Upreti, Deepak

Abstract Introduction. Voyager-V1 (VV1) is an oncolytic vesicular stomatitis virus engineered to express human interferon beta (IFNβ) to enhance cellular antitumor immune responses and tumor selectivity. VV1 also contains the human sodium iodide symporter (NIS) as an imaging gene. We report here the novel use of virus-encoded IFNβ as a PD marker using correlative data from three Phase I trials of VSV-IFNβ-NIS in patients with refractory cancers (n=46). Methods. 46 patients with solid tumors (n=34) and hematological malignancies (n=12) received 1 dose of VV1 either intratumorally (ITu) or intravenously (IV) at doses ranging from 3 x 106 to 5 x 1010 TCID50. Plasma IFNβ levels were collected pre-treatment, 4 hours post-infusion, Day 2 (24-hour), Day 3, 8, 15 and 29 (IT only). Samples were processed using a standard IFNβ specific ELISA kit. Results. ITu dose escalation is complete with 27 patients treated and no DLTs. IV escalation is ongoing at 5 x 1010 TCID50 with 19 patients treated to date. In the ITu study, plasma IFNβ levels at 24h were undetected at the lowest dose levels (up to 1 x 107 TCID50), and became detectable from 3 x 107 TCID50. In the IV patients, IFNβ was detectable at all dose levels (5 x 109 through 5 x 1010 TCID50) with the highest peak and longest duration in a patient with metastatic endometrial cancer coincident with shrinkage of multiple tumors. The IFNβ produced by virus-infected cells can be differentiated from the acute innate antiviral responses by magnitude of response and AUC as the majority of the inflammatory cytokines returned to baseline by 48h. Peak IFNβ levels were variable between patients, likely reflecting heterogeneity in tumor susceptibility to VV1, ranging from 1.4pg/mL to 656pg/mL across 6 patients (mean 153pg/ml) at the highest ITu dose. Plasma IFNβ 24 hours post-therapy of >20pg/mL appears to predict for RECIST 1.1-evaluated SD vs PD, p=0.048 in the ITu patients. Peak IFNβ ranged from 18 to 1700 pg/mL across 9 patients (mean=442pg/ml) at 1.7 x 1010 in the IV study. Peak IFNβ was highest in two cases of endometrial cancer (1500 and 1700 pg/ml). The patient with the highest IFNβ levels on the IV trial showed 16.7% tumor shrinkage at the first tumor evaluation. SPECT imaging, which shows location of viral replication, was positive in 50% of ITu injected tumors, also validating of VV1 infection of target cells. To date, SPECT images were negative in the IV trial despite IFNβ positivity, reflecting IFNβ as a more sensitive PD marker of viral infection. IHC staining of tumor biopsies collected pre-treatment and 1 month after VSV showed increased numbers of CD3, CD8, CD68, PDL1 or PD1 in some injected or noninjected tumors. Other immune markers and tumor gene signatures are also being evaluated. Conclusions. Plasma IFNβ has emerged as a simple and convenient biomarker of viral replication in tumors. IFNβ will be used in future studies as a PD marker to assess the impact of immune-modulating combination drugs with VV1. Citation Format: Timothy P. Cripe, Jamie Bakkum-Gamez, Jaime R. Merchan, Martha Q. Lacy, Manish R. Patel, Steven Powell, James Strauss, Lianwen Zhang, Toshie Sakuma, Memy Diaz, Nandakumar Packiriswamy, Deepak Upreti, Bethany Brunton, Dragan Jevremovic, Stephen J. Russell, Alice Bexon, Kah-Whye Peng. Rational design of an oncolytic virus permits use of interferon beta as a pharmacodynamic marker for clinical application [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr CT090.

Abstract Introduction Voyager-V1TM is derived from VSV, a bullet-shaped negative sense RNA virus with very low human seroprevalence; it is engineered to selectively replicate in and kill human cancer cells. Voyager-V1 encodes the human IFNβ gene to boost antitumoral immune responses and the thyroidal sodium iodide symporter NIS gene to permit noninvasive imaging of virus spread. FIH studies are underway via both IV and IT routes. Here we report safety and preliminary correlative data from the FIH IT study. Methods and objectives This is a classical 3+3 phase 1 design, using escalating single IT viral doses from 3 x 106 to 3 x 109 TCID50 into one target lesion. The primary objective is safety and tolerability, monitored by committee. Other objectives include PK by RT-PCR for viral genomes, serum IFNβ levels, Tc-99m SPECT/CT imaging to monitor virus infection in injected lesions, peripheral blood immunophenotyping with 11-color flow cytometry for activation markers on T cells, T-regs, NK cells, and MDSCs, and serial biopsies to assess the tumor microenvironment (TME). IHC was performed on tumor biopsies for CD3, CD8, CD4, FoxP3, CD68, PD-1 and PDL-1 pre and post treatment in non-injected and injected lesions. CD45 staining is ongoing. Results Dose level 3 is ongoing (n=8+). No DLTs have been observed to date. Most patients were male (75%), white (100%), with ECOG PS1 (75%) and a median of >6 lines of prior systemic therapy. AEs (in 63% patients) reported as related to study drug were mild-moderate, short-lived, and consisted of G1 fever, chills, hot flashes, nausea, vomiting, hyperhidrosis and G2 fatigue. Most patients had some mild AEs (pain, bruising, subclinical pneumothorax) related to biopsy and IT injection procedures and one had an SAE related to biopsy (G2 pneumothorax). There was no difference in incidence, intensity or duration of related AEs between dose levels so far. In patients analyzed to date, at the first three dose levels of 3 x 106, 1 x 107 and 3 x 107 TCID50 (n=7), viremia and serum IFNβ were below levels of detection after IT injection. There is evidence in all patients of an increase in PD-1 expression on CD4 and CD8 T cells, suggesting T cell activation post- single virus injection. One patient at dose level 2 had increased CD4 and CD8 T cells in the peripheral blood. TME analysis revealed increased infiltration of CD3 and CD8 cells in tumors of some patients. In addition, there are changes in numbers of FoxP3 and CD68 positive cells in the injected lesions compared to the pre-treatment and non-injected lesion. Two patients at dose level 3, one with an injected adrenal metastasis of colorectal cancer and another with a chest wall lesion from a head of pancreas primary, have positive SPECT/CTs on days 3-15 showing viral replication in tumor plus concomitant lymphocyte/neutrophil trafficking in the periphery. Conclusions IT injection of a single dose of the novel oncolytic virotherapy Voyager-V1 has proven safe at doses up to 3 x 107 TCID50. There was evidence of T cell activation with increased PD-1 expression in peripheral blood lymphocytes after a single IT injection. Tumor biopsies indicate increased lymphocyte infiltration. Two patients at the highest dose tested show signs of viral replication in the injected lesion plus evidence of systemic impact on leukocyte populations. Citation Format: Steven Powell, Manish R. Patel, Jaime R. Merchan, Timothy P. Cripe, James Strauss, Rosa M. Diaz, Nandakumar Packiriswamy, Bethany A. Brunton, Deepak Upreti, Rehan Khan, Lukkana Suksanpaisan, Rianna Vandergaast, Stephen J. Russell, Alice S. Bexon, Kah Whye Peng. Preliminary correlative and clinical data from a first-in-human (FIH) study of the intratumoral (IT) oncolytic virotherapy, Voyager-V1, in patients with solid tumors [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr CT051.

218 Background: Evidence from mouse studies shows that Oncolytic viral mode of tumor cell killing can boost the cytotoxic T lymphocyte response against tumor associated antigens (TAA), leading to ‘bystander’ killing of uninfected tumor cells. Methods: To investigate whether oncolytic virotherapy can boost immune responses to tumor antigens in human subjects we studied T cell responses to 10 different TAA reported to have high expression in patients with multiple myeloma (n = 10) before and after intravenous administration of an oncolytic measles virus (MV-NIS). Baseline and 6 week post-therapy PBMC samples were stimulated with peptides specific to each of the selected TAA and their responses to this antigenic stimulation were measured using IFNγ ELISPOT. Results: Despite their prior exposure to multiple immunosuppressive treatment regimens, some of the tested patients had higher T cell responses to some of the TAA even prior to MV-NIS. Baseline T cell responses against MAGEc1 were present in 50% of the patients, and against hTERT and NYESO-1 in 30% of the patients. Furthermore, MV-NIS treatment significantly boosted T cell responses against NYESO-1, MAGEc1, MAGEA1 and MUC1 in 40% of the patients. Interestingly, in one outlier patient who achieved a stringent and durable complete disease remission after MV-NIS therapy, strong baseline T cell responses were present against 8 of the 10 tested TAA but did not change significantly after virotherapy. Based on the T cell responses to the TAA, three distinct phenotypes can be identified. Phenotype 1 was characterized by positive T cell responses to TAA prior and post therapy, Phenotype 2 was characterized by positive T cell responses to TAA only post therapy and phenotype 3 was characterized by negative T cell responses to TAA prior and post therapy. Conclusions: Our data suggests oncolytic measles virotherapy boosted the cytotoxic T lymphocyte response against tumor associated antigens in some of the patients. Identifying patient’s phenotype based on responsiveness to the TAA could be useful marker for identifying the immune responsive state in multiple myeloma patients and identify patients who can be most responsive to oncolytic measles virotherapy. Clinical trial information: NCT00450814.