The goal of this clinical trial is to evaluate whether stopping antibiotic treatment early is safe in paediatric patients with cancer who develop high-risk febrile neutropenia but show good clinical evolution and low biomarker levels 48-72 hours after the episode.

The main questions it aims to answer are:

Is early discontinuation of antibiotics as safe as the standard strategy in terms of preventing invasive bacterial infections (such as sepsis, microbiologically documented infection, ICU admission, or death)? Does this strategy reduce the number of days on antibiotics without increasing infection-related complications?

Researchers will compare early antibiotic discontinuation with the standard care strategy to see whether the early-stop approach provides similar safety while reducing antibiotic exposure.

Participants will:

Receive standard initial antibiotic therapy for febrile neutropenia. Undergo clinical and biomarker evaluations (including CRP and PCT).

Be randomly assigned to:

Experimental group: early discontinuation of antibiotics, or Control group: continuation of the standard antibiotic strategy.

Be followed for 28 days after randomisation to monitor safety outcomes and treatment effects.

Children with cancer who develop febrile neutropenia (FN) routinely receive broad-spectrum intravenous antibiotics to prevent severe bacterial infections. Although this approach is essential in high-risk cases, prolonged antibiotic exposure increases the risk of adverse effects, antimicrobial resistance, drug toxicity, and longer hospital stays. Advances in risk stratification-particularly through clinical stability assessment and biomarkers such as C-reactive protein (CRP) and procalcitonin (PCT)-now allow earlier identification of patients who may no longer require intensive antibiotic coverage.

This study evaluates whether early discontinuation of antibiotic therapy is safe in paediatric patients with cancer who present high-risk FN but show a favourable clinical course 48-72 hours after onset. Eligible patients must meet predefined criteria indicating low risk of invasive bacterial infection (IBI), including clinical stability, absence of microbiologically documented infection, and decreasing biomarker levels. The trial uses a randomised, controlled, open-label design to compare an early-stop strategy with the current standard management.

The rationale for this approach is to determine whether antibiotic therapy can be safely shortened without compromising infection-related outcomes. Limiting unnecessary antibiotic exposure may reduce adverse drug events, improve patient comfort, minimise disruption of the microbiome, and support antimicrobial stewardship programmes. The study also incorporates biomarker measurements (CRP, PCT, IL-8) to explore their predictive value for IBI and to validate an IBI-risk prediction model for this population.

Participants will be monitored closely for signs of bacterial infection or clinical deterioration throughout the 28-day follow-up period. This design enables robust assessment of safety while reflecting real-world clinical decision-making and the potential benefits of personalised antibiotic duration in paediatric oncology care.

Inclusion Criteria:

1. Male and female patients ≤18 years of age expected to develop prolonged neutropenia (>7 days), with:

   • Acute myeloblastic leukaemia at any phase of chemotherapy
   • Acute lymphoblastic leukaemia in induction, consolidation, or intensification phases
   • Biphenotypic leukaemia at any phase of chemotherapy
   • Lymphoblastic lymphoma in induction and consolidation phases
   • B-cell and anaplastic lymphoma receiving high-intensity chemotherapy
   • Solid tumours receiving high-intensity chemotherapy
   • Relapsed leukaemia at any phase of treatment

2. Episode of febrile neutropenia (FN), defined as a single axillary temperature ≥38.0°C in a patient with an absolute neutrophil count (ANC) <500 neutrophils/mm³, or expected to fall below this value within the next 48-72 hours.

3. Antibiotic treatment initiated for the current FN episode (routine antimicrobial prophylaxis is allowed, as well as teicoplanin 3 days/week for patients with AML included in the CHIP-AML-2022 protocol and therefore in the Pro-teico study).

4. Low risk of invasive bacterial infection (IBI) at the start of the FN episode. Patients must meet all of the following:

   • CRP <9 mg/dL
   • PCT <0.5 ng/mL
   • Absence of hypotension

5. No microbiologically documented bacterial infection 48-72 hours after the FN episode.

6. Good clinical evolution 48-72 hours after the FN episode, defined as:

   • Afebrile for >48 hours (axillary temperature <38°C)
   • Haemodynamically stable
   • Stable paediatric early warning score (PEWS)

7. CRP <5 mg/dL, or CRP <9 mg/dL and PCT <0.5 ng/mL, with decreasing trend at the time of randomisation (values will be assessed on day 3 and day 5 after the FN episode).

8. ANC <500 neutrophils/mm³ at the time of randomisation.

9. Signed informed consent from the patient and/or parent(s)/legal representative(s).

10. Patient and/or parent(s)/legal representative(s) must have sufficient reading and writing skills to understand and provide consent to participate in the study.

11. Patient and/or parent(s)/legal representative(s) must be considered reliable and capable of adhering to the protocol.

Exclusion Criteria:

1. Antibiotic treatment at the time of the FN episode different from that used prophylactically.
2. Empirical antibiotic treatment different from that recommended in international guidelines.
3. Patient with poor clinical evolution during the first 12 hours (hemodynamic instability, PICU admission, death).
4. Active participation in the same study at the onset of the current FN episode.
5. Active participation in another clinical trial that, in the investigators' opinion, may interfere with the assessment of the results.
6. Any condition which, in the investigator's opinion, makes study participation unsuitable for the patient or could limit, prevent, or confound the assessments planned in the protocol.
7. Female patients who are pregnant or breastfeeding