Among a plethora of novel immunotherapeutic and targeted agents, the treatment landscapes for small cell lung cancer and non–small cell lung cancer have witnessed a rapid expansion that has led to improved survival and quality of life for patients, including a wider range of patients who harbor molecular abnormalities.
Jared Weiss, MD
Among a plethora of novel immunotherapeutic and targeted agents, the treatment landscapes for small cell lung cancer (SCLC) and non–small cell lung cancer (NSCLC) have witnessed a rapid expansion that has led to improved survival and quality of life (QOL) for patients, including a wider range of patients who harbor molecular abnormalities, said Jared Weiss, MD.
Moreover, practice patterns should reflect the growing armamentariums for lung cancer by ensuring treatment selection is informed by patient characteristics, the results of next-generation sequencing (NGS), and prognosis, added Weiss, an associate professor of medicine in the Division of Oncology of the Department of Medicine at the University of North Carolina School of Medicine, in an interview with OncLive® during an Institutional Perspectives in Cancer webinar on lung cancer.
The virtual meeting covered treatments in newly diagnosed and relapsed/refractory SCLC and advances with biomarkers and targeted therapies in NSCLC.
In the interview, Weiss, who chaired the event, discussed the evolving paradigm for patients with SCLC amid novel therapeutic advances, the importance of NGS in NSCLC, and additions to the targeted therapy armamentarium for a growing number of actionable alterations in NSCLC.
Weiss: There are absolutely no clinical drivers [to inform] the decision between atezolizumab and durvalumab. Both [agents] come across a backdrop of really no improvement in survival in frontline [SCLC] in a very long time. Historically, our standard of care before PD-L1 inhibitors was carboplatin and etoposide.
Carboplatin and etoposide supplanted CAV [cyclophosphamide/doxorubicin/vincristine], not based on a survival advantage, but on a toxicity advantage. The modest efficacy advantage provided by PD-L1 inhibitors is fresh water in a desert, but the drugs do not distinguish themselves from each other. The design of the studies [evaluating atezolizumab and durvalumab] and the inclusion criteria were more similar than different. The [survival] curves are nearly superimposable; the findings are nearly identical.
There are minor points of differentiation [between atezolizumab and durvalumab] that have arisen. Carboplatin, etoposide, and atezolizumab was the first to come out, which means that more [providers] have that [regimen] on their formulary and in their order sets. Does that mean it is better for patients? No. However, that will be a driver in the real world. The other [differentiation point] is that when trilaciclib [Cosela] was first studied, only atezolizumab [had been evaluated]. [Therefore], if we [to capitalize on] the QOL advantage that trilaciclib offers, the data exist with the combination of carboplatin, etoposide, and atezolizumab.
There is not a lot right now for patients with SCLC regarding the frontline setting. Drug development typically starts in the later-line settings and moves earlier.
We just saw data during the 2021 ASCO Annual Meeting on AMG 757, which is a bispecific T-cell engager against the target DLL3. The failure of Rova-T [rovalpituzumab tesirine], an antibody-drug conjugate, in this paradigm is somewhat famous. However, if we dig deeply into those data, it doesn’t look like DLL3 is a bad target, [rather] the conjugated toxin was toxic. The favorable data with AMG 757, particularly at higher doses, speaks positively about DLL3 as a target, as well as the idea of targeting preserved neoantigens in SCLC.
In that context, the other preserved neoantigen worthy of conversation is GD2. Clinical efforts are not nearly as advanced, but there is a promising [GD2-directed] CAR T-cell therapy moving toward clinical trials.
Beyond that, PARP inhibition, particularly with biomarker selecting by schlafen-11, looks promising. Much further down the road, but, eventually, I’m hopeful that [we will see] the subtyping of SCLC into transcription-based subtypes. That differential biology could lead to differential treatments.
It is an amazing year for SCLC. We come against the backdrop of the wonder of durvalumab and atezolizumab, the first frontline advance in survival we’ve had in years. Then, just in 2021, we have had 2 other advances for a disease that badly needed it.
Lurbinectedin is an alkylating agent approved in the second-line setting. It looks to be more effective and less toxic vs topotecan. I just saw a patient with a near complete response after only 2 treatments with lurbinectedin and it was well tolerated. That is a nice advance for our patients, particularly at the [FDA-]approved dose.
The other advantage [we have seen] is [regarding] QOL [with trilaciclib]. We all [advocate] the idea that QOL is at least as important as survival, but our patients would be clear that it is more important than survival; [QOL] is the main thing. Harms to physical QOL largely come from the cancer itself.
Advances like atezolizumab, durvalumab, and lurbinectedin are great at controlling disease, but the other half [of treatment] comes from what we do to help our patients with the infamous adverse effects [AEs] of chemotherapy. If we look at SCLC regimens, such as carboplatin, etoposide, plus a PD-L1 inhibitor, topotecan, lurbinectedin, and CAV, they share that their dominant toxicities are myelosuppression or the downstream sequelae of myelosuppression. Enter trilaciclib, a short-acting CDK4/6 inhibitor that provides myeloprotection. We have a nice story [with trilaciclib] from mechanism of action to numeric count preservation to prevention of AEs, to improvement in patient-reported QOL.
In terms of whether we should be [including] osimertinib in the adjuvant context, the ADAURA data are controversial. What is unambiguously clear to me is that it is good to [order] NGS early. The whole reason that studies like ADAURA are done is that the majority of our patients relapse, and [we are] in a world where NGS takes 4 or more weeks to come back. In addition to being able to consider osimertinib together with our patients with shared–decision making, it is extraordinarily valuable to have molecular testing done early. Should the patient recur, instead of waiting [at least] 4 weeks and watching our patients suffer psychologically and, sometimes, physically from progression, we know exactly what [treatment to give] up-front.
The results of IMpower010 were extremely favorable and are practice changing. There was a very large advantage in disease-free survival, much larger than we were used to seeing with chemotherapy. The only remaining questions are whether an adjuvant strategy is better than a neoadjuvant strategy and whether there should be combination [strategies] with other immunotherapeutic agents or cytotoxic chemotherapies.
CheckMate 816 [yielded] very exciting data with neoadjuvant nivolumab [Opdivo] in combination with chemotherapy. No surprise, the pathologic outcomes were better with the triplet compared with the doublet, but how very good [the outcomes were] was surprising.
When we look at data from [studies like] TRACERx [NCT01888601] showing the evolution of lung cancer over time—even in the absence of immunomodulation with drugs—it provides an argument that the earlier we use immunotherapy the better. That argument is strengthened by clinical data in the metastatic context. Unlike with targeted therapy where we have no sequential survival difference, we do seem to perhaps [see a difference] with immunotherapy. If it is going to work, it is more likely to work earlier. Seeing neoadjuvant data, in the context of chemotherapy, with that degree of favorability gives us high hopes.
Of course, we have to wait and see whether we need a chemotherapy combination. Is there another immunotherapeutic agent that would combine better? There are many outstanding studies, and it is going to take years to unravel the detailed questions to inform us of what we should do. [Taken together], the results of IMpower010 and CheckMate 816 clearly [show] that PD-1 is going to be integrated into attempts to cure with surgery, in addition to the existing indication of durvalumab in combination with chemoradiotherapy.
The question of whether [the paradigm] is going to be concurrent [immunotherapy and chemoradiation] is [one of] deep speculation. I speculate that the answer is no. Concurrent [treatment] is not going to be quite as good [as sequential treatment].
Of course, the 5-year results from the PACIFIC trial were amazing. In fact, although the PACIFIC strategy of chemoradiotherapy followed by a year of durvalumab was only studied in the context of unresectable stage III NSCLC, the community has expanded [the utility of the regimen] widely to be any context in which chemoradiotherapy is chosen. This has resulted in chemoradiotherapy being sometimes chosen over surgery in resectable stage III cases. That is a little bit past what the data showed us, but it is eminently reasonable.
With IMpower010 and CheckMate 816, it is clear that these same agents are helpful in the context of surgery. It will be interesting to see in which patient populations the shift back to surgery [occurs]. The only thing that is clear is that we are helping patients.
If we don’t find the change, we don’t help our patients. If we find the change but don’t give the drug, we don’t help our patients.
Providers, as well as guideline committees, are often hesitant to specify name brands of specific tests. With the understanding that this is not continuing medical education and I have no conflict with any of the lab companies, I will say, quite firmly, that it is important we have NGS with an RNA fusion panel. This gives us the broadest and most sensitive testing [potential]. As of right now, there are 3 companies offering this: Tempus, which is what I use in my practice, Caris, and Foundation[One]. If you order Foundation[One], you have to use the poorly named [FoundationOne] Heme panel, not the CDx panel. Otherwise, you won’t get a good RNA fusion analysis.
We need optimal sensitivity for RET, NTRK, and MET exon 14 skipping [mutations]. We are highly likely to miss these if we order al la carte polymerase chain reaction and fluorescence in situ hybridization testing, or even DNA-only NGS.
Another important point is that there is a dizzying array of newly actionable changes. There is also a dizzying array of changes that, on clinical trials, look like effective drugs and actionable [targets].
I need to phone a friend sometimes for help with [interpreting molecular testing results]. When I need help, I call up a molecular pathologist. In the community, [needing further interpretation] is quite reasonable because the reports can’t be trusted. [The reports] make everything look actionable in order to claim that more things are actioned vs competitors. If [a community provider] doesn’t know [whether a result is actionable], they need to call somebody like me with the understanding that if I don’t know, I am going to call a molecular pathologist or present [the community-based case] at a molecular tumor board. This rigor of insisting on getting broad, sensitive molecular testing and ensuring we understand what is actionable with an FDA-approved drug or clinical trial, makes a massive difference in a patient’s life.
We all know this story from EGFR-mutated NSCLC where 5-year survival is now routine without cytotoxic [chemotherapy]. The human advantages [of targeted therapy] are just massive, so we want to capture that for every [eligible] patient.
What [those datasets] make absolutely clear is that [pralsetinib and selpercatinib] are active against RET-rearranged lung cancer and that they are better than chemotherapy for patients with actionable RET changes. The data that we just saw updated speak more loudly to the intracranial activity of these agents, as well as the first-line efficacy of these agents.
In my mind, the minimal bar for true targeted therapy to be considered for first-line use was set by gefitinib [Iressa] more than a decade ago for patients with EGFR-mutated NSCLC. The response rate [with gefitinib] was about two-thirds and the progression-free survival [PFS] was about 10 months. If we look to what [pralsetinib and selpercatinib] are offering to patients with RET fusions, [response rates are] substantially superior [to those with gefitinib in patients with EGFR mutations]. The PFS in the frontline setting is more than doubled [with pralsetinib and selpercatinib vs gefitinib].
In my mind, both [pralsetinib and selpercatinib] can and should be first-line agents [for patients with RETfusion–positive NSCLC]. That is an opinion, not an absolute, but the updated data strongly support that, as well as the intracranial activity.
The consideration about longer survival is that it allows for more time for the AEs of brain radiation [to accrue]. Most famously, whole-brain radiation [is associated with AEs], but necrosis from stereotactic [radiation] manifests. If we are in a situation where a patient is going to be dead in 6 months, the only thing that matters is keeping them alive in the moment and optimizing QOL. If we need to do repeat whole-brain radiation for a patient with SCLC who is out of systemic options, we can with relative assurance that we will [provide] neurocognitive palliation over the next 1 to 2 months. Then, they will not be alive long enough to suffer the consequences [of whole-brain radiation].
Now, we are at the other end of that spectrum. Patients with changes in EGFR, ALK, or RET fusions may live 3 years, 5 years, or even longer. Anything we do to their brain in terms of radiation [confers] a high chance they will suffer [associated] AEs later. Happily, in addition to the increased survival, we are entering a world where we expect intracranial efficacy from target therapies. We have seen this with EGFR mutations and osimertinib, ALK mutations and alectinib [Alecensa] and the next-generation TKIs, and now, RET fusions and selpercatinib and pralsetinib.
For patients with all but the scariest of brain metastases, my preference in practice is to use the TKI to treat the brain metastases. Then, I carefully follow patients for neurologic signs and symptoms. Ideally, I’ll see improvement if there was a problem to start with. I get the first MRI early at about 6 weeks after a patient starts a TKI, and I almost always see a response. Then I get an MRI every time I [order] a CT scan, so at least every 3 months. The idea is that if we are going to rely on TKIs for intracranial control—because intracranial progression can be so symptomatic and cause suffering and death—we have to follow patients carefully so that when the disease progresses in the brain, we can [identify] it on an MRI.
The question of sotorasib and the appropriate line of therapy is not so easy. We don’t have that many frontline patients [who harbor KRAS G12C mutations]. If the efficacy data that we have seen [with sotorasib as] late-line treatment is all that we ever see, [sotorasib] is not a frontline drug. The [data from] very small numbers of frontline patients are approaching the numbers we saw with gefitinib to set that minimum bar years ago. Against that backdrop, particularly in a frail patient or one who doesn’t want an intravenous agent, it isn’t crazy to consider sotorasib for frontline treatment.
The complicating factor is that our bar for non-targeted therapy has evolved since that minimum bar was set by gefitinib. Patients with KRAS G12C mutations respond to immunotherapy, so single-agent pembrolizumab [Keytruda] or a triplet of chemoimmunotherapy are options for these patients. [Treatment selection] is rather controversial [in terms of] which patients with KRAS G12C mutations are going to best respond [to which strategy]. [This is particularly true in cases] of co-occurring STK11 mutations. Is that a negative predictor for immunotherapy?
In the most general sense, patients with KRAS G12C mutations tend to be good responders to immunotherapy. If I have a motivated patient with a KRAS G12C mutation, high PD-L1, no TKI contraindications, good performance status, and absent STK11 mutation, I’m going to treat them with single-agent pembrolizumab. If another provider wanted to give them a chemoimmunotherapy triplet, I wouldn’t say they were wrong. It’s not my practice because there are human advantages to giving a cytotoxic-free regimen. An immunotherapy-containing regimen or sotorasib will be my considerations for next-line therapy.
In a less-fit, relapsed patient with an autoimmune disease or a renal allograft where PD-L1 is the rejection pathway for organ transplant, it becomes more attractive to consider targeted therapy.
Most of the real-world lies somewhere in between [those clinical scenarios. Melanoma had to deal with this debate [of immunotherapy vs targeted therapy] in the context of BRAF mutations. For the majority of patients who fit in the non-extreme, we need to see more data.
Amivantamab is an option for patients with EGFR exon 20 insertions. It is another exciting new agent, both for patients with EGFR exon 20 insertions and perhaps, down the road, for [patients who develop] resistance to osimertinib.
The data [with amivantamab and lazertinib] were extremely favorable. Clinically, [the combination] looked like something that I would consider for a patient. There are unanswered questions. What is the contribution of lazertinib?
I moved to UNC-Chapel Hill from Philadelphia, Pennsylvania more than a decade ago. In the transition, I recognized, as have my peers who moved from lower-smoking areas into [heavy-smoking areas], that the patients I serve are going to have different tendencies than our previous patients. I love my smoking patients every bit as much as my former-smoking patients and never-smoking patients; stigma is a big problem in lung cancer that we need to address.
However, there are different types of lung cancer that patients with a smoking history tend to get. The demographics in the Southeast United States are relatively unique. More specifically, we see a lot of SCLC, squamous cell carcinoma, and within that, KRAS-mutated disease.
At UNC-Chapel Hill, we’ve focused our research on these patient populations. We are proud to have been part of some of the immunotherapy trials that have brought immunotherapy to the forefront for these populations. While we are proud of our work with agents such as selpercatinib that are focused predominantly on non-smoking patients, we are also proud of smoking-focused efforts, such as trilaciclib for SCLC.
Going forward clinically, we are hoping that 2 studies will open this year at UNC in these populations. One is for the combination of trilaciclib and lurbinectedin. We have two shiny new toys that are helping patients, but lurbinectedin [is associated with] hematologic toxicities. It makes sense to try to use trilaciclib to ameliorate that. It may also improve efficacy. There were some data at the 2021 World Conference on Lung Cancer that tried to answer why single-agent lurbinectedin was a winner but [lurbinectedin in] combination with doxorubicin was a loser. The best answer [the investigators] could come up with was that [the regimen didn’t include] enough lurbinectedin for many patients. If trilaciclib can help preserve lurbinectedin’s dose intensity, there may be a survival advantage.
The other [trial] in SCLC is a GD2-directed CAR T-cell product that was invented in-house. That phase 1 protocol is written and will go to [investigational new drug review] any day now. Hopefully, [that clinical trial] will open in late 2021.
In the more basic realm, we have some exciting work going on regarding squamous biology. [We] recently [published a] paper on mechanisms of metastasis. [We also have] work going on targeting other KRASvariants beyond KRAS G12C mutations.