Epidemiology of Bacterial Infections during Induction Chemotherapy in Adult Patients with Acute Lymphoblastic Leukemia (ALL): Analysis of the Graall-2005 Study

Background: Bacterial infections (BI) are a major cause of morbidity and mortality in patients treated for hematological malignancies, especially those with acute myeloblastic leukemia or receiving allogeneic hematopoietic stem cell transplantation. Despite severe neutropenia and prolonged treatment with corticosteroids, there are little published data on BI during induction chemotherapy in adults with acute lymphoblastic leukemia (ALL). Between 2006 and 2014, 787 adult patients were included in the GRAALL-2005 study, a prospective, randomized and multicenter phase III trial for patients newly diagnosed patients with Philadelphia chromosome-negative B-cell precursor (BCP) or T-cell ALL. We retrospectively reviewed the occurrence of BI during induction treatment in these patients.

Patients and Methods: The GRAALL-2005 study evaluated the value of hyperfractionated cyclophosphamide in the whole study population and of rituximab in patients with CD20+ BCP-ALL. All patients received a 5-drug induction therapy with corticosteroids (prednisone) for 21 days, associated with vincristine, daunorubicin, cyclophosphamide and L-asparaginase. A broad-spectrum antibiotic treatment effective on Gram-negative and positive germs was recommended when the neutrophil count was less than 0.5 G/L. Pneumocystis prophylaxis was made by trimethoprim/sulfamethoxazole or pentamidine.

Results: During induction chemotherapy, 270 of the 787 patients (34.3%) experienced a total of 376 BI episodes (1.4 BI episodes per patient). The BI incidence rate was higher in the subgroup of patients combining hyperfractionated cyclophosphamide and rituximab as compared to those who received standard-dose cyclophosphamide and no rituximab (40.7% versus 29.5%; p=0.098). The median time from the first day of induction therapy to BI diagnosis was 10 days (range, 7-14). The infection was considered as serious in 58 patients, with a diagnosis of septic shocks in 57. Forty-one patients were transferred in intensive care unit. At 50 days after induction initiation, 22 patients had died from BI: 8.1% of patients with BI and 2.8% of all patients. Bloodstream was the most common site (82.7%), followed by gastrointestinal tract (6.5%) and lungs (6.5%). In less than 2% of cases, skin and soft tissues, central venous catheter, or urinary tract was concerned. Infections with Gram-positive cocci predominated as the etiology of microbiologically documented infections (46.9%), more specifically coagulase-negative Staphylococci. E. coli and Pseudomonas species were the most common Gram-negative organisms (40.5%). The patients received a median number of 3 antibiotics. The first-line was a monotherapy in 57% of cases, with the predominant use of betalactam. In one-third of the cases, it was betalactam in combination with aminoglycoside or glycopeptide. More than 2 antibiotics were prescribed in 12% of cases.

Conclusion: Induction chemotherapy in adults with ALL is associated with a high incidence of bacterial infections and a significant related mortality. To our knowledge, this report is the only large adult ALL study dealing with bacterial infectious complications during induction chemotherapy. Despite an intensely myelosuppressive chemotherapy regimen, the infection-related mortality seems to be lower than that reported during induction in acute myeloid leukemia. Predictive risk factors for bacterial infections have to be analyzed, as well as prophylactic/empirical antibiotic strategies in order to improve care for this subset of patients.