A Temporal Candidemia Cluster in a Jordanian Pediatric Oncology Unit: Clinical Characterization, Putative Environmental Source, and Multidisciplinary Infection Control Response

1. Introduction

Candidemia is the most common invasive fungal infection in pediatric oncology wards, with consequences that are disproportionately severe in patients who are neutropenic, CVC-dependent, and receiving broad-spectrum antibiotics. Among immunocompromised hosts, Candida species account for a substantial proportion of all invasive fungal disease (IFD). Multiple host factors independently increase risk, including prolonged neutropenia, chemotherapy-related mucosal barrier disruption, CVC use, broad-spectrum antibiotic pressure, and corticosteroid administration [1,2]. In low- and middle-income country (LMIC) settings, non-culture diagnostics (β-D-glucan, PCR) are largely unavailable, antifungal formularies are often limited to fluconazole and amphotericin B deoxycholate, and infection control infrastructure is under-resourced — each independently worsening candidemia outcomes.

The epidemiology of candidemia in pediatric oncology units across the Middle East and the broader LMIC region is shaped by the predominance of non-albicans Candida species — particularly C. tropicalis — and rising azole resistance rates that approach 20–50% in some surveillance cohorts [3,4].

Beyond its epidemiological prevalence, C. tropicalis exhibits specific virulence attributes that distinguish it from C. albicans in the neutropenic host. Superior biofilm formation on vascular catheter surfaces facilitates persistent bloodstream seeding; visceral tropism drives hepatosplenic candidiasis at rates disproportionate to other species; and secreted protease and phospholipase activity promotes tissue invasion [5]. Azole resistance in C. tropicalis is predominantly mediated by ERG11 point mutations — notably Y132F and A114V — and UPC2-mediated upregulation of sterol biosynthesis genes, conferring cross-resistance between fluconazole and voriconazole and rendering standard azole prophylaxis ineffective against resistant strains [6]. Echinocandin resistance, driven by FKS1 and FKS2 gain-of-function mutations, has been reported in C. glabrata and, less commonly, in C. tropicalis, representing a serious threat to the last reliable empirical prophylaxis class [11]. Collectively, these mechanisms underscore the imperative for local antifungal resistance surveillance and stewardship-guided prophylaxis selection, rather than empirical azole use in settings where azole resistance prevalence is known to be high.

Temporal clusters of candidemia in hematology-oncology units — defined as a higher-than-expected number of cases within a defined period, location, and patient population — have been attributed to environmental sources including hospital water systems, construction activity, and failures in air handling [7,8]. Environmental water sources are of particular relevance because hospital plumbing biofilms can simultaneously harbor multiple Candida species, producing mixed-species clusters without clonal transmission [9]. Early recognition of such clusters depends on prospective microbiology surveillance and a clearly defined case definition that separates culture-confirmed from clinically suspected disease.

We report a temporal candidemia cluster at KHCC — a 330-bed tertiary oncology center and Jordan’s primary national cancer referral institution — occurring between June and September 2025 and coinciding with a documented environmental event. We describe the clinical and microbiological characteristics of culture-confirmed cases, the putative environmental context, and the multidisciplinary IPC response. To our knowledge, this represents one of the first published candidemia cluster investigations from a pediatric oncology unit in Jordan, contributing to the sparse literature on candidemia clusters from the Levant region.

2. Methods

2.1. Study Design and Setting

This was a retrospective case series conducted at KHCC, Amman, Jordan. The cluster observation period spanned June 1 to September 30, 2025. A documented background rate of 1 culture-confirmed candidemia case in the preceding 6 months (December 2024 to May 2025) served as the epidemiological baseline against which the cluster was defined, consistent with WHO criteria for a temporal cluster [10].

2.2. Case Definitions

Culture-confirmed candidemia (primary analysis): Isolation of a Candida species from at least one blood culture drawn from a peripheral vein or CVC in a patient with compatible clinical signs. These cases constitute the primary analytical cohort.

Clinically suspected IFI (secondary descriptive group): Patients with host risk factors consistent with modified ECMM/ISHAM criteria for probable IFI and compatible clinical or radiologic findings, but with negative blood cultures throughout. These patients are described separately and are explicitly excluded from all candidemia-specific analyses. The term “candidemia” is not applied to this group.

2.3. Microbiological Methods

Species identification from positive blood culture bottles was performed using the BioFire FilmArray Blood Culture Identification (BCID) panel (bioMérieux, Salt Lake City, UT, USA), a multiplex PCR-based platform that detects fungal and bacterial pathogens directly from flagged blood culture bottles. Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) was additionally available for species identification as required. Antifungal susceptibility testing was performed using Etest gradient diffusion strips (bioMérieux) applied to all culture-positive isolates, with minimum inhibitory concentrations (MICs) reported as susceptible (S), intermediate (I), or resistant (R) per the laboratory reference ranges in use at KHCC. C. krusei was classified as intrinsically resistant to fluconazole per standard microbiological criteria. No echinocandin resistance was detected in any isolate.

2.4. Data Collection

Data were extracted from electronic medical records by the infection control team. Variables included patient demographics, underlying malignancy and disease status, clinical risk factors, microbiological results, antifungal therapy regimens, and 30-day outcomes.

2.5. Cluster Investigation and IPC Response

2.5.1. Cluster Identification

The cluster was identified through prospective daily review of microbiology reports from June 1, 2025, following recognition of an above-background case frequency. All positive blood cultures were communicated to the infection control lead within 6 hours of laboratory reporting. Twice-weekly interdisciplinary review meetings were established, comprising representatives from infectious diseases, pediatric oncology, clinical pharmacy, and nursing leadership.

2.5.2. Individual Single-Room Isolation and Geographic Ward Sectioning

Given the profound immunocompromise of the patient population, in whom shared-room cohorting would itself constitute an unacceptable cross-infection risk, each confirmed case was placed in strict individual single-room isolation. Cases were housed within a geographically defined section comprising approximately half of the pediatric oncology ward, physically separated from the unaffected area by closed doors maintained throughout the cluster period.

2.5.3. Contact Precautions

Contact precautions were implemented beyond standard hospital precautions and included: mandatory donning of gloves and fluid-resistant gowns before any patient contact; dedicated equipment per patient (thermometers, blood pressure cuffs, and stethoscopes); twice-daily disinfection of high-touch surfaces with sodium hypochlorite 1,000 ppm; and a dedicated nursing assignment at a minimum 1:2 ratio, with cross-assignment between isolation and general ward patients within a single shift prohibited.

2.5.4. Hand Hygiene Promotion and Monitoring

The infection control team conducted regular hand hygiene monitoring rounds throughout the cluster period, comprising direct observation of hand hygiene practices and on-the-spot educational reinforcement for clinical and support staff. Observations focused on patient contact areas within the affected ward section, and findings were communicated verbally to ward leadership and nursing staff on the same day. No formal quantitative compliance data were prospectively recorded; hand hygiene monitoring is therefore described as a qualitative IPC reinforcement activity rather than a structured audit with reportable compliance rates.

2.5.5. Visitor Restrictions

Visitor access was restricted to one designated primary caregiver per patient; sibling and extended-family visits were suspended for the duration of the active cluster. All caregivers received instruction on hand hygiene technique and contact precautions at admission and at each nursing shift change.

2.5.6. Antifungal Prophylaxis

Antifungal prophylaxis was introduced on June 15, 2025. In response to the active cluster, institutional eligibility criteria were expanded beyond standard practice to encompass all patients meeting any of the following high-risk criteria: (1) all AML patients (any line of therapy); (2) relapsed AML; (3) ALL during induction or re-induction with ANC ≤500 cells/µL; (4) ALL currently receiving or post re-intensification or FLAG-based induction/consolidation; and (5) relapsed ALL. First-line prophylaxis was voriconazole administered per weight-based institutional dosing; micafungin 1 mg/kg/day intravenously (maximum 50 mg/day) was used when voriconazole was contraindicated. Prophylaxis was continued for the duration of severe neutropenia (ANC ≤500 cells/µL) and for a minimum of two weeks following neutrophil recovery. Following identification of azole-resistant C. tropicalis in two patients receiving voriconazole prophylaxis (Cases 4 and 6), micafungin was substituted as the preferred prophylaxis agent for all new initiations from July 1, 2025, irrespective of voriconazole contraindications.

2.5.7. Temporary Deferral of Myelosuppressive Chemotherapy

During the period of maximum concurrent case burden (operationally defined as ≥3 simultaneous active cases), initiation of new highly myelosuppressive chemotherapy cycles — defined as regimens expected to produce an absolute neutrophil count (ANC) nadir below 100 cells/µL for more than 7 consecutive days — was deferred pending joint oncology–infectious disease review. All deferral decisions were individually documented, with family counseling recorded in the medical record.

2.5.8. Environmental Investigation

Environmental sampling of ward surfaces, sink drains, and water outlet points was performed using contact plates and air samplers. A plumbing infrastructure failure with an associated water leak was identified in late August 2025 during the cluster investigation. Hospital water systems are a recognized source of nosocomial fungal contamination; biofilm accumulation in hospital plumbing can simultaneously harbor multiple Candida species, producing mixed-species clusters without clonal transmission [9]. The plumbing failure is accordingly described as a putative environmental source — temporally associated with the cluster and biologically plausible — but not microbiologically confirmed, as environmental cultures from affected surfaces were not obtained prior to structural repair. Remediation was completed in September 2025.

2.6. Antifungal Therapy

All patients received therapy in accordance with IDSA 2016 candidemia management guidelines [11] and the MASCC/ESCMID guidelines for fever and neutropenia management in children with cancer [17]. Empirical antifungal treatment was initiated within 24 hours of clinical suspicion of candidemia, before species identification results were available. Micafungin was dosed at 2 mg/kg/day IV (maximum 100 mg/day); caspofungin at a loading dose of 70 mg/m² followed by 50 mg/m²/day; and liposomal amphotericin B (L-AmB) at 3–5 mg/kg/day IV. Step-down or combination therapy was guided by species identification and susceptibility results, typically available by day 3–5 of treatment. CVC management followed current institutional and guideline-recommended protocols. Minimum treatment duration was 14 days following documented blood culture negativity and clinical resolution.

2.7. Ethics

This study was approved by the KHCC Institutional Review Board.

3. Results

3.1. Culture-Confirmed Candidemia Cases

Seven pediatric inpatients with culture-confirmed candidemia were identified between June 1 and September 23, 2025 (Table 1, Cases 1–7). The cohort comprised 6 males and 1 female, with a median age of 12 years (range 1–17 years). All 7 patients had a CVC in situ and had received broad-spectrum antibiotics within the preceding 2 weeks. Leukemia was the underlying diagnosis in 71% (5/7), comprising AML (n=3) and ALL (n=2). Four of 7 patients (57%) required PICU admission.

3.2. Clinically Suspected IFI Cases

Five additional patients presenting during the same period met criteria for clinically suspected IFI based on compatible host risk factors and clinical or radiologic findings, but had negative blood cultures throughout their clinical course (Table 1, Cases 8–12). These patients are described for clinical completeness but are excluded from all candidemia-specific microbiological and outcome analyses.

3.3. Microbiological Findings

Candida tropicalis was the predominant species among confirmed cases, isolated in 5 of 7 patients (71%). C. albicans and C. krusei were each identified in 1 patient (14% each). Azole resistance was identified in 40% of tested isolates: voriconazole resistance in 2 C. tropicalis isolates, and intrinsic fluconazole resistance in the 1 C. krusei isolate. Two of the 3 patients receiving voriconazole prophylaxis at candidemia onset harbored azole-resistant organisms (67%), representing prophylaxis breakthrough. No echinocandin resistance was detected. Full resistance profiles are presented in Table 2.

3.4. Cluster Timeline and Environmental Context

The cluster spanned June 1 to September 23, 2025. One confirmed case was identified in June (first positive blood culture June 16); no new confirmed cases occurred throughout July; and the remaining 6 cases presented between August 8 and September 23, with case frequency accelerating in August–September, coinciding with identification of the plumbing water leak in late August. Against a background rate of 1 confirmed case in the preceding 6 months, 7 confirmed cases over a 4-month period represent a 7-fold increase above background incidence.

Environmental investigation identified a plumbing infrastructure failure with an associated water leak in late August 2025. Structural repair was completed in September 2025. No new confirmed candidemia cases were identified after September 23, 2025, during the surveillance period. Environmental microbiological cultures from the affected surfaces were not obtained prior to remediation, which precludes microbiological confirmation of the water system as the source.

3.5. Treatment and Tolerability

All 7 confirmed candidemia patients received guideline-concordant antifungal therapy. L-AmB was the initial empirical agent in 4 cases (Cases 2, 3, 4, and 7); an echinocandin was used in 3 cases (Cases 1, 5, and 6). Median treatment duration was 21 days (range 21–28 days). CVC removal was achieved in all patients within 48 hours of candidemia diagnosis. Antifungal tolerability was acceptable; transient hypokalemia requiring supplementation occurred in 3 patients receiving amphotericin-based therapy, resolving without permanent renal impairment.

3.6. Outcomes

Six of 7 confirmed candidemia patients completed antifungal therapy and were discharged. One patient (Case 1 — Non-Langerhans Cell Histiocytosis, C. kefyr) died from septic shock attributable to candidemia, yielding a candidemia-attributable mortality of 14% (1/7). A second patient (Case 2 — relapsed T-cell ALL) died from disease progression unrelated to the fungal infection; 30-day all-cause mortality among confirmed cases was 29% (2/7). Among the 5 clinically suspected IFI patients, 2 deaths were adjudicated to underlying malignancy.

4. Discussion

Between June and September 2025, 7 culture-confirmed candidemia cases occurred at KHCC over 4 months, against a background of 1 confirmed case in the preceding 6 months — a 7-fold incidence increase. The investigation identified a concurrent plumbing infrastructure failure as the putative environmental source. C. tropicalis was the predominant species (71%) with 40% azole resistance, consistent with regional epidemiological patterns. A structured IPC bundle was implemented, and no new confirmed cases were identified after September 23, 2025.

4.1. Case Definition Approach

The primary analysis was deliberately restricted to 7 culture-confirmed candidemia cases, distinguishing these from 5 patients with clinically suspected IFI but negative blood cultures. This approach reflects established microbiological standards: blood culture positivity remains the accepted criterion for candidemia diagnosis, and clinical or radiologic findings alone cannot confirm Candida bloodstream infection in the absence of microbiological evidence [11]. Although this conservative stance may underestimate the true cluster magnitude — blood culture sensitivity is estimated at 63–83% for pure candidemia and as low as 21–71% when deep-seated disease predominates[12] — it ensures that all analyses and outcome data rest on verified microbiological evidence.

4.2. Cluster Epidemiology and Environmental Context

The 7-fold increase above the 6-month background rate satisfies WHO criteria for a temporal cluster warranting investigation [10]. The heterogeneous species distribution — C. tropicalis (71%), C. albicans (14%), and C. krusei (14%) — is consistent with an environmental rather than a clonal source of acquisition, as hospital water system biofilms can simultaneously harbor multiple Candida species and produce mixed-species clusters without clonal transmission [9]. The identification of a plumbing infrastructure failure during the cluster investigation provides a temporally plausible putative source. However, the absence of environmental cultures prior to remediation precludes microbiological confirmation, and the cessation of new cases following structural repair represents a temporal association rather than proof of causality.

4.3. Species Distribution and Antifungal Resistance

The predominance of C. tropicalis and a 40% azole resistance rate are consistent with published surveillance data from LMIC pediatric oncology settings [3,4]. Azole resistance in C. tropicalis characteristically involves ERG11 mutations conferring cross-resistance between fluconazole and voriconazole [5,6]. The observation that two of the three patients receiving voriconazole prophylaxis at candidemia onset harbored azole-resistant organisms — a prophylaxis-breakthrough rate of 67% — underscores the critical importance of local resistance surveillance in guiding prophylaxis selection. Furthermore, C. tropicalis demonstrates an increased propensity for disseminated disease in neutropenic hosts compared with other species,[14] which has direct clinical implications for this predominantly leukemic cohort.

4.4. IPC Response

The IPC bundle comprised individual single-room isolation within a geographically sectioned ward, enhanced contact precautions, hand hygiene monitoring with direct educational reinforcement, expanded antifungal prophylaxis, and chemotherapy deferral during the period of maximum case burden. The decision to use individual room isolation rather than conventional cohort bay grouping directly reflected the immunocompromised status of the patient population, in whom shared-room placement would expose equally vulnerable patients to unnecessary cross-infection risk. The last confirmed case was recorded on September 23, 2025, coinciding with completion of environmental remediation; no new confirmed cases were identified thereafter. Given the retrospective case series design without a concurrent control group, this temporal association cannot be attributed to any specific intervention.

4.5. Clinical Outcomes

Candidemia-attributable mortality was 14% (1/7), and overall 30-day mortality was 29% (2/7), with the second death attributable to underlying malignancy rather than invasive fungal disease. These outcomes compare favorably with published mortality rates in analogous pediatric oncology populations: 36.5% at 28 days in Vietnamese pediatric leukemia cohorts [13], 36% at 30 days in an Indian pediatric cancer series [3], 20% in a Brazilian pediatric tertiary hospital cohort [16], and up to 55% in Turkish hematology-oncology series [4].

4.6. Limitations

This study has several important limitations. First, the retrospective design meant that risk factor data were extracted from clinical documentation rather than prospective research records; CVC insertion dates and exact neutropenia duration were not uniformly captured, precluding formal risk factor analysis. Second, isolate storage and molecular typing — whether whole-genome sequencing (WGS) or multi-locus sequence typing (MLST) — were not available at KHCC, preventing determination of clonal versus polyclonal acquisition; this represents the most significant analytical gap in the cluster investigation. Third, environmental cultures from the water leak site were not collected before structural repair, preventing microbiological confirmation of the proposed environmental source. Fourth, blood culture sensitivity limitations — estimated at 63–83% for pure candidemia and 21–71% for deep-seated disease — mean the true cluster magnitude may be underestimated, particularly given the high rate of hepatosplenic involvement observed on imaging.

5. Conclusions

We describe a temporal candidemia cluster at a Jordanian pediatric oncology center, defined by a 7-fold increase in culture-confirmed candidemia above background incidence and coinciding with a novel environmental event. C. tropicalis accounted for 5 of 7 confirmed cases; 40% azole resistance and two prophylaxis-breakthrough infections in patients receiving voriconazole prompted real-time substitution of micafungin for all new prophylaxis initiations from July 1, 2025, illustrating active antifungal stewardship in a resource-limited setting. A structured IPC bundle incorporating individual single-room isolation, geographic ward sectioning, enhanced contact precautions, and expanded antifungal prophylaxis was fully implemented with defined operational criteria. Candidemia-attributable mortality among confirmed cases was 14%. These data contribute to the limited published literature on candidemia clusters in Middle Eastern pediatric oncology centers and demonstrate that structured IPC responses are operationally achievable without advanced microbiological or genomic infrastructure.: