Interview – Next from Autolus, cracking tolerance and solid tumours
With this week’s $80m raise the private CAR-T player Autolus achieved a paper valuation of $311m, making it one of the UK’s most valuable biotechs – something it has achieved just three years after inception.
This attests to the popularity of cell therapy and to the ideas being generated by the group’s founder and chief scientific officer, Dr Martin Pule. And while Autolus still does not display an R&D pipeline on its website its chief executive, Christian Itin, has started confirming which of the concepts presented at scientific meetings will be taken into the clinic (see table below).
Already the first two Autolus assets to be taken forward have been named: a dual approach targeting C19 and CD22, and a CAR using the April ligand to hit BCMA and Taci (Autolus takes its first baby steps into the clinic, September 18, 2017).
This points to a key focus for the group as it slowly emerges from stealth mode. “You want to address the issue of target escape,” Mr Itin tells EP Vantage. “That’s where we start.” Even if the targets are not novel, the approach is.
Both these initial assets, now designated Auto3 and Auto2 respectively, will be familiar to followers of presentations of Dr Pule’s work at UCL, given at scientific conferences.
For its next trick Autolus intends to take into human trials a CAR that uses the ligand Jovi-1 to target the TRBC1 isoform of the T-cell receptor’s constant region beta chain. Normal T cells carry TRBC1 and TRBC2 but malignant ones just one isoform, so hitting just TRBC1 might selectively knock out a T-cell lymphoma, Dr Pule told the International Society for Cellular Therapy meeting in May.
And after that comes what might be one of Dr Pule’s most ambitious concepts: dual-targeted, Boolean logic-gated CARs that can make “kill decisions based on the input from more than one antigen,” Mr Itin says. For activation these might require the presence of both antigens, either one, or one but not both.
Juno’s IPO document talked about a similar dual-CAR approach, but this has yet to be advanced into the clinic. “The beauty of Martin’s work is that it’s related to decision-making at the level of the synapse, which is absolutely unique,” says Mr Itin.
“Solid tumour programmes will be rolled out [with] additional cell-programming components that address the defence mechanisms. Which technology elements will be in those products is something we haven’t disclosed.”
|Autolus R&D pipeline|
|Auto3||CD19 & CD22 dual CAR (humanised)||DLBCL (with Keytruda)
|Auto2||BCMA & Taci (via April ligand)||Multiple myeloma||NCT03287804 (April)|
|Auto4||TRBC1 (via Jovi-1 ligand)||T-cell lymphoma||Starting late 2017|
|?||Undisclosed AND/OR/NOT gated CARs||Solid tumours||Starting late 2018|
|Source: Company disclosures, presentations at scientific meetings.|
As far as dual targeting goes, Auto3 – which expresses two separate CARs on the same T cell – gives a good insight into Autolus’s thinking. “We’re not using one receptor that has two binding sites on it, because [CD19 and CD22] are so different in size and structure that if you try to do that you get a suboptimal engagement,” says the chief exec.
He also dismisses the approach used by Juno in JCAR018, a CD22-directed construct for patients relapsing on anti-CD19 therapy, saying “The second time you go in with CAR-T you can still get some level of activity, but it doesn’t tend to stick.”
The problem is that cells evade attack not only by downregulating the initial antigen. “You come in with the same mechanism, just with a different set of goggles on it to see a different antigen. But that cell has learned to defend [itself], independent of the antigen – that’s why sequential use is a flawed approach.”
And that is before considering the cost and complexity of generating CAR-T cells twice and infusing them twice. But do you not hasten loss of both CD19 and CD22? “I don’t think there is evidence of that,” says Mr Itin.
Bispecifics vs CARs
As the former chief executive of Micromet, the developer of Blincyto that was sold to Amgen for $1.2bn, he says: “I had a great opportunity with Micromet to pioneer the space of bispecific T-cell engagers.
“But to create a cell that can deal with multiple layers of checkpoint inhibition it’s virtually impossible to do that with conventional pharmacological means; it is possible with genetic engineering of the cells themselves.
“We’re very early into this journey of engineered T cells,” he admits. But he summarises Autolus’s approach as controlling T-cell activity, breaking tumour defences, and improving cell survival and manufacturing. He also says technology to switch off CAR-T cells using small molecules or biological defence is an area of interest.
As for business development, “we’re focusing on CARs, but we wouldn’t exclude a move into other modalities of recognition. We’ll keep an open eye for technologies that might be additive, but right now we’ve got plenty on our plate.”