Bone and Joint Infections in Neonates

Pathology and Etiology

Bone infection, or osteomyelitis, occurs by three mechanisms: bacteremia or hematogenous spread; direct inoculation from a puncture wound, such as a heel or femoral vein stick; and a contiguous spread from an adjacent focus of infection. In neonates, as in infants and children, most bone infections are hematogenous in origin. The most common site of osteomyelitis is the metaphysis, the region of the bone immediately adjacent to the physis or growth plate. The anatomic arrangement of metaphyseal vessels and the dynamics of blood flow in this region permit bacteria to lodge and proliferate. The nutrient artery ascends to the metaphysis from a central location within the bone. When the arterioles reach the physis, they make 180-degree turns and empty into the venous sinusoids. This process creates an area of sluggish blood flow and an opportunity for bacteria to become trapped and proliferate. A separate set of vessels nourishes the epiphysis.

In childhood, there is no connection between the epiphyseal and metaphyseal blood supply. The physis creates a barrier that is seldom penetrated by the spread of infection from the metaphysis to the epiphysis or from the epiphysis to the metaphysis. However, during the first 12-18 months of life, this barrier to the spread of infection does not exist, because there are vessels passing through the physis that connect the epiphyseal and metaphyseal circulations. These vessels allow infections to cross the physis, which may account for the increased incidence of physeal damage seen in neonatal osteomyelitis when compared with childhood osteomyelitis. Peters and associates observed the late onset of growth disturbance after neonatal osteomyelitis and recommended that the affected neonates be followed to skeletal maturity to monitor for growth disturbances.

When bacteria lodge in the metaphyseal vessels and begin to proliferate, inflammation followed by abscess formation occurs. The pressure from the purulent material causes extrusion of the pus through the haversian canals to the cortex and subsequently into the subperiosteal space. The continued subperiosteal accumulation of purulent material strips the periosteum from the bone. Because the periosteum supplies blood to the cortex, this stripping process interrupts cortical blood flow. As a result, large areas of cortical bone may become devascularized. This dead bone, or sequestrum, can serve as a site for chronic infection, which is isolated from limited neonatal defense mechanisms and antibiotics. The elevated periosteum produces new bone in an attempt to repair the injured bone. This process in turn produces the involucrum, which surrounds the sequestrum.

Draining cutaneous sinuses may arise when pus ruptures through the periosteum, adjacent soft tissues, and skin. Infection may occasionally spread into an adjacent joint space, causing secondary septic arthritis. The destruction of the epiphysis may occur through the direct spread of infection into it. This process can result in subsequent shortening, angular deformity, or both, of the involved extremity. In neonates, complete physeal closure can occur, resulting in shortening. Damage to the articular surface of the joint may result in the loss of motion and ultimately degenerative osteoarthritis.

The etiologic organisms responsible for neonatal osteomyelitis are variable. Staphylococcus aureus has traditionally been the most common causative organism. Methicillin-resistant S. aureus (MRSA) has been reported in neonates, which is especially destructive. Group B streptococcus ( Streptococcus agalactiae ) and Gram-negative organisms ( E. coli and Klebsiella pneumonia ) are all seen in the neonatal period. Candida albicans can also cause osteomyelitis in those neonates at high risk for infection.

Diagnosis

The clinical manifestations of osteomyelitis in children vary with age. In previously healthy neonates, it usually occurs during the first 2 weeks of life. Limitation of spontaneous movement with pseudoparalysis of the involved extremity is the most common sign. Localized tenderness, erythema, increased warmth, and swelling may occur. Associated septic arthritis with accompanying joint effusion and increased warmth occurs in many cases. Term neonates usually appear less ill than would be expected because of the persistence of maternal antibodies. They have less fever, leukocytosis, and elevation of the C-reactive protein and erythrocyte sedimentation rate than older children with similar infections. Less commonly, neonatal osteomyelitis presents as septicemia. The presentation and course of osteomyelitis are strongly correlated with the health of the infant before presentation. Infants with multiple sites of infection are usually ill before its onset and have an increased incidence of placement of umbilical catheters or other lines. They are also more ill than those neonates presenting with only one site of infection.

Some neonates are at increased risk for osteomyelitis; they also have a more severe course of the disease. Bergdahl and colleagues identified the following risk factors in a study of 40 neonates with osteomyelitis: a birth weight of less than 2500 g or gestational age of less than 37 weeks, emergency cesarean delivery, a congenital malformation requiring neonatal surgery, respiratory distress syndrome, hyperbilirubinemia, large vessel (usually umbilical) catheterization, perinatal asphyxia, scalp laceration after vacuum extraction, and renal vein thrombosis. Twenty-one neonates were found to have risk factors but 19 did not. Of the 21, most had multiple sites of infection; 13 of the 21 (62%) neonates with risk factors had serious skeletal sequelae. In the remaining 19 neonates, multiple sites of infection were uncommon, and serious skeletal sequelae occurred in less than 20%.

The early diagnosis of osteomyelitis is based on obtaining purulent material, blood, or both for cultures and antibiotic sensitivities. Early in the course of the disease, radiographs and bone scans may be normal. Bone aspiration is usually positive. The cultures of subperiosteal metaphyseal pus yield a pathogen in about 70% of cases. The point of maximal swelling, bone tenderness, and fluctuation on physical examination is the most appropriate location for needle aspiration. The skin overlying the affected region should be prepared with an antiseptic solution and draped with sterile towels. After infiltration of the area with local anesthetic, an 18-gauge spinal needle, with the stylet in place, is passed through the skin to the bone. The subperiosteal space is aspirated first. If the tap is dry, the needle with the stylet should be gently twisted through the bone cortex into the metaphysis, which is then aspirated. The aspirated fluid or blood should be immediately Gram-stained and cultured. The organism may also be recovered from other sources. Blood cultures are positive in 60% of children with osteomyelitis. When osteomyelitis complicates meningitis, the organism may be recovered from cultures of the cerebrospinal fluid.

Imaging

See Chapter 38 . A diagnosis cannot be easily made by bone aspiration during the inflammatory phase before abscess formation. At this point, imaging can be helpful. Plain radiographs of the suspected bone are often a valuable procedure. Within a few days of the onset of infection, deep edema, joint effusion, and sometimes bone destruction can be detected. The edema is visible sooner in neonates than in children, because the neonate has very porous bones and a loosely attached periosteum ( Fig. 98.1 ), which permits the earlier collection of subperiosteal abscesses in neonates than in children. This fact also accounts for the relatively favorable results of diagnostic aspiration in neonates.

If radiographs are unproductive, ultrasound can be useful in localizing the areas of deep edema (subperiosteal fluid collection or abscess) and joint effusion, as well as periosteal thickening or elevation. Magnetic resonance imaging is very sensitive for early detection of osteomyelitis and will show details of bone involvement, abscess formation, and septic arthritis. Limitations include possible need for sedation and patient stability to transfer to the MRI unit. Technetium bone scanning is currently used less often; however, it may be helpful in neonates with suspected multifocal infection. There is some controversy over the merits of bone scanning in neonates because of their decreased inflammatory response and the large amount of isotope uptake from the very active adjacent epiphysis. In 1980, Ash and Gilday concluded that it was very unreliable. However, Bressler and co-workers and others have suggested that technetium bone scans can be helpful. Bressler's group scanned 33 infants suspected of having osteomyelitis. Each of the 25 sites of proven osteomyelitis in 15 infants was demonstrated on bone scanning. Another 10 sites that were radiographically negative were also demonstrated to be positive on bone scanning. The bone scan, if negative, does not exclude osteomyelitis.

The results of indium and gallium scans for acute osteomyelitis in neonates have not been reported. A concern about gallium scans is related to the pathology of osteomyelitis. If there is extensive septic embolization of metaphyseal vessels, labeled blood cannot penetrate the metaphysis. If the periosteum is stripped off the cortex, devascularizing the cortical bone, there can be no blood flow in the cortex. Under these circumstances, the uptake by bone may be difficult to predict. However, Demopulos and associates recommended an indium scan if the technetium bone scan in the neonate is not highly suggestive of osteomyelitis.

Another concern in the assessment of osteomyelitis is the determination of whether metaphyseal aspiration can produce a positive bone scan. Canale and colleagues assessed the effects of bone, and joint aspiration on the bone scan results in healthy dogs. They used multiple aspiration techniques on the bones and joints of 15 dogs and scanned them between 5 hours and 10 days of aspiration. In no case did joint aspiration lead to a positive bone scan. Metaphyseal drilling and periosteal scraping with needles occasionally led to positive bone scans after a 2-day delay. Thus, bone aspiration does not initially affect a technetium bone scan.