Lymphopenia Prior High-Dose Chemotherapy Is a Major Prognostic Factor for Severe Infectious Complication after Autologous Blood Stem Cell Transplantation for Lymphoma's Patients.

Abstract

Rationale: Host factors able to predict accurately the infectious risk after high dose chemotherapy (HDCT) and autologous stem cell transplantation (ASCT) would be very useful. We previously reported that lymphopenia immediately prior conventional chemotherapy (day 1) identifies a subgroup of patients at risk of early death (Ray-Coquard et al. Br J Cancer 2001). This model was then investigated for patients receiving HDCT and ASCT.Patients and Methods: Between 1993 and 2000, 146 patients with diffuse large cell lymphoma (69), burkitt lymphoma (2), T lymphoblastic lymphoma (5), mantle cell lymphoma (16), follicular lymphoma (27), MALT lymphoma (2), lymphocytic lymphoma and chronic lymphoid leukemia (7) and hodgkin lymphoma (18) received HDCT with (72) or without (74) Total Body Irradiation (TBI). Median age at time of HDCT was 50 years (range, 38–68). Patients were treated in first (35.6%), second (48.4%) and third or more (16%) line treatment. Complete and unconfirmed complete remission before HDCT was obtained in 45.2%, partial response in 47.9%, minor response in 2.1% and 4.8% of patients were treated in progressive disease. Median transplanted CD34+ cell number was 6.79 × 106/kg (range, 0.70–47.58). 25 patients received G-CSF injection after graft. The predictive value of pre HDCT (d1) lymphocyte count for the risk of infection during the first 100-days after ASCT was investigated on this series.Results: Median d1 lymphocyte count was 750/mm3 (range, 90–3900). Infectious events were 35 clinical severe infectious complications (20 pneumonia, 8 severe enterocolitis, 3 viral infectious, 2 invasive fungal infection, 1 soft tissue infection, 1 severe septicaemia), 8 (6%) intensive care unit transfert and 4 (3%) toxic deaths for infectious reason. Using logistic regression, patients with ≤ 1000 d1 lymphocyte count presented statistically more infectious events (39 vs. 8 p<0.01). 62 patients of this series (69%) had ≤ 1000 d1 lymphocyte count. Age at HDCT, disease type, tumour response before HDCT, used of TBI, performance status before HDCT, number of transplanted CD34+ cells were not predictive of risk of infection. During the hospitalization after graft, used of G-CSF, number of days of broad spectrum antibiotics and anti fungal treatment, documented bacterial infection, number of days of hospitalization, number of transfusion, duration of neutropenia under 500/μl and plaquets recovery were not statistically different between patients with ≤ or > 1000 d1 lymphocyte count.Conclusion: Pre HDCT (d1) lymphocyte count is a potent predictor of the risk of severe infectious complication. These data need to be confirmed in a validation series. As 69% of patients had ≤ 1000 d1 lymphocytes, we incline to analyse lymphocyte subset to provide more discriminative information about the risk of infectious complication after HDCT. This currently being investigated.