Autologous tumor cell vaccine induces antitumor T cell immune responses in patients with mantle cell lymphoma: A phase I/II trial

Abstract

Here, we report on the results of a phase I/II trial (NCT00490529) for patients with mantle cell lymphoma who, having achieved remission after immunochemotherapy, were vaccinated with irradiated, CpG-activated tumor cells. Subsequently, vaccine-primed lymphocytes were collected and reinfused after a standard autologous stem cell transplantation (ASCT). The primary endpoint was detection of minimal residual disease (MRD) within 1 yr after ASCT at the previously validated threshold of ≥1 malignant cell per 10,000 leukocyte equivalents. Of 45 evaluable patients, 40 (89%) were found to be MRD negative, and the MRD-positive patients experienced early subsequent relapse. The vaccination induced antitumor CD8 T cell immune responses in 40% of patients, and these were associated with favorable clinical outcomes. Patients with high tumor PD-L1 expression after in vitro exposure to CpG had inferior outcomes. Vaccination with CpG-stimulated autologous tumor cells followed by the adoptive transfer of vaccine-primed lymphocytes after ASCT is feasible and safe.

Conflict of interest statement

Disclosures: R.H. Advani reported grants from Forty Seven, Merck, Seattle Genetics, Roche/Genentech, Regenron, Kura, Pharmacyclics, and Celgene; and personal fees from Seattle Genetics, Roche/Genentech, Sanofi, Bayer, Astra Zeneca, Gilead, Autolus, Cell Medica, Celgene, and Portola outside the submitted work. A.R. Rezvani reported grants from AbbVie outside the submitted work; served on one-time scientific advisory boards with Nohla Therapeutics and Kaleido, both in 2018; served as medical expert witness for the US Department of Justice; and his brother is employed by Johnson & Johnson. J.D. Brody reported grants from Genentech, Seattle Genetics, Kite, Celldex, Acerta, Merck, Janssen, and Gilead outside the submitted work. R. Levy serves on scientific advisory boards of the following companies: Quadriga, BeiGene, GigaGen, Teneobio, Sutro, Checkmate, Nurix, Dragonfly, Abpro, Apexigen, Spotlight, Forty Seven, Inc., XCella, Immunocore, and Walking Fish. No other disclosures were reported.

© 2020 Frank et al.

Figures

Figure 1.

Schema of trial design and CONSORT diagram. (A) Prior to chemotherapy, tumor cells were collected by apheresis or biopsy, treated with CpG, radiated, and cyropreserved in single-use aliquots as described in the Materials and methods. Patients achieving at least a partial response to initial chemotherapy received three vaccine doses followed by apheresis for T cell collection. 1 d after infusion of autologous stem cells, collected T cells were reinfused and a fourth vaccination was given. After complete recovery from the ASCT, a final “booster” vaccination was given. PBMCs were collected before and after the initial three vaccine doses for immune response assessments. (B) CONSORT diagram of all patients enrolled.

Figure 2.

Clinical trial outcomes: TTP, OS, and patient outcomes by MRD analysis.(A) Swimmers plot shows time to first MRD event(s) and time to relapse from the time of ASCT. The first detectable MRD event is shown in either blue or red, and the time of relapse is shown in yellow. The time of first MRD detection >1 molecule per 104 is red, and time of first MRD detection less than this threshold is blue. (B and C) TTP (B) and OS (C) examined by Kaplan-Meier analysis and log-rank test were assessed from the date of ASCT in all patients who received the CpG-MCL vaccination and adoptive T cell transfer. L.E., leukocyte equivalents.

Figure S1.

Clinical trial outcomes from the start of immunochemotherapy and by B-MIPI scores from ASCT.(A and B) TTP (A) and OS (B) examined by Kaplan-Meier analysis and log-rank test was assessed from the initial day of immunochemotherapy in all patients who received the CpG-MCL vaccination and adoptive T cell transfer. (C and D) TTP (C) and OS (D) examined by Kaplan-Meier analysis and log-rank test was assessed from ASCT stratified by high versus low/intermediate (low/int) B-MIPI scores. Patients with low/intermediate B-MIPI scores compared with high B-MIP scores had longer median TTP (not reached vs. 6.93 yr; *, P = 0.0260) and median OS (not reached vs. 4.11; *, P = 0.0143).

Figure S2.

TTP based on source of the vaccine. TTP examined by Kaplan-Meier analysis and log-rank test was assessed from ASCT stratified by source of vaccine.

Figure 3.

Patients who generate a memory CD8 T cell vaccine response have favorable outcomes.(A) An example of a patient with an induction of tumor-reactive memory CD8 T cells after vaccination is shown. PBMCs, collected before or after vaccine administration, were simultaneously coincubated with or without CpG-stimulated tumor cells, as described in the Materials and methods. Each plot is gated on CD8+ T cells, and the percentage CD137+CD45RO+ T cells are shown. The percentage of tumor-reactive memory CD8+ T cells is determined by subtracting the percentage of CD137+CD45RO+ T cells found in the media-only sample from the percentage found in the sample with CpG-stimulated tumor cells. This patient was determined to have had a memory CD8+ T cell vaccine response since there was a higher percentage of tumor-reactive memory CD8+ T cells after vaccination compared with baseline. (B) TTP from ASCT was examined by Kaplan-Meier analysis and log-rank test in patients with a memory CD8 T cell vaccine response (blue) or no vaccine response (red). Trial patients with a memory CD8 T cell vaccine response had a significantly longer TTP than those who did not (median TTP was not reached vs. 5.8 yr; *, P = 0.0246).

Figure S3.

Patient outcomes by memory CD4 T cell vaccine response.(A) TTP was examined by Kaplan-Meier analysis and log-rank test from the ASCT in patients with a memory CD4 T cell vaccine response (blue) or no vaccine response (red). (B) Mantle cell tumors were incubated with CpG as described in the Materials and methods and assayed for PD-L1 expression. Each row shows a patient’s CpG-induced tumor PD-L1 expression from lowest to highest expression level in relation to whether that patient had a CD4 T cell vaccine response (black box) or no response (white box).

Figure S4.

Vaccine-associated immune responses.(A) Immune responses to lymphoma Ig neoantigens were tested in a patient 2 yr after completion of ASCT. PBMCs were stimulated with either a pool of T cell epitopes from common pathogens or with a peptide corresponding to a lymphoma-specific Ig neoantigen. Antigen-specific CD8 and CD4 responses were detected by CD137 expression. (B) Antigen-specific CD4 T cells as determined in A were FACS sorted and expanded ex vivo. Expanded T cells were co-cultured with labeled autologous lymphoma cells for 24 h, and killing of lymphoma cells was determined by staining with 7-amino-actinomycin D.

Figure S5.

CpG-activated tumor characteristics. Mantle cell tumors were incubated with media only or CpG as described in the Materials and methods and assayed for the expression of the indicated molecules. The mean for each indicated molecule is indicated by the horizontal line. ***, P = 0.0001–0.0003; ****, P < 0.0001 by a paired two-tailed t test. MFI, mean fluorescence intensity.

Figure 4.

High CpG–induced PD-L1 expression is associated with poor outcomes.(A) Mantle cell tumor suspensions were incubated with CpG as described in the Materials and methods and assayed for PD-L1 expression. Examples of low (upper panel) and high (lower panel) tumor PD-L1 expression are shown. Each row shows a patient’s CpG-induced tumor PD-L1 expression from lowest to highest expression level in relation to whether that patient had a CD8 T cell vaccine response (black box) or no response (white box). Expression level was determined by mean fluorescence intensity (MFI) normalized to an isotype control. The average expression level, as measured by MFI, is shown for the 36 patients with available material for analysis. Patients with expression levels above this level were considered to have high PD-L1 expression, and those below this level were considered to have low PD-L1 expression. (B and C) TTP and OS examined by Kaplan-Meier analysis and log-rank test was assessed from ASCT stratified by PD-L1 expression (red line represents patients with tumor PD-L1 expression above the mean, and blue line represents patients with tumor PD-L1 expression below the mean). Patients with lower PD-L1 expression had a significantly longer TTP (median TTP was not reached vs. 5.76 yr; *, P = 0.0453) and OS (median OS was not reached vs. 4.11 yr; *, P = 0.0462) than those who did not.