Candida haemulonii species complex: an emerging species in India and its genetic diversity assessed with multilocus sequence and amplified fragment-length polymorphism analyses

Emerging Microbes and Infections (2016) 5, e49; doi:10.1038/emi.2016.49; published online 25 May 2016

based on the concatenated sequences of the four loci was constructed using MEGA version 6 (http://www.megasoftware.net/). In addition, as previously described, an AFLP dendrogram was generated with Bionumerics v6.6 (Applied-Maths, Sint-Martens-Latem, Belgium) using the standard Pearson and unweighted pair group method with averages (UPGMA). 11 GenBank BLAST searches of the ITS region and MALDI-TOF MS confirmed the presence of 15 isolates belonging to C. duobushaemulonii (n = 8), C. haemulonii (n = 6) and C. haemulonii var. vulnera (n = 1). The ITS sequences of all of the isolates exhibited 100% homologies with the type and reference strains of the respective species in GenBank. The isolates originated from the deep-seated tissue or bone obtained during surgical debridement (n = 9), blood (n = 5) and BAL (n = 1). The isolates exhibited budding yeast cells with pseudohyphae and developed a dark pink color in CHROMagar Candida medium after 48 h of incubation at 37°C. The isolates grew well at 37°C, whereas no growth was observed at 42°C. The NJ tree consisted of three major clades ( Figure 1). Clades 1 and 2 contained C. haemulonii (n = 3 each) and a solitary C. haemulonii var. vulnera isolate in addition to the type and reference strains, and suggested strain variation. The similarity percentages among the C. haemulonii isolates were 79%-89%. However, all of the C. duobushaemulonii clustered together in clade 3. Notably, and as previously reported, the currently available loci (ITS, LSU, RPB1 and RPB2) possessed low discriminatory power for describing the intraspecies genetic diversity of C. haemulonii, C. haemulonii var. vulnera and C. duobushaemulonii. 2 Thus, the inclusion of better polymorphic genes for multilocus sequence analysis is warranted for investigations of the clonal transmissions of these species. The AFLP fingerprint analysis revealed markedly variable banding patterns with~50 bands/strain comprising 25-425 bp overall.
High geometric mean (GM) minimum inhibitory concentrations (MICs) for FLU (CLSI 12.7 mg/L; VITEK2 24.2 mg/L) and AMB (CLSI 14.6 mg/L; VITEK2 6.6 mg/L) were observed via both the CLSI and VITEK2 methods. In contrast, all isolates exhibited low CLSI GM MICs for ISA (0.02 mg/L), POS (0.08 mg/L), ITC (0.3 mg/ L) and VRC (0.134 mg/L) with the exception of a solitary C. haemulonii isolate that exhibited a high MIC for VRC (4 mg/L). Notably, a significant twofold greater CLSI MIC for FLU was observed for C. haemulonii (GM 26.2 mg/L) than for C. duobushaemulonii (GM 6.72 mg/L; Supplementary Table S1). In addition, all of the isolates exhibited low GM MICs for CAS (CLSI 0.15 mg/L; VITEK2 0.30 mg/L), MFG (CLSI 0.33 mg/L; VITEK2 0.10 mg/L) and AFG (CLSI 0.5 mg/L), with the exception being solitary isolates of C. haemulonii var. vulnera and C. duobushaemulonii, which exhibited high MICs (1 mg/L) against all of the echinocandins and high MICs (1 mg/L) for AFG and MFG, respectively, by CLSI. Furthermore, significantly lower (Mann-Whitney Po0.05) GM MICs for VRC, CAS and FC, and higher GM MICs for AMB and MFG were observed by CLSI compared with VITEK2. However, the two methods exhibited good agreement (80%-100%) for AMB, FLU, CAS and MFG, but a low agreement (20%) was noted for VRC. Therefore, standardization of VITEK2 via testing with a larger number of C. haemulonii isolates is warranted to validate the use of the VITEK2 system as an alternative method for routine use in laboratories for susceptibility testing.
The clinical features and outcomes of the 15 patients were retrieved from records (Supplementary Table S2). Nine patients had chronic foot infections, and surgically debrided bone and soft tissues yielded C. haemulonii isolates that included C. haemulonii var. vulnera in five patients and C. duobushaemulonii in four. All nine patients had the major risk factor of uncontrolled diabetes and underwent foot amputations. Two of the nine patients also received azole antifungals. Among the five candidemia patients, two had acute myeloid leukemia and the remaining three patients had the underlying conditions of renal transplant and abdominal surgery. Breakthrough candidemia while on azole prophylaxis (FLU = 3 and VRC = 1) was noted in four patients. VRC therapy was initiated in three candidemia patients and two were successfully managed. The 28-day all-cause mortality among the candidemia patients was 60%. Overall, the patients had at least three predisposing risk factors for candidiasis as previously reported, 12 and these risk factors included the following: 15 (100%) were on broad-spectrum antibiotics, ten (66%) had diabetes mellitus, six (40%) had chronic kidney conditions, five (33.3%) had coronary artery disease, four (27%) had peripheral occlusive vascular disease, three (20%) had malignancies, two (13%) had undergone renal transplant and one had undergone abdominal surgery.
Worldwide, the uncommon Candida species that cause infections are not well characterized. 1, 13 The present study reports the large series of candidiasis cases due to the C. haemulonii species complex with clinical and microbiological data that provide novel insights into this yeast. Unlike its sibling species, C. auris, which is widely prevalent in Indian hospitals, C. haemulonii accounts for only 0.8% of Candida species. Notably, previous reports regarding the occurrence of C. haemulonii in India are misleading because the identifications were performed with conventional or commercial identification systems. 14 Recently, Kathuria et al. 15 reported that of 102 C. haemulonii isolates identified with the VITEK2 system in five centers in India, 88.2% (n = 90) were subsequently confirmed as C. auris by ITS sequencing.
Notably, a high frequency of infections due to the C. haemulonii species complex was observed in the patients with diabetes mellitus (66.6%). Furthermore, as previously reported, the C. haemulonii isolates exhibited elevated MICs to AMB and FLU. 2,3 In addition, the high MICs to echinocandins noted in two isolates (C. haemulonii var. vulnera and C. duobushaemulonii) are worrisome. Thus far, two previous studies have reported high MICs for echinocandins. 2,4 In conclusion, this study highlights the significance of C. haemulonii species in chronic foot infections and candidemia, and the cases of the latter were associated with breakthrough infections, specifically in the patients on FLU therapy. Thus, correct identifications and determinations of the antifungal susceptibilities of C. haemulonii isolates are warranted to guide antifungal therapy. Because the clinical experience with C. haemulonii infections is limited, optimal treatments have not yet been defined. 16 However, the in vitro activities of VRC, POS and ISAV suggest that they can be used as effective treatment options.