In Vivo Purge with Rituximab before Harvest Does Not Alter the Graft nor the Hematologic Recovery after Transplant.

Abstract

Currently the combination of rituximab with chemotherapy is the standard for treating patients with B-cell lymphoma, particularly when high-dose therapy (HDT) plus autotranplant is scheduled, because of the in vivo purge effect of this treatment. However, the effect of rituximab on the harvest of CD34+ cells, on the hematologic recovery after transplant, and on the risk of complication was never really assessed in large series of patients. We retrospectively analyzed our data on 198 patients who were scheduled for HDT after rituximab therapy because of high failure risk, partial response, or relapse and we compared these data to a series of 207 patients treated in our center with transplant but without rituximab before stem cell harvest (Ketterer Blood1998;91:3148, BM Transplant1999;23:1309, Br J Haematol1998;103:235). Stem cells were usually harvested after chemotherapy but in case of failure were harvested at baseline. This was the case in 25 patients (15%). 29 (15%) patients failed to mobilize; all but 1 in first or second relapse and 2 patients were harvested but were not transplanted. 167 patients have been transplanted: 73 FL, 32 DLCL, 26 MCL, 15 MZL, 11 SLL, 9 MALT, and 1 BL. 44 (26%) had HDT during first line therapy. Patients who failed to mobilize had a higher number of treatment before harvest (median 3 compared to 1) and a higher frequency of bone marrow involvement at time of harvest (93% compared to 44%). The 15% failure rate is not high for these patients and was expected, suggesting that rituximab exposure was not associated with poor mobilization. Harvested patients were divided into 3 groups according to the number of CD34+ cells: low (LH) <2.5 106/kg, intermediate (IH) ≥2.5 but <10, and high (HH) ≥10 as previously described. The percentage in each subgroup was not different between patients receiving rituximab and those who did not: LH 7% and 13%, IH 77% and 62%, HH 16% and 25% for patients with and without rituximab exposure, respectively. A poorer harvest was significantly associated with fludarabine exposure (p<.0001), absence of prior rituximab exposure (p=.013), and increased number of previous treatment (p=.022). This was already the case in our previous study, except for prior rituximab exposure that seemed a favorable factor: it was observed in 48% of HH, 33% of IH, and 0% of LH. Hematologic recovery after transplant was not different from what was observed before the rituximab area and expected. Median time to ANC >1000/μl was 12 days and significantly longer in LH (12 vs. 13 days, p=.009). Median time to platelet >50 000/μl was 13 d for HH, 15 d for IH, and 22 d for LH (p=.033). Patients with LH have significantly more days with antibiotics, more platelet transfusions, and longer hospital stays but this was not different from our previous observation. 4 patients died before d100 (2.4%), all from infection. 5 other patients died before d300, all after lymphoma progression. No unexpected infection or other unexpected toxicity was observed. Rituximab as part of HDT increased the quality of response and if used for in vivo purging before stem cell harvest was not associated with a decrease quality of the harvest or with an increase of cytopenia or adverse events after transplant.