HLA-Identical Allogeneic Stem Cell Transplantation (alloSCT) for Patients With Primary Cutaneous T-Cell Lymphoma (CTCL): A Retrospective Analysis of the Lymphoma Working Party of the European Blood and Marrow Transplantation Group (EBMT-LWP)
Biology of Blood and Marrow Transplantation
than 18 years of age. All patients had acute leukemia and were transplanted in 2000-2006. Median follow-up of surviving patients was 2 years in both treatment groups. There were no significant differences in patient and disease characteristics, transplant conditioning regimen, GVHD prophylaxis and donor-recipient HLA disparity by graft type. Though the early probability of neutrophil recovery ($500/ul) was faster after transplantation of PBPC (31% vs. 10% at day-14, p\0.001) the probability of
... the probability of recovery by day-28 was similar after PBPC and BM transplants (94% and 91%, p 5 0.391). Platelet recovery ($20,000/ul) was better after PBPC transplants (86% vs. 76% at day-60, p 5 0.022). Risks of grade 2-4 (hazard ratio [HR] 1.24, p 5 0.147) and grade 3-4 (HR 1.07, p 5 0.785) acute GVHD were similar after PBPC and BM transplants. The risk of developing chronic GVHD was significantly greater after PBPC transplants compared to BM transplants (HR 2.36, p\0.001). After adjusting for disease status, donor-recipient HLA disparity and age, TRM, relapse, leukemia-free survival and overall survival were similar after PBPC and BM transplants. In parentheses is shown the day-100 probability of grade 2-4 acute GVHD (53%/49%) the 3-year probabilities of chronic GVHD (58%/33%), TRM (20%/ 24%), relapse (34%/28%), LFS (46%/48%) and overall survival (49%/49%) for the PBPC/BM groups, respectively. These results differ from transplantation of PBPC from HLA-matched siblings To conclude, in patients \18 years of age with acute leukemia, unrelated PBPC transplant compared to BM resulted in faster recovery of neutrophils and platelets and an increased risk of chronic GVHD. Acute GVHD, TRM, relapse, survival and LFS were similar using PBPC or BM. EBV-associated Hodgkin's Lymphoma (HL) and some non-Hodgkins lymphoma (NHL) have type II viral latency expressing the subdominant EBV antigens EBNA1, LMP1 and LMP2. These antigens may serve as targets for immunotherapy approaches after stem cell transplant. We hypothesized that CTL enriched for effector cells specifically targeting LMP antigens would have efficacy in HL and NHL patients. LMP-CTL were generated using dendritic cells for initial stimulations then EBV-transformed lymphoblastoid cell lines (LCL) both of which had been genetically modified to overexpress either LMP2 alone or inactive LMP1 (DLMP1) and LMP2 by transduction with an Ad5f35LMP2 (n 5 16) or Ad5f35DLMP1-I-LMP2 (n 5 14) vector respectively. All LMP-CTL lines were polyclonal comprising CD41 (mean 17618%; range 1-92) and CD81 (mean 74 6 25%; range 1-99) T-cells. Flow cytometric analysis of memory markers revealed mixed populations of CD45RA-CD62L-T-cells (45615%; range 31-63) and CD45RA-CD62L1 T-cells (3465%; range 28-41). The CTL lines had specificity for CD41 and CD81 restricted LMP2 epitopes alone (n 5 19; mean 1; range 0-7) or both LMP1 and LMP2 epitopes (n 5 13; mean 2; range 0-6) per CTL line, as determined using overlapping LMP1 and LMP2 peptide pools in ELISPOT assays. 24 patients with EBV1 HL and NHL have been treated on dose escalation studies. 16 with LMP2 CTLs and 8 with LMP1/2 CTLs. No immediate toxicity was observed. After CTL infusion, increased numbers of LMP-specific T cells were detected in the blood of 15/22 evaluable patients, (range 2 to 70 fold) persisting for up to 3 months. Additionally, two patients had lymph node biopsies 3-6 months post CTL, which showed selective accumulation of LMP2-multimer positive cells in lymph nodes. 12/13 high-risk and/or multiply relapsed patients who received LMP-CTL as adjuvant treatment after SCT or chemotherapy remained in remission for a median of 2yrs (range . 3mths to 5years) after CTL. 11 patients had detectable disease at the time of CTL, 2 of these had progressive disease by 8 wks and 9 had clinical responses. The median duration of the clinical responses is 1 yr with one stable disease (.12months), one partial response (36 months), and 7 complete responses (range 9 months to .4.5 years). In conclusion, immunotherapy with CTL targeting LMP antigens is well tolerated in patients with EBV1 lymphoma and infused LMP-CTL can accumulate at tumor sites and induce complete and sustained clinical responses. Patients with advanced stage Mycosis Fungoides (MF), Sezary syndrome (SS) and other less common subtypes of primary CTCL have a dismal prognosis with conventional therapy. A potential role for alloSCT in these patients has been suggested by a small number of case-reports and small series. We have now conducted a retrospective registry analysis of 64 recipients of an alloSCT from HLA-Identical donors (52 related, 12 unrelated) for advanced CTCL reported to EBMT between 1997 and 2006: 23 MF, 21 SS, 16 primary cutaneous anaplastic large cell lymphoma and 4 subcutaneous panniculitis-like T-cell lymphoma; 38 male/26 female; median age 46 (5-65); median follow up in survivors, 28 months). AlloSCT was performed at a median of 22 (3-313) months from initial diagnosis, following a median of 3 (1-8) prior lines of therapy, 54 PB and 10 BM as stem cell source, and using a reduced-intensity conditioning (RIC) in 41 (64%) cases. Only 23% of patients were in CR and 26% in PR at the time of al-loSCT. The rest were primary refractory (RE, 25%) or in progression (PG, 26%). Non-relapse mortality was 24% at 3 years, and appeared poorer in patients receiving myeloablative conditioning (MAC) than in those receiving RIC alloSCT (35% vs 16%; p 5 0.06). Acute graft versus host disease (GVHD) occurred in 37% of patients at risk, and chronic GVHD in 69% of patients alive at day 1100. The cumulative incidence of disease relapse (36% at 3 years) was significantly higher in patients with advanced disease status at alloSCT (RE/PG 46% vs CR/PR 21%; p 5 0.01), but similar between related and unrelated donors (p 5 0.39) and between MAC and RIC (p 5 0.58). First progression free survival (PFS) at 3 years was 41%, with a significant advantage for patients in CR/PR at alloSCT (58% vs 26%; p\0.01), but not for type of conditioning (MAC 45% vs RIC 44%; p 5 0.44) or donor type (related 46% vs unrelated 33%; p 5 0.17). Out of 22 patients who relapsed, 10 remain alive and 4 of these are in CR at last follow up, indicating that a proportion of patients can achieve a better control of their CTCL following relapse after alloSCT. Overall survival for the whole series was 55% at 3 years, with a trend towards an improved OS for patients with disease in CR/PR at transplant (65% vs 46%; p 5 0.12). Overall, this outcome appears superior to the expected median OS for these patients with conventional therapy, and reinforces, in the largest series to date, the important role of alloSCT in advanced stage CTCL.