Treatment of osteomyelitis due to nontuberculous mycobacteria in adults

INTRODUCTION — Nontuberculous mycobacteria (NTM) are a large group of organisms that are widespread in the environment (table 1). They have been isolated from numerous environmental sources, including water and soil. NTM can cause a broad range of infections that vary depending on the NTM species and on the host's immune status. In immunocompetent individuals, disease can present as chronic pneumonia, lymphadenitis, or skin, soft tissue, and/or bone infection. Immunocompromised individuals can also present with any of these findings, but disease in such patients may also manifest as disseminated infection. Unfortunately, NTM are seldom considered as a possible etiology in infections of the soft tissues and/or bones; therefore, delays in diagnosis and treatment are common.

The treatment of osteomyelitis due to NTM will be reviewed here. The epidemiology, clinical manifestations, and diagnosis of NTM osteomyelitis are discussed separately. (See "Epidemiology, clinical manifestations, and diagnosis of osteomyelitis due to nontuberculous mycobacteria".)

Other NTM infections and the management of osteomyelitis due to bacteria and Mycobacterium tuberculosis are reviewed elsewhere. (See "Epidemiology of nontuberculous mycobacterial infections" and "Microbiology of nontuberculous mycobacteria" and "Overview of nontuberculous mycobacteria (excluding MAC) in patients with HIV" and "Mycobacterium avium complex (MAC) infections in persons with HIV" and "Treatment of Mycobacterium avium complex pulmonary infection in adults" and "Rapidly growing mycobacterial infections: Mycobacteria abscessus, chelonae, and fortuitum" and "Nonvertebral osteomyelitis in adults: Clinical manifestations and diagnosis" and "Vertebral osteomyelitis and discitis in adults" and "Osteomyelitis associated with open fractures in adults" and "Hematogenous osteomyelitis in children: Management" and "Bone and joint tuberculosis".)

GENERAL PRINCIPLES

●Importance of a multidisciplinary approach – Management of osteomyelitis caused by NTM includes combination antimycobacterial therapy, often with surgical debridement or excision. For patients on immunosuppressive agents, these should be minimized whenever possible. Collaboration between all medical and surgical providers is essential because of the difficult-to-treat nature of NTM.

●Caveats to NTM isolation and susceptibility testing – Unlike isolation of M. tuberculosis, isolation of an NTM does not necessarily dictate initiation of therapy. Because NTM can be isolated as a result of contamination of a clinical specimen and because the pathogenicity of the various NTM species varies substantially, clinicians must determine the clinical significance of any isolate and whether treatment is indicated [1]. In addition, the results of in vitro antimicrobial susceptibility testing for many NTM do not correlate with clinical response, and therefore the clinician should use these data with an appreciation for the limitations of such testing. (See "Epidemiology, clinical manifestations, and diagnosis of osteomyelitis due to nontuberculous mycobacteria", section on 'Diagnosis'.)

●Limitations in supporting evidence – There are no randomized controlled trials or comparative observational studies that have evaluated the efficacy of treatment for NTM osteomyelitis. Thus, treatment recommendations are based upon case reports and small retrospective reviews, as well as on in vitro antimicrobial susceptibility results. Our recommendations for therapy are generally in keeping with the American Thoracic Society/Infectious Diseases Society of America guidelines on the management of NTM diseases [1,2], as well as our own clinical experience.

These guidelines, as well as other society guidelines, can be found through the society guideline link below. (See 'Society guideline links' below.)

ANTIMICROBIAL SUSCEPTIBILITY TESTING — In vitro antimicrobial susceptibility testing is recommended for all clinically significant NTM isolated from the lesions of patients with osteomyelitis [1].

Specific recommendations differ for each species, as detailed below. The Clinical and Laboratory Standards Institute recommends specific drugs and testing conditions for the most common species [3,4].

For most NTM species, there is no known correlation between in vitro susceptibility results and the clinical response to therapy. In vitro susceptibility testing has not been validated for all NTM species. Despite these limitations, susceptibility testing plays an important role in the management of NTM osteomyelitis. Recommendations for susceptibility testing of each NTM species are discussed below. (See 'Antimycobacterial therapy' below.)

Susceptibility testing for NTM is also discussed separately. (See "Microbiology of nontuberculous mycobacteria".)

ANTIMYCOBACTERIAL THERAPY — The specific antimycobacterial regimen depends upon the NTM species isolated and, for certain species, on the results of in vitro antimicrobial susceptibility testing. It is important not to use monotherapy, which can lead to the development of resistance.

It is also crucial to design a regimen based on the specific NTM isolated, since there are differences in antimicrobial susceptibility among the species. Specific regimens for each NTM species are discussed below.

Rapidly growing mycobacteria — Patients with osteomyelitis caused by rapidly growing mycobacterial species should receive combination antimycobacterial therapy, typically in conjunction with surgery. The rapidly growing mycobacteria tend to be more difficult to treat than slowly growing mycobacteria due to their inherent drug resistance. For this reason, surgical resection is critical for disease control and eradication. In rare cases, amputation is required for cure [5,6]. (See 'Surgery' below.)

Mycobacterium fortuitum, Mycobacterium chelonae, and Mycobacterium abscessus (including the subspecies M. abscessus, M. massiliense, and M. bolletii) are resistant to antituberculosis agents (eg, isoniazid, ethambutol, and rifampin), but they are typically susceptible to various traditional antibacterial agents [1]:

●Most rapidly growing mycobacteria are susceptible in vitro to amikacin, clofazimine, carbapenems (imipenem, meropenem), oxazolidinones (linezolid, tedizolid), bedaquiline, and newer cycline derivatives (tigecycline, eravacycline, omadacycline) [7-10]. Imipenem is the most active carbapenem against rapidly growing mycobacteria [9]. Almost all rapidly growing mycobacteria are susceptible to tigecycline and other newer cycline derivatives. Omadacycline has excellent in vitro activity against rapidly growing mycobacteria, with minimum inhibitory concentration (MIC) 90 values of 0.25 mcg/mL, 0.5 mcg/mL, and 0.25 mcg/mL for M. abscessus, M. chelonae, and M. fortuitum, respectively [11].

●Susceptibility to macrolides depends on the presence and activity of the erm gene, which confers inducible resistance to macrolides. (See "Rapidly growing mycobacterial infections: Mycobacteria abscessus, chelonae, and fortuitum", section on 'Detecting macrolide resistance'.)

●Susceptibility to fluoroquinolones (ciprofloxacin, moxifloxacin), cefoxitin, doxycycline, minocycline, and trimethoprim/sulfamethoxazole is variable [7-10]. Most strains of M. fortuitum and M. chelonae are susceptible to moxifloxacin, but M. abscessus is almost universally resistant [7].

●Susceptibility to aminoglycosides varies between species. Amikacin is generally the most active aminoglycoside; the exception is M. chelonae, against which tobramycin is more active.

In vitro antimicrobial susceptibility testing should be performed on all clinically significant isolates of rapidly growing mycobacteria (eg, M. fortuitum complex, M. chelonae, M. abscessus), including isolates that have been recovered after treatment failure or relapse [1,3,4]. The Clinical and Laboratory Standards Institute recommends that the following antimicrobial drugs be tested for in vitro susceptibility: clarithromycin, amikacin, tobramycin, imipenem, meropenem, cefoxitin, ciprofloxacin, moxifloxacin, trimethoprim/sulfamethoxazole, doxycycline, tigecycline, linezolid, and clofazimine. Moxifloxacin susceptibility testing is necessary even when ciprofloxacin is tested, since ciprofloxacin susceptibility cannot be extrapolated to moxifloxacin. (See "Rapidly growing mycobacterial infections: Mycobacteria abscessus, chelonae, and fortuitum", section on 'Susceptibility testing'.)

M. fortuitum complex — M. fortuitum complex includes at least 12 species, of which M. fortuitum is the most common [12]. For initial therapy, we suggest a combination of ≥3 agents, including one to two oral agents and one to two parenteral agents to which the isolate is susceptible, adjusted for renal dysfunction [1]. The final number of antimicrobial agents used in the regimen will depend on the in vitro susceptibility pattern, which drugs are available, the site, and the extent of disease. It is important to note that monotherapy is not considered appropriate, given the risk of inducing resistance.

●Amikacin 15 mg/kg intravenously (IV) once daily or 15 to 25 mg/kg IV three times weekly

PLUS

●One of the following IV agents:

•Imipenem 750 to 1000 mg IV every 12 hours

•Cefoxitin 3 g IV every 8 to 12 hours

PLUS

●One to two of the following oral agents:

•Moxifloxacin 400 mg once daily (or ciprofloxacin 500 to 750 mg twice daily)

•Doxycycline 100 mg twice daily (or omadacycline 300 mg daily)

•Clofazimine 100 mg daily (can be obtained through an expanded-access program or single-patient investigational new drug application)

•Linezolid 600 mg daily (or tedizolid 200 mg daily)

•Trimethoprim-sulfamethoxazole (3.5 to 4 mg/kg of the trimethoprim component every 8 to 12 hours)

These doses are for patients with normal renal function.

The duration of parenteral therapy is not well defined. We suggest continuing this initial parenteral combination regimen for six to eight weeks and then switching to an oral regimen that includes at least two agents to which the isolate is susceptible in vitro. Total duration of antimycobacterial therapy is discussed elsewhere. (See 'Duration of therapy' below.)

Oral agents that typically have in vitro activity against M. fortuitum complex include trimethoprim/sulfamethoxazole, ciprofloxacin and moxifloxacin, azithromycin and clarithromycin, and linezolid and clofazimine [12]. However, most M. fortuitum, Mycobacterium porcinum, and Mycobacterium septicum isolates carry an inducible erythromycin methylase gene, erm(39), which confers inducible macrolide resistance; macrolides should be avoided if the erm gene is present [12,13]. In contrast, approximately two-thirds of Mycobacterium peregrinum isolates are susceptible to macrolides. (See "Rapidly growing mycobacterial infections: Mycobacteria abscessus, chelonae, and fortuitum", section on 'Detecting macrolide resistance'.)

IV agents with in vitro activity include amikacin, imipenem, and tigecycline [1,14-17]. Imipenem is the preferred carbapenem, although meropenem may also be used; ertapenem and doripenem have limited activity. Many isolates are susceptible to cycline derivatives (doxycycline, omadacycline, eravacycline, tigecycline) and cefoxitin.

No randomized trials or comparative observational studies have been performed to inform treatment outcomes. In one study that included 76 patients with extrapulmonary infections due to M. fortuitum, 68 patients (90 percent) were successfully treated with antimycobacterial therapy with or without surgery [18]. Most patients had skin and soft tissue infections, but 15 had osteomyelitis. The majority of patients with extensive disease underwent surgical debridement in combination with antimycobacterial therapy. Patients received a variety of regimens, ranging from combination parenteral therapy with amikacin plus cefoxitin to oral monotherapy with a sulfonamide, doxycycline, or amikacin. However, for osteomyelitis, we do not recommend monotherapy. The mean duration of antimycobacterial therapy for patients with osteomyelitis was seven months.

M. chelonae — Treatment of M. chelonae disease is difficult due to multidrug resistance and the dearth of oral antibiotics for therapy.

For initial therapy, we suggest a combination of ≥3 agents, including a macrolide (preferably azithromycin), another one to two oral agents, and one to two parenteral agents to which the isolate is susceptible [1].

●One to two of the following IV agents:

•Tobramycin 4 to 7 mg/kg IV once daily or three times weekly

•Imipenem 750 to 1000 mg IV every 12 hours

•Tigecycline 50 mg IV once daily (or every 12 hours if tolerated)

PLUS

●Azithromycin 250 to 500 mg once daily (or clarithromycin 500 mg twice daily)

PLUS

●One to two of the following oral agents:

•Moxifloxacin 400 mg once daily (or ciprofloxacin 500 to 750 mg twice daily)

•Doxycycline 100 mg twice daily (or omadacycline 300 mg daily)

•Clofazimine 100 mg daily (can be obtained through an expanded-access program or single-patient investigational new drug application)

•Linezolid 600 mg daily (or tedizolid 200 mg daily)

•Trimethoprim-sulfamethoxazole 3.5 to 4 mg/kg of the trimethoprim component every 8 to 12 hours

These doses are for patients with normal renal function.

The duration of parenteral therapy is not well defined. We usually continue the initial parenteral combination regimen for 8 to 12 weeks until there are signs of clinical improvement, and we then switch to an oral regimen that includes two to three agents to which the isolate is susceptible. Total duration of antimycobacterial therapy is discussed elsewhere. (See 'Duration of therapy' below.)

The macrolides omadacycline and clofazimine are the only oral agents to which M. chelonae is typically susceptible [1,11,16,19,20]. Up to 20 percent of M. chelonae isolates from one case series were susceptible to doxycycline, ciprofloxacin, and sulfamethoxazole [19]. Although prior evidence suggested reasonable susceptibility to fluoroquinolones and linezolid, in a subsequent in vitro assessment of 526 strains of M. chelonae, most strains were resistant to these agents [21].

Based on in vitro susceptibility patterns, the most promising parenteral agents are tobramycin [15], imipenem [17], tigecycline, eravacycline [11,22], and omadacycline [11]. Tobramycin has greater activity against M. chelonae than amikacin, in contrast to other rapidly growing mycobacteria [15]. Cefoxitin typically does not have activity against M. chelonae [1].

In an observational study that included 47 patients with nonpulmonary infections due to M. chelonae or M. abscessus, 34 patients (72 percent) were successfully treated [18]. The outcomes for these two species were combined, so it is not possible to determine the outcomes of those infected with M. chelonae alone. Only seven patients had osteomyelitis.

M. abscessus — M. abscessus (and its subspecies, M. abscessus, M. massiliense, and M. bolletii [23-26]) is challenging to treat because it is resistant to many agents that are used to treat other NTM infections [1]. We suggest a combined medical and surgical approach for treatment of M. abscessus osteomyelitis. We also recommend subspecies identification and in vitro susceptibility testing, including assessment for inducible macrolide resistance in all clinically significant isolates. (See "Rapidly growing mycobacterial infections: Mycobacteria abscessus, chelonae, and fortuitum", section on 'Nomenclature for M. abscessus'.)

Regimen selection — Regimen selection depends on the presence or activity of the erm gene, which confers inducible macrolide resistance despite in vitro susceptibility at three to five days. The majority of subspecies M. abscessus and M. bolletii contain an erm(41) gene. However, there are some M. abscessus isolates that have a T-to-C polymorphism at nucleotide 28 of the erm(41) gene, which inactivates the gene, and these isolates are susceptible to macrolides [27]. Similarly, the erm(41) gene in M. massiliense is nonfunctional due to truncation, so inducible resistance does not occur [28]. (See "Rapidly growing mycobacterial infections: Mycobacteria abscessus, chelonae, and fortuitum", section on 'Detecting macrolide resistance'.)

●M. massiliense or M. abscessus with a nonfunctional erm gene (C28 sequevar) – For initial therapy of these subspecies, we suggest a combination of at least 3 agents, including at least two oral agents (including a macrolide) and one to two parenteral agents to which the isolate is susceptible. All of the following doses are for patients with normal renal function.

•Amikacin 10 to 15 mg/kg IV once daily or 15 to 25 mg/kg IV three times weekly

PLUS

•One of the following IV agents:

-Imipenem 750 to 1000 mg IV every 12 hours (preferred)

-Cefoxitin 3 g IV every 8 to 12 hours

-Tigecycline 50 mg IV once daily or every 12 hours

-Omadacycline IV 100 mg once daily

-Eravacycline 1 mg/kg IV every 12 hours

PLUS

•Azithromycin 250 to 500 mg once daily (or clarithromycin 500 mg twice daily)

PLUS

•At least one of the following oral agents:

-Clofazimine 100 mg once daily (can be obtained through an expanded-access program or single-patient investigational new drug application)

-Omadacycline 300 mg once daily (if not receiving an IV tetracycline derivative)

-Linezolid 600 mg once daily (or tedizolid 200 mg once daily)

The duration of parenteral therapy is not well defined. We suggest continuing one of the regimens outlined above for at least 8 to 12 weeks and then switching to an oral regimen. Factors that influence the duration of parenteral therapy include extent of disease, host immunosuppression, susceptibility of the organism, and whether surgical debridement is performed. Total duration of antimycobacterial therapy is discussed elsewhere. (See 'Duration of therapy' below.)

●M. abscessus with a functional erm gene (T28 sequevar) or M. bolletii – For initial therapy of these subspecies, we suggest a combination of at least four agents (not including a macrolide), including two parenteral agents to which the isolate is susceptible. All of the following doses are for patients with normal renal function.

•Amikacin 10 to 15 mg/kg IV once daily or 25 mg/kg IV three times weekly

PLUS

•At least one of the following IV agents:

-Imipenem 750 to 1000 mg IV every 12 hours (preferred)

-Cefoxitin 3 g IV every 8 to 12 hours

-Tigecycline 50 mg IV once daily or every 12 hours

-Omadacycline IV 100 mg once daily

-Eravacycline 1 mg/kg IV every 12 hours

PLUS

•Two of the following oral agents:

-Clofazimine 100 mg once daily (can be obtained through an expanded-access program or single-patient investigational new drug application)

-Omadacycline 300 mg once daily (if not receiving an IV tetracycline derivative)

-Linezolid 600 mg once daily (or tedizolid 200 mg once daily)

-Bedaquiline 400 mg daily for two weeks, then 200 mg three times a week

The duration of parenteral therapy is not well defined. We suggest continuing one of the regimens outlined above for at least 12 weeks and then switching to an oral regimen if possible. Some patients will need to continue parenteral therapy throughout the course of treatment if the isolate does not have at least two companion oral agents with in vitro activity. Total duration of antimycobacterial therapy is discussed elsewhere. (See 'Duration of therapy' below.)

In vitro activity and efficacy of options — Parenteral agents that typically have in vitro activity against M. abscessus include amikacin, cefoxitin, imipenem, meropenem, tigecycline, omadacycline, and eravacycline. Imipenem is the preferred carbapenem, although meropenem may also be used; ertapenem and doripenem have limited activity [9]. We prefer imipenem to cefoxitin, as it has been shown to be superior to cefoxitin in time-kill and macrophage assays and is better tolerated [29]. We also prefer imipenem to tigecycline, as the optimal dose for tigecycline in a hollow-fiber model was 200 mg per day, which would be toxic to patients [30]. In fact, most patients require a reduced dose of 25 to 50 mg per day in order to tolerate the medication [31,32]. If we use tigecycline, we typically start with 50 mg IV every 12 hours but often have to reduce the dose to once daily in patients who have substantial nausea and vomiting. Pretreatment with antiemetic therapy can be used to facilitate twice-daily dosing, but it may not be adequate in many patients [32].

The only oral agent with reliable activity against subspecies M. abscessus is clofazimine [33-36]. Other oral agents with possible activity include macrolides (azithromycin, clarithromycin), fluoroquinolones (ciprofloxacin, moxifloxacin), oxazolidinones (linezolid, tedizolid), bedaquiline, and omadacycline. One study of M. abscessus isolates noted that clarithromycin induces greater erm(41) expression and thus higher macrolide resistance than azithromycin [37]. Therefore, azithromycin may be more effective in M. abscessus infections.

Bedaquiline is an oral diarylquinoline with activity against M. tuberculosis. The MIC90 against M. abscessus has been reported to be 0.125 mcg/mL [10,12]; however, the drug is bacteriostatic. In a mouse model evaluating bedaquiline with other compounds, use with clofazimine showed improved activity against M. abscessus [38], but there are cardiovascular safety and dual cross-resistance concerns with the combination. Clinical data on bedaquiline for M. abscessus are limited [39].

Tedizolid is an oxazolidinone with MICs 4- to 16-fold lower than those of linezolid against M. abscessus [40,41]. Like linezolid, it is bacteriostatic with some increased activity when given with imipenem and amikacin in vitro [41-44]. There are no clinical case series describing its effectiveness.

Omadacycline is a novel aminomethylcycline that comes in both oral and IV formulations. Limited clinical data have described omadacycline use in extrapulmonary M. abscessus infections [45,46]. In vitro data have demonstrated activity similar to tigecycline but with better tolerance and pharmacokinetic parameters [47-51].

M. abscessus has a beta-lactamase called Bla_Mab, which efficiently hydrolyzes beta-lactams [52]. Adding clavulanate, sulbactam, or tazobactam to beta-lactams does not improve activity, but adding avibactam, relebactam, or vaborbactam can increase activity of some beta-lactam antibiotics [53]. For example, addition of avibactam significantly improves in vitro activity of ceftaroline but not imipenem or cefoxitin [54]. Furthermore, combinations of carbapenems and other beta-lactams such as ceftaroline or ceftazidime result in significant decreases in MICs against M. abscessus [55]. Dual beta-lactam agents with or without beta-lactamase inhibitors are beginning to be used clinically, but there are no clinical series or trials yet available.

The evidence for efficacy of treatment regimens for M. abscessus comes primarily from retrospective studies of M. abscessus pulmonary infections [56,57]. One study of nonpulmonary infections caused by rapidly growing mycobacteria included 47 cases of M. chelonae, 32 of which were further classified as being M. chelonae subspecies abscessus (now called M. abscessus) [18]. The overall success rate for infections caused by M. abscessus or M. chelonae was 72 percent, but the major cause of treatment failure was the absence of an oral drug with activity against M. abscessus. Another study of 20 extrapulmonary infections caused by subspecies M. abscessus and M. massiliense included six patients with bone and joint disease, all of whom received IV and oral therapy for a median of 23 and 343 days, respectively, in addition to surgery [58]. Treatment success rates were higher in patients infected with either M. massiliense or M. abscessus with a nonfunctional erm gene (the C28 sequevar) than in those infected with M. abscessus with a functional erm gene (the T28 sequevar).

Slowly growing mycobacteria

M. avium complex — For patients with localized osteomyelitis due to Mycobacterium avium complex (MAC), we suggest a combination of excisional surgery or surgical debridement and antimycobacterial therapy [1]. For MAC isolates, antimicrobial susceptibility testing should be performed for clarithromycin and amikacin [3,4]. (See 'Surgery' below.)

We suggest that patients receive the following regimen, adjusted for renal dysfunction [1]:

●Azithromycin 250 to 500 mg once daily or clarithromycin 500 mg twice daily.

PLUS

●Ethambutol 15 mg/kg orally daily.

PLUS

●Rifampin 10 mg/kg once daily (maximum daily dose 600 mg); in patients with human immunodeficiency virus (HIV) receiving antiretroviral therapy with an agent that has strong interactions with rifampin, rifabutin 150 to 300 mg once daily can be used. (See "Treatment of drug-susceptible pulmonary tuberculosis in nonpregnant adults with HIV infection: Initiation of therapy", section on 'Drug interactions'.)

IV amikacin or intramuscular streptomycin (each dosed at 10 to 15 mg/kg daily or 15 to 25 mg/kg three times weekly) should be added for severe or disseminated infection or in patients who respond poorly to the regimen outlined above. When amikacin or streptomycin is used, it is typically continued for the first two or three months. Total duration of antimycobacterial therapy is discussed elsewhere. (See 'Duration of therapy' below.)

Immunocompromised patients generally require a more complex regimen to treat disseminated disease. Immunocompromised patients also require a longer duration of therapy. These issues are discussed in greater detail separately. (See "Mycobacterium avium complex (MAC) infections in persons with HIV" and "Nontuberculous mycobacterial infections in solid organ transplant candidates and recipients", section on 'Treatment'.)

As above, clarithromycin and amikacin susceptibility testing should be performed for new (previously untreated) isolates; clarithromycin susceptibility should also be tested in isolates from patients who have not responded to macrolide therapy after several months. Since clarithromycin and azithromycin share cross-resistance and susceptibility, the results of clarithromycin susceptibility testing can be extrapolated to azithromycin [1,3,4]. In patients who fail initial macrolide-based therapy, we also suggest testing susceptibility to ethambutol, rifampin, rifabutin, and clofazimine. Susceptibility testing should also be considered for moxifloxacin and linezolid.

Bedaquiline has even better activity against MAC than M. abscessus, but there remains a very limited amount of clinical data regarding its efficacy in the treatment of patients with MAC disease [12,39].

M. kansasii — In patients with Mycobacterium kansasii osteomyelitis, antimycobacterial therapy with a combined regimen similar to a regimen used for tuberculosis (isoniazid, rifampin, and ethambutol) has been successful even without surgical debridement [59-62]. M. kansasii is highly susceptible to macrolides, and treatments with macrolide-based regimens that include rifampin and ethambutol have been shown to be effective in patients with pulmonary disease [63,64]. For the initial treatment of M. kansasii, antimicrobial susceptibility testing should be performed for rifampin and clarithromycin [3,4].

We suggest the following oral combination regimen for patients with M. kansasii osteomyelitis, adjusted for renal dysfunction [1]:

●Azithromycin 250 to 500 mg once daily or clarithromycin 500 mg twice daily

PLUS

●Rifampin 10 mg/kg once daily (maximum daily dose 600 mg)

PLUS

●Ethambutol 15 mg/kg per day

We recommend continuing antimycobacterial therapy for at least six months.

Rifamycins play a critically important role in the treatment of M. kansasii disease; therefore, it is important to use a rifamycin-based regimen even in the setting of concomitant antiretroviral therapy for HIV infection, despite the drug-drug interactions [1]. This recommendation is based upon the observation that substantial rates of relapse occurred in the pre-rifampin era and that outcomes improved dramatically after the introduction of rifampin for the treatment of pulmonary M. kansasii infections [1,65-68]. The use of rifamycins in patients receiving antiretroviral therapy is discussed in detail separately. (See "Treatment of drug-susceptible pulmonary tuberculosis in nonpregnant adults with HIV infection: Initiation of therapy", section on 'Drug interactions'.)

Isolates that demonstrate in vitro susceptibility to rifampin are also susceptible to rifabutin. Although isoniazid and ethambutol are used in the treatment of M. kansasii, MIC values correlate poorly with clinical response, and testing is not recommended [3,4]. If the isolate is rifampin resistant, susceptibility testing to the following agents is recommended: amikacin, ciprofloxacin, moxifloxacin, doxycycline, minocycline, rifabutin, and trimethoprim/sulfamethoxazole. Moxifloxacin susceptibility testing is necessary even when ciprofloxacin is tested, since ciprofloxacin susceptibility cannot be extrapolated to moxifloxacin. Because of its excellent in vitro activity against M. kansasii, moxifloxacin is a potential substitute for rifampin in resistant or intolerant cases.

M. haemophilum — Based on reported cases, combinations of antibiotics, including clarithromycin, ciprofloxacin, and a rifamycin, appear to be effective in the treatment of osteomyelitis caused by Mycobacterium haemophilum [1,69,70]. In addition to these agents, amikacin also appears to have in vitro activity against M. haemophilum [1,70,71]. In contrast, doxycycline and sulfonamides have variable activity against M. haemophilum, and all clinical isolates are resistant to ethambutol [72]. There are no standardized in vitro susceptibility methods for M. haemophilum; therefore, when in vitro antimicrobial susceptibility testing is performed, the results must be interpreted with caution [1].

For the treatment of M. haemophilum, we suggest the following oral combination regimen, adjusted for renal dysfunction [69,70]:

●Azithromycin 250 to 500 mg once daily or clarithromycin 500 mg twice daily

PLUS

●Ciprofloxacin 500 mg twice daily or moxifloxacin 400 mg daily

PLUS

●Rifampin 10 mg/kg once daily (maximum daily dose 600 mg) or rifabutin 150 to 300 mg daily

In patients with extensive or severe disease, we add amikacin 15 mg/kg IV once daily or 15 to 25 mg/kg three times weekly to the oral regimen described above. We continue the amikacin for at least 12 weeks, but we continue the three oral agents for a longer period. Therapy should continue for at least 6 months and for up to 24 months depending on the clinical response [70].

Surgery is performed in selected cases. (See 'Surgery' below.)

M. marinum — Mycobacterium marinum isolates are usually susceptible in vitro to macrolides, rifamycins, ethambutol, and sulfonamides; susceptible or intermediately susceptible to streptomycin, doxycycline, and minocycline; and resistant to isoniazid and pyrazinamide [1,73]. Antimicrobial susceptibility testing is not routinely recommended for M. marinum, but it should be performed in patients who do not respond to therapy after three months [1,3,4]. When testing is performed, we suggest testing rifampin, rifabutin, ethambutol, clarithromycin, a sulfonamide, doxycycline, minocycline, linezolid, and streptomycin.

For the treatment of osteomyelitis due to M. marinum, we suggest the following oral regimen, adjusted for renal dysfunction [1]:

●Azithromycin 250 to 500 mg once daily or clarithromycin 500 mg twice daily

PLUS

●Ethambutol 15 mg/kg per day

PLUS

●Rifampin 10 mg/kg once daily (maximum daily dose 600 mg)

We recommend continuing antimycobacterial therapy for at least six months.

The 2007 American Thoracic Society/Infectious Diseases Society of America guidelines recommended a regimen that includes at least two medications and noted that a macrolide and ethambutol may represent the best balance between efficacy and tolerability [1]. Although milder forms of cutaneous disease have sometimes been treated successfully with monotherapy, we do not recommend this approach for osteomyelitis. Most patients who have been treated successfully with more invasive infections, including osteomyelitis, received two to three drugs.

Surgery is performed in selected cases. In a study that included 63 patients with M. marinum infection, approximately half of the patients underwent surgery, and failure of therapy was related to deep structural involvement [73]. In another study that included 10 patients with osteomyelitis due to M. marinum, six recovered and three underwent amputation [74]. (See 'Surgery' below.)

M. marinum soft tissue infections are discussed in greater detail separately. (See "Soft tissue infections following water exposure".)

M. ulcerans — The standard of care for Buruli ulcer, which is caused by Mycobacterium ulcerans, has been antimycobacterial therapy in combination with surgical debridement and skin grafting, when necessary. The treatment of Buruli ulcer is discussed in greater detail elsewhere. (See "Buruli ulcer (Mycobacterium ulcerans infection)".)

Other species — Other slowly growing mycobacterial species that have been reported rarely to cause osteomyelitis include M. scrofulaceum, M. simiae, M. szulgai, M. terrae complex, and M. xenopi. The general approach to therapy is similar to the approach to the other slowly growing mycobacterial species. We suggest consultation with an infectious diseases specialist with experience treating infections with these organisms.

Duration of therapy — The duration of antimycobacterial therapy depends upon several factors, including the extent of disease, response to treatment, the causative species, the immune status of the patient, and whether surgical debridement was performed. For most NTM species, antimycobacterial therapy should be continued for a minimum of six months and is sometimes given for 12 months or longer [1].

Monitoring during therapy — Monitoring for adverse effects of antimycobacterial therapy is important given the potential toxicities of the agents used. Patients should be monitored for evidence of hepatotoxicity (isoniazid, rifamycins, macrolides, oxazolidinones); nephrotoxicity (tobramycin, amikacin, streptomycin); auditory and vestibular function (tobramycin, amikacin, streptomycin, macrolides); and hematologic toxicity (rifamycins, sulfonamides, cefoxitin, oxazolidinones). We check a complete blood count and a comprehensive metabolic panel (including liver function tests) weekly when on IV therapy and monthly thereafter on oral therapy.

For aminoglycosides, monitoring should include routine questioning about balance, ability to walk (especially in the dark), tinnitus, ear fullness, dizziness, and difficulty hearing. Baseline blood urea nitrogen and creatinine measurements should be obtained. Periodic monitoring of renal function is recommended for high-risk patients receiving an aminoglycoside, especially in patients older than 50 years or who have impairment of renal function. A baseline hearing test should be performed and then repeated monthly or sooner if signs or symptoms of eighth nerve toxicity appear. (See "Rapidly growing mycobacterial infections: Mycobacteria abscessus, chelonae, and fortuitum", section on 'Monitoring for drug toxicity'.)

Repeat imaging should be performed with any recurrence of symptoms or lapse from treatment. Repeat imaging at the end of therapy is helpful to establish a new baseline. In addition, we typically repeat imaging six months after the completion of therapy to evaluate for possible relapse.

We monitor serum inflammatory markers (erythrocyte sedimentation rate, C-reactive protein) when the diagnosis of osteomyelitis is first established and monthly during therapy.

Adverse effects — The antimycobacterial agents that are used to treat NTM osteomyelitis are often difficult to tolerate, and some have important toxicities. Important adverse effects include (but are not limited to) ototoxicity and nephrotoxicity with aminoglycosides, optic neuropathy with ethambutol and linezolid, and hepatotoxicity with isoniazid. The adverse effects of the agents used to treat NTM infections are discussed in greater detail separately. (See "Pathogenesis and prevention of aminoglycoside nephrotoxicity and ototoxicity" and "Manifestations of and risk factors for aminoglycoside nephrotoxicity" and "Ethambutol: An overview" and "Azithromycin and clarithromycin", section on 'Adverse reactions' and "Rifamycins (rifampin, rifabutin, rifapentine)" and "Isoniazid: An overview" and "Isoniazid hepatotoxicity" and "Aminoglycosides", section on 'Toxicity'.)

Drug interactions — Immunocompromised patients, such as solid organ transplant recipients, are often receiving immunosuppressive agents, such as calcineurin inhibitors that interact with certain antimycobacterial agents (eg, rifampin, clarithromycin). These interactions are discussed in detail separately. (See "Nontuberculous mycobacterial infections in solid organ transplant candidates and recipients", section on 'Treatment'.)

Certain antiretroviral agents for HIV infection also have important interactions with the rifamycins, especially rifampin. (See "Treatment of drug-susceptible pulmonary tuberculosis in nonpregnant adults with HIV infection: Initiation of therapy", section on 'Drug interactions'.)

Specific interactions between medications may be determined using the drug interactions program. This program can be accessed from the UpToDate online search page or through the individual drug information topics in the section on drug interactions.

SURGERY — Surgery plays an important role in the treatment of NTM osteomyelitis, particularly since the rate of side effects and drug interactions from antimycobacterial therapy is high and may limit the tolerance to medical therapy alone. In addition, some NTM respond poorly to antimycobacterial therapy. In our experience, more aggressive surgical intervention early in the course of the disease can often decrease the number of surgeries needed and shorten the overall course of therapy. (See 'Duration of therapy' above.)

●For patients with localized osteomyelitis due to M. avium complex (MAC), we suggest a combination of excisional surgery or surgical debridement and antimycobacterial therapy [1].

●We also suggest aggressive surgical intervention for osteomyelitis due to other NTM species when there is extensive local involvement, when an abscess is present, or when antimycobacterial therapy alone is poorly tolerated or does not result in clinical improvement.

●We recommend early surgical intervention for NTM species that are difficult to treat medically, such as M. chelonae, M. abscessus, and M. haemophilum.

●A combined medical and surgical approach is also recommended for patients with vertebral osteomyelitis in order to minimize the risk of neurologic sequelae [75]. In a review of 69 patients with NTM vertebral osteomyelitis, approximately 70 percent underwent surgery in addition to antimicrobial therapy [17,76].

●Surgery should also be performed when the closed spaces of the hand are involved; this is most likely to occur with M. marinum infection, since infection with this species is usually acquired through handling the contents of aquatic environments.

High-quality evidence demonstrating the efficacy of surgical debridement for NTM osteomyelitis is lacking, and no guidelines exist regarding which patients should undergo surgery in addition to medical therapy. In a report of 22 patients with NTM otomastoiditis treated with antibiotics for a median of five months, 20 (18 of which were due to M. abscessus) underwent surgery, and 4 required revision surgery [77]. No recurrence was noted after a median follow-up of 12 months.

Several reports have described the use of bone cement [78-82], beads [83-85], and spacers [86,87] impregnated with antibiotics [88] in the setting of osteomyelitis. However, few reports have described these procedures in patients with mycobacterial infections [82,89-91]. Systemic toxicity has been reported with antibiotic-laden bone cement [92,93]. As an example, two cases of acute kidney injury were associated with the use of tobramycin-laden cement incorporated in total hip arthroplasties, and both patients had elevated serum tobramycin concentrations [93]. When aminoglycoside-laden cement is used, serum aminoglycoside concentrations should be monitored. (See "Aminoglycosides", section on 'Toxicity'.)

MANAGING POOR TREATMENT RESPONSES — Most patients will respond to therapy with improvements in symptoms and serum inflammatory markers (C-reactive protein, erythrocyte sedimentation rate) within the first month of therapy. Improvement in imaging studies will take longer, often months after treatment initiation. If the patient does not respond to antimycobacterial therapy, we suggest surgical debridement, repeating cultures and performing (or repeating) in vitro susceptibility testing, and additional antimycobacterial therapy, such as the addition of a parenteral agent.

OUTCOMES — Given the rarity of NTM osteomyelitis and the sparsity of studies, the outcomes of NTM osteomyelitis have not been clearly defined. In our experience, most patients with osteomyelitis due to NTM can be cured with appropriate surgical debridement and antimicrobial therapy. (See 'Antimycobacterial therapy' above and 'Surgery' above.)

Outcomes depend on several factors, including the immune status of the host, the extent and location of the infection, and the causative species.

As an example of the differences in outcome based upon the causative species, in an observational study that included 76 patients with nonpulmonary infections due to M. fortuitum, 68 patients (90 percent) were successfully treated [18]. Most patients had skin and soft tissue infections, but 15 had osteomyelitis. The same study included 47 patients with nonpulmonary infections due to M. chelonae or M. abscessus, of whom 34 patients (72 percent) were successfully treated. Only seven of these patients had osteomyelitis.

Good outcomes have been described with vertebral osteomyelitis. In one review, 81 percent of patients without HIV (45 of 47) and 90 percent of patients with HIV (9 of 10) recovered, improved, or achieved a pain-free state [17,76]. Four patients died during follow-up, but the deaths were not due to NTM. Outcomes were not analyzed by species. In an earlier review of 38 patients with vertebral osteomyelitis due to M. abscessus, full recovery occurred in 79 percent and partial recovery in 11 percent [94].

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Nontuberculous mycobacteria".)

SUMMARY AND RECOMMENDATIONS

●Multidisciplinary approach – Osteomyelitis caused by nontuberculous mycobacteria (NTM) requires combination antimycobacterial therapy, often in conjunction with surgery. (See 'General principles' above.)

●Combination antimycobacterial therapy – We suggest combination antimycobacterial therapy for patients with osteomyelitis due to NTM (Grade 2C). The choice of the antimycobacterial regimen should be chosen based upon the usual regimens given for the specific NTM species isolated and, for certain species, on the results of in vitro susceptibility testing. Concomitant immunosuppression should be minimized whenever possible.

In vitro susceptibility testing is recommended for the majority of NTM species isolated from the lesions of patients with osteomyelitis. Specific testing and regimen selection for each NTM species are discussed above. (See 'Antimycobacterial therapy' above.)

●Surgical intervention – For patients with localized osteomyelitis due to Mycobacterium avium complex, we suggest a combination of excisional surgery or surgical debridement and antimycobacterial therapy (Grade 2C). We also suggest surgical debridement (in addition to antimycobacterial therapy) for patients with osteomyelitis due to other NTM species when there is extensive local involvement, an abscess, vertebral osteomyelitis, involvement of the closed spaces of the hand, or when antimycobacterial therapy alone is poorly tolerated or does not result in clinical improvement (Grade 2C). Surgery is indicated for NTM species that are difficult to treat medically, such as Mycobacterium chelonae and Mycobacterium abscessus. (See 'Surgery' above.)

●Duration of therapy – The duration of antimycobacterial therapy depends upon several factors, including the response to treatment, the causative species, and the immune status of the patient. Antimycobacterial therapy is typically continued for a minimum of 6 months and is sometimes given for 12 months or longer. (See 'Duration of therapy' above.)

●Nonresponse to therapy – If the patient does not respond to antimycobacterial therapy, then surgical debridement, repeating cultures and performing (or repeating) susceptibility testing, and additional antimycobacterial therapy, such as the addition of a parenteral agent, are the next steps. (See 'Managing poor treatment responses' above.)