Those listening closely on Roche’s quarterly call last week might have guessed that results from the Impassion-130 study in triple-negative breast cancer were not as strong as had been hoped. Data unveiled at Esmo today confirmed this to be the case, with a statistically significant but clinically marginal extension in progression-free survival in the all-comer population.
Markedly stronger responses were seen in subjects positive for PD-L1, however, and physicians described the results in these patients as practice-changing. Roche will rightfully chalk this up as a win, but the wider result will disappoint many in the financial community, who were expecting a much bigger commercial impact from this study.
The outcome also leaves the door open for others to establish that checkpoint inhibition can play a broader role in this disease. The biggest threat here is Merck & Co, which is also running an extensive late-stage programme in triple-negative breast cancer, this time with Keytruda. The neoadjuvant trial Keynote-522 is due to yield data towards the end of the year, while a Keytruda monotherapy study, Keynote-119, in a later line, could also report in the coming months.
Both of these studies are recruiting patients regardless of tumour mutation status, and their chance of showing wide utility are now lower. Given Merck’s previous successes where others have failed, however, nothing can be completely ruled out.
|A hit and a miss? Impassion-130 results|
|Intent to treat||PD-L1 positive (>1%)|
|Tecentriq + Abraxane (n=451)||Placebo + Abraxane (n=451)||Tecentriq + Abraxane (n=185)||Placebo + Abraxane (n=184)|
|PFS||7.2||5.5||HR 0.80 (p=0.025)||7.5||5.0||HR 0.62 (p<0.001)|
|OS||21.3||17.6||HR 0.84 (p=0.08)||25.0||15.5||HR 0.62*|
|*Statistical analysis not carried out owing to failure in ITT population. Source: New England Journal of Medicine.|
To Impassion-130, however, which recruited just over 900 patients with unresectable locally advanced or metastatic triple-negative breast cancer, and had overall survival and progression-free survival as joint primary endpoints. These were to be measured first in all-comers and then, if positive, in the PD-L1-high (>1%) group, which accounted for 369 of the 902 subjects enrolled.
Up to half of all triple-negative breast cancer patients are thought to express this marker.
A PFS benefit of under two months across all patients hit statistical significance, but was judged not clinically relevant by physicians discussing the data at Esmo, though the much improved hazard ratio in the PD-L1-high group was said to be important.
And, while the failure to establish an OS benefit in all-comers meant that a statistical analysis was not carried out on the results from the PD-L1-high group, the outcome in these patients was described as unprecedented in this disease.
“This is a PD-L1-positive story. Although the ITT population was positive, there’s no benefit that justifies use outside of these patients,” Professor Peter Schmid of Barts Cancer Institute in London, primary author of the study, told a press conference today.
He said that investigators had expected to see a stronger benefit in a wider patient group, based on both preclinical and clinic evidence.
“We had a clear rationale as to why the biomarker would not be so binary and why there would be a benefit in PD-L1-negatives, but the data are what they are,” Professor Schmid said.
Assuming that regulators have no quibbles with this result, a 10-month extension in survival in PD-L1-positives looks good enough for approval. Commercially, however, this result is almost certain to hit sales forecasts, as triple-negative disease was viewed as a substantial opportunity for Roche’s flagship immuno-oncology asset.
Other PD-(L)1 assets deemed to hold potential here will also be dented by this finding. Discussing the data Dr Nadia Harbeck, of Munich University, said the trial would have been designed differently had this outcome been deemed likely.
None of the ongoing pivotal studies of anti-PD-(L)1 antibodies in triple-negative breast cancer are selecting patients based on mutation status. Investigators take note.
|Phase III studies of anti-PD-(L)1 MAbs in triple-negative breast cancer|
|Treatment||Sponsor||Setting||Study||Trial ID||Primary completion|
|Keytruda + chemo||Merck & Co||Neoadjuvant||Keynote-522||NCT03036488||Nov 2018|
|Keytruda vs chemo||Merck & Co||2nd or 3rd-line||Keynote-119||NCT02555657||Dec 2018|
|Tecentriq + paclitaxel||Roche||1st-line||Impassion-131||NCT03125902||Feb 2019|
|Tecentriq + chemo||Roche||Neoadjuvant||Impassion-031||NCT03197935||Mar 2019|
|Tecentriq + chemo||Roche||1st-line||Impassion-132||NCT03371017||Jun 2019|
|Keytruda + chemo||Merck & Co||1st-line||Keynote-355||NCT02819518||Dec 2019|
|Tecentriq + Abraxane||Roche||1st-line||Impassion-130||NCT02425891||Apr 2020|
|Bavencio||Istituto Oncologico Veneto IRCCS||Adjuvant||A-Brave||NCT02926196||Jun 2021|
|Tecentriq + chemo||Roche||Adjuvant||Impassion-030||NCT03498716||Jan 2022|
|Tecentriq + Abraxane||Fondazione Michelangelo||Neoadjuvant||NeoTRIPaPDL1||NCT02620280||May 2022|
|Tecentriq + chemo||Roche||Neoadjuvant||MO39875||NCT03281954||Dec 2023|