An Intravenous Stimulus Package for Oncolytic Virotherapy

A recently published phase I clinical trial shows that an oncolytic vaccinia virus, JX-594, can reach and infect tumors following a single intravenous infusion, despite the barriers to systemic virus spread.¹ Direct evidence of virus localization in tumors, along with a portfolio of supportive data, suggests that virus can also replicate selectively in tumors. These data come extremely close to validating the basic principles underlying the development of replicating viruses for cancer therapy.^2,3

The impending collapse of the world's financial apparatus⁴ is generally unnerving, but the most concerning aspect is the radically divergent solutions proposed by the politicians and economists whose job it is to pull us back from the precipice—should we spend our way out of recession or make deep cuts to balance the budget?⁵ Although the success, or otherwise, of oncolytic virotherapy (OV) may have rather less global significance to most people than the financial meltdown, to analysts monitoring the “futures” of novel cancer therapies, the debates surrounding OV appear to have become as contradictory, and bewilderingly polarized, as those addressing the current politico-economic challenge.

The theory behind the development of oncolytic viruses is disarmingly simple. An arsenal of viruses would be developed, with genetically engineered, or natural, tropism for tumor cells. By seeding even just a few particles in a tumor, viral replication would amplify the therapeutic dose to high levels. Progeny virus would then rip through cancer cells in waves of lytic replication before reaching normal cells, at which point the infection would die away.^2,3 The result would be a novel class of drug with a high therapeutic index. In theory, OV should outperform small-molecule pharmaceuticals because of its intrinsic ability to amplify the therapeutic signal once delivered to the tumor. In addition, OV would, appealingly, use our wealth of knowledge on the molecular defects in cancer cells as the basis for the design of the tumor specificity of infection and replication.⁶

After initial concerns of creating a nightmarish new viral plague (cf. the 2007 movie I Am Legend), the market reaction to OV was favorable. However, with time, hurdles have emerged, and, as with the economy, the debates have begun. For example, is it possible for OV to be effective in the presence of a person's immune system, in that a strong innate response to the virus will inevitably lead to extinction of viral replication, spread, and cancer cell lysis?^7,8 Alternatively, and paradoxically, others have suggested that much of the efficacy of OV is directly attributable to the actions of an intact immune system,^7,9 given that antivirus innate immune reactivity can lead to concomitant clearance of tumor cells.¹⁰ Similarly, neutralizing antibodies (NAbs) to oncolytic viruses will already be present either in prevaccinated or infected subjects or will inevitably be formed de novo by virus administration. Is this good (protecting the patient from widespread dissemination and toxicity) or bad (inhibiting access of the circulating virus to disseminated tumor sites)?¹¹ The apparently mutually exclusive nature of these opposing viewpoints is alarmingly similar to that of the contradictory standoff between fiscal economists. In both cases, the successful way forward will almost certainly come from a middle-ground approach, incorporating elements of both extremes.

Compromise of opposing viewpoints is often more readily achieved when there is overall optimism about the bigger picture. In the case of OV, such optimism has been lacking because of the absence of an answer to the Great Unanswered Question: will it be possible to deliver an oncolytic virus, in a truly systemic fashion, in the face of the full force of a patient's immune system and other antiviral barriers, to metastatic sites where it can replicate specifically in tumor? Although an unequivocal answer to this question has not yet emerged, there have been stimulus packages to shore up confidence. Thus, various types of cells have been used successfully to carry viruses to tumors, even in the presence of NAb- and complement-mediated antiviral clearance mechanisms.^12,13 In addition, combination approaches have enhanced both delivery and efficacy.^14,15,16,17 Nevertheless, in the absence of successful clinical data addressing the Great Unanswered Question, the long-term value of OV as the next big anticancer drug has remained in doubt.

In their high-profile article, Breitbach et al. report on a phase I dose-escalation clinical study in which JX-594, an oncolytic poxvirus, was delivered intravenously to 23 patients with advanced treatment-refractory solid tumors.¹ This is by no means the first report of the use of OV in humans, or even of a trial using systemic delivery,¹⁸ so why, you might ask, is this such a big deal? The answer is that the study provides unequivocal evidence that intravenous delivery of this virus can result in delivery to the tumor. Ten days after a single intravenous infusion, virus was detected in tumor biopsy specimens by both polymerase chain reaction and immunohistochemistry. Interestingly, these results were obtained only at the highest dose level. Viral gene expression (actually from the β-gal transgene) was detected within tumor tissue, surrounded by very low levels of expression in normal tissue. These findings have added value in several ways. All the subjects had previously been vaccinated against smallpox, although only six had detectable antibody. Importantly, virus delivery to tumor was shown in at least one of these individuals, showing that vaccinia virus can be delivered systemically despite the presence of NAbs. Moreover, a dose-related antitumor activity was suggested, which correlated well with evidence of viral delivery to tumors. Although reports of clinical efficacy in similar trials are not uncommon,¹⁸ it is difficult not to be heartened by the overall package, which suggests that a critical threshold of intravenous virus infusion leads to intratumoral virus infection, subsequent viral replication, and possible clinical benefit.

Of course, no clinical study, especially a phase I trial such as this, can keep all of the people happy all of the time, and a nagging question remains. The authors have demonstrated systemic virus delivery to tumors but fall tantalizingly short of proving that the virus, once there, actually replicates, although there is a plethora of supportive data. Intratumoral staining is highly suggestive of “virus factories.” Intratumoral expression of β-gal from the virus is (probably) dependent on replication, and detection of antibody to β-gal in patients confirms expression of this viral transgene. It is also difficult to believe that the distribution of gene expression seen in the tumors could be achieved without viral amplification, although peaks in serum levels of granulocyte–macrophage colony-stimulating factor (the product of a second transgene expressed by JX-594) following virus infusion again suggest that viral genes were expressed in vivo. However, the findings did not include recovery of live virus from tumor biopsy specimens, meaning (at the risk of being hypercritical) that formal proof is lacking that delivery, infection, and expression in fact equate with viral replication.

Encouraging as these results are, it remains to be seen whether all oncolytic viruses will be as amenable to systemic delivery. JX-594 may be one of the best cases, in that vaccinia has evolved specific viral forms that facilitate spread through the body, thereby evading both NAb- and complement-mediated inactivation.^19,20 However, the ability of many viruses to spread through infected hosts suggests that vaccinia is unlikely to be the only successful systemic traveler following intravenous delivery.¹⁸ An exciting avenue for further studies will be to characterize, and then exploit, the precise mechanisms that viruses use to persist in the circulation long enough to reach tumors. Interestingly, the authors detected virus in the whole blood of patients, but whether those viral particles were going it alone or hitching a ride on circulating cells²¹ was not clear. Several groups have shown that OV can both “ride and hide” on blood-derived cells, and identifying cells that show an increased propensity to traffic into tumors may provide opportunities for enhancing virus delivery in the future.^12,13,21

In addition to the current findings with vaccinia, the OV portfolio recently received another important market boost from the herpes simplex virus department.²² That success, combined with the current report¹ and progress of other oncolytic viruses to phase III clinical trials,¹⁸ should instill considerable confidence in the marketplace that OV represents a sound investment. Does the article by Breitbach and colleagues¹ provide justification for a huge end-of-year bonus for OV enthusiasts? Probably not, at least not unless further positive reports emerge from randomized trials—and even then, the final market analysis is likely to remain significantly in the hands of those notoriously contrary economists.