Myeloid-derived suppressor cells (MDSCs) in metastatic castration-resistant prostate cancer (CRPC) patients (PTS)

Abstract

Myeloid-derived suppressor cells (MDSCs) in castration-resistant prostate cancer (CRPC) patients (PTS) Type: Abstract Category: Genitourinary tumours, prostate Authors: N. Mehra1, G. Seed1, M. Lambros1, A. Sharp1, M. Sousa Fontes1, M. Crespo1, S. Sumanasuriya1, W. Yuan1, G. Boysen1, R. Riisnaes1, A. Calcinotto2, S. Carreira1, J. Goodall1, Z. Zafeiriou1, D. Bianchini1, A. Morilla3, R. Morilla3, A. Alimonti2, J.S. de Bono1; 1Cancer Studies, The Institute of Cancer Research (ICR), London, GB, 2Institute of Oncology Research (IOR), Oncology Institute of Southern Switzerland (IOSI), Bellinzona, CH, 3Centre for Molecular Pathology, Institute of Cancer Research ICR, London, GB Background The relevance targeting MDSCs in CRPC is increasingly recognized; preclinical studies implicate MDSCs with senescence evasion and treatment resistance following loss of tumour suppressor (TSL) function or oncogenic activation (OA). We investigated MDSC subset levels in CRPC PTS with regard to their molecular underpinnings and associated with PSA response. Methods We prospectively evaluated MDSCs in 46 progressing CRPC PTS prior to new lines of therapy (n=65), and in 14 male healthy volunteers (HV). Following blood draw MDSCs were analysed within 24h according to a gating strategy designed to standardize MDSC phenotyping. Here we report 5 MDSC subsets (phenotypes described in table), of which 2 are monocytic (M)/granulocytic (Gr) lineage-negative (MDSC3 and MDSC9), 2 with Gr (MDSC8 and MDSC8A) and 1 with a M phenotype (MDSC4). Subsets are expressed as % of their parental population. Data were acquired with a BD Canto II with FACS-Diva and analysed using Kaluza 1.3. We tested for the association between MDSCs subsets and copy number aberrations of 8q gain, loss (het/homdel) of PTEN and RB1 in cell-free circulating DNA by targeted amplicon-based sequencing (IonTorrent) using CNVkit V0.3.5. Differences in levels of the 5 MDSC subsets were assessed using non-parametric testing and associations of dichotomized MDSC subsets (at median) with PSA response were tested by Chi-square. Results Overall, 65 baseline samples were analysed from 46 PTS with 4 of 5 MDSC subsets increased significantly in CRPC samples compared to HV (Table). Subset Phenotype CRPC (%) CRPC (N) HV (%) HV (N) P MDSC3 CD14neg CD15neg HLA- DRlow CD33pos 0.86 65 2.17 13 0.034 MDSC9 CD14neg CD15neg CD33pos 5.76 65 5.16 13 0.373 MDSC8 CD15pos CD33pos 1.99 65 0.40 14 <0.001 MDSC8A CD15pos CD33pos CD62Lpos 0.49 65 0.22 14 0.002 MDSC4 CD14pos CD15neg HLA- DRlow CD33pos 22.6 63 6.1 14 <0.001 Increased MDSCs were associated with 8q gain (MDSC8, n=20 /57 evaluable, P=0.004; MDSC4, 19 /55, p=0.027), with RB1 loss (MDSC8A; n=13 /56; P=0.017) with a trend for increased MDSCs with PTEN loss (MDSC8A; n=4 loss/57; P=0.16), with a lower likelihood of a PSA response in PTS with elevated MDSC8 (P=0.001). Conclusions MDSC subsets are increased in CRPC, with specifically Gr-MDSC populations associating with molecular aberrations and treatment resistance. Clinical trial identification http://cpaper.ctimeetingtech.com/submission/preview/print_publication.php?publication_id=2449 Page 1 of 2 11/05/2016, 15:38 Print http://cpaper.ctimeetingtech.com/submission/preview/print_publication.php?publication_id=2449 Page 2 of 2