Investors don't buy Lyell's solid tumour Car-T success

Not long after AbelZeta came to ASCO with strong liver cancer data using its GPC3-directed car-T therapy C-CAR031, Lyell Immunopharma has claimed Car-T activity in another solid tumour, triple-negative breast cancer, with its anti-ROR1 asset LYL797.

But while the AbelZeta boasted a 55% response rate among 24 patients, the LYL797 dataset amounts to just two partial responses among 16, leaving Lyell pedalling hard on “clinical benefit rate” and a 40% ORR among five TNBC patients given a particular dose. Indeed, perhaps the biggest takeaway from the limited data presented is a worrying rate of dose-limiting pneumonitis.

Among 18 safety-evaluable subjects there were four cases of dose-limiting pneumonitis or hypoxia, and the first patient who developed pneumonitis died 41 days later of respiratory failure, a revelation that sent Lyell stock down 36% today. The company said that subsequently any sign of pneumonitis was treated early.

High-dose steroids

On an investor call today Lyell argued that pneumonitis was treatable with high-dose steroids, and all patients are now given prophylactic dexamethasone. So far pneumonitis has been observed only in patients with lung involvement; the company has treated a handful of NSCLC patients with LYL797, but hasn’t seen any efficacy here.

ROR1 is a target that became prominent after Merck & Co paid $2.8bn to buy Velosbio, and Boehringer Ingelheim acquired NBE-Therapeutics for up to $1.5bn. But those deals involved ADCs, not Car-T therapies, and the latter’s clinical trial was terminated with just 12 patients enrolled.

And safety is in the spotlight after another anti-ROR1 Car-T therapy, Oncternal’s ONCT-808, caused the death of one of four B-cell lymphoma patients, from complications due to cytokine release and neurotoxicity, according to data presented at EHA. There were also two cases of severe pneumonia, and that study’s protocol has since been amended.

If toxicity can be managed any efficacy with ROR1 targeting will be intriguing, given that one of the first anti-ROR1 Car-Ts, in development at Fred Hutchinson Cancer Research Center, yielded very poor data. That project was licensed to Juno as JCAR024, and in Bristol Myers Squibb’s hands emerged as BMS-986403, but has since been terminated.

Anti-ROR1 Car-T: the data so far

Source: OncologyPipeline.

Lyell argues that without extra features Car-T cells fail in solid tumours because of exhaustion. LYL797 uses a similar construct to Hutch’s, but additionally expresses the transcription factor c-Jun, which Lyell claims “reprogrammes” the cells and helps them resist exhaustion.

This, the group argued today, has enabled it to report the first signs of Car-T cell infiltration into a solid tumour, which it said was seen in all nine LYL797-treated patients in whom on-study tumour biopsies were available. 

Ultimately, however, Lyell’s efficacy data boiled down to two confirmed partial responses in five TNBC patients given 150 million LYL797 cells. The company claimed to see a dose-response relationship, but one patient given 300 million cells didn’t respond, and neither did 10 others (at 50 or 100 million cells).

In light of the toxicity signal Lyell says it’s now moving forward with separate dosing of 75 million cells in patients with lung involvement, and 300 million in those without. The group also plans to start clinical trials of LYL797 in multiple myeloma and CLL, where it says ROR1 expression is high.

Lyell also highlighted a separate anti-ROR1 Car-T in its pipeline, LYL119, for which it is awaiting IND clearance. Some investors might see that as an admission that a plan B is needed.