ORIGINAL
ARTICLE
Diagnosis of deep cutaneous fungal infections: Correlation between skin tissue culture and histopathology Tania M. Gonzalez Santiago, MD,a Bobbi Pritt, MD,b Lawrence E. Gibson, MD,a,b and Nneka I. Comfere, MDa,b Rochester, Minnesota Background: Deep cutaneous fungal infections (DCFIs) are responsible for significant morbidity and mortality, particularly in immunocompromised patients. Although a direct correlation between histopathologic examination and culture is expected, discordant findings may be seen, presenting a unique diagnostic and therapeutic challenge. Objectives: We sought to determine the correlation between skin tissue cultures and histopathologic examination in patients with DCFI. Methods: This is a 10-year retrospective review (2003-2012) of patients with a diagnosis of DCFI seen at a single tertiary care institution. Tissue cultures and histopathologic findings were reviewed. Results: In 8 of 33 cases, fungal elements were seen on routine histopathologic sections but skin cultures were negative. Three of 8 of the discordant cases had concurrent positive noneskin tissue cultures that correlated with the pathology interpretation, and 3 of 8 patients in the discordant group died of systemic fungal infection. Limitations: This was a retrospective study design and a single tertiary care institution experience. Conclusions: The histopathologic interpretation of skin tissue specimens is critical for rapid and accurate diagnosis of DCFI. Despite the identification of fungal organisms on histopathologic assessment of skin tissue specimens, skin tissue culture may fail to show fungal growth. A diagnosis of a DCFI and initiation of appropriate treatment should always be considered in spite of discordant results. ( J Am Acad Dermatol http://dx.doi.org/10.1016/j.jaad.2014.03.042.)
D
eep cutaneous fungal infections (DCFIs) are associated with significant morbidity and mortality, especially in immunocompromised patients. Mortality rates range from 4% to 10% in localized infections and can be as high as 83% to 94% in disseminated disease.1 The clinical presentation of DCFI is variable and dependent on host-related factors, the type of fungal organism, and the mode of transmission.2,3 Because of the nonspecificity of presenting clinical symptoms, cutaneous lesions may be misdiagnosed as cutaneous neoplasms or necrotizing lesions caused by coagulation disorders. In addition, initial presentations in From the Departments of Dermatologya and Laboratory Medicine and Pathology,b Mayo Clinic, College of Medicine, Rochester. Funding sources: None. Conflicts of interest: None declared. Accepted for publication March 30, 2014. Reprints not available from the authors.
Abbreviations used: BAL: CSF: DCFI: GMS: MPA: PAS: PBSCT: SLE:
bronchoalveolar lavage cerebrospinal fluid deep cutaneous fungal infection Grocott methenamine silver microscopic polyangiitis periodic acideSchiff peripheral blood stem cell transplant systemic lupus erythematosus
the skin may herald the onset of a life-threatening systemic mycosis.4,5 Therefore, the rapid diagnosis and characterization of the offending fungus is Correspondence to: Nneka I. Comfere, MD, Department of Dermatology, Mayo Clinic, 200 First St SW, Rochester, MN 55905. E-mail: [email protected]. Published online May 15, 2014. 0190-9622/$36.00 Ó 2014 by the American Academy of Dermatology, Inc. http://dx.doi.org/10.1016/j.jaad.2014.03.042
1
J AM ACAD DERMATOL
2 Gonzalez Santiago et al
defined as those with fungal organisms identified on essential for appropriate treatment, which carries histopathology and without growth of fungal significant prognostic implications. Routine histopathologic examination of lesional organisms in skin tissue culture. The discordant skin tissue remains the primary method of confirming group was defined this way given the differences a DCFI. It permits fast, presumptive identification of in turnaround time, with return of histopathology fungi and can also provide some insight into the results preceding tissue culture results. diagnostic implication of some culture isolates. Because of the variable and RESULTS longer turnaround time on CAPSULE SUMMARY Patient characteristics tissue culture results relative Twenty-six of 33 patients to routine histopathology, Deep cutaneous fungal infections are were in an immunosupthe latter is often relied on responsible for significant morbidity and pressed state. Thirteen of 26 for the rapid diagnosis of mortality. immunosuppressed patients DCFI. Although direct were also transplant patients correlation between these 2 Discordant findings between (9/13 with a history of solid diagnostic tests is expected, histopathology and culture results are organ transplant and 4/13 discrepancies between histooften seen and present a diagnostic and with bone marrow/stem cell pathology and culture results therapeutic challenge. transplant). Five patients had are sometimes seen,6 with Increased awareness about these an underlying lymphoproliresultant treatment delays, diagnostic pitfalls may prevent adverse ferative disorder without a morbidity, and mortality. consequences associated with delays in history of transplantation. Few studies have investidiagnosis and treatment, especially in The remaining immunogated the correlation beimmunosuppressed patients. suppressed patients had tween the results of an underlying malignancy histopathologic examination (10/22). Three patients had of skin tissue specimens and a history of autoimmune disorders managed with tissue cultures. The aim of this review is to describe long-term systemic corticosteroids, including 1 case the clinical, histopathologic, and microbiologic each of microscopic polyangiitis (MPA), psoriasis, findings in a series of patients with a diagnosis and systemic lupus erythematosus (SLE). Other of DCFI and characterize features predictive of identified comorbid conditions included HIV, discordance between the lesional histopathology sarcoidosis, diabetes, and occupational exposure and skin tissue culture. (in a patient who worked at a turkey processing plant). Seven of 33 patients were otherwise healthy. METHODS Healthy individuals were immunocompetent and We performed a 10-year retrospective review of had no identifiable underlying systemic predisposiall histopathologic specimens diagnosed as DCFIs tion for opportunistic infections. Twenty-one of 33 and their corresponding skin culture results in the patients had a primary cutaneous mycosis, while 12 Department of Dermatology, Mayo Clinic, Rochester, of 33 patients had a systemic mycosis with secondary Minnesota between 2003 and 2012. Forty-six DCFI cutaneous involvement (Tables I-III). cases were identified by a search of our CoPath (pathology information system) database. Thirteen Clinical findings cases of DCFI without concurrent skin tissue culture Areas of involvement included the upper extremfor fungi were excluded from analysis. Thirty-three ities (15/33), lower extremities (10/33), trunk (3/33), patients with a diagnosis of DCFI were included. The multiple sites of skin involvement (3/33), and the electronic medical record for included cases was head and neck region (2/33). Clinical presentations reviewed and the following data were abstracted: encompassed nodules (22/33), ulcerated nodules patient age, sex, location of the sampled skin lesion, (4/33), plaques (4/33), ulcers (2/33), and erythemaunderlying medical comorbidities, histopathologic tous macules in 1 patient. Nine of 33 patients interpretation, histochemical stains, skin microbipresented with clinical evidence of tissue necrosis ology results, and antifungal therapy. The subset of (Part A of Figs 1 to 3). cases with discordance between the pathology diagnosis and tissue culture results (discordant Histopathologic findings cases) were identified, and review of the lesional In all cases (33/33), fungal elements were identiskin pathology was performed by a board-certified fied on routine hematoxylineeosin (H&E) stained dermatopathologist (N.C.). Discordant cases were d
d
d
J AM ACAD DERMATOL
Gonzalez Santiago et al 3
Table I. Patient characteristics Characteristic
Gender Male Female Median age, y Immunologic status Immunosuppressed Nonimmunosuppressed
Table II. Comorbid conditions n (%)
24 (72.7) 9 (27.3) 54 26 (78.7) 7 (21.2)
skin sections, and in 29 of 33 cases fungal stains (ie, Grocott methenamine silver and/or periodic acideSchiff) were used to confirm the diagnosis (Parts B and C of Figs 1 and 2 and Parts B to D of Fig 3). Fungi identified on H&E sections included zygomycosis, Alternaria, phaeohyphomycosis, blastomycosis, Coccidioides, hyalohyphomycosis, and Trichophyton. The predominant histopathologic patterns were granulomatous inflammation (18/33), pseudoepitheliomatous hyperplasia (2/33), perivascular and interstitial inflammation without granulomas (1/33), small vessel vasculitis (1/33), and necrosis (1/33). In 10 of 33 cases, a mixture of histopathologic patterns was seen (Table IV). Microbiologic findings In 25 of 33 cases, the skin tissue culture revealed fungal growth, while in 8 of 33 cases skin cultures were negative. Organisms identified in tissue cultures included Blastomyces dermatitidis, Alternaria, Rhizopus, Fusarium, Acremonium, Pseudoallescheria, Trichophyton mentagrophytes, Coccidioides immitis, and Aspergillus. In addition, 7 of 12 patients with systemic mycosis also had a noneskin tissue culture performed (ie, cerebrospinal fluid [CSF], brain tissue, or bronchoalveolar lavage [BAL]). Four of the 7 patients with noneskin tissue cultures had a concomitant positive skin tissue culture (Table V). Fungal identification Fungal organisms were identified within routine histologic sections in 23 of 33 cases, either of a specific fungal species or within a fungal class. Eighteen of 23 had the same identified fungus on histopathology and tissue culture results; however, 5 of 23 were misclassified on histopathology. Discordant cases In 8 of 33 cases, skin tissue cultures were negative despite the identification of fungal organisms on histopathology. Six of 8 patients were immunosuppressed, with equal numbers of cases with systemic
Condition
Oligoastrocytoma (grade 2) Lymphoproliferative disorders Chronic myelogenous leukemia (allogeneic PBSCT) Follicular lymphoma (grade 3) Hodgkin lymphoma (autologous PBSCT) Acute myelocytic leukemia (allogeneic PBSCT) Chronic lymphocytic leukemia Chronic lymphocytic leukemia (stem cell transplant) Mantle cell lymphoma B-cell acute lymphocytic leukemia Organ transplant Kidney, pancreas Heart Lung Liver Kidney Kidney, liver, and heart Iatrogenic immunosuppression* Microscopic polyangiitis Psoriasis Systemic lupus erythematosus Miscellaneousconditions and exposures Occupational (turkey meateprocessing plant employee) Sarcoidosis Diabetes HIV/AIDS Total Healthy individuals
n (%)
1 9 (27.3) 1 1 1 2 1 1 1 1 9 (27.3) 3 2 1 1 1 1 3 (9) 1 1 1 4 (12) 1 1 1 1 26 (78.7) 7 (21)
PBSCT, Peripheral blood stem cell transplant. *Iatrogenic immunosuppression includes long-term systemic steroids.
Table III. Primary versus secondary infection according to immunologic state n (%)
Primary deep cutaneous mycosis Immunosuppressed Nonimmunosuppressed Systemic mycosis with secondary cutaneous involvement Immunosuppressed Nonimmunosuppressed
21 (63.6) 15 7 12 (36.4) 7 4
mycosis and secondary cutaneous involvement (3/6) and primary cutaneous mycosis (3/6). Two of 8 patients were not immunosuppressed, and both had primary cutaneous mycosis. A noneskin tissue culture was performed in 3 of 8 cases. This was positive for all 3 cases with the following findings:
4 Gonzalez Santiago et al
Fig 1. A, Zygomycosis. Purpuric, indurated, plaques (solid arrows) on the abdomen of a patient with chronic lymphocytic leukemia, posteperipheral blood stem cell transplantation. B and C, Hematoxylineeosin staining revealed small, intravascular, hyaline, fungal hyphal forms (dashed arrows). In this case, skin cultures were negative despite the identification of intravascular hyphae. This patient died of disease. (Original magnification: B and C, 320.)
Aspergillus spp on CSF culture and a BAL, and Exophiala jeanselmei in a body fluid secretion from a skin lesion. In 2 of these 3 cases, the fungal organisms were correctly identified on histopathology (Table VI).
DISCUSSION Over a 10-year period at a single tertiary care facility, we identified 33 patients with a histopathologic diagnosis of a DCFI. While the incidence of DCFI in the general population is unknown, it is more common in the immunocompromised
J AM ACAD DERMATOL
Fig 2. A, Disseminated aspergillosis. Purpuric plaque (solid arrow) with associated bulla (dashed arrow) on the chest of a patient with heart transplant. B, Hematoxylineeosin staining revealed multiple 458 (dashed arrow), branching hyphae (solid arrows) associated with suppurative inflammation and necrosis. C, Grocott methenamine silver staining of lesional skin demonstrates branching fungal hyphae and budding (blue arrows). In this case, skin cultures were negative despite these histopathologic findings. Diagnosis was confirmed with CSF cultures. Patient died of disease. (Original magnification: B and C, 340.)
population, occurring in 20% to 30% of organ transplant recipients, with the highest rates noted within the first 2 years posttransplantation. The observed frequency of invasive fungal infections has been estimated to range between 20% and 30% in patients with acute leukemia, 10% and 15% in patients with lymphoma, and 5% in patients with other malignancies.7-11 In this study, the number of identified DCFI cases is lower than estimates in the
J AM ACAD DERMATOL
Gonzalez Santiago et al 5
Fig 3. A, Primary cutaneous alternariosis and microscopic polyangiitis. Ulcerated purpuric plaques in a sporotrichoid distribution on the leg (arrows). Patient was chronically immunosuppressed, with a history of microscopic polyangiitis. B and C, Hematoxylineeosin staining revealed suppurative and granulomatous inflammation (solid arrows) with focal tissue necrosis (dashed arrows). D, Grocott methenamine silver staining revealed fungal elements (arrows). Skin cultures were positive for Alternaria spp. This patient was treated and cured. (Original magnification: B, 320; C and D, 340.)
literature, especially in light of notable increases in the incidence of DCFI in parallel with the increased use of aggressive antimicrobial, chemotherapeutic, immunomodulatory, and transplantation-related therapies.1 The most common comorbid condition in our study population was an impaired immunologic status. Half of our patients were solid organ transplant recipients or had underlying hematologic malignancies and had received bone marrow or stem cell transplantation. Among immunosuppressed patients, neutropenia (neutrophil counts \1000 cells/mL for $ 7 days) has been described as a
predisposing factor for DCFIs. Other risk factors include diabetes mellitus, renal failure and renal transplantation, chronic corticosteroid therapy, and immunosuppressive therapy.6,12,13 A high index of suspicion for a DCFI should be maintained in immunosuppressed patients, particularly when they present with persistent fever and unresponsiveness to antimicrobial agents.8 Primary cutaneous fungal infections are most commonly attributed to traumatic inoculation of the fungi or foreign material containing the fungi into the skin.7 Most of our patients (21/33) had a primary cutaneous mycosis without systemic signs or
J AM ACAD DERMATOL
6 Gonzalez Santiago et al
Table IV. Histopathologic features n (%)
Histochemical stains* Performed Not performed Predominant histopathologic pattern Granulomatous inflammation Pseudoepitheliomatous hyperplasia Perivascular and interstitial inflammation without granulomas Vasculitis Necrosis Combination of histopathologic patterns
28 (84.8) 5 (15.2) 18 (54.5) 2 (6) 1 (3) 1 (3) 1 (3) 10 (30)
*Including Grocott methenamine silver and periodic acideSchiff stains.
Table V. Skin culture results n/N or n (%)
Positive skin culture (%) Blastomyces dermatitidis Alternaria spp Rhizopus spp Fusarium spp Acremonium spp Pseudoallescheria Trichophyton mentagrophytes Coccidioides immitis Aspergillus Negative skin culture Other, noneskin tissue culture Positive noneskin tissue culture, positive skin culture CSF culture (Alternaria spp)* BAL (Blastomyces dermatitidis)* BAL (Coccidioides immitis)* Positive noneskin tissue culture, negative skin culture Brain culture (Aspergillus spp) Body fluid secretion (Exophiala jeanselmei) BAL (Aspergillus spp)
25/33 (76) 9 (27) 5 (15) 3 (9) 3 (9) 1 (3) 1 (3) 1 (3) 1 (3) 1 (3) 8/33 (24) 7 (21) 4 (57) 1 2 1 3 (43) 1 1 1
BAL, Bronchoalveolar lavage; CSF, cerebrospinal fluid. *Same organism for both skin and noneskin tissue cultures.
symptoms. Occupational exposures may also represent important risk factors, because 1 of our patients with a diagnosis of disseminated blastomycosis worked in a turkey meateprocessing factory. The clinical diagnosis of a DCFI can be challenging. In most cases, while an infection may be suspected, a specific infectious agent cannot be implicated solely on the basis of the clinical findings. Skin lesions are nonspecific, requiring adjunctive pathologic examination and skin tissue cultures for
diagnostic confirmation. Necrosis is seen when the offending fungi have angioinvasive properties, such as Mucorales. Although histopathology is often needed to confirm the specific etiologic fungal organism, the clinical presence of necrosis should enable clinical consideration of a DCFI in the differential diagnosis. Histopathologic findings in DCFIs can vary, and close inspection of serial sections may be necessary to identify the offending organism(s). A granulomatous inflammatory pattern is often described as the most common histopathologic feature.14 This was confirmed in our study, where a majority of cases (18/33) presented primarily with this pattern. Histochemical stains (ie, Grocott methenamine silver and/or periodic acideSchiff) were used to assist with pathologic interpretation in most cases (28/33). In only 5 of 33 cases, identification of the fungal organisms (ie, hyaline and dematiaceous fungi, Zygomycetes, and Aspergillus spp) was possible based solely on histopathologic review, without the need for adjunctive histochemical staining. High rates of concordance between the pathologic diagnosis and skin tissue culture results are reflective of the relatively high diagnostic accuracy of histopathology for DCFIs in our practice. These concordance rates are supported by limited studies on culture (skin and noneskin tissue) and pathology (histopathology and cytopathology) correlations showing that overall diagnostic accuracy for microscopic morphologic techniques ranges from 20% to 80%.6,15,16 Guarner and Brandt6 attributed discordance between histopathology and tissue culture to the following: (1) alteration of fungal characteristics because of antifungal medications or host responses; (2) lack of experience of the pathologist in fungal identification; (3) differences in fungal morphology caused by fragmentation of the fungal elements with tissue processing; (4) inflammatory response obscuring fungal morphology; (5) similarities among different fungal species; and (6) only 1 fungus growing in culture in a dual infection where 1 is more abundant.17 It is also important to note that not all cases of phaeohyphomycosis feature pigmented hyphae (ie, they can appear hyaline). It is unclear to what extent $ 1 of the above factors may have contributed to the misclassification (5/33) noted between pathology and tissue culture in our case series. For discordant cases, the skin tissue culture was negative despite the identification of fungal elements in histopathologic sections. However, culture of a nonskin tissue in 3 of 8 discordant cases revealed growth of the offending fungal organism. As expected, these patients (2/3) fared poorly and
J AM ACAD DERMATOL
Table VI. Discordant cases Age (y), sex
60, M 61, M* 60, F 75, M*
70, M* 45, M*
44, M* 31, M*
Histopathologic findings Clinical findings
Underlying disease(s)
Ulcerated nodules Erythematous nodulesy Erythematous nodules Erythematous nodules
Diabetes Heart transplant None Lymphoproliferative disorder
Erythematous nodulesy Necrotic nodules
Lymphoproliferative disorder Lymphoproliferative disorder
Erythematous papules Necrotic plaquey
Heart, kidney, and liver transplant Lymphoproliferative disorder
Other noneskin tissue culture
Description
Stains
Fungi
Granulomatous inflammation PEH
Positive GMS and PAS Not performed
Zygomycetes
N/A
Aspergillus
Granulomatous inflammation Granulomatous inflammation
Positive GMS
NS
Aspergillus spp on CSF N/A
Positive GMS
NS
Panniculitis
Positive GMS
Granulomatous inflammation and vasculitis Granulomatous inflammation Vasculitis
Positive PAS
Hyaline fungus Zygomycetes
Positive GMS Negative
Dematiaceous fungus Zygomycetes
Exophiala jeanselmei on skin secretion N/A
Treatment
Outcome
Voriconazole and amphotericin B Voriconazole and amphotericin B Excision and debridement Itraconazole
Lesion resolution Death Lesion resolution Lesion resolution
Amphotericin B
Death
N/A
AmBisone and Voriconazole
Lesion resolution
N/A
Itraconazole
Aspergillus spp on BAL
AmBisone
Lesion resolution Death
Gonzalez Santiago et al 7
AmBisone is a registered trademark of Astellas Pharma (Tokyo, Japan). BAL, Bronchoalveolar lavage; CSF, cerebrospinal fluid; GMS, Grocott methenamine silver; N/A, not available; NS, not specified; PAS, periodic acideSchiff; PEH, pseudoepitheliomatous hyperplasia. *Immunosuppressed. y Systemic infection.
8 Gonzalez Santiago et al
died of systemic aspergillosis, while the remaining patient with localized E jeanselmei infection (identified on skin secretions but not on skin culture) was successfully treated. In the remaining 5 discordant cases, treatment was given empirically based on the histomorphologic description of the fungal organism. These findings suggest that the skin culture, although essential for the diagnosis and accurate identification of a DCFI, may fail to provide needed confirmation of the histopathology when negative. Our results found that skin cultures could be negative despite histopathologic evidence of fungal organisms in skin tissue sections. Possible explanations for this phenomenon have been described and include: (1) homogenization of tissue; (2) nonviable fungal organisms in the tissue, such as walled-off infections with dimorphic fungi; (3) tissue sampling from 2 different areas6; and (4) the use of preservative-containing local anesthesia.18 Homogenization of the tissue refers to the process of chemically or mechanically ‘‘reducing’’ samples into small particles. Mechanical techniques are used for most skin cultures and include grinding, shearing, beating, and sonification of the tissue before implantation on culture media.19 Some fungi, especially Mucorales and Aspergillus spp, are particularly prone to destruction with this technique, resulting in a negative skin culture.6 These techniques may have contributed to positive nontissue cultures (ie, CSF, body secretion, and BAL) in our discordant cases, because these samples are typically directly planted onto the culture medium without homogenization. In our discordant group, 3 cases were described as Mucorales on histologic sections, making this a possible explanation for the discordant skin culture result (note that the terms zygomycosis and mucormycosis are often used interchangeably). For this reason, planting larger samples of tissue on fungal culture media has been proposed.6,17 In addition, the growth pattern for each genus varies. Organisms of the Mucorales order should grow within 3 days, while others, such as Histoplasma capsulatum and Paracoccidioides brasiliensis, grow slowly in most laboratories. Therefore, in certain cases, a longer incubation time may be required to increase skin culture specificity. The clinician must convey the clinical suspicion for a DCFI to the microbiology laboratory so that the skin tissue specimen may be processed with caution and longer incubation times may be allowed, if necessary. Other possible explanations for our discordant results include sampling error when multiple biopsy specimens are obtained, with the skin tissue
J AM ACAD DERMATOL
containing the fungal organisms sent to pathology while the second sample—possibly not containing fungi or containing nonviable fungi—sent to microbiology.6 In addition, when used inadvertently, the fungicidal preservative methylparaben can directly affect the culture results.18 Clinicians should also be aware that although tissue cultures and histopathologic investigations are required for an accurate diagnosis, in some cases a diagnosis could be made immediately by staining the roof of a bulla or examining a potassium hydroxide preparation of a biopsy specimen.6 Frozen sections for faster results should be considered in severe cases. Study limitations included the fact that this was a retrospective review of a small number of cases (n = 33) from a single institution. There is a high likelihood that we underestimated the number of cases given that we only included patients with histopathologic evidence of a fungal infection and excluded those in whom a concurrent skin culture was not performed. None of our patients reported a history of traumatic injury preceding the diagnosis of a DCFI. This could be reflective of recall bias, because most of our patients were evaluated on a referral basis. We conclude that a diagnosis of DCFI should be considered in the differential diagnosis of nonspecific cutaneous lesions with supportive histopathology despite a negative skin tissue culture. In situ hybridizationebased diagnostic tests have been found to be useful for the diagnosis of invasive mycosis. Although not yet widely available, the development and implementation of these techniques may be useful. Increased awareness within the dermatology community about the histopathologic and microbiologic diagnostic pitfalls is necessary. This can prevent adverse consequences associated with delays in diagnosis and treatment, particularly in immunosuppressed patients. REFERENCES 1. Skiada A, Petrikkos G. Cutaneous zygomycosis. Clin Microbiol Infect 2009;15(suppl 5):41-5. 2. Hinshaw M, Longley JB. Fungal diseases. In: Elder DE, Elenitsas R, Johnson BL, Murphy GF, editors. Lever’s histopathology of the skin. 9th ed Philadelphia (PA): Lippincott Williams and Wilkins; 2005. pp. 601-34. 3. Vera-Cabrera L, Salinas-Carmona MC, Waksman N, Messeguer-Perez J, Ocampo-Candiani J, Welsh O. Host defenses in subcutaneous mycoses. Clin Dermatol 2012;30:382-8. 4. McNeil MM, Brown JM. The medically important aerobic actinomycetes: epidemiology and microbiology. Clin Microbiol Rev 1994;7:357-417. 5. van Burik JA, Colven R, Spach DH. Cutaneous aspergillosis. J Clin Microbiol 1998;36:3115-21. 6. Guarner J, Brandt M. Histopathologic diagnosis of fungal infections in the 21st century. Clin Microbiol Rev 2011;24: 247-80.
J AM ACAD DERMATOL
7. Kim MS, Lee SM, Sung HS, Won CH, Chang S, Lee MW, et al. Clinical analysis of deep cutaneous mycoses: a 12-year experience at a single institution. Mycoses 2012;55:501-6. 8. Singh N. Fungal infection in the recipients of solid organ transplantation. Infect Dis Clin North Am 2003;17:113. 9. Patterson JE. Epidemiology of fungal infections in solid organ transplants recipients. Transpl Infect Dis 1999;1:229. 10. Gabardi S, Kubiak DW, Chandraker AK, Tullius SG. Invasive fungal infections and antifungal therapies in solid organ transplant recipients. Transpl Int 2007;20:993. 11. Miller J, Kappe R, Kubitza D, Fessler R, Scheidecker I. The incidence of deep-seated mycoses in Freiburg. Mycoses 1988; 31(suppl 1):9-18. 12. Bassiri Jahromi S, Khaksar AA. Deep-seated fungal infections in immunocompromised patients in Iran. Iran J Allergy Asthma Immunol 2005;4:27-32. 13. Hart PD, Russell E Jr, Remington JS. The compromised host and deep fungal infection. J Infect Dis 1969;120: 169-91.
Gonzalez Santiago et al 9
14. Pang KR, Wu JJ, Huang DB, Tyring SK. Subcutaneous fungal infections. Dermatol Ther 2004;17:523-31. 15. Schofield CM, Murray CK, Horvath EE, Cancio LC, Kim SH, Wolf SE, et al. Correlation of culture with histopathology in fungal burn wound colonization and infection. Burns 2007;33:341-6. 16. Tarrand JJ, Lichterfeld M, Warraich I, Luna M, Han XY, May GS, et al. Diagnosis of invasive septate mold infections. A correlation of microbiological culture and histologic or cytologic examination. Am J Clin Pathol 2003;119:854-8. 17. Sangoi AR, Rogers WM, Longacre TA, Montoya JG, Baron EJ, Banaei N. Challenges and pitfalls of morphologic identification of fungal infections in histologic and cytologic specimens: a ten-year retrospective review at a single institution. Am J Clin Pathol 2009;131:364-75. 18. Miller MA, Shelley WB. Antibacterial properties of lidocaine on bacteria isolated from dermal lesions. Arch Dermatol 1985;121: 1157-9. 19. Burden DW. Guide to the homogenization of biological samples. Random Primers 2008;7:1-14.
ARTICLE
Diagnosis of deep cutaneous fungal infections: Correlation between skin tissue culture and histopathology Tania M. Gonzalez Santiago, MD,a Bobbi Pritt, MD,b Lawrence E. Gibson, MD,a,b and Nneka I. Comfere, MDa,b Rochester, Minnesota Background: Deep cutaneous fungal infections (DCFIs) are responsible for significant morbidity and mortality, particularly in immunocompromised patients. Although a direct correlation between histopathologic examination and culture is expected, discordant findings may be seen, presenting a unique diagnostic and therapeutic challenge. Objectives: We sought to determine the correlation between skin tissue cultures and histopathologic examination in patients with DCFI. Methods: This is a 10-year retrospective review (2003-2012) of patients with a diagnosis of DCFI seen at a single tertiary care institution. Tissue cultures and histopathologic findings were reviewed. Results: In 8 of 33 cases, fungal elements were seen on routine histopathologic sections but skin cultures were negative. Three of 8 of the discordant cases had concurrent positive noneskin tissue cultures that correlated with the pathology interpretation, and 3 of 8 patients in the discordant group died of systemic fungal infection. Limitations: This was a retrospective study design and a single tertiary care institution experience. Conclusions: The histopathologic interpretation of skin tissue specimens is critical for rapid and accurate diagnosis of DCFI. Despite the identification of fungal organisms on histopathologic assessment of skin tissue specimens, skin tissue culture may fail to show fungal growth. A diagnosis of a DCFI and initiation of appropriate treatment should always be considered in spite of discordant results. ( J Am Acad Dermatol http://dx.doi.org/10.1016/j.jaad.2014.03.042.)
D
eep cutaneous fungal infections (DCFIs) are associated with significant morbidity and mortality, especially in immunocompromised patients. Mortality rates range from 4% to 10% in localized infections and can be as high as 83% to 94% in disseminated disease.1 The clinical presentation of DCFI is variable and dependent on host-related factors, the type of fungal organism, and the mode of transmission.2,3 Because of the nonspecificity of presenting clinical symptoms, cutaneous lesions may be misdiagnosed as cutaneous neoplasms or necrotizing lesions caused by coagulation disorders. In addition, initial presentations in From the Departments of Dermatologya and Laboratory Medicine and Pathology,b Mayo Clinic, College of Medicine, Rochester. Funding sources: None. Conflicts of interest: None declared. Accepted for publication March 30, 2014. Reprints not available from the authors.
Abbreviations used: BAL: CSF: DCFI: GMS: MPA: PAS: PBSCT: SLE:
bronchoalveolar lavage cerebrospinal fluid deep cutaneous fungal infection Grocott methenamine silver microscopic polyangiitis periodic acideSchiff peripheral blood stem cell transplant systemic lupus erythematosus
the skin may herald the onset of a life-threatening systemic mycosis.4,5 Therefore, the rapid diagnosis and characterization of the offending fungus is Correspondence to: Nneka I. Comfere, MD, Department of Dermatology, Mayo Clinic, 200 First St SW, Rochester, MN 55905. E-mail: [email protected]. Published online May 15, 2014. 0190-9622/$36.00 Ó 2014 by the American Academy of Dermatology, Inc. http://dx.doi.org/10.1016/j.jaad.2014.03.042
1
J AM ACAD DERMATOL
2 Gonzalez Santiago et al
defined as those with fungal organisms identified on essential for appropriate treatment, which carries histopathology and without growth of fungal significant prognostic implications. Routine histopathologic examination of lesional organisms in skin tissue culture. The discordant skin tissue remains the primary method of confirming group was defined this way given the differences a DCFI. It permits fast, presumptive identification of in turnaround time, with return of histopathology fungi and can also provide some insight into the results preceding tissue culture results. diagnostic implication of some culture isolates. Because of the variable and RESULTS longer turnaround time on CAPSULE SUMMARY Patient characteristics tissue culture results relative Twenty-six of 33 patients to routine histopathology, Deep cutaneous fungal infections are were in an immunosupthe latter is often relied on responsible for significant morbidity and pressed state. Thirteen of 26 for the rapid diagnosis of mortality. immunosuppressed patients DCFI. Although direct were also transplant patients correlation between these 2 Discordant findings between (9/13 with a history of solid diagnostic tests is expected, histopathology and culture results are organ transplant and 4/13 discrepancies between histooften seen and present a diagnostic and with bone marrow/stem cell pathology and culture results therapeutic challenge. transplant). Five patients had are sometimes seen,6 with Increased awareness about these an underlying lymphoproliresultant treatment delays, diagnostic pitfalls may prevent adverse ferative disorder without a morbidity, and mortality. consequences associated with delays in history of transplantation. Few studies have investidiagnosis and treatment, especially in The remaining immunogated the correlation beimmunosuppressed patients. suppressed patients had tween the results of an underlying malignancy histopathologic examination (10/22). Three patients had of skin tissue specimens and a history of autoimmune disorders managed with tissue cultures. The aim of this review is to describe long-term systemic corticosteroids, including 1 case the clinical, histopathologic, and microbiologic each of microscopic polyangiitis (MPA), psoriasis, findings in a series of patients with a diagnosis and systemic lupus erythematosus (SLE). Other of DCFI and characterize features predictive of identified comorbid conditions included HIV, discordance between the lesional histopathology sarcoidosis, diabetes, and occupational exposure and skin tissue culture. (in a patient who worked at a turkey processing plant). Seven of 33 patients were otherwise healthy. METHODS Healthy individuals were immunocompetent and We performed a 10-year retrospective review of had no identifiable underlying systemic predisposiall histopathologic specimens diagnosed as DCFIs tion for opportunistic infections. Twenty-one of 33 and their corresponding skin culture results in the patients had a primary cutaneous mycosis, while 12 Department of Dermatology, Mayo Clinic, Rochester, of 33 patients had a systemic mycosis with secondary Minnesota between 2003 and 2012. Forty-six DCFI cutaneous involvement (Tables I-III). cases were identified by a search of our CoPath (pathology information system) database. Thirteen Clinical findings cases of DCFI without concurrent skin tissue culture Areas of involvement included the upper extremfor fungi were excluded from analysis. Thirty-three ities (15/33), lower extremities (10/33), trunk (3/33), patients with a diagnosis of DCFI were included. The multiple sites of skin involvement (3/33), and the electronic medical record for included cases was head and neck region (2/33). Clinical presentations reviewed and the following data were abstracted: encompassed nodules (22/33), ulcerated nodules patient age, sex, location of the sampled skin lesion, (4/33), plaques (4/33), ulcers (2/33), and erythemaunderlying medical comorbidities, histopathologic tous macules in 1 patient. Nine of 33 patients interpretation, histochemical stains, skin microbipresented with clinical evidence of tissue necrosis ology results, and antifungal therapy. The subset of (Part A of Figs 1 to 3). cases with discordance between the pathology diagnosis and tissue culture results (discordant Histopathologic findings cases) were identified, and review of the lesional In all cases (33/33), fungal elements were identiskin pathology was performed by a board-certified fied on routine hematoxylineeosin (H&E) stained dermatopathologist (N.C.). Discordant cases were d
d
d
J AM ACAD DERMATOL
Gonzalez Santiago et al 3
Table I. Patient characteristics Characteristic
Gender Male Female Median age, y Immunologic status Immunosuppressed Nonimmunosuppressed
Table II. Comorbid conditions n (%)
24 (72.7) 9 (27.3) 54 26 (78.7) 7 (21.2)
skin sections, and in 29 of 33 cases fungal stains (ie, Grocott methenamine silver and/or periodic acideSchiff) were used to confirm the diagnosis (Parts B and C of Figs 1 and 2 and Parts B to D of Fig 3). Fungi identified on H&E sections included zygomycosis, Alternaria, phaeohyphomycosis, blastomycosis, Coccidioides, hyalohyphomycosis, and Trichophyton. The predominant histopathologic patterns were granulomatous inflammation (18/33), pseudoepitheliomatous hyperplasia (2/33), perivascular and interstitial inflammation without granulomas (1/33), small vessel vasculitis (1/33), and necrosis (1/33). In 10 of 33 cases, a mixture of histopathologic patterns was seen (Table IV). Microbiologic findings In 25 of 33 cases, the skin tissue culture revealed fungal growth, while in 8 of 33 cases skin cultures were negative. Organisms identified in tissue cultures included Blastomyces dermatitidis, Alternaria, Rhizopus, Fusarium, Acremonium, Pseudoallescheria, Trichophyton mentagrophytes, Coccidioides immitis, and Aspergillus. In addition, 7 of 12 patients with systemic mycosis also had a noneskin tissue culture performed (ie, cerebrospinal fluid [CSF], brain tissue, or bronchoalveolar lavage [BAL]). Four of the 7 patients with noneskin tissue cultures had a concomitant positive skin tissue culture (Table V). Fungal identification Fungal organisms were identified within routine histologic sections in 23 of 33 cases, either of a specific fungal species or within a fungal class. Eighteen of 23 had the same identified fungus on histopathology and tissue culture results; however, 5 of 23 were misclassified on histopathology. Discordant cases In 8 of 33 cases, skin tissue cultures were negative despite the identification of fungal organisms on histopathology. Six of 8 patients were immunosuppressed, with equal numbers of cases with systemic
Condition
Oligoastrocytoma (grade 2) Lymphoproliferative disorders Chronic myelogenous leukemia (allogeneic PBSCT) Follicular lymphoma (grade 3) Hodgkin lymphoma (autologous PBSCT) Acute myelocytic leukemia (allogeneic PBSCT) Chronic lymphocytic leukemia Chronic lymphocytic leukemia (stem cell transplant) Mantle cell lymphoma B-cell acute lymphocytic leukemia Organ transplant Kidney, pancreas Heart Lung Liver Kidney Kidney, liver, and heart Iatrogenic immunosuppression* Microscopic polyangiitis Psoriasis Systemic lupus erythematosus Miscellaneousconditions and exposures Occupational (turkey meateprocessing plant employee) Sarcoidosis Diabetes HIV/AIDS Total Healthy individuals
n (%)
1 9 (27.3) 1 1 1 2 1 1 1 1 9 (27.3) 3 2 1 1 1 1 3 (9) 1 1 1 4 (12) 1 1 1 1 26 (78.7) 7 (21)
PBSCT, Peripheral blood stem cell transplant. *Iatrogenic immunosuppression includes long-term systemic steroids.
Table III. Primary versus secondary infection according to immunologic state n (%)
Primary deep cutaneous mycosis Immunosuppressed Nonimmunosuppressed Systemic mycosis with secondary cutaneous involvement Immunosuppressed Nonimmunosuppressed
21 (63.6) 15 7 12 (36.4) 7 4
mycosis and secondary cutaneous involvement (3/6) and primary cutaneous mycosis (3/6). Two of 8 patients were not immunosuppressed, and both had primary cutaneous mycosis. A noneskin tissue culture was performed in 3 of 8 cases. This was positive for all 3 cases with the following findings:
4 Gonzalez Santiago et al
Fig 1. A, Zygomycosis. Purpuric, indurated, plaques (solid arrows) on the abdomen of a patient with chronic lymphocytic leukemia, posteperipheral blood stem cell transplantation. B and C, Hematoxylineeosin staining revealed small, intravascular, hyaline, fungal hyphal forms (dashed arrows). In this case, skin cultures were negative despite the identification of intravascular hyphae. This patient died of disease. (Original magnification: B and C, 320.)
Aspergillus spp on CSF culture and a BAL, and Exophiala jeanselmei in a body fluid secretion from a skin lesion. In 2 of these 3 cases, the fungal organisms were correctly identified on histopathology (Table VI).
DISCUSSION Over a 10-year period at a single tertiary care facility, we identified 33 patients with a histopathologic diagnosis of a DCFI. While the incidence of DCFI in the general population is unknown, it is more common in the immunocompromised
J AM ACAD DERMATOL
Fig 2. A, Disseminated aspergillosis. Purpuric plaque (solid arrow) with associated bulla (dashed arrow) on the chest of a patient with heart transplant. B, Hematoxylineeosin staining revealed multiple 458 (dashed arrow), branching hyphae (solid arrows) associated with suppurative inflammation and necrosis. C, Grocott methenamine silver staining of lesional skin demonstrates branching fungal hyphae and budding (blue arrows). In this case, skin cultures were negative despite these histopathologic findings. Diagnosis was confirmed with CSF cultures. Patient died of disease. (Original magnification: B and C, 340.)
population, occurring in 20% to 30% of organ transplant recipients, with the highest rates noted within the first 2 years posttransplantation. The observed frequency of invasive fungal infections has been estimated to range between 20% and 30% in patients with acute leukemia, 10% and 15% in patients with lymphoma, and 5% in patients with other malignancies.7-11 In this study, the number of identified DCFI cases is lower than estimates in the
J AM ACAD DERMATOL
Gonzalez Santiago et al 5
Fig 3. A, Primary cutaneous alternariosis and microscopic polyangiitis. Ulcerated purpuric plaques in a sporotrichoid distribution on the leg (arrows). Patient was chronically immunosuppressed, with a history of microscopic polyangiitis. B and C, Hematoxylineeosin staining revealed suppurative and granulomatous inflammation (solid arrows) with focal tissue necrosis (dashed arrows). D, Grocott methenamine silver staining revealed fungal elements (arrows). Skin cultures were positive for Alternaria spp. This patient was treated and cured. (Original magnification: B, 320; C and D, 340.)
literature, especially in light of notable increases in the incidence of DCFI in parallel with the increased use of aggressive antimicrobial, chemotherapeutic, immunomodulatory, and transplantation-related therapies.1 The most common comorbid condition in our study population was an impaired immunologic status. Half of our patients were solid organ transplant recipients or had underlying hematologic malignancies and had received bone marrow or stem cell transplantation. Among immunosuppressed patients, neutropenia (neutrophil counts \1000 cells/mL for $ 7 days) has been described as a
predisposing factor for DCFIs. Other risk factors include diabetes mellitus, renal failure and renal transplantation, chronic corticosteroid therapy, and immunosuppressive therapy.6,12,13 A high index of suspicion for a DCFI should be maintained in immunosuppressed patients, particularly when they present with persistent fever and unresponsiveness to antimicrobial agents.8 Primary cutaneous fungal infections are most commonly attributed to traumatic inoculation of the fungi or foreign material containing the fungi into the skin.7 Most of our patients (21/33) had a primary cutaneous mycosis without systemic signs or
J AM ACAD DERMATOL
6 Gonzalez Santiago et al
Table IV. Histopathologic features n (%)
Histochemical stains* Performed Not performed Predominant histopathologic pattern Granulomatous inflammation Pseudoepitheliomatous hyperplasia Perivascular and interstitial inflammation without granulomas Vasculitis Necrosis Combination of histopathologic patterns
28 (84.8) 5 (15.2) 18 (54.5) 2 (6) 1 (3) 1 (3) 1 (3) 10 (30)
*Including Grocott methenamine silver and periodic acideSchiff stains.
Table V. Skin culture results n/N or n (%)
Positive skin culture (%) Blastomyces dermatitidis Alternaria spp Rhizopus spp Fusarium spp Acremonium spp Pseudoallescheria Trichophyton mentagrophytes Coccidioides immitis Aspergillus Negative skin culture Other, noneskin tissue culture Positive noneskin tissue culture, positive skin culture CSF culture (Alternaria spp)* BAL (Blastomyces dermatitidis)* BAL (Coccidioides immitis)* Positive noneskin tissue culture, negative skin culture Brain culture (Aspergillus spp) Body fluid secretion (Exophiala jeanselmei) BAL (Aspergillus spp)
25/33 (76) 9 (27) 5 (15) 3 (9) 3 (9) 1 (3) 1 (3) 1 (3) 1 (3) 1 (3) 8/33 (24) 7 (21) 4 (57) 1 2 1 3 (43) 1 1 1
BAL, Bronchoalveolar lavage; CSF, cerebrospinal fluid. *Same organism for both skin and noneskin tissue cultures.
symptoms. Occupational exposures may also represent important risk factors, because 1 of our patients with a diagnosis of disseminated blastomycosis worked in a turkey meateprocessing factory. The clinical diagnosis of a DCFI can be challenging. In most cases, while an infection may be suspected, a specific infectious agent cannot be implicated solely on the basis of the clinical findings. Skin lesions are nonspecific, requiring adjunctive pathologic examination and skin tissue cultures for
diagnostic confirmation. Necrosis is seen when the offending fungi have angioinvasive properties, such as Mucorales. Although histopathology is often needed to confirm the specific etiologic fungal organism, the clinical presence of necrosis should enable clinical consideration of a DCFI in the differential diagnosis. Histopathologic findings in DCFIs can vary, and close inspection of serial sections may be necessary to identify the offending organism(s). A granulomatous inflammatory pattern is often described as the most common histopathologic feature.14 This was confirmed in our study, where a majority of cases (18/33) presented primarily with this pattern. Histochemical stains (ie, Grocott methenamine silver and/or periodic acideSchiff) were used to assist with pathologic interpretation in most cases (28/33). In only 5 of 33 cases, identification of the fungal organisms (ie, hyaline and dematiaceous fungi, Zygomycetes, and Aspergillus spp) was possible based solely on histopathologic review, without the need for adjunctive histochemical staining. High rates of concordance between the pathologic diagnosis and skin tissue culture results are reflective of the relatively high diagnostic accuracy of histopathology for DCFIs in our practice. These concordance rates are supported by limited studies on culture (skin and noneskin tissue) and pathology (histopathology and cytopathology) correlations showing that overall diagnostic accuracy for microscopic morphologic techniques ranges from 20% to 80%.6,15,16 Guarner and Brandt6 attributed discordance between histopathology and tissue culture to the following: (1) alteration of fungal characteristics because of antifungal medications or host responses; (2) lack of experience of the pathologist in fungal identification; (3) differences in fungal morphology caused by fragmentation of the fungal elements with tissue processing; (4) inflammatory response obscuring fungal morphology; (5) similarities among different fungal species; and (6) only 1 fungus growing in culture in a dual infection where 1 is more abundant.17 It is also important to note that not all cases of phaeohyphomycosis feature pigmented hyphae (ie, they can appear hyaline). It is unclear to what extent $ 1 of the above factors may have contributed to the misclassification (5/33) noted between pathology and tissue culture in our case series. For discordant cases, the skin tissue culture was negative despite the identification of fungal elements in histopathologic sections. However, culture of a nonskin tissue in 3 of 8 discordant cases revealed growth of the offending fungal organism. As expected, these patients (2/3) fared poorly and
J AM ACAD DERMATOL
Table VI. Discordant cases Age (y), sex
60, M 61, M* 60, F 75, M*
70, M* 45, M*
44, M* 31, M*
Histopathologic findings Clinical findings
Underlying disease(s)
Ulcerated nodules Erythematous nodulesy Erythematous nodules Erythematous nodules
Diabetes Heart transplant None Lymphoproliferative disorder
Erythematous nodulesy Necrotic nodules
Lymphoproliferative disorder Lymphoproliferative disorder
Erythematous papules Necrotic plaquey
Heart, kidney, and liver transplant Lymphoproliferative disorder
Other noneskin tissue culture
Description
Stains
Fungi
Granulomatous inflammation PEH
Positive GMS and PAS Not performed
Zygomycetes
N/A
Aspergillus
Granulomatous inflammation Granulomatous inflammation
Positive GMS
NS
Aspergillus spp on CSF N/A
Positive GMS
NS
Panniculitis
Positive GMS
Granulomatous inflammation and vasculitis Granulomatous inflammation Vasculitis
Positive PAS
Hyaline fungus Zygomycetes
Positive GMS Negative
Dematiaceous fungus Zygomycetes
Exophiala jeanselmei on skin secretion N/A
Treatment
Outcome
Voriconazole and amphotericin B Voriconazole and amphotericin B Excision and debridement Itraconazole
Lesion resolution Death Lesion resolution Lesion resolution
Amphotericin B
Death
N/A
AmBisone and Voriconazole
Lesion resolution
N/A
Itraconazole
Aspergillus spp on BAL
AmBisone
Lesion resolution Death
Gonzalez Santiago et al 7
AmBisone is a registered trademark of Astellas Pharma (Tokyo, Japan). BAL, Bronchoalveolar lavage; CSF, cerebrospinal fluid; GMS, Grocott methenamine silver; N/A, not available; NS, not specified; PAS, periodic acideSchiff; PEH, pseudoepitheliomatous hyperplasia. *Immunosuppressed. y Systemic infection.
8 Gonzalez Santiago et al
died of systemic aspergillosis, while the remaining patient with localized E jeanselmei infection (identified on skin secretions but not on skin culture) was successfully treated. In the remaining 5 discordant cases, treatment was given empirically based on the histomorphologic description of the fungal organism. These findings suggest that the skin culture, although essential for the diagnosis and accurate identification of a DCFI, may fail to provide needed confirmation of the histopathology when negative. Our results found that skin cultures could be negative despite histopathologic evidence of fungal organisms in skin tissue sections. Possible explanations for this phenomenon have been described and include: (1) homogenization of tissue; (2) nonviable fungal organisms in the tissue, such as walled-off infections with dimorphic fungi; (3) tissue sampling from 2 different areas6; and (4) the use of preservative-containing local anesthesia.18 Homogenization of the tissue refers to the process of chemically or mechanically ‘‘reducing’’ samples into small particles. Mechanical techniques are used for most skin cultures and include grinding, shearing, beating, and sonification of the tissue before implantation on culture media.19 Some fungi, especially Mucorales and Aspergillus spp, are particularly prone to destruction with this technique, resulting in a negative skin culture.6 These techniques may have contributed to positive nontissue cultures (ie, CSF, body secretion, and BAL) in our discordant cases, because these samples are typically directly planted onto the culture medium without homogenization. In our discordant group, 3 cases were described as Mucorales on histologic sections, making this a possible explanation for the discordant skin culture result (note that the terms zygomycosis and mucormycosis are often used interchangeably). For this reason, planting larger samples of tissue on fungal culture media has been proposed.6,17 In addition, the growth pattern for each genus varies. Organisms of the Mucorales order should grow within 3 days, while others, such as Histoplasma capsulatum and Paracoccidioides brasiliensis, grow slowly in most laboratories. Therefore, in certain cases, a longer incubation time may be required to increase skin culture specificity. The clinician must convey the clinical suspicion for a DCFI to the microbiology laboratory so that the skin tissue specimen may be processed with caution and longer incubation times may be allowed, if necessary. Other possible explanations for our discordant results include sampling error when multiple biopsy specimens are obtained, with the skin tissue
J AM ACAD DERMATOL
containing the fungal organisms sent to pathology while the second sample—possibly not containing fungi or containing nonviable fungi—sent to microbiology.6 In addition, when used inadvertently, the fungicidal preservative methylparaben can directly affect the culture results.18 Clinicians should also be aware that although tissue cultures and histopathologic investigations are required for an accurate diagnosis, in some cases a diagnosis could be made immediately by staining the roof of a bulla or examining a potassium hydroxide preparation of a biopsy specimen.6 Frozen sections for faster results should be considered in severe cases. Study limitations included the fact that this was a retrospective review of a small number of cases (n = 33) from a single institution. There is a high likelihood that we underestimated the number of cases given that we only included patients with histopathologic evidence of a fungal infection and excluded those in whom a concurrent skin culture was not performed. None of our patients reported a history of traumatic injury preceding the diagnosis of a DCFI. This could be reflective of recall bias, because most of our patients were evaluated on a referral basis. We conclude that a diagnosis of DCFI should be considered in the differential diagnosis of nonspecific cutaneous lesions with supportive histopathology despite a negative skin tissue culture. In situ hybridizationebased diagnostic tests have been found to be useful for the diagnosis of invasive mycosis. Although not yet widely available, the development and implementation of these techniques may be useful. Increased awareness within the dermatology community about the histopathologic and microbiologic diagnostic pitfalls is necessary. This can prevent adverse consequences associated with delays in diagnosis and treatment, particularly in immunosuppressed patients. REFERENCES 1. Skiada A, Petrikkos G. Cutaneous zygomycosis. Clin Microbiol Infect 2009;15(suppl 5):41-5. 2. Hinshaw M, Longley JB. Fungal diseases. In: Elder DE, Elenitsas R, Johnson BL, Murphy GF, editors. Lever’s histopathology of the skin. 9th ed Philadelphia (PA): Lippincott Williams and Wilkins; 2005. pp. 601-34. 3. Vera-Cabrera L, Salinas-Carmona MC, Waksman N, Messeguer-Perez J, Ocampo-Candiani J, Welsh O. Host defenses in subcutaneous mycoses. Clin Dermatol 2012;30:382-8. 4. McNeil MM, Brown JM. The medically important aerobic actinomycetes: epidemiology and microbiology. Clin Microbiol Rev 1994;7:357-417. 5. van Burik JA, Colven R, Spach DH. Cutaneous aspergillosis. J Clin Microbiol 1998;36:3115-21. 6. Guarner J, Brandt M. Histopathologic diagnosis of fungal infections in the 21st century. Clin Microbiol Rev 2011;24: 247-80.
J AM ACAD DERMATOL
7. Kim MS, Lee SM, Sung HS, Won CH, Chang S, Lee MW, et al. Clinical analysis of deep cutaneous mycoses: a 12-year experience at a single institution. Mycoses 2012;55:501-6. 8. Singh N. Fungal infection in the recipients of solid organ transplantation. Infect Dis Clin North Am 2003;17:113. 9. Patterson JE. Epidemiology of fungal infections in solid organ transplants recipients. Transpl Infect Dis 1999;1:229. 10. Gabardi S, Kubiak DW, Chandraker AK, Tullius SG. Invasive fungal infections and antifungal therapies in solid organ transplant recipients. Transpl Int 2007;20:993. 11. Miller J, Kappe R, Kubitza D, Fessler R, Scheidecker I. The incidence of deep-seated mycoses in Freiburg. Mycoses 1988; 31(suppl 1):9-18. 12. Bassiri Jahromi S, Khaksar AA. Deep-seated fungal infections in immunocompromised patients in Iran. Iran J Allergy Asthma Immunol 2005;4:27-32. 13. Hart PD, Russell E Jr, Remington JS. The compromised host and deep fungal infection. J Infect Dis 1969;120: 169-91.
Gonzalez Santiago et al 9
14. Pang KR, Wu JJ, Huang DB, Tyring SK. Subcutaneous fungal infections. Dermatol Ther 2004;17:523-31. 15. Schofield CM, Murray CK, Horvath EE, Cancio LC, Kim SH, Wolf SE, et al. Correlation of culture with histopathology in fungal burn wound colonization and infection. Burns 2007;33:341-6. 16. Tarrand JJ, Lichterfeld M, Warraich I, Luna M, Han XY, May GS, et al. Diagnosis of invasive septate mold infections. A correlation of microbiological culture and histologic or cytologic examination. Am J Clin Pathol 2003;119:854-8. 17. Sangoi AR, Rogers WM, Longacre TA, Montoya JG, Baron EJ, Banaei N. Challenges and pitfalls of morphologic identification of fungal infections in histologic and cytologic specimens: a ten-year retrospective review at a single institution. Am J Clin Pathol 2009;131:364-75. 18. Miller MA, Shelley WB. Antibacterial properties of lidocaine on bacteria isolated from dermal lesions. Arch Dermatol 1985;121: 1157-9. 19. Burden DW. Guide to the homogenization of biological samples. Random Primers 2008;7:1-14.
