Non-Hodgkin’s lymphoma (NHL) includes heterogeneous entities of mature lymphoid malignancies that vary in biology, clinical behavior, and treatment. Chemotherapy alone remains the most utilized upfront therapy for the majority of NHL cases. Chemotherapy alone is capable of curing roughly half of aggressive NHL cases and attaining long-term control in the majority of indolent subtypes. Nevertheless, there is a well-established role for autologous (auto) hematopoietic cell transplant (HCT) in relapsed chemosensitive aggressive NHL, and occasionally, it has a role in the upfront setting such as consolidation modality for mantle cell lymphoma (MCL). Although auto-HCT is a safe procedure, and it is associated with low morbidity and mortality, it has a limited role in heavily pretreated or relapsed patients that had failed to mobilize adequate stem cells, chemorefractory relapsed lymphomas, and those who relapsed/progressed after prior auto-HCT. In those situations, allogeneic HCT (allo-HCT) may be the only curative option, and it has been shown to be effective but is also associated with high transplant-related morbidity and mortality (TRM). In this review, we will discuss the current available evidence for the role of allo-HCT in the different subtypes of NHL.

Follicular lymphoma (FL) is the prototype of low-grade lymphoma. It is characterized by indolent clinical behavior and sensitivity to chemotherapy. However, the natural history of FL involves frequent disease recurrences and shorter response durations with each relapse. Eventually, FL progresses to become chemoresistant, and portion of cases transforms to more aggressive lymphoma subtypes. Despite long survival that exceeds a decade in more recent studies, FL is considered an incurable disease with conventional chemotherapy, and there is an increasing interest in using allo-HCT as a curative modality in advanced disease and younger patients.

In relapsed/refractory (R/R) FL, phase II prospective studies from the MD Anderson Cancer Center (MDACC), the Cancer and Leukemia Group B (CALGB 10901), the Spanish group (GELTAMO), and the Canadian transplant group have established the feasibility of reduced-intensity conditioning (RIC)/non-myeloablative conditioning (NMA) allo-HCT and have documented low relapsed/progression risk (RR) as well as low non-relapse mortality (NRM) [1–4]. Investigators from MDACC prospectively treated 47 chemosensitive relapsed FL with allo-HCT following NMA conditioning with FCR (fludarabine, cyclophosphamide, and rituximab). The majority of patients transplanted from matched–related donors (MRD), and 19 % had already failed prior autograft. Long-term follow-up demonstrated low RR and NRM of 6 and 13 %, respectively, in addition to high 11-year progression-free survival (PFS) and overall survival (OS) of 72 and 78 %, respectively. Subsequently, rituximab was replaced with ⁹⁰Y-ibritumomab (YFC), and another 26 relapsed FL patients were treated prospectively. Among the YFC cohort, more patients were chemorefractory compared to the FCR cohort (38 vs. 0 %) and have undergone transplant from matched unrelated donor (MUD) (42 vs. 4 %). Relapse risk and NRM were low, and both were 8 %. Three-year PFS and OS were 85 and 88 %, respectively. Although pretransplant PET activity did not predict outcomes in either cohort, there was a trend toward decreased 3-year OS in the FCR cohort for patients with positive PET activity (57 % vs. 90 %. P = 0.06). Complete remission (CR) was achieved in the majority of patients after transplant, including 9 of 10 chemorefractory patients, and there was no significant difference in PFS or OS based on chemosensitivity. The 6-year OS for chemorefractory disease was 80 %, which is very promising in this high-risk cohort [1].

The CALGB 109901 is a prospective phase II study that enrolled 44 patients with relapsed chemosensitive indolent NHL and chronic lymphocytic leukemia (CLL), including 16 patients with FL. All patients underwent allo-HCT from MRD using RIC with FC. The RR and NRM rates were 19 and 9 %, respectively. The 3-year event-free survival (EFS) and OS rates for the FL cohort were 75 and 81 %, respectively, which were higher than their counterpart patients with CLL and other indolent lymphomas in the study [2].

The Spanish group (GELTAMO) analyzed the outcomes of 2 prospective protocols including 37 patients with FL who had undergone allo-HCT from MRD using RIC with fludarabine and melphalan (Flu/Mel). Among enrolled patients, 18 % were chemorefractory and 46 % had already failed prior autograft. The 4-year RR and NRM rates were 8 and 37 %, respectively. Chemorefractory lymphoma prior to transplant was a risk factor for increasing NRM (P = 0.04). The 4-year PFS and OS rates for the whole group were 55 and 57 %, respectively. Disease status prior to transplant was associated with a trend toward improved PFS and OS after transplant, but it did not reach statistical significance. The 4-year OS in patients with CR, PR, and refractory disease were 71, 48, and 29 %, respectively (p = 0.09) [3].

The Canadian study used a different approach to treat R/R FL by applying tandem auto-HCT followed by NMA allo-HCT from MRD. Among 27 enrolled patients, 19 % were considered chemorefractory at the time of auto-HCT and 19 % had transformed histology. The conditioning regimen was BEAM (carmustine, etoposide, cytarabine, melphalan) or BEAC (carmustine, etoposide, cytarabine, cyclophosphamide) for auto-HCT and Flu/Cy (Fludarabine, cyclophosphamide) for allo-HCT, and the median time between the 2 transplants was 133 days. The Canadian approach was proven to be active and safe. There was no reported relapse during the follow-up period, and the NRM was as low as 4 %. The 3-year PFS and OS rates were 96 % for both. No cases of therapy-related MDS/AML was observed during follow-up [4].

However, retrospective analysis of transplant registries reported the outcome of allo-HCT in R/R FL were inferior to single center experiences, but these results provide more insights regarding the role of allo-HCT in FL. The largest analysis was reported by the Center for International Blood and Marrow Transplant Research (CIBMTR), in which myeloablative (MAC) (n = 120) was compared to RIC (n = 88) in patients who had undergone allo-HCT from MRD for R/R FL. Although 3-year OS (62 % vs. 71 %, p = 0.15) and PFS (55 % vs. 67 %, p = 0.07) were not different between RIC and MAC, respectively, there was an increased risk for lymphoma relapse after RIC in multivariate analysis (Relative risk = 2.97, p = 0.04). The 3-year probability for lymphoma progression was 8 % after MAC and 17 % after RIC in univariate analysis. Poor performance status (PS) and chemorefractory disease were associated with lower PFS and OS, and higher TRM [5].

In contrast to the CIBMTR analysis, the European Group for Blood and Marrow transplantation (EBMT) analysis compared the outcome of allo-HCT in 131 patients who had undergone allo-HCT MUD for R/R FL (RIC = 87, MAC = 44). The RIC cohort was older, more frequently had undergone prior autograft, and less frequently used ATG as part of conditioning compared to the MAC cohort. The 3-year NRM, RR, PFS, and OS rates were 33, 30, 47, and 51 %, respectively, for the whole cohort. No significant difference was noticed between RIC and MAC in the risk of acute or chronic graft versus host disease (GVHD); however, the occurrence of acute GVHD (aGVHD) was associated with higher risk of NRM, and shorter PFS and OS. In multivariate analysis, lower NRM and higher PFS and OS were seen for patients who had undergone RIC, and lower PFS and OS for patients who had prior autograft. Of 61 patients with prior autograft, 3-year NRM, RR, PFS, and OS rates were 42, 19, 39, and 42 %, respectively [6].

The role of T cell-depleted RIC allo-HCT in FL was the focus of the United Kingdom (UK) analysis. Among 82 patients who were conditioned with Flu/Mel/Alemtuzumab, 9 % had chemorefractory disease, 26 % had failed prior autograft, and 54 % received transplant from MUD. The study reported low NRM (15 %), aGVHD (13 %), and chronic GVHD (cGVHD) (18 %). The 4-year PFS and OS rates were 76 % for the whole cohort. Post-transplant, 52 % of 64 evaluable patients with chimerism study had mixed chimerism (MC). Of 28 patients who received DLI for MC, 61 % achieved full chimerism (FC). The 4-year RR was higher for patients with persistent MC compared to FC cohort (36 % vs. 10 %; p = 0.03). There was a trend toward higher NRM in MUD recipients compared to MRD (22 % vs. 8 %, p = 0.08). The 4-year PFS was lower in MUD compared to MRD (64 % vs. 90 %; p = 0.012), and in patients who had prior autograft compared for those without (57 % vs. 83 %; p = 0.016), but there was no difference based on disease chemosensitivity at the time of transplant (p = 0.133). Nevertheless, in multivariate analysis, MRD was the only variable associated with improved PFS (p = 0.035) and OS (p = 0.037). Salvage DLI was effective for relapse after transplant as 10 of 13 patients (77 %) achieved CR, and this including 9 of whom had a durable response [7].

The Fukuoka BMT Group from Japan analyzed the outcome of 30 R/R FL who had undergone allo-HCT (MAC = 43 %, RIC = 57 %), including 50 % with chemorefractory disease and 7 % with prior autograft. There was no difference in outcomes between RIC and MAC. The 2-year RR, NRM, PFS, and OS rates were 20, 33, 47, and 47 %, respectively, for the whole cohort [8].

Mantle cell lymphoma (MCL) is a unique form of NHL that shares the incurability feature of low-grade lymphoma and the clinical behavior of aggressive subtypes. Auto-HCT is part of MCL treatment in the upfront and the relapsed setting. Although allo-HCT was used in those with chemorefractory disease and those relapsed after prior autograft, there are no prospective studies that define the exact role.

Data from the CIBMTR compared the outcomes of auto (n = 381)- versus RIC allo-HCT (n = 138) in chemosensitive MCL and analyzed outcomes based on the timing of transplantation (early = 299, late 220). Early transplantation was defined as those who received transplant in PR1/CR1 with no more than two lines of prior chemotherapy. Allo-HCT was associated with higher NRM but lower RR in both early and late cohorts. There was no significant difference in 5-year OS for early (61 % vs. 62 %, p = 0.95) and late (44 % vs. 31 %, p = 0.2) transplantation for auto-HCT and RIC allo-HCT cohorts, respectively. Early transplant was associated with survival benefit for both transplant cohorts in multivariate analysis [22]. Another analysis from the CIBMT focusing on the role of allo-HCT in chemorefractory MCL identified 202 patients (MAC = 74, RIC/NMA = 128). More patients in the RIC/NMA cohort had history of prior autograft, received prior rituximab, and had undergone MUD transplant. Despite this, there was no significant difference between MAC and RIC/NMA cohorts in terms of NRM (47 % vs. 43 %, p = 0.68), RR (33 % vs. 32 %, 0 = 0.89), PFS (20 % vs. 25 %, p = 0.53), and OS (25 % vs. 30 %, p = 0.45) at 3 years. This study showed that about 25–30 % of refractory MCL can be cured with allo-HCT [23].