Therapeutic Immune Modulation against Solid Cancers with Intratumoral Poly-ICLC: A Pilot Trial

Abstract

Purpose: Polyinosinic-polycytidylic acid-poly-l-lysine carboxymethylcellulose (poly-ICLC), a synthetic double-stranded RNA complex, is a ligand for toll-like receptor-3 and MDA-5 that can activate immune cells, such as dendritic cells, and trigger natural killer cells to kill tumor cells.Patients and Methods: In this pilot study, eligible patients included those with recurrent metastatic disease in whom prior systemic therapy (head and neck squamous cell cancer and melanoma) failed. Patients received 2 treatment cycles, each cycle consisting of 1 mg poly-ICLC 3× weekly intratumorally (IT) for 2 weeks followed by intramuscular (IM) boosters biweekly for 7 weeks, with a 1-week rest period. Immune response was evaluated by immunohistochemistry (IHC) and RNA sequencing (RNA-seq) in tumor and blood.Results: Two patients completed 2 cycles of IT treatments, and 1 achieved clinical benefit (stable disease, progression-free survival 6 months), whereas the remainder had progressive disease. Poly-ICLC was well tolerated, with principal side effects of fatigue and inflammation at injection site (<grade 2). In the patient with clinical benefit, IHC analysis of tumor showed increased CD4, CD8, PD1, and PD-L1 levels compared with patients with progressive disease. RNA-seq analysis of the same patient's tumor and peripheral blood mononuclear cells showed dramatic changes in response to poly-ICLC treatment, including upregulation of genes associated with chemokine activity, T-cell activation, and antigen presentation.Conclusions: Poly-ICLC was well tolerated in patients with solid cancer and generated local and systemic immune responses, as evident in the patient achieving clinical benefit. These results warrant further investigation and are currently being explored in a multicenter phase II clinical trial (NCT02423863). Clin Cancer Res; 24(20); 4937-48. ©2018 AACR.

Conflict of interest statement

Disclosures of Potential Conflicts of Interest: N B. is on the scientific advisory board of Neon, CPS companion diagnostics, Genentech and Curevac, Inc. R.S. works as clinical scientist at Genentech, Inc. A.S. is the scientific director of Oncovir, Inc. No potential conflicts of interest were disclosed by other authors.

©2018 American Association for Cancer Research.

Figures

Figure 1. Protocol schema

Patients received two cycles of poly-ICLC treatment, each cycle including a priming and boosting treatment course (Figure 1). In cycle 1, patients were treated with 1 mg of Poly-ICLC 3 times weekly during week 1-2 (“priming treatment”) into the same lesion. During weeks 3-9, patients were treated with IM maintenance boosters biweekly, followed by a rest week (week 10) without treatment. This 10-week cycle was repeated in cycle 2, followed by a “no treatment rest period” during weeks 20 – 26. At week 26, patients were assessed and response determined, and those patients with CR, PR, or SD were offered option of maintenance therapy (from week 27-36 with administration of 1 mg IM poly-ICLC twice weekly). Tumor biopsies were performed at baseline, week 3, and week 26 if possible. Pre- and post-vaccination tumors were evaluated by quantitative multiplex immunohistochemistry (IHC) and RNA sequencing. Blood samples were collected at baseline and throughout treatment cycles (as indicated) for immune response evaluations.

Figure 2. – Clinical data

(A) Post-injection site demonstrating necrosis after IT poly-ICLC treatment (B) Patient 002 was a HNSCC patient who showed clinical benefit (stable disease) with CT scans over the treatment course demonstrating stable disease (PFS over 6 months).

Figure 3

Immunohistochemistry (IHC) of tumor biopsies taken at baseline (top row) and after 6 IT injections (bottom row). In patient 008 (progressive disease, left column), quantitative IHC showed unchanged or decreased levels of CD4, CD8, PD-1, and PDL-1 over treatment period In the patient 002 with clinical benefit (stable disease, right two columns), IHC analysis of tumor showed increased CD4 (60x), CD8 (10x), PD1 (20x) and PDL1 (3x) (also see supplemental table 1), as well as marked increases in immune cells (CD86 antigen-presenting cells, CD68 macrophages / monocytes, CD16 natural killer cells, HLA encoding the major histocompatibility complex (MHC) proteins) post treatment.

Figure 4

A. Hierarchical clustering of PBMC RNA expression from patients with progressive disease PD and stable disease SD. I – baseline cluster, including samples, taken at “screen”, “rest” or during intra-muscular injection time points, samples are: 1, 4, 9, 11, 12, 14, 17. II – intermediate cluster, including samples from PD patients, taken during intratumoral injections: 13 and 15; and samples from SD patient, taken during intra-muscular injections: 3 and 8. III – inflamed cluster, including samples from SD patient, taken after intratumoral injections: 2, 5 and 6. Each column represents patient at specific treatment time point, as numbered below. For example, 002- C1W1 is patient 002 at time point Cycle 1 week1. Samples are: 1 – 002_Screen10 – 002_C2W202 – 002_C1W111 – 002_C2W263 – 002_C1W312 – 004_Screen4 – 002_C1W713 – 004_C1W25 – 002_C1W1014 – 008_Screen6 – 002_C2W1115 – 008_C1W17 – 002_C2W1216 – 008_C1W38 – 002_C2W1317 – 008_C1W79 – 002_C2W1718 – 008_C1W10 B. Scatter plot of log2-transformed fold changes, inferred from differential gene expression analyses of two comparisons: PBMC samples 15 and 13 versus cluster I, contrast called “PD”; PBMC samples 2 and 5 versus cluster I, contrast called “SD”. Statistically significant changes (false discovery rate < 0,05) are plotted.

Figure 5

BTM enrichments of selected genesets in PBMC of patients 002 with SD and patient 004 and 008 (PD) collected at different time points during poly-ICLC treatment cycle. (A) unsupervised hierarchical clustering of selected BTM genesets (false discovery rate FDR

Figure 6

Correlation of NK cell activation signatures and DC infiltration in blood with specific chemokine expression at tumor site and survival analysis of selected gene signatures. Survival analysis of head and neck squamous cell carcinoma (HNSC) and skin cutaneous melanoma (SKCM) patients from the cancer genome atlas (TCGA). Analysis is done by web-server TIMER, ranking patients by selected gene expression and using top 30% and bottom 30% of patients for survival estimation. (A) DC surface markers are highly up-regulated in blood of SD patient (RIGHT). DC presence correlates with XCL1/2 high expression in the blood of same patient (LEFT). (B) Up-regulation T/NK-cell co-stimulatory cytokines IL15 and IL23 at tumor site of patient 002 upon completion of poly-ICLC treatment (RIGHT). Patients 004 and 008 have up regulated pro-tumorigenic cytokines CXCL1 and CXCL5 at the beginning and the end of treatment (LEFT). Each column represents patient at specific treatment time point, as numbered on Figure 4. (C) Increased expression of IL15, IL23A, Clec9A, XCR1, XCL1, XCL2 correlates with better prognosis in HNSC and SKCM (p-value