Purpose/Objective(s): Immune checkpoint inhibitors (ICIs) are commonly used for treatment of advanced stage non-small cell lung cancer (NSCLC) and clinical studies are examining combinations of ICIs with radiotherapy (RT). Standard-of-care response assessment is CT imaging, which is typically performed no earlier than 2-3 months after treatment initiation due to delayed radiographic responses and occasional pseudoprogression. Improved methods of response assessment would be useful clinically and in trials combining ICIs and RT. Here we explored if CAPP-Seq ctDNA analysis might allow early identification of metastatic NSCLC patients who will benefit from immune checkpoint inhibitors. Purpose/Objective(s): We enrolled 14 patients with metastatic NSCLC (adenocarcinoma n = 10, squamous cell carcinoma n = 2, large cell carcinoma n = 2) treated with nivolumab (n = 13) or pembrolizumab (n = 1), 9 of whom had received prior RT. Nivolumab was administered every 2 weeks and pembrolizumab every 3 weeks. CT imaging and CAPP-Seq ctDNA quantitation were performed for each patient pre-treatment. Response assessment was performed by surveillance imaging every ∼2 months and CAPP-Seq ctDNA analysis at ∼6 weeks. CT scans were scored by RECIST v1.1 criteria using MINT Lesion software and blinded to the ctDNA results. Patients were classified as having durable clinical benefit (DCB; partial response or stable disease for ≥ 4 months) or no durable benefit (NDB) based on CT imaging. Results: Median follow-up was 15.6 months and median progression free survival (PFS) was 2.0 months. Four patients had DCB and 10 had NDB. Average time from treatment start to ctDNA response assessment was 41 days (range 14-71), and to imaging response was 60 days (range 37-83), which did not differ between patients with DCB vs. NDB. Median PFS in patients with DCB and NDB was 8.8 and 1.8 months, respectively. There was no significant difference in pre-treatment ctDNA allele fraction (AF) between those with DCB and NDB (average 4.8% vs. 6.0%; P=0.8). However, patients with DCB had significantly larger ctDNA drop compared to those with NDB (average post:pre fold-change +/-SEM: 0.005+/-0.005 vs. 1.3+/-0.3, P=0.001). Evaluable patients with undetectable post-treatment ctDNA all had DCB whereas no patients with NDB had undetectable post-treatment ctDNA. Another DCB patient with detectable post-treatment ctDNA displayed a 100-fold decrease in ctDNA (1% pre-treatment vs. 0.01% post-treatment). Response classification by ctDNA analysis occurred on average 19 days earlier than imaging. Evaluation of earlier time-points is ongoing. Conclusion: Our findings suggest that analysis of ctDNA potentially allows early identification of NSCLC patients who will have DCB from ICIs. Response assessment by ctDNA may therefore be useful in clinical studies examining combinations of ICIs and radiotherapy.
Merriott, D. J., Chaudhuri, A. A., Jin, M., Chabon, J. J., Newman, A., Stehr, H., … Diehn, M. (2017). Circulating Tumor DNA Quantitation for Early Response Assessment of Immune Checkpoint Inhibitors for Lung Cancer. International Journal of Radiation Oncology*Biology*Physics, 99(2), S20–S21. https://doi.org/10.1016/j.ijrobp.2017.06.061