Study supports early antibiotic suspension in kids with cancer, febrile neutropenia

The E-STOP trial presented at ESHRE 2025 supports early antibiotic suspension in paediatric patients with cancer and febrile neutropenia (FN).

“In 2010, the IDSA* recommended antibiotic therapy maintenance until neutrophil recovery for cancer patients with FN even [in the absence of] a bacterial infection,” said Dr Natalia Mendoza-Palomar from the Children’s Hospital and Vall d’Hebron Women’s Hospital, Barcelona, Spain, at ESPID 2025.

Since then, studies have shed light on the safety of early antibiotic suspension, but most were retrospective or restricted to the adult population, noted Mendoza-Palomar.

“[Hence, we sought] to evaluate the safety of early antibiotic suspension in [a paediatric cancer cohort with no] demonstrable invasive bacterial infection (IBI), good clinical evolution, and low biomarker values 48-72h after fever onset in the experimental arm group compared with the classic strategy in the control group,” she said.

The control group had no adverse event (AE) attributable to BI, which includes death, PICU** admission, sepsis, microbiologically demonstrated BI, and pneumonia. Only one case of bacteraemia due to Klebsiella (K.) pneumoniae was reported in the experimental group. A comparison between arms yielded a p-value of 0.491, suggesting no significant difference.

According to Mendoza-Palomar, the bacteraemia case occurred 12 days after antibiotic suspension and coincided with a new chemotherapy cycle.

Secondary objectives

The number of severe neutropenia (ANC*** <500 neutrophils/mm³) days was also comparable between the experimental and control arms (8 vs 9 days), as were the incidence of fever recurrence (n=25 vs 26) and the number of fever days (2 days for both).

The experimental arm had fewer antibiotic days than the control arm (3 vs 4 days; p<0.001 [first course] and 4 vs 7 days; p=0.007 [successive courses]), but there were no AEs attributable to antibiotic use in both arms. [ESPID 2025, abstract OP087]

Using data on FN patients admitted in one facility (n=133), sensitivity of the risk stratification score by Santolaya, et al was 50 percent, and specificity 60.2 percent. [Clin Infect Dis 2002;35:678-683] IBI was detected in 10 patients (five low-risk and five high-risk).

“Validation of the risk stratification score in the targeted population is mandatory before its implementation. Unfortunately, [its] performance … was lower than expected,” Mendoza-Palomar noted. “More studies are needed to find a predictive rule that better suits our population.”

Considering the IBI cutoff value for interleukin(IL)-8 (200 pg/mL), the median biomarker levels were low (2.23 mg/dL, 0.18 ng/mL, and pg/mL for C-reactive protein, procalcitonin, and IL-8, respectively). Mendoza-Palomar and colleagues recommended further comparison of biomarkers with a high-risk cohort to potentially establish cutoff values.

Of the 63 patients who underwent T2 magnetic resonance, 56 were negative, five tested positive, while two were unavailable due to technical issues. The five positive cases were due to Escherichia (E.) coli (n=2), K. pneumoniae (n=1), Enterococcus faecalis (n=1), and E. coli and K. pneumoniae (n=1); of note, all had negative blood cultures and good clinical evolution.

Study characteristics

A total of 103 paediatric patients with cancer and FN from four hospitals in Spain were screened, underwent blood test plus IL-8, and randomized 1:1 to discontinue antibiotic use within <24h (experimental arm) or maintain antibiotics until neutrophil recovery (ANC ≥500 neutrophils/mm³; control arm). Fifty patients failed the screening, leaving 53 evaluable patients.

In the experimental arm, the median age was 4 years, and nearly 60 percent were boys. Similar patterns were observed in the control arm (median age 4.88 years, 51.9 percent boys). The most common underlying condition was acute lymphoblastic B leukaemia (62 percent), followed by solid tumours (17 percent), other leukaemias (12 percent), and lymphoma (9 percent).