Peripheral Stem Cell Mobilization with Chemotherapy + G-CSF Priming in a Large Series of Patients with Lymphoma Candidates to Autologous Transplantation. An Analysis of Baseline and Ongoing Predictors of Failure or Suboptimal Stem Cell Collection

Abstract 2149 Poster Board II-126 Background Autologous stem cell transplantation (ASCT) is a potentially curative treatment for lymphoma. Adequate stem cell (SC) collection is possible in the majority of cases but poor mobilization remains a major issue in some patients (pts). New agents were recently developed which may improve the rate of SC mobilization. Aims A retrospective analysis was conducted on a large series of lymphoma pts candidates to ASCT in order to identify factors influencing SC mobilization outcome. Potential early markers of poor mobilization were also evaluated. Patients and methods A total of 415 attempts of PBSC collection consecutively performed at 7 Italian centres in 388 pts affected by lymphoma were analyzed. Their median age was 52. A collection of less than 2×106 CD34+ cells/kg was defined as “mobilization failure” and of more than 5×106 CD34+ cells/kg as “good mobilization”. The following parameters were analysed for correlation with mobilization outcome: lymphoma diagnosis, disease status at mobilization, type of mobilizing chemotherapy, bone marrow infiltration at collection, n° of previous lines of therapy, prior use of fludarabine, alkylating agents or radiotherapy. The ratio between circulating CD34+cells/nL and total WBC/μL on the first day of CD34+ count (SCratio) was also analysed, trying to predict mobilization failure. Both univariate and multivariate statistical analyses were performed with SPSS package 13.0. Results Lymphoma diagnosis was diffuse large B cell/Burkitt in 38%, follicular in 10%, mantle in 13%, Hodgkin in 26%, T or NK/T in 6%, and other in 7% of cases. Disease status at apheresis was CR in 25%, chemosensitive in 49% and refractory in 26%. Mobilization was attempted during first-line therapy in 14%, in 37% during second-line and in 26% at third- or subsequent lines of therapy. Fludarabine, alkylating agents and extended-fields radiotherapy had been used prior to mobilization in 5%, 73% and 13% of cases respectively. Marrow infiltration by lymphoma was present in 17,7% and 7,7% of patients had failed a prior mobilization attempt. In 99% of cases (411/415) mobilizing therapy included chemotherapy and G-CSF. Two pts received AMD3100 and 2 G-CSF. Several chemotherapy programs were used for mobilization. For analysis purposes they were grouped as follows: single-agent cyclophosphamide (CTX) in 62 (15,1%); high-dose cytarabine, either single agent or in combination (HD-ARAC) in 143 (34,8%), etoposide-containing regimens in 119 (30,0%), and platinum containing regimens in 108 cases (26,3%). Mobilization failure occurred in 14% (59/415) whereas a good mobilization was obtained in 72% of cases. At univariate analysis mobilization failure was significantly associated with the use of single-agent CTX as mobilizer (p=0.000), with the n° of prior lines of treatment (p=0.006) and with failure of a previous mobilization attempt (p=0.000) whereas a diagnosis of follicular lymphoma (FL) was associated with a reduced risk of mobilization failure (p=0.007). Multivariate analysis confirmed FL as protective factor for mobilization failure (p=0.008) and single agent CTX and a higher n° of prior treatment lines as negative predictive factors (p=0.000 and 0.033 respectively). A “good mobilization” was significantly predicted at univariate analysis by HD-ARAC use as mobilizer (p=0,004) whereas single-agent CTX (p=0.000), n° of prior lines of therapy (p=0.005), prior fludarabine (p=0.035) or alkylating agents (p=0.003) were adverse predictors. Multivariate analysis confirmed HD-ARAC as independent predictive factor for good mobilization (p=0.031) and single-agent CTX and n° of prior lines of therapy as independent negative predictive factors (p=0.000 and 0.003 respectively). A SCratio of less than 10 on the first day of CD34+ count was associated with an increased risk of poor mobilization (p=0.000 risk estimate 0.638). Conclusions In this large series of pts with lymphoma the standard policy of SC mobilization was the use of chemotherapy + G-CSF. Nevertheless collection failed in a significant proportion of cases (14 %) and only 72% of cases reached an optimal SC collection. Predictive factors for mobilization outcome included disease and treatment history, type of lymphoma and type of mobilizing chemotherapy. These data provide the framework to rationally explore the use of new mobilizing agents in patients with lymphoma undergoing SC mobilization with chemotherapy and G-CSF. Disclosures: No relevant conflicts of interest to declare.