DIAGNOSIS — The gold standard for the diagnosis of candidemia is a positive blood culture; blood cultures should be obtained in all patients with suspected candidemia. In patients with focal findings (eg, skin lesions or parenchymal involvement), biopsy should be performed for staining, culture, and histopathologic evaluation. When available, we suggest sending the beta-D-glucan assay because it can be a useful adjunct to blood cultures and biopsy and is particularly useful in patients with deep-seated invasive candidiasis (eg, intraabdominal candidiasis).
Culture and stain of biopsy material — Directed biopsy of sites of involvement often leads to a definitive diagnosis. Material obtained by scraping the base of a pustule should be submitted to the microbiology laboratory for Gram stain (picture 5) and culture. More commonly, a punch biopsy is performed and the tissue is submitted for culture and for histopathologic examination using special stains for fungi.
Punch biopsies of skin or tissue biopsy will show microabscesses, and special stains will show budding yeasts and often pseudohyphae or hyphae that are characteristic of Candida species.
Clinicians frequently must rely upon clinical judgment about the probability of candidemia as an explanation for a patient's symptoms while awaiting the return of blood cultures. Certain findings on physical examination, especially the presence of suggestive skin or eye lesions, can alert the clinician to the possibility of Candida infection. (See 'Clinical manifestations' above.)
Blood cultures — The most obvious method to detect candidemia is to grow the organism from the blood. Unfortunately, blood culture techniques are relatively insensitive. Studies from several decades ago showed that blood cultures were positive in only approximately 50 percent of patients who were found to have disseminated candidiasis at autopsy .
Although similar autopsy-based studies have not been reported since the advent of modern blood culture techniques, there have been many studies comparing the sensitivity of different blood culture systems for the detection of yeasts [3,4]. The introduction of the lysis-centrifugation method (Dupont Isolator tube) improved the detection of yeasts when compared with other blood culture systems. Subsequent changes in the culture media used in the BACTEC and BactiAlert systems, allowing enhanced growth of yeasts, have made these systems as sensitive as the lysis-centrifugation technique for detecting Candida species in blood.
A drawback of all blood culture systems for the diagnosis of candidemia is that one to three days are required for growth and an additional one to two days for identification of the organism after subculture onto agar medium. For a seriously ill patient, more rapid and more sensitive techniques are essential.
Several techniques have been developed that decrease the time to identification of the most common Candida species once a blood culture bottle shows growth and yeasts are seen on the smear of the bottle contents. Using the technique of peptide nucleic acid fluorescence in situ hybridization (PNA-FISH), C. albicans and C. glabrata can be identified within hours of a blood culture becoming positive for yeast [5,6]. Another technique is matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) mass spectrometry. Using this technique, proteins released from Candida spp are detected in a blood culture and are compared to a large database of proteins from many species of yeast . MALDI-TOF mass spectrometry can be used directly on material taken from a blood culture bottle that is positive for yeast, and results can be read in as short a period as 30 minutes.
A study that compared the performance of blood cultures with the polymerase chain reaction (PCR) and the beta-D-glucan assay is discussed below. (See 'Polymerase chain reaction' below.)
Non-culture methods — Over the last several decades, substantial efforts have been directed at the development of non–culture-based method for the diagnosis of candidemia.
Beta-D-glucan and other antigen assays — Detection of circulating Candida antigens or metabolites has been thought to hold more promise than antibody-based methods . Assays for several cell wall components, cytoplasmic antigens, and the metabolite arabinitol are not sufficiently sensitive for diagnostic purposes .
The most promising antigen assay is based upon the detection of beta-D-glucan, which is present in the cell wall of many fungi [9-12]. Thus, this assay is not specific forCandida. As noted above, when available, we suggest sending the beta-D-glucan assay because it can be a useful adjunct to blood cultures and biopsy and is particularly useful in patients with deep-seated invasive candidiasis (eg, intraabdominal candidiasis).
In one multicenter study of 163 patients with proven or probable invasive fungal infections and 170 patients without invasive fungal infections, 107 patients had proven candidiasis . Depending on the cutoff value chosen as positive, between 78 and 81 percent of the 107 patients with proven candidiasis had positive results using the beta-D-glucan assay.
Researchers in Japan have a longer experience with this assay, and the commercial kits available in Japan differ from those available in the United States. In an autopsy series of 456 patients that included 54 with invasive fungal infections, 41 patients had a beta-D-glucan assay performed . Using a cut-off value of 60 picograms/mL,the positive predictive value of the assay was found to be 70 percent, and the negative predictive value was 98 percent.
A study showed that a combination of blood cultures with the beta-D-glucan assay or the polymerase chain reaction increases the sensitivity of diagnostic testing compared with blood cultures alone . This is discussed in greater detail below. (See 'Polymerase chain reaction' below.)
T2Candida — In 2014, the US Food and Drug Administration (FDA) cleared the T2Candida panel and T2Dx instrument (T2Candida) for the detection of bloodstream infection caused by C. albicans, C. tropicalis, C. parapsilosis, C. glabrata, and/or C. krusei . The T2Candida assay can detect these pathogens from a blood specimen within three to five hours, far more quickly than blood cultures. The assay breaks yeast cells apart, releasing DNA; copies the target DNA; and detects the amplified DNA using magnetic resonance technology.
The FDA based their approval of T2Candida on two studies . In a study of 300 blood samples containing specific concentrations of yeast, T2Candida correctly identified the organism in 84 to 96 percent of positive specimens. In a separate study that included 1801 patients, T2Candida had an overall sensitivity of 91 percent and correctly categorized nearly 100 percent of negative specimens as negative for the presence of yeast .
The clinical utility of this technique requires further study.
Polymerase chain reaction — One current focus of non-culture methods is on the development of a PCR assay for candidemia and invasive candidiasis . As is true of cultures, PCR can identify Candida to the species level. However, to date, there is no commercially available approved PCR test to detect Candida species.
For the detection of candidemia, the sensitivity of PCR has been previously reported to be similar to that of blood cultures [17-19] and, in occasional cases of invasive candidiasis, PCR has assisted in establishing the diagnosis when cultures were negative .
A study compared the performance of a newer real-time PCR assay, the beta-D-glucan assay, and blood cultures for diagnosing invasive candidiasis in 55 patients with invasive candidiasis (17 with candidemia, 33 with deep-seated candidiasis, and 5 with both) and 73 hospitalized controls . The PCR assay was more sensitive than the beta-D-glucan assay for diagnosing invasive candidiasis (80 versus 56 percent) but had comparable specificity (70 versus 73 percent). The PCR assay and the beta-D-glucan assay had similar sensitivity for candidemia, but PCR was more sensitive for deep-seated candidiasis (89 versus 53 percent). Both PCR and the beta-D-glucan assay were more sensitive than blood cultures among patients who had invasive candidiasis. In patients with invasive candidiasis, the sensitivity of blood cultures combined with PCR or the beta-D-glucan assay was 98 and 79 percent, respectively. Thus, these two non–culture-based assays, when combined with blood cultures, have the potential to increase the sensitivity of culture methods for invasive candidiasis.
Antibody assays — To date, no sensitive and specific Candida antibody assay has been developed that is useful for diagnosis. Because Candida are part of the normal flora, healthy persons have antibodies against these organisms, thereby decreasing the specificity of the assay. On the other hand, an immunosuppressed host may be unable to generate a vigorous antibody response to Candida infection, thereby decreasing the sensitivity of the assay in this high-risk population.
SUMMARY AND RECOMMENDATIONS
●Candida in a blood culture should never be viewed as a contaminant and should always prompt a search for the source. For many patients, candidemia is a manifestation of invasive candidiasis, whereas for others it reflects colonization of an indwelling intravenous catheter. (See 'Introduction' above.)
●The clinical manifestations of candidemia vary from minimal fever to a full-blown sepsis syndrome that is indistinguishable from severe bacterial infection. Invasive candidiasis is defined as hematogenous spread to multiple viscera (eg, eye, kidney, heart valves, brain). (See 'Clinical manifestations' above.)
●Clinical clues on physical examination to hematogenous spread of Candida include characteristic eye lesions (chorioretinitis with or without vitritis), skin lesions, and, much less commonly, muscle abscesses. (See 'Clinical manifestations' above.)
●The gold standard for the diagnosis of candidemia is a positive blood culture; blood cultures should be obtained in all patients with suspected candidemia. However, blood cultures can be negative in patients with invasive candidiasis, and the diagnosis in such cases is often based on clinical suspicion. (See 'Blood cultures'above.)
●In patients with focal findings (eg, skin lesions or parenchymal involvement), biopsy should be performed for staining, culture, and histopathologic evaluation. (See'Diagnosis' above and 'Culture and stain of biopsy material' above.)
●When available, we suggest sending the beta-D-glucan assay as it may be a useful adjunct to blood cultures and biopsy. It appears to be particularly useful in patients with deep-seated candidiasis (eg, intraabdominal candidiasis) because blood cultures are insensitive in this setting. Beta-D-glucan is present in the cell wall of many fungi and the beta-D-glucan assay is therefore not specific for Candida spp. (See 'Diagnosis' above and 'Beta-D-glucan and other antigen assays' above.)