Prame looks like the real deal

When Roche launches a diagnostic for a novel cancer antigen the endorsement, in terms of target validation, counts for a lot. And this is precisely the endorsement that Prame, a target around which excitement had been building over the past year, got this morning.

True, the Roche assay is not intended to guide the type of pharmacological treatment, but rather to assess whether a tumour is malignant. But the development comes just a day after Immatics reported the biggest and most promising clinical dataset with a Prame-targeting therapy to date – a happy coincidence for the German cell therapy player.

The Immatics results came from phase 1 cohorts of a study of the anti-Prame engineered T-cell receptor IMA203, and represented an important update to a dataset that had piqued interest at last year’s SITC meeting. They were enough to send the group’s Nasdaq stock up 12% yesterday, and enable it to close a $110m secondary share offering.

Roche, meanwhile, today launched the Prame (EPR20330) Antibody, which runs on the company’s IHC/ISH instruments, and is intended for use specifically on tissue samples from patients with suspected melanoma. Roche says that, because Prame is expressed in most melanomas, the assay can be used to differentiate between benign and malignant lesions.

Focus on melanoma

A day earlier Immatics’ data pointed in a similar direction. Though Immatics cited Prame’s highly cancer-specific expression on many solid cancers, including lung and ovarian, its update suggested that IMA203’s best activity was in melanoma, including cutaneous and uveal subtypes. Prame is shorthand for "preferentially expressed antigen of melanoma".

A year ago the phase 1 trial in question had yielded eight remissions among 16 evaluable patients on the first three dose levels of IMA203; four quickly relapsed. Yesterday the result was updated for a fourth dose, and with that initial dose-escalation cohort now at 27 evaluable subjects the ORR at six weeks stood at 48%.

The bad news is that relapses here were still a problem: six weeks later the confirmed ORR was just 19%, or 29% among six subjects at dose level 4.

However, the good news came from a separate dose-expansion group using a new manufacturing process, in which three of four dose level 4 patients had confirmed remissions at week 12 – as did one patient treated with an even higher, exploratory dose.

Thus Immatics was able to make a headline claim that treating patients at dose level 4 or higher resulted in a confirmed ORR of 50%, or 80% if looking just at the second cohort, in which all responses were ongoing at the September 22 cut-off. Prame, it said, was now clinically validated as a multi-tumour target.

The other thing in Immatics’ favour, and which backs its claim that Prame is not expressed especially well in normal tissue, is safety. There are few off-tumour effects in its trial, and cytokine release and neurotoxicity appear at relatively low levels.

If this is the case one obvious question is why no one before Immatics has been able to seize on the potential of Prame, which has been studied by industry for some time. Immatics says ATR203 has improved affinity thanks to enhanced pairing of TCR chains, as well as high avidity and low off-target toxicity.

The group has not finished yet. The first patient was treated in August with an improved version of ATR203 coded IMA203CD8, and initial data are due next year. Competitors, including Immunocore, which disappointed at Esmo, and Biontech, which did a Prame deal with Medigene in February, will pay attention.