Ronald Blackman M.D.
Kelly O'Neal M.D.
George Picetti III M.D.
Kaiser Permanente Medical Center
Oakland ,Calif USA

Anterior thoracic discectomy over multiple levels has been used in the surgical treatment of scoliosis and kyphosis for more than 30 years. There is no standard as to when this procedure should be carried out but the theory has been that in those kyphotic and scoliotic curves which do not correct well on either extension (for kyphosis) or lateral bending (for scoliosis), removing the disc may enhance correction of the curve.

Anaesthesia : The success of this procedure depends on proper anaesthesia technique. Visualization and performance of surgical technique is improved if the continuous motion of the lung can be eliminated. This goal may be hard to achieve particularly in young children and those with initially poor pulmonary function. Ventilatory function of the lung is the responsibility of the anesthesiologist.

Maintenance of an adequate arterial p O 2 can be affected by patient positioning: If the patient is in the lateral decubitus position, the "good" lung is usually the down lung and is compressed increasing the ventilatory pressure. The following techniques are utilized.

A double lumen endotracheal tube is used in persons having a trachea large enough to accommodate this and get adequate oxygenation. The cut-off is around 100 pounds body weight. A fibre-optic bronchoscope is needed to visualize and intubate the left main stem bronchus. When the tube is unilaterally clamped breath sounds on the clamped side should cease. It is important to clamp the tube once or twice prior to the start of the surgical incision to be sure tube placement is correct.

Arterial oxygen desaturation may occur. We have learned that this desaturation is often due to shifting of the endotracheal tube and may be rectified by tube repositioning (again, utilizing the fibre-optic bronchoscope). There reaches a point where reoxygenation of the unaerated lung becomes imperative and if desaturation is severe the procedure should be interrupted and revert to two lung ventilation.

In those patients weighing less than 100 pounds the trachea may be too small for a double lumen tube. In those cases intubating only the bronchus of the non operative side is the best approach.

Changing of patient position can alter the position of the endotracheal tube. Initially we flexed the table for better endoscopic visualization (opening up the disc spaces). In unflexing, the endotracheal tube position changed just enough to require reintubation. Probably it is safer to maintain one position once adequate ventilation occurs. During the procedure, if a drop in oxygen saturation continues both lungs can be ventilated by pulling the tube back so the double lumen tube will function as a single lumen. A significant drop in arterial oxygen saturation occurred in 4 of our patients and 2 of them were adequately handled in this manner. One 14 year old with an 80 degree thoracic curve, significant deformity and decreased pulmonary function would not maintain arterial oxygen saturation. This occurred early in our experience and the procedure was aborted without first trying the above recommendations.

Surgical Technique : With the patient in the lateral decubitus position the upper arm is placed in a support at right angles to the body: The shoulders and hips are taped for stability.The skin is marked with the outline of the ribs and scapula and the posterior axillary line is drawn. For scoliosis the initial incision is in line with the posterior axillary line. For kyphosis, the approach is posterior to the posterior axillary line. The intent is to come onto the postero-lateral spine and avoid working directly towards the spinal cord. The right side is generally a little easier and if one has an option the right side should be chosen. Kyphosis and right thoracic scoliosis fall into this category. Left thoracic curves are more unusual but should be done from the left side.

We prefer to have the surgeon and assistant at the patients' back with a second assistant (if desired) on the side of the patients' abdomen. This keeps the operator working from posterior to anterior away from the spinal cord. Two video cameras are set up so all participants - surgeon and scrub - have a view of the procedure without turning. The endoscope is a 10 mm 45 degree 'scope. The first port is made 12 mm in length in line with the upper part of the 8th or 9th rib. A finger is inserted to be sure the lung is deflated (if that is the mode) or that the lung is away from the finger at least. An endoscope is inserted through a plastic port (stabilization thread) for looking around. If the lung is deflated one should be able to see up to the apex of the chest and count the ribs from the first thoracic to the twelfth. The diaphragm limits the inferior exposure, but it's presence is noted so it will not be inadvertently perforated as additional (caudad) ports are placed.

A second and a third incision ten to twelve millimeters in length are made either up or down from the initial cut, superior to the rib. The ports are established and secured with plastic stabilization threads, and the spine is easily seen. Additional ports may be necessary. There are no designated ports; at different times, a best exposure and working situation will be obtained by switching camera, retractor and instruments to different ports. two other ports are for purposes of retraction and working on the spine. Keeping the desired area of the spine in view a cautery hook is placed on the pleura midway between the head of the rib and the anterior spine overlying a disc space. Normally the disc space has no vessels over it and the pleura can be pulled up by the hook and cauterized in successive movements proximally and distally leaving the segmental vessels over the body of the vertebra unscathed.

The disc or discs can then be cleaned with a Kittner. Each disc can be incised with a shortened cautery blade and removal of disc material can be done in a standard manner of subchondral loosening of the disc attachment by standard Cobb elevator and its' removal by use of rongeurs, angled curettes, and ring curettes. We have tried automatic shavers and burrs and find no advantage in their use. Bleeding of the subchondral bone can be controlled by packing the disc space with Surgicel . The Surgicel is always removed prior to closure. Bleeding is magnified through the endoscope and small amounts of blood can obscure the vision. A small stab wound of 5 millimeters through which a neural type suction tip is inserted may be necessary. Obvious bleeding points can be controlled by cautery forceps. These are standard surgical endoscopic instruments - long forceps with an attachment for the electrocautery. The surgical procedure can then move cephalad or caudad using either the three initial ports or by making additional ports the disc space can be approached directly in line of sight.

The goals in kyphosis, idiopathic and neuromuscular scoliosis and congenital scoliosis may be slightly different. In kyphosis the aim is particularly to sever the anterior longitudinal ligament and to remove the anterior part of the disc. From the right, the azygos vein is immediately adjacent to the anterior longitudinal ligament but the manoeuver can be safely performed by lifting up the pleura and gently dissecting the obvious tissue plane between bone/ligament and pleura. A retractor can be placed in this interval or the tissue grasped and an interval created to place a rongeur or angled curette and work from the blind side to the visualized side.

In idiopathic scoliosis the goal is removal of the lateral disc particularly, though as much disc as possible is removed including the anterior longitudinal ligament. The rotation of the spine must be mentally visualized while the disc space immediately adjacent to the head of the rib is cleaned of annulus. Staying anterior to the head of the rib should keep the instruments out of the spinal canal.

In congenital scoliosis the goal is more limited. We visualize the hemivertebra, adjacent disc space and the growth plates. This usually looks like a transverse "Y" The disc spaces on either side of the "Y" are cleaned and deepened. We curette local bone to fill the defect. This amount should be sufficient to form a fusion since the space is quite small. If pleura is closed over these bone chips healing should, (and did in our cases) occur.

We now supplement most of the discectomy procedures by filling the defect created with local bone plus Grafton . In lieu of an appropriate delivery system we use the 4mm tube and cannula from the Craig Bone Biopsy set. The 4mm tube can be filled with Grafton from its' syringe and the tube is then placed into the disc space and its' contents expelled by using the solid cannula to push forward. Ten cc.of Grafton is usually sufficient for 5 disc spaces. Additionally rib graft can be packed into the space by taking a piece of rib internally. This is accomplished by incising the periosteum longitudinally on the inner surface and then subperiosteally cleaning the lateral surface of the rib. A long Kerison rongeur can be used to cut the rib in two places. Bleeding surfaces are packed with Gelfoam. The rib is then cut into pieces and placed in the disc space. Care is taken to be sure the disc space is cleaned to bleeding bone.

Closure of the pleura is difficult and we have experimented with numerous ways. The pleura is often quite thin and readily tears. Simple interrupted suture takes a long time. Use of an autosuture appears to offer an alternate technique, though it too is laborious. Placement of even a few sutures with partial closure may assist in covering open spaces and decrease post operative bleeding and adhesions

A chest tube is routinely used to evacuate air and blood for 24 to 48 hours depending on the amount of fluid produced. Normally one of the surgical incisions is used; correct chest tube placement into the apex is verified through the 'scope.

Our initial idea was to do an anterior release procedure followed a week later by posterior fusion. However by 72 hours most patients were able to be discharged home to await the second procedure. Our operative time for the anterior procedure has decreased so that there is usually sufficient time to do the posterior procedure on the same day.

Results We have performed anterior endoscopic procedures on 34 persons. Of these 30 had anterior releases, multiple level discectomy and/or fusions for deformity of scoliosis or kyphosis. We have analyzed our results in these thirty patients.

Not included in these figures are two patients in whom we attempted this procedure but failed due to inability to maintain adequate oxygen saturation. The first was very early in our experience. She was a 13 year old girl with a thoracic curve of 73 degrees. On attempting to deflate the lung on the right side she consistently and rapidly dropped her oxygen saturation levels. Without prior experience we were unaware that this was most probably poor endotracheal tube positioning or a shift in the tube position, and the procedure was cancelled. With experience gained, we would now re-intubate with a fibreoptic bronchoscope. Our second patient, a 13 year old retarded girl with a thoracic curve of 65 degrees underwent the initial skin incisions but we were unable to maintain oxygen saturation levels even with full two lung ventilation, during manipulation of the lung. The procedure was abandoned. Today we would attempt bronchoscopy with suctioning and removal of any mucous plugs before abandoning the procedure.

Also not included in these figures was a 9 year old girl with a thoracic curve of 100 degrees and severe rotation of the spine with a large rib hump. After selective intubation of the left bronchus and with the lung deflated, the incisions in the chest wall were made. It then became obvious in this very thin, very small and very deformed person that there was insufficient room to insert the 'scope between chest wall and spine. We elected open thoracotomy. We have learned that curves over 80 degrees with severe rotation may not be candidates for this procedure.

Long term pulmonary function should not be affected by the procedure. We have shown a statistically similar effect on FEV1,FVC and TLC FEC ) not to be any worse than as reported by others doing similar procedures by open techniques. Out of 11 patients who had 6 month postoperative pulmonary function tests, FVC dropped in 8 = 73%; FEV1 dropped in 7 (= 64%). By one year however, 6 out of 8 had increased their FVC and FEV1 above their preoperative levels. Total lung Capacity (TLC) dropped in 6 out of 8 with reversion to above preop levels in 4 out of 6; 6 out of 8 including the initial 2 that didn't drop. Obviously the relationship between pulmonary function and the endoscopic procedure cannot be scientifically evaluated since we have inserted a posterior fusion with attempted three dimensional correction, multiple partial rib resections and a fusion. It would appear though that we have similar early results to others; with an inference that we have not done harm.

Logic would have it that this procedure should be ideal for those patients with neuromuscular disease. Recovery from invasion of the pleura without rib resection and large muscle incisions should be ideal. One of our patients, a 14 year old boy with severe pulmonary restrictive lung disease secondary to a 95 degree curve coupled with a pectus deformity an anterior multiple level release. Postoperatively he failed to develop adequate respiratory function for one month and was intubated for one week and then required 3 weeks of intensive "lung work". Extensive workup showed him to have unrecognized neuromuscular disease.We thus are cautious in the use of logic in predicting outcome in individual cases.

Blood loss was likewise hard to scientifically measure. Each procedure was a little different including lysis of an adhesion which bled; but there was no significant acute bleeding and chest tube drainage subjectively appeared little changed from open surgery. No patient required additional blood because of blood loss from their anterior approach.

Having determined that we can technically perform this procedure safely we need to see if its as good as an open release, or no release. to the latter, there are once again no scientific answers but many dogmas since the procedures are continually being modified, improved upon and added to. The thoracic curves ranged from 95 degrees to 42 degrees averaging 61 degrees. Preoperative bending went from the stiff curve at 85 degrees to flexible curves correcting to 20 degrees. Bending correction averaged 43 degrees. Correction went from an average preop curve of 61 degrees to 23 degrees postoperatively averaging 61% correction and 44% more than the bending ability

Operative time has gradually decreased from 45 mins per disc space to 15 minutes per disc space. Set up time appears longer than with open procedures.

Discussion :

This approach has been shown to be effective in decreasing morbidity, decreasing hospital stay and to successfully perform the desired procedure. Of greatest significance is the substitution of three to five one centimeter scars for the long costal incision. While the adage regarding beauty is appreciated at least to these authors the costal scar is the antithesis of beauty while the small endoscopic scars heal well and are barely visible.

We have had no neuroma or neuritic symptoms from the large endoscope between two relatively small ribs and, have not had to remove any ribs. A few patients have complained of numbness over the breast which, with time, has uniformly resolved. The overall correction of the deformity was excellent

Link: Discussion of Scoliosis from an orthopedic surgeons view.

rgb@scoliosisrx.com