Interview – Gritstone hopes to inject some interest

When it comes to cancer neoantigen plays it’s fair to say that, so far, Gritstone Bio has been eclipsed by the likes of Moderna and BioNTech. But the smaller group reckons it could have a differentiated approach with its Granite programme – a fact that its chief executive, Andrew Allen, tells ApexOnco has allowed the company to go after colorectal cancer, a tricky tumour to target with immunotherapy.

Gritstone will have a chance to start proving this in early 2024, when initial phase 2 data are due from a study of Granite in first-line CRC. However, with the company focused on molecular response, rather than the more widely accepted surrogate endpoint of progression-free survival, it might have to work hard to convince investors about the strength of the results, Allen admits.

It is also currently unclear whether Gritstone will be able to gain accelerated approval based on molecular response, its preferred path to market. With the company’s market cap currently at a mere $180m it clearly has some convincing to do.

Cold colorectal

Granite is a personalised neoantigen immunotherapy designed to drive a T-cell response to a patient’s tumour. Gritstone's move to evaluate the therapy in combination with Roche’s checkpoint inhibitor Tecentriq in first-line CRC is bold – this tumour type that has so far not benefited from immunotherapy

This is deliberate, Allen says. “The whole point of our programme is to give a patient T cells that don’t exist. Therefore the logical place to go is cold tumours, because there’s huge unmet need, and the signal will be very simple to see.”

This strategy stands in contrast to Moderna and BioNTech, which are initially evaluating their neoantigen projects – the Merck-partnered mRNA-4157 and the Roche-partnered autogene cevumeran (BNT122) respectively – in melanoma, a cancer in which PD-(L)1 inhibitors are already effective.

In the adjuvant melanoma setting mRNA-4157 appears particularly promising, with Merck yesterday reporting a 49% reduction in risk of disease recurrence or death at three years for a Keytruda combo versus Keytruda alone. This is statistically significant, with the company citing a one-sided nominal p value of 0.0095, though this trial comprises just 157 subjects.

Gritstone believes that Granite might work in other solid tumours because it is designed to elicit the strongest possible CD8+ T-cell response. 

While Moderna and BioNTech’s projects are delivered via multiple doses of mRNA, Gritstone uses two different approaches in a heterologous prime-boost strategy. The initial therapy is delivered by adenovirus, which Allen says is the best system at stimulating a CD8+ T-cell response. The “boost” is given via self-amplifying mRNA, to get around the fact that adenovirus can’t be given again too soon after the first dose, as the patient will have developed an immune response against the virus.

Gritstone’s claims will soon be put to the test, with initial data due in the first quarter of next year from the first 50 patients in the phase 2 portion of its ongoing phase 2/3 trial.

All patients in the study initially get standard of care chemo for around five months, then they are randomised either to continue with chemo, or to receive chemo plus Tecentriq plus Granite. The neoantigen therapy is given as six doses over a year: via the adenoviral vector at day zero and month five; and via samRNA at months one, three, eight and 11.

Gritstone also plans to compare results to the failed Modul trial of chemo plus Tecentriq in first-line CRC, which will act as an external control.

The initial data will primarily involve molecular response, measured via circulating tumour DNA (ctDNA); a reduction in this marker is associated with tumour shrinkage. Here, Allen says, Gritstone hopes to see around a 50% molecular response rate, versus no more than 10% in the chemo control arm. 

Molecular response vs PFS

Backing the group’s focus on molecular response, Allen contends: “It’s been shown, in multiple studies, that ctDNA response is a better predictor of OS for immunotherapy than radiology.”

Indeed, Gritstone’s own phase 1 data in third-line CRC found that molecular responders to Granite had a median overall survival of over 22 months, versus 7.8 months for non-responders – however, these results came in a small number of patients.

Meanwhile, tumour shrinkage as measured by Recist, the basis for progression-free survival, is a more widely accepted surrogate for OS – and something that investors expect to see. For this reason, Gritstone also plans to release PFS data next year.

Current standard of care in this setting leads to PFS of 11 months, according to B Riley analysts, giving Gritstone a bar to aim for.

However, immunotherapy can cause the infiltration of T cells into tumours, therefore increasing the size of a lesion, says Allen – which would be picked up as progression according to Recist. A problem for Gritstone could therefore arise if the the ctDNA and PFS data don't correlate.

And investors will have a while to wait for phase 2 OS data, which are due in 2025, to confirm the relationship between ctDNA responses and survival. Currently Gritstone has enough cash to get it to the end of 2024, so a lot is riding on next year’s readout.

Accelerated approval

Another question is whether Gritstone can gain accelerated approval for Granite, as hoped. Should all go well with phase 2, the company expects to start the phase 3 portion of the trial in late 2024 or early 2025, Allen says.

Initial phase 3 molecular response data could emerge as early as 2025, and it is these, alongside phase 2 OS results, that could form the basis of an accelerated approval application.

Still, this is no means a given, Gritstone admits; the group's base case is that phase 3 OS data will be needed for approval. The company will not allow crossover, Allen says, to avoid the sort of problems that have hit Novartis’s PSMAfore trial of Pluvicto, for example.

All this is a long way off, and to get to this point will require more cash than Gritstone has on hand. However, Allen reckons that a positive result with Granite next year would allow the company to “partner up and launch five phase 3s, at least”.

He adds that big pharma is much more comfortable with molecular response data than investors might be – a claim that could be put to the test next year.

This story has been updated to clarify that approval of Granite is expected to be based on phase 3 OS data.