Epidemiology of Fungal Colonization in Children Treated at the Department of Oncology and Hematology: Single-Center Experience

Oncological patients are especially predisposed to fungal infections due to multiple risk factors and immunocompromising treatment. Epidemiological research regarding pediatric oncologic patients is still insufficient, and existing data are difficult to generalize on different populations. Therefore, we aimed to analyze fungal infections and fungal epidemiology in the Department of Oncology and Hematology of the University Children’s Hospital in Krakow with help from the Clinical Microbiology Department. During the chosen period of 2005 and 2015–2020, 2342 tests were performed in our ward on 847 patients. Analyzed samples were divided into five source groups. The amount of patients with positive test results was 62.5%. The year with the highest detection level was 2005. The most frequent pathogen was Candida albicans, with a significant decrease in tendency. An increase in non-albicans species was observed. Candida parapsilosis was not frequently observed compared to similar studies. We noticed an increase in positive results from the urinary tract material. Our results confirmed that fungal infections are still an issue, and they may indicate the efficacy of prophylaxis. The majority of our results are consistent with the literature, yet we managed to emphasize data unique to our patients’ population. Our findings are helpful in clinical work and for further studies in our center.


Introduction
Infections are a major issue in the treatment of immunocompromised pediatric patients, as they are a significant cause of morbidity and mortality. Oncological patients are at a high risk of developing complicated infections that are difficult to treat [1][2][3]. Cancer is a state of impaired innate and adaptive immunity due to both treatment effects and the malignancy itself, and it leads to increased susceptibility to infections. Additional factors contributing to a patient's vulnerability are, among others, prolonged leukopenia and neutropenia, malnutrition, disruption of physiological barriers of the mucosa and the skin, and impairment of microbiota or instilment of foreign bodies such as a central venous catheter. Early diagnosis and treatment are important since delays make the infectious process progress quickly, or even generalize. Therefore, prophylaxis, rapid detection, and effective treatment are important factors for overall survival [1].
Fungal infections are not as common as those caused by bacteria or viruses, but they are insidious and can cause severe complications, especially in patients with impaired immunity [1,4,5]. Clinical manifestations of infection may vary from a superficial or cutaneous form to a life-threatening, systemic disease [5][6][7]. Candida spp. and Aspergillus spp. are frequently isolated from immunocompromised patients [5,8,9]. The most common fungi causing sepsis during oncologic treatment are: Candida spp., Aspergillus spp., Zygomycetes, Cryptococci, and Pneumocystis jiroveci [1]. Mycological identification of pathogens is challenging due to problems with obtaining adequate material, as well as culturing itself. Furthermore, isolation of fungi from clinical material may indicate not only an infection or a disease, but also colonization. Interpretation of results is therefore difficult [5]. In addition, currently widely used anti-fungal prophylaxis can further restrict the possibility of an accurate infection diagnosis. All of the above impede evaluation of the true incidence of fungal infections.
The majority of epidemiological data on this topic has come from case studies or singlecenter studies concerning a single fungal species or examining specific populations [10][11][12][13][14][15][16]. Many researchers have investigated a single fungal species in studies published over 10 years ago [17][18][19][20][21]. Furthermore, the knowledge of etiological factors, both in general and typical for clinical facilities, is crucial for focused therapy adjustment. The need for more data in the pediatric population is still valid.
We aimed to retrospectively analyze fungal infections and fungal diagnostics in the pediatric hemato-oncologic patients of the University Children's Hospital in Krakow (UCH) during the period of 2015-2020 from a mycological point of view. We referred to historical data from the randomly selected year of 2005 in order to observe whether any change in incidence appeared.

Materials and Methods
The Clinical Microbiology Department of UCH performed a total of 23,334 mycological diagnostic tests in the year 2005 and during the years 2015-2020, which 2342 (10%) were run for the Department of Oncology and Hematology (DOH). Diagnostics involved mycological and serological testing. The main focus was to analyze the incidence of test results during the period of 2015-2020. The year 2005 was randomly selected as a historical time reference to note possible changes.
Every child admitted to the DOH who underwent mycological testing was included in the analysis, constituting a group of 847 patients (472 male). The population included children with various hemato-oncologic disorders. Samples were collected when there were clinical indications. In our ward, diagnostics were performed in patients with signs of infection. Mycological samples were collected from adequate sources if there was a possibility of fungal etiology. Prophylaxis was used according to the national and EUCAST guidelines. We were solely interested in the amount and results of mycological testing performed by our department; therefore, we did not analyze detailed clinical data. Mycological testing included: 1. direct and stained samples; 2. liquid surface culture possibly with Sabourauda agar culture with or without antibiotics in the temperature of 30 • C; and 3. identification of cultivated yeasts. Identification (no. 3) included: a. filamentation (germ tube test); b. differential surface culture (corn meal agar, differentiating presence or lack of chlamydospores and possibly pseudomycelium), c. pre-prepared biochemical tests ID 32 C (bioMerieux) (bioMérieux SSC Europe Sp.

Results
During the chosen period, 1035/2342 mycological tests were positive. Figure 1 shows the distribution of the mycological tests collected from DOH patients during the studied years, including the number of positive results.
Candida albicans was the most frequently isolated yeast species each year. The prevalence of positive Candida albicans results in the studied years was signif-      Remaining species did not exceed 11% of all cultures. Details can be found in Figure 3. Remaining species did not exceed 11% of all cultures. Details can be found in Figure  3. We additionally analyzed the collected data divided into the sources of examined clinical samples, which can be found below.  We additionally analyzed the collected data divided into the sources of examined clinical samples, which can be found below.    Figure 5.

Year (Amount of Tests
We collected 134 lower respiratory tract samples from 54 patients. Twenty-four different yeast species were isolated. More details are shown in Table 3. Table 3. Yeast species and Saccharomyces isolated from the lower respiratory tract in the studied years. We collected 261 urinary tract samples from 167 patients with 45 yeast isolates detected. More details can be found in Table 4.

Year (Amount of Tests
The prevalence of patients with positive results in the studied years was as follows:  Tables 5 and 6. No significant differences were found between the years.

Discussion
The most common pathogen detected in our patients was Candida albicans. Although this finding is consistent with previously published literature, [20,[22][23][24][25] Candida albicans had a stronger adherence capacity during testing than other Candida species, [9] which could make it easier to detect. Regardless, a significantly higher amount was detected in 2005, with a decrease in subsequent years. Additionally, we observed a significant increase in the detection of Candida glabrata and Saccharomyces in subsequent years. Candida krusei had a significantly higher detection in 2015-2018. Candida lusitaniae detection was higher only in 2016 and 2018 compared to 2005. Nonetheless, the use of prophylactic antifungal drugs could potentially be the reason for the observed shift to non-albicans species [17,23,24,26].
A prolonged and repeated therapy with wide-spectral antibiotics or steroids, use of central venous catheters, and frequent mucositis are predisposing factors to fungal infections in children with hematologic and oncologic diseases [1,27,28]. Adequate antifungal prophylaxis is therefore introduced in these patients, particularly those at higher risk, e.g., children after stem cell transplantation or patients with prolonged neutropenia, according to national and EUCAST guidelines [1,5,22]. Although often necessary, longitudinal prophylaxis can result in drug resistance. An example is Candida lusitaniae resistance and the stimulation of its colonization after treatment with amphotericin B and decreased sensitivity to the drug [29,30]. Similar effects have been observed in a prophylactic use of fluconazole with Candida glabrata and Candida krusei colonization [31,32]. In addition, probiotics used in patients with immunosuppressive disorders have been observed to cause colonization, possibly with sepsis, with Saccharomyces [33][34][35].
In contrast to other centers, we detected fewer Candida parapsilosis isolates [11,17,24,36]. In general, the majority of patients had positive fungal tests results (62.5%), meaning the decision to diagnose was justified. In the described centers, fluconazole has been routinely used as prophylaxis in majority of cases. A decrease in Candida albicans detection suggests that an anti-fungal prophylaxis was successful. Other studies have confirmed a decrease in fungal infection prevalence [13,22,37,38]. In addition, the highest number of patients was admitted for treatment in 2018. Fewer patients were admitted to the Department in 2020 due to the COVID-19 pandemic, which might have influenced the diversity of the results. Despite the fact that the percentage of positive results this year was not significantly different, it might not be fully reliable.
We noticed a significant difference in positive fungal samples detected from the urinary tract between 2015 and 2018-2019. Candida albicans was the leading pathogen in this sample source, similar to other referred studies [11]. This observation is important, suggesting that urinary tract infections could end with fungemia [39,40], and need more inspection.
Our experience indicates that blood cultures and antigen testing do not add important information to fungal diagnostics. Colonization with fungal pathogens, Candida albicans in particular, is very common, making antigen testing less useful without information on clinical presentation. Our results showed that serological detection was very limited. Nevertheless, if an increase in antibody levels was detected, it could suggest the status of fungal infection [41][42][43]. We suggest that detection of antigens can be more informative in DOH patients. Fungal antigens were detected more frequently than antibodies. Thus, antigen testing could be treated as a screening method in immunocompromised patients. Establishing a validated method for the detection of fungal antigens in clinical samples could be beneficial and helpful in predicting upcoming infections.
Several large studies on invasive fungal infections have been previously conducted [44][45][46] with the use of EORTC-IFG/NIAID-MSG classification criteria [47], which we did not use due to a lack of clinical data. Our study focused on fungal epidemiology in DOH patients mainly from a mycological point of view. This could be regarded as a limitation, as we could not be certain whether patients dealt with an infection or colonization. Nonetheless, the samples were collected from cases with occurring symptoms. In addition, the analyzed data came from a single center of a large, tertiary care pediatric referral hospital. Despite the limitations, we consider the presented data useful and a good introduction to further studies, as it is the first study presenting fungal testing results from our center. The knowledge of the particular epidemiology in our center is important for making therapeutic decisions regarding antifungal medication. Existing data from other centers should be carefully generalized, since there may be differences in patient population, risk factors, and treatment demeanor. Regardless, the prevalence of fungal colonization in oncological and hematological patients confirmed the importance of implementing prophylaxis in immunocompromised children with a high risk of infection in order to prevent generalized opportunistic infection.

Conclusions
The majority of patients who qualified for mycological diagnostics each year had positive test results, indicating that fungal infections are still a concern for our patients. Nonetheless, a significant decrease in Candida albicans compared the benchmark set at the year 2005, as well as an increase in non-albicans species, suggested that the cause was prophylaxis, despite the fact that Candida albicans was still the most frequently detected culture. Contrary to similar studies, we detected a low amount of Candida parapsilosis isolates. The differences found in the urinary tract material were significant and need more insight. We concluded that our analysis provides good suggestions for clinical work and a baseline for further research regarding this topic, emphasizing the data unique to our center.