An overall sensitivity of 10 pg DNA was determined. Negative results and no template controls were confirmed by a positive band for the internal amplification control QS. The specificity was tested with increasing amount of human DNA. No cross-reactivity was observed up to 90 ng of human DNA. Although the kit is exclusively intended for human in vitro diagnosis, DNA preparations from selected

pets and farm animals (Table 1) were tested at 2 ng. Faint cross-reactivity was only observed with PS-341 clinical trial DNA from cat although the size of the unspecific amplification products did not match the reference ladders. The robustness of the PCR was confirmed by varying the thermocycler (‘Material and

methods’) and the annealing temperatures of the PCR by ±2 °C. Mixtures of 10 pg RGFP966 fungal DNA and 300 ng human DNA were assembled and subjected to PCR 1 and 2. These experiments revealed clear pathogen-specific amplification products and no cross-reactivity with human DNA at the detection limit. In total, 253 patients treated at the Department of Dermatology (University Hospital Carl Gustav Carus, TU Dresden, Germany) and 10 healthy subjects were analysed from September 2011 to April 2012. The clinical diagnosis revealed 122 onychomycoses, 76 tinea peduum, 16 tinea manuum, 3 tinea inguinales, 21 tinea corporum et facies and 15 mucosal candidoses. According to these clinical manifestations, 122 nail clippings, 105 skin scrapings and 26 smears from mucosa and weeping skin lesions were collected and subjected to microscopy, microbial culture and multiplex PCR. The nail clippings from all healthy subjects were negative for the three diagnostic methods. These results were not included in the further calculations. Of the 253 patients, 87 (34.4%) were tested positive in microscopy, 80 (31.6%) in culture, 128 (50.6%) in culture or microscopy or both and 127 (50.2%) in PCR respectively. The compliance

find more between the technologies is shown in Table 3. 44.8% of microscopically positive samples showed positive results in culture whereas in 90.8% of these samples positive results were revealed by multiplex PCR. Positive cultures could be confirmed in 80.0% by multiplex PCR and in 48.8% by direct microscopic examination. The detected pathogens are listed in Table 4. Candida yeast were further differentiated by culture and metabolic tests into 28 C. parapsilosis, 12 C. albicans, 6 C. guilliermondii, 2 C. glabrata and 2 C. krusei. Mixed fungal infections were seen in 10 cultures. These included all Cryptococcus spp. and Trichosporum spp. isolates in combination with T. rubrum or Candida spp. respectively. A combined infection of T. rubrum and C. parapsilosis was observed in three cases. The performance of multiplex PCR 1 and 2 with clinical samples are exemplified in Fig. 2.