Pre-neutropenic fever (PNF) (fever following chemotherapy but before developing low white cells) and neutropenic fever (NF) (fever in the setting of low white cells) are very common after chemotherapy for acute leukemia, bone marrow transplantation or Chimeric Antigen Receptor T-cell (CAR T) therapy. Often, there is no bacterial cause for fever found, and in the setting of a well patient with resolved fever, some studies have shown it to be safe to cease antibiotic therapy which was commenced at the onset of fever. This reduces the overall exposure to antibiotics, which can be beneficial to the patient (reduced risk of resistant bugs emerging, reduced serious side effects). However, some subgroups of high-risk patients have been underrepresented in these studies (in particular, those who have received a bone marrow transplant from a donor, those with longer duration of low white cells) and none have been performed in Australia, hence applying this data to our setting and patient groups is indirect and further data are needed. This study plans to recruit participants who have received chemotherapy for acute leukemia or a stem cell transplant (either their own cells or a donor's cells) or CAR T-cell therapy and perform a trial to compare early stopping of antibiotics (STOP arm) to the standard of care, which traditionally involves continuing antibiotics until the white cell count reaches above a specific threshold. The primary study outcome is duration of days free of antibiotics within 28 days of study allocation. The investigators will also observe for important clinical outcomes including rates of fever recurrence, bloodstream and other infections, intensive care admission and mortality. Patients will stay in hospital during this period, even in the setting of stopping antibiotics, and these antibiotics can be recommenced urgently according to the sepsis protocol if there is concern for infection.

This study is designed to assess the safety, benefits and impacts of early cessation of empiric antibiotics in all fever (both pre-neutropenic (PNF) and neutropenic (NF)) that develops post conditioning or chemotherapy until count recovery in high-risk hematology patients who meet clear inclusion criteria. This is in recognition of the fact that both PNF and NF are often not infective in nature, and that cessation is likely an important and safe approach in both scenarios. Secondly, the patient's pre-neutropenic and neutropenic status is highly fluid and can rapidly change from one to the other (within a day), making strict definitions of neutropenia arbitrary and not particularly useful for implementation in the clinical setting. Furthermore, as a programmatic-type intervention that is embedded in clinical workflow, approaching high-risk patients with fever in a standardized way would enable consistency and inform clear and concise management protocols. Stratification will allow for assessment of each patient sub-group to provide more granular data.

As an Australian first, this study will exploit the full potential of electronic medical record (EMR) systems, embedding all key aspects of the trial including screening, randomization and data collection into standard clinical and EMR workflows. This highly novel and innovative clinical trial methodology has the potential to improve trial efficiency, data quality and transferability between healthcare centers and will systematically evaluate the barriers and enablers of embedded trials (ELSA-EMR).

The study hypothesizes that early cessation of antibiotics in adult patients with high risk fever is safe, acceptable, cost-effective and will minimize an unnecessarily prolonged health care intervention

Inclusion Criteria:

Adult patients ( ≥18 years) who are receiving either:

• Conditioning chemotherapy for an autologous or allogeneic haematopoietic cell transplant or CAR T cell therapy, OR
• Induction remission chemotherapy for acute leukaemia,

AND develop fever ( ≥38degC) between time of initiation of chemotherapy/conditioning administration and ANC recovery to ≥500 cells/mm3 post the ANC nadir,

AND fever subsequently has settled (<38degC) for ≥48 and <96h hours.

[participants will be stratified into pre-neutropenic (ANC ≥500 cells/mm3) and neutropenic (ANC<500 cells/mm3) strata based on ANC level at 48 hours post fever onset, as per international consensus definition of neutropenic fever]

Exclusion Criteria:

• - Prolonged fever prior to defervescence (documented daily temperature ≥38.0°C for ≥ 5 days)
• Documented positive blood culture for bacteria since onset of fever episode and prior to randomisation
• Documented other infection (clinically or microbiologically defined) requiring antibacterial treatment
• Grade 2 or higher mucositis (WHO) or neutropenic enterocolitis
• Clinically unstable and/or admission to ICU at time of potential randomization
• Within 28 days of last randomization
• Prior randomization during current chemotherapy/conditioning cycle
• Pregnant or breastfeeding
• Currently being treated for CRS Grade 3 or 4, and/or ICANS Grade 3 or 4 (defined as per ASTCT Consensus Guidelines, Lee et al)