World Lung – Toxicity scuppers Amgen’s first-line lung hopes
In the Kras plus checkpoint inhibitor race it looks like advantage Mirati, for now.
Moving Kras inhibitors into first-line non-small cell lung cancer will depend on getting a result with a checkpoint inhibitor combination. Following promising but early data from Mirati here, Amgen has recently been playing down the prospects of a Lumakras-PD-(L)1 tie-up – and results presented yesterday at the World Lung meeting explain why.
Response rates with Lumakras plus either Merck & Co’s Keytruda or Roche’s Tecentriq looked worse than those historically seen with Lumakras monotherapy, while liver toxicity loomed large. Amgen is still trying to nail down the best dosing regimen but, based on the data so far, its door to the first-line NSCLC space could be closing.
So it is advantage Mirati, but that group, which has fallen behind in later-line NSCLC, must do better this time around. There are reasons to be cautious about the combo results with adagrasib so far, including small patient numbers, relatively short duration of follow-up, and the fact that only three of 11 partial responses had been confirmed at last count.
| Cross-trial comparison of Lumakras and adagrasib PD-(L)1 combos in NSCLC | |||||||
|---|---|---|---|---|---|---|---|
| Product/ project | Company | Trial | PD-(L)1 combo | ORR | TRAEs | Grade ≥3 TRAEs | ALT increases |
| Lumakras | Amgen | Codebreak-100/101 | Keytruda or Tecentriq | 29% (17/58) | 88% (51/58) | 62% (34/58) | 31% (18/58) |
| Adagrasib | Mirati | Krystal-7 | Keytruda | 55% (11/20)* | 32% (12/37) | 43% (16/37) | 16% (6/37) |
| *Includes three confirmed responses. Krystal-7 enrolled 1L pts, in Codebreak-100/101 only 21% of pts 1L. Source: World Lung 2022 & company presentation. | |||||||
Another factor making for a tricky cross-trial comparison is the fact that the Krystal-7 trial of adagrasib enrolled first-line patients, while only 21% of patients in the Lumakras combo dataset were treatment-naive.
Still, one thing is clear: the Lumakras combo data disappointed, with liver enzyme elevations dominating the findings.
Presenting the data Dr Bob Li, of Memorial Sloan Kettering Cancer Center, said the low overall response rate in all-comers might have been down to the high treatment discontinuation rate, which meant that patients were not exposed to drug for a long period of time.
He pointed to ways that Amgen was trying to mitigate these effects: lowering the dose of Lumakras – the Codebreak-100/101 dataset included five doses ranging from 120mg to 960mg – and testing a lead-in period of Lumakras dosing before starting checkpoint inhibitor therapy.
The lead-in strategy did appear to help lower adverse events, but rates were still high. And a 32% ORR with lead-in Lumakras plus Keytruda is unimpressive versus a 36% ORR with Lumakras monotherapy in previously treated NSCLC.
Amgen is taking forward a 240mg Lumakras dose with a lead-in period, plus Keytruda, which showed liver enzyme elevation rates of 20%. Results will be closely watched to ascertain whether the company can salvage anything here.
Not the same?
A big question now is why such liver toxicity has not been seen with the combination of adagrasib and Keytruda.
SVB suggested that pretreatment with checkpoint inhibitors could have contributed to hepatotoxicity in Amgen’s Codebreak-101 trial. They also pointed to the late emergence of liver tox with the Lumakras/PD-(L)1 combo, adding that rates with adagrasib might increase with longer follow-up.
Meanwhile, Stifel analysts speculated that adagrasib’s longer half-life might be giving it an edge, allowing better pharmacokinetics at lower doses. They are expecting more adagrasib/Keytruda combo data in the second half of this year, which could put this theory to the test.
Amgen's slip-up might also give hope to the many other Kras developers.
