Background The anti-NKG2A monoclonal antibody (mAb), NNC141-0100, may potentially reduce inflammation in Rheumatoid Arthritis (RA) and other immune disorders via a novel mechanism of action. NNC141-0100 has been proposed to promote the cytotoxicity of natural killer (NK) cells towards pro-inflammatory cells expressing HLA-E. NNC141-0100 blocks the inhibitory interaction between HLA-E and the heterodimer CD94/NKG2A receptor on the NK-cells by binding to the NKG2A subunit. Given the novelty of the mechanism of action involving the immune system, the starting dose in the First-in-Man trial should be supported by determination of the Minimum Anticipated Biological Effect Level (MABEL)¹.

Objectives To characterise the affinity and potency of NNC141-0100 in human in vitro systems designed to facilitate the translation to humans. The results should be used to calculate the MABEL and set the starting dose following implementation in a PK/PD model.

Methods The affinity (Kd) and maximum receptor binding capacity was assessed in whole blood incubated with increasing concentrations of NNC141-0100. The number of free and bound CD94/NKG2A receptors on relevant peripheral blood leucocytes subsets was assessed by flow cytometry. For comparison, Kd was assessed by surface plasmon resonance (Biacore®) and on isolated peripheral blood mononuclear cells (PBMC). The potency for the biological effect (EC₅₀) was tested in an in vitro system using cell types resembling the in vivo situation. Cytokine-stimulated human PBMC (effector cells) were incubated with autologous CD4⁺ T-cells (target cells), and increasing concentrations of NNC141-0100. The cytotoxic potential of the effector cells was assessed by flow cytometry using the degranulation marker CD107, known as a sensitive biomarker preceding cytotoxicity. Both healthy subjects and Rheumatoid Arthritis (RA) subjects provided blood samples. In the PK/PD model, the Kd and EC₅₀ values provided the basis for simulating receptor occupancy and CD107 profiles versus time and dose levels in humans.

Results NNC141-0100 was bound with a similar, high affinity to the human CD94/NKG2A receptor in all test systems. The EC₅₀ in the CD107 assay was approximately 100-fold higher than the Kd. This difference may be due to competitive interaction from HLA-E on the target cells. Similar CD107 potencies were observed for healthy and RA subjects. By means of PK/PD simulations, the i.v. MABEL dose in humans was calculated to give a predicted CD107 response of 10% at C_max. The corresponding receptor occupancy at C_max was 90%. Due to the high occupancy and the uncertainty of the in vitro-in vivo correlation for cytotoxicity and potentiating the immune system, a safety factor was applied to the MABEL to determine the starting dose.

Conclusions Given the observed difference between affinity and potency, the MABEL dose for the First-In-Man trial in RA patients had a high predicted CD94/NKG2A receptor occupancy of 90%, while the expected cytotoxicity as measured by the degranulation marker CD107 was much lower (<1%).

1. EMA guideline: EMEA/CHMP/SWP/294648/2007

Disclosure of Interest L. Alifrangis Shareholder of: Novo Nordisk A/S, Employee of: Novo Nordisk A/S, P. André Shareholder of: Innate Pharma, Employee of: Innate Pharma, V. Pascal Shareholder of: Novo Nordisk A/S, Employee of: Novo Nordisk A/S, E. Bonnet Shareholder of: Innate Pharma, Employee of: Innate Pharma, L. Radzikowski Shareholder of: Novo Nordisk A/S, Employee of: Novo Nordisk A/S, M. Petersen Shareholder of: Novo Nordisk A/S, Employee of: Novo Nordisk A/S, M. Bléry Shareholder of: Innate Pharma, Employee of: Innate Pharma