Pancreatic ductal adenocarcinoma (PDAC) carries one of the worst prognoses among solid tumors. Even after curative-intent resection, 70-80% of patients recur within two years. Current post-operative surveillance relies on computed tomography (CT) imaging and the serum tumor marker CA 19-9, but both have limited sensitivity for detecting microscopic residual disease. This single-center, prospective observational study evaluates the use of the K-4CARE Lite platform-a tumor-informed plus tumor-agnostic circulating tumor DNA (ctDNA) assay with a limit of detection of 0.005%-for dynamic monitoring of minimal residual disease (MRD) in patients with resected PDAC. Thirty adult patients who have undergone R0 or R1 (margin \<1 mm) resection at Linkou Chang Gung Memorial Hospital will be enrolled over 24 months. Each participant will provide one baseline tumor tissue sample (formalin-fixed paraffin-embedded) and three serial blood samples at three pre-specified study timepoints: Timepoint 1 (Week 4 to Week 10 after surgery, before adjuvant chemotherapy); Timepoint 2 (12 weeks after Timepoint 1, approximately 3 months into adjuvant chemotherapy); and Timepoint 3 (at completion of adjuvant chemotherapy, approximately Month 6 after surgery). A fourth long-term follow-up phase (Timepoint 4) collects results from patient-funded ctDNA testing performed as part of routine care. The primary outcome is the cumulative MRD detection rate across the three pre-specified post-surgical timepoints (Timepoint 1, Timepoint 2, and Timepoint 3). Secondary outcomes include the association between ctDNA status and disease-free survival (DFS) and overall survival (OS), molecular clearance rate at Timepoint 3, lead time of molecular relapse over imaging, and platform performance. ctDNA results are reported back to the treating physician for reference but do not mandate treatment changes-all clinical decisions remain at the physician's discretion per standard guidelines. This study aims to generate the prospective evidence base needed to design future ctDNA-guided interventional trials in pancreatic cancer.
BACKGROUND Pancreatic ductal adenocarcinoma (PDAC) is one of the most lethal solid tumors. Even after curative resection, 70-80% of patients relapse within 2 years. Conventional surveillance with CT imaging and serum CA 19-9 detects recurrence only after macroscopic tumor burden has accumulated, missing the window for early intervention. Circulating tumor DNA (ctDNA) has emerged as a sensitive biomarker for minimal residual disease (MRD) in colorectal and lung cancers, but standardized prospective ctDNA monitoring data in pancreatic cancer remain limited. PLATFORM K-4CARE Lite (Gene Solutions JSC, Ho Chi Minh City, Vietnam) is a comprehensive genomic profiling platform combining: * 515-gene DNA panel for baseline tissue profiling * Up to 50 personalized tumor-informed tracking variants for liquid biopsy * 113-gene gastrointestinal tumor-agnostic panel as a parallel tumor-agnostic track * Hybrid capture sequencing at 5,000-10,000X depth on plasma, 200X on tissue * Limit of detection: 0.005% variant allele frequency (VAF) * Buffy coat / white blood cell sequencing in parallel to filter clonal hematopoiesis of indeterminate potential (CHIP) and germline variants * Reports homologous recombination deficiency (HRD), tumor mutational burden (TMB), microsatellite instability (MSI) status * First report turnaround: approximately 6 days DESIGN Single-center, prospective, non-interventional observational cohort study at Linkou Chang Gung Memorial Hospital. ctDNA results are returned to the treating physician for reference but do not mandate any treatment change; all therapeutic decisions follow standard adjuvant chemotherapy practice and physician discretion. POPULATION N = 30 adults (aged 20 years or older) with histologically confirmed PDAC who have undergone R0 or R1 (margin close \<1 mm) curative resection (pancreaticoduodenectomy, distal pancreatectomy, or total pancreatectomy), with post-operative pathologic stage at least pT1 N0 or any N1+, no radiologic evidence of residual or distant disease within 4 weeks before enrollment, and intent to receive adjuvant chemotherapy. SAMPLING TIMELINE * Screening: Week 4 to Week 12 after surgery; informed consent * Baseline: Formalin-fixed paraffin-embedded (FFPE) tumor tissue and buffy coat (drawn at Timepoint 1) * Timepoint 1 (mandatory): Week 4 to Week 10 after surgery, before first adjuvant chemotherapy dose; 20 mL whole blood * Timepoint 2 (mandatory): 12 weeks after Timepoint 1 (approximately 3 months into chemotherapy); 20 mL whole blood * Timepoint 3 (mandatory): At completion of adjuvant chemotherapy or premature termination (approximately Month 6 after surgery); 20 mL whole blood * Timepoint 4 (passive data collection only): Long-term follow-up after Timepoint 3; the study does not arrange or fund ctDNA testing in this phase. If a participant elects to undergo ctDNA testing as part of self-funded routine care, the study collects those results. PRIMARY OUTCOME Cumulative MRD detection rate across the three pre-specified post-surgical timepoints (Timepoint 1, Timepoint 2, and Timepoint 3), defined as the proportion of participants with at least one tracking mutation detected at variant allele frequency above the K-4CARE Lite platform's limit of detection at one or more timepoints. SECONDARY OUTCOMES * Disease-free survival (DFS): From surgery to first radiologic or pathologic recurrence or death from any cause * Overall survival (OS): From surgery to death from any cause * Molecular clearance rate at Timepoint 3: Proportion of participants with negative ctDNA result at Timepoint 3 (newly cleared or persistently negative) * Lead time of molecular relapse: Interval from first ctDNA conversion (negative to positive) to radiologic recurrence * VAF kinetics from Timepoint 1 through Timepoint 3 (slope: rising, falling, stable) * Platform performance: Assay success rate and turnaround time from sample receipt to report * Single-timepoint MRD detection rate at Timepoint 1 (reported separately for cross-study comparison with prior single-timepoint PDAC ctDNA literature) EXPLORATORY OUTCOMES * Correlation of baseline HRD, TMB, and MSI with ctDNA dynamics * Concordance of ctDNA monitoring with CA 19-9 trajectory (Cohen's kappa) * Frequency and types of newly emerging resistance mutations during follow-up STATISTICAL METHODS Primarily descriptive. MRD detection rate reported as percentage with 95% Wilson confidence interval. Survival analyses by Kaplan-Meier with log-rank tests, exploratory given the limited sample size. Categorical comparisons by Fisher's exact test. Continuous comparisons by Wilcoxon rank-sum test. Cox proportional hazards regression limited to 2 to 3 covariates given the sample size. Analyses in R (version 4.0 or higher) or SAS (version 9.4 or higher); two-sided p \< 0.05. Missing data not imputed. FUNDING Timepoint 1 through Timepoint 3 ctDNA testing is funded by the principal investigator's intramural Body of Medical Research Project (BMRP) grant from Chang Gung Memorial Hospital. Gene Solutions provides the testing service only and has no role in study design, data analysis, or publication decisions. Timepoint 4 self-funded testing by participants is not reimbursed by the study. SAMPLE LOGISTICS Samples are received in Taiwan at NEWCL and forwarded to Gene Solutions JSC, Ho Chi Minh City, Vietnam, for sequencing. Specimens carry a study code only; no identifiable patient information accompanies the sample.
Study Type
OBSERVATIONAL
Enrollment
30
Linkou Chang Gung Memorial Hospital
Taoyuan, Taoyuan, Taiwan
Cumulative Minimal Residual Disease (MRD) Detection Rate Across Three Post-Surgical Timepoints
The proportion of participants with circulating tumor DNA (ctDNA) detected at one or more of the three pre-specified post-surgical surveillance timepoints (Timepoint 1, Timepoint 2, or Timepoint 3). A timepoint is classified as ctDNA-positive if at least one tracking mutation is detected at variant allele frequency above the K-4CARE Lite platform's limit of detection (0.005%). Pancreatic ductal adenocarcinoma is recognized as a moderate-to-low ctDNA shedder, with reported single-timepoint post-resection detection rates of 30-60%; cumulative detection across serial timepoints is hypothesized to improve sensitivity for assay-defined minimal residual disease. Reported as percentage with 95% Wilson confidence interval, with stratified contribution by individual timepoint.
Time frame: From Timepoint 1 (Week 4 to Week 10 after surgery) through Timepoint 3 (approximately Month 6 after surgery, at completion of adjuvant chemotherapy), spanning up to 8 months per participant.
ctDNA Molecular Clearance Rate at Timepoint 3 (End of Adjuvant Chemotherapy)
Proportion of participants with negative ctDNA result at Timepoint 3 (no tracking mutation detected above the limit of detection), regardless of Timepoint 1 or Timepoint 2 status. Includes both newly cleared participants (positive at Timepoint 1, negative at Timepoint 3) and persistently negative participants.
Time frame: At Timepoint 3, defined as completion of adjuvant chemotherapy (approximately Month 6 after surgery) or earlier termination.
ctDNA Variant Allele Frequency (VAF) Kinetics During Adjuvant Treatment
For each participant, the slope of the highest variant allele frequency (VAF) of any tracking mutation is computed for up to two intervals: (a) Timepoint 1 to Timepoint 2 (intra-treatment, approximately 12 weeks apart, capturing early response to adjuvant chemotherapy); and (b) Timepoint 1 to Timepoint 3 (full-course, approximately 6-month span, capturing end-of-treatment status). Each interval is categorized as rising (2-fold or greater VAF increase), falling (2-fold or greater decrease), stable (no 2-fold or greater change between detectable values), undetectable-throughout (negative at both timepoints), or new emergence (negative at Timepoint 1, converting to positive at Timepoint 2 or Timepoint 3). Reported per-participant per-interval; aggregate distribution reported as proportions, with sub-analysis comparing Timepoint 1-to-Timepoint 2 versus Timepoint 1-to-Timepoint 3 trajectories. Participants missing one of the two later timepoints contribute only the available interval.
Time frame: From Timepoint 1 (Week 4 to Week 10 after surgery) through Timepoint 3 (approximately Month 6 after surgery, at completion of adjuvant chemotherapy), spanning up to 8 months per participant.
Disease-Free Survival (DFS)
Time from date of curative surgery to the first radiologically or pathologically confirmed disease recurrence, or death from any cause, whichever occurs first. Estimated by Kaplan-Meier method; participants without recurrence or death are censored at the date of last disease assessment. Stratified analysis comparing ctDNA-positive versus ctDNA-negative participants at Timepoint 1 by log-rank test.
Time frame: From date of surgery up to Month 36 (last enrolled participant followed for at least 12 months after Timepoint 3)
Overall Survival (OS)
Time from date of curative surgery to death from any cause. Participants alive at the data cut-off are censored at last known alive date. Estimated by Kaplan-Meier method.
Time frame: From date of surgery up to Month 36.
Lead Time of Molecular Relapse Over Imaging
Interval (in days) between the date of first ctDNA conversion from negative to positive (during the monitoring window spanning Timepoint 2 through Timepoint 4) and the date of radiologically confirmed disease recurrence. Reported as median with interquartile range among participants who experience both ctDNA conversion and imaging recurrence.
Time frame: From Timepoint 1 (Week 4 to Week 10 after surgery) to end of follow-up, up to Month 36
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.