Acute Hematogenous Osteomyelitis

Clarence L. Starr (Fig 1) was born in Georgetown, Ontario where his father practiced medicine. Although Canadian, Starr’s family had strong roots in the United States. His family had moved from Northern Ireland to Pennsylvania in 1717. His father had trained at Bellevue Hospital in New York City before opening his practice in Ontario, and his brother also trained at Bellevue Hospital and lived all of his life in upstate New York.Fig 1.: Dr. Clarence L. Starr (Reprinted with permission from Clarence Leslie Starr: The History of Canadian Surgery. Ottawa, Ontario, The Canadian Medical Association 109–111, 1960. © Canadian Medical Association.)Starr was a product of the University of Toronto. After his graduation in 1890, Starr interned in a private hospital in New Jersey and then went to Bellevue Hospital for surgical training. He completed his education with 1 year spent in Germany and Vienna. On returning to New York City, he briefly was an assistant to Dr. Virgil Gibney at the Hospital for the Ruptured and Crippled. In 1894, Starr opened his practice in Toronto and shortly after he was put in charge of the orthopaedic section of the Hospital for Sick Children. In 1899, he joined the staff of the Toronto General Hospital. In 1911, Starr was named chief of the Hospital for Sick Children and was forced to resign from the General Hospital. During World War I, Starr served as chief of surgery of the Granville Canadian Special Hospital at Ramsgate. After 18 months, he returned to Canada as a consultant in orthopaedic surgery to the Canadian Army and organized a chain of rehabilitation facilities across the country. In 1921, Starr returned to the Toronto General Hospital as the first full-time chief of surgery, a position he held until his untimely death from coronary artery disease at the age of 60 years. He was succeeded by his friend and colleague, W. E. Gallie. Starr’s forte was administration as was shown in his reorganization of the staffs of the Hospital for Sick Children and the General Hospital, and his work in the army. He was a past president of the American Orthopaedic Association and vice president of the American Surgical Association at the time of his death. Starr’s publications were few, but significant, especially those on the subject of hematogenous osteomyelitis. The following article is the best account of the pathogenesis, natural history, diagnosis, and treatment of this condition before the discovery of antibiotics. Although the treatment has changed, Starr’s description of the clinical picture still is relevant today. Leonard F. Peltier, MD, PhD In view of the fact that an early diagnosis is rarely made in cases of acute osteomyelitis, and that the recognized textbook treatment of this condition in the early stage is open to question, it seems desirable that the subject should be reviewed and the experience of a large hospital clinic placed before the surgical public. The condition which we call osteomyelitis is essentially an inflammation of all the structures of the bone, and really should be designated a periosteomyelitis. It seems incredible that an acute infection could be limited to the cancellous bone or to the periosteum, in view of the easy access and free communication between these by means of the haversian canals and para-epiphyseal line. Anatomic Considerations It is necessary in studying this problem to remember some of the anatomic features as they relate to bone growth and development. The circulation of an ordinary long bone, as described by Lexer, is derived from three sources: The nutrient artery enters the shaft at about its middle and divides, sending branches to either extremity, which, after numerous divisions, end in fine capillaries near the juxtaepiphyseal line or metaphysis. The epiphysis gets its blood supply from the cortical branches, which find their way in from the cortex along the epiphyseal line and are distributed to the whole of the epiphysis. The cortex of the entire shaft receives its nutrition from the periosteal vessels. The epiphysis in each long bone is, in most cases, not united by bone until early in adult life. The growth takes place on the diaphyseal side of the growing line, as can be demonstrated by an attempt to separate the epiphysis. By forcibly tearing off the epiphysis from the shaft in the growing bone, it is seen that the separation takes place at the diaphyseal side of the epiphyseal line, and leaves an irregular series of indentations. In the child, the periosteum continued down from the shaft dips into the epiphyseal line and becomes densely adherent at this point. This condition prevents, in most cases, direct extension of infection to the joint. Age Acute osteomyelitis is essentially a disease of childhood. It occurs at this period because the actively growing bone is less resistant to infection than adult tissue, and because trauma at the epiphysis is possible only at this age. It is extremely rare to find osteomyelitis in the adult, aside from that due to direct injury to bone and primary infection, such as is found in compound fractures and bullet wounds. The commonest age is from 2 to 10 years. Sex Boys are slightly more frequently affected than girls, possibly owing to increased opportunity of injury. Roentgen-Ray Findings There are several facts of importance in the study of the roentgenograms. Roentgen-ray findings are always negative in the early stages. It is not until gross destruction of cancellous bone has taken place, or until new bone begins to develop as an involucrum, that the roentgenograms reveal any definite picture. This is too late to be of any diagnostic value, and it should be emphasized here that roentgen-ray findings should not be waited for in making a diagnosis.Figure 14 (figure not shown), taken six days after the onset, shows only a slight change in density on the diaphyseal side of the epiphysis. The physician in charge of the patient had two roentgenograms made previous to this one, and no bony change was reported. On admission, on the same day that the accompanying roentgenogram was made, incision demonstrated pus in fairly large quantity under the periosteum, firm attachment of the periosteum at the epiphyseal line, and the periosteum stripped nearly to the upper end of the shaft.Figure 15 (figure not shown), taken six months later, shows the amount of the destruction of the shaft.Roentgenograms of cases of osteomyelitis which have been opened through the periosteum even moderately early, all show a fairly similar picture. The bone changes always extend farther up the shaft on the periosteal side, than in the medullary area, and for the most part the periosteum is stripped higher on one side than on the other, and consequently, the destroyed part is wedge-shaped, with the base at the epiphysis and the apex on one side of the shaft.Figure 16 (figure not shown), shows a much greater advance of the disease along the cortical area than in the medullary area. This roentgenogram was taken one month after onset, and the physician had evacuated pus from under the periosteum on the seventh day. The relief of tension under the periosteum had at once limited the extension of the disease both in the medullary area and under the periosteum.Roentgenograms of still later cases, when the tension of pus under the periosteum has not been relieved, show a characteristic “spotty” infection, as if the organisms had been poured through the cortex at various places. This is well shown in Figure 15 (figure not shown), where the periosteum was not incised until nearly entirely stripped. Figure 17 (figure not shown) was taken six months after the onset, during which time no treatment had been instituted, but the abscess had been allowed to burst. This also shows the spotty character of the infected bone, as well as areas of shaft and medulla which appear quite normal. Diagnosis The diagnosis must be made on the signs present. Severe pain in the neighborhood of a joint, with a spot of extreme tenderness, in a growing child is extremely significant. If it is accompanied by high fever and rapid pulse, with high blood count, the diagnosis is moderately certain. It is so essential that an early diagnosis be made, that one should be constantly on guard. The roentgen ray is of no assistance in early diagnosis. Valuable time is consumed, which means more and more bone destruction. Salicylates should not be administered if time is lost thereby. In rheumatism the onset is not usually so severe, nor the pain so excruciating. There is no sharp point of localized tenderness, and the pain is articular, not in the neighborhood of the joint. There is almost immediate swelling of the joint in rheumatism. An infectious arthritis is sometimes confusing; but the joint manifestations are usually sufficient to differentiate this condition. There is muscle spasm, limited movement of the joint, and early swelling of the synovial pouch. In osteomyelitis the joint is usually free of swelling, and movements are free if gently attempted. Do not wait for signs of swelling, redness and edema, as these are rather later signs. Of course, if these are present when the patient is first seen, they render the diagnosis still more certain. Treatment The diagnosis being made early, treatment should consist of incision over the area of greatest tenderness, through the skin, subcutaneous tissue and periosteum to the bone. Care should be exercised to keep the incision clearly on the diaphyseal side of the epiphysis to preserve the periosteal attachment to the epiphyseal line. This safeguards to a large extent the possibility of extension to the joint. If frank pus is encountered, our experience shows that this incision is sufficient, if a drain is kept in for a time. If no gross pus is encountered, the periosteum is stripped for a short distance on either side of the incision, to be sure that the incision is not in the wrong place. If still no frank pus is seen, a series of possibly three drill holes is made from the cortex, obliquely downward toward the epiphyseal line. These are about one fourth of an inch (6.4 mm.) apart up the cortex, and extend at least into the center of the shaft at the epiphyseal line, as shown in the diagram (figure not shown). Cultures from the debris removed from these drill holes always show infective organisms, even though free pus is not obtained. These establish clear lines of least resistance, and within twenty-four hours pus is draining freely. In no case have we opened the medullary canal, and in all cases the infected bone seems limited to the area present at the time of establishment of drainage. The treatment, usually advocated, of trephining an opening into the medullary canal, 2½ inches (6.4 cm.) or more up the shaft, seems, in the light of our experience, to be pernicious. If the periosteum is opened and holes drilled into the shaft at the metaphysis, one can hope to relieve the symptoms and prevent necrosis sufficient to produce sequestration. Even if free pus has stripped the periosteum, if it is confined to a small area, regeneration of the necrotic area may take place without sequestration. The subjoined case illustrates this type. A. B., a boy, aged 12 years, had sore throat, headache and cervical adenitis, one week before onset of symptoms in the arm. Two days before admission he had pain, which increased in severity just above the left elbow. Slight swelling was noted the day before admission. On admission he had pain just above the left elbow, some swelling and redness over the inner aspect of the humerus, with pitting on pressure and exquisite tenderness. Movement at the elbow was nearly complete in range, and only painful when the arm was extended to its extreme length. The temperature was 103 F.; the pulse 150; respirations 30; white blood cells 24,000; polymorphonuclear leukocytes 98 per cent. A free incision over the inner side of the lower end of the humerus, above the epiphyseal line, opened a cavity containing ½ ounce (15 c.c.) of pus. Culture from this pus showed a hemolytic streptococcus. A blood culture taken at the same time was sterile. The boy had a rapid convalescence, and was discharged from the hospital at the end of three weeks, with the wound healed. It has remained healed ever since, now one year. Figure 20 (figure not shown), a roentgenogram taken four months later, shows some periosteal thickening and no cancellous tissue change. In the series of patients under treatment there are six that have been similarly treated; and the wounds have healed in three or four weeks without a sequestrum being formed. There are also three patients with positive blood cultures on repeated examination, who recovered with greater or less sequestration of necrotic bone. In the later stages the plan of treatment adopted is this: After efficient drainage has been established, operation is delayed until the sequestrum is separated. Then a channel sufficiently large to remove the sequestrum is chiseled through the involucrum; the cavity is gently curetted until it is free of dirty granulations; the cavity is sponged with iodin, (sic) and packed tightly for forty-eight hours with iodoform gauze after which all packing and drains are removed. If the cavity left is too large to fill easily by granulation, the edges are made saucer-shaped, or flattened, to permit the soft tissues and periosteum to fall in and obliterate it. It is rare to find in hematogenous osteomyelitis the large central cavities that one finds in osteomyelitis near the extremity of a long bone following a gunshot wound, which can best be obliterated by a transplanted muscle flap. The practice of attempting to chisel away necrotic bone before it has delimited itself is to be condemned, as it is impossible to determine where the necrotic bone ends and where the living bone begins. Living bone is thus either taken away or necrotic bone left. In the operation of making a depressed or saucer-shaped cavity the periosteum should not be widely separated from the living bone, as this is likely to interfere with the circulation of the bone tissue immediately underneath, which may also have its endosteal circulation cut off. This would, of course, make a new necrotic area and subsequent sequestrum. The blind curetting of a bone cavity where there may or may not be a sequestrum, is not a procedure which is likely to produce any beneficial results. In our experience the sequestrotomy should be performed in a bloodless field, using a tourniquet wherever it can be safely used. Ill-considered and incomplete operations in the latter stage are largely responsible for the fact that in so many of these bone cases discharging sinuses continue for years, and the patients are subjected to operations without number.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.