Portal de Periódicos da CAPES

THE CLASSIC: Anterior Spinal Fusion

2006; Lippincott Williams & Wilkins; Volume: 444; Linguagem: Inglês

10.1097/01.blo.0000203456.67016.b7

1528-1132

A. R. Hodgson, Francis E Stock,

Orthopedic Infections and Treatments

Professor Arthur Hodgson ("Hoddy") turned 75 years old in June 1990 (Figure not shown). This was the year of the 35th anniversary of the first anterior surgical approach for spinal tuberculosis. In 1955, three patients were surgically treated with this procedure and received follow-up treatment for several months. At this stage, the patients had solid healing of the bone grafts and no sign of paraparesis. It was then decided, on the basis of these initial surgical observations, to proceed. Since then, Hodgson's contribution has been legendary. A paper that reported the surgical treatment of the first 50 patients was published in the British Journal of Surgery in 1956. During the next few years, large series of patients were reported with dramatic recovery of Pott's paraplegia and a very high rate of anterior bone graft fusion. Doctor Hodgson, his wife Monica, and their five children arrived in Hong Kong in 1951. He had left Rochester, England, where he was a senior registrar, and sailed to Hong Kong to take up the post as lecturer in orthopaedic surgery within the Department of General Surgery at Hong Kong University. During these early years in Hong Kong, Hodgson also had to deal with endemic poliomyelitis, osteomyelitis, and all the Third World orthopaedic problems. His first assistant in these early years was Dr. Harry Fang. By 1960, his work was well recognized, and he was made the first professor of the new Department of Orthopaedic Surgery, separate from the Department of General Surgery in which he began this monumental work. I was fortunate to spend eight years as lecturer in the department, leaving at the end of 1975. During this period, there was a renaissance in spinal surgery that was solely attributable to the hard work and clear thinking of Hodgson. Not only did he think and write clearly (his papers are few in number, but classical in quality), but he provided a great stimulus and influence on many leaders in spinal surgery during the past quarter of a century, including Doctors John Hall, Kenton Leatherman, and the late Alan Dwyer. It is no secret that without Hodgson's continuing support for Dwyer's screw and cable technique, it probably would have failed. All told, he spent 24 years in orthopaedic surgery at Hong Kong University, 15 of these as professor and head of the department. This volume of Clinical Orthopaedics and Related Research is a tribute to Professor Hodgson. I personally thank those who made the effort to put together their thoughts. The spinal papers in this Symposium represent much of the latest contemporary thought in spinal surgery, and would have been strongly approved of and encouraged by Professor Hodgson himself. As he looks back during his years of contribution to spinal surgery, Hodgson can be more than gratified by the knowledge that this concepts and ideas have been disseminated throughout the world. His contribution and the development of the anterior approach to spinal tuberculosis remain a milestone; from this flowed the anterior approach for so many other diseases of the vertebral column as we know it today. Professor Hodgson died on November 16, 1993 at the age of 78. His leadership in spinal surgery will be sorely missed throughout the world. John P. O'Brien, PhD, FRCS(Ed), FACS, FRACS The object of this paper is to report the early results of a method of treatment of tuberculous disease of the spine which aims at the complete extirpation of the diseased focus and its replacement by bone grafts in such a position that they are structurally sound. The authors do not claim that the method is entirely new. It is a development of ideas conceived by others in the past and made possible by advances in chemotheraphy and anaesthesia which have made an open approach through the chest feasible, and by the adoption of architectural principles in arthrodesis and especially by the recognition of the initial and increasingly developing strength of a bony strut in compression. ANTERIOR SPINAL FUSION A perusal of the literature dealing with the treatment of Pott's Disease since the introduction of spinal fusion by Hibbs and Albee forty-five years ago shows that opinion on the value of posterior fusion is divided. Some advocate conservative treatment alone, some conservative treatment with late spinal fusion and others early spinal fusion followed by conservative treatment. There is no agreement on whether spinal fusion should be reserved for the adult or the child or performed on both. This lack of unanimity suggests that there is no great advantage in posterior spinal fusion over conservative treatment. Henderson (1917) drew attention to the fact that posterior fusion does nothing to the diseased focus, to quote him, "neither the Hibbs nor the Albee type of operation for ankylosing the diseased area in tuberculosis of the spine is in any sense a radical operation for the extirpation of the focus of the disease"-an observation as true to-day as then. Girdlestone (1923) gave a lucid description of the spinal joints and emphasised that the Hibbs and Albee fusions are posterior spinal fusions and that they can only prevent flexion if the lateral columns of the spine are intact. King (1933) and McKee (1937) commented upon the fact that early spinal fusion can prevent collapse of the vertebral bodies anteriorly and so delay healing. Swett (1940) in a pertinent appraisal of the pros and cons of spinal fusion observed "the dominant role of the abscess in the healing process." Wilkinson (1955) advocates evacuation of the abscess and curettage of the lesion through a costotransversectomy approach. There is no doubt that throughout the world orthopaedic surgeons are starting to be aware of the important part played by the abscess in delaying healing. The spine was approached anteriorly by Muller (1906) before the introduction of the Hibbs and Albee operation. It would appear that the introduction of these two operations drew attention away from the lesion itself. Ito, Tsuchiya and Asami (1934) published a series of ten cases of spinal tuberculosis in which they approached the lumbar spine retroperitoneally in eight cases, introducing an anterior graft in two cases and a second stage Albee graft in five cases. One case of thoracic disease was approached through a costotransversectomy and bone chips packed into the curetted cavity. This report was labelled a preliminary report and the follow up of the earliest case was one year. We have been unable to find further reference in the literature to these cases. It is of interest to note that even in those pre-antibiotic days only two of the wounds broke down and we are told that they soon healed. Gjessing (1951) fused two cases of tuberculosis of the lumbar spine anteriorly. We believe from a review of cases of recurrence of disease in our own series of posterior fusions that Swett and others and Wilkinson are right when they emphasise that healing in this disease does not take place in the presence of an abscess. This is especially true in the thoracic region where the abscess is enclosed and subject to the motions of respiration and to the beat of the heart and pulsations of the aorta. We agree with Seddon (1935) when he states "There is a strong case for every effort being made to reduce the bone destruction during the active stage of caries of the thoracic spine." The literature on the pathology of tuberculosis of the spine is sparse when one considers the frequency of the disease. Willis (1926) drew attention to the existence of extensive lesions of the spine which only superficially erode the bone. Cleveland and Bosworth (1942) have drawn attention to the "constant recurring states of sclerosis and caseation" and have shown the sclerosis to be due to aseptic necrosis. Wherever the primary site of the lesion is, it travels by sub-ligamentous spread which strips the ligaments from the vertebral body. When the ligaments are detached the blood supply is lost and aseptic necrosis ensues. Fraser (1914) states that healthy bone has great resistance to tuberculous infection and this has been seen in our cases where healthy bleeding bone is seen close to a tuberculous focus. The bone with yellow marrow which Fraser postulated was devitalised by an exotoxin produced by the tubercle bacillus is really bone which has undergone aseptic necrosis. It would appear that the sub-ligamentous spread with resultant aseptic necrosis is the important reason for the spread in the thoracic region, and in this fact may lie part of the reason for the better prognosis in the lumbar region where the abscess gravitates down the psoas sheath. In children the periosteum and ligaments strip easily and consequently lumbar as well as thoracic lesions are liable to be extensive. The positive stripping characteristic of this abscess had led us to name it a dissecting paravertebral abscess. These observations have been made during the anterior approach to our cases and are best seen in the thoracic region where the abscess finds difficulty in escaping and where it is subject to the action of the heart, aorta and lungs. While it would appear unlikely that systemic antibiotics could penetrate an abscess (Mackenzie 1954), we have had many abscesses which have been proved sterile to the usual culture methods and to guinea pig inoculation. The very bulk of the abscess, however, forms a mechanical barrier to anything but slow healing. THE AIMS OF THE METHOD It appears to us that decompression of the abscess with removal of the pus, caseous material, granulation tissue, intervertebral discs, and sequestra is imperative to arrest further vertebral destruction where the disease is progressing, in much the same way as early decompression of an acute osteomyelitis will prevent chronic osteomyelitis with sequestration. Unless the spine is approached anteriorly the extent of the disease must remain unknown by present clinical methods for we have found almost invariably that the gross lesion is more extensive than the x-rays suggest. It is possible to leave a clean raw bleeding cavity when this stage has been completed and it remains to make the spine stable. We have felt that the obvious place to fuse the spine would be anteriorly where there is the largest portion of bone and where the graft would be in compression and with adequate protection (Brittain 1942). It should hypertrophy according to Wolffs' Law and fulfill Percival Pott's criterion of healing, that fusion of the bodies of the vertebrae "was the thing aimed at." This is Nature's way of healing the disease but it takes a long time and only occurs in about one in four patients (cleveland 1940). It is noteworthy that tuberculosis of the spine is the only tuberculous disease which does fuse spontaneously. We were dissatisfied with the exposure afforded by costotransversectomy as we found the operation difficult and the field restricted. Through this approach it was impossible to determine the extent of the lesion or to evacuate it completely, neither could the graft be inserted with any accuracy. When we commenced this study we were unaware of Ito and others article or his earlier references. OPERATIVE TECHNIQUE APPROACH The approach to the spine anteriorly presents few difficulties and differs little from the approach to other structures in the vicinity. The cervical spine is adequately reached by an incision along the anterior border of sternomastoid reflecting that muscle backwards, and oesophagus, trachea, larynx and carotid sheath forwards. If an abscess were present and pointing posteriorly the incision could be made along the posterior border of sternomastoid but this has not yet occurred in our series. The cervico-thoracic region is one of the most inaccessible levels, but may be approached by a periscapular incision similar to that used for a first stage thoracoplasty. The scapula is lifted with a mechanical retractor such as Edwards and the third rib is resected. The pleura is opened but, if necessary, or if the pleura were very adherent, an extra pleural approach could be made easily at this level as in Semb's apicolysis. Access is limited but the region from C.6 to D.4 is reached without undue difficulty. The superior intercostal artery should be divided at its origin together with the accompanying vein. The approach to these two levels may be made on either side with equal ease and the choice is determined by the site and size of the abscess, being made on the side of the larger abscess. Below the level of D.4, however, in both thoracic and lumbar regions, we have found it much more satisfactory to approach the spine from the left side where the pulsating aorta provides an easily identifiable landmark in what may be a mass of fibrous tissue and abscess. An exception is made to this rule, if radiologically the abscess appears to be penetrating the right lung, or if on account of its chronicity the abscess may be dissociated from a spinal lesion which is partly healed and replaced by much hard fibrous tissue. An abscess of almost any size can be evacuated from either side provided that the diseased spine is adequately resected. It is only in the exceptional patient that we have approached the spine on the right side where the vena cava may be very difficult to identify and particularly exposed to damage. In the thoracic region from D.4-D.12, the thoracotomy should be performed by rib resection so that the rib is available for the graft. It is much easier to work downwards than upwards on the spine and the rib to be resected should always be above the level of the spinal lesion. The correct rib may be selected from the A.P. film of the chest by counting the ribs at the extreme lateral margin of the film and taking the one which at that point is level with the upper border of the lesion. Thus, a lesion of D.8-10 will probably be most easily approached through 7th or 6th rib. The dorso-lumbar spine is another difficult region and neither the thoracic nor lumbar incision are adequate. We have, however, made use of Fey's 11th rib incision which was devised as an approach to the upper pole of the kidney and the suprarenal gland. Through this incision good access is obtained from T.11 to L.5 but if the disease extends higher the 10th rib may be resected and the diaphragm separated posteriorly so that the peritoneum may be displaced forwards. If the disease does not extend above L.1 the 12th rib approach to the kidney described by Digby (1941) and others gives good access while, if the upper limit of disease is L.3, the incision need not extend backwards beyond the tip of the 12th rib. Through this incision a large presacral abscess may be evacuated by suction and swabbing without difficulty. In order to approach the spine with ease and safety, it is desirable to enter the abscess cavity as early as possible. In the thoracic region, it is covered by pleura and intercostal vessels, and we therefore mobilize the aorta by dividing the pleura lateral to it, and ligating and dividing the intercostal arteries close to their origin together with tributaries of the hemiazygos vein. A vertical incision is then made into the wall of the abscess close to the midline and medial to the distally placed ligatures. At the upper and lower ends of this incision which extends to the visible edge of the abscess, transverse extensions are made so that a flap consisting of abscess wall, intercostal vessels and pleura may be lifted and hinged laterally. This gives better access than either a single transverse or cruciate incision which we used in early cases. These incisions are made and the flap is lifted by diathermy needle dissection because the tissue is very vascular and the minute vessels are more readily coagulated than secured by forceps. This flap is held back by a suture. In the lumbar region, the lateral aspect of the spine is covered by psoas muscle and this must be either reflected or divided before the spine is reached. In the region of T.12-L.1 the upper margins of psoas may be reflected downwards, but if the disease does not extend above L.3-4 a simple transverse incision through that muscle may be made with diathermy needle bearing in mind the trunks of the lumbar plexus which lie on its deep surface. EVACUATION OF THE DISEASED FOCUS As soon as the abscess is opened, pus is evacuated and some material saved for culture, inoculation and histological examination. All diseased tissues, pus, sequestra, aseptic necrotic bone and devitalized discs are removed. The pus can best be removed by a sucker while the rest are removed with a combination of osteotomes and a variety of nibblers. It is important to make sure that the full extent of the disease has been removed as it is easy to leave some disease behind. It is also important to remove the whole of the diseased vertebrae as far back as the spinal cord as there is usually a posterior intraspinal abscess (Swett 1940). When this toilet has been completed, a clean bleeding cavity should remain. INSERTION OF THE GRAFT It is possible to correct the kyphosis partially by direct pressure on the spine posteriorly. A mortice is made in the vertebrae at either end, the distance measured with dividers and a strut graft of the correct length introduced to keep the vertebrae sprung apart. The interval which remains between the vertebrae is filled with small match grafts. We usually fashion the strut graft from rib which has been removed during the approach, while the match grafts are made from frozen banked bone. In the cervical and lumbar regions, however, both strut and match grafts have been made from banked bone. The flap of pleura and fibrous mass is sewn back loosely. POST-OPERATIVE TREATMENT This presents no serious problem. The patient is placed in a plaster bed previously prepared with an anterior shell and is nursed for a considerable time prone. Chemotherapy commenced pre-operatively is continued in the post-operative period. When the chest has been opened, we have not usually found it necessary to use an under water-seal drain but have relied on intermittent aspiration of fluid or air. Frequently no aspiration has been required. The patient is allowed up as soon as there is clinical evidence of fusion. RESULTS Fourteen of the first seventeen patients in these series have fused rapidly and soundly despite the fact that the time interval is relatively short. In the remaining thirty-three cases fusion is incomplete although many are showing encouraging evidence of fusion. Fusion has been assessed both radiologically and by clinical methods. The spine becomes clinically stable about eight to twelve weeks after the operation as judged by the "spring" and "stethoscope" tests. The spring test consists of "springing" the area of disease with the patient lying prone when there is a definite sense of give when the spine is unstable. The stethoscope test depends on the transmission of a true note across the fused area where the stethoscope is placed at one end of the fused area and the other end is percussed with the fingers. COMPLICATIONS ACUTE DILATATION OF THE STOMACH This has occurred in two patients with disease of the lumbar spine and has responded to treatment by gastric suction. HAEMOTHORAX AND PNEUMOTHORAX This has occurred in several of the thoracic cases but has never been persistent. In all cases, it has responded to aspiration, and the chest film has been completely clear within two weeks. CONVULSIONS Two patients with cervico-dorsal lesions have developed severe convulsions approximately twelve hours after the operation lasting for twelve to twenty-nine hours. Both patients recovered. There seems to be little doubt that these resulted from sensitivity to Streptomycin which had been placed within the evacuated abscess cavity and thus in direct contact with the theca. We have now discontinued the practice of dusting the cavity with Streptomycin and merely use the drug systemically. DEATH There has been one death in the series. This child of nine years, with extensive disease from D.12 to L.5 inclusive, had been under conservative treatment for six years and had been treated by posterior fusion without success. He was extremely ill but stood up to an extensive resection of six vertebrae very well. He became pyrexial on the third day and deteriorated rapidly and died. The cause of death was bronchopneumonia, probably aggravated by over administration of intravenous saline. DISCUSSION While it is far too soon to pretend to estimate long term results, we feel satisfied that the procedure is a great advance in treatment by cutting the duration of treatment from years to months and in preventing gross vertebral destruction with the resulting deformity. The anterior approach has unveiled the pathology of the disease in the living which has not been universally appreciated, and this has led to a better understanding of the disease process. Some of these features are: THE ABSCESS The character of the pus is an indication of the activity of the lesion thin fluid greenish pus denoting activity and thick white pus chronicity. These observations have corresponded with the results of the Tuberculin test. It is impossible to aspirate the pus completely. Harris and Coulthard (1940) found abscesses in 90% of their patients. We would go further and say that every case has an abscess unless it is very early disease or it has healed. We do not feel that calcification in the pus denotes healing; sometimes this calcification is really sequestra and only indicates that the case is of some duration. In the lumbar region, the abscess may track down the front of the sacrum into the pelvis. DISKS While we were aware that disks were resistant to tuberculous infection we were surprised to find that whole disks could remain at the site of the lesion as sloughs. There is no doubt that they form a bar to bony ankylosis and in some cases can produce pressure on the spinal cord. X-RAYS The long narrow shadow which has been considered to be a paravertebral abscess may be due to thickening of the tissues covering the spine, periosteum, ligaments and pleura and may reach the thickness of one inch. A more bulbous paravertebral shadow indicates pus. Debris and calcification and ischaemic necrosis can give a dense shadow which gives a false impression of early healing. Sometimes, sequestra are difficult to pick up radiologically but on re-examination of the pictures after the operation, they can often be seen. Vertebral bodies whose outline changes from concave to convex and become slightly enlarged are affected by the disease and may be simply a bony shell containing pus. PENETRATION OF THE LUNG BY THE ABSCESS While we were aware that cases had been described where the aorta had been perforated by an abscess (Simpson and Grabbilaar 1955) and that the abscess could rupture into the pleura, it had not occurred to us that the lung was likely to be involved. In six of our cases the lung has been involved and has demanded treatment of the lung as a separate step in the operation. In these cases it has been possible to curette out the lesion in the lung which has been surrounded by a thin zone of atelectasis. The cavity has been dusted with Streptomycin and closed by suture. Resection has not been necessary. This small group of patients is being made the subject of a separate communication. CORRECTION OF THE KYPHOSIS It has been found possible to correct the kyphosis to a greater or lesser extent, greater in active disease and less in chronic disease with fibrosis. In a well healed lesion the fibrosis is extremely tough and has to be out with an osteotome at times and although there are no radiological signs of ossification, it appears to be preosseous. PARAPLEGIA "It is high time that fresh light should be thrown on the treatment of Pott's Paraplegia"-(Girdlestone 1949). In the early cases we noticed how easy it would be to decompress the spinal cord from the front. We found no difficulty in this when we operated upon our first patient with complete paraplegia (Case No. 13). We found that the cord was pressed on by retropulsion of the vertebral body into the spinal canal. The decompression was performed with ease, all the retropulsed vertebrae was removed and we could see that it had been evacuated completely. Fixation of the spine was done at the same time by the introduction of an anterior graft, care being taken to protect the cord from the graft. The second case was a much more advanced case with involvement of D.8-11 (Case No. 18) and all these vertebral bodies had to be removed. The cord was eventually decompressed from D.8-11. The whole cord had been stretched round these vertebrae but the main pressure was at the apex of the kyphosis where there was a definite groove in the cord. In cases of recent paraplegia the recovery of the paralysis is dramatic recovery taking only a few days. The longer the duration of the paralysis, the longer the recovery takes and the less complete it is. CONCLUSION The anterior approach to the spinal column is a practicable proposition at all levels. It is the only approach which allows accurate visual assessment of the extent of the disease. It offers the prospect of early radical treatment of Pott's Disease and the prevention of paraplegia by the evacuation of the abscess from within the spinal canal. It allows complete evacuation of the abscess and decompression of the spinal cord in established paraplegia. It allows a bone graft to be placed in the position where it is most likely to succeed. Because of the complete eradication of the disease it offers the hope of shortening the period of hospitalisation. It opens the study of the condition to detailed investigation in the future. SUMMARY The literature on the pathology and treatment of Potts' Disease of the spine is reviewed. A radical method of treatment consisting of complete extirpation of the disease and fixation by bone graft in compression is described in detail. The early results are tabulated and complications enumerated.