`This approach should be the first choice. The peritoneum
`is bluntly detached from the inner abdominal wall on the
`right side. The transverse fascia has to be incised to mobi-
`lize the abdominal contents adequately. The psoas muscle
`as well as the common iliac artery with the urether are
`identified. Preparation is continued towards the midline
`between the urether (displaced medially) and the artery.
`Medial to the common iliac artery, the lateral circumfer-
`ence of L5/S1 can be exposed. In this area, the superior
`hypogastric plexus is very thin with rare and small branches,
`which decreases the risk of damaging this plexus. Blunt
`dissection of the prevertebral fat tissue including the
`plexus exposes the medial sacral artery and vein, which
`can then be clipped or coagulated and dissected. Thus
`L5/S1 can be exposed easily. The left common iliac vein
`
`Table 1 Instrument and im-
`plant properties for minimally
`invasive implantation
`
`Instrument
`
`Distractor
`Probe implant
`
`Chisel
`Implant applicator/distractor
`
`Properties
`
`Slim instrument design for small approach corridors
`Same size as implant, fits through small corridor easy, coupling
`mechanism guides chisel
`Guided by probe implant, no additional space required
`Same width as implant, no additional space required, easy coupling/
`uncoupling, easy distraction and inlay insertion
`
`2 of 7
`
`

`
`S126
`
`Fig. 2a, b Three-dimensional
`computed tomography (3D CT)
`angiography of the retroperi-
`toneal blood vessels of the
`lumbar spine: a arterial branch,
`b venous branch
`
`Fig. 3 Positioning of the patient (nb hyperextension must be
`avoided!)
`
`can be retracted carefully to the left (Fig. 5). This is the
`safest and easiest approach to L5/S1.
`
`Retroperitoneal approach from the left side
`
`This approach is chosen in cases with previous abdominal
`surgery in the lower right quadrant (e.g. appendectomy, gy-
`necological operations, operation for abdominal hernia).
`The dissection process is the same as on the right side. Dis-
`section is performed across the common iliac vein to the
`disc space L5/S1, which is sometimes difficult, especially
`if the vein has a large diameter. The plexus hypogastricus
`superior has to be pushed medially with care, avoiding any
`coagulation. These two factors make this approach the
`‘second-choice approach’; however, exposure of L5/S1 can
`be achieved as completely as from the right side.
`
`Fig. 4 Skin incision for mini-laparatomy
`
`Transperitoneal approach
`
`In very obese patients, in patients who have had conven-
`tional abdominal surgery and in revision cases, the trans-
`peritoneal minimally invasive approach is the appropriate
`technique. It is the most direct way to L5/S1, and can be
`performed easily even in obese and previously operated
`patients [7].
`
`Technique for the L4/5 level
`
`Vascular anatomy determines the approach to L4/5 (see
`Fig. 2B). Due to the venous anatomy, the retroperitoneal
`approach from the left side is preferred in conventional
`surgery. Dissection can be performed across the aorta or
`the common iliac artery first. The arcuate line has to be
`incised in order to get adequate mobilization.
`
`3 of 7
`
`

`
`S127
`
`ing lumbar vein on the left side to prevent indirect injury
`to these structures.
`
`Technique for the L2/3/4 levels
`
`The approach to L3/4 and L2/3 needs modifications on
`the skin-to-spine-route. The skin incision is usually at the
`level of, or above, the umbilicus. If it is at the umbilical
`level, a small, longitudinal paramedian incision on the left
`side is preferred. Retroperitoneal exposure is much more
`difficult at these levels, since the peritoneum is adherent
`to the posterior rectus sheet. Innervation of the rectus
`muscle must be preserved and the integrity of the fascial
`indentations at these levels must be respected. It is thus
`recommended to expose the retroperitoneal space in two
`steps: (1) longitudinal midline incision of the anterior rec-
`tus sheet 5 mm lateral to the linea alba and exposure of the
`left rectus muscle, and (2) dissection anterior to the mus-
`cle to its lateral border and opening of the retroperitoneal
`space. Thus, the peritoneum can be detached from the
`posterior rectus sheet from left lateral to the midline. The
`exposure is then continued by opening of the posterior
`rectus sheet close to the midline and retroperitoneal dis-
`section from the left to the right. In obese patients, again,
`a transperitoneal route is recommended. The various op-
`tions of vascular preparation are shown in Fig. 7.
`After removal of the nucleus pulposus and after end-
`plate preparation, the implant is positioned according to
`manufacturer’s guidelines.
`
`Materials and methods
`
`Study design
`
`This was a prospective, non-randomized clinical multicenter study.
`All patients had to give written informed consent. Study documen-
`
`4 of 7
`
`Fig. 5 Three-dimensional CT angiography: direction of vascular
`mobilization and retraction at L5/S1 (r.c.i.v. right common iliac
`vein, l.c.i.v. left common vein, 1 ligation of medial sacral artery
`and vein)
`
`However, retroperitoneal exposure of L4/5 has its lim-
`itations in a minimally invasive approach. Mobilization of
`the abdominal contents is more difficult through a 4- to
`5-cm skin incision. The same is true for preparation and
`retraction of the blood vessels. Due to the lordotic curve
`of the lumbar spine, the distance between the L4/5 disc
`space and the anterior abdominal wall is quite short. This
`makes a direct transperitoneal approach reasonable. Easy
`orientation and dissection of the superior hypogastric
`plexus and the perivascular tissues are further advantages.
`Exposure of the disc space follows the vascular situation
`(Fig. 6). Care has to be taken to ligate and dissect all seg-
`mental arterial and venous branches as well as the ascend-
`
`Fig. 6a, b Three-dimensional
`CT angiography: direction of
`vascular mobilization, retrac-
`tion and branches to be ligated
`– most common variants at
`L4/5. a Approach between ar-
`terial bifurcation, lateral to ve-
`nous bifurcation (v.b. venous
`bifurcation, r.c.i.a. right com-
`mon iliac artery, l.c.i.a. left
`common iliac artery; ligations
`of 1 medial sacral vein, 2 as-
`cending lumbar vein, 3 medial
`sacral artery, 4 segmental vein
`L5 left). b Approach between
`venous and arterial bifurcation
`(v.b. venous bifurcation, r.c.i.a.
`right common iliac artery; liga-
`tion of 1 medial sacral vein
`and artery, 2 ascending lumbar
`vein, 3 right segmental artery
`L4, 4 left segmental vein L4)
`
`

`
`S128
`
`Fig. 7a, b Three-dimensional
`CT angiography: direction of
`vascular mobilization, retrac-
`tion and branches to be ligated
`– most common variants at
`L2/3/4. a Approach between
`the vena cava and the abdomi-
`nal aorta (v.c. vena cava, a.a.
`abdominal aorta; ligation of
`1 segmental vein L4 left,
`2 segmental artery L4 right,
`3 segmental vein L3 left,
`4 segmental artery L3 right).
`b Approach from the left side.
`(v.c. vena cava, a.a. abdominal
`aorta; ligation of 1 segmental
`vein L4 left, 2 segmental artery
`L3 left, 3 segmental artery L4
`left)
`
`tation was standardized, and included the visual analog scale
`(VAS), the Oswestry Disability Score, the SF36 Health Question-
`naire and numerous clinical and radiological parameters. Data
`acquisition was performed preoperatively and at 3, 6, 12 and
`24 months postoperatively. For each follow-up visit, radiographs
`of the lumbar spine in antero-posterior and lateral projection plus
`functional views in flexion and extension were acquired.
`
`Table 2 Diagnosis for total disc replacement (FBS failed back
`syndrome)
`
`Degenerative disc disease
`Degenerative disc disease + disc herniation
`FBS/postop. osteochondrosis
`Adjacent level degeneration
`Degenerative following nucleus replacement
`
`61.8% (21/34)
`11.8% (4/34)
`14.7% (5/34)
`8.8% (3/34)
`2.9% (1/34)
`
`Patient selection
`
`The indications were mono- or bisegmental lumbar disc degenera-
`tion and postoperative disc degeneration, as well as osteochondro-
`sis and degeneration of levels adjacent to a former lumbar fusion.
`In all patients, symptoms had not responded to an extensive course
`of outpatient and inpatient physiotherapy including fluoroscopy-
`guided infiltrations as part of the preoperative workup. Conserva-
`tive treatments were performed for more than 6 months in all pa-
`tients. The symptoms of the patients had to be concordant with the
`results of preoperative imaging.
`Contraindications were all kinds of translational instability
`(e.g. spondylolisthesis), spinal stenosis, significant osteoarthritis of
`the facet joints, deformities, infection or tumor, unwillingness to
`comply with study requirements regarding follow-up visits and ra-
`diological controls, previous fusion attempts in the affected levels,
`pregnancy and incomplete worker’s compensation procedures.
`All patients were operated according to the surgical philosophy
`described above. Postoperatively, the patients were mobilized on
`the day after surgery. With physiotherapeutic advice, most patients
`were able to be discharged a few days postoperatively.
`
`Results
`
`Patient population
`
`Between June 2000 and March 2002, 34 patients were op-
`erated. Gender distribution was 12 males and 22 females.
`Average age was 44.0 years, ranging from 25.2 to 65.4 years.
`The predominant diagnosis was degenerative disc disease
`in 61.8% (21 patients), while disc degeneration in combi-
`nation with a median nucleus pulposus herniation was
`found in 11.8% (four patients). Five patients (14.7%) had
`
`a postoperative osteochondrosis following disc surgery,
`three patients (8.8%) had a disc degeneration adjacent to a
`spinal fusion, and one patient had a dislocated nucleus re-
`placement device at the affected level (Table 2). The lum-
`bosacral motion segment, L5/S1, was affected in most
`cases (24 patients; 70.6%). In three patients (8.8%) it was
`L5/6, in a further three patients L4/5 was symptomatic,
`and in three patients we found a bisegmental affection in
`L4/5 and L5/S1. One patient (2.9%) had an affection of
`L2/3.
`
`Intra-operative data
`
`In 54.8% a transperitoneal approach was used and in
`45.2% a retroperitoneal midline approach was used. Mean
`operating time was 130.9 min, ranging from 88 to
`300 min, with a standard deviation of 45.9 min. Average
`blood loss was 117 ml per level (range 30–350 ml).
`The Prodisc implant is available in two sizes. The “me-
`dium” size was used in 36 of 37 affected segments, and
`“large” was used at one level. The implant with 6° lordo-
`sis angle was used in 65.4%; the 11° angle was used in
`sis angle was used in 65.4%; the 11° angle was used in
`
`
`
`
`34.6%. The polyethylene inlay of 10 mm height was used34.6%. The polyethylene inlay of 10 mm height was used34.6%. The polyethylene inlay of 10 mm height was used34.6%. The polyethylene inlay of 10 mm height was used34.6%. The polyethylene inlay of 10 mm height was used
`34.6%. The polyethylene inlay of 10 mm height was used
`
`
`
`
`in 34 segments, the 12-mm inlay in two segments and thein 34 segments, the 12-mm inlay in two segments and thein 34 segments, the 12-mm inlay in two segments and thein 34 segments, the 12-mm inlay in two segments and thein 34 segments, the 12-mm inlay in two segments and the
`in 34 segments, the 12-mm inlay in two segments and the
`14-mm inlay in one segment.
`14-mm inlay in one segment.
`
`5 of 7
`
`

`
`S129
`
`The duration of the postoperative hospital stay aver-
`aged 12.0 days (range 4–25 days).
`At the time of the latest follow-up, 60.9% of all pa-
`tients were “completely satisfied”, while 21.7% were
`“satisfied“, which gives an overall success rate according
`to the subjective rating of 82.6%. However, all of the
`17.4% who were not satisfied with the clinical result
`stated that they would undergo the operation again if they
`were again faced with that choice.
`
`3,4
`
`2,4
`
`2,2
`
`6,3
`
`7,0
`
`6,0
`
`5,0
`
`4,0
`
`3,0
`
`2,0
`
`1,0
`
`0,0
`
`Score
`
`pre
`
`3m
`
`6m
`
`12m
`
`Follow-Up
`
`Radiological results
`
`Regarding all 37 devices implanted, we saw no loosening
`of the implant and no migration. Follow-up radiographs
`showed no change in the function of the implant over
`time. Endplates of the adjacent vertebrae did not show
`any subsidence.
`
`Complications
`
`In 91.2% of all patients, we did not see any complications.
`We have seen three complications related to the surgical
`procedure. There were no intra-operative complications,
`no general complications and no deaths. No patient in our
`series had to be fused in a re-operation of the affected seg-
`ment. We did not see any superficial or deep infections.
`One patient experienced a nerve root irritation of the
`L5 nerve root several days postoperatively. Computed to-
`mography revealed an extra-foraminal protrusion of nu-
`cleus material compromising the L5 nerve root on the left
`side. Neurological examination was normal. A 3-week
`outpatient course of conservative treatment and perineural
`infiltrations led to complete and permanent pain reduc-
`tion.
`One other case showed substantial pain reduction in
`the immediate postoperative course. Five weeks later, an
`increase of pain was noted with no notable trauma. Radio-
`graphs showed an inlay dislocation anteriorly. The surgi-
`cal revision revealed an intact polyethylene inlay. The
`endplates were solidly fixed and showed no signs of loos-
`ening. They were removed and the whole device was re-
`placed by a new implant. We believe that this case was a
`technical failure at the time of the first implantation, when
`the polyethylene inlay obviously was not completely
`snapped into the inferior endplate. The ongoing clinical
`course was uneventful, pain reduction was achieved, and
`radiographic controls were normal. One patient com-
`plained of a retrograde ejaculation at 3 months follow-up.
`
`Fig. 8 Visual analog scale (VAS) pre-operatively and after total
`disc replacement
`
`Clinical results
`
`Twenty-six of 34 patients (76.5%) attended at least one
`follow-up visit for evaluation. The remaining patients had
`not yet finished the first 3-month interval postoperatively.
`Average follow-up was 5.8 months, with a standard devi-
`ation of 3.0 months.
`The VAS scale averaged 6.3 points preoperatively.
`It was reduced by 3.9 points on average, ranging from
`8.4 points reduction to 7.5 points increase (Fig. 8).
`The Oswestry score ranged from 1 to 32 points before
`surgery, with an average of 19.1 points (standard devia-
`tion 7.4 points). It was reduced postoperatively by an av-
`erage of 11.5 points. The change in the postoperative
`score ranged from 27 points reduction to an increase of
`12 points (standard deviation 9.6 points) (Fig. 9).
`While all patients had low-back pain before surgery,
`76% had no low-back pain at the time of their latest fol-
`low-up.
`Presently, we do not see any difference in clinical out-
`come between the subgroups of those patients previously
`operated or between the patients with a bisegmental ver-
`sus those with a unisegmental implantation.
`
`9,2
`
`6,3
`
`7,2
`
`19,1
`
`25,0
`
`20,0
`
`15,0
`
`10,0
`
`5,0
`
`0,0
`
`Score Points
`
`pre
`
`3m
`
`6m
`
`12m
`
`Conclusions
`
`Follow-Up
`
`Fig. 9 Oswestry Low Back Disability Index pre-operatively and
`following total disc replacement
`
`These are preliminary results with a new implant for total
`disc replacement. As compared to a first-generation com-
`
`6 of 7
`
`

`
`S130
`
`peting implant system, which has been in clinical use
`since 1984, there are striking differences with respect to
`the design and implantation technique [1, 4]. The major
`advantage of this kind of implant is that removal of the
`disc, distraction of the intervertebral space and insertion
`of the device can be performed through standardized min-
`imally invasive approaches, which have already been used
`successfully for anterior lumbar interbody fusion with
`slight modifications. Minimally invasive approaches are
`possible even in difficult anatomical regions such as L4/5
`and higher lumbar levels. The perioperative results show
`that iatrogenic morbidity is very low. All patients could
`get out of bed the day after the operation. For study and
`academic reasons, we kept most of the patients in hospital
`for more than 1 week; however, our data suggest that the
`majority of the patients will be able to leave the hospital
`after a few days where there is an uneventful periopera-
`tive course. Intra-operative data are virtually the same as
`with anterior interbody fusion, except for the fact that
`there is no co-morbidity at the donor site for bone grafts
`
`[6]. Although this was not a randomised study, the early
`clinical results are promising. There were two “specific”
`complications, which were resolved and ended up with a
`favorable clinical result. The postoperative L5 root irrita-
`tion in one patient was most probably the result of inade-
`quate removal of the nucleus pulposus. Since the implant
`is space-occupying, there might be a certain risk that re-
`maining disc tissue is pushed towards the spinal canal or
`the foramen while the implant is impacted into the disc
`space. The second complication (anterior dislocation of
`the polyethylene inlay) was definitely a technical error
`during the implantation. The snap-locking mechanism of
`the inlay prevents dislocation, but requires precision dur-
`ing the insertion. The anterior border of the inlay must be
`in line with the anterior border of the inferior endplate.
`Even slight “steps” of less than 1 mm should not be toler-
`ated.
`In summary, although preliminary, the results suggest,
`that total disc replacement for the indications mentioned
`above might be a reasonable alternative to lumbar fusion.
`
`References
`
`1. Büttner-Janz K, Schellnack K, Zippel
`H (1989) Biomechanics of the SB
`Charite lumbar intervertebral disc en-
`doprosthesis. Int Orthop 13:173–176
`2. Cinotti G, David T, Postacchini F
`(1996) Results of disc prosthesis after a
`minimum follow-up peroid of 2 years.
`Spine 21:995–1000
`3. Danish Institute for Health Technology
`Assessment (1999) Low-back-pain –
`frequency, management and prevention
`from an HTA perspective. Danish In-
`stitute for Health Technology Assess-
`ment series, vol. 1(1)
`
`4. David T (1993) Lumbar disc prosthe-
`sis: surgical technique, indications and
`clinical results in 22 patients with a
`minimum of 12 months follow-up. Eur
`Spine J 1:254–259
`5. Enker P, Steffee A, McMillin C, Kep-
`pler L, Biscup R, Miller S (1993) Arti-
`ficial disc replacement. Spine 18:1061–
`1070
`6. Mayer HM (1997) A new, microsurgi-
`cal technique for minimal invasive
`anterior lumbar interbody fusion
`(MINIALIF). Spine 22:691–700
`7. Mayer HM (2000) Microsurgical ante-
`rior interbody fusion: the transperi-
`toneal approach to L5/S1. In: Mayer
`HM (ed) Minimally invasive spine sur-
`gery. Springer, Berlin Heidelberg New
`York, pp 133–144
`8. Mayer HM (2000) The ALIF concept.
`Eur Spine J 9:S35-S43
`
`9. Mayer HM (2000) Minimally invasive
`anterior arthrodesis of the lumbar
`spine. In: Duparc J (ed) Surgical Tech-
`niques in Orthopaedics and Traumatol-
`ogy, 55–070-E-10. Elsevier, Paris
`10. Saal JS, Saal JA (2000) Management
`of chronic discogenic low back pain
`with a thermal intradiscal catheter.
`Spine 25:382–388
`11. Scott AH, Harrison DJ (2000) Increas-
`ing age does not affect good outcome
`after lumbar disc replacement. Int Or-
`thop 24:50–53
`12. Zeegers WS, Bohnen LMLJ, Laaper
`M, Verhaegen MJA (1999) Artificial
`disc replacement with the modular type
`SB Charite III: 2-year results in 50
`prospectively studied patients. Eur
`Spine J 8:210–217
`
`7 of 7