Chaft Crafts Considerations for Restarting Immunotherapy in Lung Cancer

Jamie E. Chaft, MD

When determining whether immunotherapy should be restarted in patients with lung cancer, the question should be considered in terms of immune-related adverse events (IRAEs), according to Jamie E. Chaft, MD.

“Immunotherapy absolutely has a space in the curative setting; we just need to figure out for whom and where and whether that is preoperatively or postoperatively,” said Chaft. “Immunotherapy-related toxicities need to be respected, and for that same reason, that is why we should not be giving these drugs to everyone in the perioperative setting.”

The decision of when to restart immunotherapy in patients with lung cancer can be grouped into 4 categories: always able to restart, cautious to resume treatment, rarely restart, and never restart. For example, a patient with thyroiditis whose event has been quelled, who is stable on a thyroid replacement hormone and is otherwise feeling well would be a candidate for restart, according to Chaft.

In contrast, a patient with pneumonitis or any high-grade IRAE that requires corticosteroids would fall in the last category in that they should never or almost never resume treatment with immunotherapy. “In that setting, unless the patient has exhausted all other treatment options and is willing to take the risk of a recurrent IRAE, the risk outweighs the benefit,” explained Chaft.

In an interview with OncLive, Chaft, a thoracic medical oncologist at Memorial Sloan Kettering Cancer Center, provided insight into IRAEs and immunotherapy restart, considerations for immunotherapy in the pre- and post-operative period, outlined biomarker research in lung cancer, and projected her hopes for the future use of this approach in this space.

OncLive: How do you approach restarting immunotherapy in patients with lung cancer?

Chaft: [When to restart immunotherapy is] a discussion [that should be considered in] the context of IRAEs. What are they? When do they occur? How are they mediated? How do we treat them? When is it safe to restart immunotherapy?

The initial reconsideration around immune-related AEs is that they were all T-cell mediated. We have a better sense now that not all IRAEs are created equal; some are T-cell mediated while others are likely B-cell mediated, such as thyroiditis, and perhaps we have cytokine-mediated events. Each of these events can be dealt with in a different [way]. They have different severities, duration of impact on the patient, some of which are indefinite, and a different level of consideration when determining whether to restart immunotherapy.

My take-home message is that not all IRAEs are created equal; yet, all need to be respected, thoroughly evaluated, and intervened upon to prevent harm to the patient. The most common [intervention is] corticosteroids, whether that is topical or systemic. Sometimes steroid-sparing agents are used to prevent the all-feared steroid-induced AEs of osteoporosis and gastritis, among others.

When it comes to deciding when to restart immunotherapy, I think about it in 4 categories. The first is when we are always able to restart; that is a situation as simple as thyroiditis. Once it is quelled, and someone may be rendered with hypothyroid, but they are perfectly stable on a thyroid replacement hormone and otherwise feeling well and tolerating the medicine. The [second situation is when] to very cautiously resume immunotherapy, and that is in a low-grade IRAE that has improved with conservative management. For example, a dermatitis which is sufficiently controlled with a topical steroid, or an arthritis that may be managed reasonably with 5 mg of prednisone per day.

Then, there are the trickier categories of when to rarely resume immunotherapy, and that is, for example, a low-grade liver function abnormality. Even if it gets better, there is tremendous concern that if it really was immune-related that it is going to recur. A similar example is colitis. The last category is when never or almost never to resume immunotherapy and that would be pneumonitis or any high-grade IRAE requiring corticosteroids.

What should be known about immunotherapy in the pre- and post-operative period?

The pre- or post-operative setting of immunotherapy, most of which is investigational at this point. I had the pleasure of speaking on this topic over the past few years and the most exciting element of this talk is that all the studies that we’re enrolling in the United States, as well as the Cooperative group studies internationally, that are evaluating adjuvant immunotherapy, have completed accrual. Today, we have no adjuvant opportunities for our patients outside the setting of small, early-phase clinical studies; that does mean a lot of waiting. For the adjuvant setting, there is an upcoming National Clinical Trials Network trial that will evaluate chemoimmunotherapy versus sequence versus standard chemotherapy, which I am sure will enroll robustly. 

However, we are shifting our focus to the neoadjuvant space, and this will always be a challenging space because it requires a referral from the surgeon and the willingness of the patient to potentially delay a surgery. The neoadjuvant space is looking exceptionally promising with chemoimmunotherapy, as well as immunotherapy in select patients, though we yet understand who those patients are to prescribe a monotherapy.

The overarching excitement in the perioperative space today is enrollment on studies of preoperative chemotherapy with or without immunotherapy that are likely to be practice-changing over the course of the next couple of years—if the data we have seen from earlier single-arm studies hold up.

Could you discuss the emergence of immunotherapy over the past few years and its move toward the frontline setting in stage IV disease?

Immunotherapy came in with a vengeance, and I was one of the skeptics. I wasn’t sure whether this approach would have a place in lung cancer; however, it rapidly moved from the second-line setting into the frontline in a subset of biomarker enriched patients with high PD-L1 expression. Now, we see across the board, studies showing the combination of chemoimmunotherapy are better than chemotherapy alone; this is true for small cell lung cancer and NSCLC. That being said, I am still a bit underwhelmed with the magnitude of advantage of the combination in an unselected patient population. It is very unusual for our patients to not receive an immunotherapy in the frontline. Perhaps only those with an actionable oncogene driver mutation that is known at the time of therapy will not receive immunotherapy.

The question today involves whether to use single-agent versus chemoimmunotherapy. We have come this far really quickly, but we have a long way to go in terms of understanding who needs the combination. Is there an option to sequence drugs? Who really are those durable responders to monotherapy? We have done a lot by selecting by PD-L1 expression, but we can probably do better, and we definitely owe it to our patients to invest in this biomarker research.

How do you combat resistance to immunotherapy in lung cancer?

We do not know. Resistance is a complex situation. In fact, defining resistance is a complex situation between hyperprogression theories and pseudoprogression, which is rare but real. We are trying to come up with a uniform definition of resistance and refractory; those are things that we did a decade ago in oncogene-driven lung cancer, but the same rules do not necessarily apply for immunotherapy.

Other unique caveats are oligoprogression and things like that. However, once we have a patient who is responsive to immunotherapy and then develops resistance, figuring out how to overcome that resistance is a huge unmet need. Multiple clinical trials are trying to address that question. For the primary refractory space, [this situation] is even more complex. The question of whether we will be able to find an immune-directed approach to salvage those patients remains unanswered.

Beyond PD-L1 expression, are there any other predictive biomarkers that are being investigated in this patient population?

Certainly, rare mutations detected through next-generation sequencing (NGS) have shown correlation to response in advanced disease. However, probably, the leading non–PD-L1 biomarker is tumor mutational burden (TMB). You have your supporters and skeptics of TMB in the clinic. I would say that it has a place, but the question is how to test and define it; a lot of that will need to be centrally validated.

However, for our patients who do not otherwise have high PD-L1 expression but have a high TMB, are they going to be immunotherapy responders? I am going to bet yes, but we are going to need prospective studies specifically in that patient population using monotherapy, given the toxicity concerns around ipilimumab (Yervoy) and nivolumab (Opdivo). However, now in the National Comprehensive Cancer Network guidelines, that, too, may have a place. The short answer is that TMB is the next leading best biomarker, but we certainly need better biomarkers; none of these are at the level of our oncogene predictive markers.

What does the future of immunotherapy in lung cancer look like?

I hope the future of immunotherapy in lung cancer is more refined than it is presently. We have done a good job of refining first-line monotherapy in terms of high PD-L1 expression and saving those patients the toxicity of chemotherapy up front. However, I have fears that the approval of a chemoimmunotherapy combination regardless of predictive biomarkers is just simpler and that there are many patients being exposed to chemotherapy today who perhaps do not need it in the frontline setting.

What does the future look like? I hope it looks more refined: that everyone is getting NGS and patients with an oncogene and PD-L1 expression of 90% are getting an oncogene-selected therapy and not immunotherapy. We need to have commitment to studies such as those ongoing throughout the Cooperative groups that ask and answer the questions regarding sequencing, because it is hard to justify combining therapy in patients with PD-L1–negative tumors based on the available data today, when the appropriate questions of sequencing have not been answered.