Highly effective NK cells are associated with good prognosis in patients with metastatic prostate cancer

Abstract

Clinical outcome of patients with metastatic prostate cancer (mPC) at diagnosis is heterogeneous and unpredictable; thus alternative treatments such as immunotherapy are investigated. We retrospectively analyzed natural killer (NK) cells by flow cytometry in peripheral blood from 39 mPC patients, with 5 year-follow-up, and their correlation with time to castration resistance (TCR) and overall survival (OS). In parallel, NK functionality was carried out against prostate tumor cell lines, analyzed for the expression of NK cell ligands, to identify the receptors involved in PC recognition. NK cells from patients with longer TCR and OS displayed high expression of activating receptors and high cytotoxicity. The activating receptors NKp30 and NKp46 were the most obvious predictive markers of OS and TCR in a larger cohort of mPC patients (OS: p= 0.0018 and 0.0009; TCR: p= 0.007 and < 0.0001 respectively, log-rank test). Importantly, blocking experiments revealed that NKp46, along with NKG2D and DNAM-1 and, to a lesser extent NKp30, were involved in prostate tumor recognition by NK cells. These results identify NK cells as potential predictive biomarkers to stratify patients who are likely to have longer castration response, and pave the way to explore therapies aimed at enhancing NK cells in mPC patients.

Keywords: castration resistance; metastases; natural killer cells; prostate cancer; survival.

Figure 1. NK cells from mPC patients with longer survival and response to castration have strong cytotoxic potential

mPC patients sampled within 2 months after diagnosis of metastases and without treatment at sample (n = 18) were stratified into two groups, according to the time to castration resistance with an 18-months cutoff value: LCR (long castration response, n = 8) and SCR (short castration response, n = 10) patients. A. Clinical endpoints analyzed in this study: overall survival (OS) and time to castration resistance (TCR). B. Kaplan-Meier curves of OS and TCR. Blue solid line, LCR patients; pink dashed line, SCR patients. The relative differences in survival and response distribution (χ²) and p values were determined by log-rank statistics. C. The expression of NK cell markers on peripheral NK cells was analyzed by flow cytometry in LCR (white plots) and SCR (grey plots) patients sampled at diagnosis. The y axis shows the MFI ratio or the percentage of NK cells (CD56+CD3-) positive for each marker depending on uni- or bimodal expression. Data are represented by “box and whisker (min to max; horizontal lines represent mean values)” graphs. P values were obtained using Mann-Whitney test. p < 0.05 = *; p < 0.01 = **; p < 0.001 = ***. D. Gating strategy for NK cells (CD56+CD3− among living lymphocytes) and representative histogram or dot plot for each NK cell marker.

Figure 2. NKp46high and/or NKp30high phenotype is associated with good prognosis in mPC patients

Kaplan-Meier curves for OS and TCR were performed on data from the total cohort of 39 mPC patients (n = 18 patients used for first statistical analyses and n = 21 patients initially excluded because under treatment at sample or sampled at distance from diagnosis). Patients were stratified according to the previously defined cut-off value of NKp30 A. and NKp46 B. Log-rank statistics were used to compare high (solid line) and low (dashed line) curves; χ² and p values are indicated on each graph with p < 0.05 = *; p < 0.01 = **; p < 0.001 = ***. n indicates the number of corresponding patients in each arm.

Figure 3. Receptors involved in human prostate cancer cell recognition by NK cells

A. Expression of NK cell ligands on PC3, DU145 and LNCaP prostate cell lines by flow cytometry. The score indicate: − : no detectable expression; ± : less than 5%; + : 5% to 10%; ++ : 10% to 50%; +++ : 50% to 100%; ++++ : 100% of cells express the ligand. The receptor matching with each ligand is indicated. B. PBMCs from healthy donors were activated overnight in IL-2 and IL-15, and used in a 4-hours CD107 degranulation assay against PC3, DU145 or LNCaP prostate tumor cell lines, in the presence of blocking antibodies for NKp46, NKp30, DNAM-1, NKG2D or irrelevant isotype control mAb. E:T = 10:1. The percentage of CD107 degranulation by NK cells was evaluated by flow cytometry. The decreased effect of the respective blocking antibody compared with the irrelevant isotype control mAbs was evaluated with a non-parametric Wilcoxon test. The difference between prostate cell lines and K562 cell line was evaluated with Mann-Whitney test. Results are representative of 4 independent experiments. p < 0.05 = *; p < 0.01 = **; p < 0.001 = ***.