In case 1, an extremely low-birth-weight male infant was delivered at 23 weeks and 3 days of gestation. On postnatal day 49, desaturations, neutropenia, and elevated C-reactive protein level were observed. Peripheral blood and central line (C-line) tip cultures grew C. famata, with low discrimination from C. parapsilosis. The infant was treated with fluconazole, and blood cultures obtained 1 week later were negative.

In case 2, a 41-yr-old woman had T-cell lymphoma with leptomeningeal seeding. She was undergoing chemotherapy with an Ommaya reservoir in place, and neutropenic fever developed. Meropenem, vancomycin, and caspofungin were empirically administered. Blood cultures obtained from a C-line and peripheral blood were positive for C. lusitaniae. Five days after treatment, blood cultures were negative; however, C. lusitaniae rebounded 2 days later and was identified in the cerebrospinal fluid. At this point, caspofungin was replaced by liposomal amphotericin B (AmB). Owing to persistent fevers, fluconazole was added to the treatment regimen, and a subsequent blood culture grew C. famata. Finally, owing to probable invasive pulmonary aspergillosis, voriconazole alone was administered. The patient died 2 days after the initiation of this treatment.

In case 3, a 68-yr-old woman was hospitalized for treatment of a prosthetic joint infection following a total left knee arthroplasty in 2010. She was treated unsuccessfully with several surgical and antibiotic therapies. In July 2013, C. parapsilosis was isolated from the joint fluid, and the patient subsequently underwent open debridement and insertion of an antibiotic-loaded cement spacer in February 2014. Owing to ongoing pain, her joint fluid was cultured the following month, which tested positive for C. famata, and fluconazole treatment was initiated.

Because of the discrepancy within these results or low discrimination between species, sequence analyses of the internal transcribed spacer (ITS) and D1/D2 regions of the rRNA gene were performed to identify C. famata isolates [5]. In cases 1, 2, and 3, C. parapsilosis, C. lusitaniae, and C. parapsilosis were identified, respectively (Table 1), with no isolate being identified as C. famata.

Matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS)-based identification of clinically relevant yeasts has proven to be superior to identification by phenotypic identification systems [4]. Thus, we subjected the isolates from cases 1 and 2 to MALDI-TOF MS (Bruker Biotyper, Bruker Daltonik, Bremen, Germany). The results were consistent with those obtained by sequencing; however, based on the manufacturer's breakpoints, the identification confidence score (1.70) allowed for only a genus-level identification in both cases [4].

In recent molecular identification studies, phenotypic methods initially identified isolates as C. famata, including strains of C. guilliermondii, C. lusitaniae, C. parapsilosis, and C. palmioleophila [1, 3]. Moreover, the ARTEMIS and SENTRY surveillance programs that identified 53 isolates as C. famata were disproven when sequencing analysis was performed by Castanheira et al. [4], suggesting that phenotypic identification of C. famata is almost certainly incorrect. The results of these studies are consistent with those of the three cases presented herein. Therefore, it is possible that, in the absence of confirmatory assays (e.g., sequencing), previous case reports may have misdiagnosed C. famata infection. To date, six such cases have been reported in Korea (Table 2), and the data presented here suggest the possibility of misidentification.

General antifungal susceptibility patterns of C. famata, C. parapsilosis, and C. lusitaniae are different. C. famata exhibits good susceptibility to AmB but reduced susceptibility to echinocandins and azoles. Therefore, liposomal AmB is recommended as the initial therapy [2]. C. parapsilosis is highly susceptible to most antifungal agents, but exhibits higher minimal inhibitory concentrations to echinocandins. C. lusitaniae is usually susceptible to azoles and echinocandins; however, it rapidly acquires resistance to AmB in vitro. Therefore, although misidentification of C. famata in our cases did not have a significant impact on the patients' treatment, this could have led to misinterpretation of antifungal susceptibility and incorrect selection of the initial antifungal agent.

In conclusion, it is important to be aware of the potential misidentification of C. famata by commercial systems. For more accurate identification of these isolates, sequencing analysis of ITS and D1/D2 may be helpful.