Oncolytic Virotherapy: Single Cycle Cures or Repeat Treatments? (Repeat Dosing Is Crucial!)

Main Text

Repeat dosing is essential for the success of oncolytic virotherapy because it is a form of cancer immunotherapy. Therefore, the focus of optimizing the efficacy of oncolytic viruses (OVs) needs to be maximizing activation of anti-tumor immunity, and this simply cannot be achieved by a “single shot”.

A central tenet of Polly Matzinger’s original, paradigm-shifting, 1994 “danger theory” of the immune response,¹ which proposed that the immune system does not distinguish between self and non-self, but instead discriminates between “dangerous” and “safe” by recognition of, for example, pathogens, was that the danger signal has to be repetitive to fully activate the immune response. In the context of a virus (as the pathogen) alerting the immune system to cancer (as the danger), this rule remains critical. In the current parlance, to ask a single treatment to convert an immunologically “cold” tumor to “hot” is simply too much to ask.

There are arguments proposed to justify why “once is enough”. The first is that, because OVs will replicate and spread within a tumor over time, a single treatment evolves into a self-propagating, cytotoxic, and immunogenic therapy. However, there is no evidence for the dependence of OV therapy on viral replication in humans, and the degree of replication in patients and real clinical samples (unlike in mice) is actually very limited.² Moreover, there are murine models in which replication-incompetent viruses can still be effective treatments.³ Therefore, and particularly as our field moves into an increasingly clinical context, even the term “oncolytic virotherapy” is arguably a misnomer. OVs were named at a time when therapy was thought to be inextricably linked to direct viral cytotoxicity, and, today, perhaps “immunovirotherapy” is a more accurate description of how these agents are actually working for the benefit of patients.

However, the main argument against repeat treatments has been that multiple dosing will trigger an anti-viral immune response that will restrict therapy. To address this concern, testing of the first approved clinical agent, talimogene laherparepvec (T-Vec), has been restricted to intratumoral administration, despite the fact that systemic herpes virotherapy can be effective in pre-clinical models.⁴ While it is true that direct injection does avoid the concern that systemic virus will be neutralized in the blood before it can reach its tumor target, the reality is that intratumoral injection is a challenge for widespread application in the real, clinical world. Oncologists prefer to give intravenous agents, including viruses, because this route can potentially access multiple sites of macro- and micro-metastasic disease. Critically, clinical trials have now shown that reovirus, adenovirus, vaccinia, and coxsackie OVs can all be detected in tumors after systemic administration, at least providing the basis for systemic therapy, even though definite evidence of efficacy from randomized trials is still awaited.5, 6, 7 For reovirus, this applies even in the face of a pre-existing anti-virus neutralizing antibody (NAB) response, which is present because reovirus is a ubiquitous, non-pathogenic virus to which we have all been exposed in childhood. The mechanism by which reovirus is protected from NAB in the blood seems to be protective cell carriage by blood cells, such as monocytes,⁸ and virologists will tell you that this is no surprise, as such strategies are common to pathogenic viruses, which need to evade anti-viral immunity to infect and spread. There is much more to learn about the interaction between different OVs and different components of the immune response (including, for example, anti-viral cytokines and complement as well as NAB), but the argument that the anti-viral immune response (prior to, or induced by, repeat treatments) is an inevitable barrier to successful therapy is now untenable. Even more provocatively, there is pre-clinical evidence that the anti-viral immune response can actually enhance the efficacy of OV therapy, for Newcastle disease virus, maraba, and reovirus, although the underlying mechanisms remain poorly understood.9, 10 Potential explanations of why anti-viral immunity may be paradoxically beneficial include the idea that the adjuvant-like properties of the anti-viral innate responses are critical for the initial priming of anti-tumor immune responses or that the immune response to the virus within the tumor supports recruitment of anti-tumor effector immune cells.¹¹

We still know far too little about mechanisms of action of OVs, but, as the field becomes more immunology and less pure molecular virology, the emphasis is changing. Not all viruses and routes of delivery will have the same immunological consequences, and a better understanding of how single-stranded RNA (ssRNA), double-stranded RNA (dsRNA), and DNA viruses are delivered, sensed, and shape an anti-tumor versus anti-viral immune response will be critical in optimizing treatment schedules. However, whatever mechanisms are uncovered, multiple doses will remain an essential part of ensuring that the tumor is seen as maximally “dangerous” by the immune system and that the virus immunologically heats up the tumor as much as possible. Even if NABs are problematic in some contexts, multiple virus treatments will simply need to be “front loaded”, i.e., given repeatedly and early in the schedule, before NAB levels rise. Finally, and again to be clinically pragmatic, the reality is that a majority of the lead clinical agents (herpes, vaccinia, coxsackie, adenovirus, maraba, reovirus, and others) are all being given as multiple treatments within trials, and this momentum of repeat dosing is now unstoppable.

In this issue of Molecular Therapy, the accompanying Commentary by Russell argues that patients do not like multiple treatments. However, viruses are, in fact, relatively non-toxic, and patients do not mind repeat treatments if they work. Clinical pragmatism and maximum immune activation, which is how these agents work, will require more than one treatment, although I would accept that multiple rounds of treatment over a long period of time may well run into the law of diminishing returns and not be necessary.