Adcom puts T-Vec on track to underwhelm

Serious doubts about study design and interpretation of efficacy ultimately failed to derail Amgen’s oncolytic virus project T-Vec, a US advisory panel last night recommending its approval by 22 votes to one.

That said, assuming that the FDA follows the adcom’s advice, T-Vec looks set to underwhelm in a melanoma market that is changing fast, and where its role looks to be limited to that of a supporting player. Seeing how the science has been evolving, the greatest value for oncolytic viruses will be in combinations, but for T-Vec this is still a long way away (see table below).

This is despite the Amgen project being by far the most advanced of a range of the industry’s oncolytic viruses, which are naturally occurring and work by preferentially infecting and killing cancer cells – at least in theory. Their anticancer potential had been spotted decades ago, but the only one to reach the market so far is Shanghai Sunway Biotech’s Oncorine, in China.

First ever in the US

After yesterday’s vote in favour of its use in metastatic melanoma T-Vec (talimogene laherparepvec) looks set to become the second worldwide, and the first in the US to hit the market.

Amgen had a lot riding on this project, having acquired it through its takeover of BioVex for $425m up front in 2011. But a lot has happened in the intervening four years – not least melanoma having been taken by storm by the checkpoint inhibitors Yervoy and Opdivo from Bristol-Myers Squibb and Keytruda from Merck & Co.

And well before the adcom there had been significant doubts about T-Vec’s real-world potential, not to mention niggling uncertainty over its pivotal 005/05 study (Optim). Look no further than the opinion of a melanoma expert at Asco two years ago, who deemed T-Vec to be unready for prime time (Asco – Despite promise, future of oncolytic viruses remains clouded, June 3, 2013).

Judging by the panel’s briefing documents full readout of Optim had done little to allay the concerns. The open-label trial did hit its primary efficacy measure of durable response rate, though the important secondary endpoint of survival appeared to have been narrowly missed, with median OS 4.4 months longer in the T-Vec arm versus GM-CSF control (hazard ratio: 0.79; p=0.051).

However, the FDA reviewers poured scorn on these analyses, raising doubts over clinical meaningfulness of the durable responses and disagreeing with the response assessments for two patients, citing asymmetric dropouts, leading to asymmetric bias. The agency’s own post hoc OS analysis yielded a 16% reduction in risk of death at a non-significant p value of 0.155.

T-Vec is based on HSV1 that has a single gene deletion as well as a gene insertion coding for GM-CSF. The theory is that T-Vec is inactive inside healthy cells, but replicates after infecting cancerous cells, secreting GM-CSF and lysing the cells, releasing more GM-CSF and more viruses that go on to infect other cancer cells.

The reviewers criticised the use of GM-CSF in Optim’s comparator arm, and also expressed concerns about the risk of transmission to healthcare providers via viral shedding; Amgen has proposed running a 60-patient viral shedding study.

But, while accepting these shortcomings, the panel found some solace in additional post hoc analyses, such as one supporting a survival benefit in earlier-stage patients.

And, T-Vec having shown a reasonable adverse event profile, the panel decided that its benefit/risk profile was sufficient for approval, thus sending it on its way towards the market – the ultimate arbiter of issues like therapeutic benefit and practicality of use.

Real-world use questions aside, T-Vec making it to the market could spur interest in other groups developing oncolytic viruses. As the table above shows, there are plenty.

And despite the setbacks and slow development speed there has been business development, as witnessed by a deal in January under which AstraZeneca licensed rights to a lead vesicular stomatitis virus project under way at the private US biotech Omnis Pharmaceuticals.

The key is the potential of combining oncolytic viruses, particularly with novel checkpoint inhibitors. The scientific rationale involves the defence mechanisms that infected tumour cells employ to evade immune response, such as upregulating PD-L1, which makes additional PD-1/PD-L1 blockade a particularly apt strategy.

Moreover, it has been postulated that interventions like oncolytic viruses could increase neoantigen exposure to the T-cell based immune system, acting like a vaccine to boost the tumour’s susceptibility to immunotherapy. The Finnish biotech Oncos Therapeutics, for instance, cites the “priming” effect of its ONCOS-102 project as a reason for combining it with checkpoint inhibition.

However, for all the players here a combinatorial approach is still years away. For now, oncolytic virus therapies are destined to remain an interesting concept, and T-Vec is surely about to show that securing approval is quite separate from having commercial utility.

To contact the writer of this story email Jacob Plieth in London at [email protected] or follow @JacobPlieth on Twitter