The current gold standard for diagnosis of invasive candidiasis (IC) is a blood or sterile-site culture positive for Candida species, but these tests can miss up to 50% of cases. In addition, culture often becomes positive late in the course of disease. This issue of Strategies examines a study that found a real-time polymerase chain reaction (PCR) test to be superior to both blood culture and 1,3- β-D-glucan (BDG) assays.

IC is the fourth most common cause of hospital-acquired bloodstream infections in the U.S. In hospitals, Candida represent 8-9% of all nosocomial bloodstream infections, with higher risk among patients with cancer or in intensive care units. Surgical patients, those with central venous catheters, immunosuppressed patients, and very low-birth-weight infants also are at risk. Unfortunately, blood cultures lack sensitivity for invasive and deep-seated disease, and they can take days to become positive, leaving clinicians guessing about who needs antifungal therapy. In the quest for faster, more sensitive diagnosis of IC, two promising testing methods have been recognized: PCR and BDG. A recent study investigated the ability of both of these assays to identify candidemia and deep-seated infections among patients with IC (Clin Infect Dis 2012;54:1240-1248).

Researchers have been exploring PCR testing for IC over several years. However, this new study differed from previous research in that the researchers focused on patients with deep-seated candidiasis, which occurs when candidemia—Candida in the blood stream—invades organs and causes infection in the heart, abdomen, or other normally sterile areas of the body. Deep-seated IC also is the most likely to go undetected by blood culture.

Deep-seated infections are of particular concern in the hospital setting. Antibiotic therapy can kill off the normal flora of the gut, allowing Candida to thrive. Likewise, chemotherapy sloughs off the mucosa of the gut, allowing Candida to transmigrate from the gut lumen into the bloodstream. Abdominal surgery is especially risky because the bacteria can leak out and invade vulnerable organs.

Minh-Hong Nguyen, MD, and her colleagues at the University of Pittsburgh Medical Center designed their study to see how well a PCR test would identify deep-seated infection even among blood culture- negative patients. They prospectively identified 55 patients with IC, including those with Candida recovered only from sterile sites. The study also included 75 hospitalized controls, 55 of whom had Candida colonization. The researchers found that PCR was more sensitive compared to BDG for diagnosing IC (80% versus 56%) and deep-seated candidiasis (89% versus 53%), with at a cutoff of 80 pmol/mL for BDG. Both PCR and BDG were notably superior to blood cultures among patients with deep-seated candidiasis, with sensitivity of 88% and 62% versus 17%, respectively. They found that the greatest sensitivity was achieved by combining blood culture with PCR or BDG, 98% and 79%, respectively. The Fungitell BDG and PCR testing was performed at and supported by Viracor-IBT Laboratories, the maker of the Candida PCR assay.

Nguyen emphasized the importance of her study having included patients that tested positive from sterile sites but not blood culture. “Previous studies have often looked at the wrong type of patient population,” Nguyen said. “If the blood culture has already turned positive, you don’t need PCR. Because we had patients that were blood culture-negative, but positive for deep-seated candidiasis, I think it demonstrated that PCR testing can add a lot of value and help us to treat patients. In order to decrease the usage of anti-fungal prophylaxis, we have to come up with better diagnostics, because right now, clinicians cannot rely on blood culture to start anti-fungal treatment.” Nguyen is professor of medicine and director of the transplant infectious diseases and antimicrobial management programs.

Specificity for PCR and BDG were low—70% and 73%, respectively—but this was by design. The researchers purposefully included a large number of patients in the control group who were colonized with Candida, Nguyen explained. “In our study, the specificity is much lower than the sensitivity. But this was intentional,” she said. “When we designed the study, we deliberately selected highly rigorous controls, primarily patients who were in the ICU for long periods of time and colonized with Candida or diagnosed with mucosal rather, than deeply invasive candidiasis. As such, our specificity is likely lower than would be encountered in actual clinical practice, where the percentages of heavily colonized patients or patients with mucosal candidiasis would probably be lower.”

It is also possible that some of the symptomatic false-positive PCR patients in the study may have had unrecognized invasive candidiasis, which was missed by cultures and existing diagnostic approaches, Nguyen added. For example, the study included patients with fevers who had esophageal candidiasis, but negative blood cultures for fungal or bacterial pathogens. “By our definition, these patients were considered negative controls, but it is plausible that they had unrecognized non-mucosal invasive disease,” she said.

Nguyen emphasized that more research is needed before PCR testing for IC will be implemented widely. Unlike the BDG assay, there is no FDA-approved kit, and each lab’s particular methodology varies considerably. “The big problem is that nobody has a standardized assay, so an assay we perform here at our hospital would be totally different because we use different primers, a different DNA extraction method, and so on,” Nguyen said. “Until we have a standardized, FDA-approved test, we cannot really compare between different centers.”

Although a standardized assay is lacking, the results from this study do confirm the outcomes of previous studies that have compared PCR to blood culture. In a 2011 review of 49 studies, the authors found that among patients suspected of probable or possible IC, the positivity rate of PCR was much higher than blood culture—85% versus 38% and 67% versus 29%, respectively (J. Clin. Microbiol 2011; 49: 665-670). “These results are compatible with the conclusion that PCR provides higher sensitivity than blood culture for the diagnosis of IC, with a specificity higher than 90%,” the authors wrote. They also noted that none of the studies they examined reported the effects of PCR on clinical outcomes. As a result, the authors recommended clinical trials to assess the impact of using PCR testing in conjunction with blood culture, compared to blood culture alone.