Purpose
We analyzed patient demographic factors involved in the development of nonmarinum, nontuberculous mycobacterial infections (NTMI) involving the upper extremity, and assessed diagnostic and prognostic values of commonly used preoperative laboratory and imaging studies, as well as factors related to recurrence of disease and patient outcomes.
Methods
Patients from 2 academic, tertiary facilities with culture-proven, nonmarinum NTMI involving the upper extremity were reviewed. Patient-related factors and clinical outcomes were extracted. The analysis was based on pathogen identification (rapid- vs slow-growing subspecies) and immune status.
Results
Our 76 patients had a mean age of 59 years, and 65% were male. Forty-eight percent reported an injury, and hands were frequently involved (58%). Forty-one percent were immunosuppressed (19% organ transplant recipients). The mean symptom duration prior to presentation was 203 days. The culture identification took a mean of 33 days, with 25 different species identified (subcategorized as rapid or slow growers). Seventy-seven percent had solitary lesions, with a cutaneous or subcutaneous location most common. Immunosuppressed patients were treated longer with antibiotics (243 vs 155 days in immunocompetent patients) and experienced higher rates of side effects, complications, and recurrence. All patients underwent debridement to control infection, including 4 individuals who required amputations. One-third experienced complications and/or recurrence, regardless of the organism type.
Conclusions
Upper-extremity nonmarinum NTMI is often misdiagnosed, causing management delays. Early consideration in differential diagnoses of chronic, painful swelling, nodular or inflammatory lesions, or septic arthritis is crucial. Tissue biopsy with specimens for histopathology and microbiological analysis (mycobacterial smear, cultures, and broad range polymerase chain reaction) and early involvement with an infectious disease specialist are recommended. Empiric antibiotic therapy is not standard. Debridement and prolonged, directed combination antimicrobial therapy is required; however, adverse reactions are commonly encountered.
Type of study/level of evidence Prognostic IV.
Key words
Nonmarinum nontuberculous mycobacteria (NTM) species are ubiquitous in the environment and uncommonly cause human infections.
1
The most common slow-growing NTM species include Mycobacterium avium complex, Mycobacterium kansasii, Mycobacterium xenopi, and Mycobacterium abscessus are the most common rapid-growing NTM.2
Although nontuberculous mycobacterial infections (NTMI) are not on the Centers for Disease Control and Prevention’s reportable list, the estimated annual incidence is 30,000 cases in the United States.3
A 2013 study found the incidence of cutaneous NTMI increased nearly 3-fold between 1980 and 2009.4
They theorized that this increase may be due to increases in the average age and number of solid organ transplant recipients in the United States. Nonmarinum NTMIs account for 5% to 15% and 25% to 40% of all mycobacterial infections in the general population and in transplant recipients, respectively.5
, 6
, 7
A hallmark of skeletal NTMI is an indolent, progressive infection. Often confused for a noninfectious, inflammatory process, NTMI frequently delays diagnosis and treatment. The most common manifestations of NTMI are pulmonary disease and lymphadenitis, yet the wrist and hand are theorized to be the most common sites of extrapulmonary NTMI.
8
, 9
, 10
The majority of cases present as tenosynovitis.10
, 11
, 12
, 13
Subsequently, corticosteroid injections are often used, which can exacerbate underlying NTMI.14
Once a diagnosis is confirmed by tissue culture, a multidisciplinary treatment approach is indicated.We present a retrospective study, spanning over an 8-year period, of NTMI of the upper extremity. We chose not to include Mycobacterium marinum infections because they have specific and unique risk factors, as well as classic clinical manifestations and standardized treatment regimens, that have been well described in the literature.
14
, 15
, 16
, 17
, 18
, 19
, 20
The purpose of this study was to describe patient factors, comorbidities, clinical manifestations, laboratory data, and histopathologic and microbiological information; to identify potential clinical features or risk factors; and to examine therapy-based outcomes when comparing infections caused by rapid- versus slow-growing organisms or when comparing infections in patients with and without clinical immunosuppression.Materials and Methods
This retrospective, descriptive study was approved by our Mayo Clinic institutional review board. The microbiology database from 2 referral centers of a large, tertiary-care hospital system was searched for the term “mycobacterium” between January 2011 and December 2018. From a comprehensive list of positive mycobacterial cultures, we included patients if they met the following inclusion criteria: age >18 years and NTMI involving the upper extremity (shoulder to fingertip).
For each patient, the inciting injury or event (if any) was categorized as a burn, crushing, gardening, or penetrating injury or as a medical procedure or intervention after which symptoms began. The duration of symptoms (time from onset to presentation at our institutions) was recorded in days. Pain, swelling, and constitutional symptoms, including night sweats, fever, and weight loss, were recorded. Unifocal or multifocal lesion locations (digit, hand, wrist, forearm, elbow, brachium) and characteristics (bone, tendon, synovium, joint, cutaneous, or bursal involvement and abscess formation) were noted. Interventions prior to surgical debridement, including antibiotic therapy, steroid injections, or wound care, were verified. Preoperative laboratory values consisted of the white blood cell count, erythrocyte sedimentation rate, and C-reactive protein. Underlying comorbidities were documented, including body mass index, diabetes mellitus and insulin dependence, and rheumatologic, autoimmune, and skeletal disorders. Immunosuppressive disorders and microbiological test data are shown in Table 1. Histopathology results from biopsy specimens were detailed, as were the duration and choice of antimicrobial therapy, adverse reactions, and complications.
Table 1List of Included Immunosuppressive Disorders and Microbiological Test Data
| Immunosuppressive Disorders | Microbiological Test Data |
|---|---|
| Hematologic malignancy | Acid fast smear (AFB stain) |
| Solid organ or bone marrow transplantation | Mycobacterial culture |
| Human immunodeficiency virus (HIV infection) | Broad-range PCR assay |
| Common variable immune deficiency | DNA probe |
| Immunosuppressive therapies at the time of NTMI acquisition: cancer chemotherapy, systemic corticosteroids (>20 mg/d of prednisone or equivalent for >1 mo), therapeutic interferon preparations, receipt of biologics (such as tumor necrosis factor and interleukin inhibitors), other disease-modifying antirheumatic drugs (such as azathioprine, methotrexate, calcineurin inhibitors, or mycophenolate mofetil) |
AFB, acid-fast bacteria; HIV, human immunodeficiency virus; PCR, polymerase chain reaction.
Continuous variables were summarized with the mean, standard deviation, median, and range by organism type. Wilcoxon rank-sum tests were used to test the associations between continuous variables and organism types. Categorical variables are summarized as counts and percentages by organism type. To test the associations between categorical variables and organism types, chi-square or Fisher exact tests were used.
Results
Patient demographics
A total of 76 patients with upper-extremity NTMI were identified during the study period. The mean age was 59 years (range, 18–89 years) and 65% of patients were male. The mean body mass index was 27.1 kg/m2 (range, 18.5–52.4 kg/m2). Pertinent comorbidities included diabetes mellitus (n = 23; 30%) and underlying immunosuppression (n = 31; 41%). Among immunosuppressed patients, 52 (71%) had an underlying autoimmune disorder, 84% were on active therapy with disease-modifying antirheumatic drugs, and 19% had undergone a solid organ transplant (Table 2).
Table 2Patient Demographics and Clinical History
| Characteristic | Organism Type | Total (n = 76) | P Value | |
|---|---|---|---|---|
| Fast Growers (n = 23) | Slow Growers (n = 53) | |||
| Age | ||||
| N (missing) | 23 (0) | 53 (0) | 76 (0) | .9 |
| Mean (SD) | 58.7 (16.39) | 59.5 (15.09) | 58.7 (16.21) | |
| Median | 63.0 | 61.0 | 61.0 | |
| Range | 20.0–85.0 | 27.0–89.0 | 12.0–89.0 | |
| Sex, n (%) | ||||
| Female | 7 (30.4%) | 20 (37.7%) | 27 (35.1%) | .6 |
| Male | 16 (69.6%) | 33 (62.3%) | 49 (64.9%) | |
| Body mass index | ||||
| N (missing) | 23 (0) | 53 (1) | 76 (1) | .8 |
| Mean (SD) | 27.2 (5.29) | 27.2 (6.00) | 27.1 (5.79) | |
| Median | 26.2 | 25.5 | 25.5 | |
| Range | 18.5–41.6 | 19.2–52.4 | 18.5–52.4 | |
| Diabetes mellitus, n (%) | ||||
| No | 18 (78.3%) | 45 (84.9%) | 63 (83.1%) | .5 |
| Yes | 5 (21.7%) | 8 (15.1%) | 13 (16.9%) | |
| IInsulin-Dependent DM, n (%) | ||||
| No | 18 (78.3%) | 48 (90.6%) | 64 (83.1%) | .5 |
| Yes | 5 (21.7%) | 5 (9.4%) | 13 (16.9%) | |
| HIV, n (%) | ||||
| No | 23 (100.0%) | 52 (98.1%) | 76 (98.7%) | >.99 |
| Yes | 0 (0.0%) | 1 (1.9%) | 1 (1.3%) | |
| Transplant, n (%) | ||||
| No | 20 (87.0%) | 50 (94.3%) | 71 (92.2%) | .4 |
| Yes | 3 (13.0%) | 3 (5.7%) | 6 (7.8%) | |
| Autoimmune disease, n (%) | ||||
| No | 16 (69.6%) | 38 (71.7%) | 55 (71.4%) | .8 |
| Yes | 7 (30.4%) | 15 (28.3%) | 22 (28.6%) | |
| Injury or event, n (%) | ||||
| No | 12 (52.2%) | 28 (52.8%) | 40 (51.9%) | >.99 |
| Yes | 11 (47.8%) | 25 (47.2%) | 37 (48.1%) | |
| Injury or event type (gardening, trauma, other, etc.), n (%) | 0.1.1 | |||
| Burn | 2 (18.2%) | 0 (0.0%) | 2 (5.4%) | |
| Crushing | 0 (0.0%) | 2 (8.0%) | ||
| Gardening | 2 (18.2%) | 4 (16.0%) | 6 (16.2%) | |
| Medical | 4 (36.4%) | 5 (20.0%) | 9 (24.3%) | |
| Penetrating | 3 (27.3%) | 14 (56.0%) | ||
| Missing | 12 | 28 | ||
| Location | ||||
| Digit, n (%) | .6 | |||
| No | 15 (65.2%) | 40 (74.1%) | 55 (71.4%) | |
| Yes | 8 (34.8%) | 14 (25.9%) | 22 (28.6%) | |
| Hand, n (%) | .7 | |||
| No | 20 (87.0%) | 44 (81.5%) | 64 (83.1%) | |
| Yes | 3 (13.0%) | 10 (18.5%) | 13 (16.9%) | |
| Wrist, n (%) | .03 | |||
| No | 23 (100.0%) | 44 (81.5%) | 67 (87.0%) | |
| Yes | 0 (0.0%) | 10 (18.5%) | 10 (13.0%) | |
| Forearm, n (%) | >.99 | |||
| No | 19 (82.6%) | 44 (81.5%) | 63 (81.8%) | |
| Yes | 4 (17.4%) | 10 (18.5%) | 14 (18.2%) | |
| Elbow, n (%) | .1 | |||
| No | 16 (69.6%) | 46 (85.2%) | 62 (80.5%) | |
| Yes | 7 (30.4%) | 8 (14.8%) | 15 (19.5%) | |
| Brachium, n (%) | >.99 | |||
| No | 21 (91.3%) | 49 (90.7%) | 70 (90.9%) | |
| Yes | 2 (8.7%) | 5 (9.3%) | 7 (9.1%) | |
DM, diabetes mellitus; HIV, human immunodeficiency virus.
∗ Fisher exact P value.
† Wilcoxon rank-sum P value.
Clinical characteristics
Clinical characteristics and outcomes are summarized in Table 3. Local symptoms of pain (75%) and swelling (96%) were documented more than systemic symptoms, such as fever or night sweats (13%) Additionally, isolated cutaneous or subcutaneous lesions were reported in 67% of cases, and the mean C-reactive protein value was 23 mg/L in patients with rapid-growing NTMI compared to 9 in patients with slow-growing NTMI.
Table 3Clinical Manifestations and Intraoperative Findings
| Symptom | Organism Type | Total (N = 76) | P Value | |
|---|---|---|---|---|
| Fast Growers (n = 23) | Slow Growers (n = 53) | |||
| Night sweats, fever, weight loss, n (%) | ||||
| No | 19 (82.6%) | 48 (90.6%) | 67 (87.0%) | .5 |
| Yes | 4 (17.4%) | 5 (9.4%) | 10 (13.0%) | |
| Pain, n (%) | ||||
| No | 5 (21.7%) | 14 (26.4%) | 19 (24.7%) | .8 |
| Yes | 18 (78.3%) | 39 (73.6%) | 57 (75.3%) | |
| Swelling, n (%) | ||||
| No | 0 (0.0%) | 3 (5.7%) | 3 (3.9%) | .6 |
| Yes | 23 (100.0%) | 50 (94.3%) | 73 (96.1%) | |
| Single vs multifocal lesion, n (%) | ||||
| Multifocal | 7 (30.4%) | 11 (20.8%) | 18 (23.4%) | .4 |
| Single | 16 (69.6%) | 42 (79.2%) | 58 (76.6%) | |
| Abscess formation, n (%) | ||||
| No | 12 (52.2%) | 32 (60.4%) | 44 (57.1%) | .6 |
| Yes | 11 (47.8%) | 21 (39.6%) | 32 (42.9%) | |
| Bone involvement, n (%) | ||||
| No | 17 (73.9%) | 41 (77.4%) | 58 (75.3%) | >.99 |
| Yes | 6 (26.1%) | 12 (22.6%) | 18 (24.7%) | |
| Tendon or tenosynovial involvement, n (%) | ||||
| No | 13 (56.5%) | 21 (39.6%) | 34 (45%) | .2 |
| Yes | 10 (43.5%) | 32 (60.4%) | 42 (55%) | |
| Joint involvement, n (%) | ||||
| No | 16 (69.6%) | 37 (69.8%) | 53 (70.1%) | >.99 |
| Yes | 7 (30.4%) | 16 (30.2%) | 23 (29.9%) | |
| Cutaneous or subcutaneous involvement, n (%) | ||||
| No | 5 (21.7%) | 20 (37.7%) | 25 (33%) | .3 |
| Yes | 18 (78.3%) | 33 (62.3%) | 51 (67%) | |
| Bursal involvement, n (%) | ||||
| No | 17 (73.9%) | 40 (75.5%) | 57 (75%) | >.99 |
| Yes | 6 (26.1%) | 13 (24.5%) | 19 (25%) | |
| Compressive neuropathy, n (%) | ||||
| No | 23 (100.0%) | 46 (86.8%) | 69 (90%) | .1 |
| Yes | 0 (0.0%) | 7 (13.2%) | 7 (10%) | |
| ESR | n = 11 | n = 22 | N = 33 | .97 |
| N (Missing) | 11 (0) | 22 (0) | 33 (0) | |
| Mean (SD) | 15.5 (11.43) | 21.3 (22.92) | 19.3 (19.83) | |
| Median | 10.0 | 21.0 | 15.0 | |
| Range | 2.0–36.0 | 2.0–99.0 | 2.0–99.0 | |
| CRP | n = 12 | n = 22 | N = 34 | .7 |
| N (Missing) | 12 (0) | 22 (0) | 34 (43) | |
| Mean (SD) | 22.6 (40.02) | 8.7 (11.52) | 13.6 (25.76) | |
| Median | 5.7 | 4.9 | 4.9 | |
| Range | 3.0–141.2 | 3.0–55.2 | 3.0–141.2 | |
| DMARDs or TNF, n (%) | ||||
| No | 15 (65.2%) | 35 (66.0%) | ||
| Yes | 8 (34.8%) | 18 (34.0%) | ||
∗ Abbreviations: CRP, C-reactive protein; DMARDs disease-modifying antirheumatic drugs; ESR, erythrocyte sedimentation rate; TNF, tumor necrosis factor.
† Fisher exact P value.
‡ Wilcoxon rank-sum P value.
Characteristics of pathogens and immune status
All patients had either an intralesional (debulking) or marginal excision (depending upon the anatomic location and proximity of neurovascular or other vital structures) of pathologic tissue visualized during the surgical debridement. No patients in this series were treated without surgery. There were 25 different species of NTM identified; organisms were subcategorized as rapid growers or slow growers (Table 4). The distinction between rapidly growing and slow-growing mycobacteria was made based on Runyon’s classification, which allows grouping of mycobacterium based on the growth rate and certain phenotypic characteristics, such as pigmentation.
21
,22
Notably, acid-fast bacteria stains were positive in 67% of patients, regardless of the organism type. Table 5 lists some of the differences in presentations and outcomes between immunosuppressed and immunocompetent patients.Table 4Characteristics of Pathogens
| NTMI Classification | Mean/Median Time to Culture Identification (d) | Number/Percentage of Immunosuppressed Patients | Mean/ Median Duration of Symptoms (d) | Mean/ Median HbA1c | Prior Corticosteroid Injections at Site of Infection (Percentage of Patients) | Mean/ Median Duration of Treatment With Antimicrobials (d) | Histopathology | ||
|---|---|---|---|---|---|---|---|---|---|
| Incidence of Granuloma Formation (%) | Chronic Inflammatory Changes | Tissue Necrosis | |||||||
| Rapid-growing organisms (n = 23; 30%), M. chelonae, M. fortuitum, and M. abcessus | 23/20 | 11/48 | 115/60 | 7.9/8.7 | 23 | 120/135 | 36 | 9 | 9 |
| Slow-growing organisms (n = 54; 70%), M. haemophilum (n = 7), M. kansasii (n = 3), M. avium complex (n = 19), and other (n=25) | 37/34 | 20/38 | 248/90 | 6.4/ 6.8 | 11 | 209/180 | 60 | 32 | 24 |
| P value | <.05 | .4 | .8 | .1 | .3 | .1 | .3 | .2 | .4 |
HbA1c, glycated hemoglobin.
Table 5Immune Status
| Immune Status | Inciting Event or Trauma Reported, Number/Percentage of Total | Single Lesion, Number/Percentage of Total | Corticosteroid Injection at Site Prior to Infection, Number/Percentage of Total | Complications, Number/Percentage of Total | Hand Involvement, Number/Percentage of Total | Duration of Antimicrobial Therapy, Mean/Median in d | Side Effects to Antibiotics, Number/Percentage of Total | Recurrence Rate, Number/Percentage of Total | Mean/Median Patient Age at Diagnosis |
|---|---|---|---|---|---|---|---|---|---|
| Immunocompetent (n = 45) | 31/69% | 40/89% | 10/22% | 5/16% | 12/27% | 155/180 | 12/33% | 7/21% | 56/58 |
| Immunosuppressed (n = 31) | 6/19% | 18/58% | 1/3% | 11/55% | 1/3% | 243/180 | 14/70% | 9/47% | 63/67 |
| P value | <.05 | <.05 | <.05 | <.05 | <.05 | .16 | <.05 | .07 | .07 |
Management and outcomes
Inadequate coverage and side effects were the most common reasons for a change in antimicrobial therapy, regardless of immune status. Thirty-three percent of all patients experienced complications, 67% of which were medical (nonsurgical). Medical complications were most commonly due to antibiotic-related side effects, and surgical complications most commonly involved wound-healing problems. Complications were seen more in immunosuppressed patients (55%) compared to immunocompetent patients (16%; P < .05). The 4 patients who underwent amputation (3 females, 1 male) had an average age of 71 years (range, 63–78 years). Three of these patients were immunosuppressed (rheumatoid arthritis, n = 2; transplant, n = 1) and 2 had insulin-dependent diabetes mellitus (1 with concurrent immunosuppression).
Discussion
In the United States, the reported incidence of upper-extremity NTMI is increasing, but still may be underdiagnosed.
18
In 2018, a retrospective review of culture-positive tenosynovial or intraarticular NTMIs (including M. marinum) of the upper extremity found that patients saw an average of 5 physicians prior to receiving an accurate diagnosis, with a mean time to diagnosis of 10 months. Diagnoses of all NTMIs are often delayed because of a lack of awareness or knowledge of these pathogens, an indolent clinical course, and difficulty with inadequate testing and/or organism isolation.10
,20
In our cohort, an increased frequency of infection during the study period could be due to a combination of several factors: ours is a tertiary referral center with a large population of immune-compromised patients (solid organ, stem cell transplant) and an increasing numbers of patients on novel biologics with expanding indications. Additionally, over the past 2 decades, novel molecular identification techniques have become available. Our practice has evolved to routinely send suspicious tissue specimens for bacterial, fungal, and mycobacterial cultures, and to obtain early infectious diseases consultations. We suspect that experience accumulated over the years by both hand surgeons and infectious diseases physicians at our institution has also contributed to better awareness and quicker diagnoses.Nonmarinum NTM are also now recognized as notable pathogens in immunocompetent hosts. Hellinger et al
23
published 6 cases of localized soft tissue infection with Mycobacterium avium or Mycobacterium intracellulare complex (MAC) in immunocompetent patients, concluding that soft tissue inoculation of MAC can cause a localized infection and that degenerative changes associated with aging may be important in allowing the infection to progress (average age, 64 years). Other studies have also evaluated the environment (farmers, marine workers, aquatic exposure) and reported mechanism of injury as potential risk factors.10
,24
Forty-eight percent of patients in our study reported an initial injury. Among patients with an initial injury, 8 (22%) were infected with MAC, compared to 11 (28%) without a reported injury. We did note a higher incidence of injuries in immunocompetent patients, a finding also reported by Sotello et al.15
Our study differed from this latter study because we did not include patients with M. marinum infections and all our patients underwent surgical debridement, whereas 50% of immunosuppressed patients did not have surgery in Sotello et al’s15
cohort. They also found similar treatment outcomes between immunocompetent and immunosuppressed patient groups, whereas we found a higher complication rate in immunosuppressed patients (55% vs 16% in immunocompetent patients). This may be due to greater virulence and more toxic antimicrobial side effects with treating nonmarinum NTMI in our group of patients.In our study, the average symptom duration prior to presentation was 206 days. Kozin et al
11
analyzed 33 patients with culture-positive NTMI (including M. marinum) of the upper extremity, and found an average delay between symptom onset and correct diagnosis of 1 year (range, 1 month to 8 years). Anatomically, most reported NTMIs involving the hand present as tenosynovitis, fasciitis, septic arthritis, or localized cutaneous involvement, with tenosynovitis being most frequently reported.11
,23
,25
Multiple case reports have been published that illustrate the various presentations of NTMIs, including indolent flexor tenosynovitis, osteomyelitis, abscess, and granulomatous synovitis.1
,8
,11
,23
,26
, 27
, 28
, 29
, 30
, 31
, 32
, 33
, 34
, 35
, 36
In our study, cutaneous or subcutaneous lesions (68%) were most frequent, and most commonly involved the wrist, hand, and digit (n = 46; 58%), consistent with prior studies.9
,10
Although there is a wide spectrum of clinical presentations and no clearly defined pathognomonic findings, NTMI should be considered for an inflammatory-like process that persists after several weeks or months despite treatment.10
,12
,13
There should be a low threshold for biopsy, atypical mycobacterial cultures, and polymerase chain reaction assay of a persistently painful, swollen lesion, particularly if the patient is immunosuppressed or has a history of penetrating trauma or other interventions. C-reactive protein is a very sensitive inflammatory marker and is often elevated in these patients. It was also noted to be higher in patients infected with rapid-growing subspecies (median, 5.7 vs 4.9 with slow-growing subspecies). We postulated that it might indicate a more aggressive inflammatory response in those patients; however, the erythrocyte sedimentation rate was higher in patients infected with slow-growing organisms (median, 21.3 vs 15.5 in rapid-growing organisms). Unfortunately, our small cohort size did not permit a statistical analysis of these differences. Once an NTMI is confirmed thorough debridement (if not accomplished at the initial biopsy procedure), appropriate antimicrobial therapy is required. Serial magnetic resonance imaging scans with intravenous gadolinium can also be helpful in identifying early recurrent bone, joint, or tenosynovial disease. These can be helpful in focusing redebridement, which may help improve antibiotic efficacy or support a change in the antibiotic regimen. Patients should be cautioned about a prolonged course of therapy (mean, 184 days), a high incidence of side effects (46%), the need to change antibiotic coverage (74%), a high rate of complications (32%), and a high rate of recurrent infection (31%).In our study, tissue samples were sent for histopathology and bacterial, fungal, and mycobacterial cultures, as well as broad-range polymerase chain reaction. An infectious disease specialist was consulted before surgery, and directly supervised the flow of specimens. Interestingly, the most commonly isolated mycobacteria in our study were Mycobacterium chelonae and M. abscessus (23/77; 30%), both rapidly growing NTMs. There is little published information comparing clinical features and treatment outcomes of infections caused by rapid- versus slow-growing organisms. We were able to identify some differences in our study, although not with respect to disease recurrence or complication rates. Again, since these infections are uncommon, a relatively small cohort size did not allow for statistical analyses of some comparisons.
A multidisciplinary approach involving surgeons, infectious diseases specialists, and a microbiology laboratory with capability for growth and identification of mycobacteria is key to success. In our study, 62 patients (81%) had inappropriate treatment prior to presentation at our institution, including localized corticosteroid injections and antimicrobial therapy directed toward routine bacterial infections. Current recommendations for treatment of invasive NTMIs include a combination of debridement and antibiotics.
37
, 38
, 39
, 40
In our cohort, all patients underwent surgical debridement; of those, 4 ultimately underwent amputation for disease eradication. In 2017, Lopez et al18
published a retrospective review of 34 patients with positive NTM cultures of the upper extremity, and all patients underwent surgical intervention and received antibiotic therapy for a mean duration of 5 months. In our cohort, immunocompromised patients were treated longer (mean, 243 days [8.1 months] with antimicrobial therapy versus 155 days [5.2 months] in immunocompetent patients).Several institutions have described changes in their practices due to the increasing prevalence of NTMIs in the upper extremity, even in immunocompetent patients.
4
,18
One group recommends a separate cutaneous biopsy for atypical mycobacterial tissue culture in patients who present with an upper-extremity lesion with atypical features or who have failed to resolve with appropriate antibiotic treatment for common skin organisms, arguing this facilitates a more timely diagnosis.15
Another proposed a management algorithm that would start antimicrobial therapy immediately upon initial suspicion of infection from clinical symptoms or pathologic findings.19
,27
Given the need for a prolonged duration of combination antimicrobial therapy for NTMIs that differs based on the species, and due to accompanying side effects of such a therapy, we discourage empiric antimicrobial therapy; obtaining surgical or radiographically directed specimens for culture and histopathology is crucial before initiating antimicrobials. Rapid-growing mycobacteria, with their shorter replication half-life, tend to present in a more acute fashion, with a shorter latency compared to slowly growing mycobacteria. In our series, isolated cutaneous and subcutaneous lesions were more common with rapid-growing species. This, combined with a higher proportion of immunosuppression in patients with these rapid-growing species, may have led to earlier consideration of atypical pathogens and use of more invasive procedures to confirm a diagnosis. The shorter duration of therapy was likely based on the clinical response, combined with the adequacy of debridement when performed. Since there are no clear recommendations for the duration of therapy, management needs to be individualized.Our study has several limitations. Since our study was retrospective, all aspects of the presentation, symptoms, and clinical course may not have been accurately captured, due to reliance on medical records that may not have been complete for every patient. This also prevented an a priori sample size estimate to determine the number of patients needed for many of the comparisons made in this study. Consequently, it is quite possible that a null hypothesis based on any comparisons is at risk for a type 2 error (false negative). It is quite possible that a larger cohort of patients could reveal significant differences (P < .05). Changes in laboratory techniques and NTM nomenclature may also have led to misclassification of certain species of NTM. We excluded M. marinum infections because these are known to occur often after a marine-related injury involving the hand; because of our geographical location, we do not see many of these patients. Mycobacterium marinum infections have been abundantly analyzed in published studies with respect to causative and patient-related factors, as well as treatments.
14
, 15
, 16
, 17
, 18
, 19
, 20
Finally, any inciting injury or event, as well as the durations of symptoms that patients noted in our study, are susceptible to recall bias.References
- Recurrent granulomatous tenosynovitis of the wrist and finger caused by Mycobacterium intracellulare: a case report.Diagn Microbiol Infect Dis. 2006; 56: 99-101
- Treatment of nontuberculous mycobacterial pulmonary disease: an official ATS/ERS/ESCMID/IDSA clinical practice guideline.Clin Infect Dis. 2020; 71: e1-e36
- Nontuberculous mycobacteria.in: Quah S.R. International Encyclopedia of Public Health. 2nd ed. Vol.5. Academic Press, 2017: 257-263
- Increased incidence of cutaneous nontuberculous mycobacterial infection, 1980–2009: a population-based study.Mayo Clin Proc. 2013; 88: 38-45
- Mycobacterium szulgai–a new pathogen.Tubercle. 1972; 53: 210-214
- Synchronous infection with Mycobacterium chelonae and Paecilomyces in a heart transplant patient.Transpl Infect Dis. 2011; 13: 80-83
- Mycobacterial infections in renal transplant recipients. Seven cases and a review of the literature.Arch Intern Med. 1982; 142: 888-892
- Mycobacterium avium complex tenosynovitis of the wrist and hand.Arthritis Rheum. 2004; 51: 140-142
- Extrapulmonary infections associated with nontuberculous mycobacteria in immunocompetent persons.Emerg Infect Dis. 2009; 15 (quiz 1544): 1351-1358
- Non-tuberculous mycobacterial infection of the musculoskeletal system: pattern of infection and efficacy of combined surgical/antimicrobial treatment.Bone Joint J. 2014; 96-B: 1561-1565
- Atypical mycobacterium infections of the upper extremity.J Hand Surg Am. 1994; 19: 480-487
- Non-tuberculous mycobacterial infections of the hand.Chir Main. 2015; 34: 18-23
- Non-tuberculous mycobacterial tenosynovitis: a review.Scand J Infect Dis. 1999; 31: 221-228
- Review article: Mycobacterium marinum infection of the hand and wrist.J Orthop Surg (Hong Kong). 2010; 18: 98-103
- Nontuberculous mycobacterial infections of the upper extremity: 15-year experience at a tertiary care medical center.J Hand Surg Am. 2018; 43: 387.e1-387.e8
- Twenty-eight cases of Mycobacterium marinum infection: retrospective case series and literature review.Infection. 2015; 43: 655-662
- Mycobacterium marinum: current recommended pharmacologic therapy.J Hand Surg Am. 2009; 34: 1734-1735
- Atypical mycobacterial infections of the upper extremity: becoming more atypical?.Hand (N Y). 2017; 12: 188-192
- Atypical mycobacterial infections of the upper extremity.Orthopedics. 2018; 41: e383-e388
- Mycobacterium marinum infection: a case report.J Venom Anim Toxins Incl Trop Dis. 2015; 21: 7
- Anonymous mycobacteria in pulmonary disease.Med Clin North Am. 1959; 43: 273-290
- Non-tuberculous mycobacteria and the rise of Mycobacterium abscessus.Nat Rev Microbiol. 2020; 18: 392-407
- Localized soft-tissue infections with Mycobacterium avium/Mycobacterium intracellulare complex in immunocompetent patients: granulomatous tenosynovitis of the hand or wrist.Clin Infect Dis. 1995; 21: 65-69
- Clinical manifestations, diagnosis, and treatment of Mycobacterium haemophilum infections.Clin Microbiol Rev. 2011; 24: 701-717
- Mycobacterium fortuitum midpalmar space abscess: a case report.J Hand Surg Am. 1986; 11: 438-440
- Nontuberculous mycobacterial infections: a clinical review.Infection. 2004; 32: 257-270
- Tenosynovitis due to Mycobacterium avium intracellulare and Mycobacterium chelonei: report of two cases with review of the literature.Clin Rheumatol. 1998; 17: 152-156
- Atypical mycobacterium infections of the upper extremity.Arch Orthop Trauma Surg. 2005; 125: 475-478
- Mycobacterium intracellulare as a cause of a recurrent granulomatous tenosynovitis of the hand.Diagn Microbiol Infect Dis. 2000; 38: 127-129
- Mycobacterium kansasii osteomyelitis of the scaphoid.J Hand Surg Am. 1987; 12: 1092-1094
- Osteomyelitis and skin ulcers caused by Mycobacterium szulgai in an AIDS patient.Scand J Infect Dis. 2004; 36: 883-885
- Multifocal osteomyelitis due to Mycobacterium szulgai in a patient with chronic lymphocytic leukemia.J Infect. 2008; 56: 151-154
- Osteomyelitis and synovitis produced by Mycobacterium marinum in a fisherman.J Clin Microbiol. 1990; 28: 2570-2572
- Primary nontraumatic Mycobacterium avium complex osteomyelitis of the distal phalanx.Am J Orthop (Belle Mead NJ). 2004; 33: 248-249
- Infection of prosthetic arthroplasty by Mycobacterium fortuitum. Two case reports.J Bone Joint Surg Am. 1979; 61: 300-302
- Mycobacterium fortuitum granulomatous synovitis caused by a dog bite.J Hand Surg Am. 1983; 8: 342-343
- Radical surgical debridement alone for treatment of carpal tunnel syndrome caused by Mycobacterium avium complex flexor tenosynovitis: case report.J Hand Surg Am. 2011; 36: 1047-1051
- Flexor tenosynovitis caused by Mycobacterium arupense.J Hand Surg Eur Vol. 2011; 36: 72-73
- Mycobacterium monacense: a mycobacterial pathogen that causes infection of the hand.J Hand Surg Am. 2008; 33: 94-96
- An official ATS/IDSA statement: diagnosis, treatment, and prevention of nontuberculous mycobacterial diseases.Am J Respir Crit Care Med. 2007; 175: 367-416
Article info
Publication history
Published online: May 28, 2022
Accepted:
March 16,
2022
Received:
March 23,
2021
Publication stage
In Press Corrected ProofFootnotes
No benefits in any form have been received or will be received related directly or indirectly to the subject of this article.
Identification
Copyright
© 2022 by the American Society for Surgery of the Hand. All rights reserved.
