Interpreting ctDNA dynamics in gastrointestinal stromal tumor (GIST) therapy—timing may be key

Introduction

Gastrointestinal stromal tumors (GISTs) represent the most common mesenchymal malignancy of the gastrointestinal (GI) tract, arising primarily from the stomach and small intestine (1). Most GISTs (80–90%) are driven by gain-of-function mutations in the KIT proto-oncogene (75–80%) or, less commonly, in the platelet-derived growth factor receptor alpha (PDGFRA; 10–15%) gene (1). These oncogenic mutations lead to constitutive activation of the receptor tyrosine kinases, promoting uncontrolled cell proliferation and tumor growth.

Clinical management of advanced GIST has been transformed by targeted tyrosine kinase inhibitor (TKI) therapy over the past two decades. Imatinib, as first-line therapy, has led to a remarkable improvement in patient outcomes. However, durable disease control remains challenging for most patients and acquired resistance inevitably develops in most patients, typically within 2–3 years of treatment initiation. This resistance is predominantly mediated by secondary KIT mutations that arise in the ATP-binding pocket (exons 13 and 14) or the kinase activation loop (exons 17 and 18). Current clinical management of advanced GISTs relies on the sequential use of second- and later-line TKIs, including sunitinib, regorafenib, and ripretinib, largely without real-time molecular reassessment at progression (2,3). While pragmatic, this therapeutic strategy does not fully account for molecular heterogeneity and evolutionary dynamics in advanced GIST.

Circulating tumor DNA (ctDNA) in GIST

ctDNA has emerged as an appealing tool that enables serial molecular profiling of tumors through simple blood draws. ctDNA profiling offers several distinct advantages over conventional tissue-based genotyping, especially in an advanced disease setting. It enables the detection of heterogeneous mutations from multiple metastatic sites, while tissue biopsies provide information from only a single lesion. This benefit is pertinent to advanced GIST, where prolonged TKI exposure has been shown to promote clonal evolution and subsequent emergence of spatially heterogeneous secondary resistance mutations that may be missed by single-site tissue biopsies (4,5). The minimally invasive nature of blood draws enables more frequent molecular profiling without the risks associated with tumor biopsies, especially in patients with challenging anatomical locations. This also provides an opportunity for serial ctDNA sampling, allowing for dynamic and longitudinal monitoring of mutational evolution during therapy, potentially enabling early detection of emerging resistance mechanisms.

Several studies have demonstrated the feasibility and clinical utility of next-generation sequencing-based ctDNA assays for detection of both the driver mutations as well as the secondary resistance mutations in advanced GIST (6-8). These studies have also shown good concordance with tissue-based genotyping for primary driver mutations and the ability to detect a broader spectrum of secondary mutations characteristic of polyclonal TKI resistance.

Key findings from Kelly et al. (9)

The study by Kelly and colleagues in Clinical Cancer Research provides important insights into ctDNA utility in early-line treatment for advanced GIST (9). This phase II trial evaluated the combination of imatinib plus binimetinib, a mitogen-activated protein kinase kinase 1/2 (MEK1/2) inhibitor, in patients with treatment-naïve, advanced GIST. ctDNA profiling was performed not just at baseline, but also during treatment, allowing the investigators to examine ctDNA as a dynamic biomarker shaped by treatment exposure. The study yielded several noteworthy observations regarding ctDNA detection patterns during active first-line therapy. First, baseline ctDNA detection of the primary oncogenic driver mutation was relatively low at 39% and was significantly more likely in treatment-naïve patients or those who had received less than 4 weeks of imatinib (48% versus 13%, P=0.004). This inverse correlation between treatment duration and ctDNA detectability suggests that effective TKI therapy rapidly suppresses tumor DNA shedding into circulation, consistent with what our group has shown previously (6).

When ctDNA was detectable, the authors found 100% concordance with tissue-based genotyping, supporting the analytical validity of liquid biopsy approaches. However, the low baseline detection rate, particularly in patients on imatinib, limits the clinical utility of ctDNA for treatment decision-making in a substantial fraction of patients. Importantly, serial ctDNA monitoring during treatment demonstrated that ctDNA responses often preceded radiographic responses by RECIST criteria. This temporal advantage suggests that ctDNA clearance may serve as an early pharmacodynamic indicator of treatment efficacy. Furthermore, serial ctDNA profiling detected emerging resistance mutations in KIT, highlighting a potential role for liquid biopsy in detecting clonal evolution before manifesting as clinical progression.

Considerations and limitations

The Kelly et al.’s study provides valuable insights but reveals important constraints on ctDNA utility in first-line GIST. The baseline ctDNA detection rate of only 39% in this early-line cohort is substantially lower than the 77–86% rates reported in later-line settings such as INTRIGUE and VOYAGER trials (Table 1) (3,10). This marked disparity may reflect both technical and biological factors that could limit ctDNA as a routine biomarker in treatment-naïve patients. The small sample size (n=31) and single-arm design of this phase II study also limit definitive conclusions about the prognostic or predictive value of baseline ctDNA status. Larger studies are needed to determine whether baseline ctDNA status or serial monitoring patterns correlate with meaningful clinical outcomes such as progression-free or overall survival, and to validate whether serial ctDNA monitoring can reliably guide treatment decisions in the first-line setting.

Technical platform differences

Observations made from cross-trial comparisons must be interpreted with caution for the fact that Kelly et al. employed MSK-ACCESS rather than Guardant360 CDx assay used in INTRIGUE and VOYAGER. MSK-ACCESS is a 129-gene panel with 92% sensitivity for de novo mutations at 0.5% variant allele frequency (VAF). In contrast, Guardant360 CDx is a 74-gene panel with a lower analytical limit of detection of 0.25% mutant allele fraction (4,11). These differences in panel size, sequencing depth and detection thresholds can influence sensitivity, particularly when ctDNA shedding is minimal and variant allele fractions approach detection limits. However, while platform differences may contribute to the observed disparity, they are unlikely to fully explain it. Biological factors related to disease state and treatment status could also substantially augment ctDNA detectability in early-line GIST.

Biological determinants of low baseline ctDNA detection

Several biological factors intrinsic to early-line GIST may also explain the markedly lower detection rates. For one, GIST is characterized by low tumor mutational burden, driven almost exclusively by a single activating KIT or PDGFRA mutation. In treatment-naïve disease, ctDNA detection relies predominantly on identifying the primary mutation. In contrast, later-line patients tend to harbor polyclonal resistance with multiple secondary KIT mutations accumulated through successive TKI exposures. In VOYAGER, patients averaged 2.6 KIT mutations, with some harboring up to 14 distinct resistance mutations (3). This expanded mutational landscape in later-line disease may provide more variant targets than the genomically austere first-line setting.

The mechanism of imatinib action may also influence DNA shedding, though this relationship remains incompletely understood. Evidence suggests that imatinib induces cytostasis and cellular quiescence in GIST rather than direct cytotoxicity, with treated cells entering reversible cell cycle arrest rather than consistently undergoing apoptosis (12,13). While apoptotic DNA fragmentation has been proposed as a major mechanism of ctDNA release in cancer more broadly, the extent to which this explains ctDNA dynamics in GIST specifically is unclear. If quiescent tumor cells under effective TKI control release less DNA into circulation compared to actively dying cells, this could partially explain Kelly et al.’s observation that ctDNA detection dropped from 48% in truly treatment-naïve patients to just 13% in those receiving even brief prior imatinib exposure. By contrast, in progressing TKI-resistant disease where elevated cell turnover and death may occur, ctDNA release could be enhanced, as suggested by Ko et al.’s observation of 70% detection rates in patients with active progression versus substantially lower rates in stable disease (6). However, the precise biological mechanisms linking treatment response, cell death pathways, and ctDNA shedding in GIST warrant further investigation.

Implications and future of ctDNA in GIST management

The Kelly et al. study highlights that ctDNA utility in GIST may be highly context-dependent. The findings suggest that ctDNA profiling may be most informative in later-line disease where tumor mutational complexity is greater, ctDNA shedding is more robust, and molecular stratification could meaningfully inform treatment selection among multiple TKI options. In the first-line setting, where treatment is standardized and detection rates are low, baseline ctDNA’s clinical utility appears limited. However, the ability of serial ctDNA monitoring to detect treatment responses and emerging resistance earlier than imaging suggests potential value for dynamic monitoring during therapy, particularly in patients with detectable baseline ctDNA.

Several questions warrant investigation to better define ctDNA’s role in GIST. The optimal timing of ctDNA sampling relative to treatment initiation requires clarification, as even brief TKI exposure substantially reduces detectability. The biological determinants of variable ctDNA shedding remain incompletely characterized—understanding why some patients shed detectable ctDNA while others with similar disease burdens do not could identify subpopulations where liquid biopsy is most informative. Prospective studies are needed to establish whether ctDNA-guided treatment modifications improve survival outcomes compared to standard imaging-based management.

As technology advances, newer ultra-sensitive detection platforms with lower limits of detection could improve capture rates in low-shedding early-line disease. Integration with other circulating biomarkers might provide complementary information when ctDNA alone is undetectable. One such promising avenue is the ctDNA methylome, which, unlike somatic mutations confined to limited hotspots, spans thousands of cytosine-phosphate-guanine (CpG) sites genome-wide, providing a larger signal space even when mutational burden is low (14). Moreover, methylation-specific assays have demonstrated detection at allele fractions below conventional mutation-based approaches, and genome-wide methylation profiling additionally enables tissue-of-origin deconvolution to facilitate multi-modal biomarker integration (15).

Ultimately, ctDNA profiling in GIST appears most promising as a complementary tool that provides insights into real-time tumor evolution and treatment response dynamics rather than as a replacement for tissue genotyping or radiographic monitoring. The field would benefit from prospective studies rigorously evaluating clinical outcomes associated with ctDNA-guided management strategies across different lines of therapy. Until such evidence emerges, ctDNA should be viewed as an evolving research tool with selected clinical applications, with value in later-line settings where molecular heterogeneity is greatest.

Acknowledgments

None.

Footnote

Provenance and Peer Review: This article was commissioned by the editorial office, Journal of Gastrointestinal Oncology. The article has undergone external peer review.

Peer Review File: Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2026-1-0197/prf

Funding: This work was supported by the Mr. Ho Chee Kiong through “NCCS Cancer Fund” and “The Chen Lin Trust”, Tanoto Foundation, Singapore Ministry of Health’s National Medical Research Council Research Transition Award (No. TA21jun-0005), and Clinician Scientist Individual Research Grant (No. CIRG25jan-0007).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jgo.amegroups.com/article/view/10.21037/jgo-2026-1-0197/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Blay JY, Kang YK, Nishida T, et al. Gastrointestinal stromal tumours. Nat Rev Dis Primers 2021;7:22. [Crossref] [PubMed]
2. Akahoshi K, Oya M, Koga T, et al. Current clinical management of gastrointestinal stromal tumor. World J Gastroenterol 2018;24:2806-17. [Crossref] [PubMed]
3. Serrano C, Bauer S, Gómez-Peregrina D, et al. Circulating tumor DNA analysis of the phase III VOYAGER trial: KIT mutational landscape and outcomes in patients with advanced gastrointestinal stromal tumor treated with avapritinib or regorafenib. Ann Oncol 2023;34:615-25. [Crossref] [PubMed]
4. Rose Brannon A, Jayakumaran G, Diosdado M, et al. Enhanced specificity of clinical high-sensitivity tumor mutation profiling in cell-free DNA via paired normal sequencing using MSK-ACCESS. Nat Commun 2021;12:3770. [Crossref] [PubMed]
5. Xie F, Luo S, Liu D, et al. Genomic and transcriptomic landscape of human gastrointestinal stromal tumors. Nat Commun 2024;15:9495. [Crossref] [PubMed]
6. Ko TK, Lee E, Ng CC, et al. Circulating Tumor DNA Mutations in Progressive Gastrointestinal Stromal Tumors Identify Biomarkers of Treatment Resistance and Uncover Potential Therapeutic Strategies. Front Oncol 2022;12:840843. [Crossref] [PubMed]
7. Arshad J, Roberts A, Ahmed J, et al. Utility of Circulating Tumor DNA in the Management of Patients With GI Stromal Tumor: Analysis of 243 Patients. JCO Precis Oncol 2020;4:66-73. [Crossref] [PubMed]
8. George S, von Mehren M, Fletcher JA, et al. Phase II Study of Ponatinib in Advanced Gastrointestinal Stromal Tumors: Efficacy, Safety, and Impact of Liquid Biopsy and Other Biomarkers. Clin Cancer Res 2022;28:1268-76. [Crossref] [PubMed]
9. Kelly CM, Bradic M, Saunds S, et al. ctDNA as a Molecular Biomarker in the Phase II Trial of Imatinib plus Binimetinib in Patients with Advanced Gastrointestinal Stromal Tumor. Clin Cancer Res 2025;31:4110-21. [Crossref] [PubMed]
10. Heinrich MC, Jones RL, George S, et al. Ripretinib versus sunitinib in gastrointestinal stromal tumor: ctDNA biomarker analysis of the phase 3 INTRIGUE trial. Nat Med 2024;30:498-506. [Crossref] [PubMed]
11. Lanman RB, Mortimer SA, Zill OA, et al. Analytical and Clinical Validation of a Digital Sequencing Panel for Quantitative, Highly Accurate Evaluation of Cell-Free Circulating Tumor DNA. PLoS One 2015;10:e0140712. [Crossref] [PubMed]
12. Hayashi Y, Nguyen VTT. A narrative review of imatinib-resistant gastrointestinal stromal tumors. Gastrointest Stromal Tumor 2021;4:6. [Crossref] [PubMed]
13. Liu Y, Perdreau SA, Chatterjee P, Wang L, Kuan SF, Duensing A. Imatinib mesylate induces quiescence in gastrointestinal stromal tumor cells through the CDH1-SKP2-p27Kip1 signaling axis. Cancer Res 2008;68:9015-23. [Crossref] [PubMed]
14. Pharo H, Vedeld HM, Sjurgard IV, et al. From concept to clinic: a roadmap for DNA methylation biomarkers in liquid biopsies. Oncogene 2025;44:4814-31. [Crossref] [PubMed]
15. Neefs I, Ibrahim J, Peeters M, et al. The technology landscape for detection of DNA methylation in cancer liquid biopsies. Epigenetics 2025;20:2453273. [Crossref] [PubMed]