Characteristics of relapsed/refractory diffuse large B‐cell lymphoma patients with durable responses to maveropepimut‐S, pembrolizumab, and cyclophosphamide: Long‐term follow‐up from the SPiReL trial

Abstract

Introduction

Patients with relapsed/refractory diffuse large B‐cell lymphoma (R/R DLBCL) have an unmet medical need. The objective of this trial was to assess the efficacy and toxicities of a novel triple immunotherapy regimen—pembrolizumab, low‐dose cyclophosphamide, and maveropepimut‐S (MVP‐S). This regimen was designed to activate tumor‐specific T cells by targeting the tumor‐associated antigen survivin with MVP‐S and reducing two important T cell inhibitory pathways: T cell exhaustion and regulatory T cells with pembrolizumab and metronomic cyclophosphamide, respectively.

Methods

This was a single‐arm Phase II clinical trial in 25 participants with R/R DLBCL‐SPiReL trial (NCT03349450).

Results

The median overall survival was 10.1 months and a third of participants survived over 2 years. Enhanced long‐term survival was associated with favorable clinical characteristics and enhanced immune reactivity, as assessed by ELISpot and ISR‐immune reactive responses. The regimen was well‐tolerated with minimal Grade 3–4 toxicities.

Conclusion

Combination immunotherapy regimens such as this could offer a promising alternative to other treatments with significant toxicities for select patients.

1.

Patients with relapsed/refractory (R/R) diffuse large B‐cell lymphoma (DLBCL) who are ineligible for, or who relapse after, autologous stem cell transplant (ASCT) have a very poor prognosis and are often resistant to standard cytotoxic therapies [1, 2, 3]. In the SCHOLAR‐1 study, the median overall survival of participants with refractory DLBCL was 6.3 months; only 20% of participants survived 2 years [1]. In the KEYNOTE‐13 and CHECKMATE 139 trials, response rates to single‐agent pembrolizumab and nivolumab in R/R DLBCL respectively were reported to be low [4, 5]. In recent years, novel immunotherapies including bispecific antibodies and chimeric antigen receptor T‐cell (CAR‐T) therapy have shown improved outcomes in this patient population. While promising, limitations of these therapies include significant toxicities such as cytokine release syndrome (CRS), Immune effector cell‐associated neurotoxicity syndrome (ICANS), and B‐cell depletion, and there remain significant difficulties related to logistics, cost, and access to these therapies [6, 7, 8]. In the Phase II SPiReL trial (NCT03349450) we studied the efficacy and safety of a novel triple immunotherapy regimen with pembrolizumab, low‐dose cyclophosphamide, and maveropepimut‐S (MVP‐S, a DPX‐based T cell educating therapy directed against survivin) in 25 participants with R/R DLBCL [9].

The methodology of the SPiReL trial has been previously described in detail [9]. Adult participants from six Canadian centers with histologically confirmed R/R DLBCL transformed lymphoma, or other subtypes of high‐grade B‐cell lymphoma with survivin expression of >10% as assessed by central immunohistochemistry on a recent tumor biopsy were eligible for inclusion. Participants received two doses of 0.5 mL MVP‐S by subcutaneous injection 3 weeks apart followed by up to 6 additional doses every 8 weeks together with intermittent low‐dose oral cyclophosphamide (50 mg BID; 7 days on, 7 days off) and pembrolizumab 200 mg intravenously every 3 weeks for 1 year or until disease progression, whichever occurred first [9]. The primary outcome of the objective response rate using the modified Cheson criteria was 28% and an additional six patients had stable disease. The regimen was well‐tolerated, with only one discontinuation in addition to minor dose interruptions/modifications. Self‐limited Grade 1 or 2 injection site reactions were the most common treatment‐related adverse event and only three (12%) Grade 3 or greater neutropenia events occurred. Four related serious adverse events (two episodes of pancreatitis and two episodes of neutropenia) occurred in two participants, all related to cyclophosphamide/pembrolizumab. Full safety/adverse event data has been reported previously [9]. In this letter, we present the final data of the SPiReL trial from participants followed up to 40 months including an analysis of the clinical characteristics of participants with the longest survival.

The median age of participants was 74 years (range: 50–82); 56% were female. All participants had Eastern Cooperative Oncology Group performance status 0 (44%) or 1 (56%). 56%, 20%, and 24% had attempted 1, 2, and 3 prior lines of systemic therapy respectively; all of whom had attempted a CHOP ± rituximab regimen. Four participants had prior ASCT. In this final analysis, the median progression‐free survival (PFS) was 4.7 months (95% confidence interval [95% CI]: 3.0, 14.1), and overall survival (OS) was 10.1 months (95% CI: 8.3, NR) (Figure 1). Seven participants (33.8%) were alive and followed for longer than 2 years (range: 2.1–3.3 years); three of these participants were progression‐free. Time to next treatment, defined as the time from trial enrollment to the start of a new DLBCL treatment, was available for 16/25 participants and was 135.5 days (interquartile range: 209.5 days, range 37–909 days). No participants went on to CAR‐T or bi‐specific therapy after documented progression.

FIGURE 1.

Progression‐free survival and overall survival among 25 participants in the SPiReL trial.

We wished to identify clinical characteristics associated with enhanced PFS or OS, including survival longer than the study median of 10.1 months (n = 12) or in the participants with survival ≥2 years (n = 7). Although the number of prior systemic lymphoma treatments (p = 0.6) and prior history of ASCT (p = 0.97) were not associated with the duration of PFS, participants above the median time from the last treatment to the enrollment in the study (p = 0.04) had enhanced PFS. Antigen‐specific T‐cell responses to wild‐type (WT) survivin and immunizing peptides (detected using the interferon‐γ ELISpot assay with peripheral blood mononuclear cells) obtained before (p = 0.02) and during (p = 0.02) treatment were also associated with longer PFS. Among tumor biomarkers, only lack of MYC (p < 0.05) expression and lack of CD5 (p < 0.01) were associated with longer PFS. CD20+/PD‐L1 expression at cutoffs of both ≥ 5% and ≥ 10% (p < 0.01) as well as a CD20+/PD‐L1 H‐Score above a cutoff of ≥30% (p < 0.01) were all significantly associated with longer duration of PFS. Relapsed vs. refractory status (p = 0.45) and median time from diagnosis to enrollment (p = 0.45) were not associated with PFS. Regarding overall survival (Table 1), IPI score (p = 0.017), as well as immune responsiveness based on ISR (p = 0.003) and ELISpot Response (p = 0.03), were significantly associated with longer survival than the median. The sample size was insufficient to demonstrate statistical significance for the association between OS and other clinical/tumor‐related characteristics including age, tumor bulk, CD20+/PD‐L1+, and LDH in this exploratory analysis. Among the seven participants who survived longer than 2 years, the median age was 74.5 with a median IPI of 2; all seven participants (100%) had an ISR, and six (86%) also had an ELISpot Response. The sample size was insufficient to demonstrate statistical significance for associations between survival ≥ 2 years and baseline demographic, tumor, or treatment‐related characteristics.

TABLE 1.

Patient, tumor, and treatment characteristics of participants based on duration of overall survival.

	<10.1 months (n = 13)	≥10.1 months (n = 12)	p‐Value
Patient characteristics
Median days from last treatment to recurrence (days)	189	180	0.53
Median time from diagnosis (years)	1.34	4.13	0.78
Median age (years)	77	72.5	0.27
Median lines of systemic treatment	1	2	0.67
Median LDH	308	201	0.19
Median IPI	3	2	0.017
Tumor characteristics
CD20+/PD‐L1+	6/13 (1*)	7/12 (1**)	0.51
CD5+	3/13	0/12 (2*)	0.99
Median tumour bulk (cm2)	61.52	16.19	0.12
Sub‐class (Non‐GCB)	6/13	5/12	0.82
Transformed	5/13	3/12	0.47
Ki‐67 (>50)	11/13 (2*	8/12 (2*)	0.99
MYC	7/13(1*)	4/12 (1***)	0.29
Treatment characteristics
BOR/DCR (CR+PR+SD)	3/13	10/12	0.005
ISR	3/13	11/12	0.003
ELISpot Response.	2/13(5*)	9/12.	0.03

^*Number of participants without available data.

In summary, the SPiReL trial examined a triple immunotherapy combination regimen with an antigen‐specific T‐cell educating therapy in addition to inhibitors of regulatory T‐cells and T‐cell exhaustion in participants with R/R DLBCL. Survival in the SPiReL trial was 33.8% at two years with a median overall survival of 10.1 months (95% CI: 8.3, NR). Interestingly, enhanced immune reactivity as assessed by ELISpot response to the survivin‐immunizing peptides or ISR were each associated with significantly longer PFS as well as OS in the SPiReL trial. We postulate that the participants with a more durable response and long‐term survival may have a more intact immune system perhaps related to their more favorable clinical characteristics and may have been able to better respond to the combination immune therapy studied.

This triple immunotherapy treatment regimen studied in SPiReL showed activity in R/R DLBCL patients ineligible for or refractory to ASCT. Given that the response rate to pembrolizumab monotherapy in the KEYNOTE‐13 trial was low, we suspect that the encouraging activity of our regimen is related to the combination of immune active agents (including cyclophosphamide and MVP‐S) rather than any one agent alone [5]. This regimen was also very well‐tolerated in a relatively elderly cohort of patients which is a notable finding when compared with other novel immunotherapies such as CAR‐T cells and bispecifics which are also associated with improved outcomes in R/R DLBCL patients but may be associated with significant toxicities [9]. In two prior trials, the median overall survival (OS) was 11.5 months (95% CI: 7.9–15.7 months) with Glofitamab and not reached (95% CI: 11.3 months–NR) with Epcoritamab; both with significant rates of Grade 3–4 cytopenias including neutropenia, Grade 2 or higher CRS as well as neurologic toxicity consistent with ICANS [6, 8, 10]. The CAR‐T cell therapy Tisangenlecleucel yielded a median OS of 11.1 months in the JULIET trial but had similarly significant rates of Grade 3–4 cytopenias including neutropenia and Grade 3–4 CRS [7]. Of note, however, while these regimens were also studied in R/R DLBCL, most participants in these trials had received three or more lines of therapy whereas in the SPiReL trial, 75% of participants had only received 1–2 lines of treatment. Thus, comparisons of OS and toxicities of these trials with the results of SPiReL may not be directly applicable.

Limitations to this analysis include the lack of direct comparison with novel therapies including CAR‐T and bispecific antibody therapy, small sample size, and potential lack of generalizability to other subsets/cohorts of participants such as those who have already received CAR‐T cell or bispecific therapy. That being said, given the expected median OS of 6.3 months as defined in the SCHOLAR‐1 data, the median OS (10.1 months) seen in the SPiReL trial appeared to be promising, within the limits of non‐randomized comparison. [1]. Our results suggest that the selection of R/R DLBCL patients with favorable clinical characteristics or with the potential for immune reactivity may identify those who are better able to respond to well‐tolerated, active combination immunotherapies such as the regimen studied in SPiReL, to achieve a longer duration of responses.

AUTHOR CONTRIBUTIONS

Arjun Pandey analyzed the data and wrote the paper. Kim Roos analyzed the data and wrote the paper. Yidi Jiang analyzed the data. Kathryn Mangoff analyzed the data. Gail Klein analyzed the data. Neil L. Berinstein designed the research study, performed the research, analyzed the data, and wrote the paper. Nick Forward performed the research and wrote the paper. Douglas Stewart performed the research and wrote the paper. Pierre Laneuville performed the research and wrote the paper. Joy Mangel performed the research and wrote the paper. Isabelle Bence‐Bruckler performed the research and wrote the paper. George Tomlinson analyzed the data.

CONFLICT OF INTEREST STATEMENT

Arjun Pandey declares no conflict of interest; Kim Roos declares no conflict of interest; Yidi Jiang declares no conflict of interest; Kathryn Mangoff declares no conflict of interest; Gail Klein declares no conflict of interest; Neil L. Berinstein declares no conflict of interest; Nick Forward Speaker Fees: Pfizer; Advisory Board Membership: AstraZeneca, AbbVie, BMS, Celgene, IMV, Janssen, Pfizer, Roche, Servier. Research Support: Astellas, IMV, Merck, Seattle Genetics, MorphoSys; Douglas Stewart Consulting fees for Hoffman La Roche, SeaGen Canada, Novartis, Gilead, AbbVie, Exact Sciences Canada; PL Speaker Fees: Novartis, Janssen, Paladin, Teva, Amgen. Advisory Boards: Novartis, Pfizer, Janssen, Paladin, AbbVie, Roche, BMS. Research Support: Novartis, Gilead, AstraZeneca, Paladin, BMS; IBB Merck Advisory Board; Joy Mangel declares no conflict of interest; George Tomlinson declares no conflict of interest.

FUNDING INFORMATION

The per‐patient funding, funding for the translation studies, and MVP‐S were provided by IMV Inc., pembrolizumab was provided by Merck Canada.

ETHICS STATEMENT

The study was conducted in compliance with the International Conference on Harmonisation Good Clinical Practice E6 (ICH‐GCP), World Medical Association Declaration of Helsinki: Ethical Principles for Medical Research Involving Human Participants, as well as applicable regulatory and research ethics board requirements and research policies. OCREB, MUHC, HREBA, and NSHAREB.

PATIENT CONSENT STATEMENT

All participants enrolled provided informed consent.

CLINICAL TRIAL REGISTRATION

ClinicalTrials.gov Identifier: NCT03349450

Pandey A, Roos K, Jiang Y, Mangoff K, Klein G, Forward N, et al. Characteristics of relapsed/refractory diffuse large B‐cell lymphoma patients with durable responses to maveropepimut‐S, pembrolizumab, and cyclophosphamide: Long‐term follow‐up from the SPiReL trial. eJHaem. 2025;6:e1062. 10.1002/jha2.1062

DATA AVAILABILITY STATEMENT

For original data, contact SPiReL@sunnybrook.ca. Individual patient data will not be identified, however, data can be shared for a period of up to 5 years after trial closure with formal Confidentiality and Data Transfer agreements in place as appropriate. A data dictionary will not be shared, but the study protocol can be found (https://clinicaltrials.gov/ct2/show/NCT03349450?term=spirel&draw=2&rank=1) and the Statistical Analysis Plan can be made available with publication.