Using ctDNA Next-Generation Sequencing to Predict Response in Patients with Advanced NSCLC

During an interview with <em>Targeted Oncology</em>, Charu Agarwal, MD, reviewed the findings of a prospective study of pembrolizumab monotherapy in patients with advanced NSCLC.

Charu Aggarwal, MD

Is it feasible to use circulating tumor DNA (ctDNA) next-generation sequencing (NGS) as a predictive biomarker of response and progression-free survival (PFS)? A group at the University of Pennsylvania aimed to answer this question in a prospective study. Researchers monitored response and survival in patients with non-small cell lung cancer (NSCLC) after treatment with pembrolizumab (Keytruda) by collecting serial plasma samples.

In this study, plasma was collected at week 0, 9, and 18, at which point, researchers performed ctDNA NGS using the 730-gene panel, Guardant360. At baseline, mutations (MTs) and variant allele fraction (VAF) numbers were determined. Serial determinations were also calculated based on a change in the mean between T1 and T0 and T2 and T0.

Of the 95 plasma samples collected, MT was detected in 32 patients, with the most common type of MTs being tumor suppressor P53 (n = 21). The remaining patients showed non-synonymous MT (n = 3) or protein arginine methyltransferases (n = 9).

Results also showed that smokers had a worse prognosis with a higher amount of MTs at T0. However, the response rate was not affected by whether not a patient was a smoker.

The median follow-up was 9.26 months, accompanied by a PFS rate of 7.4 months and an overall survival (OS) of 10.5 months. Patients who had a decrease in ctDNA VAF in both T1 and T0 showed a longer PFS than other patients.

These data show that ctDNA NGS is a feasible predictive biomarker revealing response to pembrolizumab monotherapy and PFS after therapy in patients with advanced NSCLC.

During an interview withTargeted Oncology, Charu Aggarwal, MD, MPH, Assistant Professor for Lung Cancer Excellence, Abramson Cancer Center, Penn Medicine, reviewed the findings of a prospective study of pembrolizumab monotherapy in patients with advanced NSCLC.

TARGETED ONCOLOGY: What was the rationale for this prospective study?

Aggarwal: I was very pleased to be able to present these data as a poster at this year's ASCO meeting. The rationale for this study was to look at dynamic changes in ctDNA NGS over time in patients receiving immunotherapy. We have a lot of literature and data supporting the use of this technology for detection of patients with therapeutically targetable mutations. However, we don't quite know what to look for or what to follow in patients with immunotherapy. We started with a simple design, [first asking the question], is this really feasible to do this? We set off by collecting blood at baseline in patients that were starting standard of care immunotherapy.

TARGETED ONCOLOGY: Can you explain the study design and scale?

Aggarwal: We started off by assessing feasibility by collecting 2 tubes of blood in patients starting therapy with metastatic NSCLC at baseline. This was an Institutional Review Board approved protocol. We also collected blood at week 9, which coincided with the time of first staging after 3 cycles of pembrolizumab monotherapy. At an additional time point, [we collected more blood] at 18 weeks which coincided with the second re-imaging time point after 6 cycles of immunotherapy.

We looked at comparing the VAF at T-1 or the 9-week time point versus T-0 and T-2 or the 18-week time point versus T-0. We anticipated enrollment of 30 patient over a year, and we were successfully able to do that.

TARGETED ONCOLOGY: What were the study findings?

Aggarwal: We found 2 interesting things. One, we found that it was feasible and easy to enroll patients and consent them to the study. Secondly, we found that most of the patients on our study, at baseline, had detectable mutations; they were not actionable. We actually excluded patients with actionable mutations since these were the patients that were destined to get immunotherapy.

We found the difference or change in VAF from T-1 to T-0 or T-2 to T-0. If the difference was negative, it implied that the VAF was going down, and if it was positive, it implied that the VAF was going up or going in the wrong direction. We found that the negative VAF delta actually predicted for a superior PFS and an OS advantage that was statistically significant at the 9-week time point.

Additionally, patients that had a negative VAF were more likely to have partial responses as compared to patients with stable disease or progressive disease.

TARGETED ONCOLOGY: What are the key takeaways of the study?

Aggarwal: I think the [conclusions] of the study are very cool; we now know that chemoimmunotherapy can be used and should be used as the standard of care in first-line therapy for metastatic NSCLC. I will say that about 75% of the patients in our study underwent first-line pembrolizumab monotherapy as standard of care.

How would I apply these data? I would say that patients with a high delta VAF, meaning their VAF is actually going up, I know that at the 9-week time point, even if their scan is stable or equivocal and they're not really progressing, maybe those are the patients that would benefit from the addition of chemotherapy to pembrolizumab, even though they have PD-L1 greater than 50%. This biomarker gives us an additional way to amplify their therapy to fit their needs exactly. In the same vein, I think if somebody has a negative delta VAF and is not doing well on their scans, I would probably have much more comfort in saying, "Let me give you another 3 cycles, and let's just see what happens."

I think this will serve as a complimentary biomarker or a complimentary surrogate to radiographic imaging for us to better deliver precise immunotherapy.

TARGETED ONCOLOGY:Do you believe immunotherapy should be combined with standard chemotherapy?

Aggarwal: I think it should be done with standard chemotherapy. In fact, we are already doing that, and we look forward to presenting the data soon.


Aggarwal C, Thompson J, Chien A, et al. Dynamic monitoring of circulating tumor DNA next-generation gene sequencing as a predictive biomarker of response and progression-free survival after pembrolizumab monotherapy in patients with advanced NSCLC.J Clin Oncol. 2019;37(suppl 15; abstr 3040). doi: 10.1200/JCO.2019.37.15_suppl.3040.