In Nevada, the first clinical case of C. auris was identified in August 2021. As of June 2022, a total of 300 cases had been reported and 22 healthcare facilities affected, including 10 acute care hospitals, 8 skilled nursing facilities, and 4 long-term acute care facilities. Clade III C. auris comprises most cases in Nevada, but clades I, II, and IV have also been identified. This sustained pathogen transmission represented an ideal scenario to explore the potential of wastewater surveillance for C. auris.

We collected wastewater influent samples (i.e., raw sewage) from 2 WWTPs (WWTP1 and WWTP2) in southern Nevada. WWTP1 treats an average daily flow of ≈100 million gallons/day; it serves a local population of nearly 1 million people, in addition to ≈800,000 weekly visitors. WWTP2 treats an average daily flow of ≈5 million gallons/day and serves a primarily residential sewershed with ≈90,000 people. We collected 50 mL aliquots of wastewater influent grab samples from each facility on the mornings of May 23 and 31 and June 6 and 13, 2022. We shipped samples on ice overnight to the Utah Public Health Laboratory, Utah Department of Health and Human Services (Taylorsville, UT, USA), where they were processed immediately upon receipt. We centrifuged each 50 mL sample at 5,000 × g for 10 minutes and resuspended the resulting pellet in 1 mL of 0.9% saline. We used 100 μL of this suspension to inoculate 2 mL of S² Media salt Sabouraud dulcitol broth (SSDB; Thomas Scientific, https://www.thomassci.com) (9) containing fluconazole (SSDBF) at a final concentration of 8 μg/mL. We incubated inoculated SSDBF tubes at 42°C for up to 7 d with vigorous agitation at 250 rpm. After incubation, we plated 100 μL of the inoculated SSDBF broth on HardyCHROM Candida medium (Hardy Diagnostics, https://hardydiagnostics.com) or BBL CHROMagar Candida medium (Becton, Dickinson and Company, https://www.bd.com). Alternatively, we used a 1 μL loop for quadrant streaking. We used matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (Bruker, https://www.bruker.com) to identify species of growing colonies. We performed whole-genome sequencing on NextSeq 2000 or NovaSeq platforms (Illumina, https://www.illumina.com), and performed genome assembly as described elsewhere (10). We mapped reads to a reference C. auris genome and performed single-nucleotide polymorphism (SNP) and phylogenetic tree analyses using the MycoSNP pipeline (https://github.com/CDCgov/mycosnp-nf) (10).

Figure 1

Figure 1. Culture-based isolation of Candida aurisfrom wastewater, southern Nevada, USA, 2022, showing the effects of incubation temperature and fluconazole selection on competing organisms. A, B) Incubating the enrichment broth...

Culture-based recovery of C. auris from marine environmental samples has been recently reported (11), but isolation from wastewater has not been documented. Following the protocol developed to detect C. auris in human specimens, use of SSDB as an enrichment medium was first explored (9). However, the selection criteria used for clinical samples in that study proved not stringent enough because fast-growing filamentous fungi and halo- and stress-tolerant Candida spp. could easily overgrow during the enrichment step (12). Inhibition of competing filamentous fungi, such as Mucor spp. and Geotrichum spp. (6), was mitigated by increasing the incubation temperature of the enrichment broth from 40°C (suggested temperature for patient colonization screening) to 42°C (Figure 1, panels A, B). Suppressing halo- and stress-tolerant Candida spp. required adding other selective agents to SSDB. Given that modal MIC for fluconazole for C. auris isolates is significantly higher than for other Candida spp. (13), we pursued an azole-based selection strategy.

Figure 2

Figure 2. Genetic relatedness of clinical and wastewater isolates identified in the southern Nevada, USA, 2022, to reference clinical isolates. Neighbor-joining phylogenetic tree shows clade III C. aurisisolates recovered from...

Preliminary experiments using samples from wastewater spiked with a fluconazole-resistant clinical strain of C. auris confirmed that adding fluconazole at a final concentration of 8 μg/mL suppressed growth of nontarget Candida spp. while ensuring recovery of C. auris (data not shown). We then prospectively validated this approach using wastewater samples collected in southern Nevada, which was experiencing an intense C. auris outbreak. Of the 8 samples we collected and analyzed during May 23–June 13, we recovered well-isolated C. auris colonies from a single WWTP1 sample from May 23 using SSDB broth enriched in the presence of fluconazole (Figure 1, panel D). Without azole selection, the target organism was overgrown by an unidentified yeast species (Figure 1, panel C). Of note, genomic analysis of 2 separate wastewater-derived C. auris isolates revealed phylogenetic relatedness to clade III isolates identified in 3 acute care hospitals within the WWTP1 sewershed (Figure 2). The 2 wastewater isolate genomes differed by 4 SNPs from each other and by 13–20 SNPs from the outbreak-associated isolates from the acute care hospitals (Figure 2).

We report recovery of C. auris isolates from wastewater that demonstrated an epidemiologic link to healthcare facilities within the WWTP sewershed in southern Nevada. Although our findings highlight the potential utility of community-level wastewater surveillance for C. auris, the methods and the data presented here represent only early-stage implementation. More recent prospective testing of samples in Nevada using both culture and quantitative PCR monitoring (D. Gerrity, unpub. data) has shown considerable variability between samples for the presence of competing organisms, which required increased selection stringency. Indeed, in August 2022, we could repeat isolation of C. auris from Nevada samples only by increasing the fluconazole concentration to 32 μg/mL (data not shown). This experience illustrates one limitation of the culture-based approach we used in this study: C. auris strains with a fluconazole MIC lower than the concentration of the drug in the broth might not be detected. In addition, the actual sensitivity of this surveillance approach is still unknown, because it can be influenced by the degree of competition between organisms during the enrichment step. PCR-based quantification of C. auris genome equivalents in Nevada wastewater indicates that culture-based isolation has occurred at organism concentrations as low as ≈100 CFU/mL (D. Gerrity, unpub. data). In addition, the full extent of C. auris colonization, transmission within sewersheds, and organism shedding have not been sufficiently studied. Therefore, it remains to be determined whether C. auris could be detected in geographic areas by wastewater surveillance before being recognized in clinical settings. Regardless, wastewater surveillance of pooled samples at the community level might effectively complement clinical surveillance of individual patients for detecting and characterizing C. auris outbreaks.