Less is more: Improving outcomes in haemato-oncology with optimal antibiotic use for febrile neutropenia Free

Background:

Fluoroquinolone (FQ) prophylaxis for febrile neutropenia (FN) is supported by international guidelines and over half of surveyed transplant centres reported using prophylaxis in 2020. Even with this approach, rates of FN are high and empirical antibiotic (Abx) therapy for FN is universal. However, FQ prophylaxis leads to emergence of multidrug-resistant organisms and reduces blood culture (BC) sensitivity. This study reports findings from a haemato-oncology unit where all inpatient FQ prophylaxis was prohibited.

Study Design: A prospective, observational study enrolled 86 consecutive patients from June to August 2024.

Objectives: 1) Investigate Abx usage; 2) Abx sequencing; 3) Compliance with guidelines; 4) Microbiological and clinical outcomes.

Patient population:

Patients with acute myeloid or lymphoblastic leukemia (AML/ALL), and myelodysplastic syndrome (MDS) undergoing intensive chemotherapy and patients undergoing a haematopoietic stem cell transplant (HSCT).

Management of FN: Rapid administration of piperacillin-tazobactam and amikacin (Piptaz/Ami) for fever (≥38°C) with neutropenia (≤0.5x10⁹/L or expected neutropenia) after taking BCs. If no gram-negative organisms are detected, amikacin is stopped after 48 hours. Abx can be stopped after 5–7 days in stable patients. Abx changes are guided by culture and susceptibility data; empirical changes are discouraged. Vancomycin is reserved for resistant gram-positive infections, MRSA colonisation, or line/soft tissue infections. In deteriorating patients, escalation to meropenem is advised.

Data Collection: Demographics; microbiological results; antibiotic therapy; admission to intensive care (ICU) and inpatient mortality.

Results: Patient population: 49% had AML/MDS; 66% had active disease (presentation or relapse). Treatment modalities: intensive chemotherapy (37%) autologous (37%) or allogeneic HSCT (13%). Median age was 57 years (23-85).

Abx Treatment for FN: Empirical Abx therapy for FN was required in 83/86 (97%) patients. No patients received FQ prophylaxis. 63/83 (76%) received Piptaz/Ami; 6 (7%) Piptaz monotherapy (due to renal impairment); 5 (6%) oral co-amoxiclav (not neutropenic); 8 (10%) meropenem or alternative regimens (due to penicillin allergy).

Abx sequencing based on initial empirical treatment: Among those started on Piptaz/Ami, 26 (41%) had no Abx changes, 24 (38%) changed to meropenem (of these, 9 also received vancomycin), 10 (16%) switched to other regimens, 3 received vancomycin alone. Among patients who received oral co-amoxiclav, 3 empirically changed to Piptaz/Ami and 1 to meropenem. 6 patients on Piptaz alone required no further changes; 2/6 on initial meropenem received additional abx.

Total days of antibiotics: mean duration in the whole cohort was 13 days

- AML and ALL chemotherapy,17 days and 12 days, respectively; allogeneic HSCT, 20 days

Microbiology: BCs positive in 35/83 (42%), with 49 isolates: 22 Gram-negative (e.g. E. coli, K. pneumoniae, Ps. aeruginosa, E. hormaechei), 27 Gram-positive (including coagulase-negative staphylococci [CoNS]). Five sputum and 3 urine cultures contributed additional isolates.

Outcomes: 5 (6%) inpatients died; 3 from infection, 1 intracerebral hemorrhage, 1 cardiac arrest (unclear cause). Two patients had relapsed AML, 2 newly diagnosed AML (on Venetoclax/Azacytidine and one on palliative Hydroxycarbamide), and 1 post-haploidentical HSCT for lymphoma. ICU admission occurred in 11 patients (13%).

Discussion: FN management is fundamental to patient outcomes in Haemato-Oncology. This small study suggests that less maybe more, with good outcomes using a strategy based on: no prophylactic Abx, rapid administration of empirical Abx; Abx changes guided by microbiological data; and avoidance of escalation in stable patients. Compliance with guidelines was high with no FQ prophylaxis given. Although 42% of patients with FN had positive blood cultures, this did not translate into poor outcomes. In a mixed population of consecutive patients - half receiving AML chemotherapy or allogeneic HSCT, and two thirds with active disease - the mortality rate was low (6%) and admissions to ICU were as expected (13%). Total days of Abx use, and overall use of meropenem and vancomycin, were low compared to published data. Delivering and maintaining this approach is challenging and requires close working between the antimicrobial stewardship and haemato-oncology teams.