Nonmyeloablative Allogeneic (Allo) BMT for B-Cell Lymphoma with Posttransplant Rituximab: Donor Selection by Prioritizing FCGR3A-158 Polymorphism Over HLA Matching.

In nonmyeloablative, HLA-haploidentical (haplo) related donor BMT with high-dose posttransplant cyclophosphamide (PT/Cy), we found no detrimental effect of greater HLA disparity on event-free survival (EFS; BBMT 2010;16:482–9). This raised the possibility that donor selection need not be dictated by the extent of HLA matching and that non-HLA graft characteristics that enhance antitumor activity may instead be prioritized. The Fc receptor polymorphism FCGR3A-158VV (population frequency 15–20%) is associated with enhanced ADCC and greater sensitivity to rituximab than 158FF. We hypothesized that donor selection based on permissive Fc receptor polymorphisms coupled with posttransplant rituximab would reduce relapse after BMT.

A phase 2 study of nonmyeloablative, haplo or matched related donor BMT with PT/Cy and rituximab was developed for poor-risk B-cell lymphomas. The study uniquely prioritized donors with the most favorable FCGR3A-158 polymorphism (in order of priority, VV, VF, FF), thereby potentially conferring greater sensitivity to rituximab in the host. Donor polymorphism was prioritized over HLA matching and other variables. The primary endpoint was 1-y EFS. Donors were at least haplo first-degree relatives or half-siblings with high-resolution typing at HLA A, B, Cw, DRB1 and DQB1. All patients (pts) received Cy (14.5 mg/kg IV d -6, -5), fludarabine (30 mg/m² IV d -6 to -2), TBI (200 cGy d -1) and T-cell replete bone marrow infusion. GVHD prophylaxis consisted of PT/Cy (50 mg/kg IV d 3, 4), mycophenolate mofetil (d 5–35) and tacrolimus (d 5–180) with filgrastim. Rituximab (375 mg/m² weekly × 8) began d 30, contingent on ANC ≥ 500/μL and could be postponed for GVHD, infection or neutropenia.

In the 67 transplants (57 haplo) performed through 7/2012, 44 pts (66%) maintained or improved polymorphism quality. Thirteen donors (19%) had VV, 36 (54%) VF, 18 (27%) FF. In 11 cases (16%) a relative with a more favorable polymorphism was identified but unsuitable to donate. In 4 cases haplo donors were chosen over full matches due to polymorphism. Outcomes of the 49 pts (median age 60, range 34–74) transplanted through 12/2011 are presented, 39 (80%) having haplo BMT. Diagnoses were DLBCL without transformation (8), transformed lymphoma (12), high-grade large cell lymphoma (1), follicular lymphoma (10), mantle cell lymphoma (11), CLL (5), other (2). This was a poor-risk cohort including 10 (20%) with prior auto BMT, 9 (18%) with induction failure or resistant relapse and 14 (29%) in ≥ 3^rd PR or CR. Yet, with 17-month median follow-up (range 6–33), the estimated 1-y EFS was 74% and 1-y overall survival (OS) was 90% (fig A). With haplo BMT the 1-y EFS and OS were 71% and 87%. In contrast, historically for these diseases, our 1-y EFS is 48% after haplo BMT with PT/Cy and no posttransplant rituximab. Assessment of EFS differences as a function of donor polymorphism was limited by the success of the overall approach and by sample size (fig B). On competing risk analysis, estimated 1-y cumulative incidences (CIs) of nonrelapse mortality (NRM) and relapse were 8% and 17%. Hematopoietic recovery was brisk, with CIs of ANC and platelet recovery of 98% and 88% by d 30. After posttransplant rituximab, ≥1 episode of late-onset grade 4 neutropenia was documented in 26/49 (53%) pts; 17 (35%) had ≥1 episode without apparent cause (e.g. viral infection, GVHD). Most cases had no major clinical consequence. On competing risk analysis, estimated CIs of acute grade II-IV and acute grade III-IV GVHD were 49% and 2%; the 1-y CI of chronic GVHD was 15%. At d 200 the CI of systemic steroid initiation for GVHD was 31%, with 8/24 cases (33%) of grade II GVHD requiring no systemic therapy.

Posttransplant rituximab with nonmyeloablative, haplo or matched BMT and PT/Cy is associated with excellent short-term EFS and OS, without evident increase in severe GVHD or NRM. Efficacy was so high that it is not yet possible to determine whether donor selection by FCGR3A-158 polymorphism is beneficial. Conversely, no adverse consequences of this selection strategy were appreciated. This study suggests that donor prioritization based on non-HLA polymorphisms does not compromise EFS or OS and opens the way for further investigation of donor selection.

View largeDownload slide

View largeDownload slide

Close modal

Kasamon:Genentech: Research Funding. Off Label Use: cyclophosphamide post-transplant for GVHD prophylaxis. Swinnen:Genentech: Honoraria, Research Funding.