Background Gram-negative bacteremia (GNB) is a major cause of illness and death after hematopoietic stem cell transplantation (HSCT), and updated epidemiological investigation is advisable. Methods We prospectively evaluated the epidemiology of pre-engraftment GNB in 1118 allogeneic HSCTs (allo-HSCTs) and 1625 autologous HSCTs (auto-HSCTs) among 54 transplant centers during 2014 (SIGNB-GITMO-AMCLI study). Using logistic regression methods. we identified risk factors for GNB and evaluated the impact of GNB on the 4-month overall-survival after transplant. Results The cumulative incidence of pre-engraftment GNB was 17.3% in allo-HSCT and 9% in auto-HSCT. Escherichia coli, Klebsiella pneumoniae, and Pseudomonas aeruginosa were the most common isolates. By multivariate analysis, variables associated with GNB were a diagnosis of acute leukemia, a transplant from a HLA-mismatched donor and from cord blood, older age, and duration of severe neutropenia in allo-HSCT, and a diagnosis of lymphoma, older age, and no antibacterial prophylaxis in auto-HSCT. A pretransplant infection by a resistant pathogen was significantly associated with an increased risk of posttransplant infection by the same microorganism in allo-HSCT. Colonization by resistant gram-negative bacteria was significantly associated with an increased rate of infection by the same pathogen in both transplant procedures. GNB was independently associated with increased mortality at 4 months both in allo-HSCT (hazard ratio, 2.13; 95% confidence interval, 1.45–3.13; P <.001) and auto-HSCT (2.43; 1.22–4.84; P = .01). Conclusions Pre-engraftment GNB is an independent factor associated with increased mortality rate at 4 months after auto-HSCT and allo-HSCT. Previous infectious history and colonization monitoring represent major indicators of GNB. Clinical Trials registration NCT02088840.

Summary Specific trials on adult Burkitt lymphoma (BL) and ‘unclassifiable’ lymphomas with features intermediate between BL and diffuse large B‐cell lymphoma (BL/DLBCL) are advocated which include substantial numbers of older patients, to improve treatment feasibility, while countering risks of systemic and central nervous system (CNS) recurrences. We prospectively evaluated a modified CODOX‐M/IVAC (CODOX‐M: cyclophosphamide, vincristine, doxorubicin, high‐dose methotrexate; IVAC: ifosfamide, etoposide and high‐dose cytarabine) regimen by the addition of rituximab (R) and liposome‐encapsulated cytarabine (D) to increase antitumour activity and halve the number of intrathecal treatments. Thirty adults (40% >60 years) with BL (n = 15) and BL/DLBCL (n = 15) were accrued. Primary endpoints were progression‐free survival (PFS), CNS recurrence, and liposomal cytarabine‐associated toxicity. Eighty percent of patients received the whole treatment programme, the remaining cases received at least three full courses. Application of the RD‐CODOX‐M/IVAC regimen resulted in remarkable 4‐year PFS (78%) and complete remission (CR) rates (93%). However, PFS was significantly lower in patients older than 60 years as compared to younger ones (49%vs 93%, P = 0·03; median, 36 months), despite high actual dose‐intensity, CR rate and tolerability. Reduced‐intensity intratechal prophylaxis through liposomal cytarabine was effective because the CNS failure rate was low (3·4%) and without severe neurological toxicities. The RD‐CODOX‐M/IVAC strategy is feasible and highly effective, but improving outcomes in elderly patients remains a priority.

Both regimes showed good tolerability and appealing response rates. FFS was more prolonged in R-GEMOX, but patients continuously relapsed without a clear plateau on survival curves.

This Phase II study assessed feasibility and efficacy of a biweekly R-COMP-14 regimen (rituximab, cyclophosphamide, non-pegylated liposome-encapsulated doxorubicin, vincristine and prednisone) in untreated elderly patients with poor-risk diffuse large B-cell lymphoma (DLBCL) and moderate to high ‘life threat’ impact NIA/NCI cardiac comorbidity. A total of 208 courses were delivered, with close cardiac monitoring, to 41 patients (median age: 73 years, range: 62–82; 37% >75 years) at a median interval of 15·6 (range, 13–29) days; 67% completed all six scheduled courses. Response rate was 73%, with 68% complete responses (CR); 4-year disease-free survival (DFS) and time to treatment failure (TTF) were 72% and 49%, respectively. Failures were due to early death (n = 3), therapy discontinuations (no-response n = 2; toxicity n = 6), relapse (n = 6) and death in CR (n = 3). Incidence of cardiac grade 3–5 adverse events was 7/41 (17%; 95% confidence interval: 8–31%). Time to progression and overall survival at 4-years were 77% and 67%, respectively. The Age-adjusted Charlson Comorbidity Index (aaCCI) correlated with failures (P = 0·007) with patients scoring ≤7 having a longer TTF (66% vs. 29%; P = 0·009). R-COMP-14 is feasible and ensures a substantial DFS to poor-risk DLBCL patients who would have been denied anthracycline-based treatment due to cardiac morbidity. The aaCCI predicted both treatment discontinuation rate and TTF.