Allogeneic Stem Cell Transplantation in Mantle Cell Lymphoma; Insights into Its Potential Role in the Era of New Immunotherapeutic and Targeted Therapies: The GETH/GELTAMO Experience

Simple Summary We present the long-term results of patients receiving allogeneic stem cell transplantation (allo-SCT) for relapsed/refractory mantle cell lymphoma (R/R MCL) in the last 25 years in Spain. We conclude that allo-SCT may be a curative option in R/R MCL with a low cumulative incidence (CI) of relapse, although non-relapse mortality (NRM) is still high, which is mainly secondary to acute graft-versus-host disease (aGVHD). Results are better for fit patients, using HLA-identical (related or unrelated) or haploidentical related donors and without previous ASCT. However, the arrival of new highly effective and low toxic immunotherapeutic or targeted therapies inevitably will relegate allo-SCT to those fit patients who fail these therapies, being administered far away from the optimal timing. Abstract Allo-SCT is a curative option for selected patients with relapsed/refractory (R/R) MCL, but with significant NRM. We present the long-term results of patients receiving allo-SCT in Spain from March 1995 to February 2020. The primary endpoints were EFS, OS, and cumulative incidence (CI) of NRM, relapse, and GVHD. We included 135 patients, most (85%) receiving RIC. After a median follow-up of 68 months, 5-year EFS and OS were 47 and 50%, respectively. Overall and CR rates were 86 and 80%. The CI of relapse at 1 and 3 years were 7 and 12%. NRM at day 100 and 1 year were 17 and 32%. Previous ASCT and Grade 3–4 aGVHD were associated with a higher NRM. Grade 3–4 aGVHD, donor type (mismatch non-related), and the time-period 2006–2020 were independently related to worse EFS. Patients from 1995–2005 were younger, most from HLA-identical sibling donors, and were pretreated less. Our data confirmed that allo-SCT may be a curative option in R/R MCL with low a CI of relapse, although NRM is still high, being mainly secondary to aGVHD. The arrival of new, highly effective and low toxic immunotherapeutic or targeted therapies inevitably will relegate allo-SCT to those fit patients who fail these therapies, far away from the optimal timing of treatment.


Introduction
Mantle cell lymphoma (MCL) is an uncommon B-cell lymphoma that generally has a poor prognosis, with high rates of chemorefractoriness and an advanced median age at diagnosis [1]. For transplant-eligible patients, clinical outcome improves using intensive cytarabine-based induction chemotherapy, followed by autologous stem cell transplant (ASCT) [2,3] consolidation. Recently, new approaches with anti-CD20 maintenance [4,5], new target drugs [6,7], or new ways of immunotherapy such as CARTs [8] are changing the front and salvage therapeutic lines.
Allogeneic stem cell transplantation (allo-SCT) is a potential curative option for selected patients, mediated by a well-demonstrated graft versus lymphoma (GVL) effect in this lymphoma [9]. However, allo-SCT in MCL has also been associated with significant non-relapse mortality (NRM) [10,11]. The efficacy and toxicity of allo-SCT should be well-balanced, considering that MCL typically affects an older population with potentially higher rates of comorbidities, and there is an increasing number of emerging effective and manageable alternative therapeutic approaches. For these reasons, there is a need to clarify candidates' selection, and which is the optimal target population for allo-SCT. We should especially consider that most previously published studies that focused on investigating the role of allo-SCT in MCL are retrospective and conclusions are limited by a reduced sample size of patients. Furthermore, this is particularly important with the outstanding efficacy and manageable toxicity associated with new immunotherapies based on CAR-Ts in relapsed/refractory (R/R) MCL [8,12].
Our objective was to analyze the long-term results of MCL patients undergoing allo-SCT in Spain, trying to define better its current role in the era of new immunotherapeutic and targeted therapies, focusing on candidates' selection.

Study Design and Patient Eligibility
We designed a retrospective multicenter study including all registered patients from centers' members of the Spanish Group of Hematopoietic Transplantation (GETH)/Spanish Group of Lymphoma (GELTAMO) with relapsed or refractory (R/R) MCL consolidated with allo-SCT. For this purpose, all patients who had undergone an allo-SCT in the abovementioned centers from March 1995 to February 2020 that was reported to the EBMT registry were eligible. The primary endpoints were event-free survival (EFS) and overall survival (OS). Secondary outcomes were cumulative incidence (CI) of NRM, relapse, and graft versus host disease (GVHD). The study was performed in compliance with the Declaration of Helsinki and approved by research ethics committees and institutional review boards at each participating institution. As part of the EBMT registration, all patients signed informed consent.

Data Recovery and Study Definitions
The histological diagnosis was based on a local review, and patients were staged according to the Ann Arbor system. Disease status was assessed by the local team according to the Revised Response Criteria for Malignant Lymphoma [13] and/or Lugano Classification [14]. Myeloablative conditioning was defined as a regimen containing either total body irradiation (TBI) with a dose greater than 6 Gy, a total dose of oral busulfan greater than 8 mg/kg, or a total dose of intravenous busulfan greater than 6.4 mg/kg. All other regimens were defined as reduced-intensity conditioning (RIC) [15]. The diagnosis and grading of acute and chronic graft versus host disease (aGVHD and cGVHD) were performed by the transplant centers using the standard criteria [16].

Statistical Analysis
All outcome measures were assessed from the time of allo-SCT. OS was defined as the time to death. Event-free survival (EFS) was defined as the time to relapse, progression, or death from any cause. NRM was defined as the time to death without previous disease relapse or progression (considering relapse as a competing event). CI of relapse was defined as the time from relapse or progression (considering death without relapse as a competing event).
Qualitative or binomial variables are expressed as frequencies and percentages. Comparisons between qualitative variables were made using the Fisher Exact Test or the Chisquared test. Comparisons between quantitative and qualitative variables were performed through non-parametric tests (U of Mann-Whitney or Kruskal-Wallis). The binary logistic regression was used to find out the risk factors associated with day 100 complete response (CR) or NRM. Time to event variables were estimated according to the Kaplan-Meier method and comparisons between variables of interest were performed by the log-rank test. Multivariate analysis with the variables that appeared to be significant in the univariate analysis was carried out according to the Cox proportional hazard regression model (stepwise forward likelihood ratio selection). Those variables not available at transplant were included as time-dependent variables. All p values reported were 2-sided, and statistical significance was defined at p < 0.05. To analyze the impact of time periods on the survival of transplanted patients, we segmented the full range of follow-up (1995-2020) of our series using MAXTAT for disease progression or death. The statistical analysis was performed using SPSS software (SPSS version 28.0; IBM, Chicago, IL, USA) and RStudio (Version 1.3.959; RStudio, PBC, Boston, MA, USA).

Patient and Allo-SCT Characteristics
A total of 135 patients with R/R MCL that fulfilled the inclusion criteria were included in the study. Table 1 summarizes the main patient and allo-SCT information. Briefly, the median age of the study cohort at the time of the allo-SCT was 56 years (32-70), with 27% of patients being older than 60 years. A proportion of 66% of patients had classic, 27% blastic, and 7% indolent MCL. The median time from diagnosis to allo-SCT was 33 months , and the median number of previous lines of therapy before allo-SCT was two (one to eight), including previous autologous SCT (ASCT) in 49% of patients. Disease status before allo-SCT was as follows: complete response (CR) in 86 patients (64%), partial response (PR) in 35 (26%) patients, and SD/PD in 13 (10%).

Impact in Results of Allo-SCT Candidates' Selection along Decades
We focused on the impact of time periods on the survival of transplanted patients. For this purpose, we segmented the full range of follow-up (1995-2020) of our series using MAXTAT for EFS, obtaining two cutoffs at 15 and 9 years: 2005 and 2011. As shown in Table 3 and Figure 2, the best outcomes in terms of EFS were obtained from 1995 to 2005; with the worst observed from 2006 to 2011. Since 2012, the results slightly improved again but still did not achieve previous levels. However, the 5-year CIs of relapse were similar between all three time periods from 1995 to 2020: 14, 15, and 19% (p = 0.89).
We compared patient characteristics along time periods to discover the causes of these results. As shown in Table 4, patients from 1995 to 2005 were younger, with a much shorter interval from diagnosis to allo-SCT; most of them were from HLA-identical sibling donors and with much less previous therapy. In other words, from 1995 to 2005, patients were much more selected for and transplanted earlier than after 2005. This translated into a lower NRM and a better EFS and OS in these patients (Tables 3 and 4).

Discussion
Our study presents real-world evidence from GETH and GELTAMO Spanish centers about the role of allo-SCT in MCL, confirming its efficacy as a potentially curative option but, at the same time, highlighting its major handicap in terms of potential toxicity and high NRM. However, the most important point is that our series provides a great insight into the importance of candidates' selection for allo-SCT, which limits its current role in the era of new immunotherapeutic and targeted therapies in MCL.
Most previously published series about the role of allo-SCT in MCL are retrospective and generally small. Our retrospective series with 135 patients compares favorably with most of them in terms of size. It was obtained from all patients reported to the EBMT registry from Spanish centers, with a 5-year EFS and OS of 47 and 50%, respectively. This is in the range between 30-60% of the previously published studies [9,18,19].
All these works and our series provide convincing evidence of the existence of an allogeneic GVL effect, suggesting a curative potential, although this is weaker than in indolent lymphoma [9,[18][19][20]. This is also illustrated by the better EFS in patients having cGVHD and the high number of responses in our patients: overall 80% CR, which was higher in patients with previous CR (95%) but also in cases with previous PR (73%) or even SD/PD (50%), demonstrating the graft-versus-MCL effect. Furthermore, the incidence of cGVHD was significantly higher in those patients with less than a CR at pretransplant, who then obtained a CR posttransplant (43%), vs those not achieving a CR (0%) (p = 0.027). Other works have reported that chemorefractoriness is not a major risk factor for disease control in MCL after allo-SCT [21]. In our series, the relapse rate was not a major challenge (only 7 and 12% at 1 and 3 years, respectively). This contrasts with other RIC series in which relapse was reported in up to 40% [18]. However, the high rate of NRM might reduce the number of patients at risk of relapse in our series.
Most previously published studies in allo-SCT in MCL share high rates of toxicity in terms of 10 to more than 50% of NRM, as well as high rates (30-40%) of acute or chronic GVHD. Of course, the higher rates of NRM have been reported to be associated with myeloablative conditioning regimens as well as in more pretreated patients, particularly those failing ASCT [10,11], as was seen in our patients. These high rates of NRM may be lower (below 30%) with a similar efficacy using RIC, as shown in several retrospective [18,19,[21][22][23] or even prospective clinical trials [24]. New strategies of GVHD prophylaxis such as cyclophosphamide post-allo-SCT could also improve NRM results in these patients [25,26]. However, only 9% of our patients received this prophylaxis in some of our last haploidentical transplants. In our series, we had 17 and 32% CI of NRM at 100 days and 1 year, respectively, which was independently associated with Grade 3-4 aGVHD incidence and previous ASCT, but not with other transplant characteristics. This may be influenced by the fact that most of our patients received an RIC regimen (85%).
For these reasons, once one demonstrates efficacy, this should be balanced against significant rates of NRM, and is when patient selection and other therapeutic options should be considered. In our series, we observed worse outcome results in terms of EFS and OS in patients with Grade 3-4 aGVHD who were transplanted with unrelated mismatched donors (related mismatched transplants, including haploidentical ones, fared similarly to HLA-identical procedures) and, unexpectedly, in the period 2006 to 2020, this was clearly related to a change in the pattern of candidate selection towards older and more pretreated patients beyond 2006. This is the logical consequence of having much better alternatives to allo-SCT in this population of patients, with also better results.
Nowadays, frontline high-dose cytarabine-containing programs followed by ASCT have been considered the standard of care for young and fit patients [3,4]. When comparing the abovementioned time-periods, in the older cohort (1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005), less than half of the patients received an ASCT (19 vs. 56% beyond 2006) as the value of frontline ASCT in MCL was first reported in 2005 [2]. Maintenance with rituximab has been shown to prolong PFS and OS, both in fit or unfit patients [4,5]. New, non-cross resistant chemotherapeutic drugs have shown interesting activity in MCL such as oxaliplatin [27][28][29], bortezomib [30], or bendamustine [31][32][33]. Targeted therapies such as BTK inhibitors [6,34,35] or venetoclax [7,36] obtain impressive results with manageable toxicity, representing good salvage options that may also delay the decision of allo-SCT. Finally, outstanding results have been reported with new immunotherapies with anti-CD19 CAR-T [8,12], which led to the approval of brexucabtagene autoleucel as salvage therapy for R/R MCL patients. Points favoring CAR-T cell therapy are that it is more effective than allo-SCT in patients with active disease, with much lower rates of toxicity and NRM. On the other hand, the follow-up with CAR-T cell therapy is still short when compared with allo-SCT, so it is not known for its long-term curative potential; there is also a relevant economic impact, and limited accessibility. Considering the sequencing of both approaches, if CAR-T cell therapy fails, patients could still receive an allo-SCT but in a more pretreated status that we know would further reduce the efficacy and increase the toxicity of this procedure.
The most important updated guidelines from the American Society of Transplantation and Cellular Therapy, Center of International Blood and Marrow Transplant Research, European Society for Blood and Marrow Transplantation [37], and National Comprehensive Cancer Network (NCCN) recommend CAR-T cell therapies in MCL from the third line for patients who are intolerant to or relapse after at least one BTKi. NCCN guidelines also recommend considering allo-SCT as consolidation for high-risk, young, and fit responders to a second line. Consequently, if there is an increase in the accessibility of CAR-T cell therapy, we hypothesize that CAR-T cell therapies will probably delay allo-SCT to later lines of therapy, far away from the optimal timing evidenced in our series.
Our work has several limitations. First, it is a retrospective study that implies real world evidence but lacks homogeneity in terms of frontline or salvage therapies, supportive care strategies, and GVHD prophylaxis and management. We do not have some important diagnostic or prognostic information, such as molecular high-risk profiles or TP53 mutation status that could modify therapeutic decisions for allo-SCT candidates [38,39]. Addressing the abovementioned lack of some new strategies for GVHD prophylaxis such as cyclophosphamide post-allo-SCT, we could also mention a few patients receiving ibrutinib before and after allo-SCT, which may provide a benefit in terms of survival [40]. However, in our series, we only had 12 (9%) patients with cyclophosphamide post-allo-SCT prophylaxis and 19 (10%) having received ibrutinib.
Of note, while the older cohort (1995-2005) comprised essentially of identical siblings in contrast with the later cohorts, GVHD incidence was not significantly different and Grade 3-4 aGVHD was only 10% lower in the older cohort. This may illustrate the improvement in GVHD prophylaxis and therapy with the time. At the same time, in our study there were several patients at a high risk for poor outcomes, in which allo-SCT usually is contraindicated: ECOG PS 2-4 (5%) and pretransplant SD/PD (10%). In our series, 5 y-PFS was respectively 0 and 15%, illustrating the potential for disease control of allo-SCT even in selected cases with active disease at transplant, as well as the need of a good ECOG PS to avoid NRM.

Conclusions
Taking all this data together, allo-SCT is a feasible and effective therapy in MCL with a well-demonstrated GVL effect favored by cGVHD even in situations of active disease at allo-SCT, but still with high rates of toxicity and NRM. As concluded from our time-period analysis, allo-SCT may be a better approach for young, fit, high-risk patients consolidated early (i.e., second line) that probably is linked to lower rates of severe aGVHD and NRM. Improved outcomes may be obtained using HLA-identical (related or unrelated) or haploidentical related donors, which are better than mismatched, unrelated donors. However, the arrival of new highly effective and low toxic immunotherapeutic or targeted therapies inevitably will relegate allo-SCT to fit patients who fail these treatments, which would then be administered far away from the optimal timing.

Institutional Review Board Statement:
The study was performed in compliance with the Declaration of Helsinki and approved by research ethics committees and institutional review boards at each participating institution. While the inclusion of data in the EBMT registry had a different committee for every institution, this study was approved by Comité de Ética de la Investigación de las Islas Baleares, 25 July 2018. Reference: IB 3746/18 EPA.
Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.