Diagnosis of Invasive Fungal Disease in Coronavirus Disease 2019

Approaches and Pitfalls

P. Lewis White


Curr Opin Infect Dis. 2021;34(6):573-580. 

In This Article

Diagnosis of Coronavirus Disease-2019-associated Candidosis

CAC has been described globally at varying incidences (<1 to 23.5%), mostly presenting as candidaemia 1–2 weeks post admission, caused primarily by Candida albicans and Candida glabrata, although outbreaks caused by multidrug-resistant C. auris continue to be reported.[3,10,18,40–42] Diagnosis is primarily through the recovery of Candida spp. through blood culture. Although no performance data specific for CAC is currently available, it is likely that blood culture sensitivity is comparable to that in non-COVID-19 patients, detecting approximately 50% of all forms of IC, reduced when the organism causes deep-seated infection in the absence of fungaemia.[43] With higher levels of Candida intravenous line infection reported, regular culture of line tips may be beneficial in the deteriorating patient, and while a risk factor for deep-seated infection, additional mycological evidence is required to confirm invasive candidosis.[10] The presence of Candida spp. cultured from the respiratory tract likely reflect commensal organisms rather than Candida pneumonia, diagnosis requires histological/microscopic evidence of pseudohyphae/hyphae invading lung tissue.

Nonculture diagnostics in the form of Candida PCR, BDG, Candida antigen and antibody EIA can aid the diagnosis of invasive candidosis but performance data specific to CAC is lacking. A pre-COVID-19 meta-analysis of Candida PCR testing of blood generated very high sensitivity and specificity (>90%) and the development of the T2 Candida assay allows fully automated testing, with promising performance (Se: 91%/Sp: 94%) and commercial PCR assays for the detection of C. auris are available.[44–46] The performance of serum BDG for detection of invasive candidosis generates sensitivity and specificity of approximately 80% but an understanding of the discussed limitations of BDG testing is critical, along with combining BDG testing with Candida specific assays.[18,47] BDG testing of respiratory samples is not recommended, even when respiratory fungal infection is suspected, as commensal Candida spp. and other colonizing fungi will compromise assay specificity. The individual meta-analytical performance of Candida antibody and Candida antigen testing provide moderate pooled sensitivity (approximately 60%) but good pooled specificity (approximately 83–93%). Combining these two tests enhances sensitivity (approximately 83%, when either test is positive) while maintaining specificity (approximately 86%, when both tests are positive).[48]

Although Candida risk/colonization scores have shown potential for identifying patients at increased risk of IC, the significant number of clinical interventions and the necessary duration of admission in the ICU mean that critical-care COVID-19 patients are likely exposed to prolonged risk of invasive candidosis.[18,40,49,50] It remains unclear what combination of tests will prove optimal for the diagnosis of invasive candidosis in either the COVID-19 or non-COVID-19 population, but a likely combination of molecular, serological and conventional testing will cover the range of targets potentially available in cases of invasive candidosis and when all tests are negative invasive candidosis will be unlikely.[51] In candidaemic patients, particularly those with immunosuppression or persisting blood culture positivity, transoesophageal echocardiography and fundoscopy are recommended for the diagnosis of endocarditis or ocular candidosis.[52]