Clinical accuracy of cryptococcal antigen lateral flow assay for diagnosis of canine and feline cryptococcosis

Fungal infections with Cryptococcus spp. are commonly diagnosed by serological detection of circulating antigen. Available assays are the latex cryptococcal antigen agglutination test (LCAT) offered by commercial laboratories and the cryptococcal antigen lateral flow assay (CLFA) as a point of care test. Two recent veterinary studies revealed an excellent sensitivity but variable specificity of the CLFA when compared to LCAT. The accuracy of the CLFA for diagnosing canine and feline cryptococcosis in a clinical setting is unknown. The aim of the study was to investigate clinical performance of a CLFA for dogs and cats presented to a referral hospital. It was hypothesized that clinical data demonstrate suitability of the CLFA as a screening test to exclude cryptococcosis but would not allow reliable confirmation of the disease. A hospital patient database was searched for dogs and cats that had a CLFA (IMMY, Immuno-Mycologics Inc, Oklahoma) performed between August 2012 and December 2020. Eligible animals were divided into cryptococcosis positive (Cr+) and negative (Cr-) based on their clinical records. All Cr+ patients had confirmation of diagnosis by cytology, histopathology and/or culture. The Cr- patients had a follow-up of at least 3 months with stable, improved or resolved clinical signs and no treatment with antifungal medication. A total of 170 patients were included. Nine patients (4 dogs, 5 cats) were assigned to the Cr+ group and all had a positive CLFA result. The remaining 161 patients (120 dogs, 41 cats) were classified as Crwith a median follow-up time of 15 months (range 3-101) for dogs and 14 months (range 3-55) for cats. Of 120 Cr- dogs, 110 had a negative CLFA result and 10 a positive CLFA result (median follow-up 15 months, range 3-31). Two positive CLFA results were recorded for the cats in the Cr- group (follow-up 28 months each). Disease prevalence in this study was 3.3% and 11.1% for dogs and cats respectively. Sensitivity and negative predictive value of the CLFA were 100% for both dogs and cats. Specificities were 91.7% (CI 85.2-95.3%) for dogs and 95.1% (CI 83.5-99.4%) for cats with corresponding positive predictive values of 29% (CI 18.5-42.6%) and 71.9% (CI 39.9-90.8%). The clinical performance of the CLFA reflects the previously established laboratory data and enables the clinician to reliably exclude cryptococcal infections in dogs and cats. However, a positive result requires further confirmation of the presence of the disease in both dogs and cats