Scholar Rock’s quest for a better TGF-β drug

When it comes to biological targets with promise to improve the efficacy of PD-(L)1 blockade few have been hotter than TGF-β. And Scholar Rock, a group not primarily focused on oncology, is one of many biotechs ploughing the TGF-β furrow despite the failure of this approach’s most advanced candidate, Merck KGaA/Glaxosmithkline’s bintrafusp alfa.

Scholar has a two-pronged strategy to give it a better shot at success, its chief medical officer, Yung Chyung, tells Evaluate Vantage. It claims to have elucidated key aspects of TGF-β that have given its candidate, SRK-181, a broad therapeutic window, and at a scientific summit today revealed three key biomarkers that will inform SRK-181’s future trial design.

These biomarkers are cytotoxic (CD8+) T-cell infiltration into the tumour, phosphorylation of the protein Smad2, and the presence of myeloid-derived suppressor cells, Scholar Rock told today’s TGF-β Summit.

The second part of SRK-181’s phase 1 Dragon study is evaluating all these, with a view to a biomarker-driven strategy to assess and predict activity in trials. The findings are based on preclinical work suggesting that SRK-181 increases tumoural cytotoxic T cells, reduces P-Smad2, and reduces myeloid-derived suppressor cells; the company is working on assays to determine levels of these markers.

It is also important to stress that the potential of TGF-β blockade lies in combination with anti-PD-(L)1 MAbs. There is emerging evidence implicating TGF-β1 as a key driver of resistance by stopping immune system cells from penetrating into the tumour, says Mr Chyung, and SRK-181 could overcome resistance to anti-PD-(L)1 therapy.

Though Dragon initially tested SRK-181 monotherapy this was mainly to show safety with escalating doses. The study’s second part will investigate the anti-PD-(L)1 combo approach in tumours with primary resistance to PD-(L)1 blockade, effectively to show whether SRK-181 can turn such “cold” tumours “hot”.

Doing it wrong

This is something bintrafusp, a fusion protein combing TGF-β and PD-L1 blockade, did not show; even though bintrafusp remains in a few trials, its failures in NSCLC and cholangiocarcinoma resulted in Glaxo pulling the plug on a deal last September.

Mr Chyung reckons he knows where Merck KGaA went wrong, however. While not referring specifically to bintrafusp he says one problem with having both modalities in one molecule is limitation of dosing, in that the optimal dose for blocking PD-(L)1 is unlikely to be the same as that for blocking TGF-β1.

“Moreover, the molecule may be distributed locally within the tumour environment to areas where there is PD-1 or PD-L1 expression, which may not be the same sites where TGF-β1 is relevant, and vice versa,” he states.

A separate point that goes to SRK-181’s second supposed advantage is specificity. While bintrafusp and several other industry projects hit TGF-β SRK-181 has specificity for TGF-β1, a subtype that is “very challenging to drug. TGF-β comes as three different isoforms [that] are structurally quite similar, making it difficult to hit TGF-β1 without also hitting the others,” Mr Chyung explains.

The problem is that non-selective TGFβ blockade has potential for off-target toxicity. Scholar reckons it has achieved the desired specificity, allowing relatively high dosing, thanks to targeting the latent rather than the mature form of TGFβ1.

The competitor pipeline reveals just one other clinical asset with TGFβ1 subtype specificity, Pfizer’s PF-06952229, while Agenus and Jiangsu Hengrui are continuing with bispecifics. It is notable that Gilead abandoned Agenus’s AGEN1423, a bifunctional anti-CD73/TGFβ trap, over a year ago.

“It is important that SRK-181 is a MAb rather than a bispecific,” says Mr Chyung.