Trinity sets Abbvie on a collision course with inevitability

The writing had been on the wall for the Trinity study of Abbvie’s Rova-T for some time, yet the group still managed to fall 15% after confirmation of the trial’s failure this morning. It seems that there are some in today’s market who still cannot face up to inevitability.

Perhaps the problem is that close inspection of the Trinity data, in third-line small-cell lung cancer, show Rova-T to be even worse than the bears had warned. There is now an even stronger case for expecting the failure of the two remaining Rova-T studies, in first-line maintenance and second-line SCLC.

Remarkably, Abbvie still touts the possibility of these two trials, Meru and Tahoe respectively, delivering positive data, and today said it continued to believe that Rova-T had “potential for patients with SCLC and other DLL3-expressing cancers”. Then again, having shelled out $5.8bn to buy Rova-T’s originator, Stemcentrx, the company risks losing face if it throws in the towel.


So what does Trinity tell us about Rova-T? The project is an antibody-drug conjugate targeting DLL3, a ligand seen in over 80% of SCLC tumours, yet the result suggests that, even in high expressers, targeting this ligand is futile.

This much is revealed by how Abbvie has cut the data to try and flatter the result. Trinity enrolled SCLC subjects whose tumours expressed DLL3 at 1% or above – in itself a dubious strategy, since an earlier study had shown an ostensibly positive effect in >50% DLL3 expressers.

But today’s announcement contains a footnote revealing that the 16% overall remission rate, which deemed Trinity a bust, was not even recorded in the >1% DLL3 population, but rather in a subgroup of 177 subjects “with high DLL3 expression”. It does not quantify “high”, but it seems reasonable to suppose that the number might be 50%.

A separate point is what might theoretically have constituted a positive readout in Trinity – controversially designed as a single-arm trial. Abbvie recently claimed that an 18% remission rate was the benchmark set by retrospective analyses, yet a perusal of available scientific literature suggests that rates as high as 31% have been achieved with third-line chemo.

So what Trinity comes down to is this: even after first enriching the SCLC population, and then carving out an especially responsive subgroup post hoc, Abbvie has still failed to meet the low historical bar it had set itself. And this is before considering the toxicity seen in Trinity, which Leerink analysts today described as “unacceptable for virtually any cancer drug”.

Investors still holding out for the results of Meru and Tahoe should remember that these settings will soon face challenges from immuno-oncology: Bristol-Myers Squibb’s Opdivo plus Yervoy showed promise in a SCLC subset of the Checkmate-032 study, and Roche’s Impower-133 trial of Tecentriq specifically in first-line SCLC will read out imminently.

A final reflection is that, through the mounting doubts about Trinity – an open-label study – Abbvie continued to pump up hopes, for instance saying on its fourth-quarter call that it planned to file for accelerated approval on the Trinity data.

Of course, that plan was today scrapped – five months after an FDA request for longer follow-up as good as confirmed Trinity to be a bust (Why the Rova-T delay is anything but irrelevant, October 30, 2017). It will not be long before Rova-T heads to the scrapheap too.

Study  Setting Detail Trial ID  Data?
Trinity  3L and later SCLC expressing DLL3 >1%  Rova-T (no comparator) NCT02674568  Failed, even in DLL3 "high" subgroup
Meru  SCLC maintenance after 1L Pt chemo  Rova-T + dexamethasone, vs placebo NCT03033511  Nov 2019
Tahoe  High-DLL3 and chemo-progressed SCLC   Rova-T vs topotecan NCT03061812  Sep 2019

To contact the writer of this story email Jacob Plieth in London at or follow @JacobPlieth on Twitter

Share This Article