Further Research May Delineate Immunotherapy Resistance Mechanisms in Lung Cancer

Solange Peters, MD, PhD

Defining mechanisms of resistance to immunotherapy in patients with lung cancer is the first step toward determining patients who are likely to benefit from immunotherapy and developing more effective targeted combinations for patients who are at a high risk of recurrence, according to Solange Peters, MD, PhD.

“Everything we do in lung cancer should improve the survival of patients because they still have a dismal prognosis,” Peters said in an interview with OncLive^® during the 24th Annual International Lung Cancer Congress.

In the interview, Peters discussed the successes and shortcomings of immunotherapy in patients with lung cancer; current challenges with defining resistance mechanisms to immunotherapy in this disease; and the potential applications of circulating tumor DNA (ctDNA) monitoring in the early and metastatic disease settings.

Peters is a full professor and chair of Medical Oncology and the Thoracic Malignancies Programme in the Department of Oncology at the University Hospital of Lausanne in Switzerland.

OncLive: What was the primary focus of your presentation on the mechanisms of resistance to immunotherapy in lung cancer?

Peters: I discussed the known and unknown mechanisms of resistance to immunotherapy. I summarized our current knowledge and [discussed potential] ways to overcome these mechanisms of resistance.

[There is a] lack of definition of resistance today. The resistance setting [includes] resistance [during] immunotherapy and after immunotherapy. Additionally, we have been using combinations of chemotherapy and immunotherapy, for example, in many patients, which makes the definition of resistance difficult to be sure about. We don’t know if [a patient is] resistant to immunotherapy, chemotherapy, or both.

More importantly, what does resistance mean vs refractoriness? When [is a patient] resistant? If they stop immunotherapy, what is the time elapsed [before] the sensitivity, if still present, is regained, for example? [There are many] definitions, and [resistance] should be defined by disease type, probably not across solid tumors. [This information is] still lacking.

[This is a] problem [preventing us from] making clinical trials with homogeneous definitions. We are still not there. We have been defining resistance in immunotherapy trials. We are proud to describe that 15% to 20% of patients have long-term benefits with immunotherapy. However, that means 80% [of patients] don’t [experience that benefit]. [Patients with] acquired or de novo resistance or refractoriness represent most patients.

The main question is: Why? A range of [resistance] mechanisms [have been] described, always coming from laboratory models or melanoma patients, because melanoma is a big step ahead of us in this field. We’ve been showing mechanisms that can be related to the tumor itself, or resulting from the microenvironment, an extrinsic mechanism in the host, to define why immunotherapy doesn’t work most of the time [but] works sometimes. Knowing these mechanisms, it’s important that the neoantigen definition, the [disease] presentation, and [the patient’s] own immune system are all described. Still, in lung cancer, we don’t know if these apply, because we have few patients with rebiopsies. Few patients have been studied for this specific explanation of refractoriness or resistance.

However, theoretically, we have a range of mechanisms that can explain why immunotherapy, PD-1 and PD-L1 inhibitors, mainly, do not work in [many] patients. We also have many mechanisms explaining how to proceed. Where are we with these potential options for treating patients better in the front line or, at resistance, trying to treat them [after they have received immunotherapy] for a time? We have many explanations, but still completely lack success there.

We have been running clinical trials. We are still running phase 3 trials, for example, in the second line. However, the standard of care is still docetaxel or docetaxel plus ramucirumab [Cyramza] because none of these trials have defined an option to treat immunotherapy resistance. In the front line, we have still not identified a better immunotherapy compound, a PD-1 or PD-L1 inhibitor. Some patients receive CTLA4 inhibitors, but [this is not] innovative. [We want to make patients’ immune systems stronger so they can] resist the predominant immune evasion we see. We are still not [able to apply] the nice translational datasets that we have across diseases.

[Apart from in] lung cancer, we know much about how [resistance] can happen. We should try to rebiopsy [patients] and define how this theoretical mechanism applies to lung cancer. Maybe we should try to see how we can make new compounds addressing specific mechanisms. All this is huge and just starting.

Still, immunotherapy has changed the fate of patients and prolonged survival. [Some patients] have been cured. Although we like to say [patients] can be cured of their advanced cancer, this is still only a reality for a tiny minority of patients.

How has the movement of immunotherapy into the adjuvant setting affected efforts to define resistance mechanisms?

[Frontline immunotherapy] has fostered these efforts. Now, we give immunotherapy to almost all patients receiving radical treatment [who have] a certain risk of relapse. [Patients with] large N0 disease through stage III disease after surgery will receive immunotherapy, probably for 1 year, [similarly to the regimen of] radiochemotherapy followed by durvalumab in the [phase 3] PACIFIC trial [NCT02125461].

Most of these patients, unfortunately, have a risk of relapse. If [the relapse occurs] during immunotherapy, what do we do next? They have advanced disease, so [we would just use] chemotherapy. If they relapse 3 months after the end of immunotherapy, in which setting are they resistant? Should we rechallenge immunotherapy? This is subjective. We sometimes suggest 6 months or 3 months because of the half-life of the compound, but we don’t know the mechanisms.

[Defining these resistance mechanisms would provide] a paradigm of opportunities for continuing or restarting immunotherapy, or reinforcing immunotherapy by adding a CTLA4 inhibitor, [for example]. [This needs to be] studied. The effort we have been making for late lines in advanced disease has suddenly moved to frontline therapy for relapse after radical treatment. [This may benefit] more than half of cancer patients, who [may] be cured with a local strategy.

What are the implications of retreatment with a checkpoint inhibitor or another class of agents after progression on immunotherapy?

If we [observe] a certain interval between the end of checkpoint inhibitor therapy and the rechallenge, we see benefits. [These responses are not as common as those in] the treatment-naïve patients. [However, this rechallenge] often translates to disease stabilization or a long duration of disease control, as has been shown in pooled datasets of pembrolizumab [Keytruda] rechallenge.

This is an opportunity. I would potentially, subjectively, put the bar at 3 months. If a patient has been living 3 months after stopping checkpoint inhibitors, for example, after a certain timeframe, it would probably be nice and relevant to rechallenge the immunotherapy component as monotherapy. Or [responses] may be even better with chemotherapy or a combination. We all do rechallenges, and we usually try to consider a certain interval of 3 months or 6 months after the end of [the patient’s] last [dose of] immunotherapy.

How might ctDNA testing help inform immunotherapy-related treatment decisions?

[Using ctDNA to] detect remaining minimal residual disease [(MRD) in early disease] should help us understand the patients who need more immunotherapy. Immunotherapy seems useful in most patients, but the question is: How much and how long? If we give neoadjuvant immunotherapy for 3 months or 2 and a half months, and the patient has a successful surgery with a pathological complete response, do we need to give 1 more year of immunotherapy?

[ctDNA] assays may be sufficiently sensitive to help us predict the patients who will need more immunotherapy, but we need to [conduct] these tests. Currently, they are not sensitive enough. They are prognostic; the more ctDNA a patient has, the more disease they have. However, they are not predictive. In lung cancer, contrary to in breast cancer, it’s not time for deescalating. We will need to move [forward] with these tests to understand and select the best treatments.

However, these tests can also [determine whether we should escalate treatment]. [After] 1 year of immunotherapy, should we stop? Maybe the patient will prefer to continue. These tests, if they have positive results, may help us [decide to] go beyond what we do in clinical trials.

How may ctDNA testing be used in the metastatic setting?

[In metastatic disease, ctDNA testing is] more complex because it is quantitative. Usually, in metastatic disease, we use a median variant allele frequency [VAF]. We try to [determine] the quantity of ctDNA [and see whether a patient’s VAF is] above or below [the median]. In this patient population, it’s difficult to understand [which patients are treatable]. On an individual basis, we should quantify wild-type [genes], as well as the increases or decreases of allele variants. [Determining] how much [of an] increase [is worrisome] and how much of a decline is reassuring is difficult. Would we get these data [from ctDNA testing]? We need much accuracy to use ctDNA for treatment decisions.

Clinical trials are defining evidence-based medicine. [Upon] progression [on a computed tomography (CT) scan], [patients should] receive docetaxel. Would patients have a doubling of their VAF in a ctDNA test? Would we anticipate the switch to docetaxel before the CT scan tells us [the disease] is progressing? I don’t think so.

Treatment decisions based on quantitative ctDNA assessments need consistent accumulated data and maybe prospective assessment. This is particularly true for EGFR mutations, for example, where we have some early data. Are you willing to switch to another treatment strategy just based on ctDNA, or do you want to prolong the [patient’s current] treatment when the CT scan is not worrisome? We will need to address this in the future.

We will probably define the use of MRD faster and earlier and use quantitative assessments of ctDNA in metastatic disease, which is probably also less of an unmet need for sustainability. Anticipating the next line of treatment is probably good for a reduction of clonal divergence of the tumor, where the more it grows, the more it [becomes] divergent. However, either way, this is in the metastatic setting. [When evaluating treatments with] curative intent, [if our] methods are strong, the results can be huge for patients.

What clinical trials may move the needle for patients with metastatic disease who have progressed on frontline therapies?

Many trials are in development, but there’s 1 we have our eyes on. I’m the chair of the [phase 2 ARC-7 trial (NCT04262856) with] TIGIT. We are almost all convinced that TIGIT has an immune effect. The question is: Will the magnitude be enough without a biomarker? It’s not biomarker selective. Will it be enough to define new standards and change the way we treat patients in combination with chemotherapy like in the [phase 3] SKYSCRAPER-01 trial [NCT04294810]?

[Another area of interest is with] PARP inhibitors. A series of trials are adding a PARP inhibitor to a checkpoint inhibitor in the maintenance setting to potentially make the visibility of the tumor higher. It makes sense for revealing a higher burden of neoantigens. It doesn’t make sense if you think about data from Charles Swanton, MD, PhD, showing that the only mutations that can be seen are the clonal ones, not the ones we create by default. Who knows? Science might help us. Three phase 3 trials, ZEAL-1L [NCT04475939] and the 2 KEYLINK trials, have completed accrual and will let us know whether this is a good strategy.

A large set of trials are using antibody-drug conjugates. Will we use them combined with chemotherapy, immunotherapy, or both, in the first or second lines? I’m convinced [these combinations] will come in the second line. In the [phase 3] TROPION-LUNG-01 trial [NCT04656652], datopotamab deruxtecan [(DS-1062a) elicited a survival benefit] vs docetaxel [in patients with advanced or metastatic non–small cell lung cancer]. This is the next generation of trials to come. We started with [many trials] and have reduced that number.

Regarding resistance mechanisms, 1 mechanism is VEGFR. In the lab, VEGF inhibitors have been shown to be strong in resettling and increasing the immune system’s efficacy against tumors. Additionally, many of the compounds used [in these studies] have immune manipulations. For example, sitravatinib [MGCD516] plus cabozantinib [Cabometyx] helps change the macrophage microenvironment to an immune-sensitive environment. Cabozantinib plus sitravatinib was tested as a rational, new immunotherapy at resistance, but the phase 3 trial failed. We have a bunch of trials [with immunotherapy], but when we try to show a survival benefit in patients, we fail.

Reference

Peters S. Mechanisms of resistance to immunotherapy and how to overcome them. Presented at: 24th Annual International Lung Cancer Congress; July 27-29, 2023; Huntington Beach, CA.