The study was carried out in a tertiary care Pediatric Cancer Ward, Al-Sabah Hospital, Kuwait between July 1, 2007 to December 31, 2008. Sixty-three cancer patients, 57 (90%) with acute lymphoblastic leukemia (ALL) and 6 (10%) with acute myeloid leukemia (AML) were followed-up by weekly surveillance cultures for varying periods for assessing the extent of Candida colonization. Forty-five patients were males. Their age ranged from 1 to 16 years. A child was considered as colonized if Candida sp. was isolated from one or more anatomic sites. A patient yielding Candida sp. on repeat cultures at least from one site was considered as persistently colonized [¹⁵]. The study was approved by the Ethics Committees of the Faculty of Medicine, Kuwait University and Ministry of Health, Kuwait. Informed consent of the patients was obtained before collecting the clinical samples.

A total of 1075 swabs originating from oropharynx (n = 294), nasal (n = 600), rectum (n = 28), groin (n = 50), ear (n = 54), and axilla (n = 49) of 63 pediatric cancer patients were cultured on Sabouraud dextrose agar supplemented with chloramphenicol (Table 1). The germ tube test was performed on all the Candida spp. isolates for the presumptive identification of C. albicans or C. dubliniensis. Subsequently, Candida isolates were also identified by Vitek2 yeast identification system (BioMerieux, France). The identification of Candida spp. isolates was also confirmed by species-specific amplification and/or sequencing of internally transcribed spacer (ITS) region of rDNA.

Five ml of blood was collected in sterile BDG-free clotting tubes and serum was separated for the detection of (1-3)-β-D-glucan (BDG), Candida mannan and species-specific Candida DNA at the time of surveillance culture. The serum was kept frozen at -20°C until used. Thirty-six serum samples from 20 colonized patients and 11 serum samples from nine non-colonized patients were tested.

The BDG levels in serum samples were determined using a Fungitell kit (Associates of Cape Cod Inc., East Falmouth, MA, USA) according to the procedure described by the manufacturer. BDG levels were assayed against a purified Pachyman standard, which included a five-point two-fold curve ranging from 31 pg/ml to 500 pg/ml. In brief, 5 μl of serum was dispensed per well in duplicate and pretreated with 20 μl of 0.25 M KOH and 1.2 M KCl for 10 min at 37°C. This step inactivated protease and other inhibitors present in human serum. The Fungitell BG reagent was then reconstituted and dispensed according to the instructions supplied by the manufacturer of the Fungitell kit. A Microplate Spectrphotometer (Bio-Tek Instruments, Inc., Winooski, VT, USA) with Gen5™ software onboard was used to accomplish kinetic analysis of the microtiter plate. The BDG value of ≥80 pg/ml was considered as positive and a value between 60-79 pg/ml as intermediate.

Mannan antigen was measured by Platelia Candida Ag (BioRad, Marnes La Coquette, France). The test was performed according to the instructions of the manufacturer. Briefly, each test serum (300 μl) was mixed with 100 μl of treatment solution and placed in a boiling water bath for 3 minutes. After centrifugation, the supernatant was used for further testing. Fifty-μl of the conjugate and an equal amount of the treated serum supernatant was introduced into micro-titer plate wells pre-coated with anti-mannan monoclonal antibody. After incubation at 37°C for 90 min and 5 washing steps, 200 μl of the substrate buffer was added to each well, and the plates were incubated for 30 min at room temperature. The enzymatic reaction was terminated by adding stopping solution and the optical density was read at 450 nm using a Microplate Spectrphotometer (Bio-Tek Instruments, Inc.). The reactions were performed in duplicates and each experiment included positive and negative controls as well as a calibration curve. The calibration curve was made with a pool of normal human serum supplemented with known concentrations of mannan ranging from 0.1 to 2 ng/ml. A value of ≥0.5 ng/ml was taken as positive and ≥0.25 ng/ml but ≤0.5 ng/ml as doubtful.

DNA from cultured Candida spp. was isolated as described in detail previously [¹⁶,¹⁷]. DNA from serum was extracted using the QIAamp DNA kit (QIAGEN, Hilden, Germany) by following the instructions supplied by the manufacturer. DNA sequences of pan-fungal and species-specific forward and reverse primers and DNA amplification protocol were same as described previously [¹⁶,¹⁸]. Cultures of C. albicans ATCC 90029, C. parapsilosis ATCC 10233, C. tropicalis ATCC 750, C. glabrata ATCC 90030 and C. dubliniensis CBS 7987/CD36 were used as reference for amplification of specific products. The amplified DNA fragments were detected by agarose gel electrophoresis using 2% agarose gels as described previously [¹⁹].

The DNA isolated from selected isolates was also subjected to direct DNA sequencing of ITS region of rDNA (containing the ITS-1, 5.8S rRNA and ITS-2) to confirm species-specific identification by PCR. The ITS region was amplified by using ITS1 and ITS4 primers and both strands of amplified DNA were sequenced as described previously [²⁰,²¹]. The sequencing primers, in addition to the amplification primers, included ITS1FS, ITS2, ITS3 and ITS4RS and the sequences were assembled as described previously [²²]. GenBank basic local alignment search tool (BLAST) searches (http://www.ncbi.nlm.nih.gov/BLAST/Blast.cgi?) were performed for species identification

Mann-Whitney 2-tailed test was applied to determine the significance of differences that existed between BDG or Candida mannan levels between different patient groups who were colonized with Candida spp. on single occasion or persistently and non-colonized subjects. The Spearman correlation test was performed to determine the correlation between BDG and Candida mannan. A P value of <0.05 was considered as significant.

The results of the surveillance cultures for Candida spp. are presented in Tables 1^-2. Of 1075 swabs cultured from different anatomic sites of 63 pediatric cancer patients, 75 (7.8%) were positive for Candida spp. These isolates included C. albicans (n = 62), C. dubliniensis (n = 8), C. glabrata and C. tropicalis (n = 2 each) and C. krusei (n = 1) (Tables 1^-2). Apart from phenotypic identification, all Candida spp. isolates were also identified by species-specific amplification of ITS region of rDNA (data not shown). The identity of six selected isolates was further confirmed by direct DNA sequencing of ITS region of rDNA (EMBL Accession Nos. FN652297, FN652298, FN652301 to FN652304). The distribution of anatomic sites yielding Candida spp. in culture was as follows: rectum, 36%; oropharynx, 18%; groin, 10%; ear, 2% and nasal, 1% (Table 1). Seventeen patients were colonized at one site and 18 at two or more sites. None of the swabs taken from axilla were positive for Candida spp. Of the 8 C. dubliniensis isolates, 5 came from oropharynx, and one each from nose, groin and rectum of six patients. Two patients yielded C. glabrata (nose and rectum) and two others yielded C. tropicalis (oropharynx and rectal). A single isolate of C. krusei was recovered from the oropharynx (Table 2).

In colonized patients, the mean Candida mannan value was 0.16 ± 0.044 ng/ml, which was not significantly different from those that were not colonized (p = 0.660) (Table 3). There was also no significant difference in the mean Candida mannan values between patients colonized at one anatomic site or two or more sites (p = 0.665) or between those that were colonized once or those yielding Candida spp. repeatedly (p = 0.474) (Table 3). Although mean BDG values in colonized patients were nearly same as non-colonized patients (46.98 pg/ml vs. 36.77 pg/ml), 8 serum samples from 6 colonized patients were positive for BDG (range 82 pg/ml to 141 pg/ml, mean = 98.3 pg/ml). Three of these serum samples were obtained from the same patient within a span of 40 days and the BDG levels varied between 85 pg/ml to 115 pg/ml. However, none of these 8 serum samples yielded mannan levels >0.25 ng/ml. Additionally, 5 serum samples from two colonized patients yielded BDG values in the intermediate range (Table 3). There was no significant difference in the mean BDG values between patients that were colonized once or those that yielded Candida spp. persistently (p = 1.0) (Table 3). The mean and standard deviation of BDG and Candida mannan levels in serum samples collected from the patients at the time of surveillance samples were 42.74 ± 30.18 pg/ml and 0.173 ± 0.03 ng/ml, respectively. No correlation was observed between BDG and Candida mannan values among patients that were colonized with Candida species by Spearman correlation test (p = 0.531, R = 0.108). None of the serum sample from the colonized patients was positive for the detection of Candida DNA by PCR.

Two patients whose oropharyngeal or rectal samples yielded C. tropicalis (in addition to C. dubliniensis in one and C. albicans in the other) developed candidemia due to C. tropicalis. Besides blood culture positivity, both these patients showed presence of C. tropicalis DNA in serum by species-specific PCR amplification (Fig. 1) and also showed elevated levels of BDG (238.6 and 406 pg/ml), whereas Candida mannan was positive in one (0.77 ng/ml) and border line (0.28 ng/ml) in the other.

GM, Geometric mean; SD, Standard deviation