Combinations Targeting EGFR and KRAS G12C May Advance NSCLC Treatment

Although KRAS G12C inhibitors like sotorasib (Lumakras) and adagrasib (Krazati) have launched the non–small cell lung cancer (NSCLC) treatment arena forward, combining these drugs with other agents that deactivate KRAS could make these inhibitors more effective, according to Timothy Burns, MD, PhD.

“We’ve been trying for 30 years to target KRAS, including [KRAS G12C], which [represents] 11% of lung adenocarcinomas. To put that in perspective, ALK-[mutant disease represents] 7% [of lung adenocarcinomas],” Burns said in an interview with OncLive® following an OncLive^® State of the Science Summit^™ on lung cancer, which he chaired.

In the interview, Burns discussed the evolution of EGFR-targeted NSCLC treatments, the importance of delineating small cell lung cancer (SCLC) subtypes, and the advantages of continued antibody-drug conjugate (ADC) development across all areas of lung cancer.

Initial data from the phase 3 ADAURA trial (NCT02511106) showed that treatment with osimertinib (Tagrisso) led to a 90% 2-year disease-free survival (DFS) rate in patients with stage II to IIIA EGFR-mutated NSCLC.¹ Burns explained how these favorable data have shifted perspectives surrounding osimertinib. He also highlighted potential treatments for patients with KRAS G12C–mutant NSCLC and active, untreated brain metastases, in whom preliminary data from the phase 1/2 KRYSTAL-1 trial (NCT03785249) showed that adagrasib demonstrated a systemic overall response rate (ORR) of 35.0%.²

Burns is an associate professor of medicine, the associate program director for Research, and the associate program director for the Hematology/Oncology Fellowship Program in the Department of Medicine in the Division of Hematology-Oncology at the UPMC Hillman Cancer Center in Pittsburgh, Pennsylvania.

OncLive^®: How has the treatment landscape for EGFR-mutant NSCLC expanded beyond TKIs?

Burns: EGFR-mutant disease covers the spectrum of patients we treat [with this mutation]. [EGFR-targeted therapy] was the first targeted therapy approved for lung cancer, and we [now have] better drugs that provide longer progression-free survival [PFS] and better overall survival [OS]. Along the way, we’ve learned that treating brain metastases is important. The drugs have evolved to treat that.

Our metastatic standard of care [SOC] that current efforts are based on is the third-generation TKI osimertinib. Bringing it to the adjuvant setting [is exciting]. [We have] recent data there. Unfortunately, in the metastatic setting, with [our available] EGFR TKIs, although patients can live for years, we’re not doing a great job of treating them. Treating them in the earlier stages can have the most benefit.

In ADAURA, patients with stage IB, II, or IIIA resected disease [received] chemotherapy followed by 3 years of osimertinib. The initial data were so positive that an independent data-monitoring committee closed the trial early because they didn’t think it was ethical to keep patients on placebo. [This complicates] how we look at survival.

Now, we have 4-year survival data. The 4-year [median] DFS was [65.8 months] in the stage II and IIIA population vs [21.9 months] in those who did not receive osimertinib, with an HR of 0.23, almost a 4-year DFS improvement.

We’re still waiting for the OS data. Is 3 years [of treatment] enough [to improve survival]? As the patients stop [treatment] at 3 years, will these curves come together? Is this delaying recurrence, or curing the patients? If it’s delaying [recurrence], the difference is so great that we may still consider using this [drug], but we may consider [continuing] it beyond 3 years. We need to wait for the OS data, but it’s a discussion that many of us will have with our patients now.

Additionally, in the early-stage setting, neoadjuvant therapies, [such as] chemotherapy and in some cases, immunotherapy, have been exciting. This is another setting where EGFR has come into play. Patients with EGFR mutations were excluded from [the] trials [with immunotherapy].

[The phase 3] CheckMate 816 trial [NCT02998528] regimen increased pathologic complete response rates by 21.8% in patients with EGFR wild-type disease. Studies [such as the phase 3] NeoADAURA trial [NCT04351555] are evaluating osimertinib alone vs osimertinib with chemotherapy or chemotherapy alone. We’re hoping to see up-front osimertinib improve survival [in this population].

For patients with definitive unresectable disease, the SOC is chemoradiotherapy. We now have 5-year survival data with that [approach]. However, patients with EGFR-mutant disease didn’t seem to benefit [from that strategy]. There is an open debate about whether we should give chemoradiotherapy followed by immunotherapy in these patients. They don’t seem to benefit as much, and there may be adverse effects [AEs] if you then give them a TKI.

The [phase 3] LAURA trial [NCT03521154] is investigating osimertinib after chemoradiotherapy. What do we do in the metastatic setting after patients progress on osimertinib? Although this is the first targeted therapy [in NSCLC], it is not the most successful. Patients get an average of 19 months before they progress on osimertinib. The OS is about 3 years. However, with other therapies targeting ALK and ROS1, the median OS is over 5 years.

There’s much excitement in the metastatic setting about what we do after osimertinib. In some studies that we’ve participated in or that are ongoing, we’re trying chemotherapy [combinations] after progression on osimertinib or chemotherapy up front with osimertinib. That’s based on older data suggesting that patients do better with TKIs combined with chemotherapy. [We’re 1 of the sites participating in the phase 3] FLAURA2 study [NCT04035486], which is accrued. We’re looking forward to seeing if that [approach] will improve PFS and OS.

After years of depending on the TKIs alone, we finally have other EGFR agents that are not TKIs. One of the most exciting is amivantamab-vmjw [Rybrevant], a bispecific antibody targeting EGFR and MET. Data from several years ago showed that even alone, this agent had activity in patients who progressed on osimertinib. Now, presentations have shown that combining it with another TKI, lazertinib [Leclaza], [leads to] response rates in the 30% range in patients who have progressed on osimertinib, regardless of whether they have a MET mutation. Excitingly, [we’re also] combining amivantamab with chemotherapy after osimertinib, with similar response rates of about 50%. In the next few years, we may see our first post-osimertinib approvals.

Whenever a patient progresses on osimertinib, [our standard approach] is biopsy. Subcategories, like MET, RET, and other targetable alterations or second-site mutations in EGFR, can be targeted. We’ve done this for a while, using the biopsy to decide the best therapy. Most of that is [under investigation], but we may see approvals of amivantamab and lazertinib, as well as of combinations with osimertinib that include either chemotherapy or MET inhibitors.

How might combination therapies expand the KRAS G12C–mutant NSCLC treatment arena?

We’re excited to have targeted agents for KRAS, 1 of the first oncogenic drivers, which we’ve known about for decades. For about 40% of these mutations in lung cancer, KRAS G12C mutations, we have 2 FDA-approved drugs.

These drugs work when KRAS is in its inactive form. When KRAS becomes mutated, it is primarily in its active form, which binds to GTP. We thought for years that KRAS always remains in its active form. However, rarely, it goes back and forth from its inactive form. Both approved drugs, sotorasib, which was approved [based on the phase 1/2] CodeBreaK 100 study [NCT03600883], and adagrasib, which was approved [based on] KRYSTAL-1, bind to KRAS in its inactive form.

The [response rates associated with these] drugs are not what we expect from targeted therapies. The response rates are anywhere between 37% and 40% in the second line and beyond, with a disease-control rate of 80%. That’s still a high response rate. But with targeted therapies, we usually expect [response rates of] 60% to 70%. [The fact that] these drugs bind to the inactive form of KRAS [may be] what’s holding them back.

Other drugs in clinical trials may be more active. However, [to evolve this field], we may start evaluating combinations with these drugs to try to overcome resistance and make them work better up front. It’s a different way of looking at combinations. Often, you’re adding a second drug to inhibit a pathway upstream of KRAS that will [force it into its] active form. Drugs that shuttle KRAS into its inactive form will allow the initial drug to bind and work better. These drugs have a decent response rate, a PFS of about 7 months and an OS of 12 months. [However, we need to] make these drugs work better. Other inhibitors are in clinical trials. These drugs [can be associated with] gastrointestinal AEs and liver abnormalities but are generally well tolerated.

We have phase 3 data for sotorasib, the first [KRAS G12C inhibitor] that was approved and that we have phase 3 data for, showing that it is better than docetaxel. [These data] confirmed that the response rate in the second-line setting is not as good as we’d expect for most targeted therapies, but is at least better than docetaxel, which established sotorasib as a SOC.

Adagrasib, the second [KRAS G12C inhibitor] approved, has a similar mechanism to sotorasib, with similar response rates, duration of response, PFS, and OS. It seems to have more GI toxicities, which may be from the [study] doses. [As evidenced at the 2022 ASCO Annual Meeting], adagrasib works against brain metastases. In lung adenocarcinoma, about 30% to 40% of patients will get brain metastases, including in the KRAS G12C–mutant population.

What about putting this [agent] in the first line? Most of our targeted therapies are in the first line. Right now, the answer is no. These drugs don’t work as well as chemoimmunotherapy. Patients with KRAS G12C [mutations], unlike other targeted alterations, like EGFR [mutations], respond well to immunotherapy. In the first-line setting, [outside of] clinical trials, we’re still giving chemotherapy or pembrolizumab [Keytruda] alone or chemotherapy with pembrolizumab or other PD-1 and PD-L1 agents, because these patients do better than those with wild-type disease.

How do we get to where we’re putting this in the first-line setting? Combinations. There are trials of newer drugs, and maybe they’ll be better, but they’re all inhibiting KRAS in its inactive form. If you’re not getting KRAS into the inactive form, none of these drugs can bind.

Interestingly, resistance mechanisms may be different for these drugs. For sotorasib, the resistance mechanisms are primarily upstream of KRAS. For adagrasib, they’re primarily in the KRAS gene itself. This indicates the value of the combination studies we’re running.

Can you combine these [agents] with immunotherapy? Some interesting data suggested that in mouse tumors, KRAS G12C inhibitors inflame the immune system. When we add an immunotherapy agent, we can, in some cases, cure mice that we can’t cure with a KRAS G12C inhibitor or a PD-1 inhibitor alone.

Sotorasib was the first to do this. I [was an investigator on] this study. Unfortunately, in 1 of the cohorts from [the phase 1/2] CodeBreak 101 trial [NCT04185883] that was presented at the International Association for the Study of Lung Cancer 2022 World Conference on Lung Cancer, we couldn’t safely combine sotorasib at reasonable doses with immunotherapy without getting high rates of hepatitis. That resulted in a response rate similar to or less than that with a single agent alone and hurt the durability of those responses.

There are now promising data from KRYSTAL-1 and [the phase 2/3] KRYSTAL-7 trial [NCT04613596] with adagrasib plus pembrolizumab in the first-line setting. Pasi Jänne, MD, PhD, [of Dana-Farber Cancer Institute in Boston, Massachusetts], presented these results at the 2022 ESMO Immuno-Oncology Annual Congress, showing in the first-line setting an ORR of 49%, which may be better than the combination of the 2 agents or either agent alone in the first-line setting. There was activity in the PD-L1–negative population, about 30%, which [may be similar to that] with adagrasib alone.

However, the investigators didn’t see high rates of toxicity [with the combination], so they could combine [these agents]. The response rate is probably less important than the fact that it’s safe. We’ll need to see the durability of these responses, [which if prolonged will] hopefully lead to increased survival.

There’s much excitement in the KRAS G12C area. We’re not at the beginning of the end, we’re at the end of the beginning. We’re just getting started with these agents, and I expect to see more approvals of other agents. However, what will be a game-changer is the approval of combinations that will make these [agents] work more like the targeted therapies for other alterations.

Regarding the presentation by Stephen Liu, MD, of Georgetown Lombardi Comprehensive Cancer Center in Washington, DC, how do the developments with ADCs in NSCLC compare with the benefits seen with this class of agents in breast cancer?

ADCs have been around for several years. [There have] been advances in the antibodies and in the drugs they carry along. We’ve finally seen approvals of ADCs in [lung cancer], which have been approved in breast cancer for the past 5 years.

Several drugs are under investigation, but we have our first approval of fam-trastuzumab deruxtecan-nxki [Enhertu], a combination of trastuzumab [Herceptin] plus a chemotherapy moiety. This drug is the first to show activity in HER2-mutant [NSCLC], which is a relatively small percentage of lung cancers, about 2% to 3%. The response rates, even in heavily pretreated patients, were around 50%. These responses were durable. [This is] our first approved non-TKI targeted therapy for HER2-mutant disease.

There’s much excitement around this [treatment class], especially in HER2-mutant disease, as we’re looking to extrapolate from breast cancer and bring these chemotherapies to tumors expressing HER2. [The phase 2] DESTINY-Lung01 trial [NCT03505710] evaluated patients expressing HER2 2+ or 3+ [via immunohistochemistry] and showed [an ORR of 55%], which [is a relatively good ORR beyond] the first-line setting.

Several targets beyond HER2, some oncogenic drivers, and some populations of SCLC or mesothelioma may respond [to ADCs]. Although these drugs are not new, there are new ways of linking drugs to antibodies. Even in the HER2 space, an older drug, ado-trastuzumab emtansine [Kadcyla], was quite successful in breast cancer, but didn’t do much in lung cancer in the same population. [We need] the right antibody and the right drug. There can be AEs with ADCs. However, this is another way of targeting oncogenic drivers. EGFR, MET, and many oncogenic drivers or receptor tyrosine kinases could be targeted with these technologies.

Regarding the presentation by Yvonne Chao, MD, PhD, of UPMC Hillman Cancer Center, which patients with NSCLC may respond best to immunotherapy?

A subset of patients does well with immunotherapy. I’ve treated patients in the first-line setting who are still alive 7 years later and are not being treated [anymore]. Patients can have durable responses. We’ve had many immunotherapy regimens approved, [like] monotherapies, immunotherapy combinations such as nivolumab [Opdivo] plus ipilimumab [Yervoy], and chemotherapy combinations. However, we need to do a better job of defining which patients will benefit.

The biomarkers we currently have, including PD-L1, tumor mutational burden, and gene signatures that reflect inflammation in the tumor, are poor biomarkers and have not helped us select these patients. Our best selection is excluding all patients with oncogenic drivers except for BRAF and KRAS. By excluding those [populations], we can enrich for patients who can respond to these therapies. This is an active research area. Who are those patients? Then, for the other patients, can we add other agents [to these immunotherapies]? We’re exploring several agents, like TIGIT, LAG-3, and TIM-3 [inhibitors], in the first-line setting and in patients who respond and then progress. That’s where [we] may make the most progress.

Regarding the presentation by Antoinette Wozniak, MD, FACP, FASCO, of UPMC Hillman Cancer Center, where is the treatment of SCLC headed?

The progress in SCLC has been glacial. The first-line SOC was carboplatin or cisplatin and VP-16, [which produced] high response rates, although almost every patient relapsed. Now, we have 2 [immunotherapies], both PD-L1 inhibitors. One is atezolizumab [Tecentriq], and the [other is] durvalumab [Imfinzi] given with chemotherapy, and then as maintenance, which has shown a several-month survival improvement.

Despite many trials, SCLC has had few advances. We have 2 new approvals in the second-line setting and beyond, 1 with lurbinectedin [Zepzelca]. There are still open questions regarding who the right patients are for [lurbinectedin]. Some data [indicate] that it may benefit patients who don’t respond well to chemotherapy, but we still have a long way to go.

We’re starting to divide SCLC into several biological subtypes based on immunohistochemistry staining and gene signatures. We’re testing drugs that make sense in those subtypes, including drugs that may have failed in the past because we gave them to every patient with SCLC. That may be where we will make progress in years rather than decades.

What ongoing lung cancer research at UPMC Hillman Cancer Center would you like to highlight?

We’re about to open [a trial] we’re excited about. Around 30% of patients will have brain metastases at diagnosis, and 40% will have them by the end of their disease. The current standard is radiation. After radiation, in some patients with targets, targeted therapies can help. However, for those without oncogenic drivers, we don’t have good treatments for brain metastases.

As a team effort at UPMC and the University of Pittsburgh, we’ve identified that MET amplifications, targetable with the FDA-approved MET TKIs crizotinib [Xalkori], tepotinib [Tepmetko], and capmatinib [Tabrecta], seem to be enriched in brain metastases, even when we don’t find them in the primary tumor. This is a fundamental change. We normally biopsy a site to see what the patient has, but patients with wild-type disease will often have other alterations at other sites. MET amplifications [are found in the brain] in 16% to 20% of patients.

We’re doing preclinical and translational work now, but we’re about to open an investigator-initiated trial selecting patients using circulating tumor DNA [ctDNA]. From preliminary data, we can detect about half of patients who will have these alterations in the brain. [We will try to] treat these patients with a MET TKI. This is a population where we often gain systemic disease control but can’t control the brain metastases. It’s a sad state when patients are otherwise doing well. This [trial] is tackling that population and using ctDNA to identify them because if we biopsy the primary lung, it may not be MET amplified, but we know [the MET amplifications may be] out there. We’re excited about this trial because it would [lead to] the first targeted therapy specific to brain metastases. Beyond [the whole tumor having] 1 oncogenic driver, this is looking at site-specific vulnerabilities.

[Additionally], our group and several others are running combination trials with KRAS G12C inhibitors. We may move beyond what those initial drugs can do.

Editor's Note: Dr Burns is on advisory boards for Amgen, AstraZeneca, Blueprint Medicines Corporation, EMD Serono Inc, Janssen, Mirati Therapeutics, and Takeda; is on the data and safety monitoring board for Advarra, Inc; and has received research funding for an investigator-initiated trial from Novartis (Inst.).

References

1. Wu YL, Tsuboi M, He J, et al. Osimertinib in resected EGFR-mutated non-small-cell lung cancer. N Engl J Med. 2020;29;383(18):1711-1723. doi:10.1056/NELMoa2027071
2. Sabari JK, Spira AI, Heist RS, et al. Activity of adagrasib (MRTX849) in patients with KRASG12C-mutated NSCLC and active, untreated CNS metastases in the KRYSTAL-1 trial. J Clin Oncol. 2022;40(suppl 17):LBA9009. doi:10.1200/JCO.2022.40.17_suppl.LBA9009