Biomarkers Light the Way for Personalized Treatment in NSCLC

Although comprehensive biomarker testing should be made available to every patient with advanced non–small cell lung cancer (NSCLC) up front to personalize care, not all patients with alterations that have matched targeted therapies should receive them in the frontline setting, said Jyoti D. Patel, MD, citing KRAS G12C as a prime example.

“Personalized care continues to be an important part of lung cancer treatment. Certainly, we have a lot of tools to look at molecular markers and blood and tissue biopsies to delineate best care. The group of patients with appropriate targeted therapies continues to increase, and we continue to do early trials on [alterations] like NRG fusions and continue to hone what we’re doing for [patients with] common [alterations] like EGFR [sensitizing] mutations,” Patel, a professor of medicine (hematology and oncology) and associate vice chair for clinical research in the Department of Medicine at Feinberg School of Medicine, said in an interview with OncLive® during an Institutional Perspectives in Cancer webinar on lung cancer.

“We have a long way to go, and it’s likely that biomarkers will lead that way,” Patel added.

In the interview, Patel, also medical director of thoracic oncology and assistant director for clinical research in the Robert H. Lurie Comprehensive Cancer Center at Northwestern University, discussed the main highlights from the meeting, which focused on biomarker testing, interpreting molecular result reports, and the growing armamentarium for patients with oncogene-driven disease.

OncLive®: What has been your experience with the approved targeted agents for patients with RET alterations and KRAS G12C mutations?

Patel: Both oncogenic drivers now have multiple drugs associated with them. RET translocations are found in a small subset of patients that drive a clinical course in never smokers and younger patients. That [population] has been an unmet need for a long time despite several TKIs that had MET activity. It’s truly these much more specific drugs, selpercatinib [Retevmo] and pralsetinib [Gavretto] that are much more potent, with deep response rates, good tolerability, and long duration of response [DOR].

In terms of toxicity profiles, these drugs are well tolerated although there are some toxicities like hypertension and some transaminitis with selpercatinib, but generally patients do not have to discontinue the drug and can expect good responses.

In addition, both agents have CNS [central nervous system] response, which is a huge unmet need for our patients. Certainly, these drugs are appropriate in the frontline setting. If a patient needs to start some treatment prior to getting NGS [next-generation sequencing] results, which would be a minority of patients, [these drugs] would be appropriate in the second-line setting. We’re just now starting to understand resistance mechanisms, so that area will continue to evolve.

I think [selpercatinib and pralsetinib are] equivalent drugs. I’ve had more experience with selpercatinib, so I tend to use that a little bit more, but to me, there’s no discernible difference in efficacy or tolerability [between the agents].

For KRAS G12C, we now have sotorasib [Lumakras] approved. KRAS G12C is [found in] a substantial subset of patients with NSCLC, and sotorasib was approved based on findings from a phase 1/2 study in which patients who had prior treatment received sotorasib once daily. Responses were seen in about one-third of patients, and the progression-free survival and DOR of up to 11 months really solidifies this as a good second-line agent.

At this juncture it’s unlikely we’d move sotorasib alone into the frontline setting given higher response rates with triplet chemotherapy and immunotherapy. Certainly, in the second-line setting, we see some impressive responses [with that agent].

The KRAS G12C story is particularly exciting for us because it was the undruggable target for so long.

In terms of toxicity, the drug is well tolerated.

More recently, the Mirati compound adagrasib has also been shown to be effective in a similar patient population. It’s not yet FDA approved, but we hope we will see its approval soon. One area in which we’re really learning a lot more about these drugs is the influence of co-mutations. Things like STK, KEAP1, and P53 may change response rates, which is why it’s important to enroll patients on Lung-MAP [NCT03851445].

Jeremy Segal, MD, PhD, of the University of Chicago, discussed the larger framework of biomarker testing. What did he emphasize?

Dr Segal’s presentation really helped introduce us to much of the technology that is available and the need for comprehensive testing and different assays that we can use from DNA to RNA signaling. He stressed just how scalable this technology has become. Something that was much more cumbersome and took a long time and was much more expensive now is something that every patient should have up front.

In terms of biomarker pearls, [he emphasized that] this idea of a larger data set is important. How do we really harness co-mutations and look more comprehensively in a larger number of patients to make decisions and to understand resistance mechanisms? The technology and the platforms are ever evolving, but this piece of rapidly moving data means we probably need more of it. [Dr Segal] is part of that collaboration of sites, looking to pool data and to understand some of these best practices.

Let’s move to the presentation Christine Bestvina, MD of the University of Chicago, gave on interpreting molecular testing results. What are some of the main barriers we’re seeing?

Dr Bestvina stressed a couple points, including how we order [tests], so this piece of reflex ordering and delays to ordering. Leveraging what we can in the EMR [electronic medical record] and developing workflow within institutions to get rapid resulting is important. Another piece that Dr Bestvina stressed was how we can use send out labs within our EMR, [avoiding] the pitfalls of having outside scans uploaded in various tabs and not having timely results. How to develop an infrastructure for each working group is important.

The piece that really struck home was the importance of understanding what we’re reading in terms of different mutations and levels of evidence, in which we’d be able to ascertain a particular marker. That piece of understanding the difference between an EGFR exon 20 insertion or 19 deletion and having the context around that [is critical]. How we can improve reporting to clinicians is certainly important. [She also highlighted] the multitude of ways that we can use these molecular tests. For example, do we use minimal residual disease [MRD] at diagnosis? Do we consider it when we’re thinking about intensifying therapy? Considering all those other research pieces, it’s an exciting topic.

She also highlighted the IT infrastructure that one needs as well as through the team-based care because there are multiple touch points from IR [interventional radiology] or IP [interventional pathology] to pathology to the medical oncologist that need to work in concert.

How could some of these best practices trickle down to community practices?

Significant communication with a pathologist and deciding whether that pathologist is going to reflexively order [the test] or whether it’s going to come from medical oncology is important. Most of us think that the pathologists should own that ordering piece. Certainly, that has been a significant change over the past several years, but really, the pathologists can move this [piece] most quickly, rather than waiting until it ends up in an EPIC basket.

Angel Qin, MD, of the University of Michigan, spoke about frontline immunotherapy. Are treatment recommendations largely dictated by PD-L1 expression?

For immunotherapy, there are a portion of patients that do very well with immunotherapy alone, but many of them require a combination of chemotherapy and immunotherapy and judging the best strategy is very patient centered and [should include] shared decision making.

Dr Qin gave a nice description of the significant number of options we have right now. She broke the options up by high PD-L1 expression in mutation-negative cancers––those patients by and large, should move forward with pembrolizumab [Keytruda]. She then broke down the less than 50% population by different clinical trials, whether it was CheckMate 227 [NCT02477826] with ipilimumab [Yervoy] and nivolumab [Opdivo], CheckMate 9LA [NCT03215706] with ipilimumab/nivolumab and chemotherapy, or the KEYNOTE studies [with pembrolizumab], and looked at toxicity, DOR, and response rates.

For most patients, doing a chemotherapy and pembrolizumab combination in those with PD-L1 expression under 50% is the most appropriate.

She also gave us a comprehensive view of different options, and then one that was very patient focused. What toxicities does that patient care about? How do you weigh response rate and toxicity? She was also able to give a longer timeline, so [one that encompassed] more mature overall survival data. There are a multitude of options, but primarily, what we are seeing in the United States is a combination of chemotherapy and PD-1 inhibitor for patients with low PD-L1 expression.

[Now, we’re wrestling with what to do] after progression on frontline PD-L1 inhibition. In an appropriate patient population at progression, do you add chemotherapy to pembrolizumab? INSIGNIA [NCT03793179] will be an important study for us to really understand [that]. We’ll further define how we look at patients with less than 50% PD-L1 expression and the large number of patients who have PD-L1 expression over 50% who don’t respond as long as we would like [to frontline therapy]. I certainly think that will be important.

Other studies that Dr Bestvina talked about that may have clinical applicability soon is looking at MRD and intensification of therapy. Those are the 2 biggest practice-changing and widely applicable trials that we have right now.