Efficacy and safety of adjusting Antimicrobial Therapy in episodes of Fever and Neutropenia catalogued as Fever of unknown origin in children with Cancer: a randomized clinical trial

INTERVENTION: This is a randomized, controlled, two‐arm, double‐blind clinical trial. Children were evaluated at admission in order to a establish their risk for Invasive Bacterial Infection. After this initial evaluation, all children were treated according to the Latin American and international Consensus for a Rational Approach of children with Febrile Neutropenia. After 48 hours of hospitalization for episodes categorized as Low risk and 72 hours for High risk episodes were evaluated and clinical evolution were classified as favourable or unfavourable. Children with an unfavourable evolution were excluded from the study receiving standard of care in the network Hospitals. Children with episodes of Low risk and High risk with favourable evolution were subject to one of the three following interventions. (1)Viral etiology. Children positive for a Respiratory Viral Infections, negative for a bacterial pathogen with a favourable evolution at 48 ‐72 hours after admission (good clinical condition, temperature = 38.5ºC, without new clinical foci, with two consecutive levels of CRP < 90 mg/L) were subject to a 1:1 simply randomization by the blinded study coordinator using statistical software (GraphPad Prism, San Diego, Ca, USA) were randomized to two treatment groups: 1.a) Low risk episodes: One group received the current Standard Antimicrobial Treatment based on outpatient management after day 2 and oral Antimicrobial for 7 days. The second group based on Antimicrobial stewardship, withhold Antimicrobial treatment, under close observation as outpatients. 1.b) High risk episodes: One group received the current standard strategy based on hospitalized management with intravenous Antimicrobial until the end of the episode, at least for 7 days; the second group, will be treated withold antimicrobials, under close observation as hospitalized patients.(2)No etiology. Children with Low and High risk febrile neutropenia with no etiology, with a favourable evolution at 48 ‐72 hours after admission and good clinical condition, were subject to a 1:1 simply randomization by the blinded study coordinator using statistical software. 2.a) In Low risk episodes: The current management strategy with standard antimicrobial treatment based on inpatient or outpatient management after day 2 and intravenous/oral Antimicrobial for 7 days, and the second novel strategy group based on Antimicrobial stewardship, adjusting to a Narrow Antimicrobial therapy, under close observation as in or outpatients. 2.b) In High risk episodes: The current management strategy with Standard antimicrobial treatment, based on hospitalized management with intravenous Antimicrobial until the end of the episode, at least for 7 days and the Antimicrobial stewardship strategy, stopping vancomycin and meropenem (if previously indicated) and adjusting to a Narrow Antimicrobial Treatment, under observation in hospitalized patients. (3) bacterial etiology Children with High risk episodes with an Invasive Bacterial Infection, with a bacterial pathogen identified in blood or other sterile site, with a fav CONDITION: Febrile Neutropenia ; Neoplasms INCLUSION CRITERIA: Patients under eighteen years of age; of both genders; undergoing chemotherapy treatment for cancer; hospitalized for febrile neutropenia in any of the participating hospitals; were invited to participate and included in the study after informed consent from the parents and the child (or assent from the child, if older than 8 years of age) PRIMARY OUTCOME: Expected outcome 1: the primary outcome will be the number (%) of episodes with uneventul resolution. Outcome found 1: for viral episodes, the frequency of uneventful resolution was 106 (96%) in the group that maintained antimicrobial therapy and 114 (95%) in the adjusted group (p=1.00); relative risk 1.01 (95% CI 0.95–1.06); absolute risk difference 0.01 (95% CI ‐0.05–0.06). Children with episodes of febrile neutropenia and fever of unknown origin randomized to either maintain or adjust antimicrobial therapy showed the same percentage of uneventful resolution: 106 (96%) in the group that maintained antimicrobial therapy and 114 (95%) in the adjusted group (p=1.00); relative risk 1.01 (95% CI 0.95–1.06); absolute risk difference 0.01 (95% CI ‐0.05–0.06). Analysis of intervention group 3 is pending due to insufficient sample size. SECONDARY OUTCOME: Expected outcome 2: The expected secondary outcome is fewer days of fever (number), fewer days of hospitalization, (fewer days of AM therapy, fewer antimicrobials after randomization, fewer days of antimicrobial therapy (AM) therapy for vancomycin and meropenem per 1000 days of neutropenia, for the adjusted group. ; Outcome found 2: For the viral group the median days of antimicrobial therapy were significantly lower in the group that withdrew the antimicrobials (3 versus 5 days; p < 0.001). The days of fever were significantly lower in the withholding group (1 versus 2 days; p= 0.04). The days of hospitalization (5 versus 5), days with absolute neutrophil count (ANC) <500/mm3 (4 versus 4), and the percentage of patients with oxygen requirement (3% versus 2%) were similar between the groups. No cases of PICU admission, sepsis, mechanical ventilation requirement, or death were reported. In the fever unknown origin group: Children with adjusted antimicrobial therapy had less days of hospitalization, 7 days (5‐8) versus 5 days, interquartil range (IQR) (4‐8) (p<0.001) and less days of antimicrobial therapy, 6 (5‐8) versus 5 (4‐7) (p<0.001). The median number of antimicrobials after randomization was 2, IQR (1 ‐ 2) in the group that maintain antimicrobial (AM) therapy versus 1, IQR (1‐1) in the adjusted group, (p<0.001). The number of days of vancomycin and meropenem per 1,000 days of neutropenia showed a decreased in the vancomycin use in the adjusted group, from 93 to 59 days (p=0.04). The percentage of Pediatric intensive care unit (PICU) admission was 1% in the group that maintained AM therapy, and 2% in the adjusted group (p=1.00), there were no cases of sepsis nor deaths in either group. In the bacterial group: The most prevalent microorganisms were Enterobacter cloacae, Escherichia coli, and viridans group Streptococcus. Clinical course was similar in both groups, assessed by days of fever (p=0.13), days of hospitalization (p=0.24), development of sepsis (p=0.5), and admission to the ICU (p=0.69). There were no deaths in either group. The median number of antimicrobials used by patients decreased from three (2–3) to one (1–2) post‐randomization (p=0.002).