49.

Chapter 42 Osteomyelitis

Manual of Rheumatology and Outpatient Orthopedic Disorders

---

Chapter 42 Osteomyelitis

---

Barry D. Brause

Etiology , epidemiology, and pathogenesis
Clinical features and presentation
Diagnosis
Treatment

Although the distinction made between the acute and chronic forms of osteomyelitis has some value in therapy and prognosis , no abrupt shift from one category to the other occurs in most cases. A more useful classification, based on the pathogenetic route of infection, divides cases into three types: (a) hematogenous osteomyelitis; (b) introduced infection, which results from contamination accompanying surgical and nonsurgical trauma; and (c) contiguous infection, which results from the spread of microorganisms from adjacent infected tissue and includes osteomyelitis associated with peripheral vascular disease.

I. Etiology, epidemiology, and pathogenesis. Osteomyelitis represents the invasion of microorganisms into bone, and virtually all microbes can infect bone. Bacteria are the usual pathogens, and staphylococci are the most common etiologic agents in all three types of osseous infection. Staphylococcus aureus causes approximately 60% of cases of hematogenous and introduced osteomyelitis and is the most prominent pathogen when osseous infection develops from sepsis in contiguous tissue. Staphylococcus epidermidis and the other coagulase-negative staphylococci have become the major pathogens in bone infections associated with indwelling prosthetic materials and foreign bodies, such as joint replacement implants and fracture fixation devices, which are responsible for 30% of these cases. Streptococci, gram-negative bacilli, anaerobes, mycobacteria, and fungi are causative agents in a variety of clinical situations (Table 42-1).

Table 42-1. Predispositions, anatomic sites, and prominent pathogens in forms of osteomyelitis

1. Hematogenous osteomyelitis. The anatomic osseous site of involvement in hematogenous infection is age-dependent. From birth through puberty, the metaphyseal regions in the long bones of the extremities (tibia, femur, humerus) are most frequently involved owing to their large blood flow during these developmental years . In adults, blood- borne pathogens preferentially infect the spine (lumbosacral, thoracic) because vertebrae receive relatively more blood flow with maturation . In bacteremia, the more vascular anterior end plates are seeded , and osteomyelitis commonly involves two adjacent vertebral bodies and the intervertebral disk space. The septic process compromises the nutrient supply to the intervertebral disk, resulting in disk necrosis and disk space narrowing, which is often the earliest radiographic sign of vertebral osteomyelitis. The etiologic pathogens in hematogenous osteomyelitis reflect the microorganisms associated with bacteremia in specific patient populations (see Table 42-1).
2. Introduced osteomyelitis. Patients are at risk for the introduced form of osteomyelitis whenever the skin and soft tissues overlying and protecting bone are breached by trauma or surgery. Approximately 70% of compound fractures are contaminated by skin and soil microflora, but thanks to effective debridement and perioperative antibiotic therapy, infection develops in only 2% to 9%. Prophylactic antibiotics and extensive antiseptic operative techniques allow large foreign bodies to be inserted into bones during reparative and reconstructive orthopedic surgery with infection rates below 2%. Indwelling foreign bodies decrease the magnitude of a bacterial inoculum necessary to establish infection in bone, and they permit pathogens to persist on the surface of avascular material, often within host- or microbe-derived biofilms, sequestered from circulating immune factors and systemic antibiotics.
3. Osteomyelitis by contiguous spread (including vascular insufficiency). Osteomyelitis develops by contiguous spread in one-third to two- thirds of diabetic patients with long-standing foot ulcers, and more hospital days are utilized to treat foot infections than any other complication of diabetes mellitus. Osseous involvement reflects unsuccessful reversal of or compensation for underlying severe neuropathy and vascular insufficiency, which prevent a skin ulcer from healing, so that progressively deeper microbial invasion culminates in spread to contiguous bone. This clinical scenario is also seen in patients with chronic skin ulcerations resulting from other conditions associated with severe sensory neuropathy (e.g., meningomyelocele) or vascular insufficiency (e.g., decubitus ulcers, vasculitis, atherosclerosis, and arteriosclerosis). The most common pathogens are staphylococci, streptococci, gram-negative bacilli, and anaerobes. Multiple organisms are isolated in more than 60% of cases.

   In all three forms of osteomyelitis, the microorganisms induce local metabolic changes and inflammatory reactions that produce osseous edema. As infection spreads within the bone, local thrombophlebitis develops, increasing edema and intraosseous pressure, which can result in ischemic necrosis of large areas of bone, called sequestra. If the osseous cortex is breached, subperiosteal abscesses can develop, with periosteal inflammation and periosteal formation of new bone in adjacent soft tissue, called an involucrum.

4. Chronic osteomyelitis. Unsuccessful therapy of acute osteomyelitis results in relapsing or chronic infection. Microorganisms persist in foci of gross or microscopic necrotic bone (sequestra) and intermittently invade surrounding tissues, causing the acute exacerbations of chronic osteomyelitis. Because of the avascular nature of sequestra, persistent pathogens are not eradicated by systemic antibiotics; however, local immune responses may control or eliminate the septic process. Recrudescent infection usually involves only local tissue but may spread through the overlying soft tissue to produce a drainage tract or sinus that eventually reaches the skin surface, creating a cutaneous draining sinus orifice. The sinus commonly drains sporadically from an osseous origin (often at a site of necrotic bone, or sequestrum) and reflects the degree of active infection and inflammation in the bone. Rare complications of chronic osteomyelitis include squamous cell carcinoma in overlying draining sinus tracts and secondary amyloidosis.

II. Clinical features and presentation

1. Acute hematogenous osteomyelitis in childhood. The majority of children with acute hematogenous osteomyelitis show evidence of systemic illness fever , chills, malaise, leukocytosis, elevated erythrocyte sedimentation ratealong with local symptoms. Focal bone pain is a characteristic feature, with overlying erythema, warmth, and swelling variably seen. Limb motion may be limited by local pain (pseudoparalysis) when the infection is near an articulation. Adjacent joint effusions can occur but are usually sterile when the epiphyseal cartilage is intact.
2. Adult hematogenous osteomyelitis (including vertebral osteo-myelitis). Vertebral osteomyelitis most often presents subacutely with prominent back pain and spine tenderness following urinary tract instrumentation or infection (30%), skin infection (13%), or respiratory infection (11%). In one-third of cases, the source of the bacteremia is unidentifiable. Fever is present in fewer than 50% of patients and is commonly low-grade. The infection can extend beyond the vertebral column and produce suppuration at the particular spinal level of involvement, such as a retropharyngeal abscess, mediastinitis, empyema, or subdiaphragmatic and iliopsoas abscesses, in addition to meningitis. If paresis, sensory deficits, or bowel or bladder dysfunction develop, spinal epidural abscess, the most feared complication of vertebral osteomyelitis, should be considered and evaluated immediately. Tuberculous infection should be considered in relatively indolent infections of the spine, hip, and knee.
3. Introduced osteomyelitis. Osteomyelitis following trauma or bone surgery commonly presents with persistent or recurrent fever, increasing pain at the operative site, and poor wound healing. Incisional healing difficulties include protracted drainage, cellulitis, suture abscesses, wound hematomas or seromas, and dehiscence. Bone infection should also be considered in cases of fracture non-union and failed arthrodesis. Prosthetic joint infection is associated with joint pain (95%), fever (43%), or cutaneous sinus drainage (32%).
4. Osteomyelitis by contiguous spread. Osseous involvement by spread of infection from an overlying chronic ischemic or neuropathic foot ulcer typically presents in patients with long-standing insulin -dependent diabetes mellitus or other vascular or neuropathic disease and involves the metatarsals (44%) or the proximal phalanges (32%). It is characterized by erythema (86%), swelling (75%), cellulitis, and necrosis, but pain is variable because of the frequent presence of sensory neuropathy (94%). Features correlated with the development of osteomyelitis in patients with such ulcers include the duration of unhealed ulceration (mean, 4 months), size of the ulcer (area 2 cm ² and depth >3mm), and the presence of exposed bone. Additional presentations of osteomyelitis caused by contiguous spread of infection are listed in Table 42-1.
5. Chronic osteomyelitis. Acute exacerbations of chronic osteomyelitis present with local bone pain and tenderness in the area of previous osseous involvement. Fever with focal swelling, erythema, and increased warmth may also occur. Purulent drainage through an old or new cutaneous sinus frequently develops and usually is accompanied by defervescence as the inflammatory process is decompressed.

III. Diagnosis. The diagnosis of osteomyelitis is based on the compilation of clinical observations (history and physical examination) and findings from well- chosen imaging studies, and on the exercise of sound clinical judgment to give appropriate weight to the data collected. It is essential to be mindful of the pathogenesis attributed to the specific clinical situation being evaluated. Establishing the presence of osteomyelitis includes both confirming the site of bone involvement and identifying the causative microorganism(s). Osseous infection should be differentiated from septic arthritis and bursitis, cellulitis, soft-tissue abscesses, bone fractures, and neoplasms, and from bone infarcts seen with sickle cell hemoglobinopathy and Gaucher's disease. Anatomic delineation of bone infection depends substantially on imaging techniques.

1. Radiology. The earliest osseous roentgenographic changes in hematogenous infection are medullary areas of lucency, which require 30% to 50% decalcification to be seen and take 2 to 4 weeks to develop. As sepsis progresses, periosteal elevation, thickening, and new bone formation may be seen, with sequestra and sclerosis occurring in chronic cases. Vertebral osteomyelitis appears initially as disk space narrowing with subsequent cortical degradation at the adjacent end plates. With the introduced form of osteomyelitis, bone resorption is evident at the site of the fracture, fixation device, or bone- cement interface of a joint prosthesis. When bone infection develops by spread from contiguous tissue, subperiosteal bone lucencies and cortical erosions are demonstrable and followed by lytic medullary lesions. Periosteal reaction is commonly seen in patients with chronic, deep ulcerations overlying bone but is not diagnostic of osteomyelitis. There is commonly a delay in radiographic improvement during the healing phase of bone infection, and 30% of patients have worsening roentgenographic findings while they are improving clinically on therapy. Computed tomography (CT) can be useful in demonstrating small osseous changes, sequestra, and extraosseous extension of infection. Roentgenography is the most specific and least expensive imaging technique for diagnosing osteomyelitis. In patients with very acute, fulminant infection (such as childhood hematogenous osteomyelitis), roentgenograms may not be sufficiently sensitive to reveal diagnostic abnormalities at the time of presentation, and additional studies are needed. However, most cases of bone infection present subacutely, with adequate duration of infection for demineralization to occur, and can be well evaluated by radiologic techniques.
2. Radionuclide scans . Technetium diphosphonate bone scans, gallium citrate scans, and indium-labeled leukocyte scintigraphy are much more sensitive than roentgenography and usually demonstrate increased radionuclide uptake at the onset of symptoms. However, these imaging methods are plagued by inadequate specificity and spatial resolution, so they cannot be relied on to be definitively diagnostic.

   Inflammatory and degenerative processes in surrounding soft tissues, areas recently subjected to orthopedic surgery or trauma, bone fractures, and neoplasms produce abnormal findings on nuclide scans in the absence of osteomyelitis. Table 42-2 lists the approximate sensitivity and specificity of various imaging techniques in clinical situations in which the diagnosis is ambiguous because of the presence of a pathologic process in adjacent tissues (e.g., cellulitis, edema) or a comorbid state in the osseous tissue (e.g., diabetic osteoarthropathy, bone infarction, recent trauma).

   
   

   Table 42-2. Estimated relative value of different imaging techniques for the diagnosis of osteomyelitis in complicated cases

   
   
   
3. Magnetic resonance imaging (MRI) can detect osteomyelitis earlier and with a sensitivity equal to or greater than that of roentgenography, and its spatial resolution is much better than that of scintigraphy. Negative MRI findings are strong evidence against the presence of osteomyelitis. However, this technique is too nonspecific to be the optimal determinant for the presence of bone infection. An increased MRI signal represents small increases in tissue water content. Gadolinium is deposited in the extracellular fluid compartment in areas of increased vascular permeability and thereby increases the sensitivity of MRI for inflammation of all etiologies. MRI detects the bone edema of osteomyelitis; however, differentiation of nonspecific reactive marrow edema from adjacent foci of infection or other causes of soft-tissue edema is often not possible. The specificity of MRI for osteomyelitis is significantly diminished because of false-positives related to sterile inflammation, edema in tissues adjacent to bone, bone infarction, recent trauma, diabetic osteoarthropathy, heterotopic bone formation, neoplasm, and local radiation therapy (see Table 42-2). A recently developed technique that uses fat suppression with MRI scans has increased the specificity, but not yet to the level needed for confidence.
4. Microbiologic studies. The specific etiologic pathogen in osteomyelitis should be delineated because the microbe is never sufficiently predictable to allow routine presumptive therapy (see Table 42-1). Moreover, knowledge of the antimicrobial sensitivity of the isolated causative bacterium is essential to design optimal therapy. Blood cultures are positive in 25% to 50% of children with acute hematogenous osteomyelitis but in fewer than 10% of the other forms of bone infection. If septic arthritis or soft-tissue abscess accompanies the osseous infection, arthrocentesis or abscess aspiration cultures can be diagnostic. However, superficial cultures of open wounds, wound drainage, or skin ulcerations and cultures of cutaneous sinus tracts do not delineate the true bone pathogen(s). Sinus tract and sinus drainage cultures often reflect colonizing flora and cannot be relied on to identify the actual cause of the deep osseous infection. In patients with deep, chronic skin ulcers by which infection has spread to bone, curettage cultures from the base of the ulcer have a 75% correlation with osseous tissue cultures. Bone aspirate cultures are positive in 50% to 60% of patients, whereas bone biopsy cultures are positive in 70% to 93% of cases and should be sought (percutaneously or by operative debridement) when there is no overlying skin ulcer and the microbiologic etiology has not been otherwise determined. Specific cultures for mycobacteria, fungi, and anaerobes should be considered when routine bacterial cultures are negative.
5. Guidelines for diagnosing osteomyelitis (based on pathogenesis)
   1. Childhood acute hematogenous osteomyelitis
      1. Plain roentgenograms of the suspect bone. CT if additional detail is needed.
      2. Technetium bone scan, if the roentgenograms are not diagnostic.
      3. MRI with gadolinium if roentgenograms and technetium bone scan are not diagnostic but clinical suspicion is strong.
      4. Blood cultures.
      5. Bone biopsy for cultures and histopathology.
   2. Adult acute hematogenous osteomyelitis (vertebral osteomyelitis)
      1. Plain roentgenograms of the suspect bones. CT if additional detail is needed.
      2. MRI with gadolinium if roentgenograms are not diagnostic.
      3. Blood cultures.
      4. Bone biopsy for cultures and histopathology.
   3. Introduced form of osteomyelitis
      1. Plain roentgenograms of the suspect bones. CT if additional detail is needed. Serial roentgenography or CT at intervals during a 4-week period to observe any meaningful osseous changes.
      2. Arthrocentesis (if a prosthetic joint is present) for cultures.
      3. Bone biopsy, usually with debridement surgery, for cultures.
   4. Osteomyelitis caused by contiguous spread of infection
      1. Plain roentgenograms of the suspect bones. CT if additional detail is needed. Serial roentgenography or CT at intervals during a 4-week period to observe any meaningful osseous changes.
      2. Indium-labeled leukocyte scan (or sequential indium or technetium scan), if suspicion is not strong, for the negative predictive value of a negative scan result. (A positive scan result would not be as diagnostic as meaningful osseous changes on serial roentgenograms.)
      3. MRI with gadolinium and fat suppression if clinical suspicion is strong.
      4. Culture base of soft-tissue ulcer before starting antibiotic therapy. Prepare ulcer base with iodine, then alcohol (allow alcohol to evaporate). Abrade ulcer base with a culture swab.
      5. Avoid percutaneous bone biopsy, if possible, because (a) this trauma may induce bone necrosis and infection, (b) the histopathology of neuropathic osteopathy may resemble that of osteomyelitis, and (c) the bone biopsy specimen can be contaminated with organisms from the contiguous infected soft tissue.
      6. If soft-tissue infection is present, treat with antibiotics and obtain serial roentgenograms with the patient off antibiotic therapy at intervals during a 4-week period to determine subsequently the presence or absence of osteomyelitis.
   5. Chronic osteomyelitis
      1. Plain roentgenograms of the suspect bone.
      2. CT or MRI to delineate sequestra.
      3. Sinogram (roentgenogram with dye inserted into a cutaneous sinus tract) to demonstrate the osseous site of sinus tract origin, which often is the site of sequestrum.
      4. Bone biopsy (percutaneously or at debridement surgery) for cultures and histopathology.

         Note: Sinus tract cultures are not reliable for delineating the true etiologic pathogen.

IV. Treatment

1. Antibiotics. Acute osteomyelitis is curable with adequate antimicrobial therapy accompanied by surgical debridement when necessary. IV antibiotics are commonly used, but oral agents are also effective when the quantitative susceptibility of the pathogen is sufficient and when gastrointestinal absorption and compliance are ensured. The exact potency and duration of treatment necessary to eradicate bone infections are not known. Antimicrobial agents that produce trough serum bactericidal activity at a 1:2 titer have been associated with highly successful outcomes . Therapy should be administered for at least 4 to 6 weeks. Designing a therapeutic regimen for osteomyelitis depends substantially on the sensitivity of the isolated pathogen to specific antimicrobial agents, which underscores the essential importance of delineating the pathogen in these infections.

   Chronic osteomyelitis is treatable but not curable with systemic antibiotics. Acute exacerbations of these persistent infections can be suppressed successfully by debridement of identifiable sequestra followed by protracted courses of parenteral or oral antimicrobial agents. Theoretically, depot administration of appropriate antibiotics to the site of infected, avascular bone via antibiotic-loaded polymethylmethacrylate beads could eradicate the persistent bacteria unreachable by systemic (blood vessel-dependent) therapy.

2. Surgery. When acute osteomyelitis is complicated by abscess formation or extensive necrosis, surgical debridement is an important adjunct to antibiotics. In cases associated with prosthetic joints or internal fixation devices, the foreign body usually needs to be removed to effect a cure. In chronic osteomyelitis, surgery is often the primary therapeutic treatment. Necrotic tissue and sequestra should be removed when possible, and dead spaces should be eliminated with packing, bone grafts, muscle pedicles, or skin grafts. Surgery is also important when neurologic structures are threatened (e.g., cord compression in vertebral osteomyelitis). In osteomyelitis associated with peripheral vascular disease, amputation of the affected area is frequently required if the infection does not respond to parenteral antibiotics.

Bibliography

Brause BD. Infections with prostheses in bones and joints. In: Mandell GL, Bennett JE, Dolin R, eds. Principles and practice of infectious diseases. New York: Churchill Livingstone, 1994:1051.

Crim JR, Seeger LL. Imaging evaluation of osteomyelitis. Crit Rev Diagn Imaging 1994;35:201.

Haas DW, McAndrew MP. Bacterial osteomyelitis in adults: evolving considerations in diagnosis and treatment. Am J Med 1996;101:550.

Lew DP, Waldvogel FA. Osteomyelitis. N Engl J Med 1997;336:999.

Lipsky BA. Osteomyelitis of the foot in diabetic patients. Clin Infect Dis 1997;25:1318.

Morrison WB, et al. Osteomyelitis in feet of diabetics: clinical accuracy, surgical utility, and cost-effectiveness of MR imaging. Radiology 1995;196:557.

Sapico FL, et al. The infected foot of the diabetic patient: quantitative microbiology and analysis of clinical features. Rev Infect Dis 1984;6:S171.

Torda AJ, Gottlieb T, Bradbury R. Pyogenic vertebral osteomyelitis: analysis of 20 cases and review. Clin Infect Dis 1995;20:320.

Weinstein M, et al. Multicenter collaborative evaluation of a standardized serum bactericidal test as a predictor of therapeutic efficacy in acute and chronic osteomyelitis. Am J Med 1987;83:218.

Books@Ovid
Copyright 2000 by Lippincott Williams & Wilkins
Stephen A. Paget, M.D., Allan Gibofsky, M.D., J.D. and John F. Beary, III, M.D.
Manual of Rheumatology and Outpatient Orthopedic Disorders

-->