Is rituximab retro in mantle cell lymphoma?

In this issue of Blood, Sarkozy et al report long-term follow-up of the LYMA-101 clinical trial and present results from an indirect comparison with the LYMA study to interrogate obinutuzumab (O)- vs rituximab (R)-based frontline management of transplant-eligible patients with mantle cell lymphoma (MCL).^1,2 Patients in both trials received 4 cycles of anti-CD20 antibody in combination with dexamethasone, high-dose cytarabine, and a platinum agent (O/R-DHAP) followed by autologous stem cell transplantation (ASCT) in first remission and anti-CD20 antibody maintenance. The indirect comparison results using a propensity score matching (PSM) method demonstrated superior outcomes with obinutuzumab (LYMA-101) compared with rituximab (LYMA), with higher rates of minimal residual disease (MRD) negativity at end of induction (EOI) as well as longer 5-year progression-free survival (PFS) and overall survival (OS). Given the largely incurable nature of MCL, the durability of first remission is key to improved long-term outcomes in this disease; if obinutuzumab can achieve superior outcomes to rituximab in the frontline setting, these results raise important questions regarding whether obinutuzumab should replace rituximab as the anti-CD20 antibody of choice in MCL management. As the authors highlight, it is unlikely that a randomized controlled trial of R vs O in MCL will occur, and thus the field must decide how best to incorporate the results of this indirect comparison study.

In this study, the authors used a PSM method in an attempt to account for characteristics that may otherwise be confounding variables. The authors included a fairly comprehensive set of variables in their analysis; however, as they note, there are some important missing parameters, including Ki-67 and TP53 aberrancy status, both of which have been demonstrated to confer higher-risk disease.³ Furthermore, there is increasing recognition that additional molecular characteristics correlate with disease outcomes, and although it is understandable that such information was not available to include in the PSM method for these studies, the potential imbalance of these missing variables is thus a limitation of the PSM method. Another limitation is the extent of missing MRD results for patients in the LYMA trial (MRD was a primary end point in the LYMA-101 trial and thus more data were collected). Of those patients included in the PSM method, nearly 85% of LYMA-101 responders had available bone marrow MRD data compared with only ∼60% of LYMA responders. Despite these limitations, this PSM-based indirect comparison is likely the closest the field may come to a quasi-randomized comparison, and thus it is our opinion that the overall trend of the results should be taken under strong consideration.

The data do strongly suggest a significant benefit of obinutuzumab compared with rituximab with higher MRD negativity (83% vs 63%) and improved long-term survival. The 5-year PFS and OS were 83% vs 67% and 86% vs 71% in favor of obinutuzumab, respectively, representing a large potential survival benefit. Placing these results in the context of previously published data, we find a consistent pattern across other randomized comparisons of obinutuzumab vs rituximab in other indolent B-cell non-Hodgkin lymphomas (see table). In the GALLIUM trial, which randomized previously untreated patients with follicular lymphoma to obinutuzumab- vs rituximab-based chemoimmunotherapy,⁴ and in the chronic lymphocytic leukemia (CLL)11⁵ and CLL13 trials,⁶ which randomized previously untreated patients with chronic lymphocytic leukemia to obinutuzumab- vs rituximab-based combination therapy, the PFS rates all favored obinutuzumab. In contrast, in the GOYA trial in diffuse large B-cell lymphoma, obinutuzumab did not demonstrate superiority when combined with a standard-of-care chemotherapy backbone in place of rituximab.⁷ Interestingly, the PSM results did not demonstrate superiority of obinutuzumab vs rituximab maintenance for those patients who were MRD negative at EOI. However, the sample size in each group was small (n = 43), and it is possible such patients represent those with the most treatment-sensitive disease, thereby making it more challenging to distinguish a differential effect between the 2 antibodies.

If we accept the results as potentially actionable, the question then arises regarding how to implement such a finding in the current MCL treatment landscape, and toxicity considerations become key. Notably, in these LYMA comparisons, the authors do not report significantly increased toxicity with obinutuzumab vs rituximab maintenance, but as noted, the sample size in the 2 PSM groups was small. There is a consistent trend across other studies comparing obinutuzumab- vs rituximab-based regimens toward higher toxicity with obinutuzumab (see table). For the other common frontline MCL regimens of alternating R-cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP)/R-DHAP, bendamustine/rituximab (BR), and BR alternating with R-cytarabine, it is challenging to simply substitute obinutuzumab for rituximab given the paucity of toxicity data of obinutuzumab with those other chemotherapy partners in MCL. In particular, when considering bendamustine-based regimens, results from the GALLIUM trial are cautionary, as a higher rate of grade ≥3 infections was seen with bendamustine/obinutuzumab compared with bendamustine/rituximab,⁴ particularly in the maintenance phase, which is of key consideration in MCL. The blanket implementation of obinutuzumab in frontline traditional MCL chemo-immunotherapy induction at present is therefore limited. Practically, a major barrier to widespread applicability of the results of this study will be insurance reimbursement. Such a change in the United States would require inclusion in the National Comprehensive Cancer Network guidelines, and in Europe, authorization from the European Medicines Agency would be needed.

Currently, a significant paradigm shift is underway in frontline MCL, with the likely incorporation of targeted agents into induction regimens and the omission of ASCT in first remission based on the TRIANGLE study.⁸ In our opinion, this study provides the rationale for incorporation, when possible, of obinutuzumab as the preferred anti-CD20 antibody into novel treatment regimens and future studies where toxicity can also be assessed thoroughly. Results from both the OAsIs study⁹ and the BOVen study,¹⁰ which investigate obinutuzumab in combination with a Bruton tyrosine kinase inhibitor and venetoclax as a frontline approach in MCL, have demonstrated highly promising efficacy and tolerability (and a similar triplet study, acalabrutinib, venetoclax, and obinutuzumab, is ongoing [NCT04855695]). These studies and future similar ones would lay the foundation for subsequent inclusion of obinutuzumab in advanced phase clinical trials that could then potentially establish a new obinutuzumab-containing standard of care.

Conflict-of-interest disclosure: C.E.R. has received honoraria from AstraZeneca Brazil and Research To Practice. J.B.C. has served as an advisor for HutchMed, Loxo/Lilly, ADCT, Kite/Gilead, Astra Zeneca, and Beigene; and has received research funding from Takeda, Novartis, Loxo/Lilly, Genentech, Astra Zeneca, and Bristol Myers Squibb/Celgene.