Navigating Neoadjuvant vs Adjuvant Immunotherapy Approaches in NSCLC Requires Further Elucidation

Sid Devarakonda, MD

New treatment strategies in both the neoadjuvant and adjuvant settings are being implemented in the non–small cell lung cancer (NSCLC) arena, according to Sid Devarakonda, MD, who added that this creates debate among oncologists over the benefits of neoadjuvant vs adjuvant immuno-oncology (IO) approaches. Although there are no definitive clinical trials whose findings favor one over the other, Devarakonda says that emerging data suggest an advantage with neoadjuvant therapy.

“The evidence points in the direction of there being some benefit for a neoadjuvant approach with caveats. Neoadjuvant therapy is probably going to be the standard of care for the right population moving forward. [Whether or not] adjuvant immunotherapy adds any [benefit] is a good question for which we don’t have an answer,” Devarakonda explained in an interview following an OncLive^® State of the Science Summit on lung cancer, which he co-chaired.

In the interview, Devarakonda, a thoracic medical oncologist at Swedish Cancer Institute in Seattle, Washington, discussed the evolution of the treatment landscape for patients with NSCLC as IO agents are integrated into practice. He also detailed considerations regarding patient- and disease-specific factors such as mutations and tumor characteristics and highlighted the oncology-wide debate on the benefit derived from neoadjuvant treatment strategies.

OncLive: How do particular patient and disease factors affect your decision to use perioperative IO agents?

Devarakonda: This is a field where a lot of data are coming out in a relatively short time period from multiple clinical trials that have all been designed with subtle differences between them. As we’re starting to piece all [the data] together, there are some things that we know for certain and some things where we are free to interpret the data on how we would fit [specific therapeutic approaches] into the treatment paradigm.

For example, if a patient has an EGFR or ALK mutation or someone is a never smoker with a known driver alteration such as HER2, we don’t have granularity from the data for some alterations. However, at least these are known [disease characteristics predicting a] poor response to immunotherapy. In the adjuvant space, these patients are treated with a tyrosine kinase inhibitor, so the risk of toxicity is quite significant if you give them an immunotherapy agent up front. These are the patients [you would not] give neoadjuvant chemoimmunotherapy [to].

Although we don’t really know, some of these [data appear to be reflective] of the trial design itself. When you look at the number of patients who start induction [therapy] and how many [receive] an operation and an R0 resection, there is a bit of attrition. That always makes us anxious because you’re talking about a curative intent setting. You ideally would want to take every patient for a curative intent resection because systemic therapy by itself, at least as of today, does not cure patients [with NSCLC]. That does drive some level of physician anxiety and patient anxiety as well, [and it] raises the question of which subset [of patients will] benefit the most [from this therapy].

Right now, the attrition in that percentage of patients is not a very big number at approximately 10% to 15% if you look across different trials; whether some patients benefit [from] surgery is a very different question. However, there is still that emotional component of losing [the] window for surgical resection.

When taking all of this together, patients who have known oncogene drivers that do not confer sensitivity to immunotherapy [are] not the population you’re looking at. The benefit seems to be [disease] stage-driven, and we knew that from our days of [treating patients with] chemotherapy. [When there is a] higher stage of disease, the question of resectability comes in. If a surgeon thinks that a tumor and the nodal disease is resectable, then the higher the stage of disease and the higher the PD-L1 expression, the more likely you are to lean toward giving patients chemotherapy and immunotherapy.

Which data from clinical trials have shown the benefit of neoadjuvant approaches vs adjuvant approaches alone?

[Nearly all data] that we have are an extrapolation from cross-trial comparisons; we don’t have a good trial that is going to answer this question, but hopefully sometime soon we will. If we look at the data available, there is a theoretical framework that says there is some advantage to giving immunotherapy when a tumor is intact—the tumor microenvironment is important for response.

Patients who have a pathological complete response [CR] do extremely well. Their event-free survival is [north of approximately 90%] if you look at results from subset analyses in general across 5 or 6 different clinical trials. If a patient has a pathological CR, you probably don’t have to [administer] adjuvant immunotherapy. The question is, what if you don’t? Here is where you’re stuck with subset analyses; for example, [data from] the phase 3 KEYNOTE-671 trial [NCT03425643] show that the [survival] curves are a little different in the nonpathological CR population for the study and the control regimens.

That could lead one to interpret and take the stance that maybe there is some benefit in giving immunotherapy after surgery. However, is that just a question of neoadjuvant exposure—is that enough to drive that benefit? Do you need [to administer] more adjuvant therapy? Do patients need it for one year [and] does it add any more [benefit]? We don’t have an answer to [these] questions. You could take the glass-half-empty approach and say 4 cycles of chemoimmunotherapy did not [result in] a pathological CR, so why would giving more [therapy] work? Or you could take the other approach [believing that] more [therapy] is sometimes better because there’s nothing else to do [in this circumstance]. It’s a discussion that I have with my patients, and in patients who do not experience a pathological CR, I tend to lean toward giving them adjuvant immunotherapy; it’s an option on the table. That’s the way I interpret these data.

How do you determine a treatment approach when a patient wants to go right to surgery but neoadjuvant immunotherapy appears to be a good option?

The emotional component [is something to] talk about. Patients [sometimes say,] “If there is a tumor, cut it out; why are we talking about treatments up front?” But for surgery to lead to a cure or for a cure to occur, I want to emphasize 2 things. One is that despite a surgical resection––and [cure rates are] highly dependent on [disease] stage––if only half the patients are cured, we don’t want to put our patients through an operation where the cancer has a 50% chance of coming back. We should be doing everything we can to minimize [recurrence], and often it’s systemic [therapy] failure that is the problem. The answer for that [issue] is giving chemotherapy plus immunotherapy, [as these are] the best data we have, or chemotherapy followed by immunotherapy. The data currently point to the fact that doing [treatment] before [surgery] is better than doing [it] after.

The second [consideration] is that there is a chance that we might lose our window for an operation. However, when you look at adjuvant therapy [and the data survival] curves but you don’t know what happened before—how many patients entered the surgeon’s clinic and how many went through the operation without complications—you’re starting your clock at a different point. The attrition in that sense for the adjuvant [treatment of] patients is [somewhat] unknown. The risks are real that they could develop a toxicity or progress on treatment, and that’s where I tell patients that there is a possibility that we could lose our opportunity for a surgical resection.

However, the counterpoint is that it’s not enough to cut a tumor out. Sometimes, if the disease is likely to come back in 4 or 6 months if there is poor biology, the surgery is not doing anything except adding another procedure for our patients. What proportion of patients are in the poor biology category in that attrition rate, and what percentage develop a complication [where] a treatment compromises their ability to get an operation needs to be teased out a little better.