Abstract
Background: Patients with high grade B cell lymphoma (HGBL) who harbor a MYC rearrangement with BCL2 and/or BCL6 rearrangements (HGBL-DH/TH) have a poor prognosis upon treatment with R-CHOP [Rosenwald, JCO 2019]. Although disease free survival may be prolonged with intensification of chemotherapy with DA-EPOCH-R, improvement of overall survival has not been demonstrated [Dunleavy, Lancet Haematol 2018]. Aims: Preclinical data suggest that high MYC expression impairs anti-tumor immune responses [Casey, Science 2016]. Therefore, we designed the HOVON-152 trial to investigate whether survival can be improved by nivolumab consolidation after DA-EPOCH-R induction. We also investigated whether DA-EPOCH-R induction influences the immune system to have an impact on the response. Methods: In the HOVON-152 single arm, phase II trial (NCT03620578) HGBL-DH/TH patients received 1 cycle of R-CHOP followed by 5 cycles of DA-EPOCH-R induction treatment. Patients in complete metabolic remission (CMR) after induction (Deauville score (DS) 1-3 or a negative lymphoma biopsy in case of DS 4) proceeded with nivolumab consolidation (480 mg iv every 4 weeks) for one year. Peripheral blood was sampled after one cycle of R-CHOP (enroll), before start of the 3rd DA-EPOCH-R cycle (midterm) and after the last DA-EPOCH-R cycle (end-of-induction, EOI). As of January 2022, 55/90 enrolled patients completed DA-EPOCH-R induction, of whom 31 achieved PET-CT based CMR, 24 did not achieve CMR or went off study prematurely. Multiparameter flow cytometry of unseparated full blood was used to enumerate major immune cell subsets. Cryopreserved peripheral blood mononuclear cells were used to determine the frequencies of T cell subsets. Computational flow cytometry analyses (i.e UMAP and FlowSOM) and appropriate unpaired (Mann-Whitney U) or repeated (Friedman) statistical tests were performed in R version 4.0.3. Results: In full blood, CD3 T cell frequencies and numbers decreased midterm DA-EPOCH-R treatment and restored at EOI. Remarkably, the proportion of CD4 T cells decreased (p=0.048) and that of CD8 T cells increased (p=0.01), resulting in lower CD4/CD8 ratios at EOI. In depth analysis revealed a significant decrease of CD4 naïve cells (p=0.00002) and increase of CD4 central memory (CM) cells (p=0.0005). Over time, CD8 cells showed a more senescent phenotype (CD28-KLRG1+CD57+, p=0.013), but the proportion of PD-1 positive exhausted CD8 cells decreased (p=0.007). When stratified for response, it appeared that patients achieving CMR after DA-EPOCH-R had significantly higher CD3 T cells (p=0.04) at enroll than patients not achieving CMR. T cell differentiation states (CD4 or CD8 naïve cells, stem cell memory cells, CM, effector memory (EM) or effector memory RA (EMRA) cells) and the proportion of TIGIT and TIM3 positive cells did not significantly differ between patients achieving CMR or not achieving CMR. Most interestingly, the decrease in PD-1 positive exhausted CD8 T cells was most outspoken in the CMR group (p=0.0008) (Figure 1) in CD8 CM, EM and EMRA cells (all p<0.008). Thus, at EOI, a lower proportion of CD8 PD-1 cells was associated with achieving CMR (p=0.04). Image:Summary/Conclusion: Treatment of HGBL-DH/TH patients with DA-EPOCH-R results in a shift towards a lower CD4/CD8 ratio. Response to DA-EPOCH-R is associated with higher T cell frequencies at enroll and a significant decrease in PD-1 positive exhausted CD8 T cells during DA-EPOCH-R. Additional analyses on NK cells are ongoing. Future analyses of the nivolumab consolidation phase of this study will indicate whether additional blockade of PD-1 prolongs survival.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have