Immunotherapy Response Monitoring in Lung Cancer - Episode 5

Recap: Monitoring With ctDNA for Immunotherapy Response in Lung Cancer

Roy S. Herbst, MD, PhD, discussed implementing immunotherapy and seeing its effectiveness in patients with lung cancer.

In a recent OncView™discussion, Roy S. Herbst, MD, PhD, chief of medical oncology at Yale Cancer Center and Smilow Cancer Hospital in New Haven, Connecticut, reviewed current strategies for determining success of immunotherapy in the treatment of non–small cell lung cancer (NSCLC).

Recent advances in diagnostic and treatment options are improving the outlook for this disease. Immune checkpoint inhibitors, for example, represented a fundamental change in the treatment of NSCLC and have led to improved patient outcomes with less toxicity than conventional chemotherapy.

In this program, Herbst discussed key considerations involved in selecting a treatment regimen for individual patients. He also described the monitoring parameters, such as circulating tumor DNA (ctDNA), which are measured to determine a patient’s response to therapy. Finally, Herbst offered his clinical expertise in determining treatment duration and when to give patients a drug holiday.

Selecting an Immune Checkpoint Inhibitor

Each year, multiple new drugs for the treatment of metastatic NSCLC are approved, with many others in clinical trials, according to Herbst. Treatment options depend on multiple factors, such as tumor size and location, involvement of pleura, tumor grade, lymphovascular invasion, and others.

Immunotherapies that are currently approved for NSCLC include nivolumab (Opdivo), pembrolizumab (Keytruda), durvalumab (Imfinzi), atezolizumab (Tecentriq), and cemiplimab (Libtayo).1 Herbst noted that almost “every patient who can get immunotherapy will get immunotherapy, because in metastatic lung cancer it really is the only way to see a long-term outcome.”

The selection of an immunotherapy depends on a patient’s PD-L1 expression, or tumor proportion score (TPS). Herbst described single-agent pembrolizumab, atezolizumab, and cemiplimab as “frontline” options in patients with a PD-L1 expression of 50% or greater. In patients whose PD-L1 expression is less than 50%, “most are using chemotherapy plus immunotherapy in that setting,” he noted. Pembrolizumab, he further explained, can be used in patients with at least 1% PD-L1 expression; however, “those data suggest that the result is probably as good as chemotherapy, but not better.”

Disease Monitoring

Once started on immunotherapy, patients must undergo routine monitoring to ensure they are responding to therapy or at least not progressing, said Herbst. A multimodal approach may be used for patient monitoring, including evaluation of radiographs and circulating tumor DNA (ctDNA) levels, as well as the patient’s quality of life, their performance status, and any reported adverse effects. Herbst said he typically monitors patients every 2 months depending on their drug regimen, adding that “you usually do your restaging [again] before you decide to give another round of therapy.”

Radiographs are one form of monitoring that may be used for determining partial response, stable disease, or progressive disease. RECIST criteria are typically used to characterize tumors. According to Herbst, patients receiving immunotherapy can be difficult to evaluate using RECIST due to the possibility of pseudoprogression evident on images. The tumor may look bigger due to the influx of immune cells, or it may grow before it shrinks, he explained. By contrast, pseudoprogression, which he estimated to occur in about 5% of patients, is not very common. Finally, images may also be complicated by inflammation from infection.

Another “more sophisticated” form of monitoring is measuring ctDNA, which is released into the bloodstream when apoptosis occurs in tumor cells, Herbst stated. This minimally invasive method allows for early diagnosis, prognosis prediction, detecting mutations, identifying minimal residual disease, and monitoring therapy response. In addition to being a real-time monitoring technique, ctDNA also has high sensitivity and specificity. “It can give you a quantitative measurement [because] you’re actually measuring the tumor,” Herbst explained. He added that data from a recent study conducted by the Friends of Cancer Research out of Washington, DC, showed that a decrease in ctDNA correlated with survival in patients with lung cancer receiving immunotherapy.

Herbst further explained how ctDNA levels can be measured throughout the course of treatment to determine a patient’s response to therapy. “If you see ctDNA going down,” he stated, “it tells you the patient is most likely responding.” It can also show when the disease is stabilized. In this case, he said the ctDNA results may be a good indicator to stop therapy if the patient is no longer seeing any tumor shrinkage, adding that taking an x-ray is “a way to validate those markers.”

Furthermore, ctDNA measurements can help differentiate between pseudoprogression and progression. “In pseudoprogression, you would not see ctDNA going up. In progression, you would,” Herbst explained. Thus, if the tumor looks bigger but the ctDNA is stable, “that’s telling you it’s probably pseudoprogression.”

Measuring ctDNA also has benefits in minimal residual disease monitoring, Herbst added. “Why treat patients with adjuvant therapy after surgery or after chemoradiation if they don’t need it?” He cited the IMpower010 study (NCT02486718), whose results showed that ctDNA can be used to predict which patients will benefit from adjuvant therapy with atezolizumab after surgery. The findings demonstrated an improvement in disease-free survival in patients with stage II to IIIA NSCLC who were treated with adjuvant atezolizumab after surgery, with more pronounced benefit in patients with PD-L1 expression greater than 1% (HR, 0.66; 95% CI, 0.50-0.88; P = .004).2

Data from 2 recent studies showed that ctDNA was predictive of disease progression and immunotherapy response. Investigators in one study evaluated the use of ctDNA in 3 clinical trials—the phase 1/2 CD-ON-MEDI4736-1108 trial (NCT01693562), phase 2 ATLANTIC trial (NCT02087423), and phase 1 D4190C00010 trial (NCT02261220)—with the immune checkpoint inhibitor durvalumab, with or without tremelimumab.3 They found that pretreatment ctDNA could be used as a prognostic biomarker, and on-treatment ctDNA dynamics appeared to predict the benefit of immunotherapy. Similarly, findings from the phase 2 INSPIRE study (NCT02644369), examining patients with solid tumors who were treated with pembrolizumab, demonstrated that a reduction in ctDNA from the baseline was associated with immunotherapy benefit.4

The FDA recently approved 2 ctDNA blood tests, which are also called liquid biopsies: Guardant360 CDx and FoundationOne Liquid CDx.5,6 These are for use in patients with solid tumors and are the first FDA-approved assays to assess for multiple cancer-related genetic changes rather than a single gene mutation. “Most of those are looking at mutational profiles, but I believe that they can also look at quantitative amounts of DNA and correlate that with outcomes,” said Herbst.

Determining Treatment Duration

Patients who have stable disease after long-term treatment with immunotherapy may be eligible for a drug holiday. In patients who have undergone long-term treatment, Herbst said, “coming every 3 weeks is tough, and these drugs do produce some fatigue and adverse effects. So why not stop? If someone has gone for a year or 2 [on therapy] and they’re stable, have no evidence of disease, and they want to take a holiday, I’ll usually let them do it.”

Herbst then pivoted to discuss the KEYNOTE-010 trial (NCT01905657), from which the 5-year outcome data were recently published in the Journal of Thoracic Oncology.7 Of the patients who underwent 2 years of treatment with pembrolizumab on the study, only 10% to 15% relapsed after a drug holiday. “Of those that do relapse, more than half of them respond again…if you give them pembrolizumab,” Herbst said. “My thought is that many of these patients could be, dare I say, cured after 2 years and if they’re not, we know we can rescue them or at least stabilize the disease by giving the immunotherapy again. So I am a proponent of stopping therapy if we can.”

In some instances, patients may have responded to therapy but still have an area that’s growing. “I’ve had some cases where I’ve actually either radiated or done surgery and then either stopped the drug or gone on with immunotherapy a little longer until things are stable,” Herbst said. He concluded that each patient needs to be treated on a case-by-case basis.

Choosing when to give patients a drug holiday is a difficult decision, as “there are very few studies,” Herbst admitted. Although he cited a study which favored treating patients longer with nivolumab therapy, Herbst said more data is needed to support these findings.8 However, conducting clinical trials to study drug holidays is difficult in this patient population and Herbst surmised that research may be done on a molecular level with ctDNA, “looking at how much tumor burden there is in a more molecular way.”

Treatment duration is patient specific, said Herbst, adding that some of his patients have stopped immunotherapy after 3 or 4 months due to toxicity. “They’ve had great responses, but they’ve also had toxicity and have had to stop pembrolizumab; [some are still] alive 5 or 6 years later.” On the other hand, he noted, “I’ve had patients I’ve treated for several years where they’ve never had a partial response, but they’ve been stable, and I [have] some people who I might have on therapy more than 3 or 4 years.

Looking Ahead

In addition to ctDNA, other options are needed to determine appropriate treatment duration. One of those includes imaging, according to Herbst, such as PET scanning. Furthermore, future technology may allow for more advanced molecular testing. “I would predict that we’ll have ways to file T cells in the immune microenvironment and actually label CD8-positive T cells and look at look at those…or label PD-L1 so we can actually look at the immune microenvironment in a patient. That might help us determine sites and areas of tumors that have not been totally eradicated.”

After about 30 years in this field, Herbst remarked: “Only in the last decade have I seen the amazing results of immunotherapy. I would predict that in the next few years we’re going to start seeing new immune combinations that can be used in these settings.”


1. NCCN. Clinical Practice Guidelines in Oncology. Non-small cell lung cancer, version 7.2021. Accessed October 29, 2021.

2. Wakelee HA, Altorki NK, Zhou C, et al. IMpower010: primary results of a phase III global study of atezolizumab versus best supportive care after adjuvant chemotherapy in resected stage IB-IIIA non-small cell lung cancer (NSCLC). J Clin Oncol. 2021;39(suppl 15):8500. doi:10.1200/JCO.2021.39.15_suppl.8500

3. Zhang Q, Luo J, Wu S, et al. Prognostic and predictive impact of circulating tumor DNA in patients with advanced cancers treated with immune checkpoint blockade. Cancer Discov. 2020;10(12):1842-1853. doi:10.1158/2159-8290.CD-20-0047

4. Bratman SV, Yang SYC, Iafolla MAJ, et al. Personalized circulating tumor DNA analysis as a predictive biomarker in solid tumor patients treated with pembrolizumab. Nature Cancer. 2020;1:873-881. doi:10.1038/s43018-020-0096-5

5. FDA approves first liquid biopsy next-generation sequencing companion diagnostic test. News release. FDA. August 7, 2020. Updated August 11, 2020. Accessed November 22, 2021.

6. Roche announces FDA approval of FoundationOne Liquid CDx, a comprehensive pan-tumour liquid biopsy test. News release. Roche. August 28, 2020. Accessed November 22, 2021.

7. Herbst RS, Garon EB, Kim DW, et al. Five year survival update from KEYNOTE-010: pembrolizumab versus docetaxel for previously treated, programmed death-ligand 1–positive advanced NSCLC. J Thorac Oncol. 2021;16(10):1718-1732. doi:10.1016/j.jtho.2021.05.001

8. Dudnik E, Moskovitz M, Agbarya A, et al. Alternative nivolumab (N) duration and scheduling in advanced non–small cell lung cancer (aNSCLC): real-life data. J Thorac Oncol. 2019;14(suppl 10):S235-S236. doi:10.1016/j.jtho.2019.08.469