Pharmacology Consult

Rapid diagnosis of invasive candidiasis and application to antifungal stewardship

Candida spp. are among the most common pathogens in nosocomial bloodstream infections. Crude mortality rates remain high, at around 40% — comparable to Pseudomonas and higher than Staphylococcus aureus and bacteria belonging to Enterobacteriaceae. Additionally, candidemia is more common in high-risk patients with underlying comorbid conditions, including those with immunocompromised status, central venous catheters, ICU admission or recent surgery.

The traditional blood culture method of detection has numerous limitations to timely and accurate identification of Candida spp. Erratic release of Candida cells into the bloodstream, variable microbial burden, hepatic filtration of candidemia secondary to gut translocation and rapid clearance from the bloodstream contribute to the notoriously poor sensitivity of fungal blood cultures. Even when Candida cells are collected in a blood culture sample, it may take 2 to 3 days before a blood culture bottle is flagged positive by automated blood culture systems. These delays and limitations are of particular concern because delayed time to antifungal administration is associated with increased mortality.

Evan Abrass
Leah Molloy

As a result, there is great interest in rapid diagnostic assays that offer increased sensitivity to detection and faster pathogen identification. This article addresses selected platforms and their key features, including sensitivity and specificity, which vary slightly between different sources available in the literature.

PNA-FISH

Peptide nucleic acid fluorescence in situ hybridization (PNA-FISH) technology targets ribosomal RNA for rapid identification from positive blood cultures. The Yeast Traffic Light PNA-FISH (OpGen) assay uses a pool of different probes to identify and differentiate Candida spp. and can be applied after a blood culture becomes positive. In this method, C. albicans and C. parapsilosis are observed with green fluorescence; C. tropicalis with yellow; and C. glabrata and C. krusei with red fluorescence under fluorescent microscope. Limitations of this test include an inability to further differentiate between Candida spp. that stain the same color and to identify other Candida spp. However, Candida identification remains accurate, even in blood cultures that are positive for both bacteria and yeast.

MALDI-TOF

Matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) technology identifies pathogens by ionizing microbial cells with a laser to energize small protein particles. The proteins fly in a vacuum chamber and are detected by mass spectrometry. This gives a protein signature that is matched against a database to identify the pathogen. Therefore, the number of pathogens that MALDI-TOF can identify is limited only by its database, which can grow with updates to the system. Owing to its molecular method, MALDI-TOF can identify a large variety of specific Candida spp., unlike the relatively fewer number of species that can be identified by PNA-FISH, which cannot differentiate these species from others with similar fluconazole susceptibility. Although the sensitivity and specificity of MALDI-TOF for identification of C. albicans are high and comparable to other technologies, researchers have reported varied accuracy across all Candida spp.

Multiplex PCR

The FilmArray Blood Culture Identification Panel, or BCID (bioMérieux), uses multiplex PCR to amplify and detect pathogens from positive blood cultures. Like PNA-FISH, BCID detects only the five main pathogenic Candida spp. In contrast to traditional single PCR platforms, multiplex PCR technology can screen for multiple organisms simultaneously. However, the multiplex PCR requires a positive blood culture bottle, which is a disadvantage compared with single PCR tests.

T2MR

A common limitation of all the tests described earlier is the need for a positive blood culture before they can be applied. As discussed previously, there are several barriers to identifying candidemia using traditional blood culture methods. Additionally, even when viable Candida cells are collected in sufficient concentrations in a blood sample for culture, at least 24 hours for incubation will elapse between blood culture collection and first positive result. T2 magnetic resonance (T2MR) is the first technology capable of detecting candidemia independent of a positive blood culture. The T2Candida panel of the T2Dx instrument (T2 Biosystems) uses T2MR to lyse red blood cells, concentrate and lyse any present Candida cells, and then amplify and detect Candida DNA. This test can detect five different Candida spp. and reports them as: C. albicans/C. tropicalis, C. krusei/C. glabrata or C. parapsilosis. Species-specific lower limits of detection are as low as 1 CFU/mL.

Although none of the currently available fungal rapid diagnostic tests provide antifungal susceptibility information, this is not necessarily a critical limitation. Unlike bacterial infections for which antibiotic susceptibility can vary greatly, antifungal susceptibility among Candida spp. is typically more uniform and predictable. Rapid and accurate identification of candidemia to the species level is of greater clinical relevance.

Future technologies under development

A more rapid PNA-FISH platform, QuickFISH (OpGen), is approved in Europe but used for research only in the United States. The technology is similar to those described earlier but has a shorter turnaround time of 20 minutes, compared with 90 minutes with the currently available Yeast Traffic Light PNA-FISH.

An appealing novel characteristic of new rapid diagnostic tests is the opportunity to circumvent the rate-limiting step of a positive blood culture. The Field Activated Sample Treatment (FAST) ID System (Qvella) functions to automatically isolate, concentrate and lyse microbial cells, which are then subjected to real-time (RT)-PCR for detection. The anticipated processing time from whole blood to pathogen identification is 45 minutes. This test has not been validated outside the laboratory and requires further research to confirm its sensitivity and specificity and time to results.

Application to antifungal stewardship

A Monte Carlo simulation of invasive candidiasis compared the anticipated time from culture collection to initiation of therapy using different platforms. The model accounted for test-specific sensitivity, specificity and identifiable species, as well as baseline disease incidence and prompt stewardship intervention. Compared with an average of 3.5 ± 2.1 days that elapsed from culture collection to initiation of therapy in a clinical population in which rapid diagnostics were not used, the researchers predicted that PNA-FISH would reduce the time to 2.6 ± 1.3 days, MALDI-TOF to 2.5 ± 1.4 days, and T2Candida to 0.6 ± 0.2 days.

Multiple other stewardship studies have evaluated available rapid testing platforms, although many of these combined candidemia and bacteremia, with candidemia typically in the minority. One such evaluation characterized the impact of BCID with stewardship intervention. Although candidemia comprised just over 10% of isolates, it was analyzed separately, and BCID shortened time to preliminary identification of Candida by a median time of 24.7 hours. Although researchers observed an even greater reduction in time to final identification and susceptibility results, this did not translate into reduced time to effective therapy, owing to frequent empiric treatment and fluconazole susceptibility, underscoring the greater importance of simply identifying yeast than determining precise susceptibility information.

Multiple factors should be considered when selecting a rapid diagnostic platform, including cost, time, labor and accuracy. Also of importance is the consideration of clinical implementation and workflow. Timely communication of results, clinician understanding and interpretation, and prompt and appropriate therapy changes are essential for these novel technologies to impact patient outcomes, creating yet another important opportunity for multidisciplinary collaboration to improve patient care.

Acknowledgment:
The authors thank Hossein Salimnia, PhD, professor in the department of pathology at Wayne State University School of Medicine, for his content expertise and assistance.

Disclosures: Abrass and Molloy report no relevant financial disclosures.

Candida spp. are among the most common pathogens in nosocomial bloodstream infections. Crude mortality rates remain high, at around 40% — comparable to Pseudomonas and higher than Staphylococcus aureus and bacteria belonging to Enterobacteriaceae. Additionally, candidemia is more common in high-risk patients with underlying comorbid conditions, including those with immunocompromised status, central venous catheters, ICU admission or recent surgery.

The traditional blood culture method of detection has numerous limitations to timely and accurate identification of Candida spp. Erratic release of Candida cells into the bloodstream, variable microbial burden, hepatic filtration of candidemia secondary to gut translocation and rapid clearance from the bloodstream contribute to the notoriously poor sensitivity of fungal blood cultures. Even when Candida cells are collected in a blood culture sample, it may take 2 to 3 days before a blood culture bottle is flagged positive by automated blood culture systems. These delays and limitations are of particular concern because delayed time to antifungal administration is associated with increased mortality.

Evan Abrass
Leah Molloy

As a result, there is great interest in rapid diagnostic assays that offer increased sensitivity to detection and faster pathogen identification. This article addresses selected platforms and their key features, including sensitivity and specificity, which vary slightly between different sources available in the literature.

PNA-FISH

Peptide nucleic acid fluorescence in situ hybridization (PNA-FISH) technology targets ribosomal RNA for rapid identification from positive blood cultures. The Yeast Traffic Light PNA-FISH (OpGen) assay uses a pool of different probes to identify and differentiate Candida spp. and can be applied after a blood culture becomes positive. In this method, C. albicans and C. parapsilosis are observed with green fluorescence; C. tropicalis with yellow; and C. glabrata and C. krusei with red fluorescence under fluorescent microscope. Limitations of this test include an inability to further differentiate between Candida spp. that stain the same color and to identify other Candida spp. However, Candida identification remains accurate, even in blood cultures that are positive for both bacteria and yeast.

MALDI-TOF

Matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) technology identifies pathogens by ionizing microbial cells with a laser to energize small protein particles. The proteins fly in a vacuum chamber and are detected by mass spectrometry. This gives a protein signature that is matched against a database to identify the pathogen. Therefore, the number of pathogens that MALDI-TOF can identify is limited only by its database, which can grow with updates to the system. Owing to its molecular method, MALDI-TOF can identify a large variety of specific Candida spp., unlike the relatively fewer number of species that can be identified by PNA-FISH, which cannot differentiate these species from others with similar fluconazole susceptibility. Although the sensitivity and specificity of MALDI-TOF for identification of C. albicans are high and comparable to other technologies, researchers have reported varied accuracy across all Candida spp.

PAGE BREAK

Multiplex PCR

The FilmArray Blood Culture Identification Panel, or BCID (bioMérieux), uses multiplex PCR to amplify and detect pathogens from positive blood cultures. Like PNA-FISH, BCID detects only the five main pathogenic Candida spp. In contrast to traditional single PCR platforms, multiplex PCR technology can screen for multiple organisms simultaneously. However, the multiplex PCR requires a positive blood culture bottle, which is a disadvantage compared with single PCR tests.

T2MR

A common limitation of all the tests described earlier is the need for a positive blood culture before they can be applied. As discussed previously, there are several barriers to identifying candidemia using traditional blood culture methods. Additionally, even when viable Candida cells are collected in sufficient concentrations in a blood sample for culture, at least 24 hours for incubation will elapse between blood culture collection and first positive result. T2 magnetic resonance (T2MR) is the first technology capable of detecting candidemia independent of a positive blood culture. The T2Candida panel of the T2Dx instrument (T2 Biosystems) uses T2MR to lyse red blood cells, concentrate and lyse any present Candida cells, and then amplify and detect Candida DNA. This test can detect five different Candida spp. and reports them as: C. albicans/C. tropicalis, C. krusei/C. glabrata or C. parapsilosis. Species-specific lower limits of detection are as low as 1 CFU/mL.

Although none of the currently available fungal rapid diagnostic tests provide antifungal susceptibility information, this is not necessarily a critical limitation. Unlike bacterial infections for which antibiotic susceptibility can vary greatly, antifungal susceptibility among Candida spp. is typically more uniform and predictable. Rapid and accurate identification of candidemia to the species level is of greater clinical relevance.

Future technologies under development

A more rapid PNA-FISH platform, QuickFISH (OpGen), is approved in Europe but used for research only in the United States. The technology is similar to those described earlier but has a shorter turnaround time of 20 minutes, compared with 90 minutes with the currently available Yeast Traffic Light PNA-FISH.

An appealing novel characteristic of new rapid diagnostic tests is the opportunity to circumvent the rate-limiting step of a positive blood culture. The Field Activated Sample Treatment (FAST) ID System (Qvella) functions to automatically isolate, concentrate and lyse microbial cells, which are then subjected to real-time (RT)-PCR for detection. The anticipated processing time from whole blood to pathogen identification is 45 minutes. This test has not been validated outside the laboratory and requires further research to confirm its sensitivity and specificity and time to results.

PAGE BREAK

Application to antifungal stewardship

A Monte Carlo simulation of invasive candidiasis compared the anticipated time from culture collection to initiation of therapy using different platforms. The model accounted for test-specific sensitivity, specificity and identifiable species, as well as baseline disease incidence and prompt stewardship intervention. Compared with an average of 3.5 ± 2.1 days that elapsed from culture collection to initiation of therapy in a clinical population in which rapid diagnostics were not used, the researchers predicted that PNA-FISH would reduce the time to 2.6 ± 1.3 days, MALDI-TOF to 2.5 ± 1.4 days, and T2Candida to 0.6 ± 0.2 days.

Multiple other stewardship studies have evaluated available rapid testing platforms, although many of these combined candidemia and bacteremia, with candidemia typically in the minority. One such evaluation characterized the impact of BCID with stewardship intervention. Although candidemia comprised just over 10% of isolates, it was analyzed separately, and BCID shortened time to preliminary identification of Candida by a median time of 24.7 hours. Although researchers observed an even greater reduction in time to final identification and susceptibility results, this did not translate into reduced time to effective therapy, owing to frequent empiric treatment and fluconazole susceptibility, underscoring the greater importance of simply identifying yeast than determining precise susceptibility information.

Multiple factors should be considered when selecting a rapid diagnostic platform, including cost, time, labor and accuracy. Also of importance is the consideration of clinical implementation and workflow. Timely communication of results, clinician understanding and interpretation, and prompt and appropriate therapy changes are essential for these novel technologies to impact patient outcomes, creating yet another important opportunity for multidisciplinary collaboration to improve patient care.

Acknowledgment:
The authors thank Hossein Salimnia, PhD, professor in the department of pathology at Wayne State University School of Medicine, for his content expertise and assistance.

Disclosures: Abrass and Molloy report no relevant financial disclosures.

    See more from Outbreak Tracker