Early ctDNA Kinetics Are Identified as Potential Biomarker Across Settings

DNA Molecule: © Design Cells - stock.adobe.com

Data from a recent analysis suggested that preliminary circulating tumor DNA (ctDNA) kinetics could serve as an early pharmacodynamic biomarker, especially for targeted therapies, offering potential for early biologic proof of concept. Supporting data were presented by Aaron Tan, MBBS, PhD, FRACP, at the European Society for Medical Oncology Targeted Anticancer Therapies Congress 2024.¹

“In 100% [of] patients with partial response prior to imaging…. [Among] 8 patients with stable disease/progressive disease but early ctDNA response, 4 had serial time points for ctDNA kinetics [and] in all 4 patients ctDNA burden subsequently increased, [which was] consistent with imaging,” Tan said in a presentation of the data. Tan is a medical oncologist at the National Cancer Centre, Singapore.

There were 7 patients in the analysis with partial response and all had a ctDNA response. Of the 12 patients with stable disease, 17% had ctDNA response and 83% had ctDNA progression. Of the 10 patients with progressive disease, 60% had ctDNA response and 40% had ctDNA progression.

The objective response in the analysis was determined by RECIST 1.1 and showed that among evaluated patients 7 (21%) had a partial response, 12 (35%) patients had stable disease, 10 (29%) had progressive disease, and 5 (15%) patients were not evaluable. There were 17 (50%) patients who received targeted therapy, 15 (44%) who received immunotherapy, and 2 (6%) patients who received both.

The phase 1 prospective, single-center, serial ctDNA analysis included patients with advanced cancer and a variation of the following tumor types: lung (32%), colorectal (15%), cervical (9%), and other (32%). In the “other” category, the cancer types represented included adrenocortical carcinoma, cholangiocarcinoma, endometrial, head and neck squamous cell carcinoma, mesothelioma, esophageal, and pancreatic. Most patients were male (62%), and the median age was 56 years (range 33-77).

At the interim analysis, spanning 2021 to 2023, investigators evaluated 34 patients and 102 plasma samples. The median number of samples collected per participant was 3 (range, 2-5). According to investigators, the tumor burden was measured using low-pass whole genome sequencing (lpWGS), which involved examining both the germline DNA (found in the buffy coat) and targeted sequencing. When compared with targeted sequencing, lpWGS had a faster turnaround and required lower plasma volumes resulting in lower cost, they explained.

There were 2 methods used to determine the limit of detection for ctDNA in the study: Fragle method and the ichorCNA method. The Fragle method is a rapid deep learning technique used to measure levels of ctDNA directly from cell-free DNA fragmentomic profiles. Unlike traditional methods, Fragle doesn’t need a tumor biopsy or a matched normal sample, making it quicker and more accessible for assessing ctDNA levels.² When using the Fragle method, prediction of ctDNA was 100% (n = 34) in all baseline samples and was consistent at high (3% or greater) and low (below 3%) ctDNA levels. With the ichorCNA method, ctDNA was detected in 53% (n = 18) of baseline samples.

The use of this new method has in recent years impacted regulatory decisions. The FDA has given its approval to several companion diagnostic tests based on ctDNA to ensure the safe and efficient utilization of targeted therapies.³ There are ongoing efforts to develop ctDNA-based assays for use alongside immuno-oncology treatments. This method is also increasingly being used for patient selection and stratification in clinical trials to improve efficiency. These findings could allow for optimal biological dosing and inform on recommended phase 2 dose decisions, according to investigators.

REFERENCES:

1. Tan AC, Guo YA, Rahman CR, et al. 850 - Low-pass whole genome sequencing (lpWGS) to quantify circulating tumor DNA (ctDNA) burden as a pharmacodynamic response biomarker in patients on early phase trials. Presented at: ESMO Targeted Anticancer Therapies Congress 2024; February 26-28, 2024; Paris, France. Accessed March 27, 2024. https://tinyurl.com/2exk69v6

2. Zhu G, Chowdhury Rafeed R, Getty V, et al. Fragle: Universal ctDNA quantification using deep learning of fragmentomic profiles. bioRxiv. Preprint posted online July 28, 2023. doi:10.1101/2023.07.28.550922

3. Vellanki PJ, Ghosh S, Pathak A, et al. Regulatory implications of ctDNA in immuno-oncology for solid tumors. J Immunother Cancer. 2023;11(2): e005344. doi:10.1136/jitc-2022-005344