Phase I Trial of Systemic Administration of Vesicular Stomatitis Virus Genetically Engineered to Express NIS and Human Interferon Beta, in Patients with Relapsed or Refractory Multiple Myeloma (MM), Acute Myeloid Leukemia (AML), and T-Cell Neoplasms (TCL)

Background: Oncolytic virotherapy is a novel immunomodulatory therapeutic approach for relapsed refractory hematologic malignancies. The Indiana strain of Vesicular Stomatitis Virus (VSV) has been engineered to encode interferon beta (IFNβ) and sodium iodine symporter (NIS) to produce VSV-IFNβ-NIS. The virally encoded IFNβ serves as an index of viral proliferation and enhances host anti-tumor immunity. The NIS is inserted as a reporter gene into the viral genome to enable noninvasive monitoring of viral spread using PET/CT imaging. Here we present results of Arm A (patients with low tumor burden) of the Phase I clinical trial (NCT03017820) of intravenous administration of VSV-IFNβ-NIS for patients (pts) with relapsed refractory hematological malignancies: MM, AML, and TCL.

Methods: This was a classical 3+3 phase I trial, starting at 5x10^9 TCID50 (50% tissue culture infectious dose) level 1 through 1.7x10^11 TCID50 dose level 4 given as a single IV dose. The primary objective was determining the maximum tolerated dose of VSV-IFNβ-NIS alone; secondary objectives include estimating the safety profile and preliminary efficacy. Correlative objectives include monitoring the pharmacodynamics of viral replication through SPECT/CT imaging with NIS gene, viremia, virus shedding and changes in the immune profile of peripheral blood leukocytes. Adverse events (AEs) are reported based on CTCAE v4; cytokine release syndrome (CRS) grading was based on Lee (Blood 2014; 124(2):188-195) criteria.

Results: From April 2017, 12 pts have received IV VSV-IFNβ-NIS in Arm A: MM (7), TCL(4) and AML(1); 3 pts were treated at each dose level (DL) 1 to 4. At DL1, there were five grade 3 hematologic AEs (leukopenia [1], neutropenia [1], lymphopenia [3]) and no grade 3+ non-hematologic AEs. At DL2, three grade 3 hematologic AEs (anemia [2], lymphopenia [1]), four grade 4 hematologic AEs (neutropenia [1], leukopenia [1], lymphopenia [1] and thrombocytopenia [1]). There was 1 event of grade 2 CRS at DL2, two grade 3 non-hematologic AEs (nausea [1], dehydration [1]). At DL3, there was one grade 4 hematologic AE (neutropenia [1]), three grade 3 hematologic AEs (lymphopenia [2], leukopenia [1]), and one grade 3 non-hematologic AE (vasovagal reaction [1]). One grade 2 and grade 1 CRS by Lee criteria were observed. At DL4, there were three grade 3 hematologic AEs (lymphopenia [1], neutropenia [1], thrombocytopenia [1]), and three grade 3 non-hematologic AEs (AST increase [1], skin infection [1], soft tissue infection [1])and two grade 2 CRS. There were no dose limiting toxicities. There was one partial response (DL2) and one complete response (DL4); 3 pts had stable disease, 7 pts progressed. The MM pts receiving VSV-IFNβ-NIS at DL4 had stable disease. Six of the 7 MM pts demonstrated transient reduction in serum level of involved free light chain (LC) after VSV-IFNβ-NIS infusion, with median 20.4% reduction of the involved LC within 48 hrs. Viremia was detectable in all pts at the end of infusion, to varying levels at 30 mins, 1, 2, 4, 24, 48h or 72 hrs post infusion. No persistent viremia was observed. No infectious virus was recovered in buccal swabs or urine and neutralizing anti-VSV antibodies were present by day 29. The IFN levels were detectable within 30 minutes of infusion and peaked between 4 and 48 hrs. The pts with a diagnosis of TCL mounted higher hIFNβ levels within 48 hrs. Remarkably, the TCL pt with CR (DL4) mounted a peak hIFNβ response of 18213.3pg/ml at 48 hrs post infusion, approximately 15-fold higher than any other pt enrolled in the study; and highest neutralizing antibody titer (1:40960).

Conclusion: VSV-IFNβ-NIS can be safely administered by intravenous infusion in pts with hematological malignancies, up to a dose level of 1.7 x 10¹¹ TCID₅₀, without dose limiting toxicities and no CRS higher than grade 2. In this small phase 1 study, pts with T-cell lymphoma demonstrated the strongest signal of response, with one confirmed partial response at DL2 and one complete response at DL4. IFNβ is an early biomarker of tumor susceptibility; the magnitude of IFNβ elevation appears to correlate with treatment response. There were no confirmed responses in pts with MM, but the 20.4% transient reduction in LC suggests that combination strategies with VSV-IFNβ-NIS may result in deeper and sustained responses in MM. Future trials investigating combination strategies with immune modulatory drugs are being planned.