Krascendo 1 Survival Data Position Divarasib to Reset Second-Line Standards and Fuel First-Line Combinations in KRAS G12C–Mutated NSCLC

The phase 3 Krascendo 1 trial (NCT06497556) showed that divarasib improved progression-free survival (PFS) and overall survival (OS) compared with sotorasib (Lumakras) or adagrasib (Krazati) in previously treated KRAS G12C–mutated non–small cell lung cancer (NSCLC). The survival advantage positions the next-generation KRAS G12C inhibitor as a potential new standard in this setting, according to Biagio Ricciuti, MD, PhD.^1,4

"These are remarkable results, and in my mind, these are potentially practice changing," Ricciuti said in an exclusive interview with OncLive®. "What stood out to me the most is not just the improvement in [PFS], which was already pretty robust and impressive, but also the magnitude of the [OS] benefit."

Median PFS per blinded independent central review was 9.8 months with divarasib (n = 167) vs 5.7 months with sotorasib or adagrasib (n = 171; HR, 0.58; 95% CI, 0.44-0.76; P < .0001).¹ At the first interim OS analysis, published in The New England Journal of Medicine, median OS was 17.7 months vs 12.0 months, respectively (HR, 0.56; 95% CI, 0.41-0.78; P = .0005). The confirmed objective response rate was 58% (n = 94 of 161) vs 32% (n = 54 of 167) among response-evaluable patients.

Treatment-related adverse effects led to discontinuation in 4.8% of patients receiving divarasib vs 11.7% of those in the control arm, despite a median treatment duration of 11.3 months vs 5.0 months. Diarrhea, nausea, and vomiting were more frequent with divarasib, whereas transaminase elevations were less frequent.

In the interview, Ricciuti discussed how co-mutations inform the sequencing of immunotherapy and KRAS G12C inhibitors, the rationale for first-line combinations, emerging RAS(ON) inhibitors such as daraxonrasib (Rasonque), which has produced durable responses in previously treated RAS-mutant NSCLC,⁵ and management of progression on a KRAS G12C inhibitor. Ricciuti is a thoracic medical oncologist and scientist at the Lowe Center for Thoracic Oncology at Dana-Farber Cancer Institute, where he is affiliated with the Center for RAS Therapeutics, and an instructor in medicine at Harvard Medical School in Boston, Massachusetts.

OncLive: How has direct KRAS G12C inhibition changed practice in NSCLC, and where have sotorasib and adagrasib fallen short?

Ricciuti: The development of KRAS [inhibitors] has probably been one of the most relevant breakthroughs in precision oncology over the last decades. As you mentioned, over the last 40 years, KRAS has been largely considered [undruggable], primarily because of [its] structural properties and its high affinity to GTP. The discovery of [a] druggable pocket has been transformative for our community. Sotorasib and adagrasib, [which] are 2 first-generation [G12C OFF] inhibitors, have shown the proof that we could directly block mutant KRAS, specifically G12C, in patients, and induce meaningful benefit in tumor responses. These have changed the way we treat our patients and have changed the therapeutic landscape for patients with KRAS G12C–mutant lung cancer.

These are first-generation drugs, and as you also mentioned, there are certainly limitations to their use. First of all, if we think about the response rate of these therapies, that's ranging from 28% to 40%, depending on the study we look at, which means that 60% to 70% of patients don't have an objective response to these therapies, and even among patients who respond, acquired resistance develops. If we look at the [PFS] with these therapies, which ultimately is a measure of time to resistance, that is usually in the 5- to 6-month range, [which is] certainly not ideal for our patients and for us as providers.

What we learned through these trials is that KRAS-[mutant] lung cancer is also highly heterogeneous. It's not a single entity. The presence of [a] G12C mutation alone doesn't predict how well a patient will respond to these therapies. There are co-mutation patterns, immune phenotypes, and many factors that we are still learning about that may influence how these therapies work. This has been transformative, but [there is] still a long way to go to improve outcomes for these populations.

How should clinicians weigh STK11 and KEAP1 co-mutations when sequencing immunotherapy and a KRAS G12C inhibitor?

This is an important question because over the last years we have realized that, as I mentioned, KRAS-[mutant] lung cancer is highly heterogeneous from a genetic standpoint, particularly if we think about tumor suppressor genes. Mutations in STK11, KEAP1, and also [SMARCA4] can occur in a large fraction of patients with KRAS-mutant lung cancers, particularly in patients with KRAS G12C mutations, along with G12V and [G12A]; this seems to be less common in other KRAS variants. A large fraction of patients, 30% to 40%, may have one of these tumor suppressor genes mutated on genetic profiling, and this has multiple implications.

First of all, these mutations tend to [confer] an immune-cold phenotype to these tumors. We see [a] low density of T cells, low or negative [PD-L1] expression levels, and deficient antigen presentation machinery, [which] explain why immunotherapies don't necessarily work well for patients with KRAS mutations who also have concurrent STK11 or KEAP1 loss. We know that this is true for PD-1 monotherapies. We also have data from multiple retrospective studies and even prospective trials that have shown that even when we combine chemotherapy and PD-1 blockade, outcomes tend to be inferior in patients with [these co-mutation] patterns.

Today, chemoimmunotherapy or immunotherapies remain the standard-of-care first-line option for patients with KRAS [G12C] mutations or KRAS mutations in general; therefore, the way we sequence these therapies today is still using immunotherapy-based treatments first and RAS-based therapies next. However, this might change in the future, as first-line trials are now exploring the combination of KRAS inhibitors with immunotherapies. These strategies, which are not yet approved but are being explored in multiple clinical trials, will ultimately redefine the treatment landscape for patients with these co-mutation patterns.

Blocking RAS can be extremely synergistic with immunotherapies. When we block KRAS mutations with RAS inhibitors, and this has been shown across a wide range of inhibitors, we can increase PD-L1 expression levels. We can bring T cells to the tumor. We can increase HLA class I expression. [These are] all things that ultimately can make these tumors more sensitive to immunotherapy. There is synergy, and this is going to be particularly relevant for patients with co-mutations in STK11 or KEAP1, because we know that KEAP1 in particular confers resistance not only to immunotherapies, but also to RAS inhibitors, at least [G12C OFF] inhibitors. We have seen this with sotorasib and adagrasib, for example, at least in phase 2 and phase 3 studies when these agents were used as monotherapies.

Something worth mentioning here is that dual immunotherapy combining PD-1 blockade with CTLA-4 inhibition, with or without chemotherapy, can be beneficial for patients with co-mutations in STK11 and KEAP1. We have many retrospective data, post hoc analyses from randomized clinical trials, and meta-analyses of multiple randomized studies that have consistently shown how CTLA-4 blockade can potentially benefit these patients, who are largely immune depleted with an immune-cold microenvironment. These are patients who may benefit from T-cell priming, which is what we achieve by giving CTLA-4 blockade.

What stood out about the efficacy of divarasib in Krascendo 1, and how should clinicians read the size of the benefit?

[The OS benefit is notable] at least in terms of reduction in the risk of death with divarasib vs the first-generation [G12C OFF inhibitors] that were included in the control arm of this study, sotorasib [and adagrasib]. I will note that most of these patients received sotorasib, which was the most widely available [G12C OFF] inhibitor when the trial was run. We are talking about a [PFS] of 9.8 months with divarasib compared with 5.7 months with either sotorasib or adagrasib; the [HR] was 0.58, suggesting a 40% decrease in risk of progression or death with divarasib.

Divarasib is structurally possibly more potent [than other G12C OFF] inhibitors and has consistently been shown to produce better response rates and longer [PFS] in phase 1 and phase 2 studies. It's now great to see, in a randomized fashion, a head-to-head comparison of this drug with the existing [G12C OFF] inhibitors, and [to see] such a benefit in terms of response rate, [PFS], and [OS]. [This is] probably practice changing in my mind, also because if we think about sotorasib and adagrasib, those 2 drugs did not improve [OS] compared with docetaxel in their respective randomized phase 3 studies. Here we are finally seeing [an OS] benefit, which is one of the ultimate end points for our patients in randomized studies, suggesting that these patients may live longer because of this drug.

Something relevant, of course, is the toxicity. We've seen in Krascendo 1 a bit more [gastrointestinal] (GI) [adverse effects] with divarasib. That's something that differentiates these drugs as well, not only efficacy but also safety. On the contrary, [hepatotoxicity], which is something we have seen more commonly with sotorasib [or adagrasib], was less frequent with divarasib. Perhaps the most important finding was that the discontinuation rate [due] to adverse [effects] was also lower, 5% vs 12%. This is important because, for example, liver toxicity may lead to permanent discontinuation, but [we] can probably manage grade 1 or 2 diarrhea and keep patients on therapy longer. The spectrum of toxicity certainly differentiates these drugs.

Something we need to keep in mind [is that] because divarasib extends [PFS] and also [OS], we expect these patients to be on therapy longer, so we need to familiarize [ourselves] with the safety profile of this drug and come up with strategies to proactively manage adverse [effects].

Should KRAS G12C inhibitors be combined with immunotherapy up front, and what data are needed from first-line and adjuvant trials before practice changes?

That's probably the most exciting area where we are headed. We know that [KRAS G12C inhibitors] have been active in previously treated patients who have received prior chemotherapy and immunotherapies, but biologically, there are many reasons to believe that moving these agents into the first-line setting would ultimately improve [outcomes] even further. That's a lesson we've learned from many other targeted therapies in lung cancer and other tumor types: whenever we can bring these therapies to the first-line setting, outcomes are better.

In my mind, I don't think [a RAS inhibitor] will ever make it to the first-line setting as a single agent. Most studies being run in that setting are [evaluating] combination therapies, and the most reasonable, rational, and biologically sound companion for RAS inhibition right now is immunotherapy. As I mentioned, RAS inhibition is synergistic with immunotherapy [and] can potentially convert immune-cold tumors to immune-hot tumors. Because KRAS-[mutant] lung cancers are so heterogeneous, tackling that heterogeneity with combination strategies is extremely important. We have preclinical evidence that RAS inhibition can modulate the tumor microenvironment, increasing antigen presentation and T-cell infiltration, and can make these tumors more sensitive to checkpoint blockade. For these reasons, this is the most rational combination.

We have early clinical data from combination therapies. For example, [in the phase 1b/2] Krascendo 170 [trial; NCT05789082], divarasib combined with pembrolizumab [Keytruda] has shown [a] 70% response rate [and a PFS] of 19 months.² We also have data with adagrasib, which is another [G12C] inhibitor we discussed today, in combination with pembrolizumab in patients who have high [PD-L1] expression levels, and we've seen [PFS] approaching 27 months.³ [This is] certainly unprecedented benefit that compares favorably with the efficacy we see in general with immunotherapies alone in the first-line setting.

That's where we are headed. There are phase 3 studies now looking at this question with divarasib, [olomorasib], and other [RAS inhibitors] that are being combined with immunotherapies. Of course, we need to be careful about the safety profile. As a community, we had a fairly bad experience with the combination of sotorasib with immunotherapies. Many patients had serious grade 3 or 4 liver toxicity events, so we certainly need to learn more. However, what we know about the next-generation RAS inhibitors, including not only divarasib or [olomorasib] but also pan-RAS inhibitors, is that these combinations are potentially feasible. Liver toxicity remains a major challenge, but we are learning how to manage it. By paying attention to the safety profile of these combinations and coming up with better strategies to prevent and manage this toxicity, we will be able to bring these therapies together in the first-line setting.

Where is the field headed with RAS(ON) inhibitors, multiselective inhibitors, and new combinations?

This field is growing exponentially. As you mentioned, we now have [G12C OFF] inhibitors, but we also learned that we can target RAS in its active ON state, GTP bound, and that's particularly relevant for certain KRAS variants. What we discussed today was primarily the targeting of G12C mutations, but the truth is that across tumor types, G12D is the single most common KRAS variant, and we have G12V, [G12R] [UNCLEAR — please review], codon 13 mutations, and codon 61 mutations. There are now multiple companies and drugs being developed to target RAS variants. We have pan-RAS inhibitors [and] pan-KRAS inhibitors. Some of these are ON inhibitors, some are OFF inhibitors, and some are dual ON/OFF inhibitors. A lot is happening, and targeting [KRAS] in its active form is one of the most recent breakthroughs in drug development for these patients.

We have seen interesting data from drugs such as [daraxonrasib] [Rasonque], or potentially even G12C mutant–selective RAS(ON) inhibitors, that have shown efficacy not only in patients with other KRAS variants, but also in patients with G12C mutations who have previously received [G12C OFF] inhibitors. There is a reason behind this: [daraxonrasib], for example, by targeting not only the mutant allele but potentially having broader coverage for other RAS [isoforms], [HRAS] and [NRAS], and wild-type RAS, can potentially overcome some of the adaptive resistance that can develop to [G12C OFF] inhibitors. That's why we think this may work in a population previously treated with [G12C OFF] drugs. This next class of drugs is particularly appealing, with promising preclinical and clinical efficacy seen across a wide spectrum of RAS mutations.

Speaking about pan-RAS or pan-KRAS inhibitors, these are now being combined with immunotherapies as well, and we are also seeing a growing number of early-phase studies looking at combinations with bispecific antibodies, [such as] PD-1/[VEGF bispecifics], and potentially in the future antibody-drug conjugates [ADCs]. These are all exciting scenarios where we hopefully [will] bring new therapeutics to our patients. Lastly, there is an effort to develop [adoptive] cell therapy, such as TCR-T therapies, for KRAS-mutant cancers, and we're still learning about these in early-phase studies, but [it is] certainly something to look forward to in the future.

What should clinicians do when a patient progresses on a KRAS G12C inhibitor?

In my practice, for patients who have experienced disease progression on [a] KRAS inhibitor, I profile those patients with repeat genotyping. I tend to do a repeat tissue biopsy [and a] liquid biopsy. We know there may be actionable resistance mechanisms in the literature. We've seen reported actionable fusions, for example [RET] fusions and [ALK] fusions, for which targeted therapies are available, so we can redirect certain patients to other targeted agents. We've also seen squamous cell transformation as a mechanism of resistance to sotorasib or [adagrasib]. This means we can redirect patients to histology-specific chemotherapy backbones.

We have also seen that there are second-site KRAS mutations that may confer resistance to one drug but not another. An example [is] H95 mutations. These are secondary mutations that may confer resistance to adagrasib, but in preclinical models, they can still retain sensitivity to sotorasib. Reprofiling our patients at the time of resistance is important. We are used to doing that for patients with EGFR mutations [and] ALK or ROS1 fusions, among others, but we need to get into the habit of doing this for our patients with KRAS [mutations] because there may be actionable mechanisms of resistance.

If we don't see anything that's actionable, in my practice I now tend to test my patients for new targets such as HER2 and MET by [immunohistochemistry]. We have FDA approval for 2 [ADCs] targeting these 2 proteins, so these are available on label for our patients. I'm increasingly testing for [MTAP] because [of] clinical trials, and we know that [MTAP] loss is relatively common in patients with KRAS mutations based on recent literature. Paying attention to these biomarkers is critical to inform subsequent therapies.

What is known so far about mechanisms of resistance to daraxonrasib?

While [daraxonrasib] has been developed, and we have seen the first published data from the phase 1/2 study, we are also already learning about the mechanisms of resistance to [the drug]. Even though this drug is not yet approved [in lung cancer], we are already learning what's potentially causing resistance. In the first report from the phase 1 study, a fraction of patients have been profiled with repeat genotyping, and we are learning that, for example, KRAS amplification can [mediate] resistance.

We are also learning that there are specific mutations, for example [KRAS] Y64 or Y71, that can confer resistance to [daraxonrasib] through divergent mechanisms. We've also seen that BRAF class 3 mutations are extremely common at the time of resistance to [daraxonrasib]; [they are] probably the most common genetically defined mechanism of resistance. Because we have [BRAF] dimer breakers or new agents that are [now] targeting BRAF class 3 mutations [UNCLEAR — please review], reprofiling patients upon progression [on daraxonrasib] may also be critical, because in the future we can envision a setting where we will have therapies that may target specific mechanisms of resistance, such as class 3 BRAF mutations. More is to come on this topic, but it's good to know that we are already learning what's causing resistance to this next generation of therapies.

Editor's Note: This transcript has been edited for grammar and clarity using artificial intelligence tools.

References

1. Skoulidis F, Ahn MJ, Gadgeel S, et al. Divarasib as compared with sotorasib or adagrasib in KRAS G12C–mutated NSCLC. N Engl J Med. Published online October 25, 2026. doi:10.1056/NEJMoa2605703
2. Skoulidis F, Cuppens K, Theelen W, et al. First-line (1L) divarasib plus pembrolizumab (pembro) in advanced or metastatic KRAS G12C+ non–small cell lung cancer (NSCLC): results from the Krascendo-170 study. J Clin Oncol. 2026;44(suppl 16):8510. doi:10.1200/JCO.2026.44.16_suppl.8510
3. Garassino MC, Theelen W, Jänne PA, et al. First-line adagrasib with pembrolizumab in patients with advanced/metastatic KRASG12C-mutated non-small cell lung cancer and PD-L1 ≥50% from the phase II portion of KRYSTAL-7. J Thorac Oncol. 2025;20(suppl 1):S8-S10. doi:10.1016/S1556-0864(25)00200-X
4. OncLive Staff. Divarasib improves survival and paves way for new SOC in pretreated KRAS G12C–mutated NSCLC. OncLive. Published October 25, 2026. Accessed October 30, 2026. https://www.onclive.com/view/divarasib-improves-pfs-and-os-vs-sotorasib-or-adagrasib-in-pretreated-kras-g12c-mutated-nsclc
5. OncLive Staff. Daraxonrasib yields durable responses in previously treated RAS-mutant NSCLC. OncLive. Published September 3, 2026. Accessed October 30, 2026. https://www.onclive.com/view/daraxonrasib-yields-durable-responses-in-previously-treated-ras-mutant-nsclc