Etiopathogenesis

Gram-positive cocci (Staphylococcus aureus and Streptococcus pyogenes) are by far the most common organisms to cause vertebral infection, although other infective organisms, including Escherichia coli, Pseudomonas, Klebsiella, and Proteus, have also been isolated. ³ In children with sickle cell anemia, Salmonella infection is common. Most infections occur due to bacterial spread from a distant site (dermal, respiratory tract, and genitourinary tract) to the spinal column through the bloodstream (hematogenous spread). Several risk factors such as immune deficiency states, long-term systemic administration of steroids, juvenile diabetes mellitus, organ transplantation, malnutrition, chemotherapy for malignancy, infective endocarditis, renal failure, and sickle cell disease have been identified for the development of vertebral infections.

Although the arterial route is the common route of bacterial spread to a vertebra, retrograde seeding of venous blood via Batson’s venous plexus and rarely contiguous spread of infection from a nearby infected focus to the vertebra and disk can also produce infective spondylitis. The intervertebral disks in children are vascular until 8 years of age. Hence, unlike adult vertebral infections, primary disk infection is more common in children. In older children, the infection also spreads easily from the subchondral region of the vertebral body due to the abundant blood supply of the trabecular cancellous bone and its rich, cellular marrow. As the blood flow stagnates in the metaphyseal arterial loops just beneath the vertebral end plates, the circulating bacteria readily colonize there, subsequently invading the disk. In pyogenic spondylitis, the involvement is usually focal, but multiple site involvement can occur in immune-compromised patients. As vertebral and diskal destruction proceeds, the vertebral canal can be involved by pus and granulation tissue, which can cause cord compression resulting in a rapid-onset neurologic deficit. Other sequelae include meningitis, sepsis, and rarely death. Severe vertebral deformity and excess abscess formation are uncommon in pyogenic spondylitis.

Clinical Symptoms and Signs

Toddlers and infants with pyogenic spondylitis have a varied clinical presentation, including difficulty in walking, abdominal pain, hip and thigh pain, and refusal to eat. General features of infection, such as fever, malaise, and fatigability, are also present. A high degree of clinical suspicion and a meticulous clinical examination to note the presence of spinal guarding and local spinal tenderness can help in making the diagnosis. In older children, the usual presentation is one of insidious onset back pain and fever. The pain is initially localized to the level of the infection, but vague distribution to the paraspinal areas is also common. Because young children may not directly complain of back pain, a “coin test or quarter test” has been described in which the child is unable to pick up a coin from the floor due to painful restriction of spinal movements.

Neurologic involvement can occur early when compared with tubercular infection, even in the presence of minimal vertebral body collapse and thin epidural abscess. If neurologic involvement is suspected, a meticulous neurologic examination, including per-rectal examination to detect early cauda equina compression, is essential. Ideally the neurologic examination should be repeated and documented at regular intervals as the child will be unable to complain about sudden neurologic worsening.

Investigations

Blood tests to evaluate the presence of infection including total cell count, differential cell count, erythrocyte sedimentation rate (ESR), and C-reactive protein (CRP) are performed. These investigations provide only a glimpse of the diagnosis and are neither specific nor sensitive for pyogenic spondylitis. Leukocytosis (> 15,000 cells/mm ³ ) is present in < 50% in patients with pyogenic spondylitis. Elevated ESR and CRP levels are the most common laboratory abnormality and are excellent indicators of acute inflammation but remain nonspecific in the diagnosis of infective spondylitis. Both are more helpful in assessing the response to treatment. Decreasing values would indicate a good response to antibiotic treatment.

Urine and blood cultures are performed in all patients, but are positive in only 40 to 50% of patients with spinal infections. Blood cultures should always be obtained during a febrile episode and prior to administration of antibiotics. Positive results are of immense value in choosing the appropriate antibiotic treatment. If all cultures turn out to be negative, then biopsy is required to isolate tissue specimen for culture.

Radiological Studies

Plain radiographs are notoriously normal in the first 2 to 3 weeks after infection. Loss of delineation of subchondral bone, vertebral lucencies, and destruction of the end plates with narrowing of the disk space are the earliest changes but are usually seen by the end of 3 weeks. In advanced disease, destruction of ver tebral bodies, collapse with kyphosis and features of spinal instability may be evident. Magnetic resonance imaging (MRI) is the investigation of choice due to its ability to depict changes even in the early stages of the disease ( Fig. 11.1 ). Hence, MRI should be performed with a low threshold to diagnose infection in the early stages. MRI has a high sensitivity of 96%, a specificity of 92%, and an accuracy of 94% in patients with disk space infections. ⁴ T1-weighted images show decreased signal intensity changes in the vertebral bodies and disk spaces. T2-weighted images show increased signal intensity in the vertebral disk and body. Short tau inversion recovery (STIR) sequences and contrast images are also highly useful in diagnosing infection in the early stages. It also clearly documents the location and size of epidural abscess, the presence of sequestrum within the canal, the extent of compromise of the spinal canal, the degree of compression of the spinal cord, and any signal intensity changes in the cord.

Computed tomography (CT) is performed in select situations and is considered to be more sensitive in assessing the degree of bone destruction and in examining the surrounding soft tissues. It is also used as a guide for accurate placement of the Jamshidi needle while performing a percutaneous biopsy. A radionuclear bone scan with technetium-99m is a sensitive test (> 90%) for the early diagnosis of pyogenic vertebral osteomyelitis. Although expensive, a radioactive gallium scan is more specific, with 80 to 85% specificity rates. Gallium localizes inflammatory lesions well and, when combined with technetium, demonstrates virtually all pyogenic vertebral infections. But the lesions are not well localized by radionuclide scans, and the scans do not demonstrate the true extent of infection. Thus, MRI has superseded radionuclide scans and remains the gold standard imaging test.

Histopathology

Computed tomography guided or fluoroscopy enabled percutaneous biopsy of the infected vertebra or disk is advised where a tissue biopsy is required. Although cultures are positive in only 50 to 60% of patients with an infection, histological findings are invariably confirmative. Trocar biopsies are better than fine-needle aspiration because a larger amount of material from the infected area can be obtained. If blood cultures and percutaneous biopsy techniques fail to identify the infecting organism, open surgical debridement and biopsy should be performed. An open surgical biopsy has the highest success rate for positive culture findings (up to 90%) and helps in diagnostic confirmation. ⁵