KRAS inhibitors arrived on the market in May 2021 and December 2022 to much fanfare—but, if you ask Eli Lilly’s Loxo unit, the therapies have left much to be desired.
Amgen’s Lumakras and Mirati Therapeutics’ Krazati are “modestly effective” and have not led to the “dramatic advance for patients” that was initially thought possible when the once undruggable target was finally hacked, according to Jake Van Naarden, executive vice president at Lilly and CEO of Loxo@Lilly. Specifically, to make a real dent in the market, he said, KRAS inhibitors need to combine with standard-of-care PD-1 treatment—which in Loxo's opinion has not been smooth sailing for the first-to-market winners.
With the first clinical data from Loxo's LY3537982 ready for prime time, Van Naarden thinks the up-and-coming KRAS G12C inhibitor will easily find its foothold in the market thanks to a cleaner safety profile and early efficacy results on par with the competitors'.
“The reason we're excited is that I think we see that path; and we see that path in a way that appears—despite it being early—differentiated from frankly, the whole rest of the field,” Van Naarden said.
The Lilly unit presented at the American Association for Cancer Research (AACR) meeting in Orlando Monday afternoon on phase 1 data from previously treated KRAS G12C-mutant advanced solid tumors. The study tested LY3537982 alone and in combo with PD-1 inhibitors such as Merck & Co.’s Keytruda.
In the 84-patient monotherapy arm, LY3537982 led to preliminary efficacy across all dose levels in multiple tumor types, according to a slide presentation shown at AACR. The efficacy results were on par with the existing KRAS inhibitors, according to David Hyman, M.D., Loxo’s chief medical officer. LY3537982 achieved a 38% overall response rate and an 88% disease control rate in the eight patients with non-small cell lung cancer (NSCLC) who had not previously received a KRAS inhibitor. In the 14 patients who had previously received a KRAS inhibitor, the ORR was 7% and the DCR was 64%. These were the smallest populations in the trial and the indication that Lumakras and Krazati are approved in.
Because KRAS inhibitors are now available for NSCLC, Van Naarden said fewer of those patients ended up in Loxo's trial. That makes for a smaller data set to make comparisons but provides a glimpse at LY3537982's efficacy in a range of other tumor types that do not yet have an approved KRAS inhibitor. LY3537982 led to a 10% ORR and a 90% DCR in the 20 colorectal cancer patients and 42% ORR and 92% DCR in pancreatic cancer. In a category labeled "other" that included ovarian, and head and neck cancers, the ORR was 52% and the DCR was 95%.
“As a monotherapy, the drug is doing what it ought to do. It's causing single-agent responses at about the right level that you'd expect for a drug like this,” Van Naarden said.
In the combo arm, the ORR was 78% in the nine efficacy evaluable patients who received LY3537982 and Keytruda and had not previously been treated with a KRAS inhibitor. In a very small dataset of just four patients, there was an ORR of one, or 25%.
Hyman admits the efficacy data in the combo arm represent “early days,” but Loxo is nonetheless encouraged. What the unit was most looking for were the safety data, because Lumakras and Krazati have both struggled in showing efficacy and most importantly, safety, when paired with standard of care PD-1 inhibitors.
“We really view the combination of pembrolizumab [Keytruda] as almost a pure safety experiment,” Hyman said. To really show efficacy in this population, a randomized phase 3 trial is needed. “It's a great outcome. We could have seen much lower response rates, and still been OK with it,” he said.
There were no dose-limiting toxicities in the monotherapy arm, and the treatment-emergent adverse events were predominantly grade 1 and 2. There were a few grade 3 events including diarrhea and one instance of elevated liver enzymes, but no high-grade liver toxicity.
Both Amgen and Mirati have been testing their therapies in PD-1 combos, all the while waging a back and forth war of data and milestones. Scientists believe that a combo approach to solid tumors is best to avoid resistance. Mirati has stepped ahead of Amgen slightly by showing that Krazati combined with Keytruda shrank tumors in 49% of 53 patients with newly diagnosed KRAS G12C-mutated NSCLC during a phase 2 trial. For safety, Mirati showed a relatively low rate of liver toxicity at about 5% grade 3/4 events. The data were enough for Mirati CEO David Meek to declare: "we’re the KRAS leader, and we will set the direction where this KRAS agent goes."
Amgen, meanwhile, saw a safety red flag crop up in a phase 1b trial of Lumakras with either Keytruda or Roche’s Tecentriq. The combo led to a marked increase in liver toxicities at a rate of roughly 50% of patients experiencing severe liver toxicity at grade 3 or 4. On efficacy, the Lumakras combo triggered a response in 29% of 58 patients.
“Sotorasib [Lumakras] looks like it's just unable to combine with PD-1 at any dose or schedule,” Hyman said of the liver toxicity data. “Adagrasib [Krazati] does claim to be able to combine with PD-1 based therapy … but the absolute frequency of high-grade adverse events seems very high.” Mirati is currently conducting a randomized phase 3 PD-1 combo study.
Hyman continued: “When we speak to physicians, they're concerned about prescribing a regimen with that overall burden of toxicity in patients.”
To the front of the line
If the early positive data for LY3537982 continues, Van Naarden said Loxo will head for the first-line of treatment, which has so far eluded Amgen and Mirati.
“If this holds up, we'll be in a position, I hope, to launch a suite of phase 3 randomized studies to explore the medicine in first-line lung cancer, and I think advance the field in a way that others haven't yet been able to do,” he said.
Another key is that LY3537982 had a favorable safety profile in patients who had previously taken a KRAS G12C inhibitor. “The efficacy and safety are interlinked. If you can't stay on the medicine long enough because of safety issues, your efficacy will be compromised,” Van Naarden said.
Hyman said that Loxo believes LY3537982 is seeing better results because it is 10 times more potent, meaning the drug can be administered at lower doses with less drug in the bloodstream and still inhibit the target to the same degree as other drugs in the class. Higher potency also means fewer off-target consequences. Van Naarden said LY3537982's dose is about a fifth of the existing agents.
LY3537982’s dosing was from the beginning optimized to be combined with a PD-1 inhibitor, specifically Keytruda, Van Naarden said. Loxo's competitors, meanwhile, tried to squeeze the last bit of efficacy possible from their candidates early on.
“They were trying to just maximize the dose, they didn't want to leave any efficacy on the table. And we said, we already know that that strategy leads you to trouble,” Hyman said. “We're going to be very, very rigorous about defining a safety program profile with pembrolisimab [Keytruda] and the dose that we select and I think that that strategy is paying off.”
The FDA has asked both Amgen and Mirati to perform confirmatory studies of their candidates with lower doses.
Van Naarden said Loxo has its eye on a first-line indication in KRAS G12C-mutated lung cancer. That would allow the unit to leapfrog over the initial market entrants, which are still sitting in the second-line setting.
So what seems like a crowded market is "much less crowded than it seems," Hyman said.
As for next steps, the study will continue so Loxo can compile a larger data set. The AACR data had a cutoff in January, but Van Naarden said investigators are “seeing a similar story” in the months that have passed.
“We continue to feel pretty good, but we’re not done yet,” he said. “We have the next set of pivotal studies all sketched out, and we'll hit the go button as quickly as we can once we feel like we're ready.”