Frontline therapy and role of high-dose consolidation in mantle cell lymphoma

As with any aggressive form of lymphoma, the cornerstone of therapy begins with CHOP (cyclophosphamide, doxorubicin, vincristine, and prednisone). Although this is clearly active with high response rates in this disease, these are rarely complete or very durable compared with those observed with other aggressive lymphomas. The major advance came with the incorporation of cytarabine into the treatment algorithm (Table 1). There were 2 broad approaches. First, the Hyper-CVAD (hyper-fractionated cyclophosphamide, vincristine, doxorubicin, and dexamethasone, alternating with cytarabine and methotrexate) regimen, pioneered at the MD Anderson Cancer Center, used high-dose cytarabine in combination with a number of chemotherapeutic agents in a dose-intense schedule. This lead to unprecedented results, with extremely high complete response (CR) rates and durable responses.¹  In Europe, the DHAP (dexamethasone, cytarabine, and cisplatin) regimen was used after CHOP, again showing a marked improvement in responses and the durability of responses.² 

The use of autologous stem cell transplantation was widely used in the context of relapsed mantle cell lymphoma (MCL) with good evidence that the earlier it was applied the better the subsequent outcome. As a consequence, a study randomizing patients to transplant or interferon following CHOP therapy was performed. This showed a benefit initially with respect to progression-free survival (PFS) and subsequently overall survival (OS) and was adopted as a new standard of care.³  It is important to realize that this is the only randomized study, and it was performed in the pre-cytarabine era, but the results have been widely applied including following more intensive induction regimens. A further advance was the incorporation of rituximab into common regimens for MCL, which has now been shown to improve OS.⁴  Probably the most important trial randomized almost 500 patients to an autologous transplant following R-CHOP or R-CHOP/R-DHAP.⁵  This clearly showed that the best and most durable responses following an autograft required cytarabine as part of the induction therapy. In other words, an autograft does not compensate for the inferior response rates observed with R-CHOP alone. Therefore, the best therapy today includes high-dose cytarabine with rituximab with the addition of an autograft as consolidation unless the R-Hyper-CVAD approach is used. The question arises as to what needs to be added to these 2 drugs. The R-Hyper-CVAD has received some criticism as the excellent results seen initially have failed to be reproduced in subsequent multicenter studies,^6,7  and one could argue that some components of this, eg, methotrexate, add nothing but toxicity. The Nordic group adopted the Hyper-CVAD dose of cytarabine, added rituximab, and alternated with an augmented R-CHOP regimen. This led to arguably the best published results from a multicenter setting with a median OS in excess of 10 years,⁸  but as with all of these regimens, there is an ongoing pattern of relapse.

Minimal residual disease (MRD) assessment is emerging as a very important prognostic factor in MCL. Analysis of 259 patients treated within the European MCL network elderly and younger trials showed that 56% were MRD negative after induction therapy.⁹  Of these, 87% remained in remission at 2 years compared with 61% who were MRD positive. MRD negativity was an independent prognostic factor for response duration and was also independent of the clinical response. This has been confirmed in another study where MRD positivity in patients who were in a conventional CR before autologous transplantation predicted for a shorter remission duration.¹⁰ 

As such, it is conceivable that MRD analysis could be used to adapt therapy, either by augmenting treatment prior to stem cell transplantation or conversely using that as a rationale to avoid the toxicity of a high-dose procedure. ⁹⁰Y-ibritumomab-tiuxetan (zevalin) has demonstrable activity in relapsed/refractory MCL. The Nordic group incorporated zevalin into their standard protocol high-dose AraC/R-CHOP, where it was given prior to the autograft in those patients who were not in a CR. Unfortunately this approach failed to demonstrate any benefit following the addition of this agent,¹¹  and the use of zevalin as consolidation following R-Hyper-CVAD led to substantial and unacceptable toxicity.¹²  A recent publication using a high-dose anti-CD20 radioimmunotherapy (using ¹³¹I-tositumomab)–based autograft in a nonrandomized comparison against standard conditioning suggested an improvement in outcomes following radioimmunotherapy.¹³  Although this therapeutic option is no longer available, this study suggested that the higher dose of antibody applied might have been an important factor.

Given the improvement in PFS seen with bendamustine plus rituximab (BR) over R-CHOP in patients with MCL, an alternative approach is to substitute bendamustine as part of the induction therapy prior to transplantation. A recent small phase 2 study used 3 cycles of BR followed by 3 cycles of high-dose cytarabine prior to an autologous transplant. The follow-up is short, but 21 of 23 patients were successfully transplanted, and after a median follow-up of 13 months, the PFS was 96% with a very high incidence of MRD negativity in evaluable patients.¹⁴  These results are better than what might be expected with BR alone, again showing the importance of the cytarabine component of the therapy.

A similar approach using BR followed by an autograft was 1 arm of a recently reported US intergroup study (S1106).¹⁵  This study randomized against an abbreviated Hyper-CVAD followed by an autologous transplant for young patients. The trial was stopped early because of difficulties in mobilizing stem cells in the Hyper-CVAD arm. This was unfortunate as only 53 patients were randomized in total, and there was no difference between the 2 arms with respect to response, PFS, and OS at 2 years.

A recently presented multicenter study from France¹⁶  shows potentially comparable results with those of the Nordic group when R-DHAP alone is used prior to the transplant. This study uses 4 cycles of therapy and importantly had a randomization to rituximab maintenance or observation following the transplant. This demonstrated a large improvement in PFS in favor of the rituximab arm, suggesting that this should be adopted as a new standard. Such an approach is far more pragmatic than adopting a preemptive approach to the use of rituximab by targeting molecular relapse following high-dose therapy as has been used in the Nordic style of management.¹⁷  Taken together, it is not obvious that current radio-immunotherapy approaches offer any significant benefit over rituximab alone, but the application of maintenance is clearly important.

With the widespread adoption of the less toxic reduced intensity conditioning (RIC) allogeneic procedures at relapse coupled with a potential for cure, is there a case for adopting this approach as part of the consolidation of frontline therapy? The largest experience comes from an analysis of 519 patients on the Centre for International Blood and Marrow Research database who received either an autologous or RIC allogeneic transplant as first consolidation for chemosensitive MCL.¹⁸  Patients were analyzed by time of transplantation into early (in first partial response/CR after no more than 2 lines of prior therapy) or late (all other patients). The study concluded that early transplantation was superior to late for both auto–hematopoietic cell transplantation (auto-HCT) or RIC allo-HCT. For early transplantation after 5 years, relapse was significantly lower in favor of RIC allo-HCT (15% vs 32%), but OS was identical (∼60%) due to higher non–relapse-related mortality in the patients receiving an RIC allo.

There have been 2 prospective studies using allogeneic transplants as part of frontline therapy. The final report of the East German Study Group of Haematology and Oncology¹⁹  combined data on 2 studies: one frontline and the other in relapsed patients. Twenty-one of 24 frontline patients proceeded to an allogeneic procedure following either R-CHOP or R-CHOP/R-DHAP therapy. At transplant, only 43% were in a complete remission, but at 5 years the OS was 73%. The outcomes appeared the same whether the transplant was performed up front or at relapse, and the remission status bore no relationship to outcome. The other prospective trial is the recently presented UK study that performed a RIC allo-HCT using the carmustine (BCNU), etoposide, cytarabine, melphalan Campath regimen in 25 patients with untreated MCL.²⁰  Fewer than 50% of patients were in CR at the time of transplant following a variety of induction regimens. The 2-year treatment-related mortality was only 8.6%, with 2-year PFS and OS of 68% and 80%, respectively, but with a 50% incidence of chronic graft-versus-host disease. In keeping with the German study, the remission status before transplant did not affect outcome. Although immature, the results are encouraging and provide data to support frontline allogeneic transplant for some patients; however, in the era of exciting highly active and relatively nontoxic novel agents, where might this approach be considered? It is now possible to predict those young patients who are likely to get the least benefit from an autologous procedure. The recently published paper from the European MCL group adapted the well-established Mantle Cell Lymphoma International Prognostic Index by incorporating KI67 into the algorithm using a cutoff at 30% as a prognostic factor.²¹  This allows the identification of a very high-risk group of patients where an alternative approach could be considered. As such, serious consideration should be given to an RIC allo-HCT as consolidation in the very few young patients who present with high risk (combined Mantle Cell Lymphoma International Prognostic Index) disease or who are primary refractory to initial cytarabine therapy and subsequently achieve a remission with an alternative strategy.

For older patients where autologous stem cell transplantation or intensive cytarabine-based therapy is not feasible, there are a number of conventional options (Table 2). Although there is no convincing evidence in favor of the use of anthracyclines, CHOP-based therapy has been the mainstay of therapy in this group of patients. There are few randomized trials in this setting, but the largest (485 patients) compared the use of R-CHOP with rituximab, fludarabine, and cyclophosphamide.²²  Although the time to treatment failure was identical in both arms, OS was significantly improved in the R-CHOP arm, reflecting the challenge in treating patients following purine analog-based therapy and the complications that accompany it. The second randomization in this study was to maintenance with either rituximab or interferon. The addition of rituximab produced a significant improvement in OS in the R-CHOP arm, leading to an unprecedented 4-year OS of 87%, which was not mirrored in the FCR arm.

The other major chemotherapeutic approach is the use of bendamustine. There have been 2 small randomized studies comparing BR with R-CHOP in MCL. The first demonstrated an improvement in PFS in favor of BR from 22.1 to 35.4 months, coupled with an improved side effect profile.²³  The second compared R-CHOP or R-CVP (rituximab, cyclophosphamide, vincristine, and prednisone) with BR and showed no significant difference between the arms with respect to outcome.²⁴  Both of these trials were small (<100 patients) and reported on the MCL patients as a subset analysis of a larger “indolent” lymphoma trial, and neither study involved any maintenance. As such, it is difficult to definitively recommend R-CHOP or BR until it is clear what maintenance adds, if anything, following the use of BR.

An optional approach with bendamustine is to incorporate this together with low-dose cytarabine and rituximab (the R-BAC regimen).²⁵  In the relapse setting, this produces responses of 80%, albeit in a largely older population and produced a CR rate of 95% when adopted as frontline therapy. However, this was accompanied by significant hematologic toxicity, but by reducing the dose of cytarabine, this efficacy has been maintained but with less toxicity.²⁶ 

There are currently 4 drugs licensed for use in MCL across the world: bortezomib (Velcade; Janssen), temsirolimus (Torisel; Pfizer), lenalidomide (Revlimid; Celgene), and ibrutinib (Imbruvica Pharmacyclics Inc). As single agents, the ORRs for these drugs are 33% (8% CR) with bortezomib, 22% (2% CR) with temsirolimus, 28% (8% CR) with lenalidomide, and 68% with ibrutinib (21% CR).²⁷  With the possible exception of ibrutinib, it seems unlikely that these drugs will be used as single agents for the treatment of MCL outside of maintenance strategies, but they may have a role as part of combination therapy.

The incorporation of bortezomib either within the original R-Hyper-CVAD regimen (90% CR)²⁸  or within a less intensive variation (VcR-CVAD [Velcade, rituximab, cyclophosphamide, vincristine, doxorubicin, and dexamethasone])²⁹  (77% CR/CR uncertain) looks highly promising without significantly increasing toxicity. An extension of the initial VcR-CVAD study³⁰  included 75 patients offered a choice of consolidation between an autograft and rituximab maintenance. Although not a randomized trial, there was no obvious difference between these 2 approaches, questioning the value of an autograft following such an intensive induction and vindicating the original protocol. The use of bortezomib has also been explored in the context of maintenance therapy following frontline treatment. A recent phase 2 study added bortezomib to standard R-CHOP therapy followed by 3 monthly maintenance blocks for 2 years.³¹  The results appear superior to R-CHOP alone, but a recent randomized trial shows no benefit for the addition of bortezomib following autologous transplantation,³²  suggesting that it seems unlikely that there is any real benefit for the use of bortezomib in this setting. For the transplant-ineligible patients, bortezomib has been incorporated into the R-CHOP regimen by substituting it for vincristine (VR-CAP). A large randomized trial has compared these 2 regimens as frontline therapy for older patients.³³  This demonstrated a significant benefit in favor of VR-CAP over R-CHOP with respect to CR (53% vs 42%) and PFS (24.7 vs 14.4 months); however, this was coupled with more hematologic toxicity and no significant OS benefit as yet. Unfortunately, neither arm included rituximab maintenance, which confuses the interpretation of the observed outcomes.

Lenalidomide when given together with rituximab is highly active in the setting of relapsed MCL, and a recent study evaluated this regimen as frontline therapy in this disease.³⁴  Of the 38 patients in this trial, the median age was 65 years, but a third were <60 years of age. The overall response rate of evaluable patients was 92% (CR 64%), and the median PFS has yet to be reached with a median follow-up of 30 months. Although highly active, this “chemotherapy-free” combination was associated with 50% grade 3/4 hematologic toxicity. In this trial, patients received an induction phase followed by maintenance with the 2 agents. The role of lenalidomide as a maintenance agent following induction chemotherapy or autologous transplantation is the subject of 2 ongoing large randomized trials (ClinicalTrials.gov number, #NCT01415752, EudraCT no. 2009-012807-25). The addition of bendamustine to lenalidomide and rituximab as part of frontline therapy is highly active (78% CR) but is associated with unacceptable toxicity.³⁵  Clearly, lenalidomide is a highly active agent in combination in MCL, but where this agent should be placed in the treatment algorithm is not yet clear.

Ibrutinib is the most active single agent in the context of relapsed MCL. The initial phase 2 study demonstrated a 70% response rate in heavily pretreated patients irrespective of baseline risk factors,³⁶  but with an average PFS of ∼14 months. Ibrutinib has a very modest side effect profile in comparison with the other novel agents, which lends itself to potential combination therapies. The addition of rituximab to ibrutinib significantly increases remission rates and almost doubles complete remission rates³⁷  without a significant change in the side effect profile. A large randomized trial adding ibrutinib to bendamustine and rituximab as part of frontline therapy (#NCT01776840) has completed recruitment and may set a new benchmark for this disease.

In older patients where the toxicity of chemotherapy can be significant and quality of life is of prime importance, it seems logical to consider Bruton tyrosine kinase (BTK) inhibition as part of frontline therapy. However, in younger patients, the current frontline approaches produce very durable long-term outcomes in significant numbers of patients. There are no published data on frontline therapy in MCL, but the ongoing trials are predominantly in older patients. As BTK inhibition is so active, it will be incorporated within trials to attempt to augment chemotherapy outcomes, but it also has the potential to fundamentally challenge existing paradigms. The European Mantle Cell Network has recently commenced a large trial that incorporates ibrutinib into the R-DHAP/R-CHOP + autograft regimen (EudraCT 2014-001363-12; the Triangle study). This trial has 3 arms: a control arm without ibrutinib, one with it given as part of the entire regimen, and the interesting third arm randomizes to the patient not receiving the autograft.

There are a host of other agents with activity in MCL³⁸  and a lot of excitement around the potential application of chimeric antigen receptor T-cell therapy.³⁹  Although review of these is beyond the scope of this article, probably the most promising agent beyond ibrutinib is the B-cell lymphoma 2 inhibitor Navitoclax (ABT-199/GDC-0199).⁴⁰  Response rates and the durability of remissions are on a par with those seen with BTK inhibition, albeit with very few patients having been treated. As the mechanisms of action of both ibrutinib and navitoclax are completely different and they are extremely well-tolerated oral agents, there is a rationale to combining them, and in the context of MCL, there are currently 3 trials exploring this. Some early encouraging results suggest that this may be the first novel combination that could be seen in the frontline setting.

For young patients with MCL who require therapy, the optimal treatment regimen currently includes high-dose cytarabine with rituximab and consolidation with an autologous stem cell transplant probably improves outcomes following that. There is no clear optimal regimen in this setting, and the randomized data in support of autologous transplantation were in the pre-cytarabine era. Bortezomib may improve outcomes when incorporated into cytarabine regimens, but rituximab maintenance after autograft significantly improves PFS and should be adopted. Allogeneic transplantation has a limited role in MCL, but in relapsed disease it can lead to long-term disease-free survival and perhaps in some very high-risk young patients could be used as first-line consolidation. There is a lot of excitement around the novel agents that have activity in this disease, and in the context of relapse, ibrutinib is probably the treatment of choice. This drug is the subject of trials in the front-line setting, but it is important to remember that we have long-term outcome data with conventional therapy, and we should not be quick to deviate from that before mature data to suggest better outcomes exists with alternative approaches.

Simon Rule, Department of Hematology, Derriford Hospital, Plymouth PL6 8DH, United Kingdom; e-mail: simon.rule@nhs.net.