Direct detection and differentiation of causative fungi of onychomycosis by multiplex polymerase chain reaction-based assay

ARTICLE

Auteur(s) : Xiao-fang LI¹, Wei TIAN², Hong WANG², Hui CHEN¹, Yong-nian SHEN¹, Gui-xia LV¹, Wei-da LIU¹ liumyco@hotmail.com

¹ Department of Mycology, Institute of Dermatology, Chinese Academy of Medical Sciences & Peking Union Medical College, Jiang Wangmiao jie 14, Nanjing 210042, China

² Department of Dermatology, Provincial Hospital Affiliated to Shandong University, Jinan 250021, China

Reprints: W. LIU

Onychomycosis is a common nail disease that is responsible for up to 50% of diseases of the nail and 30% of all cases of dermatophytosis. The incidence of onychomycosis has increased over the last few years due to an increase in the number of immunocompromised patients, the extensive use of immunosuppressive chemotherapy, and increasing age and lifespan. This disease is increasingly recognized as a medical disorder that poses various physical, psychosocial, and occupational problems [1, 2]. Over 90% of onychomycoses are caused by dermatophytes (mainly Trichophyton rubrum) and, rarely, yeast and nondermatophytic molds [3, 4].

The treatment of onychomycosis has undoubtedly improved in recent years; many patients can now expect a complete and lasting cure following pharmacological therapy. Numerous nail conditions may, however, mimic onychomycosis. Long-term systemic antifungal treatment has potential side-effects and is expensive. Therefore, treatment should not be commenced before mycological confirmation of infection [5]. The three drugs currently licensed for general use in onychomycosis are griseofulvin, terbinafine and itraconazole, and the usage of antifungal agents with differing spectra of activity. Griseofulvin is effective only for dermatophytic infections, with no activity against yeast and non-dermophytic molds. Terbinafine and itraconazole are the most common drugs used for onychomycosis. Terbinafine is the most active agent currently available with fungicidal activity against dermatophytes, however, it must be considered in combination with other management for non-dermatophytic refractory or resistant yeast/mold infections as well. Itraconazole is the most effective agent for the treatment of onychomycosis caused by yeast [5-9]. Thus, diagnosis of onychomycosis should be closely linked to the identity of the causative agent, particularly in terms of whether it is a dermatophyte, or yeast, or non-dermatophytic mold.

The conventional methods used to diagnose onychomycosis are microscopy and fungal culture. Microscopy, although fast and economical, has a sensitivity dependent upon many factors, including the skill of the operator and the quality and quantity of nail samples obtained; besides, it does not provide genus or species identification. Culture isolates of pathogenic fungi are routinely identified by examination of colonial and microscopic morphology as well as several physiological properties. Unfortunately, all dermatophytes grow slowly, thus culturing is time-consuming, and difficulties are sometimes encountered due to phenotypic variations between strains [10, 11]. In fact, some isolates require the assistance of an experienced mycologist.

Because of those problems, many clinicians choose to initiate treatment without confirming diagnosis. This is costly and may expose the patient to the risk of side-effects unnecessarily. Therefore, there is a need for new laboratory techniques that are quick and highly specific.

Nail histopathology, immunochemistry, and flow cytometry have recently emerged as valuable tools. But these methods cannot be applied routinely because of their complexity, their cost and the fact that access is limited to the rare centers equipped for them [12].

Molecular biology-based techniques, especially PCR-based techniques, have frequently been used as reliable and sensitive tools for the early and accurate identification of important pathogens. The application of PCR technology directly to clinical specimens would allow early and accurate identification of agents of onychomycosis. This would permit prompt and targeted initiation of antifungal therapy. Recently, many attempts have been made to collect fungal DNA directly from affected nail samples, however, most identification systems are based on nested PCR or PCR-RFLP or sequencing techniques, et al., which lack the convenience of allowing identification of the causative agent by a single PCR and electrophoresis, and, thus, may increase the time taken for the laboratory identification of pathogens. Furthermore, the practical role of these methods in the diagnostic mycology laboratory is uncertain [13-18].

We describe here an alternative multiplex PCR-based method especially developed for the detection of dermatophyte, yeast and non-dermatophytic molds. By an extraction procedure directly from nail samples, followed by a single triplex PCR and electrophoresis, the method enables a rapid differentiation of those three groups of fungal strains, which could be of significance not only at diagnosis, but also for subsequent clinical treatment decision-making.

Materials and methods

Clinical specimens

Patients with nail disease at Institute of Dermatology, Chinese Academy of Medical Sciences & Peking Union Medical College were enrolled in the present study. Patients who had used an oral antifungal agent in the past 3 months or a topical antifungal agent in the past 2 weeks were excluded from the study.

The overall severity of the affected nails was presented according to the Scoring Clinical Index for Onychomycosis (SCIO) [19]. Nail shavings from the suspected portions were used as clinical specimens after swabbing liberally with alcohol (75%) to exclude the possibility of the presence of attached contaminants. The scrapings were separated into three parts. The first part of the specimen was examined by direct microscopy with 15% KOH. The second part was cultured onto SCCA (Sabouraud dextrose agar with chloramphenicol plus cycloheximide) and SCA (Sabouraud dextrose agar with chloramphenicol), respectively. The third part of the specimen was taken to the sterile centrifuge tubes for DNA extraction.

Media were incubated at 25 ̊C for 2 weeks. The medical mycology lab of China Culture Collection Center of Medical Mycology (CCCCM) performed the routine strain identification.

Organisms

Reference organisms selected to assay primer specificity include ten dermatophytes: T. rubrum (CCCCM T_1c), T. mentagrophytes (CCCCM T_5b), T. violaceum (CCCCM T_3e), T. tonsurans (CCCCM T_4b), T. schoenleini (CCCCM T_2a), T. concentricum (CCCCM T_6a) Microsporum canis (CCCCM M_3a), M. gypseum (CCCCM M_2b), M. ferrugineum (CCCCM M₁), Epidermophyton floccosum (CCCCM E_1d); eight yeasts: Candida albicans (CCCCM C_1c), C. glabrata (CCCCM S₂), C. parapsilosis (CCCCM C_4a), C. tropicalis (CCCCM C_2a), C. krusei (CCCCM C_6a), C. gualliermondii (CCCCM C_5a), Malassezia furfur (clinical isolate), Saccharomyces cerevisiae (CCCCM Y_8b); eight molds: Aspergillus fumigatus (CCCCM A₁), Penicillium islandicum (CCCCM B₂₇), Scopulariopsis brevicaulis (CCCCM B_4b), Fusarium solani (CCCCM B₂₄), Rhizomucor variabilis (CCCCM B₃), Rhizopus nigricans (CCCCM B₂), Alternaria alternata (CCCCM B₁), Fonsecaea pedrosoi (CCCCM D_6a); one mammal (human) and two prokaryotes: Escherichia coli (ATCC 25922) and Staphylococcus aureus (ATCC 29213). The fungal isolates with CCCCM accession numbers were obtained from the China Culture Collection Center of Medical Mycology; ATCC isolates were from the American Type Culture Collection.

DNA extraction

Genomic DNA from clinical specimens was extracted according to the method described by Turin and Makimura et al. [18, 20] with some modifications. Briefly, nail clippings were cut into small pieces, and washed three times with distilled water, then suspended in 500 μL lysis buffer (200 mmol L^-1 Tris-HCl [pH 8.0], 1% [wt/vol] sodium dodecyl sulfate, 250 mmol L^-1 NaCl, 25 mmol L^-1 EDTA) in 1.5 mL sterile Eppindorf tubes, and digested by incubation with proteinase K (100 μg mL^-1) (Sigma, Chemical Co.) for 1 h at 55 ̊C, then incubated at 100 ̊C for 15 min and mixed with 50 μL of 3.0 mol L^-1 sodium acetate, kept at –20 ̊C for 20 min, and then centrifuged at 10,000 g for 15 min. After phenol-chloroform extraction, DNA was precipitated with an equal volume of isopropanol at –20 ̊C for 10 min, washed with 99% ethanol, dried, and suspended in 50 μL of ultra pure water.

Reference strains underwent a similar procedure to extract DNA, and the washing step was optional, not mandatory. Negative control DNAs from prokaryotes (E. coli and S. aureus) and eukaryotes (human) were, respectively, extracted with MiniBEST Bacterial Genomic DNA Extraction Kit (TaKaRa) and Blood Genome DNA Extraction Kit (TaKaRa).

Primers

A careful primer selection for multiplex PCR application was done, assessing critical factors such as compatibility, in terms of not producing any additional bands or spurious hybridizations of primer pairs to each other in amplification reactions. Three pairs of primers were finally selected to perform the triplex PCR: universal (NS, forward 5′-AACTTAAAGGAATTGACGGAAG-3′, reverse 5′-GCATCACAGACCTGTTATTGCCTC-3′, amplified a 310-bp fragment) [21], dermatophyte specific (CHS1, forward 5′-CATCGAGTACATGTGCTCGC-3′, reverse 5′-CTCGAGGTCAAAAGCACGCC-3′, amplified a 450-bp fragment) [22] and yeast specific (ACT1, forward 5′-GATTTTGTCTGAACGTGGTAACA-3′, reverse 5′-GGAGTTGAAAGTGGTTTGGTCAATAC-3′, amplified a 271 bp fragment) [23].

The specificity of these primers was tested by PCR of DNA extracted from previously described organisms.

Multiplex PCR amplification

DNA templates prepared from three fungal strains: T. rubrum (T_1c), C. albicans (C_1b) and S. brevicaulis (B_4b) which representing dermatophyte, yeast and mold respectively, were used to assess the triplex PCR amplifications, and the optimal condition was as follows: 5 μL of 10 × reaction buffer, 3 μL MgCl₂ (25 mmol L^-1), 4 μL dNTPmix (2.5 mmol L^-1), the content of each primer (10 μmol l^-1) was NS 1 μL, CHS1 2 μL and ACT₁ 0.25 μL, 5 μL DNA template solution, 2.5 U of Hot Start Taq polymerase (Takara Shuzo Co. Ltd., Shiga, Japan). Ultrapure water was added to increase the volume to 50 μL. The PCR amplifications were performed in a GeneAmp 2400 PCR System (Perkin-Elmer, Norwalk, CT, USA). Each reaction mixture was heated to 94 ̊C for 5 min, and then was followed the following conditions: 94 ̊C for 30 s, 54 ̊C for 60 s, and 72 ̊C for 90 s for 36 cycles. The thermal cycles were terminated by polymerization at 72 ̊C for 10 min. DNA templates of clinical samples were amplified by the same procedure.

In order to detect PCR amplified fragments, 4 μL of PCR products were electrophoresed in 4% agarose gel in the presence of ethidium bromide and visualized under UV light.

The interpretation of the band pattern results was as follows: (i) bands of NS+ CHS1 + ACT₁ indicated the infection caused by dermatophyte + yeast + mold or dermatophyte + yeast; (ii) NS+ CHS1 indicated dermatophyte + mold or simple infection of dermatophyte; (iii) NS + ACT₁ indicated yeast + mold or yeast; (iv) NS alone indicated mold.

Sensitivity of the mPCR system on detection of simulated clinical samples

T. rubrum (T_1c) suspension was serially diluted by 10-fold, to give concentrations ranging from 10⁵ cfu/mL–10¹ cfu/mL, and mixed with 5 mg sterile human nail debris to produce simulated clinical samples. DNA extraction and PCR were performed by the aforementioned method to determine the lowest limit of detection (sensitivity).

Statistical analysis

In order to increase the reliability of the assay, the “gold standard” for the diagnosis of onychomycosis in this study was established if two of the three methods, e.g. microscopy, culture or PCR were positive, namely, true positive [24]. The data of sensitivity, specificity, positive predictive value, negative predictive value and accuracy for each method were calculated based on this diagnostic criterion. Statistical evaluation of those data was performed on each method using u-test (z-test). Agreement of each method with the “gold standard” was estimated by the value of Kappa. A value of P < 0.05 was considered to reflect statistically significant differences. All P values reported were two-sided.

Results

Assessment of mPCR

The specificity of those multiplex PCR primers (NS, CHS1, ACT1) was determined using DNA extracted from the dermatophytes, yeasts, molds, bacterials and human cells (see materials and methods). Specific amplicons were detected, and no non specific or cross-amplification was observed. No amplification product was seen without a DNA template.

This triplex PCR assay could successfully amplify the target genes when the aforementioned three DNA templates were put in different combinations (figure 1), and it was able to detect as little as 10² cfu/mL of T. rubrum by visualization of DNA bands on ethidium bromide-stained agarose gel.

Clinical evaluation

One hundred and four clinical specimens were collected. According to the criteria as described in materials and methods, forty five (43.3%) were diagnosed as onychomycosis. Among them, thirty three were distal lateral onychomycosis, one was proximal subungual onychomycosis, five were superficial onychomycosis, and six were total dystrophic onychomycosis.

Forty five (100%) were positive on microscopy, and 29 (64.4%) were positive on culture. In all culture-positive samples, T. rubrum was identified as the pathogen. The detection rate of mPCR was 93.3% (42/45) and all of them showed the pattern of NS+ CHS1, and NS bands were faint in five cases (figure 2). PCR was positive in 100% of cases where a culture was positive. In 45 cases of confirmed onychomycosis, fourteen had SCIO scores between 0∼10, 22 between 10∼20, and 9 between 20∼30. No significant differences of detection rates by microscopy, culture and mPCR related with the severity of the infected nails were seen (P > 0.05). The agreement of the results of these three methods with the “gold standard” is shown in table 1. The data of sensitivity, specificity, positive predictive value, negative predictive value, accuracy for each method are shown in table 2.

Table 1 The agreement of three diagnostic methods with “gold standard”.

P, positive; N, negative; TP, true positive; TN, true negative.

Table 2 Comparison of indexes for three diagnostic methods: microscopy, cultural assay and triplex PCR.

u₁ is the U value from the comparison between microscopy and culture, u₂ is from the comparison between microscopy and PCR, u₃ is from the comparison between culture and PCR.

**P < 0.01.

*P < 0.05.

Discussion

PCR assays have dramatically increased the sensitivity and specificity of methods for identifying pathogenic fungi in onychomycosis. Multiplex PCR has the unique advantage that more than one target sequence can be amplified by including more than one pair of primers in the reaction [25-26], and it is clinically useful as an efficient and fast procedure for the detection of pathogens, thus, it complements the traditional diagnostic analysis in a useful manner. Based on the pathogens of onychomycosis and the antifungal spectrum of drugs used in onychomycosis therapy, three primer pairs were used in our mPCR system, which were pan-fungal primer (NS), dermatophyte specific primer (CHS1) and yeast specific primer (ACT1). Although it does not allow species identification, the possible pathogen group could be determined and appropriate therapeutics could be adapted.

In the present study, four subtypes of onychomycosis were included, and these subtypes cause a large number of possible changes in the nail apparatus. Various SCIO values were covered. The results obtained from the three diagnostic techniques were reliable, with the accuracy of PCR and microscopy at more than 90%, and that of culture at nearly 85% (table 2). The results of microscopy and PCR agreed with the ‘gold standard’ described in Materials and methods (kappa > 0.75), and the results of culture also have a reasonable level of agreement with the “gold standard” (kappa = 0.67) (table 1).

The detection rate of 93.3% obtained by mPCR in our study is extremely high compared with that of the conventional cultures (64.4%). Although it was less sensitive than microscopy (100%), no significant difference was observed. Similar results were obtained by Arca et al. [27] who conducted one of the first studies on PCR in the diagnosis of onychomycosis. They examined 52 nail specimens by PCR, culture and microscopy, and the results indicated the sensitivity of microscopy was higher than that of PCR, and sensitivity of the culture was the lowest. In our study, the specificity of mPCR and culture was 100%, which was clearly superior to microscopy (P < 0.01).

As with other diagnostic techniques, a good interpretation of these results is necessary to adapt the therapeutics. In the present mPCR assay, different band patterns indicated different types of infection: (i) simple infection caused by dermatophyte (NS+ CHS1), or non dermatophytic mold, (NS) or yeast (NS + ACT); (ii) mixed infection caused by two groups of pathogenic fungi, i.e., dermatophyte plus yeast (NS+ +ACT +CHS1), or dermatophyte plus mold (NS+ CHS1), or yeast plus mold (NS + ACT); (iii) mixed infection caused by three groups of pathogenic fungi, i.e., dermatophyte plus yeast and mold (NS + ACT + CHS1). Some different types of infection share the same pattern, e.g. NS+ CHS1 may indicate simple dermatophyte infection or mixed infection caused by dermatophyte and mold, which is probably not clinically significant and does not affect the choice of treatment.

The band pattern (NS + CHS1) of mPCR demonstrated in our study indicated that a dermatophyte was very likely to be the sole infectious organism of those cases. Firstly, dermatophytes are by far the commonest causal organisms. In onychomycosis, representing more than 90% of infectious fungi in nails, the cases of mixed infection involving non dermatophytic molds are relatively rare. Secondly, no definitive evidence supported non dermatophytes in either direct microscopic observation or macroscopic observation of the affected nails (data not shown). Finally, in all culture-positive samples, T. rubrum was identified as the pathogen.

If necessary, epidemiology, clinical presentation, direct microscopic observation, culture and sequencing results should be taken together to determine the causative organisms. PCR-based identification to the species level could also be performed. Brillowska-Dabrowska et al. [28-29] developed a multiplex PCR for the detection of dermatophytes in general and Trichophyton rubrum specifically, and they also evaluated the diagnostic PCR tests for Microsporum audouinii, M. canis and Trichophyton infections. Ding et al. [30] employed a multiplex PCR and PCR-RFLP identification procedure targeting the DNA topoisomerase II gene to identify seven common dermatophytes in clinical practice. A recent study based on a novel multiplex real-time PCR detection format to diagnose common dermatophyte infections allows very sensitive detection and quantification of defined nucleic acid sequences, which sheds a new light on improvement of the multiplex PCR system [31].

The higher cost of PCR and the requirement of lab facilities, compared with conventional mycological diagnostics based on microscopy and culture, are the major hurdles to the clinical application of PCR. However, conventional morphological identification takes weeks and is prone to be influenced by the skill and experience of the personnel. Besides, by obtaining results within 8 h and interpreting results easily, this mPCR method has the potential to produce considerable savings of time and effort and even cost within the laboratory, without compromising test utility.

In conclusion, this mPCR assay may provide remarkable clinical benefits which offer the unique advantage of being directly applicable to clinical samples, obtaining results within 8 h. Because this method is simple and impersonal, it might be used in laboratories with no mycological specialization for rapid etiologic diagnosis and treatment selection. Further research including more types of fungal infections and larger numbers of clinical samples is required to define the practical value of such an approach in the mycology laboratory.

Disclosure

Acknowledgments: This work was supported by a grant from Institute of Dermatology, Chinese Academy of Medical Sciences & Peking Union Medical College (received by Xiaofang Li). Financial support: none. Conflict of interest: none.

References

1 R. Hay Literature review. Onychomycosis J Eur Acad Dermatol Venereol 2005; 19: Suppl. 1 1-7.

2 R.K. Scher Onychomycosis: a significant medical disorder J Am Acad Dermatol 1996; 35: S2-S5.

3 J Faergemann, R. Baran Epidemiology, clinical presentation and diagnosis of onychomycosis Br J Dermatol 2003; 149: Suppl. 65 1-4.

4 MA Ghannoum, RA Hajjeh, R Scher et al. A large-scale North American study of fungal isolates from nails: the frequency of onychomycosis, fungal distribution, and antifungal susceptibility patterns J Am Acad Dermatol 2000; 43: 641-648.

5 DT Roberts, WD Taylor, J BoyleBritish Association of Dermatologists Guidelines for treatment of onychomycosis Br J Dermatol 2003; 148: 402-410.

6 AK Gupta, E.A. Cooper Update in antifungal therapy of dermatophytosis Mycopathologia 2008; 166: 353-367.

7 JJ Finch, E.M. Warshaw Toenail onychomycosis: current and future treatment options Dermatol Ther 2007; 20: 31-46.

8 D. de Berker Clinical practice. Fungal nail disease N Engl J Med 2009; 360: 2108-2116.

9 S. Krishnan-Natesan Terbinafine: a pharmacological and clinical review Expert Opin Pharmacother 2009; 10: 2723-2733.

10 Y.M. Clayton Clinical and mycological diagnostic aspects of onychomycoses and dermatomycoses Clin Exp Dermatol 1992; 17: Suppl. 1 37-40.

11 XF Li, YN Shen, W Chen, H Chen, GX Lv, W.D. Liu A new medium for diagnosis of dermatophyte infection Eur J Dermatol 2009; 19: 34-37.

12 JE Arrese, C Pierard-Franchimont, G.E. Pierard Facing up to the diagnostic uncertainty and management of onychomycosis Int J Dermatol 1999; 38: Suppl. 2 1-6.

13 M Walberg, C M⊘rk, P Sandven, A Tomine Jorde, M Bj⊘råsand, P. Gaustad 18S rDNA Polymerase Chain Reaction and Sequencing in Onychomycosis Diagnostics Acta Derm Venereol 2006; 86: 223-226.

14 J Garg, R Tilak, S Singh et al. Evaluation of Pan-Dermatophyte Nested PCR in Diagnosis of Onychomycosis J Clin Microbiol 2007; 45: 3443-3445.

15 AK Gupta, M Zaman, J. Singh Fast and sensitive detection of Trichophyton rubrum DNA from the nail samples of patients with onychomycosis by a double-round polymerase chain reaction-based assay Br J Dermatol 2007; 157: 698-703.

16 M Ebihara, K Makimura, K Sato, S Abe, R. Tsuboi Molecular detection of dermatophytes and nondermatophytes in onychomycosis by nested polymerase chain reaction based on 28S ribosomal RNA gene sequences Br J Dermatol 2009; 161: 1038-1044.

17 O Bontems, PM Hauser, M. Monod Evaluation of a polymerase chain reaction-restriction fragment length polymorphism assay for dermatophyte and nondermatophyte identification in onychomycosis Br J Dermatol 2009; 161: 791-796.

18 L Turin, F Riva, G Galbiati, T. Cainelli Fast, simple and highly sensitive double-rounded polymerase chain reaction assay to detect medically relevant fungi in dermatological specimens Eur J Clin Invest 2000; 30: 511-518.

19 AY Sergeev, AK Gupta, Y.V. Sergeev The Scoring Clinical Index for Onychomycosis (SCIO Index) Skin Therapy Lett 2002; 7: Suppl. 1 6-7.

20 K Makimura, Y Tamura, T Mochizuki et al. Phylogenetic classification and species identification of dermatophyte strains based on DNA sequences of nuclear ribosomal internal transcribed spacer 1 regions J Clin Microbiol 1999; 37: 920-924.

21 S Baek, CHJ Chae, D Houh, DG Byun, B.K. Cho Detection and differentiation of causative fungi of onychomycosis using PCR amplification and restriction enzyme analysis Int J Dermatol 1998; 7: 682-686.

22 R Kano, A Hirai, M Muramatsu, T Watari, A. Hasegawa Direct detection of dermatophyte in skin samples based on sequences of the chitin synthase 1 (CHS1) gene J Vet Med Sci 2003; 65: 267-270.

23 M Kretschmar, A Felk, P Staib et al. Individual acid aspartic proteinases (Saps) 1-6 of Candida albicans are not essential for invasion and colonization of the gastrointestinal tract in mice Microb Pathog 2002; 32: 61-70.

24 CL Fletcher, RJ Hay, N.C. Smeeton Clinical and laboratory investigations onychomycosis: the development of a clinical diagnostic aid for toenail disease. Part I. Establishing discriminating historical and clinical features Br J Dermatol 2004; 150: 701-705.

25 G Luo, T.G. Mitchell Rapid identification of pathogenic fungi directly from cultures by using multiplex PCR J Clin Microbiol 2002; 40: 2860-2865.

26 EM Elnifro, AM Ashshi, RJ Cooper, P.E. Klapper Multiplex PCR: Optimization and Application in Diagnostic Virology Clin Microbiol Rev 2000; 13: 559-570.

27 E Arca, MA Saracli, A Akar, ST Yildiran, Z Kurumlu, A.R. Gur Polymerase chain reaction in the diagnosis of onychomycosis Eur J Dermatol 2004; 14: 52-55.

28 A Brillowska-Dabrowska, DM Saunte, M.C. Arendrup Five-hour diagnosis of dermatophyte nail infections with specific detection of Trichophyton rubrum J Clin Microbiol 2007; 45: 1200-1204.

29 A Brillowska-Dabrowska, A Swierkowska, DM Lindhardt Saunte, M.C. Arendrup Diagnostic PCR tests for Microsporum audouinii, M. canis and Trichophyton infections Medical Mycology 2010; 48: 486-490.

30 J Ding, J Li, Z Liu, Z. Tan Clinical identification of common species of dermatophytes by PCR and PCR-RFLP J Huazhong Univ Sci Technolog Med Sci 2004; 24: 642-644.

31 M Arabatzis, LE Bruijnesteijn van Coppenraet, EJ Kuijper et al. Diagnosis of common dermatophyte infections by a novel multiplex real-time polymerase chain reaction detection/identification scheme Br J Dermatol 2007; 157: 681-689.