Research With ctDNA Continues to Pave the Way Toward Optimal Breast Cancer Management

With a half-life of minutes to hours, circulating tumor DNA (ctDNA) is in constant flux, rendering its evaluation incredibly useful in metastatic breast cancer management provided that it is applied correctly, according to Pedram Razavi, MD, PhD, who added that ongoing research continues to unfold regarding the technology’s capacity to provide a comprehensive assessment of the genomic underpinnings of the disease with sufficient sensitivity and specificity.

“Estimating the ctDNA fraction is required for accurate interpretation of the ctDNA results,” Razavi said during a presentation at the 23rd Annual International Congress on the Future of Breast Cancer® East.¹ “Incorporating ctDNA fraction assessment assessed by the same assay or orthogonal tests [is needed, and it’s] critical to interpret results in the context of ctDNA fraction and to include ctDNA fraction as an essential part of the report.”

Razavi added that despite its faster turnaround time and greater feasibility, ctDNA profiling still lacks the sensitivity and specificity required to detect certain alterations. Moreover, he added that ctDNA should not be used as a catch-all tool given the low abundance and fragmented nature of certain classes of genomic alterations including structural variants, large insertion-deletions (indels), and copy number alterations. As a result, Razavi recommended reaching for both ctDNA and tissue profiling in first-line metastatic breast cancer, and, upon progression, starting with ctDNA and reflexing to tissue in subsequent settings.

Razavi is the director of the Breast Translational Program and Breast Molecular Tumor Board, and director of Liquid Biopsy and Genomics within the MSK Global Biomarker Development Program at Memorial Sloan Kettering Cancer Center (MSK) in New York, New York.

Going Beyond Genotyping: Using ctDNA to Study Spatial Tumor Heterogeneity

The use of broad-based genomic analysis in breast cancer has helped define the germline and somatic alterations that are found in each molecularly distinct subgroup. However, less is known about the genomic makeup of metastatic tumors. As such, investigators at MSK launched a large genomic dataset of patients with metastatic cancer, from which they extracted data on patients with the largest breast cancer subset: hormone receptor (HR)–positive disease. Prospective targeted sequencing was performed on 1,918 tumors from 1,756 patients with breast cancer. Tumors and matched normal DNA were evaluated with the MSK-Integrated Mutation Profiling of Actionable Cancer Targets platform, which can identify somatic mutations, DNA copy-number alterations, and select rearrangements in up to 468 cancer-associated genes. Sequencing was performed with high coverage depth, allowing for greater sensitivity vs common broader-scale sequencing approaches for subclonal mutational events.²

Results showed that, overall, mutations in 32 genes were significantly more common in metastases compared with primary tumors (Q < .05). Additionally, a case study of a 60-year-old female with metastatic HR-positive, HER2-negative breast cancer who participated in the study was shared during the meeting. Razavi explained that several alterations were found in this patient at various metastatic sites, including TP53, CDH1, ERBB3, ERBB2, PIK3CA, and ESR1. This further indicates the extensive spatial heterogeneity in advanced tumor samples and resulting polyclonal resistance through convergent evolution.

“We had premortem samples from the autopsy from that patient and we analyzed the cell-free DNA [cfDNA]. Very interestingly we found that cfDNA can identify most of the alterations that we found in multiple tumor sites,” Razavi said. “This provides another opportunity for us to potentially design clinical trials that address the polyclonal resistance.”

Of note, the ctDNA level for each mutation was associated with the variant allele fraction (VAF) of the tumor and the number of sites harboring the mutation, according to Razavi. “Most importantly, identification and level of the mutations in the bloodstream was [highly] associated with how many of the disease sites had that mutation. So, the higher the number of sites that had it, the higher the likelihood and the higher the level of ctDNA,” Razavi said.

Findings published in Nature Medicine in 2019 further illustrated that the use of a high-intensity sequencing assay of cfDNA and matched white blood cell DNA spanning 508 genes could identify tumor-derived mutations with high sensitivity and specificity. Moreover, investigators found that 53.2% of mutations identified in 124 patients with metastatic cancer through cfDNA had clonal hematopoiesis commonalities, further illustrating the importance of a sequencing tool capable of accurate variant interpretation.³

Taking this technology one step further, investigators evaluated pretreatment and posttreatment ctDNA samples of patients who received alpelisib (Piqray) plus aromatase inhibitors in a phase 1/2 trial (NCT01870505). Results from the longitudinal analysis demonstrated that loss-of-function PTEN mutations occurred in 25% of patients with resistance and that ESR1 activating mutations, in addition to correlating with resistance, also expanded in number and allele fraction during treatment.⁴

“We can [also] potentially learn from the PADA-1 [NCT03079011] experience going after the resistant clone early and trying to address polyclonal resistance, which is a major problem and one of the reasons you cannot cure metastatic disease in breast cancer,” Razavi said.

Considerations for MRD ctDNA Assays

The next phase of development that investigators believe could take ctDNA from bench to bedside is proof of their utility in detecting clones below current thresholds, which is where minimal residual disease (MRD)–informed assays may come into play, explained Rizavi.

“Another problem is [the resistant clones] tend to [occur] even earlier than the genotyping thresholds that we have. The genotype threshold that we have now with most of the assays is around 0.05% to 0.1%, but now have multiple assays that can go below that... ctDNA changes can happen much earlier than what we can identify with those genotyping assays, [so we get] false negatives in the current paradigm using these types of assays. But we have been developing a lot of MRD assays, and these MRD assays allow us to identify ctDNA at an extremely low level,” Razavi said.

For example, findings from a study presented at the 2024 ASCO Annual Meeting demonstrated the feasibility of using an ultra-sensitive, whole-genome sequencing–based, tumor-informed ctDNA platform to identify patients at high risk of relapse. A total of 598 plasma samples were evaluated from 76 patients with early breast cancer including triple-negative (n = 23), HER2-positive (n = 33), and HR-positive disease (n = 16). Samples were collected at baseline, cycle 2 of neoadjuvant chemotherapy, post-surgery, and every 3 months during follow-up for the first year, and every 6 months thereafter for up to 5 years.⁵

Results showed a wide range of ctDNA levels, with a median value of 336 parts per million (PPM; range, 3.73-112,011). Notably, 39% of all ctDNA detections fell into the ultra-low range of less than 100 PPM (below approximately 0.01% tumor fractions). Regarding the correlation with outcomes, at a median follow-up of 76 months (range 5-113), the presence of ctDNA was associated with high risk of future relapse (P <.001) and shortened overall survival (P <.001). Notably, the median lead-time from ctDNA detection to clinical relapse was 15 months (range, 4-41), showcasing the potential utility of using the technology to monitor disease burden and intercept resistance before it can be genotyped.⁵

Show Me the Data: ctDNA Monitoring in First-Line HR-Positive Metastatic Breast Cancer

Additional data from the 2024 ASCO Annual Meeting presented by Jesus Fuentes Antras, MD, of Princess Margaret Cancer Centre, and colleagues reinforce earlier findings showing the benefit of longitudinal ctDNA evaluation. In this specific study, investigators used a tumor-informed ctDNA assay that tracked patient-specific variants identified by tumor sequencing. Eligible patients included those with HR-positive, HER2-negative metastatic breast cancer receiving standard endocrine therapy in combination with a CDK4/6 inhibitor. Plasma samples were collected at baseline, within 30 days, and approximately every 3 months with restaging scans. Archival tumor whole-genome sequencing was used to create individualized panels for ctDNA monitoring.⁶

A total of 51 patients were analyzed. Patients had a median age of 60 years (range, 38-88) and most were receiving frontline treatment (75%) as opposed to second-line therapy (20%). Most patients had visceral disease (67%) and the breakdown of CDK4/6 inhibitor use was as follows: palbociclib (Ibrance; 76%), ribociclib (Kisqali; 22%), and abemaciclib (Verzenio; 2%).

Results showed that higher estimated VAF at baseline was associated with shorter time to treatment failure (TTF; hazard ratio, 1.14; 95% CI, 1.05-1.23; P <.01). However, early increases above baseline did not significantly correlate with TTF, with 3 of 8 cases showing prolonged responses. Additional findings demonstrated that ctDNA clearance was associated with longer TTF (hazard ratio, 0.06; 95% CI, 0.01-0.45; P < .01), with a median value that was not reached vs 14.5 months for patients with and without ctDNA clearance, respectively.

“There is a role for doing ctDNA-guided vs image-guided monitoring in metastatic disease and these are the types of clinical trials that we should do. I’m not saying that we should completely abandon imaging in patients, but there is the potential for us to monitor patients with ctDNA and have ctDNA-triggered imaging in patients,” Razavi concluded.

Disclosures: Dr Razavi reported institutional grant/funding from the National Institutes of Health, Komen, Breast Cancer Alliance, Grail, Illumina, Novartis, AstraZeneca, Epic Sciences, Invitae/ArcherDx, Biotheranostics, Tempus, Inivata, Biovica, Guardant, Personalis, Myriad, Foresight; and serving as a consultant/advisory board advisor for Novartis, AstraZeneca, Pfizer, Lilly/Loxo, Prelude Therapeutics, Epic Sciences, Foundation Medicine, Inivata, Natera, Tempus, SAGA Diagnostics, Paige.ai, Guardant, and Myriad.

References

1. Razavi P, et al. Understanding ctDNA results in metastatic breast cancer exploring potential utilities in clinical practice and trial design. Presented at: 23rd Annual International Congress on the Future of Breast Cancer East; July 19-20, 2024; New York, NY.
2. Razavi P, Chang MT, Xu G, et al. The genomic landscape of endocrine-resistant advanced breast cancers. Cancer Cell. 2018;34(3):427-438.e6. doi:10.1016/j.ccell.2018.08.008
3. Razavi P, Li BT, Brown DN, et al. High-intensity sequencing reveals the sources of plasma circulating cell-free DNA variants. Nat Med. 2019;25(12):1928-1937. doi:10.1038/s41591-019-0652-7
4. Razavi P, Dickler MN, Shah PD, et al. Alterations in PTEN and ESR1 promote clinical resistance to alpelisib plus aromatase inhibitors. Nat Cancer. 2020;1(4):382-393. doi:10.1038/s43018-020-0047-1
5. Garcia-Murillas I, Cutts R, Abbott C, et al. Ultra-sensitive ctDNA mutation tracking to identify molecular residual disease and predict relapse in patients with early breast cancer. J Clin Oncol. 2024;42(suppl 16):1010. doi:10.1200/JCO.2024.42.16_suppl.1010
6. Antras JF, Elliot M, Main S, et al. Characterization of molecular response and progression in patients with metastatic HR+/HER2- breast cancer receiving endocrine therapy and CDK4/6 inhibitors using a high-sensitivity tumor-informed assay. J Clin Oncol. 2024;42(suppl 16):3052. doi:10.1200/JCO.2024.42.16_suppl.3052