Infections caused by Candida spp. are the major cause of morbidity and mortality among seriously ill patients. Prior colonization with Candida spp. has been regarded as an essential step for the development of invasive disease [10, 23]. The colonization index could be helpful in predicting risk of developing systemic infection in critically ill patients, and thus, may offer opportunities for early therapeutic or prophylactic interventions [6, 10]. In the present study, although 11 of 35 (31%) patients were colonized with Candida spp. At ≥3 sites, none of them, despite being leukemic, developed candidemia or invasive candidiasis. Two patients with ALL who subsequently developed candidemia due to C. tropicalis were colonized at two sites with two different Candida spp. (C. albicans with C. tropicalis or C. dubliniensis with C. tropicalis) and their serum samples also yielded positive results for C. tropicalis DNA, mannan and BDG. Recently, Leon et al.  conducted a prospective observational study in a cohort of non-neutropenic patients to assess the value of "Candida score" for the probability of developing invasive candidiasis. Since invasive candidiasis occurred only in <5% of patients who had Candida colonization score of <3, the likelihood of developing invasive candidiasis in such patients was considered very low.
Since early diagnosis of invasive candidiasis is challenging, the role of surrogate markers, such as Candida species-specific DNA, mannan, and BDG in predicting the onset of invasive candidiasis has attracted considerable attention [13, 24]. None of the colonized patients in the present study were found positive for Candida mannan or Candida DNA. However, BDG levels were positive in eight serum samples from six patients with values ranging from 82 pg/ml to 141 pg/ml. These observations are generally in agreement with previous studies showing that patients colonized at single or multiple sites yield BDG levels below the cut-off value recommended by the manufacturer [12, 25–27]. The BDG levels above the cut-off values (80 pg/ml) in 6 of 20 (30%) colonized pediatric cancer patients in our study may either result from absorption of BDG through the gut due to mucositis  or due to contamination with cellulose , gauze , bacterial sepsis [30, 31] or intravenous therapy with amoxicillin-calvulanic acid in these subjects . Despite the above limitations of the test, several studies have used BDG monitoring to identify patients at risk of developing invasive candidiasis to improve therapeutic outcome [25, 33, 34]. A wide range of sensitivities and specificities have been obtained in different study populations [25, 34, 35], probably due to use of different cut-off values for a positive BDG test or due to use of different brand of kits that may react differently to BDG present in the clinical samples [36, 37]. Unlike BDG, Candida mannan levels in serum seem to be less susceptible to the extent of Candida colonization. As stated above, none of our colonized patient was found to have positive serum levels (>0.5 ng/ml) for Candida mannan (mean 0.16 ± 0.04 ng/ml). In two patients who had Candida mannan in the intermediate range (0.308 and 0.287 ng/ml), sera were negative for BDG as well as Candida DNA. These findings are in conformity with the results of previous studies [12, 14].
The normal BDG values in healthy pediatric population are not yet established. In a preliminary study, Brian Smith et al.  estimated BDG levels in serum samples from 120 non-immunocompromised children (0.5 to 18 year-old) and found higher levels (mean 68 ± 128 pg/ml) than those reported earlier in adult population [25, 35]. A noteworthy observation of this study was that 18 of 120 (15%) children had BDG levels >80 pg/ml and 8 of 120 (7%) had BDG levels between 60-79 pg/ml . Thus, higher BDG levels in a minority (6 of 20, 30%) of pediatric cancer patients in our study is consistent with the BDG data reported by Brian Smith et al. . These observations warrant further studies in pediatric population for establishing BDG cut-off values for a positive test to validate the diagnostic utility of this marker for invasive fungal infections.