Quick viewing(Text Mode)

Gene Expression Profiles Associated with Stable Disease in Metastatic Castration- Resistant Prostate Cancer Patients Treated with ADXS-PSA Immunotherapy Sandra M

expression profiles associated with stable disease in metastatic castration- resistant prostate patients treated with ADXS-PSA Sandra M. Hayes1, David Balli1, Rachelle E. Kosoff1, Robert G. Petit1, Mark Stein2, Ronald Tutrone3, Anthony Mega4, Manish Agarwal5, Lawrence Fong6, Naomi Haas7 1Advaxis, Inc., Princeton, NJ; 2The Cancer Institute of New Jersey CINJ Rutgers, Inc., New Brunswick, NJ; 3Chesapeake Urology Research Associates; Towson, MD; 4Lifespan Oncology Clinical Re- search, Rhode Island Hospital, Providence, RI; 5Associates in Oncology/Hematology PC, Rockville, MD; 6UCSF University of California San Francisco, San Francisco, CA; 7University of Pennsylvania Abramson Cancer Center; Philadelphia, PA INTRODUCTION  The baseline levels of CD4+ T cells were comparable in stable disease and non-stable dis- Figure 7. Functional categories of differentially expressed in sta- ease patients, but their levels on treatment were 2-fold higher at week 6 and 2.6-fold ble disease patients on ADXS-PSA treatment • Active , such as ADXS-PSA, are designed to stimulate an antitumor response higher at week 9 in stable disease patients than in non-stable disease patients (P<.001 for T CELL ACTIVATION both time points). Upregulated AND FUNCTION by directly targeting and engaging the immune system. 13% ANTIGEN PROCESSING AND PRESENTATION • ADXS-PSA, a highly attenuated Listeria monocytogenes (Lm)-based immunotherapy that tar- Figure 3. ADXS-PSA treatment increases expression levels of genes identi- 18% ADHESION gets prostate-specific antigen (PSA), is currently being evaluated as a treatment for previous- fying T cells and major subsets in stable disease patients ANTIGEN PROCESSING AND PRESENTATION ly treated metastatic castration-resistant prostate cancer (mCRPC) in the phase 1/2 KEYNOTE NK CELL FUNCTION AUTOPHAGY -046 trial as a monotherapy (Part A, presented here) and in combination with KEYTRUDA® Total T cells CD8+ T cells CD4+ T cells 6% T CELL ACTIVATION AND NKNFSIGNAL -CELLKB SIGNALING FUNCTION TRANSDUCTIONFUNCTIONANTIGENADHESION PROCESSING AND CANCER TESTIS/TUMOR-ASSOCIATED ANTIGENS 1 AUTOPHAGY (pembrolizumab) (Part B, enrolling). INTERFERON SIGNALING0% 6% 0%PRESENTATION6% 175 SD NF-KB SIGNALING CANCER1% TESTIS/TUMOR- 5% 8% 1% CELLULAR FUNCTIONS ASSOCIATED ANTIGENS • Advaxis’ Lm-based immunotherapies act by stimulating innate immunity through multiple 2000 Non-SD CHEMOKINES AND 300 8% CHEMOKINE RECEPTORS CHEMOKINES AND CHEMOKINE RECEPTORS mechanisms including the STING pathway, by reducing the numbers and activities of immuno- 9% 150 * INNATE IMMUNE * FUNCTION CYTOKINES AND COMPLEMENT PATHWAY suppressive cells in the tumor microenvironment, and by inducing the generation of antigen- * 11% CYTOKINE RECEPTORS 2 CELLULAR FUNCTIONS9% specific T cells that infiltrate and destroy the tumor. * 1500 15% CYTOKINES AND CYTOKINE RECEPTORS 200 INNATEDownregulated IMMUNE FUNCTION 3,4 125 • Because of their ability to suppress the immune system, tumors and standard cancer treat- 29% INTERFERON SIGNALING IMMUNE REGULATION 8% CHEMOKINES ANDCANCER TESTIS/TUMOR- ments, such as chemotherapy and radiation therapy, may have an impact on the clinical out- ASSOCIATED ANTIGENS INNATE IMMUNE FUNCTION 1000 CHEMOKINE RECEPTORS8% come of cancer patients receiving active immunotherapies. 5% 100 COMPLEMENT PATHWAY INTERFERON SIGNALING 100 IMMUNECYTOKINES REGULATION AND CYTOKINE • For this reason, we assessed the immune status of previously treated mCRPC patients partici- ExpressionLevels NanoString 8% 1% RECEPTORS NF-KB SIGNALING pating in the KEYNOTE-046 trial. 500 12% 75 CELLULAR FUNCTIONS NK CELL FUNCTION Week 1 Week 3 Week 6 Week 9 Week 1 Week 3 Week 6 Week 9 Week 1 Week 3 Week 6 Week 9 INNATE IMMUNE 15% • Immune status was assessed by profiling and quantifying immune-related in FUNCTION pre-dose pre-dose pre-dose 29% SIGNAL TRANSDUCTION peripheral blood mononuclear cells (PBMCs) before and after ADXS-PSA treatment. *P<.001 • Part A, the ADXS-PSA dose-determining phase of the KEYNOTE-046 trial, evaluated 3 dose T CELL ACTIVATION AND FUNCTION

levels for safety and tolerability. COMPLEMENT PATHWAY OBJECTIVES 8% CYTOKINES AND CYTOKINE • To determine whether the dose level of ADXS-PSA has a bearing on post-treatment immune RECEPTORS 12% • Assess immune status of mCRPC patients who have been previously treated with standard status, we compared the expression levels of T cell-specific genes post-ADXS-PSA treatment lated genes in ADXS-PSA-treated stable disease patients versus ADXS-PSA-treated non- cancer therapies by profiling and quantifying immune-related gene expression in their among the 3 dosing cohorts (Figure 4). PBMCs before and after ADXS-PSA treatment. stable disease patients. Further analysis of the differentially expressed genes within this func-  The greatest fold-change above baseline in T cell-specific gene expression levels was ob- tional category revealed that (Figure 8): • Determine whether any gene expression profiles were associated with clinical response to served in the cohort that received the highest dose (1 x 1010 CFU) (P<.05). Interestingly,  On ADXS-PSA treatment, stable disease patients expressed significantly higher levels of ADXS-PSA monotherapy. this cohort did not include any patients who achieved clinical activity. genes indicative of M1 and plasmacytoid dendritic cells (pDCs), both of MATERIALS AND METHODS Figure 4. Highest ADXS-PSA dose affects expression levels of T cell- which have pro-inflammatory antitumor activities.6-10 The M1 -related genes that were upregulated in stable patients are IFNGR1, CD86, HLA-DR, IL6, IL12B, IL15, and specific genes • The KEYNOTE-046 trial (NCT02325557) is a phase 1/2 evaluation of ADXS-PSA alone (Part CXCL10.6,11,12 The pDC-related genes that were upregulated in stable disease patients are 13 A), and in combination with KEYTRUDA® (pembrolizumab) (Part B), in the treatment of mCRPC. 175 CD86, IL18R1, and IL3RA. 1 x 109 CFU 1 The study design for KEYNOTE-046 trial is summarized in Figure 1. 5 x 109 CFU  On ADXS-PSA treatment, stable disease patients expressed significantly lower levels of 1 x 1010 CFU genes indicative of M2 macrophages and myeloid-derived suppressor cells (MDSCs), both 150 160 Figure 1. Study design for KEYNOTE-046 trial SD patients of which have immunosuppressive protumor activities.9,10,14,15 The M2-macrophage-related Non-SD patients genes that were downregulated in stable disease patients are IL13RA, CD163, CD36, 125 Inclusion Criteria: LGAL3, and STAT6.6,11,12 The MDSC-related genes that were downregulated in stable dis- Progressive metastatic castration-resistant prostate 120 All patients ▪ 10,15,16 cancer (mCRPC) on androgen deprivation therapy ease patients are IL13RA1, PTSG2, STAT6, ARG1, ARG2, S100A8, and S100A12. ▪ <3 prior systemic treatment regimens, or >1 prior 100 regimen in the metastatic setting, with chemotherapy Figure 8. Differentially expressed genes involved in innate immune function 80 WEEK 1 WEEK 4 WEEK 7 REPEAT Q12 WEEK CYCLES NanoString ExpressionLevels 75 stable disease vs non-stable disease Downregulated Upregulated Endpoints: Part A Up to PD ▪ Safety/tolerability ADXS-PSA Monotherapy 50 or 2 ▪ RP2D of the 40 ▪ n=12 years combination (Part A) Week 1 Week 3 Week 6 Week 9 Week 1 Week 3 Week 6 Week 9 ▪ Dose escalation (3 dose levels) pre-dose pre-dose ▪ Antitumor activity ADXS-PSA ADXS-PSA ADXS-PSA and progression-free • To determine whether the increases in the levels of T cell-specific genes were associated with Part B Up to PD survival ADXS-PSA + KEYTRUDA® ▪ Peripheral or 2 changes in the state of T cell activation/differentiation, we measured the expression levels of ▪ n=30 immunologic value) years responses genes encoding T cell activation and differentiation markers in PBMCs of stable disease and P KEYTRUDA® KEYTRUDA® KEYTRUDA® KEYTRUDA® non-stable disease patients before and on ADXS-PSA treatment (Figure 5).

PD, progressive disease; RP2D, recommended phase 2 dose (adjusted

 At baseline, stable disease patients expressed: 10 log • Immune-related gene expression was performed on RNA from PBMCs isolated at 4 time  significantly higher levels of CXCR3 than non-stable disease patients (P<.001) - points during the first 9 weeks of treatment from mCRPC patients participating in the Part A  similar levels of TNFRSF9 (CD137) as non-stable disease patients ADXS-PSA dose-determining stage of the KEYNOTE-046 trial (Figure 2).  significantly lower levels of PDCD1 (PD-1), ENTPD1 (CD39), CD48, CD83, TBX21 (Tbet), • The NanoString nCounter PanCancer Immune Profiling Panel was used to quantitate gene IFNG, and GZMB (granzyme B) than non-stable disease patients (P<.001) expression levels.  On ADXS-PSA treatment, stable disease patients expressed significantly higher levels of: log2-fold change • Normalized NanoString gene-level counts were compared against NanoString’s immune cell-  PDCD1, ENTPD1, and TBX21 than non-stable disease patients at week 3 (P<.001) • The top 5 signaling pathways identified by the IPA software program that represent the 162 specific gene signatures to identify which immune cell types were detected in PBMCs before differentially expressed genes between ADXS-PSA-treated stable disease patients and non- and after ADX-PSA treatment.  PDCD1, ENTPD1, CD48, CD83, TNFRSF9, TBX21, and GZMB than non-stable disease patients at week 6 (P<.001). stable disease patients are: • Differential expression analysis was conducted on normalized NanoString count data by  Th1/Th2 activation pathway (P<.001, z-score=0) stratifying patient samples into groups defined as stable disease or non-stable disease (eg,  PDCD1, CD48, CD83, TBX21, and GZMB than non-stable disease patients at week 9 progressive disease), based on their antitumor response. To identify genes that were differ- (P<.001).  T helper cell differentiation pathway (P<.001, z-score=0) entially expressed (ie, adjusted P value ≤.1), we performed linear modelling as implement-  The fold-change increases above baseline in IFNG expression levels were comparable be-  Th2 pathway (P<.001, z-score=-0.5) ed in the R/Bioconductor package limma with voom transformation (version 3.30.13) in an R tween stable disease and non-stable disease patients after ADXS-PSA treatment.  Th1 pathway (P<.001, z-score=+3.9) statistical programming environment (R version 3.3.2).5 Figure 5. Stable disease patients express higher levels of T cell activation  maturation pathway (P<.001, z-score=+4.1) • Ingenuity Pathway Analysis (IPA) (QIAGEN) was used to identify signaling pathways and net- markers than non-stable disease patients on ADXS-PSA treatment • Th1/Th2 activation and dendritic cell maturation pathways are shown in Figure 9. Upregulat- works that are activated in stable disease patients on ADXS-PSA treatment. ed genes are shown in red tones while downregulated genes are shown in green tones. • Statistical analyses were performed using GraphPad Prism software. The Mann-Whitney U A PDCD1 CD48 CD83 ENTPD1 TNFRSF9 test was used to analyze nonparametric data. * SD * Figure 9. Top signaling pathways by the differentially expressed genes Non-SD Figure 2. Blood draw schedule during ADXS-PSA monotherapy * * DL1 1 x 109 CFU * * * * 9 * DL2 5 x 10 CFU * * 10 1 x 10 CFU ExpressionLevels NanoString DL3 * * Week 1 4 5 6 7 8 9 * * * 2 3 Week 1 Week 3 Week 6 Week 9 Week 1 Week 3 Week 6 Week 9 Week 1 Week 3 Week 6 Week 9 Week 1 Week 3 Week 6 Week 9 Week 1 Week 3 Week 6 Week 9 pre-dose pre-dose pre-dose pre-dose pre-dose

B CXCR3 TBX21 IFNG GZMB * SD * Non-SD ADXS-PSA infusion Blood draw for immune-related gene expression profiling DL1, dose level 1; DL2, dose level 2; DL3, dose level 3; CFU, colony -orming untils * * • RECIST v1.1 criteria were applied to assess the antitumor response (ie, clinical activity) of * ADXS-PSA monotherapy. * • Table 1 summarizes the baseline demographics of the patients who did and did not

achieve clinical activity while undergoing treatment with ADXS-PSA monotherapy. All pa- ExpressionLevels NanoString tients who achieved clinical activity had stable disease. * * * Week 1 Week 3 Week 6 Week 9 Week 1 Week 3 Week 6 Week 9 Week 1 Week 3 Week 6 Week 9 Week 1 Week 3 Week 6 Week 9 Table 1. Baseline demographics of mCRPC patients pre-dose pre-dose pre-dose pre-dose *P<.001 Patients with clinical Patients with no • To identify meaningful differences between the large gene expression data sets generated activity1 clinical activity for ADXS-PSA-treated stable disease and non-stable disease patients, we created a volcano (n=4) (n=9) plot, which plots significance on the y-axis and log2-fold change on the x-axis (Figure 6). The threshold for significance was an adjusted P value of ≤.1. Median Age (range) 69.5 (65-74) 66 (57-80)  In ADXS-PSA-treated stable disease patients, 97 genes were upregulated while 65 genes Dose Level DL1 3 4 were significantly downregulated compared to ADXS-PSA-treated non-stable disease pa- DL2 1 2 tients. DL3 0 3 Figure 6. Significantly differentially expressed genes between stable Race disease and non-stable disease patients on ADXS-PSA treatment CONCLUSIONS White 3 9 stable disease vs non-stable disease • By profiling and quantifying immune-related gene expression in previously treated mCRPC Black or African American 0 0 Downregulated Upregulated Asian 1 0 patients before and after ADXS-PSA monotherapy, we have noted major differences between the pre– and post-treatment immune status of stable disease and non-stable disease patients. ECOG Performance Status 0 3 4  At baseline, stable disease patients have significantly lower expression levels of PDCD1, 1 1 5 ENTPD1, CD48, CD83, TBX21, IFNG, and GZMB than non-stable disease patients. This find-

Prior Therapies ing suggests that low expression levels of genes encoding T cell activation and differentia- value)

Hormonal 3 8 P tion markers at baseline may be associated with stable disease. Radiation 3 5  On ADXS-PSA treatment, stable disease patients have Chemotherapy 1 3 (adjusted

10  significantly higher expression levels of gene profiles indicative of CD4+ T cells, specifi- log

Immunotherapy 1 4 - None 1 0 cally those of the Th1 lineage, as well as of gene profiles indicative of mature antigen presenting cells. These findings underscore the importance of peripheral CD4+ T cell re- Disease site 17 Lymph nodes 2 4 sponses in the generation of effective antitumor immunity by immunotherapies. Bone 2 6  significantly higher expression levels of pDC– and M1 macrophage-related genes and Viscera 0 4 log2-fold change significantly lower expression levels of MDSC– and M2 macrophage-related genes. Laboratory Test Results • To gain a better understanding of the genes that are differentially expressed between  significantly lower expression levels of gene encoding cancer testis/tumor-associated an- Median baseline PSA, ng/mL (range) 26 (4.2-32.2) 19 (5.8-2456) stable disease and non-stable disease patients on ADXS-PSA-treatment, we compared the tigens. This finding suggests that stable disease patients have fewer circulating tumor cells White blood cell count, cells/µl (range) 5150 (3300-6090) 5700 (3200-7000) differentially expressed genes by their functional categories (Figure 7). and thus less tumor burden than non-stable disease patients. Absolute neutrophil count, cells/µl (range) 3195 (2010-3660) 3630 (2110-5226) 1All patients who achieved clinical activity had stable disease  The top functional categories of the upregulated genes in ADXS-treated stable disease • The gene profiles identified in this study not only provide insight into the mechanisms of action patients were antigen processing and presentation (18%), T cell activation and function of ADXS-PSA, but also identify potential predictive and pharmacodynamic biomarkers of clin- RESULTS (13%), and innate immune function (11%). Interestingly, T cell activation and function, NK ical response to ADXS-PSA monotherapy. cell function, and NF-B signaling were functional categories that were unique to upreg- • Using established immune cell-specific gene signatures to identify major T cell effector sub- ACKNOWLEDGMENTS populations, we measured changes in their expression levels in PBMCs of stable disease and ulated genes. non-stable disease patients pre– and post-ADXS-PSA treatment.  The top functional categories of the downregulated genes in ADXS-PSA-treated stable • The patients and families who participated in Part A of the KEYNOTE-046 trial. disease patients were innate immune function (29%), cellular functions (15%), and cyto- • Dynamic changes in the levels of major T cell effector subpopulations were observed in all • The staff from the various clinical sites who are involved in the KEYNOTE-046 trial. kine and cytokine receptors (12%). The complement pathway was the only functional patients on ADXS-monotherapy (Figure 3). • Part A of the KEYNOTE-046 trial was funded by Advaxis, Inc. category unique to down regulated genes.  Although the baseline levels of total T cells were significantly higher in stable disease pa- REFERENCES tients than in non-stable disease patients (P<.001), the fold-change above baseline was  Notably, although all patients had progressive metastatic disease at the start of treat- ment, stable disease patients expressed significantly fewer genes encoding cancer tes- 1. Haas NB, et al. Phase I-II study of ADXS31-142 alone and in combination with pembrolizumab in patients with previously treated metastatic castration-resistant prostate cancer (mCRPC): the KEYNOTE-046 trial. J Immunother Cancer. 2015;3(Suppl 2):P153. similar at each time point between stable disease and non-stable disease patients. 2. Wood LM, Paterson Y. Attenuated Listeria monoctyogenes: a powerful and versatile vector for the future of tumor immunotherapy. Front Cell Infect Microbiol. 2014: https://doi.org/10.3389/fcimb.2014.00051. 3. Rini B. Future approaches in immunotherapy. Semin Oncol. 2014;41(Suppl 5):S30-S40. tis/tumor-associated antigens than non-stable disease patients. This finding suggests that 4. LawrenceYR, Dicker AP. Radiation therapy and the immune system: learning to live together. Future Oncol. 2014;10:777-780. 5. Ritchie ME, et al. (2015). limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. + 2015;43:e47.  The baseline levels of CD8 T cells were also significantly higher in stable disease patients 6. Martinez FO, Gordon S. The M1 and M2 paradigm of macrophage activation: time for reassessment. F1000Prime Rep. 2014;6:3. stable disease patients have less circulating metastatic tumor cells than non-stable dis- 7. Tel J, et al. Human plasmacytoid dendritic cells are equipped with antigen-presenting and tumoricidal capacities. Blood. 120:3936-3944. 8. Kini Bailur J, et al. Prognostic impact of high levels of circulating plasmacytoid dendritic cells in breast cancer. J Transl Med. 2016;14:151. + 9. Chevolet I, et al. Clinical significance of plasmacytoid dendritic cells and myeloid-derived suppressor cells. J Transl Med. 2015;13:9. than in non-stable disease patients (P<.001). CD8 T cell levels increased on treatment in 10. Mantovani A, et al. Tumour-associated macrophages as treatment targets in oncology. Nat Rev Clin Oncol. 2017;14:399-416. 11. Martinez FO, et al. Transcriptional profiling of the human -to-macrophage differentiation and polarization: new molecules and patterns of ease patients. gene expression. J. Immunol. 2006;177:7303-7311. + 12. Rőszer T. Understanding the mysterious M2 macrophage through activation markers and effector mechanisms. Mediators Inflamm. all patients, with CD8 T cell levels peaking at week 3 in stable disease patients and at 2015;2015:816460. 13. Linstedt M, et al. Gene family clustering identifies functionally associated subsets of human in vivo blood and tonsillar dendritic ce lls. J. Immunol. 2005;175:4839-4846. • The innate immune function category was a top category for upregulated and downregu- 14. Vuk-Pavlovic S, et al. Immunosuppressive CD14+HLA-DRlow/- in prostate cancer. Prostate. 2010;70:443-455. 15. Monu NR, Frey AB. Myeloid-derived suppressor cells and anti-tumor cells: a complex relationship. Immunol Invest. 2012;41:595-613. week 6 in non-stable disease patients. 16. Zhao F, et al. S100A9 a new marker for monocytic human myeloid derived suppressor cells. Immunology. 2012;136:176-183. 17. Spitzer MH, et al. Systemic immunity is required for effective . Cell. 2017;168:487-502.

Top View