Candida parapsilosis candidemia in children admitted to a tertiary hospital in Turkey: clinical features and antifungal susceptibility

ABSTRACT In recent years, the incidence and drug resistance of Candida parapsilosis have increased. Our study aimed to determine the antifungal sensitivity of C. parapsilosis and the clinical and demographic characteristics of children with candidemia. Two hundred pediatric patients with C. parapsilosis candidemia were included in the study between 1 January 2010 and 1 August 2023. Clinical samples were evaluated on a BACTEC-FX-40 automatic blood culture device (Becton Dickinson, USA). Yeast isolates were identified to the species level via identification cards (YST) using the VITEK 2 Compact (bioMeriéux, France) system. Antifungal susceptibility was performed using antifungal cell cards (AST-YST01). Approval for the study was received from the “University Faculty of Medicine” Hospital Clinical Research Ethics Committee. Non-catheter candidemia was detected in 127 (63.5%) patients, and catheter-related candidemia was detected in 73 (36.5%) patients. It was observed that the patients’ history of malignancy, mechanical ventilation, urinary catheter, nasogastric tube, and intensive care unit stay was associated with C. parapsilosis mortality. The mortality rate from candidemia was 9.5%. The most frequently preferred antifungal agents were amphotericin B and fluconazole. The fluconazole drug resistance rate was found to be 6%, and the amphotericin B drug resistance rate was 4%. Because C. parapsilosis candidemia mortality rates can be high depending on risk factors and clinical characteristics, it is important to initiate appropriate and timely antifungal therapy. We think that our study can provide important information about the clinical profiles, distributions, susceptibility profiles, and control of antifungal resistance of C. parapsilosis isolates. IMPORTANCE It has been observed that the frequency and antifungal resistance of Candida parapsilosis have increased recently. In our study, we aimed to determine the antifungal sensitivity of C. parapsilosis and the clinical and demographic characteristics of children with candidemia. It was observed that the patients’ history of malignancy, mechanical ventilation, urinary catheter, nasogastric tube, and intensive care stay was associated with C. parapsilosis mortality. The mortality rate from candidemia was 9.5%. The most frequently preferred antifungal agents were amphotericin B and fluconazole. The fluconazole drug resistance rate was found to be 6%, and the amphotericin B drug resistance rate was 4%. Because C. parapsilosis candidemia mortality rates can be high depending on risk factors and clinical characteristics, it is important to initiate appropriate and timely antifungal therapy.

T he incidence of invasive fungal infections is increasing worldwide, and these infections are associated with high morbidity and mortality.Candida species are the most common cause of invasive fungal infections (1).
More than 90% of invasive candidiasis is caused by Candida albicans, Candida parapsilosis, Candida glabrata, Candida tropicalis, and Candida krusei.The epidemiology of invasive candidiasis has been changing in recent years, and the most common non-albicans strain is C. parapsilosis (2,3).C. parapsilosis is frequently isolated from patients with malignancies, immunocompromised patients, and patients hospitalized for long periods in the intensive care unit (4).Additionally, the frequency of non-Candida albicans species causing nosocomial infections has increased in the last few decades (5).
Conditions such as multiple antibiotic therapy, total parenteral nutrition (TPN), nasogastric tube, mechanical ventilation, admission to intensive care, and neutropenia are risk factors for candidemia (1).Since mortality rates in candidemia (40%-60%) may be high depending on risk factors and clinical characteristics, it is important to initiate appropriate and timely antifungal treatment (6).For effective antifungal treatment, it is important to correctly identify the candidemia agent and know its sensitivity to antifungal agents (6,7).
There are very few studies and information regarding antifungal drug resistance in C. parapsilosis.In particular, the increasing use of antifungal agents may lead to resistant Candida species (8).Antifungal susceptibility results of C. parapsilosis strains may affect treatment choices and, as a result, morbidity and mortality rates may change.
This study aimed to determine the antifungal sensitivity of C. parapsilosis and the clinical and demographic characteristics of pediatric patients with candidemia.

MATERIALS AND METHODS
This is a retrospective study conducted in a single-center tertiary hospital.Pediatric patients whose blood and catheter cultures showed C. parapsilosis growth and whose signs and symptoms were compatible with candidemia between 1 January 2010 and 1 August 2023 were included in our study.Patients older than 18 years and patients with insufficient data were excluded.
If the patient with a bloodstream infection and catheter does not have another focus of infection, the same microorganism is produced from peripheral blood culture and semiquantitative (>15 cfu) or quantitative (>100 cfu) culture of the catheter tip or in simultaneous quantitative blood cultures.A catheter-related infection was defined as a growth rate of ≥3/1 in the central venous catheter/peripheral blood culture or detection of growth >2 hours earlier in the blood culture taken from the central venous catheter than in the simultaneously taken peripheral blood culture.Candidemia cases that did not meet this criterion were defined as non-catheter-associated candidemia (9,10).In patients with previous candidemia, Candida overgrowth 30 days after the first negative culture was considered a new episode of candidemia.Culture and antibiogram data regarding candidemia were obtained retrospectively from microbiology labora tory records.Demographic and clinical characteristics, underlying disease, risk factors, antifungal susceptibility test results, treatment, and prognosis of infections caused by C. parapsilosis were obtained from patient files.
The time from hospital admission to the growth of Candida species in the blood culture was recorded.Mortality within 30 days after Candida overgrowth was defined as the overall mortality and mortality rate due to candida.The overall mortality rate was calculated as the number of patients who died/all patients, regardless of the cause of death.The candidemia-related mortality rate was calculated as the percentage of patients who died from candidemia/all patients (11).Accordingly, 7-day and 30-day mortality rates were calculated.
Organ scanning was performed on all patients with abdominal ultrasonography, abdominal tomography if necessary, eye examination, and echocardiography.
Clinical samples sent to the microbiology laboratory with appropriate blood culture bottles were incubated in the BACTEC-FX-40 automatic blood culture device (Becton Dickinson, USA) and monitored for 7 days.Growth signals were seen in the smear preparations taken from the bottles, and Gram staining was performed.In addition, these samples were passed through sheep blood agar, MacConkey agar, and Sabour aud dextrose agar.Colony morphology was evaluated after 18-24 hours incubation at 37°C.Isolates compatible with yeast morphology were identified to species level using identification cards (YST) with the VITEK 2 Compact (bioMeriéux, France) system.Due to laboratory conditions, C. parapsilosis complex subspecies could not be identified.C. parapsilosis refers to all strains of the C. parapsilosis complex.At the same time, ampho tericin B, caspofungin, micafungin, fluconazole, flucytosine, and voriconazole sensitivi ties were evaluated using antifungal sensitivity cards (AST-YST01).Minimum inhibitory concentration (MIC) limit values were determined in the antibiotic sensitivity test.C. parapsilosis MIC breakpoint was evaluated according to the CLSI M60 2017 guideline (12).MIC ≥ 8 mg/L for fluconazole, MIC ≥ 8 mg/L for caspofungin, MIC ≥ 8 mg/L for micafungin, and MIC ≥ 1 mg/L for voriconazole were considered resistant (13).
Numerical measurements were summarized as mean ± standard deviation and quarter range, and categorical measurements were summarized as number (percentage).χ 2 test statistics were used to compare categorical measures.IBM SPSS Statistics Version 20.0 (IBM Corp., Armonk, NY, USA) package program was used for statistical data analysis.Statistical significance level in the tests was taken as 0.05 (14).Approval for the study was received from "University Faculty of Medicine" Hospital Clinical Research Ethics Committee (decision number 18, meeting dated 5 December 2023).

RESULTS
Between 1 January 2010 and 1 August 2023, 471 C. parapsilosis strains were detected in blood-catheter cultures at University Hospital.Additionally, C. parapsilosis was found to be the most common species after C. albicans.Of these patients, 200 were included in the study, excluding patients who had C. parapsilosis growth again without culture negativity within 1 month and whose data could not be accessed.
Of the patients included in the study, 126 (63%) were male.The mean age of the patients was 47.6 months.The distribution of the patients regarding the services in which they were hospitalized is presented in Fig. 1.Most patients were hospitalized in pediatric hematology-oncology and pediatric intensive care units.Eight (4%) patients did not have any chronic disease.Symptoms included fever in 150 (75%), hypothermia in 1 (0.5%), vomiting in 41 (20.5%), diarrhea in 24 (12%), cough in 10 (5%), respiratory distress in 88 (44%), sleepiness in 17 (8.5%),and abdominal pain in 28 (14%).One hundred forty-three of the patients had an intravenous catheter.Catheter-related candidemia was detected in 73 (36.5%) patients, and non-catheter-related candidemia was detected in 127 (63.5%) patients.
All patients with candidemia had at least one of the risk factors examined.Clinical and demographic characteristics, underlying chronic diseases, and risk factors for candidemia are presented in Table 1.A significant association was found between malignancy and candidemia-related mortality (P = 0.048).In addition, TPN, presence of a central venous catheter, and steroid use were significantly associated with overall mortality (P = 0.006, P = 0.011, P = 0.019, respectively).Nasogastric catheter use did not affect overall mortality, but it was found to affect candidemia-related mortality (P = 0.011).The relationship between risk factors, underlying diseases, and overall and candidemia-related mortality is presented in Table 2.The distribution of C. parapsilosis strains by years was variable, and the highest rate (14%) was found in 2018.The rate of C. parapsilosis candidemia per 1,000 admissions is shown in Fig. 2.
Fluconazole prophylaxis was administered before treatment to 22 (11%) patients with newborns weighing <1,500 g and hematological-oncological diseases.No resistance to antifungal agents was detected when the antibiogram susceptibility tests of the patients who received prophylaxis were analyzed.
C. parapsilosis growth was accompanied by other microorganisms in 27 (13.5%)patients.These were serious life-threatening microorganisms such as Acinetobacter The mean duration of C. parapsilosis growth after hospitalization was 41.2 ± 60.3 (3-408) days.The mean duration of C. parapsilosis growth after catheterization was 5.3 ± 5 (0-28) days.Nine (4.5%) patients received antifungal amphotericin B lock and systemic therapy.Since the catheters of these patients could not be removed at that time due to their clinical conditions, they were given lock therapy with systemic treatment, and the catheters of all patients were removed as soon as their clinical conditions improved.Of the 73 patients with catheter-associated candidemia, 71 (97.9%) had catheter removal, and 2 (2.7%) did not.It was found that the families of these two patients did not allow port removal.One of the two patients whose port could not be removed died.No ocular or intra-abdominal involvement was detected in C. parapsilosis organ involvement.Endocarditis was detected in two (1%) patients.
Of the patients, 194 (97%) had received antifungal treatment for candidemia.Six (3%) patients did not receive antifungal treatment.Of these six patients, two died and four are alive.One hundred sixty-five (82.5%) patients received monotherapy.The most preferred antifungal agents were amphotericin B and fluconazole.The patients' treatment regimens are shown in Table 3. Amphotericin B was most commonly used in the treatment of catheter-associated candidemia.The mean duration of treatment was 19 ± 9.4 (1-64) days.
The antifungal drug MIC values of C. parapsilosis strains are presented in Table 4.The susceptibility-resistance rates of antifungal drugs are also shown in Table 5.Looking at the resistance rates by year, amphotericin B and fluconazole resistance rates were highest a The disease is characterized by the inability to synthesize biochemical substances, abnormal accumulation, and damage to brain tissue as a result of mutations that cause biochemical changes in metabolic pathways.combined treatment, 1 received sequential treatment, and 2 did not receive antifungal treatment.These two patients died before the culture results were obtained.

DISCUSSION
C. parapsilosis is one of the most common invasive fungal infections among non-albicans Candida species.In South America and some European countries, C. parapsilosis was the most frequently isolated non-albicans Candida species in cases of candidemia (15,16).C. parapsilosis has a high reproduction rate, adheres to intravenous devices and prosthetic materials, and has the ability to form biofilms, gastrointestinal colonization, and the possibility of transmission from the colonized hands of healthcare personnel to the patient (17,18).In a study conducted in the pediatric population, the presence of surgical intervention, Candida colonization, chronic lung disease, the presence of indwelling urinary and central venous catheters, and total parenteral nutrition were shown to be risk factors for C. parapsilosis candidemia.In this study, mortality rates in patients with C. albicans, C. parapsilosis, and non-albicans growth were 15.8%, 26%, and 9.5%, respectively (19).In a cohort study, the mortality rate of candidemia was found to be 10%-49% (20).Studies have shown that the most important risk factors for candide mia due to C. parapsilosis are the use of central catheters and parenteral nutrition (21,22).In our study, malignancy, mechanical ventilation, urinary catheter, and nasogastric catheter use were important factors in candidemia mortality, and C. parapsilosis mortality rate was 9.5%.a For three patients, the amphotericin B MIC value was 2 and they were resistant.
b For two patients, the voriconazole MIC value was 1 and they were resistant.
When the distribution of C. parapsilosis growth by years was analyzed, it was observed that it was the highest in 2018 and decreased in the following years.We think that this situation is related to the decrease in the admission rates of non-COVID (non-coronavirus disease) diseases to the hospital with the COVID epidemic in the world.In addition, we think that the reason for the low number of cases in 2023 is that our hospital was damaged in our region by the 6 February 2023 earthquake and was partially active at that time.
Accurate identification of the causative agent of fungal infection and determination of the in vitro antifungal susceptibility profile make important contributions to the treatment of the patient (23).The emergence of antifungal-resistant strains in nonalbicans Candida species has again emphasized the importance of in vitro antifungal susceptibility testing.Sutcu et al. evaluated 54 pediatric patients with invasive candide mia between 2012 and 2016 at Istanbul Medical Faculty Hospital, a tertiary care facility.C. parapsilosis isolates were detected in 13 of them, and amphotericin B and fluconazole resistance was not found in any of them (24).Resistance may not have been detected in this study due to the low number of isolates.In the pediatric study conducted at the tertiary care Izmir Tepecik Hospital, a total of 126 candidemia cases, most commonly C. parapsilosis, were detected from 2012 to 2018; within a 7-year period, 33.8% of   (26).
In their study, Fidan et al. found the fluconazole resistance rate of C. parapsilosis grown in blood cultures of patients hospitalized in the intensive care unit to be 78.6%(27).We think that this high fluconazole resistance rate is related to the very small number of patients in the study, the fact that the patients were intensive care patients, or the hospital's own resistant flora.In our study, the fluconazole resistance rate of C. parapsilosis grown in blood cultures of patients hospitalized in the intensive care unit was 5.7%, which is very low.We think the wide range in this drug resistance rate is due to differences in age group, risk factors, region, and hospital-acquired C. parapsilosis strains.In addition, our study found no increase in resistance rate in patients receiving fluconazole prophylaxis.Close follow-up of antifungal susceptibility profiles in each hospital and even between different wards within the hospital is of great importance in empirical treatment selection.
Etiz et al. found that the resistance rate of C. parapsilosis to voriconazole was 1%, as in our study.In this study, C. parapsilosis flucytosine resistance was not detected (28).In our study, flucytosine drug resistance was found in 0.5%.Pfaller et al. found that the resistance rate to voriconazole in C. parapsilosis was 4.3% among 280 non-albicans Candida species in the SENTRY antimicrobial surveillance program.In addition, in this study, 1.1% of C. parapsilosis species showed intermediate susceptibility to voriconazole (29).Similar results were found in our study.
In invasive candidemia, length of hospital stay, underlying malignancy and immu nodeficiency, cardiac failure, use of indwelling urinary catheter and central venous catheter, total parenteral nutrition, and dialysis are factors associated with mortality (19).Knowing local epidemiological data and risk factors in invasive Candida infections is very important for empirical treatment.
Catheter removal is of vital importance in catheter-associated bloodstream infections.In our study, one of the two patients whose catheter could not be removed died, and it is seen how important catheter removal is.In a study by Tsai et al., it was shown that only lock treatment in pediatric malignancy patients caused poor results on catheter-associ ated candidemia mortality (30).Another study suggests that adding an antimicrobial lock solution to systemic antibiotics in pediatric patients may benefit catheter salvage, depending on the etiology (31).
In the treatment of C. parapsilosis, fluconazole or amphotericin B is recommended in neutropenic patients, fluconazole can be used in non-neutropenic patients, and if echinocandin is started empirically, it can be continued if there is a clinical and microbio logical response; in case of intolerance to other agents or limited benefit, amphotericin B treatment should be given (21).
As a result, it appears that the risk of C. parapsilosis candidemia is high in patients who are monitored using mechanical ventilation devices and have nasogastric and urinary catheters and central catheters.It should be considered that response to treatment will be difficult, especially in patients with malignancy and these risk factors.
Considering that antifungal drugs develop resistance against C. parapsilosis, we think that performing antifungal sensitivity tests in every center will guide empirical treatment and that empirical antifungal use policies in clinics should be re-evaluated in the light of this information.The retrospective nature of our study and the lack of isolate subtyping of C. parapsilosis complex members created limitations in our study.We believe that for treating catheter-related candidemia, the catheter should be removed first; if it cannot be removed, lock therapy should be administered along with systemic therapy and the catheter should be removed as soon as possible.
baumannii, Stenotrophomonas maltophilia, Serratia marcescens, Enterococcus faecium, Staphylococcus hominis, Staphylococcus epidermidis, coagulase negative staphylococci, Candida famata, C. glabrata, and C. pellucida.Microorganism growth in addition to C. parapsilosis growth was observed in 3 of 19 patients (15.8%) who died due to candide mia.These were Acinetobacter baumannii in one patient, Stenotrophomonas maltophilia and Serratia marcescens in one patient, and Candida pellucida in one patient.

a
Data are summarized as number (percentage).b Data are summarized as mean ± standard deviation and median (IQR, interquartile range).

TABLE 2
Comparison between alive and dead patients ab (Continued on next page) in 2021.No correlation was found between antifungal drug resistance and mortality (Table6).The overall mortality rate was 23.5%, and the mortality rate of candidemia was 9.5%.Of the 19 patients who died from candidemia, 14 received monotherapy, 2 received

TABLE 2
Comparison between alive and dead patients ab (Continued)

TABLE 3
Summary of patient treatment regimens

TABLE 4
Candida parapsilosis MIC (mg/L) distribution of strains

TABLE 5
Distribution of Candida parapsilosis according to antifungal susceptibility b The patient for whom drug resistance testing to the antifungal agent was not performed.b Data are summarized as number (percentage). a

TABLE 6
Distribution of antifungal susceptibility according to mortality a,c

Alive (n = 153) Overall mortality (n = 47) P b Candidemia-related mortality (n = 19) P b
(7))arapsilosis strains were found to be resistant to fluconazole, which is a high rate(25).In the study conducted by Lotfali et al., in which 120 C. parapsilosis isolates were examined, they showed that there was 2.5% resistance to fluconazole and 1.7% resistance to amphotericin B(7).In our study, the fluconazole resistance rate was found to be 6%, and the amphotericin B resistance rate was 4%.Blood cultures sent to Kayseri Research Hospital Microbiology Laboratory from various wards and intensive care units between January 2012 and June 2014 were analyzed bySarıgüzel etal.It was examined retrospectively.Of all the yeast strains, 67 (38.3%) were C. parapsilosis.It was determined that fluconazole resistance was 10.4%, and amphotericin B resistance was 8.9% in C. parapsilosis strains a Data are summarized as number (percentage).b P < 0.05 was considered significant.c The χ 2 test was performed to compare categorical variables between groups.d S, sensitive; R, resistance; I, intermediate.71