Pellini Discusses Immunotherapies and Current Limitations of ctDNA in Perioperative NSCLC


Bruna Pellini, MD, highlights the multiple immunotherapies currently available for patients with NSCLC in the perioperative setting and elucidates ctDNA’s role in the treatment paradigm.

Bruna Pellini, MD

Bruna Pellini, MD

There are currently a variety of immunotherapeutic options available for patients with non–small cell lung cancer (NSCLC) in the perioperative setting, however minimal residual disease (MRD) testing via ctDNA assays is not yet ready to be used by clinicians to exclusively inform treatment decisions in this setting, according to Bruna Pellini, MD.

“I don’t believe ctDNA [detection of] MRD is ready for primetime; I would be cautious if you’re using this test,” Pellini said. “If you are compelled to use it, make sure you get a blood sample before you give any neoadjuvant treatment and at least 1 sample after surgery to see the ctDNA dynamics of your patient. It’s dealer’s choice when it comes to regimens in the neoadjuvant and adjuvant settings. There’s no right or wrong, so choose the regimen that you’re most comfortable with.”

Findings from the second interim analysis of the phase 3 KEYNOTE-671 trial (NCT03425643) showed that the median overall survival (OS) among patients with early-stage NSCLC who received neoadjuvant pembrolizumab (Keytruda) plus chemotherapy followed by adjuvant pembrolizumab (n = 397) was not reached (NR; 95% CI, NR-NR) compared with 52.4 months (95% CI, 45.7-NR) among those who received placebo as adjuvant therapy (HR, 0.72; 95% CI, 0.56-0.93; P = .00517). Additionally, the pathologic complete response (pCR) rates were 18.1% (95% CI, 14.5%-22.3%) vs 4.0% (95% CI, 2.3%-6.4%), respectively (P < .00001). The median event-free survival (EFS) at this second interim analysis was 47.2 months (95% CI, 32.9-NR) vs 18.3 months (95% CI, 14.8-22.1), respectively (HR, 0.59; 95% CI, 0.48-0.72).1

In another phase 3 study, CheckMate 816 (NCT02998528), investigators examined neoadjuvant nivolumab (Opdivo) plus platinum doublet chemotherapy (n = 179) vs chemotherapy alone (n = 179) in patients with resectable NSCLC. A 3-year data update revealed that the median EFS was NR (95% CI, 31.6-NR) vs 21.1 months (95% CI, 14.8-42.1), respectively (HR, 0.68; 95% CI, 0.49-0.93).2

In an interview with OncLive, Pellini, an assistant member in the Department of Thoracic Oncology at Moffitt Cancer Center and Research Institute and an assistant professor at Morsani College of Medicine at the University of South Florida in Tampa, discussed key takeaways from a recent OncLive Scientific Interchange and Workshop in February. The workshop centered around advancements in perioperative management of NSCLC, including the progress observed with immunotherapies, as well as the current and potential future roles of ctDNA.

OncLive: What is the current standing of immunotherapy in the perioperative NSCLC treatment landscape?

Pellini: We have options nowadays. We can do solely neoadjuvant chemoimmunotherapy, solely adjuvant, or both—the perioperative paradigm is that you do 4 cycles of chemoimmunotherapy upfront followed by a year of adjuvant immunotherapy monotherapy. The studies [of these regimens] were done in parallel and not compared against each other in a head-to-head fashion, so we don’t know what the best approach is. As oncologists, we are having to make our decisions based on clinical characteristics and personal choice.

The outcomes seem very similar [with] neoadjuvant chemoimmunotherapy alone vs a perioperative [approach]. However, the patient populations were distinct in each trial, so we have to do this comparison with a grain of salt and understand the differences. We don’t know what is best. I do know some colleagues who are opting for [the KEYNOTE-671 regimen] because of the OS data [seen thus far].

At the same time, the CheckMate 816 [regimen] offers an option that is shorter in terms of duration for patients, and the disease-free survival outcomes are very good and even comparable if you’re looking across trials. A lot of us still use that regimen of a neoadjuvant chemoimmunotherapy approach only followed by surgery with no adjuvant treatment. There’s no right or wrong decision, it’s a dealer’s choice.

What are the current data indicating in terms of the accuracy, sensitivity, and potential predictive vs prognostic utility of ctDNA in early NSCLC?

ctDNA is not ready for clinical practice in my opinion because the assays don’t have a good sensitivity to determine who truly has positive MRD. A lot of the assays weren’t designed for monitoring multiple time points post-surgery that we would interrogate and see if they correlate with DFS and OS. It’s an excellent prognostic biomarker, but there are too many false negatives.

To risk stratify patients, you need a better assay and we currently do not have that. A lot of better technologies or improved assays are under development. I do not use [ctDNA] in my clinical practice to make decisions because the sensitivity is not there and the technology is not there. The only time where it would be helpful is if it’s positive, then you know your patient has a high risk for disease recurrence.

I doubt that if chemoimmunotherapy and surgery did not clear ctDNA, [that] adjuvant immunotherapy for a year would. With those data we need to develop better treatment strategies in the adjuvant setting to escalate therapy for patients who have micrometastatic disease detected by ctDNA. I know some colleagues who use ctDNA and to those that use it I would caution them about the technology because I don’t believe a negative ctDNA test at 1 time point can be trusted because it could be a false negative.

What improvements need to be made to ctDNA testing to make it clinically useful? Do you think it’ll have a potential role in the future?

There’s absolutely a role, we just need an assay that has a higher rate of truly false negatives vs false positives with better sensitivity that we can trust. I believe that it has to be at least 80%-plus sensitivity for us to be able to adjust our decision making.

I’ve had conversations with other colleagues, and I do believe this will be more of a combined biomarker decision making process. We’re going to look at major pathological response, pCR, and then ctDNA. With this information combined, we can gauge what we should do and make a decision.

For us to understand the risk [status] of our patients, we’re going to need to perform a ctDNA collection before we start any neoadjuvant chemoimmunotherapy so that we know the dynamics of the ctDNA throughout neoadjuvant and surgical treatment. You want to know if ctDNA was positive to begin with and if it cleared, if it was always negative and will remain negative or if it was positive and continues to be positive. We can risk stratify patients; we have already been able to do that in the stage IV setting. This technology is already under study to escalate treatment for stage IV disease. But, the dynamics are very important and that’s what is being used in the clinical trial setting.

Time points need to be standardized, the technology needs to be improved, and we’re going to need to know the rates of patients who have positive ctDNA after chemoimmunotherapy and surgery vs negative [ctDNA]. The technology is evolving in parallel with treatment changes and it’s hard to catch up, which is the issue.

What biomarkers do you test for in the clinic and what is the role of biomarker testing pre-and post-surgery?

I do next generation sequencing [NGS] testing for all my patients with advanced disease. In early-stage disease, if we lived in a perfect world where insurances and Medicare would reimburse, I would absolutely do the testing in every patient. It’s useful information, but currently because of financial insurance reasons, I’m doing limited testing for patients [who are surgical candidates]. [I test for] EGFR, ALK, and PD-L1 [status], because this was mandated to be tested for [in] all of the clinical trials that were approved in the perioperative setting.

Also, prior to surgery we don’t have a good sample of tissue for many patients. We have situations where we don’t have enough tissue to test. We need policies to expand NGS testing so all patients can have it if they’re undergoing surgery, especially to help us decide whether to give chemoimmunotherapy prior to surgery or to give something different. It becomes attractive to think that targeted therapy for other oncogenes would be superior to chemoimmunotherapy as we have seen in patients with EGFR mutations and in patients with ALK fusions.

There are clinical trials assessing different targeted therapies in the neoadjuvant and adjuvant settings for patients with oncogene driver positive tumors. As of now, because it does not change what I do and I don’t want my patients to get a bill and have financial toxicity, I do the limited testing for resectable [disease in this setting].


  1. Spicer JD, Gao S, Liberman M, et al. Overall survival in the KEYNOTE-671 study of perioperative pembrolizumab for early-stage non-small-cell lung cancer (NSCLC). Ann Oncol. 2023;34(suppl 2):1254-1335. doi:10.1016/S0923-7534(23)04149-2
  2. Forde PM, Girard N, Spicer J, et al. Neoadjuvant nivolumab (N) + platinum-doublet chemotherapy (C) for resectable NSCLC: 3-y update from CheckMate 816. J Thorac Oncol. 2023;18(4):89-90. doi:10.1016/S1556-0864(23)00338-6
Related Videos
Hannah Walker-Mimms, MS
Hye Sook Chon, MD, gynecologic oncologist, Gynecologic Oncology Program, Moffitt Cancer Center
Roger Li, MD, of Moffitt Cancer Center
Zeynep Eroglu, MD
Francesco Di Meo, PhD
A panel of 6 experts on colorectal cancer
A panel of 6 experts on colorectal cancer
Ariel Grajales-Cruz, MD, assistant member, Department of Malignant Hematology, Multiple Myeloma Section, Moffitt Cancer Center; assistant professor, University of South Florida
Richard Kim, MD, an expert on colorectal cancer
A panel of 6 experts on colorectal cancer