AstraZeneca goes early with Gracell

Almost exactly two years after buying China’s Gracell for $1bn, AstraZeneca is taking the lead asset from that acquisition into phase 3. Astra highlighted the upcoming Durga-4 trial, testing the anti-BCMA x CD19 Car-T therapy now coded AZD0120, at its fourth-quarter presentation this week, and the study has just been listed on clinicaltrials.gov.

Perhaps the most interesting aspect of this listing, not previously highlighted by Astra, is that Durga-4 will enrol multiple myeloma patients as early as after just one prior line of therapy. This puts the company at odds with other multiple myeloma players, which have pursued later settings as an initial way in, suggesting that Astra sees no need for such caution.

One reason for setting an aggressive development plan could be the recent regulatory changes in multiple myeloma for competing Car-T and other therapies. Johnson & Johnson/Legend’s Carvykti, for instance, was approved for second-line use in April 2024, while J&J’s anti-BCMA T-cell engager Tecvayli faces an imminent FDA decision for this use.

It’s possible that Astra sees limited potential in aiming for an initial late-stage approval, given the slew of BCMA-targeting therapies already with such labels, and the moves of some into early lines. A notable BCMA-targeting Car-T therapy still gunning for a salvage setting is Gilead/Arcellx’s anito-cel, which has a second-half PDUFA date for fourth-line or later multiple myeloma.

Two improvements?

There are two aspects of AZD0120, which Gracell had coded GC012F, that could help this autologous cell therapy stand out from the incumbents.

Firstly it targets CD19 as well as BCMA. That’s unusual, as CD19 is most commonly associated with B-cell lymphomas and leukaemias. However, there is evidence that this protein is expressed on some populations of plasma cells, albeit at low density; hitting both targets might broaden AZD0120’s efficacy.

Secondly AZD0120 uses a rapid manufacturing system. This might shorten the time taken to generate AZD0120 Car-T cells versus traditional methods, but more importantly it could give rise to younger and fitter T cells, resulting in what Astra will hope is ultimately a better product.

Durga-4, which has an estimated start date of 9 February, seeks to enrol 508 multiple myeloma patients who have received one to three lines of prior therapy, including an imid (likely Revlimid) and either a proteasome inhibitor (like Velcade) or an anti-CD38 MAb (Darzalex). 

AZD0120 monotherapy will be compared against standard regimens – these will depend on the line of therapy – on co-primary endpoints of PFS and nine-month MRD negativity. The latter has recently been promoted in FDA guidance as an appropriate endpoint in relatively early lines of multiple myeloma therapy.

Key AstraZeneca trials of AZD0120/GC012F

Source: OncologyPipeline & Astra Q4 2025 disclosure.

There are two key published results that support Astra’s enthusiasm. The first was touted by Gracell before its acquisition as being a first-line dataset showing an 86% stringent complete response rate; however, most of the 22 patients involved were already in remission at enrolment, so this was more akin to a maintenance setting designed at deepening responses.

The second came from Astra’s own phase 1/2 trial, Durga-1, which showed a 100% response rate among 15 patients in the relapsed/refractory (fourth-line or later) setting.

Buying Gracell is one of several ways in which Astra, a relative latecomer to cell therapy, has tried to make up lost ground. Other approaches include work undertaken with AbelZeta on dnTGFbRII-armoured constructs, and the EsoBiotec acquisition, which took Astra into in vivo Car-T; Astra is also understood to be working on CD8-guided allogeneic Car-T therapy.