The potential for chemotherapy-free strategies in mantle cell lymphoma

The goal of therapy in mantle cell lymphoma (MCL) is to extend overall survival (OS) with as few lymphoma-related symptoms and treatment-related side effects as possible. Historically, this has been accomplished through the sequential administration of chemotherapy or immunochemotherapy regimens, with drugs and doses individualized to best suit the specific scenario. Fortunately, immunochemotherapy is active in most patients. Unfortunately, immunochemotherapy is rarely if ever curative, and highly proliferative or TP53-mutated lymphomas tend to respond poorly regardless of treatment intensity.^1,2  Moreover, chemotherapy comes with well-documented toxicities that can limit its use and detract from quality of life. Over the past decade, 4 nonchemotherapy options (bortezomib, temsirolimus, ibrutinib, lenalidomide; Table 1) have been approved by the US Food and Drug Administration (FDA) and/or the European Medicines Agency (EMA) for treatment of MCL, and several others are in development. In some cases, these nonchemotherapy drugs appear to be active where traditional cytotoxic chemotherapy might fail, and in many cases the side-effect profile is distinct from chemotherapy. Despite the failure of any of these drugs to improve OS in randomized trials, their introduction corresponds with better outcomes in population studies^3-5  suggesting that the availability of new options that can be used sequentially may be extending survival. Increasingly, clinicians, patients, and other interested parties are raising the possibility of a future without chemotherapy.

There is clearly value in more effective, better-tolerated treatments for people with MCL; whether these new treatments are chemotherapy or not is less important than whether they work. Indeed, many experts take issue with the term “nonchemotherapy,” questioning its provenance and, no doubt, concerned that it could be construed as a gimmick promoted by those looking to advance the use of new drugs to a group of patients anxious about their diagnosis and eager to believe that progress is at hand. “Nonchemotherapy” often implies, rightly or wrongly, “more effective” and “less toxic,” generally with a mechanism of action that preferentially targets tumor cells. Although there are clearly issues with the term, it is broadly understood because it simply describes what it is not: traditional cytotoxic chemotherapy. With roughly 20 years of experience with rituximab, 10 years since the approval of bortezomib and temsirolimus, and roughly 5 years of experience (including clinical trials) with lenalidomide and ibrutinib (Table 1), we are already in the nonchemotherapy era, and there are data that allow us to evaluate whether the sentiments attached to nonchemotherapeutic approaches are true.

Without question, chemotherapy is active in MCL. Administered appropriately, combinations including high-dose cytarabine and autologous stem cell transplantation can result in remissions nearly a decade long in some younger patients.⁶  Less-intensive regimens have also been reported to produce durable remissions, with emerging evidence that patient-related factors and disease biology trump choice of therapy in impacting outcomes.^1,7,8  Moreover, chemotherapy may have some desirable effects beyond simple cytotoxicity: cyclophosphamide may inhibit regulatory T cells (Tregs),⁹  doxorubicin may impair myeloid-derived suppressor cells,¹⁰  and low-dose metronomic therapy may impact tumor angiogenesis.¹¹  Nonetheless, the most effective immunochemotherapy regimens are restricted to young, healthy patients who account for <25% of all patients.³  Cytotoxic chemotherapy is associated with well-documented toxicities. In addition to usually modest nuisance-type side effects that can sometimes be more significant, such as alopecia, nausea, and fatigue, chemotherapy can cause significant hematological toxicities, resulting in neutropenic fever and other opportunistic infections. Rarely, it can be associated with cardiac toxicity, renal toxicity, and peripheral nerve or central nervous system toxicity. Long-term side effects resulting from myelotoxicity include secondary malignancies and, more relevantly, impaired bone marrow function that can limit the ability to tolerate subsequent treatments. These factors likely contribute to the discordance between outcomes described in an optimized clinical trial population, which have improved significantly over the past decade, and real-world outcomes, which remain disappointing.³ 

Multiagent chemotherapy regimens are standard treatment of most cancers and are based on the concept that the combination of drugs with distinct mechanisms of action and toxicity profiles should be more able to eradicate a broader spectrum of subclones that might result in treatment resistance and disease recurrence. It is no surprise that incorporation of novel, nonchemotherapy drugs into standard chemotherapy regimens might be the most natural first step. Table 2 describes several trials where nonchemotherapy agents have been used in front-line treatment of MCL. Shortly after bortezomib was approved as a single agent for previously treated MCL, it was already being combined with rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP) in patients with diffuse large B-cell lymphoma and MCL, and results appeared worthy of further evaluation.^12,13  In a subsequent phase 3 trial, substitution of vincristine in R-CHOP for bortezomib (VR-CAP) resulted in a significant improvement in progression-free survival (PFS; 24.7 vs 14.4 months; hazard ratio [HR], 0.63; P < .001) and subsequent approval of the drug in the setting.¹⁴ 

The bendamustine-rituximab (BR) regimen has served as a backbone for multiple chemotherapy/nonchemotherapy combinations. The addition of bortezomib to BR is currently being evaluated in the randomized phase 2 US Intergroup E1411 trial (NCT01415752), which completed accrual in 2016. The Nordic MCL4 trial combined BR with lenalidomide (Lena-Berit).¹⁵  After early experience with the regimen resulted in unexpectedly high rates of cutaneous reactions and serious infections, the doses and timing of bendamustine and lenalidomide were modified, bringing the regimen in line with what might be clinically acceptable. The finding that 2 treatments that are relatively well tolerated when given separately (ie, BR and lenalidomide in this case) result in unexpected toxicities when combined is observed repeatedly and highlights the importance of prospective clinical trials. In a small study of patients with MCL, the combination of ibrutinib and BR demonstrated not only promising activity but also a surprisingly high rate of cutaneous reactions.¹⁶  The combination is being further evaluated in the double-blind, placebo-controlled, randomized phase 3 SHINE trial (NCT01776840), with PFS as the primary end point.

To date, there are no published data describing the combination of intensive front-line regimens with nonchemotherapy agents. The ongoing European MCL Network Triangle (NCT02858258) trial is evaluating the addition of ibrutinib to a high-dose cytarabine-based immunochemotherapy induction, and we are likely to see similar trials in the future.

Critical to the interpretation of chemotherapy/nonchemotherapy trials will be the assessment of whether the PFS benefit justifies the added toxicity if there is the absence of an OS benefit. Part of that question includes an estimate of whether the PFS benefit would be similar if the chemotherapy and nonchemotherapy regimens were given sequentially rather than concurrently. For example, if the addition of ibrutinib to BR results in a PFS benefit that is similar to what might be expected from ibrutinib administered as second-line therapy following BR failure (likely 18 months or more based on pooled data from 3 trials),¹⁷  it would likely be less toxic and less expensive to administer the drugs sequentially. Despite these issues, we are likely to see adoption of these combination regimens in the near future and should move quickly to try to understand which patients are most likely to benefit from this approach.

In the absence of true synergy arising from the combination of chemotherapy and nonchemotherapy, it may be rational to replace the chemotherapy component altogether. For example, a common criticism of the HELIOS trial,¹⁸  which evaluated BR with or without ibrutinib in patients with previously untreated chronic lymphocytic leukemia (CLL), is the question of whether patients who were randomized to the BR plus ibrutinib arm would have done equally well had they received ibrutinib without BR. Similarly, there may be specific scenarios where chemotherapy is unlikely to be effective and a nonchemotherapy approach might have a higher probability of benefit. Again, experience with CLL may provide clues to scenarios in MCL where standard chemotherapy should be avoided. Patients with CLL and a 17p deletion/TP53 mutation have poor outcomes with chemotherapy and appear to do significantly better when treated with ibrutinib alone in the front-line setting.

Patients with MCL and TP53 mutations have poor outcomes. In the Nordic MCL2 trial, which evaluated intensive induction therapy followed by autologous stem cell transplantation, patients with a TP53 mutation had a median PFS of 1 year compared with 9.9 years in patients without a TP53 mutation (P < .0001).¹⁹  Similar results were seen in an Italian trial using a similar treatment regimen.²⁰  MCL, however, is not CLL, and the answer is likely to be considerably more nuanced. Other variables, including blastoid morphology, high Ki67, KMT2D mutations, a complex karyotype, among other high-risk features have also been associated with poor outcomes,^1,2,20,21  and whether nonchemotherapy agents can overcome some, all, or any of these adverse factors is unclear. In previously treated MCL, the combination of ibrutinib, lenalidomide, and rituximab did not appear to be adversely impacted by the presence of a TP53 mutation.²²  On the other hand, blastoid MCL has been associated with worse outcomes among patients treated with single-agent ibrutinib or temsirolimus.¹⁷  Whether specific combinations can control these highly proliferative tumors remains to be seen. It is likely that some patients with MCL might do well regardless of the choice of chemotherapy or nonchemotherapy, some patients might do better receiving nonchemotherapy alone, and some patients will continue to have poor outcomes despite the availability of new, nonchemotherapy agents. Without more informative data, omitting chemotherapy altogether in very-high risk, highly proliferative MCL patients cannot be recommended. There may, however, be a role of nonchemotherapy regimens in less proliferative tumors with TP53 mutations.

Based on the promising efficacy/toxicity profile of ibrutinib plus rituximab in patients with previously treated MCL, the combination is being tested in the front-line setting prior to an abbreviated course of intensive chemotherapy (NCT02427620). Preliminary results suggest a high complete response rate (72%), although the addition of chemotherapy makes durability difficult to assess.²³  The randomized, phase 3 ENRICH trial will compare ibrutinib plus rituximab to rituximab plus chemotherapy as front-line therapy in older patients with MCL (CRUK/14/026). The most mature data with a nonchemotherapy regimen in the front-line setting comes from a multicenter phase 2 trial of rituximab-lenalidomide.²⁴  The treatment, which continued until progression or unacceptable toxicity, resulted in an overall response rate of 87%, including 61% complete responses, and a 2-year PFS of 85%, and results continue to appear promising beyond 5 years (J.R., unpublished data). Interestingly, neither the MCL International Prognostic Index (MIPI) score nor Ki67 had an impact on PFS, although only 21% of patients had a Ki67 of >30% at baseline and the study was not powered to detect a difference in this unplanned subgroup analysis. At this point, it is unclear whether novel regimens might have activity in patients expected to do poorly with chemotherapy; future trials with nonchemotherapy agents should pay particular attention to these patient subgroups.

Other interesting data to arise from the lenalidomide-rituximab study include an adverse event profile that is distinct from what might typically be seen with chemotherapy. Although 50% of patients experienced grade 3-4 neutropenia at any time during treatment, only 5% of patients experienced febrile neutropenia. However, low-grade infections were relatively common albeit over a long and frequent observation period, including 39% of patients with grade 1-2 upper respiratory tract infections. Unlike the relapsed/refractory setting, there was a high rate of adverse events that may have been immunologically mediated, including tumor flare in 34% and rash in 66%. These observations should be taken seriously. “Nonchemotherapy” is not synonymous with “no side effects.” Long-term exposure to lenalidomide has been associated with increased risk of secondary cancers.^25,26  Ibrutinib, which has limited long-term follow-up thus far, has been associated with opportunistic infections, bleeding, and atrial fibrillation.^27-29  In phase 2 and 3 trials with ibrutinib in patients with MCL and CLL, roughly 10% of patients stopped ibrutinib because of adverse events.^30,31  Unlike chemotherapy, which is administered for a finite period of time followed by a treatment-free interval, nonchemotherapy drugs are frequently given until progression or unacceptable toxicity, and the duration is likely to increase with earlier use. New methodologies for reporting the chronicity as well as the severity of adverse events will be needed.³² 

Currently, the primary approval and most common use of all nonchemotherapy agents in MCL is in patients who have received prior therapy (Table 1). Both temsirolimus and lenalidomide were superior to investigators’ choice (primarily chemotherapy drugs that would not commonly be used as single agents in standard practice) in heavily pretreated patients.^33,34 Table 3 describes several recent and ongoing studies evaluating nonchemotherapy agents in the previously treated setting. The addition of temsirolimus to bendamustine and rituximab resulted in an overall response rate of 88%, including 39% complete responses among 29 patients with previously treated MCL.³⁵  The addition of lenalidomide to bendamustine and rituximab (the R2B regimen) yielded a median PFS of 20 months.³⁶  In the absence of randomized data, it is difficult to know whether the apparent improvement in response outweighs the apparent increase in toxicity of these regimens. Interestingly, the rate of cutaneous reactions with the combination of lenalidomide and bendamustine appears to be significantly lower in the relapsed setting compared with the front-line setting, suggesting that context is as important as the treatment regimen itself.

With the possible exception of bendamustine (or bendamustine-based combinations), ibrutinib appears to have the highest response rate of all available options; it was significantly more effective and better tolerated than temsirolimus in a randomized phase 3 trial.³⁷  Most patients currently receive bendamustine plus rituximab as front-line therapy in the United States, leaving ibrutinib as a de facto standard therapy in the relapsed/refractory setting. Despite its unprecedented single-agent activity, roughly one-third of patients do not respond to ibrutinib, and relapse appears to be universal, usually within the first 12 to 18 months. Depth of response and response duration can likely be improved by combining ibrutinib with rituximab, a regimen that yielded a 44% complete response rate with an apparent improvement in response duration in a single-center phase 2 trial.³⁸  As is the case with chemotherapy/nonchemotherapy combinations, rational combinations of nonchemotherapy drugs with the potential for synergy, not just additive benefit, are particularly interesting, and preclinical data for several such ibrutinib-based regimens have been reported.^39,40  Preliminary data from combination studies with lenalidomide/rituximab (the Nordic MCL6 PHILEMON study),²²  the bcl2 inhibitor venetoclax,^41,42  and the cdk4/6 inhibitor palbociclib⁴³  all suggest that combinations are feasible. Among 50 patients treated on the PHILEMON study, 83% responded, including 41% with a complete response. Unlike the front-line setting in follicular lymphoma, where the same regimen resulted in grade 3-4 rash in 36%, only 10% of previously treated MCL patients developed grade 3 rash, consistent with the hypothesis that immune-mediated effects are more common in the front-line setting. Among 8 patients treated in the ibrutinib plus venetoclax (AIM) trial, 5 achieved a complete response with eradication of minimal residual disease and only 1 did not respond; promising responses have been observed in a second study of the same regimen.⁴¹  The OAsIs trial will evaluate the combination of ibrutinib, venetoclax, and obinutuzumab (NCT02558816). The combination of ibrutinib and palbociclib has been evaluated in 23 patients so far in a multicenter phase 1 trial.⁴³  Grade 3 rash was experienced in 2 patients at the highest dose level and was considered dose limiting. The overall response rate was 64%, including 43% complete responses, and the estimated response duration at 1 year was 90%; that is, only 1 of 13 responding patients experienced progression. Notably, although these combinations have produced impressive results, not 1 appears to completely overcome ibrutinib resistance and none have reported identification of a biomarker to help target patients to the right therapy. Additionally, with small numbers, it is possible that patient selection is partially responsible for the apparent improvement. More work will be required to determine which patients are best treated by single-agent ibrutinib, a specific combination, or a nonibrutinib approach. These and other studies set the stage for a promising future, but it is worth noting that development of novel combinations may be negatively impacted by the widespread reliance on single-agent ibrutinib and the possible adoption of nonrational ibrutinib plus chemotherapy combinations in the front-line setting. Ideally, rational combinations will be based on cutting-edge science produced by laboratories around the world, but it is naive to believe that the complexity of MCL could be completely recreated in a laboratory environment. There is an urgent need for small, rapid combination studies with carefully designed, longitudinal laboratory correlates.

As ibrutinib and ibrutinib-based combinations become more widely used, an important emerging question is how best to treat patients who experience ibrutinib failure. Retrospectively collected data from a group of high-risk patients who had progressed while receiving ibrutinib suggested that outcomes were poor, although it was unclear whether that was due to the fact that most patients were heavily pretreated and tended to have short responses to ibrutinib, or whether ibrutinib might somehow promote the growth of more aggressive subclones.⁴⁴  More importantly, the study failed to identify a specific treatment that might be most likely to produce responses following ibrutinib failure. A small, retrospective series evaluated responses to lenalidomide and lenalidomide-based combinations following ibrutinib failure and suggested that there was activity, albeit of short duration, in a sizeable minority of patients.²³  As more effective ibrutinib combinations are developed, it is likely that patients who experience treatment failure will have even more aggressive disease courses, and the question of how to treat these patients will become more acute.

Allotransplant has been reported to cure roughly 30% of patients with previously treated MCL in retrospective series.^45-47  Despite its curative potential, allotransplant relies on donor availability, fails in most cases, and is associated with significant treatment-related morbidity and mortality, limiting its use to patients without significant comorbidities. Allotransplant should be considered in patients felt to be at high risk of treatment failure with few acceptable alternatives. Unfortunately, it is not always easy to identify these patients at a time when allotransplant might be most successful.

Although chimeric antigen receptor (CAR)-T cells have demonstrated activity in MCL,^48,49  their future is still relatively unknown. In theory, MCL is the ideal setting for the implementation of CAR-T-cell therapies: it is incurable with a relatively short survival compared with other incurable lymphomas, but remissions of sufficient depth and duration can be obtained in most patients, allowing for manufacturing and administration of CAR-T cells. Possible exceptions include patients with chemorefractory MCL who experience explosive progression on ibrutinib, but hopefully this will change as we get better at predicting ibrutinib resistance and as we develop strategies to cope, at least transiently, with ibrutinib failure. That said, we already have an effective cellular/immunotherapy in MCL: allogeneic stem cell transplantation. In order for CAR-T cells to be more widely adopted in MCL, they will have to demonstrate either an ability to safely induce responses in patients who would not be eligible for allotransplant due to comorbid conditions or rapidly progressing disease, or they will need to be associated with a relatively superior efficacy and tolerability profile, that is, inducing durable remissions in at least a third of patients based on an intent-to-treat analysis with a treatment-related mortality of <15% to 20%. Additionally, the toxicities associated with CAR-T cells at present, including neurologic issues related to cytokine release syndrome, may limit applicability to MCL patients who are frequently older and have comorbid conditions.

Existing data with immune checkpoint inhibitors in MCL are limited and unimpressive to date. In published trials that included MCL, there have been no reported responses.⁵⁰  Reasons for resistance are likely multiple but may include limited expression of programmed death ligand 1 on MCL cells or tumor-derived immunosuppression. Nonetheless, given the remarkable promise these therapies have demonstrated in other cancers, it would be unfortunate to abandon further investigation at this early stage.

Vaccines have long been held out as a promising, if somewhat fantastic, strategy in B-cell lymphomas given the existence of a tumor-specific idiotype, but phase 3 trials in follicular lymphoma were mostly disappointing.^51-53  Recently, T cells capable of recognizing tumor peptides derived from the idiotype and presented on the tumor cell major histocompatibility complex have been described, sparking renewed interest in vaccines.⁵⁴  Perhaps the development of immune checkpoint inhibitors will prove serendipitous to the new generation of vaccine scientists.

Bendamustine or cytarabine-based therapies are clearly active in MCL. At a time when the novelty of chemotherapy has eroded, several recently published clinical trials have demonstrated the surprising activity of chemotherapy-based regimens, and we should not be too eager to discount these “old-fashioned” treatments. Unfortunately, no chemotherapy regimens have produced cures and the contributions of novel chemotherapeutic agents to improving survival rates is unclear. Thus far, no chemotherapy regimens have demonstrated any sign of overcoming the risks associated with highly proliferative and TP53-mutated/deleted tumors, and all chemotherapy regimens are associated with significant acute and cumulative toxicities.

The addition of nonchemotherapy to standard chemotherapy regimens has already demonstrated an ability to improve outcomes, and more studies are ongoing. Moreover, early results from trials testing regimens that consist of nonchemotherapy agents alone have suggested that the strategy may have some advantages. Patients with previously treated MCL are already standardly receiving nonchemotherapy. Despite the apparent progress of nonchemotherapy-based treatments, several challenges remain. Nonchemotherapy agents have distinctive toxicities, particularly when combined with chemotherapy or other agents with novel mechanisms of action. The duration of therapy, and therefore duration of toxicity, is clearly a factor that may impact quality of life. Importantly, the price of new drugs has real-world implications with respect to accessibility and sustainability. It remains unclear whether existing nonchemotherapy agents can overcome the highly proliferative, TP53-mutated lymphomas that are most in need of new drugs. Finally, patients who are refractory to novel agents like ibrutinib appear to have especially poor outcomes. Rational combinations are poised to improve duration of response to novel agents but it remains to be seen whether the differences will be synergistic vs additive and whether the benefits will outweigh the risks. CAR-T cells are attractive but may not be feasible in patients whose disease cannot be stabilized for a sufficient duration of time to produce and administer the treatment.

Currently, our approach is to prioritize clinical trials with nonchemotherapy agents and embedded correlative studies that include putative biomarkers for risk stratification. Existing data are either too limited or unconvincing to make recommendations regarding standard practice. Outside of clinical trials and the VR-CAP regimen (which we do not use), current data do not support the use of nonchemotherapy agents (other than rituximab) either alone or in combination with chemotherapy in the front-line setting, except for a small number of patients with TP53-mutated, leukemic, nonnodal MCL who we have treated with lenalidomide-rituximab. We expect that the evolving appreciation for the heterogeneous biology of MCL and the mechanisms of action of various novel agents are critical to the interpretation of ongoing trials and the design of future studies that incorporate nonchemotherapy agents. Our goal should be to cure MCL with treatments that are broadly available. To achieve that aim, we will need rational combinations of new drugs, regardless of how we call them.

Contribution: P.M., J.R., and J.P.L. coauthored the paper.

Conflict-of-interest disclosure: P.M. has consulted for Janssen, Pharmacyclics, AstraZeneca, Celgene, Gilead, Bayer, Novartis, KiTE, and Verastem. J.R. has consulted for Celgene. J.P.L. has consulted for Celgene, Genentech, Gilead, and Bayer.

Correspondence: Peter Martin, Weill Cornell Medicine, 525 East 68th St, New York, NY 10065; e-mail: pem9019@med.cornell.edu.