PB1672 REVIEW OF BLOODSTREAM INFECTION IN ADULT PATIENTS OF ACUTE LYMPHOBLASTIC LEUKEMIA TREATED WITH HYPERCVAD – A TERTIARY CENTRE EXPERIENCE

Abstract

Background:Aside from high relapse rate, the toxicity of HyperCVAD also renders the management of adult patients with acute lymphoblastic leukemia (ALL) challenging. Cycle A of HyperCVAD includes cyclophosphamide, vincristine and doxorubicin while cytarabine and methotrexate are the key components in cycle B.Aims:This study aims to evaluate the bloodstream infection (BSI) of HyperCVAD in ALL patients at Tuen Mun Hospital (TMH).Methods:In this retrospective single centre study from January 2013 to June 2017, 18 de novo ALL patients (7 Philadelphia‐positive, 7 Philadelphia‐negative and 4 T‐lineage) were treated with HyperCVAD +/‐ tyrosine kinase inhibitor (TKI). Anti‐bacterial prophylaxis was not routinely prescribed in our centre. Complete blood count was checked once every two days during chemotherapy and patients would receive granulocyte colony‐stimulating factor (G‐CSF) if there was grade 3 or above neutropenia. Septic workup would be performed when body temperature was ≥ 38 oC. Intravenous piperacillin‐tazobactam or levofloxacin in case of penicillin allergy would be started and intravenous vancomycin would be added in patients with history of methicillin‐resistant Staphylococcus aureus (MRSA) infection. Antibiotics would be adjusted in accordance with culture results and sensitivity profiles. BSI was defined as positive blood culture of a recognized pathogen from a single blood culture or of a common commensal from multiple blood cultures. The primary outcome is the measure of BSIs in incidence density, defined as the number of BSIs per patient‐day. Secondary outcomes include periods of neutropenia in each cycle, organism types and sensitivity profilesResults:The median age of this cohort was 50 years old (21‐67 years old) and 13 (72%) patients were of good past health. Among 88 admissions for HyperCVAD, 31 episodes of BSIs were documented in 16 patients. Three (16.7%) patients had more than one BSI in a single admission. The overall incidence density was 17.9 BSIs per 1,000 patient‐days. The incidence density of BSIs was significantly higher in induction phase (23.8 BSIs per 1,000 patient‐days) and cycle B (Median: 20.6 BSIs per 1,000 patient‐days) when compared with that of cycle A (Median: 5.9 BSIs per 1,000 patient‐days) (p = 0.008). Patients with BSIs were associated with significantly longer duration of neutropenia than those without BSIs (9 days vs. 5 days, p < 0.001). The addition of TKI in Philadelphia‐positive patients did not result in more BSI compared with Philadelphia‐negative or T‐lineage ALL patients (p = 0.294). Gram‐negative bacilli (GNB) accounted for 71% of BSIs and half of the GNB were Escherichia coli. 45.5% of E. coli were extended spectrum beta‐lactamase (ESBL) producing. The remaining 29% of BSIs were caused by gram‐positive cocci (GPC) with Streptococcus mitis as the most common GPC. Majority of cases (61%) did not have an identifiable source of infection. There were three deaths directly related to BSIs.Summary/Conclusion:Bloodstream infection is more common in induction phase and cycle B of HyperCVAD and gram negative bacilli are the main causative agents. Antibacterial prophylaxis can be considered but the emerging drug resistance imposes great challenges. More studies on the effectiveness of antibacterial prophylaxis in the era of multidrug resistant organisms are necessary.