Oncolytic Poliovirus Shows Sustained Survival Rates for Recurrent GBM

Darell D. Bigner, MD, PhD

Treatment with the recombinant oncolytic poliovirus PVSRIPO elicited a sustained 2- and 3-year overall survival (OS) rate of 21% (95% CI, 11%-33%) for patients with recurrent grade IV malignant glioblastoma (GBM), according to phase I findings published in the New England Journal of Medicine (NEJM).¹

After 27.6 months of follow-up for 61 patients, the median OS was 12.5 months with PVSRIPO (95% CI, 9.9-15.2) compared with 11.3 months in a historical control group (95% CI, 9.8-12.5). The OS rate continued to decline beyond 1 year in the historical control group but remained consistent at 21% for PVSRIPO, representing a plateau of the curve. The 2-year OS rate in the historical arm was 14% and the 3-year rate was 4%.

"With the survival rates in this early phase [trial] of the poliovirus therapy, we are encouraged and eager to continue with the additional studies that are already underway or planned," senior study author Darell D. Bigner, MD, PhD, emeritus director of The Preston Robert Tisch Brain Tumor Center at Duke, said in a statement. "Glioblastoma remains a lethal and devastating disease, despite advances in surgical and radiation therapies, as well as new chemotherapy and targeted agents. There is a tremendous need for fundamentally different approaches."

PVSRIPO, which is manufactured at the NCI’s Frederick facility, consists of a genetically modified nonpathogenic version of the oral poliovirus Sabin type 1. In the therapy, the internal ribosomal entry site (IRES) on the poliovirus was replaced with the IRES from human rhinovirus type 2 (HRV2), which eliminates neurovirulence. Once administered, the therapy enters and begins duplicating within cells that express CD155/Necl5, which is an onco-fetal cell adhesion molecular that is common across solid tumors.

Based on earlier findings from the phase I study,2 in May 2016 the FDA granted PVSRIPO a breakthrough therapy designation as a potential treatment for patients with recurrent GBM. In this earlier analysis of the data, the 2-year OS rate 24% for 24 patients treated with the oncolytic virus. The 3-year OS rate at this analysis was 8%.

"The Duke and NCI teams collaborated extensively on the preclinical work, which has culminated in these clinical trial results," Jim Doroshow, MD, deputy director for Clinical and Translational Research at NCI, said in a statement. "This promising approach in glioblastoma therapy exemplifies the strategic investment made by the NCI to support the development of new therapies such as this one from research discovery into clinical trials."

In the phase I study, 61 patients were treated with PVSRIPO at a variety of doses. Treatment was administered directly into the tumor by a catheter, which was implanted utilizing stereotactic guidance. Patients in the trial received a poliovirus immunization booster 2 weeks prior to treatment with PVSRIPO.

The initial 9 enrolled patients had escalating doses of PVSRIPO, which led to inflammation, seizures, cognitive, disturbances, and other adverse events (AEs) at the highest dose levels. Subsequently, the dose of the virus therapy was reduced, to mitigate AEs in a dose expansion cohort, with a dose of 5.0 x 107 50% tissue-culture infectious doses (TCID) selected for a phase II study.

Across all 61 enrolled patients, 41% were female and the primary Karnofsky performance status (KPS) was 90 (69%), with 26% having a KPS of 80 and 1 patient having a KPS of 70. Overall, 77% of patients had a gross total resection at diagnosis and 23% had received prior bevacizumab. IDH1 was not mutated for 74% of patients and MGMT was methylated for 23% of patients at the time of treatment. Similar characteristics were selected for the control group, with fewer having a gross total resection (66%) and a higher rate of prior bevacizumab treatment (41%).

At data cutoff in March 20, 2018, durable radiographic response to PVSRIPO was experienced by 8 patients (13.1%). Of these responses, 2 patients had a complete response (CR). The first patient with a CR continued to remain alive at 70.4 months following infusion. The second patient was alive at 15.1 months following PVSRIPO infusion. Stable and partial responses ranged in duration from 26 months to 60 months.

Of responding patients, 3 went on to receive a short course of bevacizumab at 7.5 mg/kg every 3 weeks. These patients remained disease free at the analysis. For each of these patients, the current duration of ongoing disease-free survival was 34.1+, 27.1+, and 15.4+ months.

PVSRIPO was not associated encephalomyelitis, poliomyelitis, meningitis, or systemic autoimmune reactions. Localized inflammation was managed using glucocorticoids, if neurologic AEs were present.

At the lower doses explored in the dose expansion portion of the study, grade 3/4 treatment-related AEs (TRAEs) were relatively rare, with 17% having a grade 3 event. The most common grade 3 TRAEs were pyramidal tract syndrome (8%) and gait disturbance, dystonia, headache, seizure, confusion, and delusions in 2% of patients each.

There was one possibly PVSRIPO-related death from grade 4 cerebral edema and grade 5 seizure in a patient receiving a lower dose. This was in the context of confirmed progression at 4.8 months. Additionally, one patient treated with the phase II dose of PVSRIPO went on to receive anticoagulation and bevacizumab, which resulted in a fatal intracranial bleed that was attributed to the VEGF inhibitor.

"Similar to many immunotherapies, it appears that some patients don't respond for one reason or another, but if they respond, they often become long-term survivors," study co-author Annick Desjardins, MD, Department of Neurosurgery, Duke, said in a statement. "The big question is, how can we make sure that everybody responds?"

A phase II study is currently enrolling patients to explore PVSRIPO with or without lomustine for adult patients with grade IV malignant glioma. The study plans to enroll 62 participants, with an anticipated completion date in 2021 (NCT02986178). A phase Ib study is also exploring PVSRIPO alone for children with recurrent malignant glioma (NCT03043391). Additionally, given the availability of CD155/Necl5 on solid tumors, trials are anticipated across other types of cancer.

References

1. Desjardins A, Gromeier M, Herndon JE, et al. Recurrent Glioblastoma Treated with Recombinant Poliovirus. N Engl J Med. DOI: 10.1056/NEJMoa1716435.
2. Desjardins A, Sampson JH, Peters KB, et al. Oncolytic polio/rhinovirus recombinant (PVSRIPO) against recurrent glioblastoma (GBM): Optimal dose determination. J Clin Oncol. 2015;33 (suppl; abstr 2068).