Impact of antibiotic use in early fever after haploidentical hematopoietic stem cell transplantation with post-transplant cyclophosphamide

Abstract Introduction: Early fever is a frequent complication following haploidentical hematopoietic stem cell transplantation (haplo-HSCT) with post-transplant cyclophosphamide (PTCY), most commonly associated with cytokine release syndrome (CRS). However, infectious etiologies must always be carefully excluded. The characterization and clinical impact of broad-spectrum antibiotic (BSA) use in this context remain poorly defined. This study investigates the incidence of early fever after haplo-HSCT, associated infectious events, and transplant-related outcomes. Patients and methods: A single-center, retrospective study was conducted including a consecutive cohort of 144 adult patients who underwent haplo-HSCT with PTCY between January 2019 and March 2024. All patients received PTCY 50 mg/kg on days +3 and +4, followed by tacrolimus and mycophenolate mofetil as graft-versus-host disease (GVHD) prophylaxis. Early fever was defined as a body temperature ≥37.8°C occurring between the stem cell infusion (day 0) and day +4. Twenty-three patients with fever and/or on antibiotics before stem cell infusion were excluded from the final analysis. Results: The median patient age was 53 years, and 60.3% were male. The most frequent diagnosis was AML (38.8%). Most patients had intermediate disease risk (65.3%) and a high age-adjusted HCT-CI (76%). Peripheral blood was used as the stem cell source in 99% of cases, and the most common relationship was a child as donor (48.8%). The incidence of early fever was 85% (103/121), with no significant differences by conditioning regimen. Fever most commonly began on day 1 post-infusion (median 1 day; range 0–4). Median peak temperature was 38.5 °C; median C-reactive protein (CRP) level reached 142.8 mg/L (range 1.1–301; normal <0.5). Median fever duration was 4 days (interquartile range [IQR] 3–5). Early fever met criteria for CRS in 92.2% of cases (95/103), mostly grade 1 (86/95, 90.5%). Eleven patients received tocilizumab for CRS grade ≥2. Levofloxacin prophylaxis was used in 72.8% of cases. Most (85/103, 82.5%) received empirical BSA, mainly piperacillin/tazobactam (67%). A documented microbiological isolate was identified as the primary cause of fever in 8 (7.8%), all catheter-related: three catheter-related bloodstream infections (S. haemolyticus, S. epidermidis, S. oralis) and five catheter colonizations (non-specified coagulase-negative Staphylococcus, methicillin-resistant, Bacilllus spp, S. epidermidis, S. maltophyilia). In three others, despite positive isolates, fever was attributed to concomitant CRS: one had CMV reactivation with low viral load, another had CRS grade ≥2 requiring tocilizumab and a positive stool culture for C. jejuni, and one had asymptomatic bacteriuria due to K. pneumoniae. No differences were observed in fever onset, peak, duration, or CRP levels between CRS and microbiological documented cases. In patients with a microbiological isolate, antibiotic therapy was adjusted accordingly. Subsequently, 86.4% (89/103) developed significant infections, and 15.5% (16/103) experienced new colonizations, all caused by multidrug-resistant organisms. Two patients developed infections caused by the same organism previously identified as colonization: one had catheter colonization, and the other, a catheter-related bloodstream infection. The remaining patients did not develop infections with these microorganisms. With a median follow-up of 22 months (IQR 11.2–47.4), overall survival (OS) and event-free survival (EFS) at 12 months for the whole cohort were 73.5% (95% CI, 64.7–80.5) and 61.1% (95% CI, 51.9–69.2) respectively. No statistically significant differences were observed between patients with and without early fever in terms of OS (p=0.082) or EFS (p=0.487). Early fever was not significantly associated with neutrophil or platelet engraftment, or with the incidence of acute or chronic GVHD. Conclusions: Early fever is a common complication after haplo-HSCT with PTCY and is associated with CRS in the majority of cases. Only a minority of episodes are attributed to infection. However, most patients received empirical BSA, which may have contributed to pathogen selection and subsequent infectious complications. Although in this cohort overall survival was comparable between patients with CRS-related fever and those with documented microbiological infections, our findings highlight the need for biomarkers and standardized criteria to guide the empirical use of BSA in this setting.