Monocytes Are Associated with Impaired T-Cell Immunity and Residual Interim-PET/CT Avidity After 4 Cycles of CHOP-R In Patients with High-Risk DLBCL,

Impaired systemic immunity is an established risk factor for the development of B-cell lymphoma. However current understanding of the underlying mechanisms of cell-mediated immune evasion and how this varies during chemo-immunotherapy remains rudimentary. Studies have been hampered by retrospective analysis of heterogenous cohorts containing both new and relapsed patients, with a variety of indolent/aggressive histological subtypes, treated with different treatment regimens.

We performed detailed analysis of circulating immune cell populations in 33 patients (mean age 53 years, range 31 – 69, F:M 11:22) with newly diagnosed high-risk diffuse large B-cell lymphoma (DLBCL) enrolled in a prospective multi-centre Australasian Leukaemia & Lymphoma Group trial. They received 3 cycles of CHOP-R-14, and then after a 4^th cycle underwent an interim-PET/CT with centralized review performed at day 17–20 (to enable risk-stratification). Pegylated-GCSF was given on day 2 after each cycle. PET/CT metabolic response categorisation used the harmonisation criteria. 10/33 evaluable patients remained interim-PET/CT avid. Blood samples were taken pre-treatment and on day 21 post-cycle 4, and compared to 23 age/gender-matched healthy control subjects. Results were correlated with interim-PET/CT response status. Laboratory investigators were blinded to clinical results.

CD3⁺, CD8⁺ and CD4⁺ T-cells (p=00.21, p=0.005, p=0.0017) among patients with DLBCL were all reduced at pre-treatment compared to controls. After 4 cycles, CD4⁺ lymphocytes were reduced further (p=0.0003) whereas CD3⁺ and CD8⁺ counts remained unchanged. However, CD3⁺, CD8⁺ and CD4⁺ proliferative responses to CD28/CD3 stimulation were equivalent to healthy subjects at both time-points, indicating the absence of an intrinsic functional T-cell defect in response to stimulation. Nor were regulatory T-cells (CD4⁺CD25^hiCD127⁻) or myeloid-derived suppressor cells (CD33⁺CD14⁻HLA-DR⁻) altered between time-points or as compared to healthy subjects.

Circulating monocytes are key regulators of inflammation/ immunity. Absolute monocyte counts were strikingly elevated at pre-treatment compared to healthy subjects (p=0.007), and had a predominantly CD14⁺HLA-DR^−/lo phenotype. Importantly, pre-treatment values for absolute monocytes (p=0.04), monocyte % (p=0.004), CD14⁺HLA-DR^−/lo monocytes (p=0.03), mean fluorescent intensity of HLA-DR on CD14⁺ monocytes (p=0.025) and CD3⁺ T-cells (p=0.0374), were all associated with persistent interim-PET/CT avidity. However, all were independent of the international prognostic index (IPI) score, and IPI did not segregate with interim-PET/CT status. Post-cycle 4 monocyte % (p=0.048) and CD3⁺ T-cells (p= 0.029) were also related to interim-PET/CT avidity.

The CD14⁺HLA-DR^−/lo phenotype has been described as immunosuppressive (Lin, Blood 2011). Consistent with this, in-vitro depletion of monocytes from pre-treatment (p=0.03) and post-cycle 4 (p=0.04) samples enhanced CD4⁺ proliferation. By contrast, monocyte depletion from control samples did not influence CD4⁺ proliferation. Post-cycle 4 monocytes rose above pre-treatment levels (p=0.0301) as did fluorescent intensity of HLA-DR on CD14⁺ monocytes (p=0.0027). In line with the post-cycle 4 reduction of the CD14⁺HLA-DR^−/lo ratio (p=0.0135), plasma arginase was reduced relative to pre-treatment (p=0.0035), suggesting that suppression was mediated in part through arginine metabolism. HLA-DR expression on monocytes at the interim time-point was up-regulated from pre-treatment (p=0.01), and accompanied by a relative increase in ex-vivo blood myeloid dendritic cells (BMDCs; p=0.044), suggesting the CD14⁺HLA-DR^−/lo phenotype is associated with an impaired ability to differentiate into BMDCs in-vivo.

Monocytes in DLBCL mediate immunosuppression of T-cells and both appear to be associated with interim-PET/CT avidity. Although the role of interim-PET/CT in DLBCL patients treated with chemo-immunotherapy is an area of considerable activity, the study of biological predictors of metabolic response has been relatively neglected. To the best of our knowledge, this is the first prospective study to find an association between cellular markers of inflammation/ immunity and interim-PET/CT avidity in a cohort of uniformly treated high-risk DLBCL patients.

No relevant conflicts of interest to declare.