A daily pill from Roche called divarasib helped people with advanced lung cancer live longer than two drugs already approved for the same gene change, according to a large trial presented at the European Society for Medical Oncology (ESMO) Congress in Madrid1. The results were presented on Sunday, October 25, as a late-breaking study in one of the meeting’s main sessions2, and were published at the same time in the New England Journal of Medicine1.
Half of the people who took divarasib were still alive at 17.7 months, compared with 12 months for those on the older drugs1. That figure is a median: the point at which half the patients in a group had died and half were still alive. The drug is still experimental and has not received the approvals needed for routine use1.
What the trial tested
The study, called Krascendo 1, enrolled 338 people at 151 hospitals in 26 countries1. All had a common form of lung cancer, non-small cell lung cancer, carrying a gene change known as KRAS G12C, and all had already been treated before1.
KRAS is a gene that makes a protein working like a switch inside cells. It tells cells to grow and divide, then turns off when that signal is no longer needed1. The G12C change locks the switch in the “on” position, so cells keep dividing1. Drugs aimed at KRAS G12C bind the faulty protein and hold it in an inactive state1. The change appears in about 14% of non-small cell lung cancers1.
About half the patients got divarasib, taken by mouth once a day. The rest got one of the two approved first-generation drugs for the same gene change: sotorasib (sold as Lumakras) or adagrasib (sold as Krazati)1. Most people in the comparison group got sotorasib, because adagrasib was available in only one of the participating countries1. Genentech, Roche’s US arm, describes Krascendo 1 as the only global trial to pit a newer drug of this kind directly against the approved ones2.
The results in numbers
The trial’s main goal was progression-free survival, meaning how long people lived before their cancer grew or spread. That was 9.8 months on divarasib versus 5.7 months on the older drugs, measured at the median1. Roche had first announced in July that the trial met this goal3. The new data add survival:
- Overall survival: 17.7 months versus 12 months (median), which the researchers reported as a 44% lower relative risk of death during the study1. “Relative” means compared with the other group, not an absolute chance.
- Alive after one year: 67.1% on divarasib versus 49.2% on the older drugs1.
- Tumors shrinking or disappearing: 58.4% versus 32.3%1.
- How long that shrinkage lasted: a median of 11.9 months versus 6.9 months1.
Dr. Mor Moskovitz, head of the thoracic oncology service at the Davidoff Cancer Center at Beilinson Hospital in Israel, commented on the findings to Ynetnews1:
“The results are very exciting. This is the first time we have seen a significant and statistically meaningful improvement in a subgroup of lung cancer patients whose treatment options have been limited until now.”
What it does not show yet
These are early survival numbers. They come from the first planned look at survival, after patients had been followed for a median of 14.3 months1. Later analyses could make the gap larger or smaller.
The comparison is mostly against one drug. Because adagrasib was used in only one country, the control group was largely sotorasib1, so the trial says less about how divarasib compares with adagrasib specifically.
The trial covered only people whose cancer had already been treated1. Whether the drug helps earlier in the disease is a separate question that Roche is testing in another trial2.
Side effects matter for any cancer drug taken every day. The sources Biohack reviewed for this article did not give detailed side-effect numbers from Krascendo 1. Trade publication Scrip, reporting from the meeting, described the drug’s safety profile as “impressive”3, but that is a summary, not the data itself. Claims that divarasib is more potent and selective than the older drugs come from laboratory studies1, not from this trial.
Who is behind it
Divarasib belongs to Roche, and Scrip described the head-to-head design as Roche’s own decision to test its drug against the two approved rivals: Amgen’s Lumakras and Bristol Myers Squibb’s Krazati3. Genentech presented the results at ESMO2. Its chief medical officer, Levi Garraway, framed the result as “establishing head-to-head superiority with divarasib in KRAS G12C-mutated lung cancer”2, and the company calls the drug a potential “best-in-class” therapy2. Those are the company’s words. The sources reviewed here do not list the trial’s funding details or the investigators’ financial ties to Roche.
Outside comment came from Israeli cancer doctors quoted by Ynetnews, including Dr. Abed Agbarya of Bnai Zion Medical Center in Haifa, who said newer drugs of this type “may demonstrate greater inhibitory potency and cover a broader range of variants and molecular states associated with the mutation”1.
How it fits the field
KRAS was long seen as hard to target, and the first drugs aimed at the G12C change reached patients only in recent years. Sotorasib received an accelerated approval from the US Food and Drug Administration (FDA) on May 28, 20214. An accelerated approval lets a drug reach patients early, on the condition that later studies confirm a benefit. In December 2023 the FDA declined Amgen’s request for full approval with a “complete response letter,” and asked for another confirmatory study to be finished no later than February 20284.
Part of the background is that the older drugs had not clearly helped people live longer than chemotherapy. In an earlier large trial comparing sotorasib with chemotherapy, median overall survival was 10.6 months versus 11.3 months, without a meaningful survival advantage1. Survival data from a similar adagrasib trial have not been published1.
In earlier single-group studies, without a comparison arm, median survival was 12.5 months for sotorasib5 and 12.6 months for adagrasib6. Those numbers are close to the 12 months seen in Krascendo 1’s comparison group, though studies with different patients cannot be compared directly.
The divarasib results arrived a day after another lung cancer result drew a standing ovation at the same meeting: a combination of tarlatamab (Imdelltra) and durvalumab (Imfinzi) in small cell lung cancer1, which Biohack covered in an earlier article. Scrip has also noted that other rival KRAS drugs are in development3.
What to watch next
- Regulators. Divarasib is not approved anywhere for routine use1. Filings with regulators would be the next step to watch.
- Earlier use. A second large trial, Krascendo 3, is testing divarasib against immunotherapy or watchful waiting in people whose stage II to IIIB tumors were removed by surgery2.
- Longer follow-up. Later survival analyses and fuller side-effect data from Krascendo 1 will show whether the gap holds over time.
