126

Osteomyelitis: Introduction

Osteomyelitis, an infection of bone that leads to tissue destruction and often to debility, can be caused by a wide variety of bacteria (including mycobacteria) and fungi and may be associated with viral infections. Its management must be individualized and depends on numerous factors, including the causative organism, the specific bone involved, vascular supply, nerve function, foreign bodies, recent injury, the physiologic status of the host, and associated comorbidities. The spectrum of the disease can range from extensive (e.g., tibial and vertebral osteomyelitis) to localized (e.g., bone invasion associated with a tooth abscess). Two major classification systems for osteomyelitis are used in making decisions about medical therapy and surgery. Lee and Waldvogel categorized cases as acute or chronic, hematogenous or contiguous, and with or without vascular compromise. The Cierny and Mader classification system for long-bone osteomyelitis encompasses the location and extent of the infection as well as a number of other factors.

Etiology (Table 126–1)

The foremost bacterial cause of osteomyelitis is Staphylococcus aureus. Gram-negative organisms such as Pseudomonas aeruginosa and Escherichia coli, coagulase-negative staphylococci, enterococci, and propionibacteria may also be involved. Mycobacterium tuberculosis is a common cause of osteomyelitis in countries with limited medical resources; other mycobacterial species that infect bone include M. marinum, M.chelonei, and M. fortuitum. Fungal etiologies include Candida, Coccidioides, Histoplasma, and Aspergillus species. Noninfectious pathogenic mechanisms that may cause disease mimicking osteomyelitis include avascular necrosis, rheumatoid diseases, neuropathy with chronic trauma, gout, and malignancies.

The precipitating event(s) for osteomyelitis vary greatly. The prosthetic joint implants and stabilization devices that are increasingly being used in orthopedic surgery are associated with complex infections. Trauma is also a common cause of infection, especially when a wound is involved and there is contamination of bone or surrounding tissue along with significant tissue damage or destruction. Even in the absence of an open wound or a compound fracture, damaged tissue and extravasated blood may slow the circulation, establishing a favorable medium for the growth of bacteria that may reach the area through low-level bacteremia from the peripheral venous circulation or from distal lymphatic channels. Bacteremia—whether due to endocarditis or due to seeding from other sites of infection (e.g., abscesses, boils, or vascular devices)—is also a frequent etiologic factor in osteomyelitis. Studies of S. aureus bacteremia indicate a rate of metastatic osteomyelitis approaching 28% if there is a prosthetic joint in place; S. aureus bacteremia can be complicated by the involvement of methicillin-resistant strains (MRSA), which are progressively replacing strains that are more susceptible to antibiotics. The overlapping circulations of the urinary tract and the spine may be a source of vertebral osteomyelitis due to urinary tract pathogens such as E. coli and Klebsiella. Additional predisposing factors include a poor arterial and venous supply, which may limit perfusion to bone to the point of an inadequate response and poor healing, even in patients with normal function. Host factors such as diabetes and its consequences contribute significantly to the development of osteomyelitis through impaired immunity with hyperglycemia, loss of sensation, vascular disease, and renal failure.

Epidemiology

In the United States, acute osteomyelitis affects 0.1–1.8% of the otherwise healthy adult population. After a foot puncture, 30–40% of adults with diabetes develop osteomyelitis. In this country, there has been a major change in the profile of certain bacterial pathogens, with the emergence of MRSA strains over the last decade. MRSA has become a source of great concern in hospitals, especially after surgery. The morbidity and economic consequences appear to be greater for MRSA osteomyelitis than for osteomyelitis caused by methicillin-sensitive S. aureus strains. However, it is not clear that these poorer outcomes for MRSA are due to new or more destructive virulence factors. Rather, they may simply be the result of a delay in effective antimicrobial treatment.

The types and etiologies of osteomyelitis vary by region and with time. The United States has seen a rise in infections related to the increasing use of orthopedic surgery for correction of deformities and implantation of screws, pins, rods, plates, and prosthetic joints. With the aging of populations and the epidemics of obesity and diabetes in some countries, the frequency of these predisposing factors continues to increase, requiring adaptations in treatment approaches. Any type of instrumentation may lead to infection in a small proportion of cases. Osteomyelitis attributable to orthopedic devices and surgical interventions is considerably less common in countries with limited medical resources, where tuberculosis may be the dominant infection and brucellosis is not unusual. In many of these areas, agricultural injuries, industrial accidents, and war wounds are much more common than in wealthy countries, and the pathogens causing infection reflect those injuries. Osteomyelitis is more common in situations where wounds cannot promptly be debrided and repaired, microbiology laboratories are not readily available, and effective antimicrobial agents are in short supply.

Pathogenesis

The most common predisposing factor for osteomyelitis is an area of bone or contiguous surrounding tissue that is abnormal in terms of viability, blood supply, sensation, or edema. The damaged tissue not only compromises healthy circulation to the area but may slow the flow of venous blood and lymph, thereby providing nutrients to bacteria and fueling ongoing damage. Host factors such as poor nutrition and immunosuppression may also be relevant. Diabetes in adults poses the most significant risk. Diabetic neuropathy adds to the progression of osteomyelitis as the patient may be unaware of infection as it spreads into the bone; the consequences include thousands of amputations each year. Additional sources of immunosuppression, such as chemotherapy and treatment with glucocorticoids or tumor necrosis factor (TNF) inhibitors, also inhibit normal defense mechanisms and thus predispose to more frequent and serious infections whose symptoms are diminished because of reduced inflammatory responses.

The bacteria involved in osteomyelitis perpetuate themselves by elaborating toxins that further damage tissues, including bone. S. aureus is particularly adept in this respect; it colonizes the nasal area in about one-third of healthy individuals and can produce a wide variety of cytokines, enzymes, and toxins that destroy tissue and affect neutrophil response. Some S. aureus bacteria survive uptake into the phagocytic vacuoles of macrophages and continue to cause disease and recrudescence by persistently eluding the usual defense mechanisms. This capacity for "hibernation" and persistence may allow S. aureus to remain dormant for decades before infection erupts at the sites of old injuries (e.g., shrapnel or other penetrating wounds).

Coagulase-negative staphylococci are generally not as virulent as S. aureus but have been found to persist by producing a biofilm that protects them from the host and apparently allows them to exist for many years on prosthetic joints, with minimal symptoms. The extent to which other organisms use biofilm to their advantage is unclear, but biofilm production probably plays a significant role in osteomyelitis, especially the chronic forms.

Multiple bacteria may be recovered from cultures, especially when there is an entry wound. Decisions about which ones to target in antibiotic therapy are often difficult. Common skin-dwelling and colonizing microbes usually do not need to be treated, and overtreatment in fact results in unnecessary toxicity and increases antimicrobial resistance among the organisms that survive. Anaerobic bacteria can often be recovered and may play a synergistic role with usual or unusual pathogens; specific therapy is sometimes beneficial in these situations.

The intrinsic factors of organisms that are responsible for persistence and bone destruction have not yet been identified. However, there is probably strain-to-strain variation in virulence factors produced by particular clones, with some strains consequently much more virulent than others. The prevention of biofilm production merits investigation in this regard.

Approach to the Patient: Osteomyelitis

The best approach to the care of a patient with significant osteomyelitis is to assemble a team of providers who can work together in considering the microbiology of the infection and make sound decisions about antibiotic therapy and surgery. The most effective program will include evaluation and management of antibiotics, microbiology, pharmacology, glucose levels, vascular disease, neuropathy, and renal function, with close follow-up by a knowledgeable physician who is interested in leading the team in coordinating care.

When osteomyelitis is suspected, a careful, methodical approach is needed (see "Clinical Manifestations and Diagnosis," below). Patients should be educated about the significance of an infection that involves bone, especially if risk factors cannot be eliminated. Blood tests, cultures, standard radiography, scans, biopsies, and surgery may all be necessary for a clear diagnosis and full delineation of the pathogen. Collection of this baseline information can be very important in both early and late decision-making.

Initial evaluations for osteomyelitis must be aggressive, as the infection can progress rapidly in the absence of antibiotic therapy effective against the wide variety of potential pathogens. Inadequacies in cultures, surgery, or temporizing measures may greatly exacerbate the damage caused by the infection. Hospitalization may be indicated for rapid multispecialty evaluation, imaging, and stabilization of complex infections such as with a diabetic foot. Outpatient therapy may not be adequate for the teamwork and interventions needed. Early admission and procedures may actually shorten the length of hospital stay.

The physician should inform the patient about the value of all the necessary evaluations, the implications of surgery, and the possibility of a prolonged course of IV antibiotic therapy, whether in the hospital or at home. A patient's fear of amputation can lead to inordinate delays in seeking treatment that allow the infection to progress. Moreover, it is not unusual for a patient to refuse surgery and amputation even though such treatments will clearly increase the likelihood of a functional lifestyle. Therefore, it is best to prepare patients early on if there may be negative outcomes such as amputation and perhaps to set criteria and timelines for success or failure of therapy and interventions.

Clinical Manifestations and Diagnosis

Diagnosis of acute osteomyelitis within the first few weeks of onset is important and is usually relatively easy. If the diagnosis is missed, however, the symptoms may become chronic, with slow progression or a dormant phase of several years.

A thorough history and physical examination are the mainstays of evaluation for osteomyelitis. A clear pattern of pain, swelling, and possibly drainage after surgery or injury should raise suspicion, but such indicators may not all be present, even in a patient with neuropathy, compromised circulation, chronic edema, organ failure, diabetes, or other predisposing factors. Direct questions about previous injuries, infections, surgeries, or hardware implantation—even decades earlier—can yield information critical in guiding empirical antibiotic therapy and surgery. A history of injury is particularly important, even if the skin was not broken and there were no clinical signs of bacteremia. It is not unusual for a soft tissue injury to serve as a nidus of secondary bone infection, presumably seeded by low-level bacteremia and often occurring without symptoms. Other sources of seeding may include boils, abscesses, cellulitis, or injection sites. A careful examination is essential in identifying additional predisposing factors and assessing the role of comorbidities such as neuropathy, arterial disease, venous insufficiency, and chronic trauma that can lead to severe accumulation of callus in insensate feet.

Careful consideration and assessment of disorders that may mimic or accompany osteomyelitis are essential. Arthritis, gout, ischemia, neuropathies, and recent surgery may be diagnosed when osteomyelitis is the real cause of symptoms on a cofactor. For example, chronic back pain may be attributed to degenerative arthritis, but there can be a substantial loss of neurologic function if the pain is actually due to diskitis with vertebral osteomyelitis.

Correctly diagnosing osteomyelitis early has crucial implications for later function, disability, treatment cost, and risk of a fatal outcome. A variety of tools must be used to definitively diagnose or conclusively rule out an infection. A standard x-ray is a good starting point that can reveal a variety of abnormalities (Fig. 126-1A) and may eliminate the need for further imaging studies. Bone loss, sequestra, periosteal elevation or swelling (which can develop early on), and shadows around foreign bodies are hallmarks of bone infection. However, these findings may also be found with other disorders, such as tumors, trauma, avascular necrosis, and gout. Standard two-dimensional images can be of limited value in assessing complex bones. The value of radiology may be limited by the time required for an infection to become apparent; actual dissolution or resorption of bone due to infection may not be apparent for several weeks or more.

Depending on the results of the initial x-ray, further investigations with invasive techniques may be appropriate. Collection of pus by needle aspiration through a clean area from a closed pocket not only documents bone infection but also permits recovery and evaluation of the pathogen(s). A culture of a wound swab may be of some value but is clearly less reliable in identifying the real culprit(s), which may be present in the bone but absent from its surface. Biopsy provides more accurate microbiologic information than needle aspiration and supplies tissue for pathology studies, which may be helpful. Some organisms that usually are not recovered (in a timely fashion or at all) by standard cultures may be rendered visible with special staining of tissue samples. Unfortunately, the size of the needle used for needle biopsy may not be appropriate for small bones of the hands or feet. Open surgical exploration, biopsy, and drainage, which can provide high-quality tissue samples for culture and pathology and offer a view of the infected bone and surrounding area, should also be considered. Necrotic tissue can be removed and circulation assessed with one procedure. Polymerase chain reaction and other sequencing technologies are increasingly being used to detect and identify specific organisms—and even to determine their susceptibilities—within hours instead of days or weeks. Information on specific strains of unusual organisms may be of value, especially in difficult cases.

Laboratory tests are useful in assessing osteomyelitis but usually do not yield specific information relevant to etiology or severity. Leukocytosis may be noted in acute infection but is less likely in chronic infection, which may also be associated with anemia. Determination of the erythrocyte sedimentation rate (ESR) is a simple, inexpensive aid to diagnosis; it serves as an indicator of response with S. aureus infections but is not as useful for gram-negative infections because the cytokines and inflammatory elements that result in elevations are different for gram-positive (S. aureus) than for gram-negative infections. C-reactive protein (CRP) measurement may be helpful, especially in the evaluation of children, but may not be as useful as an ESR determination in some cases. CRP changes occur earlier in response to bacterial infection. Both ESR and CRP determinations have significant limitations in multifactorial diseases, with elevated values reflecting conditions other than osteomyelitis. Additional laboratory tests for diseases associated with bone loss that may mimic or complicate osteomyelitis should include measurement of glucose levels and tests for renal failure, gout, vasculitis, and rheumatoid diseases.

Additional imaging studies may be of value if the diagnosis remains unclear. CT can delineate bone more clearly than standard radiography and offers three-dimensional displays that can be extremely useful in detecting abnormalities and devising a surgical approach. MRI (Fig. 126-1B–D) provides high-quality images of the soft tissue around the bone abnormality and may be essential in diagnosing an epidural abscess related to vertebral osteomyelitis. Technetium and leukocyte isotope scans offer insight into the activity of the disease process and the affected site(s). Although these additional screening tools may be helpful in evaluation and decision-making, they may not be cost-effective.

Treatment: Osteomyelitis

Therapy for osteomyelitis is challenging because of the variety of causative organisms, the usual comorbidities, the need for a prolonged course and IV administration, the common physical limitations of the patient, and high costs. An aggressive therapeutic approach is warranted given the dire consequences of failure of medical therapy, which can include loss of limbs. The sooner the infection is diagnosed and treated, the better the outcome and the less damage done during delays in intervention. Antibiotic therapy should be used aggressively to stop disease progression and should be designed to avoid the development of resistant organisms. Early surgical intervention (e.g., debridement) can confirm the infection, identify and characterize the etiologic agent(s), and remove dead or devitalized tissue that may be providing bacteria with nutrients and allowing them to spread. A variety of antibiotics are available for most of the likely pathogens (Table 126–2), although the most common pathogen—S. aureus—continues to evolve mechanisms to elude these drugs. MRSA strains represent an increasing problem in both the hospital and the community. Staphylococci and Enterobacteriaceae resistant to even more antibiotics than MRSA appear to be evolving.

The most common targets for empirical antibiotic therapy are staphylococci, which are carried asymptomatically in and around the nares by nearly one-third of healthy people. The common -lactam antibiotics provide excellent results against methicillin-sensitive S. aureus strains. Oxacillin and nafcillin are first-line agents but may elicit more adverse reactions than cephalosporins. Cefazolin is a reasonable alternative in the hospital, but ceftriaxone is preferred as an outpatient drug because it can be given (by the IV or IM route) only once a day.

MRSA strains have been controlled with vancomycin for many years, but this drug appears to be losing its effectiveness against these microbes. New antibiotics have been designed to fill this need, although their efficacy has not been documented. In an outpatient setting, vancomycin does not appear to be as effective against methicillin-susceptible staphylococcal osteomyelitis as oxacillin or ceftriaxone. Publications about the value of daptomycin for osteomyelitis are encouraging. Tigecycline is active against MRSA but is only bacteriostatic and does not yet have a well-established outcomes record. Telavancin may also be of value against vancomycin-resistant staphylococci but has not yet been adequately tested for bone infections.

Additional antimicrobial agents for use against staphylococcal infections include linezolid, which offers the advantage of both oral and IV formulations but is bacteriostatic and has not yet been well studied. Moreover, its use—although apparently less expensive than that of other parenteral drugs—is limited by its cost. Clindamycin can also be used as both an IV and an oral agent, although antimicrobial resistance is a growing problem. Rifampin, a potential adjunct to other antistaphylococcal agents, is highly active in vitro and can penetrate phagocytic vacuoles to reach staphylococci therein. Unfortunately, resistance develops rapidly if rifampin is used alone, and clinical outcomes are not always as good as anticipated. Other agents, such as aminoglycosides, folic acid inhibitors, and macrolides, may play a limited role; they generally are neither as effective nor as toxic as other available agents.

Fluoroquinolone antibiotics offer both IV and oral therapy options and are often included in the standard recommendation for treatment of many susceptible strains of Enterobacteriaceae and Pseudomonas species. Drugs of this class do, however, have some limitations in terms of emerging resistance (even during therapy) and may exert some adverse neuromuscular effects (e.g., tendon rupture and impaired healing) that may be particularly relevant to the prolonged courses of antibiotics usually needed to cure the infection. In general, fluoroquinolones should not be used to treat S. aureus infections because of these limitations and the availability of better-studied antibiotics.

The optimal route and duration of therapy for osteomyelitis remain controversial. The usual recommendations stem from a 1970 study in which cases of osteomyelitis were characterized and outcomes were evaluated in relation to the duration of IV therapy. Better outcomes appeared to be related to a course of 4 weeks in some types of infection. Even though the characteristics of the bacteria and the available antibiotics were quite different at that time, a 4- to 6-week course of IV therapy remains the standard and is the usual recommended minimum. This recommendation has been challenged in pediatric studies in light of increasing evidence that oral agents and shorter courses may be adequate. Because some of the active agents reach comparable levels when given by mouth, a switch from the recommended IV administration to oral therapy may be appropriate in some situations. The proper duration of antimicrobial therapy depends on a variety of factors, including the infecting organism, the bone involved, surgical procedures, and drug tolerance and safety. Prolonged courses may be justified by extensive disease, immunocompromise, poor clinical response, and vertebral osteomyelitis. Whether a bone infection has truly been cured becomes clear only over time; relapse is not uncommon and may occur years later, especially in patients with ongoing risk factors and comorbidities. The literature suggests that a 6-month follow-up period is adequate to determine the success of treatment. Patients should be followed for at least that long, even though antibiotics have been discontinued. The possibility of relapses and the potential for their prevention should not be overlooked.

Surgery is an important tool in the treatment of osteomyelitis, offering the benefits of direct observation, prompt removal of all devitalized tissue and bone, and drainage of the infection site. Nevertheless, it is not without risk, and loss of bone or other tissue may adversely affect function. In addition, because bone may regenerate to some degree when infection is eradicated, surgery is not always needed. Surgical approaches vary with the bone involved and the extent of disease. The Cierny-Mader classification system is helpful when three-dimensional imaging is done, and MRI may help determine the viability of bone or marrow. Residual dead spaces are a source of concern and may require tissue flaps and closure. Local antibiotics and impregnated cement or beads may be of value but not should not replace IV antibiotic therapy without further study. If surgery is performed and most or all of the infected bone is removed, a full 4- to 6-week course of IV therapy probably is not necessary. However, the precise duration that is required is not clear and most likely depends primarily on the other factors involved in individual cases. One week of IV therapy after surgery may be justified to ensure pathogen eradication and healing.

Outpatient parenteral antibiotic therapy (OPAT) is a valuable means of providing the long course of IV antibiotics that is considered the standard of care and has been proven efficacious over decades. Despite potential risks outside the hospital that patients and their providers must consider, OPAT is safe and effective when properly managed and administered. This approach is conducive to a better quality of life in a familiar setting, is considered safer because of the lack of exposure to hospital-related infections (which affect 1 patient in every 20 admitted), is much less expensive than treatment administered in the hospital, and generally facilitates recovery, often allowing the patient to return to work or resume other day-to-day activities during the treatment course.

Complications

The complications of osteomyelitis are numerous and are most commonly related to loss of full function of the bone or supporting tissues. Fractures are more likely with progressive disease. Local spread and dissemination of infection are also possible. Misdiagnosis is particularly likely when another disease is complicating the infection. In rare instances, chronic inflammation and infection may lead to malignant transformation into squamous cell carcinoma or sarcoma.

Prognosis

The outcomes of osteomyelitis vary tremendously depending on the bone involved, the predisposing factors, the underlying diseases, and the treatment provided. Standard guidelines cannot be applied uniformly; e.g., a case of mandible infection arising from a tooth abscess may be cured with an extraction alone, whereas a case of vertebral osteomyelitis may require a prolonged course of IV therapy as it cannot be approached surgically without neurologic sequelae. For large bones, the 4- to 6-week course of IV therapy still seems reasonable, although recent studies suggest that with some new antimicrobial agents a shorter course of IV therapy, possibly with an early switch to oral therapy, may be sufficient. Determining the outcome even of long-bone osteomyelitis is complicated by uncertainty as to the duration of follow-up needed. The actual outcome in terms of debility and limb salvage may be as dependent on underlying and complicating factors and care as it is on antibiotic therapy.

Prevention

Osteomyelitis can be prevented in some instances by better infection-control measures, especially before surgery. Both mupirocin and chlorhexidine are of proven value in preventing operative infections, which are an increasing cause of bone infections associated with implanted material. Prompt treatment of bacteremia and elimination of sources of infection (e.g., boils or folliculitis) before surgery and in other situations may prevent infections. Aggressive surgical management of injuries may also help avoid the constellation of factors that lead to bone infections.

Awareness of persistent sites of infection and reasonable attempts at eradication may promote prevention. Many persistent infections that do not initially impair function or cause pain are ignored by patients; an example is provided by the classic problem of diabetic foot infections, with ulcers that burrow into the soles of insensate feet and often reach bones. Likewise, sacral ulcers are often overlooked or ignored both by physicians and by patients with neurologic impairment. Attempts to eradicate or close entry wounds are critical and should be undertaken early on.