Nivolumab for Relapsed/Refractory Diffuse Large B-Cell Lymphoma in Patients Ineligible for or Having Failed Autologous Transplantation

A Single-Arm, Phase II Study

Stephen M. Ansell, MD, PhD; Monique C. Minnema, MD, PhD; Peter Johnson, MD; John M. Timmerman, MD; Philippe Armand, MD, PhD; Margaret A. Shipp, MD; Scott J. Rodig, MD, PhD; Azra H. Ligon, PhD; Margaretha G.M. Roemer, PhD; Nishitha Reddy, MD; Jonathon B. Cohen, MD; Sarit Assouline, MD; Michelle Poon, MD; Manish Sharma, MD; Kazunobu Kato, MD, PhD; Selda Samakoglu, MD, PhD; Anne Sumbul, MS; Andrew Grigg, MD


J Clin Oncol. 2019;37(6):481-489. 

In This Article


In this phase II study, nivolumab monotherapy was associated with a low incidence of objective response in patients with relapsed/refractory DLBCL, and the study thus did not meet the primary end point. No new safety concerns were identified compared with previous studies that involved patients with solid tumors, HL, and non-HL, acknowledging that most patients received a limited number of doses.[19,21,27]

ORR was lower in the current phase II study than in the phase I study of patients with DLBCL who had received nivolumab monotherapy.21 This result may be attributed, in part, to differences in study design and patient population. We observed a higher response rate by investigator assessment than by IRC assessment in the phase II study population (Table 2). Centralized IRC assessment on the basis of imaging scans offers standardized and more objective evaluation. In addition, the number of patients with relapsed/refractory DLBCL (n = 11) was low in the phase I study, leading to wide CIs around observed response rates.[21]

Although ORR was low in this study, three patients who achieved CR had DORs of 17 months, 11 or more months, and 14 or more months; none of them had experienced progression at the data cutoff. Two were still alive at 30 and 38 months from first dose with their most recent survival status; however, the number of responders was too low to permit analysis of OS in this group. Whereas PD-L1/PD-L2 amplification in malignant cells was observed for one of these patients, the biologic basis for response in the other two patients is unclear.

Median number of nivolumab doses received (four in the auto-HCT–failed group and three in the auto-HCT–ineligible group) by patients with DLBCL was much lower than those received (16 and 17 doses) in studies that involved patients with relapsed/refractory HL,[19,20] as the majority of the patients with DLBCL stopped treatment because of disease progression. Although it is possible that the rapid progression of DLBCL did not allow enough time for nivolumab to show an effect, we did not observe clear improvement of responses in the limited number of patients who had been treated beyond disease progression.

The study protocol included a prespecified stopping rule for futility but, on the basis of the safety profile of nivolumab and promising efficacy results observed in the phase I study, allowed enrollment to continue while the interim analysis result was pending. As a result of rapid accrual and delays related to IRC, the planned enrollment was reached before the futility analysis was concluded. Nevertheless, the majority of patients received only a short duration of nivolumab treatment and were allowed to switch to other cancer therapies upon discontinuation of the study treatment; there was no new safety signal observed. In the future, we shall ensure that study procedures are closely monitored to avoid potential delays in the timely completion of the interim analysis.

Previously reported data suggest that PD-L1 expression is associated with poor prognosis in patients with DLBCL;[8] however, the prevalence of PD-L1 expression on DLBCL tumor cells seemed to be low (11%; with a 30% threshold) in that series,[8] as well as in a EBV-negative DLBCL series (11%; with a 5% threshold).[10] The 9p24.1 FISH analyses on a large number of DLBCL trial cases in this study indicate that the incidence and magnitude of 9p24.1 alterations are significantly lower in DLBCL than in cHL, as is the level of PD-L1 expression[20,23] (H-score; Appendix Figure A2, online only). These observations are consistent with the observed differences in the efficacy of PD-1 blockade in DLBCL and cHL and provide a possible explanation for why only a small proportion of patients with relapsed/refractory DLBCL benefit from nivolumab monotherapy. As a result of the low number of responders, there were insufficient data to accurately assess double-hit or c-Myc association with responses in this study.

Figure A2.

Comparison of the incidence and magnitude of 9p24.1 alterations in diffuse large B-cell lymphoma (DLBCL; this study) and classic Hodgkin lymphoma (cHL).23 (A) Prevalence and (B) spectrum of 9p24.1 alterations in DLBCL. (C) Percentage of tumor cells with residual 9p24.1 disomy in DLBCLs classified by 9p24.1 genetic categories. (D) Prevalence and (E) spectrum of 9p24.1 alterations in cHL.23 (F) Percentage of tumor cells with residual 9p24.1 disomy in cHLs classified by 9p24.1 genetic categories.23 As noted, DLBCLs have a much lower incidence of 9p24.1 copy gain and amplification than cHLs. In addition, DLBCLs with low-level 9p24.1 alterations have a higher percentage of residual disomy than cHLs. Ampl, amplification; Coampl, coamplification; PD-L2, programmed death ligand 2; Rearr, rearrangement. [Reprinted with permission. © 2018 American Society of Clinical Oncology.]

Whereas the single-agent activity of PD-1 blockade was low in this unselected series of patients with DLBCL, it may be useful to evaluate the approach in select LBCL subtypes with increased PD-L1 expression, such as primary mediastinal B-cell lymphoma, EBV-positive LBCL, and T-cell/histiocyte-rich LBCL.[10] Recent reports suggest that combination treatment strategies, such as chimeric antigen receptor T cells and nivolumab[28] and PD-1 inhibition with concomitant radiotherapy,[29] might also be considered.

In conclusion, the results of this phase II study indicate that nivolumab monotherapy has a low response rate but may provide benefit in a small number of patients with DLBCL who have experienced failure with auto-HCT or who are ineligible for auto-HCT and is associated with a favorable safety profile. Additional research on the influence of biologic factors, including the tumor microenvironment, is warranted to characterize whether a subset of patients with DLBCL may be more likely to respond to nivolumab-based treatment, if combination therapy can improve the therapeutic activity of PD-1 blockade, or if earlier use of these agents before intensive immunoablative chemotherapy may be more effective.