Uslan DZ, Kowalski TJ, Wengenack NL, Virk A, Wilson JW. Skin and Soft Tissue Infections Due to Rapidly Growing MycobacteriaComparison of Clinical Features, Treatment, and Susceptibility. Arch Dermatol. 2006;142(10):1287-1292. doi:10.1001/archderm.142.10.1287
To compare the demographics, clinical features, susceptibility patterns, and treatment for skin and soft tissue infections due to Mycobacterium fortuitum and Mycobacterium chelonae or Mycobacterium abscessus.
Retrospective medical record review.
Mayo Clinic, Rochester, Minn.
All patients seen at our institution with a positive culture for M chelonae, M abscessus, or M fortuitum from skin or soft tissue sources between January 1, 1987, and October 31, 2004.
Main Outcome Measures
Patient demographics, clinical characteristics, therapeutic data, microbiological data, and outcomes.
The medical records of 63 patients with skin or soft tissue infections due to rapidly growing mycobacteria were reviewed. Patients with M chelonae or M abscessus were older (61.5 vs 45.9 years, P<.001) and more likely to be taking immunosuppressive medications (60% vs 17%, P = .002) than patients with M fortuitum. Mycobacterium fortuitum tended to manifest as a single lesion (89% vs 38%, P<.001), while most M chelonae or M abscessus manifested as multiple lesions (62% vs 11%, P<.001). More patients with M fortuitum had a prior invasive surgical procedure at the infected site (56% vs 27%, P = .04). Patients with multiple lesions were more likely to be taking immunosuppressive medications than those with single lesions (67% vs 30%, P = .006). Seven patients failed treatment, several of whom were immunocompromised and had multiple comorbidities.
Skin and soft tissue infections due to rapidly growing mycobacteria are associated with systemic comorbidities, including the use of immunosuppressive medications. There are significant differences in the demographic and clinical features of patients who acquire specific organisms, including association with immunosuppression and surgical procedures.
Mycobacterium fortuitum, Mycobacterium chelonae, and Mycobacterium abscessus are environmental mycobacteria that can cause chronic infections of the skin, soft tissues, and lungs. These organisms are characterized by rapid growth on standard media and by lack of pigmentation.1 Clinical manifestations include localized abscess formation and chronic ulcers.1,2 Disseminated infections, especially in immunocompromised hosts, have been widely described.1,3,4 There are multiple reports of infection after trauma and surgical or other procedures, including acupuncture,5 liposuction,6- 8 silicone injection,9 breast implantation,10- 12 intravenous catheter use,13 dermatologic surgery,14,15 pedicures,16- 18 exposure to prosthetic material,19 pacemaker placement,20 and subcutaneous injections.21- 25 Pulmonary disease due to M fortuitum has been described following hot tub use.26
Although the spectrum and features of clinical disease due to M chelonae have been described,3 no reports have directly compared the clinical features of skin and soft tissue infections due to M chelonae, M abscessus, and M fortuitum, to our knowledge. We retrospectively reviewed all patients with skin and soft tissue isolates of rapidly growing mycobacteria at our institution to compare the demographics, clinical features, susceptibility patterns, and treatment for M fortuitum and M chelonae or M abscessus.
The medical records of all consecutive patients having clinical isolates of M chelonae, M abscessus, and M fortuitum from skin and soft tissue seen at our institution between January 1, 1987, and October 31, 2004, were reviewed retrospectively. All patients had given consent for the use of their medical records for research purposes, and our institutional review board approved the study. Information was collected regarding demographics, immune status (including the use of immunosuppressive medications), infection characteristics, site and clinical description of the lesions, antimicrobial therapy, surgical procedures, and outcome. Definite infection was defined as culture positivity in the presence of a compatible clinical syndrome, such as chronic ulcers, wound drainage, or abscess. Patients with isolated single colonies in their cultures, in the absence of a compatible clinical syndrome, were excluded from the analysis.
Organisms were identified by our laboratory as rapidly growing mycobacteria by standard criteria that included growth rate, morphologic structure, mycolic acid analysis, and biochemical test results (arylsulfatase and nitrate reduction).27 On November 9, 1999, our microbiology laboratory began identifying rapidly growing mycobacteria by 16S ribosomal RNA (rRNA) gene sequencing.28Mycobacterium chelonae and M abscessus are indistinguishable by 16S rRNA gene sequencing and are reported together (as M chelonae or M abscessus) by most laboratories, as well as herein.
Minimum inhibitory concentrations (MICs) of antimicrobials were determined using serial 2-fold broth microdilution in cation-supplemented Mueller-Hinton broth at 30°C, as per guidelines from the Clinical and Laboratory Standards Institute (CLSI).29 Minimum inhibitory concentration break points from the CLSI for interpretive criteria for rapidly growing mycobacteria were used.
Data were analyzed using JMP software (version 5.1.2; SAS Institute, Cary, NC). Comparisons between categorical variables were analyzed using the Fisher exact test. Comparisons between continuous variables were analyzed using the Wilcoxon rank sum test or independent t test, as appropriate.
Seventy-six patients with skin or soft tissue infections due to rapidly growing mycobacteria were identified. Of these, 63 were classified as having definite infection and were included in the analysis. Clinical history was available for all patients. Susceptibility data were reviewed when available; given the 17-year interval, not all isolates had been tested against all antimicrobials in the current CLSI guidelines.
The patients included 35 women (56%) and 28 men (44%), with a median age of 60 years (age range, 19-88 years). Patients came from 12 different states, with most from Minnesota (29 patients), Wisconsin (7 patients), Iowa (4 patients), and South Dakota (3 patients), which reflects the general patient population at our institution. Five patients (8%) were current smokers, and 6 patients (10%) were former smokers. In view of the retrospective nature of the study design and the lack of specific information available in the medical records, no consistent exposure, occupational, or avocational history (eg, hot tub exposure) could be identified that might represent risk factors for acquisition of these organisms.
Seven cultures (11%) were obtained via superficial swab of a lesion, 34 (55%) via biopsy of a lesion, and 21 (34%) at the time of surgical debridement or resection. The method of culture was unavailable for 1 patient. Smears for acid-fast bacilli were performed in 50 patients; 24 (48%) were positive for organisms. Sixteen cultures (25%) yielded other bacteria in addition to rapidly growing mycobacteria, including coagulase-negative staphylococci, Staphylococcus aureus, and gram-negative organisms (Klebsiella species, Escherichia coli, and Proteus species). No patients' cultures grew multiple different species of mycobacteria. Skin specimens from 40 patients had been sent for pathological examination; granulomatous inflammation was observed in 21 patients (53%).
Demographics of the 63 patients by organism (M fortuitum or M chelonae or M abscessus) are given in Table 1. There were 18 M fortuitum infections (29%) and 45 M chelonae or M abscessus infections (71%). There were significant differences noted between patients with M fortuitum and M chelonae or M abscessus in their mean age (45.9 vs 61.5 years, P<.001), use of immunosuppressive medications at the time of infection (17% vs 60%, P = .002), and history of an invasive procedure at the site of diagnosis (56% vs 27%, P = .04), as well as in the presence of systemic comorbidities (including active malignancies, end-stage renal disease, congestive heart failure, solid organ transplantation, and connective tissue disease). None of the patients had human immunodeficiency virus infection. The median time between the recorded onset of symptoms and the culture of the lesion was 86 days (intraquartile range [IQR], 38.5-202.5 days). There was no difference in the duration of symptoms before diagnosis by organism, systemic comorbidities, or immunosuppression status. Patients with multiple lesions had a significantly longer duration of symptoms before diagnosis (median, 125 days [IQR, 52-342 days]) compared with patients with single lesions (median, 65 days [IQR, 17-165 days]) (P = .03). The median duration of patient follow-up after diagnosis was 189 days (range, 4 to 5477 days [IQR, 93-831 days]).
The clinical features by organism are given in Table 2. Thirty-three patients (52%) had a single lesion, and 30 patients (48%) had multiple lesions. Patients with M fortuitum were more likely to initially be seen with single lesions than patients with M chelonae or M abscessus (89% vs 38%, P<.001). Mycobacterium fortuitum was more frequently isolated from the chest or back (39% vs 4%, P = .002), while M chelonae or M abscessus was more frequently isolated from the lower extremities (64% vs 22%, P = .002). There were no differences in leukocyte count, maximum temperature, or clinical description of skin findings between patients in the 2 groups. Given the retrospective nature of the study and the reliance on medical records for clinical information, detailed descriptions of lesions were often unavailable. Reported skin findings varied widely, including sinus tracts, nonhealing ulcers, subcutaneous abscesses, subcutaneous fluctuant or firm nodules of varying size, and erythema in association with ulcers or chronic drainage from prior surgical wounds. Specific descriptions of skin color were not typically documented, although several patients' nodules were described as “violaceous.” Only 2 patients (3%) had culture-proved pulmonary coinfection.
Factors associated with single vs multiple lesions are given in Table 3. Patients with multiple lesions were more likely than those with single lesions to be taking immunosuppressive medications (67% vs 30%, P = .006) or to have a systemic comorbidity (such as a malignant neoplasm, diabetes mellitus, or connective tissue disease) (63% vs 27%, P = .006). Patients with single lesions were more likely to have a history of penetrating trauma (18% vs 0%, P = .03) or of an invasive surgical procedure (52% vs 17%, P = .007) at the site of infection.
Forty-five patients were treated with antimicrobial agents. Of the remaining 18 patients, 9 were treated with surgery alone. Fourteen patients received parenteral antimicrobial agents, with a median duration of 25 days (range, 6-158 days). Thirty-eight patients received oral antimicrobial agents, with a median duration of 127 days (range, 9-891 days). Patients with multiple lesions were treated for a median duration of 135 days (IQR, 38.5-230.75 days), while patients with single lesions were treated for a median duration of 181 days (IQR, 109.5-311.5 days), although this difference was not statistically significant (P = .33).
The 45 patients who were treated received a total of 102 courses of antimicrobial agents (with a course defined as ≥7 days of receiving antimicrobial agents). A single antimicrobial was prescribed for 20 courses (20%). Combination therapy (consisting of ≥2 antimicrobial agents) was prescribed for 82 courses (80%). The most frequently prescribed antimicrobial agents were macrolides (31 courses), fluoroquinolones (26 courses), aminoglycosides (14 courses), cephalosporins (11 courses), and tetracyclines (9 courses). Antimicrobial agents most frequently used for monotherapy included macrolides (12 courses) and fluoroquinolones (6 courses).
Thirty-seven patients (59%) underwent surgical procedures. Patients with single lesions (76% [25/33]) were more likely to undergo surgery than those with multiple lesions (40% [12/30]) (P = .005). Surgical procedures were also more likely in patients without systemic comorbidities (71% [25/35]) than in patients with systemic comorbidities (43% [12/28] (P = .04). Surgery was not associated with other characteristics (organism, patient age, sex, or lesion location).
Given the short median duration of follow-up, we were unable to assess outcome by specific therapy or by demographic factors. Treatment failure was defined as lack of clinical improvement with therapy or relapse of the infection during follow-up. Seven patients had documented treatment failure. All patients with M chelonae or M abscessus who failed treatment had multiple lesions or deep soft tissue infection. All were immunocompromised, including 3 patients who were solid organ transplant recipients. The 2 patients with M fortuitum who failed treatment had infection at the site of prior surgical procedures (removal of skin cancer in one and abdominoplasty in the other). Only 1 patient died of infection; this patient had disseminated disease, including 2 infected prosthetic joints, skin nodules, and lung infection. The median time to failure was 154 days (range, 20 to 1898 days). None of the patients who failed treatment developed acquired antimicrobial resistance.
Susceptibility data (according to 2003 CLSI standards29) and MICs for available isolates are given in Table 4. Minimum inhibitory concentrations were reviewed retrospectively and were available for different antimicrobials depending on the year the culture was performed. In general, MICs were determined for amikacin, cefoxitin, ciprofloxacin, clarithromycin, doxycycline, imipenem, linezolid, sulfamethoxazole, and tobramycin. A few isolates were sent to referral laboratories for clofazimine susceptibility testing; those results are not included given the absence of formal CLSI break points. Isolates of M fortuitum were universally susceptible to amikacin. Isolates of M chelonae or M abscessus were universally susceptible to clarithromycin but had varying resistance to amikacin (36% intermediate or resistant), tobramycin (21% intermediate or resistant), and imipenem (94% intermediate or resistant). Acquired resistance to an antimicrobial agent, defined as a change in susceptibility category from the original culture, was not observed in strains isolated from any patient, regardless of his or her outcome.
To our knowledge, these 63 patients with skin and soft tissue infection due to M fortuitum and M chelonae or M abscessus represent the largest case series directly comparing clinical characteristics of patients with rapidly growing mycobacteria infection. These emerging infections are increasingly important, especially because of their association with surgical procedures2 and their tendency to disseminate in immunocompromised hosts.3 Diagnosis is often delayed, as mycobacterial cultures are not routinely performed on skin biopsy specimens or surgical wound infections. The median time between the onset of symptoms and the microbiological diagnosis was 86 days. Therefore, a high index of suspicion is imperative for the diagnosis to be made. Because our series included only culture-positive patients, it is possible that patients with clinically mild disease (who did not prompt a culture) or patients with lesions cultured elsewhere but who were referred to our institution for management were missed.
The clinical features and demographics in our study are similar to those of other case series of rapidly growing mycobacteria,2,3,30 although we found significant differences in age, systemic comorbidities, use of immunosuppressive medications, and history of invasive procedures between patients with M fortuitum and M chelonae or M abscessus. In general, the patients infected with M chelonae or M abscessus were older and in poorer health, while the patients infected with M fortuitum were more likely to have experienced trauma or to have had a surgical procedure at the site. This is also reflected in the significant difference between the organisms in the association of multiple vs single lesions (11% of M fortuitum infections had multiple lesions compared with 62% of M chelonae or M abscessus). Most patients with single lesions had a history of trauma or a medical procedure at the subsequent site of infection, similar to previous findings.2 Because the 63 patients in our study spanned a 17-year interval and were evaluated retrospectively, details of physical examinations were obtained from the patients' medical records. Therefore, further clinical descriptions of skin lesions were limited to what had been recorded; specific details such as lesion size, color, ulcer appearance, and presence of fluctuance were not typically available.
Optimal treatment of rapidly growing mycobacterial infections remains poorly established. No study has compared different antimicrobial regimens, to our knowledge. Current guidelines recommend susceptibility testing of all isolates, with use of empirical therapy suggested until susceptibilities are known.1,31 A limitation of the present study is the small number of isolates for which susceptibility testing results were available, especially for some of the newer antimicrobial agents such as linezolid. Current guidelines recommend empirical amikacin for treatment of M fortuitum; 100% of our isolates were susceptible to this agent.1 In contrast to other published studies,1,3,32- 34 our M fortuitum isolates were not universally susceptible to ciprofloxacin (14% resistant), sulfamethoxazole (33% resistant), or imipenem (30% intermediate or resistant). In contrast to previously published in vitro data,33 most of our isolates were intermediate or resistant to cefoxitin (81% of M fortuitum and 98% of M chelonae or M abscessus). Clarithromycin appears to be reliably active against M chelonae or M abscessus, although its activity against M fortuitum is less predictable.34 Our laboratory, as many others, uses 16S rRNA gene sequencing for the identification of rapidly growing mycobacteria and does not distinguish between M chelonae and M abscessus. These organisms are known to have diverging in vitro antimicrobial susceptibilities to certain drugs, including tobramycin, amikacin, and cefoxitin. This may contribute to the broadened range of antimicrobial susceptibility data within our M chelonae or M abscessus group, but it remains unclear whether further microbial speciation (such as via citrate utilization) would affect patient management.
The single clinical trial using clarithromycin monotherapy noted only 1 case of acquired clarithromycin resistance (8% of patients), in a noncompliant patient who prematurely discontinued therapy.35 There are other case reports of clarithromycin resistance developing in patients with M chelonae infection while receiving therapy,36- 39 and monotherapy with this agent should be undertaken with caution. Although there are other case reports of effective use of single agents,15,32,40 combination therapy seems prudent initially because of concerns about acquired resistance (particularly in disseminated disease, with increased organism burden and heightened risk of resistance). None of 7 patients in our series who failed therapy developed acquired antimicrobial resistance. Death from cutaneous rapidly growing mycobacterial infection seems to be rare; most patients have had symptoms for a median of 3 months; and susceptibility testing is usually available promptly. In isolated nondisseminated disease, it may be reasonable to await results of susceptibility testing rather than to empirically start antimicrobials such as aminoglycosides, which may have serious adverse effects, especially in older patients or in those with other comorbidities such as renal disease. Given the short median follow-up in our series, we were unable to draw any conclusions about the duration of treatment or the effectiveness of a particular antimicrobial regimen.
Surgical therapy is an important adjunctive tool in treating these infections. In our series, patients with single lesions were far more likely than those with multiple lesions to undergo surgical debridement (76% vs 40%). Extensive surgical debridement for widespread infection may be technically challenging; however, patients with multiple lesions are more likely to be immunosuppressed and may be less likely to cure their infection with antimicrobial agents alone. Combined medical and surgical therapy is likely to produce optimal results, in accord with current guidelines.1 In 2 reports of outbreaks of sternal wound infections due to M abscessus, approximately one third of the patients died of uncontrolled infection,41,42 suggesting that aggressive therapy may be indicated for deep infection. Our study did not specifically address the issue of reduction of immunosuppression, but it is likely that this may be beneficial, if tolerable by a patient relative to his or her underlying comorbidities.
The prevalence of infections due to rapidly growing mycobacteria is unknown but is likely underestimated given the wide range of presenting syndromes and the lack of knowledge about these organisms by many clinicians. Clinicians caring for patients with chronic skin or soft tissue infections, especially in immunosuppressed patients or following a surgical procedure, should have a high index of suspicion for rapidly growing mycobacterial disease. Further studies establishing the optimal duration of therapy and comparing specific regimens are warranted.
Correspondence: Daniel Z. Uslan, MD, Division of Infectious Diseases, Department of Medicine, Mayo Clinic College of Medicine, 200 First St SW, Rochester, MN 55905 (firstname.lastname@example.org).
Financial Disclosure: None reported.
Previous Presentation: This study was presented in part at the American College of Physicians 2006 Annual Session; April 7, 2006; Philadelphia, Pa.
Accepted for Publication: March 26, 2006.
Author Contributions:Study concept and design: Uslan, Kowalski, Virk, and Wilson. Acquisition of data: Uslan, Kowalski, and Wegenack. Analysis and interpretation of data: Uslan, Kowalski, Wegenack, Virk, and Wilson. Drafting of the manuscript: Uslan and Wilson. Critical revision of the manuscript for important intellectual content: Uslan, Kowalski, Wegenack, Virk, and Wilson. Statistical analysis: Uslan and Kowalski. Administrative, technical, and material support: Virk and Wilson. Study supervision: Virk and Wilson.