Mycobacterium Bolletii/Mycobacterium Massiliense Furunculosis Associated with Pedicure Footbaths a Report of 3 Cases

OBSERVATION Mycobacterium bolletii/Mycobacterium massiliense Furunculosis Associated With Pedicure Footbaths A Report of 3 Cases

Rebecca Wertman, BA; Melissa Miller, PhD; Pamela Groben, MD; Dean S. Morrell, MD; Donna A. Culton, MD, PhD

Background: Mycobacterium bolletii and Mycobacte- rative granulomatous dermatitis. Mycobacterium bolletii/M rium massiliense are recently described species of non- massiliense was identified by sequencing the 16S ribo- tuberculous mycobacteria. Footbaths preceding pedi- somal RNA (rRNA) and hsp65 genes. All 3 patients re- cures at nail salons have been implicated as reservoirs of sponded to different combinations of clarithromycin, doxy- infection with nontuberculous mycobacteria. To our cycline hydrochloride, azithromycin, and moxifloxacin knowledge, this case series represents the first docu- hydrochloride for complete lesion resolution. mented outbreak of M bolletii/M massiliense furunculosis, identified by heat-shock protein 65 gene, hsp65, se- Conclusions: Clinicians should elicit a history of pedi- quencing, occurring in immunocompetent patrons of a cure footbaths and maintain a high level of suspicion when North Carolina nail salon. faced with skin lesions of the lower extremities that are culture negative or are refractory to conventional anti- Observations: We describe 3 cases of lower extremity biotic therapy. Accurate identification and discrimina- furunculosis caused by M bolletii/M massiliense associated tion of M massiliense and M bolletii is difficult and re- with pedicure footbaths from the same North Carolina nail quires sequencing of multiple gene targets beyond their salon. Lesions developed within 1 month of the salon visit identical 16S rRNA sequences. and were characterized by erythematous, indurated papules and plaques. Histologic examination revealed suppu- Arch Dermatol. 2011;147(4):454-458

ONTUBERCULOUS MYCO- We report 3 cases of lower extremity fu- bacteria(NTM)aresignifi- runculosis caused by Mycobacterium bolletii/ cant human pathogens Mycobacterium massiliense, both recently de- causing systemic and cu- scribed species.13 The patients were patrons taneous disease. Recently, ofthesamenailsaloninNorthCarolinawhere theN incidence of cutaneous infections asso- they received footbaths and pedicures prior ciated with NTM has risen because of in- to the development of skin lesions. Mycobac- creased incidence of immunosuppression, teriumbolletii/Mmassiliensewasisolatedfrom increasednumberofsurgicalprocedures,and the lesions on 2 of the patients and identi- improved detection methods.1,2 The most fied by sequencing the 16S ribosomal RNA commonNTMassociatedwithcutaneousin- (rRNA) and hsp65 (65-kDa heat-shock pro- fections are Mycobacterium marinum and the tein) genes. Mycobacterial cultures from the rapidlygrowingmycobacteriaMycobacterium third patient showed no growth; however, fortuitum,Mycobacterium chelonae, and My- similarlesionsandpatronageofthesamenail cobacteriumabscessus.3-5 Rapidlygrowingmy- salon as the first 2 patients suggested cobacteria are ubiquitous in soil and water6,7 M bolletii/M massiliense as the likely infec- and have been detected in chlorinated mu- tious agent. nicipal water systems in the United States.8 Theycanbedifficulttoeradicatebecausethey REPORT OF CASES areresistanttomanydisinfectants.2 Infections have been most commonly reported following invasive cosmetic procedures such as CASE 1 Author Affiliations: liposuction and mammoplasty,9 intramus- Departments of Dermatology cular injections such as mesotherapy,10 and A 42-year-old white woman was referred (Ms Wertman and Drs Groben, penetrating trauma predisposing to environ- to our dermatology clinic for a 2-month Morrell, and Culton) and 7,9,11 Pathology and Laboratory mental contamination ; immunocompe- history of a plaque on her lower left leg. 7 Medicine (Dr Miller), tent patients are increasingly affected. The The lesion began as a small area that pro- University of North diagnosisisoftenoverlooked,2 andprolonged gressively enlarged. By the time of her visit Carolina–Chapel Hill. treatment can be necessary.4,12 to our clinic, she had developed a second

ARCH DERMATOL/ VOL 147 (NO. 4), APR 2011 WWW.ARCHDERMATOL.COM 454

©2011 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/26/2021 Figure 1. Lower leg of a 42-year-old white woman (case 1). The patient had a 2-month history of a 2-cm verrucous erythematous plaque and a more recent 4-mm erythematous papule.

Figure 2. Histologic images from the shin of a 23-year-old white woman lesion on the leg. She denied any fever, drainage from (case 2). Histologic specimen of a lesion shows a microabscess in the deep the site, or other skin lesions; the remainder of the re- dermis with surrounding mixed inflammatory infiltrate (hematoxylin-eosin, original magnification ϫ100). Inset image shows a Fite stain, which revealed view of systems was negative and she was in good health. acid-fast organisms within the microabscess (original magnification ϫ600). She reported having had 3 pedicures with footbaths on a monthly basis at the same nail salon before the lesions appeared, but could not recall the exact timing of the le- 2 weeks before the appearance of the pustules. She re- sion’s appearance. She reported shaving her legs approxi- ported shaving her legs regularly and had continued to mately once per week. shave despite the pustules. On physical examination, a 2-cm verrucous erythem- Approximately 8 papules in various stages of devel- atous plaque on her left shin and a more distal 4-mm ery- opment were present on both legs from her knees to her thematous papule (Figure 1) were present. No other ankles. The papules were firm, nonfluctuant, and with- lesions were noted. An 8-mm punch biopsy revealed sup- out drainage, with an osteumlike opening and surround- purative and granulomatous dermatitis with dilation and ing erythema. The remainder of the physical examina- rupture of multiple follicles. Special stains did not show tion was unremarkable. Treatment with clobetasol any fungal or acid-fast organisms. Mycobacterial cul- ointment, 0.05%, and clindamycin lotion, 1%, was ini- ture was performed on the tissue, and M bolletii/M tiated; however, the patient returned to the clinic ap- massiliense was isolated and identified by sequencing the proximately 1 month later reporting no clinical improve- 16S rRNA and hsp65 genes. ment. At that time multiple indurated, erythematous to Clarithromycin treatment was initiated, but no improve- violaceous nodules and plaques were noted, with some ment was apparent after 3 weeks. The woman was subse- central ulceration on the bilateral distal portions of her quently seen by an infectious disease physician and was in- legs. An 8-mm punch biopsy was performed, revealing structed to stop taking clarithromycin, as M bolletii is an abscess in the deep dermis with surrounding granu- frequently resistant to that drug; moxifloxacin hydrochlo- lomatous inflammation shown in the hematoxylin-eosin– ride, doxycycline hydrochloride, and azithromycin were stained sections (Figure 2). Several acid-fast rods were prescribed at that time. She stopped shaving her legs and found in the microabscess (Figure 2, inset). Mycobacte- refrained from pedicures. Results of antimicrobial suscep- rial cultures were performed on the tissues, and azithro- tibility tests showed the organism to be resistant to doxy- mycin was prescribed. cycline; it was therefore discontinued. Moxifloxacin and The patient returned to our clinic 2 weeks later with azithromycin therapy was continued for 6 months, dur- fluctuant nodules, some of which had spontaneously ing which time the patient’s lesions completely resolved. drained. Three of the nodules were incised and drained, expressing thick, purulent material. Mycobacterium CASE 2 bolletii/M massiliense was isolated from the original tissue culture and identified by sequencing the 16S rRNA A 23-year-old white woman was referred to our derma- and hsp65 genes. Doxycycline was added to the azithro- tology clinic for lesions that had been present on the bi- mycin treatment course. She subsequently visited an in- lateral lower legs for 4 months. The lesions began with fectious disease physician, who added moxifloxacin to 1 small pustule, which gradually expanded into mul- her medication regimen to be continued for 6 months. tiple pustules. These pustules did not drain and re- At the time of her last visit at completion of therapy, the solved spontaneously within a few weeks, with residual patient reported that the lesions were healing well. erythematous macules. The woman applied a corticosteroid cream without improvement. She reported no pru- CASE 3 ritus or drainage but stated that the lesions were tender. She was otherwise in good health. The patient reported A 14-year-old African American girl was referred to our regularly patronizing a nail salon for pedicures with foot- clinic for eruptions on her bilateral lower extremities. The baths and recalled that her most recent pedicure had been patient was initially seen by an outside dermatologist with

ARCH DERMATOL/ VOL 147 (NO. 4), APR 2011 WWW.ARCHDERMATOL.COM 455

Figure 3. Lower leg of 14-year-old African American girl (case 3). A, The patient presented with multiple papules on bilateral lower extremities with surrounding erythema and crusting. B, Postinflammatory hyperpigmentation at sites of previous involvement following multidrug treatment.

a 1-month history of tender, pruritic red papules. She first identified in 2006 and reported in 2009,15 isolated stated that she was continuously developing new le- from sputum from a patient with chronic pneumonia and sions that spontaneously resolved; at one time she had later from a group of patients with cystic fibrosis. The approximately 20 on each leg. She was treated with mi- species was initially described13 as resistant to clarithro- nocycline for presumed furunculosis but had no clini- mycin. Mycobacterium bolletii shares 100% 16S rRNA and cal improvement after 2 weeks. A biopsy was per- 95.6% rpoB gene sequence similarity with M abscessus.13 formed, and histologic examination revealed a necrotizing It is also closely related to M massiliense, which was iden- suppurative granulomatous dermatitis; acid-fast and fun- tified and classified as a member of the M chelonae–M ab- gal stains, as well as all cultures (bacterial, fungal, and scessus complex in 2004.16 Accurate identification and mycobacterial), were negative. The patient next tried clo- discrimination between M abscessus, M massiliense, and betasol ointment, 0.05%, without improvement and then M bolletii is difficult and requires sequencing of mul- topical corticosteroid tape, which flattened the existing tiple gene targets beyond their identical 16S rRNA se- lesions but did not prevent new lesions. quences.17 Zelazny et al17 demonstrated a multilocus se- At the patient’s initial visit to our clinic, the lesions quence analysis technique that included sequencing of had been present for approximately 4 months. She was rpoB, hsp65, and secA for the molecular identification of otherwise healthy. She reported patronizing a nail salon 42 clinical isolates that had previously been classified as for pedicures with footbaths about 1 month before the M abscessus. The authors found that 7 isolates were M appearance of the lesions. massiliense and 2 were M bolletii. Clinical manifesta- Multiple papules on both legs with surrounding ery- tions were similar among the patients. The authors rec- thema and crusting were noted, as well as several hyper- ommended this approach for identification of M absces- pigmented macules in locations where lesions had been pre- sus and closely related species, and it has recently been sent (Figure 3A). Empirical treatment with clarithromycin used10,18 in the classification of nosocomial outbreaks of and doxycycline was begun for a suspected mycobacterial M bolletii and M massiliense. infection until final results from the cultures performed by Few reported cases of M bolletii and M massiliense in- an outside laboratory were available. fection exist because of their recent identification. In 2008,14 The patient returned to our clinic 1 month later and both organisms were identified among South Korean iso- reported overall improvement, including flattening of the lates that were previously misclassified as M chelonae or M lesions and no further lesion development. She was given abscessus; of 144 isolates studied, 2 were reidentified as M hydroquinone, 4% cream, to decrease pigmentation of bolletii and 59 were reidentified as M massiliense. Identifi- the macules. After 4 months of treatment, she returned cation required sequencing of rpoB and hsp65 regions.14 In to our clinic and the lesions had resolved (Figure 3B). Brazil, 311 immunocompetent individuals developed cu- Although all cultures were negative, M bolletii/M taneous infections attributed to M abscessus after under- massiliense was suspected as the most likely source of the going invasive cosmetic procedures.10 Reidentification of lesions because of the similarity of her clinical presen- the isolates using hsp65 and rpoB sequencing classified 59 tation to that of the previously described 2 patients and of 67 isolates, all from surgical and postsurgical abscesses, patronage of the same nail salon for pedicures. as M massiliense and the remaining 8, all from patients who had undergone mesotherapy, as M bolletii. To our knowl- COMMENT edge, our case series represents the first documented outbreak of M bolletii/M massiliense furunculosis, identified by Mycobacterium bolletii is a recently described13 species of hsp65 sequencing, occurring in immunocompetent pa- NTM classified as a member of the M chelonae–M absces- trons of a nail salon in North Carolina. Additional sequenc- sus complex of rapidly growing mycobacteria. This com- ing studies would have to be performed to further identify plex is frequently a cause of cutaneous infections, nota- these isolates and differentiate between M bolletii and M bly after injections or minor surgical procedures,1,14 as massiliense. well as an opportunistic pathogen in patients with un- Mycobacterial furunculosis is characterized clini- derlying pulmonary disorders.1 Mycobacterium bolletii was cally by an initially benign nodular appearance, with a

ARCH DERMATOL/ VOL 147 (NO. 4), APR 2011 WWW.ARCHDERMATOL.COM 456

Shaved Patients, Legs, Mycobacterium species Source No. No. (%) (No. of Patients) Diagnostic Method Sensitivity Results Present study 3 3 (100) M bolletii/M massiliense AFB cultures followed by S: Clarithromycin, azithromycin sequencing of hsp65 R: TMP-SMX, ciprofloxacin, doxycycline, moxifloxacin Christie,20 2006 1 1 (100) M fortuitum AFB culture Unknown Redbord et al,19 2006 4 4 (100) M fortuitum/M peregrinum AFB culture S: Amikacin, imipenem, doxycycline, ciprofloxacin, gatifloxacin, moxifloxacin, TMP-SMX I: Clarithromycin, cefoxitin R: Tobramycin, azithromycin Grubb,21 2005 1 1 (100) M chelonae AFB culture S: Clarithromycin, amikacin Gira et al,11 2004 2 2 (100) M mageritense Bacterial culture (non-AFB) S: Amikacin, imipenem, sulfamethoxazole, fluoroquinolones, linezolid R: Clarithromycin Sniezek et al,2 2003 3 2 (67) M chelonae Bacterial and AFB cultures S: Clarithromycin, ciprofloxacin M fortuitum I: TMP-SMX R: Clarithromycin, TMP-SMX, amikacin, ciprofloxacin Winthrop et al,8 2002 34 24 (71) M fortuitum (32) Bacterial and AFB cultures Unknown Unidentified (2) followed by HPLC, PFGE, and MLEE

Abbreviations: AFB, acid-fast bacilli; HPLC, high-performance liquid chromatography; hsp65, 65-kDa heat-shock protein gene; I, intermediate; MLEE, multilocus enzyme electrophoresis; PFGE, pulsed-field gel electrophoresis; R, resistant; S, sensitive; TMP-SMX, trimethoprim-sulfamethoxazole.

protracted course that can result in scarring.19 Diagno- Skin microtrauma from shaving facilitates entry of the or- sis is often delayed by patients not seeking medical treat- ganisms from the contaminated footbaths. All 3 of our pa- ment soon after appearance of the lesions and by a lack tients reported leg shaving prior to their pedicures. of suspicion by the physician.19 Lesions typically de- We report 3 cases of furunculosis caused by M velop 1 to 2 months after introduction of the organ- bolletii/M massiliense from contaminated whirlpool foot- ism.19 Rapidly growing mycobacteria are ubiquitous in baths. Our report further documents the pathogenicity municipal water supplies,6-8 and recently, footbaths pre- of these newly identified species of rapidly growing ceding pedicures at nail salons have been implicated as mycobacteria. Accurate identification of M bolletii and reservoirs of infection, where the organisms thrive in the M massiliense and differentiation from the other species nutrient-rich water (Table).2,8,11,19-21 In 2002, Winthrop of the M chelonae–M abscessus complex often depends et al8 documented a community outbreak of lower ex- on multilocus gene sequencing and is important given tremity furunculosis in 110 patrons of a nail salon in Cali- the differences in antibiotic resistance patterns. Myco- fornia. The outbreak was linked to whirlpool footbaths bacterium bolletii is known to be multidrug resistant; contaminated with M fortuitum used before pedicures. thus, cultures and sensitivities should be performed. Sniezek et al2 later reported 2 cases of M fortuitum and 1 Punch biopsy samples are thought to be more sensitive case of M abscessus infection of the lower extremities fol- in capturing organisms for culture and are preferred lowing whirlpool footbaths and pedicures at 2 nail sa- over wound swab samples.22 Notably, cultures in 2 of lons in southern California. Gira et al11 reported 2 cases our 3 patients identified the organisms, but culture in of Mycobacterium mageritense furunculosis in women re- the third patient did not; antimicrobial susceptibility ceiving footbaths and pedicures at a salon in Atlanta. Iso- tests showed the organisms in the first 2 patients to be lates recovered from the footbath drains at the salon were susceptible to clarithromycin, although M bolletii is determined, by pulsed-field gel electrophoresis, to match thought to be resistant. However, in vitro sensitivities the patients’ isolates. Case reports in 200521 and 200620 may not accurately reflect in vivo activity, as the lesions documented lower extremity furunculosis resulting from in one of our patients did not clinically improve during footbaths and pedicures attributed to M chelonae and M clarithromycin treatment. Current treatment guidelines fortuitum, respectively. Redbord et al19 reported 4 cases recommend debridement of abscesses and 2-agent treat- of M fortuitum furunculosis after footbaths and pedi- ment with clarithromycin in combination with cipro- cures from 3 different salons around Cincinnati, Ohio. floxacin, doxycycline, cefoxitin, or amikacin if the Leg shaving is known to be a risk factor for acquiring organisms are shown to be susceptible to those drugs.1,2 infection from footbaths. Winthrop et al8 found that leg However, Winthrop et al22 demonstrated that infections shaving is significantly associated with infection (odds ra- can be treated effectively with oral antibiotic therapy tio, 4.8; 95% confidence interval, 2.1-11.1). Sniezek et al2 tailored to the organism and surgical resection of the reported that 2 of 3 patients had shaved their legs prior to area can be avoided. Clinicians should prescribe 2-agent the pedicure and Gira et al11 and Redbord et al19 reported antimicrobial coverage to avoid acquired drug resis- that 100% of patients had shaved prior to the pedicures. tance.22 All patients in our series responded to a combi-

ARCH DERMATOL/ VOL 147 (NO. 4), APR 2011 WWW.ARCHDERMATOL.COM 457

©2011 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/26/2021 nation of clarithromycin, doxycycline, azithromycin, nov., a novel rapidly growing species isolated from a cosmetic infection and from and moxifloxacin. a nail salon. Int J Syst Evol Microbiol. 2004;54(pt 6):2385-2391. 4. Dodiuk-Gad R, Dyachenko P, Ziv M, et al. Nontuberculous mycobacterial infec- It is likely that mycobacterial infections from whirl- tions of the skin: a retrospective study of 25 cases. J Am Acad Dermatol. 2007; pool footbaths are underrecognized and underreported. 57(3):413-420. Vugia et al23 conducted a survey of nail salons in the 5 5. Street ML, Umbert-Millet IJ, Roberts GD, Su WP. Nontuberculous mycobacte- largest counties across California and found that 97% of rial infections of the skin: report of fourteen cases and review of the literature. footbaths (29 of 30) were contaminated with NTM and J Am Acad Dermatol. 1991;24(2 pt 1):208-215. 6. Fabroni C, Buggiani G, Lotti T. Therapy of environmental mycobacterial infections. 50% carried more than 1 species. Mycobacterium fortui- Dermatol Ther. 2008;21(3):162-166. tum was the most prevalent. The study resulted in the 7. Palenque E. Skin disease and nontuberculous atypical mycobacteria. Int J Dermatol. adoption of regulations by the California Board of Bar- 2000;39(9):659-666. bering and Cosmetology for disinfection of the foot- 8. Winthrop KL, Abrams M, Yakrus M, et al. An outbreak of mycobacterial furun- 23 culosis associated with footbaths at a nail salon. N Engl J Med. 2002;346(18): baths. In light of the prevalence of NTM organisms in 1366-1371. whirlpool footbaths, clinicians must have a high level of 9. Murillo J, Torres J, Bofill L, et al; Venezuelan Collaborative Infectious and Tropi- suspicion when patients develop skin lesions of the lower cal Diseases Study Group. Skin and wound infection by rapidly growing myco- extremities that show no organisms on culture or are re- bacteria: an unexpected complication of liposuction and liposculpture. Arch fractory to conventional antibiotic therapy and must cor- Dermatol. 2000;136(11):1347-1352. 10. Viana-Niero C, Lima KV, Lopes ML, et al. Molecular characterization of Myco- rectly identify any NTM species cultured from lesions. bacterium massiliense and Mycobacterium bolletii in isolates collected from out- Public health measures, including educating the public breaks of infections after laparoscopic surgeries and cosmetic procedures. J Clin about the risks of infection after pedicure and increased Microbiol. 2008;46(3):850-855. regulation of the nail salon industry, can have a major 11. Gira AK, Reisenauer AH, Hammock L, et al. Furunculosis due to Mycobacterium role in the prevention of mycobacterial outbreaks. mageritense associated with footbaths at a nail salon. J Clin Microbiol. 2004; 42(4):1813-1817. 12. Escalonilla P, Esteban J, Soriano ML, et al. Cutaneous manifestations of infec- Accepted for Publication: July 1, 2010. tion by nontuberculous mycobacteria. Clin Exp Dermatol. 1998;23(5):214- Correspondence: Donna A. Culton, MD, PhD, Depart- 221. ment of Dermatology, University of North Carolina– 13. Ade´kambi T, Berger P, Raoult D, Drancourt M. rpoB gene sequence–based characterization of emerging non-tuberculous mycobacteria with descriptions of Chapel Hill, 402 Mary Ellen Jones Bldg, Campus Box Mycobacterium bolletii sp. nov., Mycobacterium phocaicum sp. nov. and My- 7287, Chapel Hill, NC 27599 ([email protected]). cobacterium aubagnense sp. nov. Int J Syst Evol Microbiol. 2006;56(pt 1): Author Contributions: Ms Wertman and Drs Morrell and 133-143. Culton had full access to all the data in the study and 14. Kim HY, Kook Y, Yun YJ, et al. Proportions of Mycobacterium massiliense and take responsibility for the integrity of the data and the Mycobacterium bolletii strains among Korean Mycobacterium chelonae- Mycobacterium abscessus group isolates. J Clin Microbiol. 2008;46(10):3384- accuracy of the data analysis. Study concept and design: 3390. Morrell and Culton. Acquisition of data: Wertman, Miller, 15. Ade´kambi T, Drancourt M. Mycobacterium bolletii respiratory infections. Emerg Morrell, Groben, and Culton. Analysis and interpreta- Infect Dis. 2009;15(2):302-305. tion of data: Wertman, Miller, Groben, and Culton. Draft- 16. Ade´kambi T, Reynaud-Gaubert M, Greub G, et al. Amoebal coculture of “Myco- bacterium massiliense” sp. nov. from the sputum of a patient with hemoptoic ing of the manuscript: Wertman, Morrell, and Culton. Criti- pneumonia. J Clin Microbiol. 2004;42(12):5493-5501. cal revision of the manuscript for important intellectual 17. Zelazny AM, Root JM, Shea YR, et al. Cohort study of molecular identification content: Miller, Morrell, Groben, and Culton. Adminis- and typing of Mycobacterium abscessus, Mycobacterium massiliense, and My- trative, technical, or material support: Miller, Morrell, Gro- cobacterium bolletii. J Clin Microbiol. 2009;47(7):1985-1995. ben, and Culton. Study supervision: Morrell, Groben, and 18. Kim HY, Yun YJ, Park CG, et al. Outbreak of Mycobacterium massiliense infection associated with intramuscular injections. J Clin Microbiol. 2007;45(9): Culton. 3127-3130. Financial Disclosure: None reported. 19. Redbord KP, Shearer DA, Gloster H, et al. Atypical Mycobacterium furunculosis occurring after pedicures. J Am Acad Dermatol. 2006;54(3):520-524. 20. Christie L. Lower extremity furunculosis in a female adolescent. Pediatr Infect REFERENCES Dis J. 2006;25(5):469, 473-474. 21. Grubb B. Think again about that pedicure. JAAPA. 2005;18(9):62. 1. Weitzul S, Eichhorn PJ, Pandya AG. Nontuberculous mycobacterial infections of 22. Winthrop KL, Albridge K, South D, et al. The clinical management and outcome the skin. Dermatol Clin. 2000;18(2):359-377, xi-xii. of nail salon-acquired Mycobacterium fortuitum skin infection. Clin Infect Dis. 2. Sniezek PJ, Graham BS, Busch HB, et al. Rapidly growing mycobacterial infec- 2004;38(1):38-44. tions after pedicures. Arch Dermatol. 2003;139(5):629-634. 23. Vugia DJ, Jang Y, Zizek C, Ely J, Winthrop KL, Desmond E. Mycobacteria in nail 3. Cooksey RC, de Waard JH, Yakrus MA, et al. Mycobacterium cosmeticum sp. salon whirlpool footbaths, California. Emerg Infect Dis. 2005;11(4):616-618.

ARCH DERMATOL/ VOL 147 (NO. 4), APR 2011 WWW.ARCHDERMATOL.COM 458