Voir la version française

Minimum-access techniques have been introduced throughout surgery, including orthopaedics and traumatology, where, since the early 80s, arthroscopy has revolutionized the treatment of joint disorders.

For a long time, the spine did not benefit from these innovations. This situation was changed when, in 1991, Obenchain performed a laparoscopic L5-S1 discectomy, followed, in 1992, by Thomas Zdeblick’s L5-S1 fusion by laparascopic placement of an interbody cage. In 1994, Rosenthal et al reported the first excision of a herniated thoracic disc by thoracoscopic surgery. At the same time, Le Huec and Husson performed the first endoscopic retroperitoneal approach to the lumbar spine. Together, these three techniques provide access to the thoracic and lumbar spine in its entirety.

Initially, these endoscopic techniques gave limited access that allowed simple treatments, since the new form of surgery was difficult and involved a learning curve. Working with their colleagues in general and thoracic surgery, and using the instruments already developed in the other subspecialties for minimally invasive surgery, allowed spinal surgeons rapidly to evolve minimal-access techniques in their discipline.

At the same time, other surgeons were developing minimal-access surgery with endoscopic visualization. This video-assisted surgery proved an important stage, since it reduced the size of the incisions and gave a better view of the surgical field, while still preserving the landmarks familiar to the surgeon from the days of open surgery. Video-assisted surgery is now establishing itself in a niche of its own, between conventional open surgery and fully endoscopic techniques.

We have been practising the various techniques for the past 4 years, and are presenting our experience concerning the advantages and disadvantages, the capabilities and limitations, of these modern modalities.

I - SURGERY OF THE THORACIC SPINE

As a rule, orthopaedic surgeons will start practising minimum-access surgery of the thoracic spine using a thoracoscopic technique: once the pneumothorax has been established, the approach is ready to be used, and the landmarks are readily recognizable.

o TECHNICAL ASPECTS

The first trocar is inserted at the usual site, in the 5th or 6th intercostal space. To prevent damage to the lung, a check should always be made during this insertion to rule out local pleural adhesions (Fig. 1).

The other trocars are then inserted under direct visualization, to allow optimum placement for the intended procedure. Where the procedure permits, thin (5 mm) trocars may be used, which will cause less damage to the intercostal vessels. Disposable trocars and instruments may be used, to reduce the risk of infection. However, the increased use of minimum-access procedures has led to the introduction of reusable instruments, in the interest of cost containment.

Photo 1 : Outside view of the chest of a patient positioned for thoracoscopy.

Photo 2 : Reusable instruments for endoscopic surgery. Instruments disassemble easily to facilitate cleaning.

The instruments available for minimum-access surgery have, by now, reached a high standard of technical development, and are easily cleaned and maintained (Fig. 2). Since they are designed for work in the depth of the thoracic cavity, these instruments are long, and therefore somewhat fragile. However, many manufacturers now offer ranges of very efficient instruments. Specific surgical instruments have also been adapted to suit the requirements of minimum-access orthopaedic techniques. Much thought has been given to the weight of the instruments involved, so as to prevent fatigue on the part of the surgeon.

The instruments currently available for endoscopic surgery are continually being perfected. They now make it possible to perform interbody fusion with metal cages, interbody dowels, or other devices (Fig. 3), plating (Fig. 4), or fixation using screws and rods. The hardware has been assessed in clinical use. Most of the instruments have been evolved from conventional ones, and should, therefore, look reasonably familiar.

Photo 3 : a) Conventional cylindrical metal cage

Photo 3 : b) Threaded interbody bone dowel allograft

Photo 4 : a) Z-plate

Photo 4 : b) The Dynalock plate (Sofamor-Danek) can be more readily fitted to the upper or lower T-spine

o ESTABLISHED INDICATIONS

The treatment of symptomatic thoracic herniated discs using a conventional approach is notoriously dangerous. The enlarged posterolateral approach described by Roy-Camille has been gradually superseded by an anterior approach. This anterior approach through a thoracotomy (Zdeblick) has proved to be superior. The endoscopic approach first described by Rosenthal has made this procedure less invasive, while optimizing visualization and, consequently, cord decompression.

While soft herniated discs do not present any particular problems (Fig. 5), calcified discs still harbour major risks.

Photo 5 : Soft herniated thoracic disc; such disc are easy to excise.

- With endoscopic techniques, the intercostal vessels can be avoided, which means a reduction in the risk to the blood supply of the cord via the artery of Adamkiewicz (Fig. 6).

 

Photo 6 : Preoperative arteriography of the cord

Photo 7 : Very large calcified herniated disc, resulting in major neurological compromise. Such discs are difficult to excise.

- Calcified herniated discs (Fig. 7) are invariably associated with an induration of the dura mater; often, the vessels within the cord will be extremely fragile and likely to decompensate (sometimes irreversibly) as the compression is relieved. Endoscopy allows the surgical approach to be minimized; however, the indications for surgery and the potential complications of calcified herniated discs remain unchanged. The only difference is the reduced risk to the regional blood supply to the cord.

The anterior release of stiff thoracic scoliotic deformities (Fig. 8) is another indication for an endoscopic approach. In conventional surgery through a thoracotomy, with removal of a rib, disc excision at several levels soon becomes difficult, and with curvatures > 90°, visualization is far from optimal. Thoracoscopy allows anterior releases to be extended over 7, 8, or 9 levels; switching the scope around the trocar sites will ensure that vision is always optimal. The instruments will always be aligned with the disc to be tackled, and the endplates may be reamed under endoscopic vision. The segmental vessels can always be preserved (Fig. 9), and the release of the rib head gives better mobilization of the spine as well as providing an interbody graft where this is required. The indications of this anterior release have not yet been definitively established. In principle, they should be the same as the indications of conventional surgery; however, the less invasive nature of the endoscopic technique should allow more patients to be selected for an anterior release. The release of thoracic lordosis makes it possible to restore the physiological curvature of the thoracic spine (Fig. 10); the release of stiff scoliotic deformities with a Cobb angle > 50° and reducing less than 25 or 30% on side-bending films would also be a good indication.

Photo 8 : Stiff thoracic scoliosis, reducing less than 30% (from 92° to 87°) on bending film	Photo 9 : Endoscopic view of scoliotic deformity managed with 8-level disc release. The segmental vessels have been well preserved.

Photo 10 : Preoperative radiograph of scoliosis with thoracic lordosis; and postoperative radiograph showing good restoration of thoracic curvature after thoracoscopic anterior disc release

o POSSIBLE INDICATIONS

- Unstable fractures of the T-spine are usually managed with posterior fixation. Anterior decompression may be useful in patients with an incomplete division of the cord. The anterior endoscopic approach to the thoracic spine for anterior decompression of the cord is time-consuming, difficult, and involves a substantial amount of bleeding. All these factors need to be borne in mind in patient selection. In these cases, we prefer to perform video-assisted thoracic surgery (VATS) through a 7-8 cm minithoracotomy. The rib fragment may be used for grafting. Multiply injured patients often have bilateral lung contusion, which rules out ipsilateral lung collapse by the anaesthetist. VATS makes it easier for the lung to be retracted without being collapsed. With the instruments currently available, supplemental fixation is quite straightforward.

Anterior interbody grafting without the need for decompression of the spinal canal, in patients that have already undergone conventional posterior stabilization, is an indication of choice (Fig. 11). The procedure is less time-consuming than through a thoracotomy, and allows sound cortico-cancellous grafting to be performed. Supplemental internal fixation may be provided without extending the operating time.

Photo 11 : a) Postoperative radiograph after T-spine fracture fixation with CD Horizon hardware, showing major anterior vertebral body bone loss, putting patient at risk of malunion	Photo 11 : b) Postoperative radiograph after T-spine fracture fixation with CD Horizon hardware, showing major anterior vertebral body bone loss, putting patient at risk of malunion

- Malunion of the thoracic spine is uncommon, but could lead to slow compression of the cord and may, therefore, require anterior surgery. Sites half-way down the T-spine may be accessed, with equal ease, via a conventional or an endoscopic approach. However, access to the upper T-spine (T2-T4) is extremely difficult in conventional surgery, because of the proximity of the great vessels; these difficulties are observed even with rib resection (Fig. 12). Endoscopic surgery makes it possible to resect the vertebral body and disc fragment that is causing cord compromise, without completely destabilizing the thoracic spine, while allowing additional grafting and fixation to be performed under excellent conditions (Fig. 13).

Photo 12 : MR scan of T3 compression fracture, with posterior displacement of disc causing slow cord compression with pyramidal signs	Photo 13 : Postoperative MR scan of the patient in Fig. 12, showing good decompression of the cord. The artefacts over the vertebral bodies were caused by the titanium plate used for supplemental fixation.

Thus, the anterior approach could be useful in the management of fractures and malunion in all cases where patient age, operating time, the expected blood loss, the quality of the cord decompression effected, and the expected quality of fusion and fixation are important considerations.

o COMPLICATIONS

There are not, as yet, many reports of the use of an endoscopic approach to the T-spine, and patient numbers in such studies as have been performed are small. However, virtually all the reports thus far published show endoscopic surgery to be associated with fewer, and with less severe, complications as compared with conventional surgery. It must, however, be remembered that endoscopic surgery is proper surgery, and demands the same precautions as conventional procedures. The treatment given must be the same as that provided through more extensive approaches. There is a learning curve, and training must be obtained from teams that have experience with minimum-access techniques, as well as from work on animals such as pigs (Fig. 14).

Photo 14 : Endoscopic spine surgery practice in a porcine model. Postgraduate Diploma course in the Laboratory of Experimental Surgery, Department of Surgical Technique Research and Applied Surgery, Bordeaux University Hospital

o FUTURE CONSIDERATIONS

The outlook of endoscopic surgery of the thoracic spine is very promising. Technically, the approach is easy. In practice, however, fewer patients would be managed with minimum-access surgery at the thoracic than at the lumbar level. Wholly endoscopic surgery and VATS complement each other. Surgeons will need to be familiar with both modalities.

The surgical management of scoliosis, of some fracture patterns, and of herniated discs of the T-spine, would appear to be the indications of choice. As things now stand, an anterior release of some thoracic scoliotic deformities may be followed by entirely endoscopic anterior fixation, with correction of the Cobb angle. However, larger numbers of patients will need to be studied in order to determine the benefits that may be obtained in this condition. World-wide, some 50 cases have been done to date, and the indications are beginning to be more clearly defined.

I I- SURGERY OF THE LUMBAR SPINE

Obenchain was the first to report a laparoscopic L5-S1 discectomy. However, this approach is not very relevant in the treatment of disc herniations at this level, and the technique was not, therefore, widely adopted. Zdeblick, Regan, and McAfee performed the first L5-S1 fusions with interbody cages. The success of the procedure helped popularize this form of surgery, and rekindled interest in the anterior fusion advocated by René Louis.

o LAPAROSCOPIC APPROACH

1 - Technical aspects Laparoscopy is performed with the patient positioned supine, legs together. (The so-called French position, with the patient’s legs spread, was abandoned because it was awkward for the assistant and made it difficult to position the fluoroscopy C-arm.)

The first trocar is always inserted at the umbilicus, after the creation of a CO2 pneumoperitoneum using a Palmer needle (Fig. 15). The so-called mini-open technique allows a safer approach for the insertion of the first trocar if there is a possibility of peritoneal adhesions; sealing is then provided by a special trocar (Blunt-tip; Origin, Rueil-Malmaison, France).

Photo 15 : Insertion of Palmer needle for establishment of CO2 pneumoperitoneum	Photo 16 : Portals used for access to the promontory, for L5-S1 fusion

The placement of the two trocars for exposure has undergone some change, since the conventional pattern recommended by the Americans does not make it easier for the surgeon to dissect over the sacral promontory. We place the trocars 10 cm apart in the right lower quadrant, so as to allow exposure to be performed under optimum conditions, with good control of the sigmoid colon (Fig. 16). The assistant handling the scope stands on the left, and thus does not get in the way of the surgeon standing on the right. The fourth trocar is used for the introduction of the interbody devices; it is always inserted at a suprapubic site. The exact insertion site will need to be established on the skin after a fluoroscopic check of the L5-S1 disc slope, so as to allow the trocar to be inserted without tensioning the skin (Fig. 17).

Photo 17 : Intraoperative view of trocar sites. Note position of suprapubic portal.

Photo 18 : Endoscopic view of dissection of median sacral artery, which is divided between clips.

The problem when approaching the L5-S1 disc consists in the dissection of the left iliac vein, which is easily damaged and must be held out of the way with great care. In a patient without a history of intra-abdominal surgery involving the pelvis, this dissection is comparatively straightforward. The dissection, ligation, and division of the median sacral artery does not present any major problems (Fig. 18). The only structure that requires special attention is the presacral autonomic plexus. In females, the plexus may be swept off the midline, without any risk of postoperative complications. In males, great care must be taken in the handling of this structure. Endoscopic magnification is certainly useful; however, the surgeon must also ensure that the plexus is swept aside as atraumatically as possible, avoiding the use of monopolar cautery. On occasions, the accessory sigmoid colon mesentery may hamper the retraction of the sigmoid, and will need taking down. Pins inserted into the body of L5 will allow the vessels to be retracted on either side of the L5-S1 disc space (Fig. 19).

The final technical difficulty concerns the placement of the two cages, which must finish up perfectly parallel and at the same depth in the intervertebral space (Fig. 20). At the time of writing, trials are going on to improve the technique; these studies will be reported in the near future.

Photo 19 : Intraoperative fluoroscopic check of retraction pins at the great vessel bifurcation over the promontory	Photo 20 : Postoperative radiograph showing two perfectly parallel and well-aligned tapered cages in the L5-S1 space

2 - Established indications

Disabling low back pain, either from degenerative disc disease or after discectomy, without cord or nerve root compression, is a prime indication. Anterior fusion with cage insertion has the great advantage of not destabilizing the lumbar paraspinal muscles, and of permitting single-level fusion without affecting the facet joints of the vertebra above the fusion level, as often happens when the procedure is performed through a posterior approach. The use of two parallel cages, with cortico-cancellous grafting between or in front of the cages, allows sound fusion to be obtained.

The technique has been used by a large number of authors, who report similar results, with a very high rate of fusion. Patient selection is a problem in its own right; endoscopic surgery merely offers a reliable and minimally invasive technique for the management of appropriately selected patients. We feel that the availability of a minimally invasive procedure should not lead to an excessive extension of the indications for this surgery.

3 - Other indications In other disorders, the utility of the technique will have to be assessed on a case by case basis, since the value of the endoscopic technique has not yet been proved in these cases. These other indications would mainly be Grade I and II painful spondylolisthesis. If there is no associated lytic lesion, the technique may certainly be used, since reduction is achieved as the cages are placed.

If, however, the condition is one of spondylolisthesis with separation, the retensioning of the ligaments as a result of anterior cage placement will not procure the same stabilization as with an intact neural arch. We therefore feel that, at present, the utility of the technique in such cases is very uncertain. However, attempts are now being made to provide supplemental internal fixation, which should allow the extension of the procedure to a larger number of patients. Surgery of spondylolisthesis with neural arch separation could be of particular importance in athletes. It is known that posterior surgery in athletes will, in many cases, prevent a return to sport at the former level. Anterior endoscopic surgery may make it possible to obtain pain relief without compromising the career of the athlete. However, this assumption is, at present, purely speculative, and larger multicentre studies will be required to prove or disprove this point.

4 - Complications of anterior endoscopic surgery at the L5-S1 level are inherently the same as those associated with a conventional approach.

Vascular compromise is as much of a risk as with conventional surgery; however, endoscopy often provides better visualization, and, hence, better control of the vascular hazard. If major vascular damage has occurred, conversion to an open procedure is always possible; the operating room must be equipped for standard open surgery. The assistance of an experienced vascular surgeon may be helpful. This means having a vascular surgeon as a member of the surgical team, or on stand-by. This principle applies also to conventional (open) surgery.

Damage to the ureter or the large or small intestine may occur in any intra-abdominal surgical procedure. With endoscopic surgery, great care should be taken when using monopolar cautery. Unless metal instruments are fully visualized at all times, tissue contact may result in burns. If such injuries occur, they must be treated forthwith.

The main risk in males is retrograde ejaculation as a result of damage to the presacral plexus. Little information concerning this risk is available from vascular surgeons, who perform major surgery around the aortic bifurcation in patients who are often elderly. Orthopaedic patients tend to be much younger, and the risk has been assessed in a very small number of studies. Thus, Zdeblick, in 112 patients, found 6% of males to have retrograde ejaculation after surgery; however, in most of these patients the condition resolved, since, at one year, only 1% had permanent retrograde ejaculation. It is, therefore, vital for male patients to be informed of this risk. Present legislation prescribes that the risk must be explained in the presence of the patient’s partner or of a witness, as the case may be; and that a form must be signed to the effect that the risk has been properly explained. We have found that more than 50% of the men who have fathered children and who are over 45 will accept the risk. Younger men without children are advised by us to deposit sperm preoperatively if they wish to undergo endoscopic surgery.

5 - Future developments of laparoscopic procedures in anterior L5-S1 disc surgery will be concerned with improved stabilization at that level, probably by the use of supplemental internal fixation hardware. This should make it possible to select more patients, in particular those with spondylolysis, for this surgery. Laparoscopic surgery of the L4-L5 level is being abandoned, because of the major vascular hazards involved, and because of the frequent need for time-consuming and difficult dissection.

o RETROPERITONEOSCOPY

1 - Technically, retroperitoneoscopy (RPS) would appear to be more difficult, since the retroperitoneal space is only a virtual cavity that has to be developed by the surgeon. However, this approach is now being used extensively by general, vascular, and urological surgeons, who use it for the treatment of inguinal hernias, for lumbar sympathectomies, and for some forms of nephrectomy. The cleavage plane is easy to identify, and the approach is ideal for access to the discs from L1 to L5. We started using this approach back in 1994, and reported our experience with the first fusions using interbody cages, in 1995.

Since then, we have performed 70 procedures. The establishment of the retroperitoneal space with an inflatable balloon (Origin) is straightforward (Fig. 21). Once the space has been developed, surgery may be performed safely. Instead of using a balloon trocar, one may also develop the space by simple CO2 insufflation. Dissection will take a little longer, but is customarily performed in this way by general surgeons repairing inguinal hernias. It is mainly surgeons who are not skilled in endoscopic techniques that have misgivings about the use of CO2. If the pressure is kept low (< 12 mm Hg), and the trocars are properly sealed, this approach is very sound. The only problem working with CO2 is loss of insufflation when sustained suctioning has to be used, e.g. in the management of certain fractures. In such a situation, we use a video-assisted technique for complicated procedures or for bulkier internal fixation. The approach and dissection may be started off with CO2, after which thin retractors (Sofamor-Danek; Aesculap) are placed against the vertebral bodies, under visual control, to allow the surgeon to continue the procedure safely (Fig. 22).

Photo 21 : Origin balloon dissector for introduction of endoscope and monitoring of establishment of retroperitoneal space	Photo 22 : Video-assisted approach of lumbar spine, with mini-retractors to maintain the retroperitoneal space. The patient is lying on his side.

The patient may be positioned in a variety of ways. Surgery may be performed with the patient supine (this is our customary practice); equally, patients may be positioned on their side. Either position has its advantages. With the patient supine, laparoscopy is straightforward; this allows close monitoring of the progress of dissection in the retroperitoneal space if there is a risk of adhesions, especially in patients with a history of trauma or of previous surgeries. With the patient on his or her side, laparoscopy is more difficult to perform, but the abdominal viscera drop forward, and video-assisted surgery may be performed under good conditions. The lateral approach involves a certain amount of bending of the spine, which may be bothersome in some indications. We use the lateral position for video-assisted surgery of the spine in trauma cases, since bleeding from the bone is easier to control this way.

In RPS, the psoas will often hamper access to the L4-L5 disc. However, retracting pins may be inserted to obtain good exposure of the discs and the vertebral body. Great care should be taken to ensure that the psoas is not torn, since that may lead to haematoma formation which may irritate nerve roots, especially those of the femoral nerves. The iliac graft donor site may be used for the introduction of orthopaedic targeting devices in line with the L4-L5 disc. Through this approach, a large (14, 16, or 18 mm diameter) fusion cage with an obliquity in the coronal plane of between 0° and 30° may be introduced (Fig. 23). Equally, two cages may be used - a 12 mm diameter posterior one, and a 16 mm diameter anterior one. Adjunctive lateral plating may be provided, to enhance the stability of the fusion construct (Fig. 24). In all of these cases, cortico-cancellous graft material is packed in front of and around the cage or cages, in order to fill the interbody space.

Photo 23 : Postoperative CT scan following insertion of a cage in the disc space, with an obliquity of 22°	Photo 24 : If sufficient stability cannot be obtained with the use of a single interbody cage, additional plating will provide optimum stabilization.

2 - Established indications
Low back pain resulting from degenerative disc disease or secondary to discectomy is a prime indication. In patients who have not undergone previous posterior surgery, the implantation of a single cage with an obliquity of 20-30° gives good stability that should dispense with the need for supplemental anterior internal fixation. If there has been previous surgery of the neural arch, or if the paraspinal muscles are poor (with evidence of fatty atrophy on the CT scan), we think that interbody fusion should be supplemented by anterior plating. We have always found this anterior plate easy to insert with the instruments now available, and are increasingly adding these plates as a routine measure (Fig. 25), since the paraspinal muscles of the patients undergoing this procedure cannot be relied upon to brace the spine adequately.

Photo 25 : Postoperative radiograph following interbody fusion with cage + plating, using an endoscopic retroperitoneal approach

The obliquity of the cage is difficult to measure during surgery, and, especially in women, the quality of the bony endplates is often less than optimal. Supplemental internal fixation enhances stability, and may, therefore, be expected to improve the fusion rate. In our first series of 25 patients (managed with single cages) reviewed at a minimum of 2 years, two required additional stabilization for persistent pain, which we attributed to failure of fusion. These two patients had previously undergone posterior surgery of the spine (at least two procedures). It is probable that prior surgery of the neural arch leaves a rotational instability that is poorly controlled by the implantation of a single cage. The stress views did not show any abnormal mobility.

The anterior grafting of burst fractures that have been posteriorly stabilized is also a good indication (Fig. 26). If reduction has been performed at the expense of the disc space, with a migrated disc fragment still inside the vertebral body, there is a high risk of nonunion (18-21%, according to reports in the literature). In such cases, the retroperitoneal approach allows supplemental cortico-cancellous grafting to be performed (Fig. 27).

Photo 26 : a) Lumbar spine burst fracture managed with posterior reduction and fixation; anterior bone defect remains. This is a good indication for filling of defect with an anterior graft, using an endoscopic retroperitoneal technique.	Photo 26 : b) Postoperative radiograph following grafting

Photo 27 : Interbody graft at 1 year, showing bony bridge linking the vertebral bodies (posterior fixation hardware has been removed).

The quality of posterior fixation is usually such as to obviate the need for associated anterior internal fixation.

The same approach may be used for the anterior release of some lumbar scoliotic deformities, since the retroperitoneal route gives access to the L1 to L5 disc levels. Supplemental fusion using cortico-cancellous grafts from the iliac crest improves the prospect of a good outcome of the pedicle screw procedure subsequently carried out in customary fashion.

3 - Indications undergoing evaluation
Cortico-cancellous interbody grafting for fusion, after posterior stabilization, in degenerative disease. In patients who have undergone several posterior procedures, the approach of the spinal canal is likely to increase fibrosis in the canal; also, in elderly or postmenopausal patients, the quality of the posterolateral fusion is often poor. In such cases, the association of posterior fixation with a simple anterior cortico-cancellous interbody fusion graft should produce a better outcome. This indication is now being studied.

Malunion should also be amenable to corrective surgery from a retroperitoneal approach. To date, too few cases have been treated for any definitive conclusions to be drawn. We have managed three cases with good results. Video-assisted surgery is useful in such cases.

4 - Complications
In theory, the possible complications using a retroperitoneal approach in endoscopic surgery should be the same as those encountered in standard open surgery; in our practice, we have gained the impression that the rate and severity of complications are markedly reduced compared with conventional procedures. In 70 patients operated on with retroperitoneoscopic surgery, there were no major vascular, urological (bladder or ureteric), or irreversible neurological complications. However, there were two psoas haematomas that gave rise to femoral neuralgia; fortunately, these problems resolved. It is important for surgeons to realize that psoas haematoma is a possible complication, and that prevention of psoas injury - by retraction with two pins - is the best policy.

“Coincidental” lumbar sympathectomy is possible, especially in the management of spinal trauma. This sympathectomy does not produce any adverse effects; however, the patient must be informed prior to surgery. With conventional open surgery, this sympathectomy is commonplace. When performing a retroperitoneal approach, the surgeon may also perforate the peritoneum, especially in patients with a history of trauma, since posttraumatic adhesions will make it more difficult to dissect the retroperitoneal space. In such cases, we perform a primary laparoscopy and monitor the passage of the retroperitoneal balloon dissector with laparoscopy. If the peritoneum is violated, laparoscopy allows a repair to be performed and to continue with the procedure.

There is a risk of iatrogenic discitis. Among our patients, we had one case of Staphylococcus epidermidis discitis, in which 3 months’ treatment with antibiotics was followed by fusion and a pain-free spine.

o FUTURE CONSIDERATIONS

The use of the retroperitoneal approach for therapeutic procedures is very promising, since this tissue-sparing anterior surgery allows most spinal procedures that do not require cord decompression to be performed. The implantation of an artificial disc such as the Husson coil (Synthes-Protek, Switzerland) could also be done using this approach, in the management of painful disc disease.

III - SURGERY OF THE THORACOLUMBAR JUNCTION

The T12-L1 disc spaces are still difficult to access, because of obstruction by the crura of the diaphragm, and because of the tight space available when using the thoracoscopic approach. These disc levels may, however, be reached with endoscopic techniques combining the different modalities.

The T11-T12 disc is reached by thoracoscopy; it is more easily accessed from the left, since the diaphragm dips lower on that side, and since, at that level, the aorta is in the midline and does not obstruct access to the disc. Excision of the T11-T12 disc may be performed, and interbody grafting may be done, especially in fracture surgery. Internal fixation is a little more difficult, but not unfeasible. Fixation would take the form of simple hardware for graft stabilization.

Access to the T12-L1 disc is frequently required in trauma surgery. Using the retroperitoneal route with gas dissection is difficult, because the kidney is in the way. Thoracoscopy will permit access; however, the crus will need to be dissected. The pleura will need to be incised over the vertebral body, the segmental vessels will need to be ligated, and the crus of the diaphragm gradually retracted towards the midline. This will allow the excision of the T12-L1 disc and partial corpectomy to be performed, and a cortico-cancellous graft to be inserted at this level. However, internal fixation is much more difficult to provide. Preoperative arteriography to establish the pattern of the artery of Adamkiewicz is mandatory. We have done ten anterior grafts at this level, after fixation had been performed from a posterior approach.

Argenson and Hovorka recommend video-assisted retroperitoneal surgery. This is an interesting technique, which allows minimum-access surgery for grafting and any anterior internal fixation that may be required.

CONCLUSION

Surgeons are slowly becoming less reluctant to use endoscopic surgery of the spine. Endoscopic visualization is but one of the techniques available to the surgeon. Where it is used, results should be the same as those of standard open procedures. The instruments now available allow minimum-access surgery to be performed. Video-assisted surgery is an in-between stage, technically speaking, that is of great interest. While endoscopy gives enhanced visualization, it involves a learning experience, since the surgeon needs to monitor the progress of surgery on a screen, without being able to check the instruments directly. However, in arthroscopy none of this is a problem any more. Spine surgeons do not necessarily do arthroscopy, which may, to some extent, account for their reluctance to adopt endoscopic spine surgery. The advantages of minimum-access spine surgery are becoming ever more apparent. What is required now is multicentre studies in order better to identify the indications of these new techniques.

(Transl KRMB)