Background: Candidemia is a major infectious complication of seriously immunocompromised patients. In the absence of specific signs and symptoms, there is a need to evolve an appropriate diagnostic approach. A number of methods based on the detection of Candida mannan, nucleic acid and (1,3)-beta-D-glucan (BDG) have been used with varying specificities and sensitivities. In this retrospective study, attention has been focused to evaluate the usefulness of two or more disease markers in the diagnosis of candidemia. Methods: Diagnostic usefulness of Platelia Candida Ag for the detection of mannan, Platelia Candida Ab for the detection of anti-mannan antibodies, Fungitell for the detection of BDG, and of a semi-nested PCR (snPCR) for the detection Candida species-specific DNA have been retrospectively evaluated using 32 sera from 27 patients with culture-proven candidemia, 51 sera from 39 patients with clinically suspected candidemia, sera of 10 women with C. albicans vaginitis, and sera of 16 healthy controls. Results: Using cut-off values recommended by the manufacturers, the sensitivity of the assays for candidemia patients were as follows: Candida snPCR 88%, BDG 47%, mannan 41%, anti-mannan antibodies 47%, respectively. snPCR detected 5 patients who had candidemia due to more than one Candida species. The sensitivities of the combined tests were as follows: Candida mannan and anti-mannan antibodies 75%, and Candida mannan and BDG 56%. Addition of snPCR data improved the sensitivity further to 88%, thus adding 10 sera that were negative by BDG and/or mannan. In clinically suspected, blood culture negative patients; the positivities of the tests were as follows: Candida DNA was positive in 53%, BDG in 29%, mannan in 16%, and anti-mannan antibodies in 29%. The combined detection of mannan and BDG, and mannan, BDG and Candida DNA enhanced the positivity to 36% and 54%, respectively. None of the sera from Candida vaginitis patients and healthy subjects were positive for Candida DNA and mannan. Conclusion: The observations made in this study reinforce the diagnostic value of snPCR in the sensitive and specific diagnosis of candidemia and detection of more than one Candida species in a given patient. Additionally, in the absence of a positive blood culture, snPCR detected Candida DNA in sera of more than half of the clinically suspected patients. While detection of BDG, mannan and anti-mannan antibodies singly or in combination could help enhancing sensitivity and eliminating false positive tests, a more extensive evaluation of these assays in sequentially collected serum samples is required to assess their value in the early diagnosis of candidemia. © 2007 Alam et al; licensee BioMed Central Ltd.
