Current Diagnosis (Figure 1)
• A bone biopsy remains the gold standard when diagnosing osteomyelitis.
• All patients should have blood cultures performed.
• No one imaging modality is routinely recommended to diagnose osteomyelitis.
• Plain film radiography should always be the initial study and can be diagnostic if positive.
• Magnetic resonance imaging (MRI) can detect acute osteomyelitis as early as 3 days.
• The most important factor in any treatment is identification of the organism.
• The optimal duration of antibiotic therapy remains undefined, with most authorities recommending treatment for about 6 weeks.
• Hematogenous osteomyelitis is usually monomicrobial, whereas contiguous infections are usually polymicrobial.
Osteomyelitis is an inflammation of the bone as a result of a bacterial infection, resulting in acute or chronic inflammation caused mainly by a bacterial infection. It can primarily occur by hematogenous spread, contiguous spread to the bone from adjacent tissues, or direct inoculation usually from trauma or surgery. Patients usually complain of pain, swelling, and tenderness over the affected area of bone.
Osteomyelitis has changed over the years from a primarily hematogenous disease with high mortality to one of high morbidity as a result of infection from contiguous or direct inoculation, especially in those with diabetes mellitus and/or peripheral vascular disease.
Contiguous or local extension may occur from spread from adjacent structures or direct implantation of organisms as seen in cases of trauma, placement of internal fixation devices for fracture repair, or other surgical procedures. Direct inoculation osteomyelitis tends to involve multiple organisms and results in more clinical localized manifestations than infections from hematogenous spread, where the osteomyelitis is usually monomicrobial. Hematogenous osteomyelitis most commonly involves the vertebrae, with the lumbar spine most commonly affected, and is also more common in infants and children, involving metaphysis of their long bones.
There are several methods used to classify osteomyelitis, separating the illness into acute or chronic types; into hematogenous or contiguous types; and according to the mode of transmission—local extension from adjacent structures or direct implantation of organisms as seen in cases of trauma or surgical procedures.
The Cierny-Mader system assigns the patient to one of several groups based on the anatomy of the infection and health of the host, including the increased risk for patients who have peripheral arterial obstructive disease or diabetes mellitus. Patients are further separated according to systemic and local factors that may influence disease progression or healing, such as diabetes, extremes of age, and tobacco use. In patients with diabetes, osteomyelitis is most commonly caused by overlying lower limb cellulitis. In the nondiabetic adult patient, vertebral osteomyelitis is most common.
The Cierny-Mader classification allows the clinician to use tested treatment protocols, including chemotherapy, surgery, and adjunctive therapies, that are most effective for a specific class of disease and for the host status. The Cierny-Mader system divides the patients into four anatomic groups. Stage 1, or medullary osteomyelitis, is confined to the endosteum of the bone and is often hematogenous. State 2, or superficial osteomyelitis, is localized to the surface of the bone. This is a true contiguous lesion. Stage 3, or localized osteomyelitis, involves cortical sequestration or cavitation, or both, and is a full-thickness lesion that extends into the medullary region. Stage 4, or diffuse osteomyelitis, involves the hard and soft tissues (“through and through”), and it requires surgical debridement of the affected bone to remove all the infected tissue.
For treatment to be successful, the patient must be physiologically able to heal any wounds, defend against contamination or infection, and tolerate the stress of treatment. The hosts are classified as A, B, or C, depending on the ability to resist infections. Those with good immunity are classified as an A host, whereas a B host is compromised locally (BL) or systemically (BS); Table 1 shows the physiologic classifications. The final class assigns a C rating to patients whose treatment is more detrimental than the disability from the disease. These patients may require suppressive or no treatment. The clinical stages are adjusted during the course of therapy as conditions change, allowing adjustment of the treatment protocol to optimize therapy.
Cierny-Mader Classification System
A much simpler system described by Waldvogel classifies the patient by duration (i.e., acute or chronic) and the mechanism of inoculation (i.e., hematogenous or contiguous). Contiguous infections are further classified as those with or without vascular insufficiency. However, this classification does not provide guidance for specific surgical or antibiotic therapy.
Acute hematogenous osteomyelitis is usually seen in male children of lower socioeconomic class before the age of 2 years or between 8 and 12 years. There may be some genetic influences. Aboriginal children in Western Australia are known to suffer from acute hematogenous osteomyelitis at a rate nearly four times that of Western European children living in the same neighborhood. An acute infection will progress to chronic osteomyelitis if it is not treated.
” Some advocate doing plain radiograph, CBC, sed rate, and blood cultures as part of the initial evaluation of osteomyelitis
FIGURE 1 Algorithm for treatment of suspected osteomyelitis. (Courtesy Edward T. Bope.)
Chronic osteomyelitis is usually the result of direct inoculation (e.g., trauma, surgery). Its epidemiology is less clearly described, except in diabetic foot infections.
Kremers et al., in a population-based study, described trends in the incidence of osteomyelitis over a four-decade period from 1969 to 2009. The study demonstrated that the incidence of osteomyelitis has increased over time and is generally higher in males than in females, and that the incidence rates remained relatively stable among children and young adults (< 50 years old). However, the incidence has almost tripled among the elderly, driven significantly by an increase in diabetes-related cases.
Because chronic infection can persist for life, it is important to have early identification and treatment to ensure the best possible outcome.
The presence of bacteria in an open wound is not sufficient to cause infection. It is the compromised blood supply of traumatized tissue leading to necrosis and subsequent bacterial adherence that promotes the infection. Trauma can delay the inflammatory response to bacteria, depress cell-mediated immunity, and impair chemotaxis, superoxide production, and the microbial killing capacity of polymorphonuclear neutrophils (PMNs). Osteomyelitis usually involves the metaphysis, which is well vascularized and has significant bone growth.
In acute osteomyelitis, signs or symptoms are usually abrupt. Local infection is characterized by edema, vascular congestion, and small- vessel thrombosis. This leads to increased pressure within the intramedullary canal, allowing extravasation through the Havers and Volkmann canals to the periosteum. In children, the periosteum is usually more flexible and easier to detect radiographically; in adults, the bone matrix is more firmly attached to the periosteum. Untreated, the suppurative infection can reach adjacent soft tissue, leading to a cellulitis. The presence of a Brodie or intraosseous abscess without extravasation into surrounding tissue is classified as subacute pyogenic osteomyelitis. The resulting infection can lead to sequestration involving large areas of bone destruction and dead bone, with reactive bone formation leading ultimately to chronic osteomyelitis.
Chronic osteomyelitis is usually polymicrobial and is characterized by the presence of necrotic bone, new bone growth, and exudation of polymorphonuclear leukocytes, plasma cells, and other infection- fighting cells. The involucrum (i.e., layer of reactive competent bone that covers dead bone) is often dotted with tracts that allow pus to pass into surrounding tissue or to form a sinus tract to the skin surface. These sinus tracts are often contaminated with numerous organisms that do not reflect those found with direct sampling. This repetitive process of bone loss and growth and the involucrum explains why chronic osteomyelitis is difficult to eradicate with antibiotics alone. The antibiotics cannot penetrate avascular areas.
In children of all ages, the most common bacterial pathogen is Staphylococcus aureus, followed by Streptococcus pneumoniae and Kingella kingae. However, age and chronic illness allow other organisms to flourish; Salmonella and pneumococcal disease (S. pneumoniae) are common in patients with sickle cell disease. Pasteurella multocida, Streptococcus species, and anaerobes often are identified after animal or human bites. In children, most cases arise hematogenously and are characteristically seen in the metaphysis of long bones (i.e., femur, tibia, and humerus), accounting for 68% of childhood infections.
The exact mechanism is unclear, but it is thought that the extensive branching of the nutrient-rich arteries at the metaphyses of the long bones leads to sluggish blood flow and ultimate bacterial seeding.
Possible routes include the formation of small hematomas in the metaphysis, allowing microbial seeding after transient bacteremia; penetrating injuries or surgical manipulation, causing direct inoculation of bacteria into bone; and local invasion from a contiguous focus of infection.
Most infections in adults arise by direct inoculation from sources such as trauma, prosthetic joints, open fractures, and diabetic foot infections. The most common organism remains S. aureus. Other organisms to consider include Staphylococcus epidermis, Pseudomonas aeruginosa, Escherichia coli, and Serratia marcescens. Most contiguous, related infections are polymicrobial.
The severity of the signs and symptoms depends on the location of the infection, the patient’s age, and any comorbidities. Patients may experience many, few, or no symptoms. Classically, there are marked pain, erythema, tenderness, and swelling. Fever and leukocytosis are also common.
The probe-to-bone test with a sterile blunt metal tool should be included in the initial assessment of all patients with diabetes with infected pedal ulcers. A positive result consists of detection of a hard, gritty surface through the ulcerated skin surface. The probe-to-bone test along with other diagnostic tests (such as radiographic imaging or bone biopsy) can assist in the therapeutic decision-making process.
Vertebral osteomyelitis can occur either by hematogenous spread, direct inoculation from trauma or spinal surgery, or contiguous spread from adjacent soft tissue infection. Over 50% of cases are due to S. aureus. The most common presentation is back or neck pain, with or without fever, with noted local tenderness to percussion over the involved posterior spinous processes.
Vertebral osteomyelitis often causes severe pain, fever, and disability, whereas osteomyelitis of the foot rarely causes pain. An epidural abscess causes pain and neurologic deficits, whereas vertebral osteomyelitis without abscess formation has no neurologic deficits.
Children often present with systemic symptoms (e.g., fever, weight loss, pain). Pseudoparalysis may be the only sign in a newborn, but toddlers often exhibit pain, fever, erythema, edema, or warmth, or they may suddenly stop walking. In contrast, patients with chronic osteomyelitis may exhibit localized signs and symptoms, including nonhealing ulcers, purulence from sinus tracts, soft tissue edema and pain, abscesses, erythema, pain, and fatigue. Generalized signs and symptoms may be seen early in the disease, but they are unlikely in the later or chronic stages.
There is usually an elevated erythrocyte sedimentation rate and C- reactive protein. Magnetic resonance imaging (MRI) is the most sensitive radiographic technique, but the infection is confirmed by biopsy of the infected intervertebral disc space or vertebral bone. The antimicrobial therapy should be withheld until a microbiologic diagnosis is confirmed unless there is neurologic compromise and/or sepsis present; in patients with negative culture results, empiric treatment is warranted based on the most likely organisms to cause infection. Treatment is usually for a minimum duration of 6 weeks, but the duration should be based on the individual patient’s circumstances. In 2015, the Infectious Diseases Society of America (IDSA) issued new guidelines on vertebral osteomyelitis, and these should be used for treatment and management.
A bone biopsy remains the gold standard when diagnosing osteomyelitis. However, there is a high false-negative rate, and the negative predictive value is close to 65%, mainly owing to the organism’s patchy distribution. No specific laboratory test can be recommended. However, if osteomyelitis is suspected based on clinical history and physical findings, plain radiographs of the involved site should be ordered along with laboratory testing (acute- phase reactants, e.g., erythrocyte sedimentation rate [ESR], C-reactive protein [CRP], leukocyte count with a differential count). These labs can strongly suggest (positive predictive value of 100%) osteomyelitis if the ESR value is greater than 70 mm/h in the absence of an inflamed ulcer. However, these tests lack specificity, and it may take several days to demonstrate significantly elevated levels. Bone biopsy (open being preferred over a percutaneous approach) should be performed for pathogen identification and in cases when the diagnosis of osteomyelitis is uncertain but suspected. All patients should have blood cultures performed. A positive blood culture with a suspicious physical finding can suggest a bone infection, but only one half of the cases have a positive test result. Blood cultures, like bone biopsies, can be affected by recent antibiotic exposure. However, patients receiving antibiotics prior to the bone biopsy still may have a reasonable microbiologic yield because areas of infection also have associated bony infarction or ischemia.
There is no one imaging modality routinely recommended to diagnose osteomyelitis, so diagnosis often requires more than one technique. Plain film radiography should always be the initial study, and the result can be diagnostic if positive. However, changes (usually along the metaphysis) typically require at least 1 to 2 weeks to be seen radiographically.
Positron emission tomography (PET) is the most specific and sensitive of imaging techniques, but its high cost and lack of availability make it impractical for most clinicians. The best choice for imaging depends on the age of the patient, duration of symptoms, suspected location of infection (if known), and concurrent or previous medical conditions.
Plain radiographs are the most available, least expensive, and easiest to obtain, but they lack sensitivity (43%–75%), and a negative result does not exclude the diagnosis. However, they have reasonable specificity (75%–83%), and a positive finding can confirm the diagnosis or provide clues to alternative pathology, such as a tumor. Bony changes can take between 10 and 21 days to become visible on plain films; however, soft tissue changes can be seen in as little as 3 days. The soft tissue changes are especially important in neonates and children because focal soft tissue swelling around the bony metaphysis may be the first sign of bone involvement.
Radionuclide imaging (i.e., triple-phase bone scan, leukocyte scintigraphy, and PET) is a preferred method of advanced imaging, and it has several advantages compared with other techniques. Young children often complete the examination without sedation, and prosthetic joints do not produce the artifact commonly seen on MRI and computed tomography (CT) scans. Positive results can be seen 24 to 48 hours after onset of symptoms, and a negative examination result effectively rules out osteomyelitis.
The triple-phase bone scan (technetium 99 m diphosphonate) is often the examination of choice (sensitivity of 73%–100%), and it can distinguish between cellulitis and osteomyelitis when complications are absent. However, the sensitivity decreases dramatically when other conditions are present (i.e., trauma, diabetes, or recent surgery), and it has been reported to be as low as 38%. Bone scans usually lack the specificity (25%–90%) of other modalities, fail to provide detailed pictures of complex anatomy, and can be influenced by poor circulation. The examination may take up to 24 hours to complete and often requires the patient to make many trips to the facility. In the early phase, uptake is greatest in areas of acute inflammation. In the next phase, uptake occurs in areas of soft tissue inflammation, and in the late (delayed) phase, uptake remains in the presence of osteomyelitis.
Leukocyte scintigraphy using gallium 67 has a higher specificity (80%–90%) than triple-phase scanning (67%) in the peripheral skeleton, but it decreases to 25% when looking at the axial skeleton. Leukocyte scintigraphy is the preferred method when evaluating patients with previous joint replacements, diabetes, or trauma.
MRI can detect acute osteomyelitis as early as 3 days. It is nearly as sensitive (82%–100%) and specific (75%–96%) as radionuclide studies. MRI allows tracking of disease progress and response to treatment.
MRI can be used to date osteomyelitis. Some patients with osteomyelitis are treated and later develop another episode. The MRI can distinguish whether the second episode is a new infection or a recurrence of the previous infection. It also provides detailed visualization of complex anatomy and critical structures, allowing surgeons to map any planned surgical intervention.
CT is rarely used for osteomyelitis, except when sequestered bone is suspected or for interventional procedures. Sinography can be used to map sinus tracts with fluoroscopy or combined with CT. Ultrasound is sometimes used in children and can be helpful in differentiating acute from chronic infections. It provides guidance during drainage, aspirations, or biopsies of the affected bone, and it is a noninvasive method to monitor soft tissue involvement in chronic illness.
The Cierny-Mader classification system provides a straightforward algorithm for treatment. However, surgical débridement of necrotic tissue and proper antimicrobial therapy according to the identification of the causative organism and its susceptibilities remain the most important factors in any treatment.
Unlike chronic infections, acute infections require hospitalization for initiation of therapy and supportive care. Serial examinations should be undertaken to assess the success of treatment and monitor for systemic signs or symptoms. Cultures of blood and bone should be obtained to guide therapy. Laboratory studies can be followed, but other than CRP levels, they fail to provide significant data. The CRP value can be expected to decrease 24 to 48 hours after initiation of appropriate antibiotic therapy. A lack of response may indicate inappropriate therapy or an occult abscess, and the physician should reconsider surgery if previously delayed.
The medical literature remains inconclusive about the antibiotic treatment of osteomyelitis, especially when trying to determine the best agents, route, or duration of antibiotic therapy.
Initial treatment should cover for methicillin-resistant (MRSA) and methicillin-sensitive S. aureus (MSSA) until cultures are completed with susceptibilities.
Although the optimal duration of antibiotic therapy remains undefined, most authorities recommend treatment for about 6 weeks (Table 2). After the infection is under control, the physician may switch the patient to an oral antibiotic for 3 to 12 months (i.e., a fluoroquinolone with or without rifampin [Rifadin]1). However, treatment can be as short as 3 weeks for uncomplicated, acute, hematogenous osteomyelitis. Management can include oral preparations after a short parenteral course, provided the drug has high bioavailability and the organism is susceptible. A microbiologic diagnosis (preferably by bone biopsy) is essential so that the choice of antibiotic accounts for the specific organism, the host status, and the least toxic medication for the individual. Hematogenous osteomyelitis is usually monomicrobial, whereas contiguous infections are usually polymicrobial and may include Pseudomonas in certain populations.
For this population, fluoroquinolones are an excellent choice.
Antibiotics for Osteomyelitis
|Organism||Preferred Drug||Alternative Drugs|
|Staphylococcus aureus||Nafcillin (Unipen) 1–2 g IV or IM q4h||Cefazolin, vancomycin, clindamycin|
|Methicillin-resistant S. aureus||Vancomycin (Vancocin) 1 g q8h||Trimethoprim-sulfamethoxazole (Bactrim)* plus rifampin (Rifadin)*|
|Streptococcus pneumoniae, group A β-hemolytic streptococci||Penicillin G (Pfizerpen)* 2 million units IV q4h||Cefazolin, vancomycin, clindamycin|
|Enterococci, Haemophilus influenzae β-lactamase negative||Cefotaxime (Claforan) 2 g q6h||Trimethoprim-sulfamethoxazole, ceftriaxone|
|H. influenzae β-lactamase positive, Klebsiella pneumoniae||Ceftriaxone (Rocephin) 2 g q24h||Trimethoprim-sulfamethoxazole, ciprofloxacin, piperacillin (Pipracil), imipenem (Primaxin)|
|Escherichia coli||Cefazolin (Ancef) 2 g q8h||Ciprofloxacin, ceftriaxone, imipenem|
|Pseudomonas aeruginosa||Ciprofloxacin (Cipro) 400 mg q12h||Piperacillin plus aminoglycoside, aztreonam (Azactam)*|
|Salmonella||Choose ampicillin, ceftriaxone, imipenem (Primaxin), or ciprofloxacin, depending on sensitivities|
|Bacteroides spp.||Clindamycin (Cleocin) 600 mg q6h||Imipenem, metronidazole (Flagyl)|
|Serratia marcescens||Ceftriaxone (Rocephin) 2 g q24h||Imipenem, trimethoprim-sulfamethoxazole, ciprofloxacin|
(Modified from Cohen J, Powderly WG, editors. Infectious Diseases. 2nd ed. St Louis: Mosby; 2003.)
* Not FDA approved for this indication.
For MSSA, nafcillin (Unipen) or a first- or second-generation cephalosporin, cefazolin, can be implemented. For MRSA, vancomycin (Vancocin) is recommended. For an anaerobic infection, clindamycin (Cleocin) is a good choice.
Bones can heal in the presence of active infection. However, in the presence of obvious signs of infection, such as an abscess or Cierny- Mader stage 3 or 4 disease, acute surgical débridement and irrigation are warranted. The goals of surgery include drainage, débridement, and stabilization. After successful débridement and stabilization, antibiotic therapy is initiated and continued until adequate healing has occurred, usually in 6 weeks. If débridement is unsuccessful, inert substances must be completely removed and tissue débrided.
Antibiotics should be placed in contact with the bone using a polymethylmethacrylate antibiotic (PMMA) bead chain or other biodegradable delivery system to achieve higher local antibiotic concentrations. The site needs to be stabilized with an external fixator, and staged reconstruction should be initiated.
1. Berbari E.F., Kanj S.S., Kowalski T.J., et al. 2015 Infectious Diseases Society of America (IDSA) clinical practice guidelines for the diagnosis and treatment of native vertebral osteomyelitis in adults. Clin Infect Dis. 2015;61:e26.
2. Berendt A., Norden C. Acute and chronic osteomyelitis. In: Cohen J., Powderly W.G., eds. Infectious Diseases. 2nd ed. St Louis: Mosby; 2003.
3. Cierny III G., Mader J.T., Pennick J.J. A clinical staging system for adult osteomyelitis. Contemp Orthop. 1985;10:17–37.
4. Kaplan S.L. Osteomyelitis in children. Infect Dis Clin North Am. 2005;19:787–797.
5. Kremers H.M., Nwojo M.E., Ransom J.E., et al. Trends in the epidemiology of osteomyelitis: A population-based study, 1969 to 2009. J Bone Joint Surg. 2015;97(10):837–845.
6. Krogstad P. Osteomyelitis and septic arthritis. In: Feigin R.D., Cherry J.D., Demmler G.J., et al., eds. Textbook of Pediatric Infectious Diseases. 5th ed Philadelphia: Saunders; 2004:713– 736.
7. Lampe R.M. Osteomyelitis. In: Behrman R.E., Kliegman R.M., Jenson H.B., Stanton B.F., eds. Nelson Textbook of Pediatrics. 18th ed. Philadelphia: Saunders; 2007.
8. Lazzarini L., Lipsky B.A., Mader J.T. Antibiotic treatment of osteomyelitis: What have we learned from 30 years of clinical trials? Int J Infect Dis. 2005;9:127–138.
9. Lipsky B., Weigelt J., Gupta V., et al. Skin, soft tissue, bone, and joint infections in hospitalized patients: Epidemiology and microbiological, clinical, and economic outcomes. Infect Control Hosp Epidemiol. 2007;28:1290–1298.
10. Pineda C., Vargas A., Rodriguez A. Imaging of osteomyelitis: Current concepts. Infect Dis Clin North Am. 2006;20:789–825.
11. Waldvogel F.A., Medoff G., Swartz M.N. Osteomyelitis: A review of clinical features, therapeutic considerations and unusual aspects. N Engl J Med. 1970;282:198–206.
12. White L.M., Schweitzer M.E., Deely D.M., Gannon F. Study of osteomyelitis: Utility of combined histologic and microbiologic evaluation of percutaneous biopsy samples. Radiology. 1995;197:840–842.
13. Ziran B.H. Osteomyelitis. J Trauma. 2007;62(Suppl):S59–S60.
1 Not FDA approved for this indication.