Detailed Assessment of Blood Culture and Clinical Factors in Febrile Neutropenia Patients with High-Risk Hematological Malignancy in Japan: A Subgroup Analysis of the Cedmic Trial

Abstract

Background: During chemotherapy for hematological malignancies, febrile neutropenia (FN) is a frequent and serious complication. The causative organisms of FN vary from era to era and depending on bacterial resistance in each country. Therefore, epidemiological data are required for each era and country from prospective large clinical trials. We previously reported a prospective clinical trial indicating the importance of the D-index, a measure of neutropenia severity (CEDMIC trial, J Clin Oncol. 2020 10; 38: 815). We hypothesized that analyzing these data would provide accurate contemporary epidemiological data for hematological malignancies. Furthermore, as encountered in practice, we suspected that gram-negative (GN) bloodstream infections (BSIs) are more common at high body temperatures (BTs) than at low BTs and examined the relationship between causative bacteria and BT. Patients and Methods: Between June 2013 and April 2017, we enrolled patients with hematological malignancies aged 16-79 years post-chemotherapy or -stem cell transplantation (SCT) with expected neutropenia durations >7 days. Neutrophil counts <500/mL reflected neutropenia, and all fevers had axillary temperatures ≥37.5°C. Prophylactic antibiotics (Abx) were drugs administered post-chemotherapy and pre-FN onset. The FN incidence and blood culture (BC) positivity rate were examined in 4 treatment groups: allogeneic SCT (Allo-SCT), autologous SCT (Auto-SCT), induction chemotherapy for acute leukemia, and other treatments. Results: In total, 423 patients were enrolled, and 413 (median age: 56 years) were included in the intention-to-treat analyses (Table 1). The treatments included induction chemotherapy (n=67), Allo-SCT (n=106), Auto-SCT (n=158), and others (n=82). Prophylactic Abx were administered in 281 (68.5%) patients (only quinolone, n=222; with β-lactam, n=59). Prophylaxis rates were significantly higher in the Allo-SCT (n=94, 88.7%, p<0.001) and induction chemotherapy (n=49, 73.1%, p<0.017) groups. FN occurred in 356/413 patients (86.4%, Table 1). The Allo-SCT group had a higher FN incidence than that in the other treatments group (93.4% vs. 72%, p<0.001). D-indices ≥5500 and <5500 were associated with FN incidences of 100% and 81.8%, respectively (p<0.001). Prophylaxis did not reduce the FN incidence. In total, 1272 BC sets were collected from 356 patients with FN. At FN onset, BCs were collected from 346 patients; 72 (20.8%) were positive (Table 2). Among initial BCs, 46 (63.9%) harbored gram-positive (GP) and 28 (38.9%) harbored GN bacteria. Three patients had multiple bacterial infections. The 2 most frequent GP bacteria were coagulase-negative Staphylococcus (n=21, 44.7%) and viridans streptococci (n=14, 29.8%). The 3 most frequent GN bacteria were Escherichia coli (n=13, 44.9%), Klebsiella pneumoniae (n=4, 13.8%), and Pseudomonas aeruginosa (n=4, 13.8%). Among initial BCs, prophylactic Abx use did not reduce BC positivity. However, the GN BSI incidence was significantly lower in the prevention group vs. non-prevention group (23.8% vs. 56.7%, p=0.007). Among patients who received prophylactic β-lactam, GN BSIs did not occur. Maximum BTs (maxBTs) were significantly higher in patients with positive BCs than negative BCs (38.4°C vs. 38.0°C, p<0.001). The BTs associated with GN BSIs were significantly higher than those associated with negative BCs and GP BSIs (38.7°C vs. 38.0°C vs. 38.2°C, p<0.001, Figure 1). The maxBTs and BC positivity rates were strongly related (Table 3). The BC positivity rates were significantly higher in patients with BTs ≥38°C vs. <38°C (26.5% vs. 14.5%, p=0.008), 38.5°C vs. >38.5°C/<38.5°C (39.7% vs. 15.3%, p<0.001), and ≥39°C vs. <39°C (52.6% vs. 16.9%, p<0.001). Among patients who did not receive prophylactic Abx (n=106), the GN BSI positivity rate was significantly lower in patients with maxBTs <38°C vs. ≥38°C (14.3% vs. 76.2%, p=0.007). Conclusion: In FN patients with high-risk hematological malignancies in Japan, the frequency of GP BSIs is high. Prophylactic Abx use reduces the GN BSI risk but not the FN or BSI risk. Fever severity and BC positivity at FN onset are strongly associated. The BTs of patients with GN BSIs are high. Disclosures Kanda: Astellas Pharma: Honoraria, Research Funding; Sanofi: Honoraria, Research Funding; Meiji Seika Kaisha: Honoraria; Merck Sharp & Dohme: Honoraria; Pfizer: Honoraria, Research Funding; Shionogi: Research Funding; Celgene: Honoraria; Sumitomo Dainippon Pharma: Honoraria; Eisai: Honoraria, Research Funding; Novartis: Honoraria; Kyowa Kirin: Honoraria, Research Funding; Bristol-Myers Squibb: Honoraria; Takeda Pharmaceuticals: Honoraria; Alexion Pharmaceuticals: Honoraria; Shire: Honoraria; Daiichi Sankyo: Honoraria; Ono Pharmaceutical: Honoraria; Nippon Shinyaku: Honoraria, Research Funding; Mochida Pharmaceutical: Honoraria; Otsuka: Honoraria, Research Funding; Chugai Pharma: Honoraria, Research Funding; Janssen: Honoraria; Mundipharma: Honoraria. Kimura:Takeda Pharmaceuticals: Honoraria; Astellas Pharma: Honoraria; Pfizer: Honoraria; Sumitomo Dainippon Pharma: Honoraria; Merck Sharp & Dohme: Honoraria; Nippon Kayaku: Honoraria; Kyowa Hakko Kirin: Honoraria; Ono Pharmaceutical: Honoraria; Celgene: Honoraria; Asahi Kasei: Honoraria. Oyake:Astellas Pharma: Honoraria; Bayer: Research Funding; Chugai Pharma: Honoraria, Research Funding; Celgene: Honoraria; Kyowa Hakko Kirin: Honoraria, Research Funding; Sumitomo Dainippon Pharma: Honoraria; Bristol-Myers Squibb Japan: Honoraria; Pfizer: Honoraria; Sionogi: Research Funding. Tamura:Asahi Kasei: Honoraria; Ono Pharmaceutical: Honoraria; Eisai: Honoraria; Kyowa Hakko Kirin: Honoraria.