From clinical insight to trial design: advancing circulating tumor DNA-informed care in pancreatic cancer

Introduction

Background

Accurate measurement of disease burden is critical for cancer management. While historically based on clinical signs and imaging, circulating tumor DNA (ctDNA) is emerging as a powerful blood-based biomarker for prognosis, determining response to therapy (1,2), and detecting minimal residual disease (MRD) (3). The level of evidence in support of ctDNA varies by disease. This article examines the strengths and challenges of using ctDNA to inform management of pancreatic cancer, a devastating disease with a 13% 5-year survival rate (4).

Rationale and knowledge gap

Current measures of pancreatic cancer detection and disease response are inadequate: CA 19-9 is non-informative in at least 5–10% of patients (5), desmoplastic stroma in pancreatic cancer attenuates the ability to detect radiographic changes in disease in responders, microscopic seeding of localized disease is common and is invisible to imaging, and the imaging cadence of every two to three months recommended per standard guidelines is too slow, as patients with this aggressive disease can decompensate quickly from progressive disease. In view of these limitations and seeking more data to inform patient decision-making, our pancreatic cancer clinic was an early adopter of ctDNA technology, which we began testing for many patients along with CA 19-9 and carcinoembryonic antigen (CEA) tumor markers, at approximately monthly intervals, in a variety of settings. The ctDNA platform used was the commercially available Natera Signatera assay, a tumor-informed approach, which reports the ctDNA level as measured in mean tumor molecules per mL (MTM/mL) if detected (positive; ctDNA+) or 0.0 if not detected (negative; ctDNA−) (6). In this work, ‘ctDNA clearance’ denotes a negative test result following a positive result in a patient undergoing treatment. We present anecdotal cases showcasing ctDNA’s role in decision-making across clinical settings often encountered in the clinic. Within each setting (Figure 1), we outline the standard approach of disease monitoring, present clinical cases describing ctDNA’s integration, and illustrate clinical trial designs that may help demonstrate the impact of ctDNA in that setting (Figure 2).

Figure 1 Pancreas cancer treatment settings. Each circle references a case category presented in this work.

Figure 2 ctDNA trial schemata. ctDNA, circulating tumor DNA; POD, progression of disease.

Objective

This manuscript aims to examine the strengths, limitations, and potential clinical applications of ctDNA in pancreatic cancer management. Through illustrative cases from a high-volume pancreatic cancer clinic, we highlight how ctDNA may inform disease assessment across multiple treatment settings and discuss ongoing and proposed trial designs incorporating ctDNA-guided strategies.

Use of ctDNA by disease setting

Neoadjuvant

Standard monitoring approach

The use of neoadjuvant systemic therapy is increasing in patients with stage I–II pancreatic ductal adenocarcinoma (PDAC) (7). National Comprehensive Cancer Network (NCCN) guidelines (8) recommend imaging [abdominal computed tomography (CT) scan or magnetic resonance imaging (MRI), chest/pelvis CT, and consider positron emission tomography (PET)] after initiation of neoadjuvant chemotherapy to restage and assess response to systemic therapy. At many institutions, including our own, the neoadjuvant therapy period is generally four months in duration, with repeat imaging performed at two and four months. We hypothesize that ctDNA may be used to help medical and surgical oncologists as they decide whether to continue neoadjuvant treatment, switch to alternate systemic therapies, or proceed to surgery.

Clinical vignette: ctDNA response informs neoadjuvant therapy and surgical management

A 59-year-old woman presented with hematuria, and on CT evaluation for renal calculi was found incidentally to have a pancreatic body mass, which was biopsied and showed PDAC. During the initial multidisciplinary review, her tumor was considered locally advanced and unresectable. A baseline ctDNA test was positive, 0.07 MTM/mL. Following three cycles (6 weeks) of neoadjuvant FOLFIRINOX, repeat ctDNA showed clearance, while repeat imaging showed stable disease, though haziness around her superior mesenteric artery (SMA) and celiac artery was indeterminate for tumor involvement versus fibrotic treatment changes. At tumor board evaluation of her restaging scans, ctDNA clearance was discussed as potentially a good indicator of molecular response and accordingly, of good clinical response while the haziness around the SMA was interpreted as more likely treatment-related change. She proceeded with an additional five cycles of FOLFIRINOX with continued ctDNA clearance and stable scans. Given the stability of the disease, favorable molecular response, and the patient’s young age, she underwent surgical exploration and ultimately a partial hepatectomy, distal pancreatectomy, and splenectomy.

Toward clinical integration of ctDNA

In the setting of ambiguous perivascular findings, ctDNA clearance was interpreted as a reassuring sign that the therapy was working, and in this case neoadjuvant FOLFIRINOX was continued. To better inform clinical decision making in the neoadjuvant setting, additional pancreatic cancer-specific data evaluating the association of ctDNA clearance after two- and four-months of neoadjuvant therapy with pathologic response, disease-free survival (DFS) and overall survival (OS), would add valuable insight into whether ctDNA has utility in the neoadjuvant setting. Because current imaging techniques may fail to detect micrometastatic disease, ctDNA may play an important role in guiding whether to extend or switch neoadjuvant therapy and if and when to proceed with surgery. While decisions regarding resectability remain nuanced and require multidisciplinary collaboration, ctDNA may provide more confidence about when to time a surgery.

To evaluate the impact of ctDNA-guided decision-making in the neoadjuvant setting, a prospective trial evaluating ‘switch’ therapy could be conducted. The study would enroll patients with stable imaging but unfavorable ctDNA response and randomize them to continuation of therapy vs. switch to a different neoadjuvant therapy (Figure 2). Using ctDNA clearance as an endpoint for such studies can target large effect sizes and requires fewer patients with faster read-out than a DFS endpoint. Lastly, lack of ctDNA response during neoadjuvant therapy may help identify patients who are less likely to benefit from surgery.

Adjuvant

Standard monitoring approach

Adjuvant therapy is a treatment standard for upfront resectable pancreatic cancer to reduce the risk of recurrence. Yet, due to poor recovery, postoperative complications, early disease progression after resection, and patient preference, many patients do not receive or complete adjuvant chemotherapy (9). In the adjuvant setting, in which the macroscopic cancer has been evaluated and is absent, the current standard of evaluating for radiographic changes is highly limited, because there is no cancer to measure, and further, post-operative inflammatory changes confound interpretation of the resection bed for signs of local recurrence. In contrast, Signatera ctDNA has demonstrated substantial prognostic impact in the adjuvant and surveillance period in pancreatic cancer with ctDNA positivity, proving to be a clear and strong predictor of recurrence (10). For patients who have substantial residual disease at surgery despite neoadjuvant therapy, there remains equipoise whether to continue the previous neoadjuvant regimen. ctDNA could guide peri-operative treatment by indicating whether to continue the same adjuvant regimen or switch to an alternative in the setting of unfavorable surgical pathologic criteria or persistent ctDNA following surgery.

Clinical vignette: ctDNA positivity after resection guides therapy initiation and leads to subsequent ctDNA clearance

An 82-year-old man underwent chest CT to evaluate a cough. Imaging showed pericardial and bilateral pleural effusions (with no mediastinal mass or lymphadenopathy), and a pancreatic lesion in the uncinate process was also incidentally noted. The pancreas lesion was biopsied, and localized PDAC was diagnosed. A few weeks later, he underwent Whipple surgery with R1 disease (invasive carcinoma was present at uncinate margin); pathology confirmed stage pT2N0 invasive adenocarcinoma, arising from intraductal papillary mucinous neoplasm (IPMN) with high-grade dysplasia, lymphovascular, and perineural invasion. In the post-operative period, he had two readmissions for complications, including anastomotic leak and pleural effusion, thus surveillance was considered as an alternative to adjuvant chemotherapy. Given the presence of both favorable features such as its IPMN origin and node negativity (11), as well as less favorable features, such as his age, R1 resection, and comorbidities, observation was initially favored at the tumor board. However, a few weeks later, post-operative ctDNA resulted and was positive, 0.03 MTM/mL. We discussed with the patient that this was a concerning finding of persistent disease, and while we lack data to prove that initiation of adjuvant treatment would be beneficial based on a single positive ctDNA test, we were nevertheless inclined to offer adjuvant treatment on the basis of his less favorable clinical features and confirmatory postoperative positive ctDNA, reinforcing that he is likely to relapse without treatment. Repeat CT showed persistent pericardial and pleural effusions (which were negative for malignant cells on cytology), but no overt evidence of tumor, and his ctDNA cleared a few days prior to starting adjuvant chemoradiation with capecitabine on days of radiation (15 fractions over three weeks). After completing radiation, he transitioned to capecitabine alone. At one year after his surgery, his ctDNA remains undetectable and imaging shows no clear evidence of disease recurrence. He has subsequently been observed off therapy and ctDNA and imaging surveillance continue at quarterly intervals.

Toward clinical integration of ctDNA

For a patient considered at lower risk of cancer recurrence, ctDNA positivity is a concerning sign. Tumor-informed ctDNA tests have high positive predictive value for pancreatic cancer recurrence; one study showed that ctDNA-positivity in the post-operative or surveillance setting was associated with shorter DFS and was the most significant indicator of recurrence (not CA 19-9) (10). In colon cancer, the GALAXY study demonstrated that patients who were ctDNA-positive appeared to derive more benefit from chemotherapy than ctDNA-negative patients (1). In pancreatic cancer, the AGITG DYNAMIC-Pancreas trial (12) enrolled 102 patients with early-stage pancreatic cancer who received upfront resection, and used a ctDNA-guided approach to determine duration of adjuvant chemotherapy (six months vs. option to deescalate to three months at clinician’s discretion in ctDNA-positive vs. ctDNA-negative patients, respectively). Of the 53% of patients who were ctDNA-negative post-operatively, 44% were de-escalated to receive three months of adjuvant chemotherapy. Median recurrence free survival (RFS) in ctDNA-positive patients was 13 months compared to 22 months in ctDNA-negative patients. This trial thus showed not only that ctDNA positivity was associated with earlier recurrence, but also that a ctDNA-informed approach to adjuvant chemotherapy was feasible. Future trials will further elucidate how ctDNA trends throughout the course of therapy reflect underlying tumor biology and how to use them to navigate the type and timing of different therapeutic options, particularly in the perioperative setting. An adjuvant trial for patients whose ctDNA is not clearing, with patients randomized to continuation of current therapy vs. switching to alternative therapy, would help inform how to act on unfavorable ctDNA characteristics during the adjuvant setting. Alternatively, randomizing ctDNA-negative patients in the adjuvant setting to standard of care adjuvant chemotherapy vs. dose-reduced regimens may help guide how to act on favorable ctDNA characteristics to improve quality of life without compromising clinical outcomes.

Surveillance

Standard monitoring approach

Surveillance, defined as longitudinal monitoring for detection of recurrence, is standard management for patients with pancreatic cancer who have completed definitive therapy for localized disease. Current NCCN guidelines recommend surveillance, at unspecified intervals, via history and exam, CA 19-9 measurement, and imaging. Of note, data from the National Cancer Institute’s Surveillance, Epidemiology, and End Results Program (SEER) have not shown any significant survival benefit for patients who received regular surveillance CT scans (13), nor have any data shown that early treatment of recurrences due to increased CA 19-9 levels leads to better patient outcomes. Patients are monitored to assess their symptoms, with periodic CA 19-9 measurement and imaging, generally with CT scans, commonly every three months for the first two years after surgery and every six months for an additional three years. CtDNA is increasingly used to measure MRD in this setting, influenced by the large volume of evidence of ctDNA as a prognostic biomarker in colon cancer [e.g., the GALAXY study (1)]. Understanding the evidence that it holds prognostic value in the surveillance setting in PDAC (10), we routinely incorporate ctDNA monitoring to help inform the detection of recurrence. If ctDNA evidence is found without CT evidence of recurrence, we offer the patient a PET/CT scan or an earlier follow-up CT scan. We base this discussion on the potential to offer earlier detection of recurrence, while noting that data are yet insufficient to conclude that earlier treatment of recurrence based on this approach improves long-term outcomes compared to an approach lacking ctDNA data. Further, we are reassured with recent data suggesting that the overwhelming majority of patients value ctDNA testing and are less anxious when ctDNA testing is negative (14,15).

Clinical vignette: ctDNA-positivity expedited discovery and treatment of recurrent disease

A 71-year-old woman presented to her primary care doctor with painless jaundice and fatigue, and on CT was found to have a pancreatic head mass with no evidence of metastatic disease. Baseline ctDNA was positive, 0.13 MTM/mL (Figure 3). She underwent Whipple surgery with R1 resection (tumor was present at less than 1 mm from uncinate margin), and pathologic analysis showed pT2N2 moderately to poorly differentiated PDAC with focal squamous features, as well as adenocarcinoma in 6 of 40 lymph nodes. Germline testing identified a BRCA1 mutation. Post-surgery scans showed no evidence of local recurrence or new metastatic lesions, and she was started on adjuvant FOLFIRINOX. Due to diarrhea, weakness, and fatigue, irinotecan was dropped after one cycle, and she was continued on FOLFOX for four cycles. Repeat imaging showed no evidence of disease and repeat ctDNA showed clearance. She received another five cycles of FOLFOX, followed by imaging showing no evidence of disease and another negative ctDNA test. Given her BRCA mutation, she was offered enrollment into the APOLLO trial, a randomized phase two trial evaluating extended adjuvant therapy with olaparib vs. placebo. She declined, favoring observation. Her next CT chest/abdomen/pelvis showed mild haziness/soft tissue along the superior mesenteric artery, possibly postsurgical changes, but with no convincing evidence of disease. During ongoing surveillance, the patient felt gradual clinical improvement off chemotherapy. However, ctDNA within a month of her CT scan was positive at 0.31 MTM/mL, prompting further evaluation with PET/CT, which showed an FDG-avid lymph node in the root of her small bowel mesentery suspicious for metastatic disease. She remained asymptomatic; however, she agreed to undergo biopsy of her FDG-avid lymph node, which initially was indeterminate, however on repeat biopsy confirmed metastatic adenocarcinoma. She then agreed to starting GnP, which she has tolerated well, and subsequent scans after two cycles showed stability in her small bowel mesenteric lymphadenopathy that was suspicious for locally recurrent carcinoma. Repeat imaging after six cycles of GnP again showed stable disease, however due to slowly uptrending ctDNA levels, initiation of alternative therapy vs. clinical trials are being discussed.

Figure 3 Surveillance timeline with biomarker and radiographic correlates. CT CAP, computed tomography chest, abdomen, pelvis; LN, lymph node; MTM, mean tumor molecules; NED, no evidence of disease.

Toward clinical integration of ctDNA

This case reflects a patient with high-risk R1 disease, found to have ctDNA positivity in the MRD setting, which prompted PET imaging, histologic confirmation of disease recurrence, and initiation of chemotherapy. Oncologists routinely offer treatment to patients with asymptomatic disease based on the premise that early treatment improves outcomes. Whether this practice improves OS of patients, and how it affects long term quality of life, are unclear. A phase three trial in ovarian cancer demonstrated no improvement in OS with early initiation of treatment for MRD+ patients based on CA-125 (16), and a more recent study of patients with stage II-III colorectal cancer found that combined CEA and CT surveillance did not provide any additional 5-year OS benefit over CT surveillance alone after curative surgery (17). On the other hand, early initiation of pembrolizumab for MRD+ MSI-H solid tumors appears to improve RFS but it is not clear whether this improves OS (18). Increases in health care utilization do not always translate into improved healthcare outcomes, even in the case of PDAC surveillance, where routine imaging in patients who undergo resection does not necessarily lead to any significant survival benefit (13). This highlights the importance of refining our current surveillance tools, especially less invasive options such as biomarkers like ctDNA, which may detect micrometastatic lesions better than imaging (19), and how these can be used alongside CA 19-9 and imaging to actually improve outcomes (20).

Further trials are needed to prove the clinical utility and applicability of ctDNA, and to standardize ctDNA collection protocols in the surveillance setting. The FDA’s November 2024 Guidance for Industry outlines recommendations for incorporating ctDNA into clinical trial design for early-stage, curative-intent cancer settings when imaging may fail to capture subtle or microscopic changes in tumor biology, similar to the surveillance setting where imaging may lag behind blood-based detection methods (21). The emergence of more effective therapeutic options and improved ctDNA assays offering earlier detection of MRD are promising advances that may demonstrate the utility of early initiation of treatment. Study designs evaluating early initiation of treatment in the MRD setting could use ctDNA clearance as a surrogate endpoint to compare different therapeutic approaches, which then could be confirmed using DFS and OS endpoints. The MRD setting is an excellent opportunity to evaluate immunomodulatory approaches such as vaccine therapy, as micrometastatic disease may be more accessible to the immune system, and such treatments may be more tolerable than cytotoxic chemotherapy.

It remains uncertain in pancreatic cancer whether treatment of patients at the time of ctDNA-detected disease (MRD) reduces (or delays) recurrence or has a beneficial impact on OS compared to the current standard of initiating treatment at time of radiographic progression. Though at present ctDNA functions primarily as a supportive biomarker, with strong data supporting its role as a prognostic biomarker in the MRD window (10), future prospective studies may better clarify its role as a potential decision-defining tool in PDAC surveillance. For example, a prospective trial comparing early vs. late initiation of treatment based on ctDNA positivity using an OS endpoint could help answer this question conclusively.

Recurrent/metastatic disease

Standard monitoring approach

Given the limited chemotherapy options in pancreatic cancer, recurrent disease often entails resumption of systemic therapy. In these patients, progression of disease (POD), particularly after surgery and systemic therapy, often presents with metastatic disease discovered via symptoms, imaging, and/or labs (22-24). For local recurrence in the pancreas after resection, NCCN guidelines recommend surgical consultation and multidisciplinary review. For recurrence to the broader pancreatic operative bed, several options exist, including clinical trials, systemic therapy with or without chemoradiation or stereotactic body radiotherapy (SBRT), SBRT alone, or best supportive care. For metastatic disease with or without local recurrence, various systemic therapies are appropriate, including repeating the systemic therapy previously used if greater than six months have elapsed since completion of primary therapy. NCCN guidelines recommend serial imaging as indicated to assess disease response. Clinical trials and best supportive care should also be considered.

Clinical vignettes

Response to targeted therapy of metastatic disease

A 71-year-old woman developed ear pain and mandibular numbness, which led to a mandibular biopsy showing metastatic adenocarcinoma suggestive of upper gastrointestinal (GI) primary, possibly of pancreaticobiliary origin. Further workup revealed a pancreatic tail mass with hepatic metastases. Next Generation Sequencing (NGS) testing revealed a BRAF V487 mutation. She was started on GnP and received radiation to the jaw. After eight cycles of therapy, she had POD, at which time use of BRAF/MEK inhibition was considered but not pursued due to limited data and insurance denial. She was switched to nanoliposomal-irinotecan (nal-IRI) in combination with 5-fluorouracil (5-FU)/leucovorin; however, due to GI toxicity after two cycles, she was switched to CAPOX. After six cycles of CAPOX, she again progressed, at which time a ctDNA test showed low positivity. At this point, she was able to start BRAF/MEK therapy; ctDNA continued to be monitored, given the limited data for BRAF/MEK inhibition use. She cleared her ctDNA within approximately two months, supporting ongoing use of BRAF/MEK inhibition. Her CA 19-9 also trended down, and scans showed stable disease. Roughly two months after her ctDNA had first cleared, a repeat ctDNA showed a mild increase; CA 19-9 had always remained positive but now showed a modest increase of ~17%. A month later, repeat imaging showed stable disease except for a new liver lesion; she underwent wedge resection of this lesion, which returned positive for metastatic adenocarcinoma. This example clearly demonstrates how ctDNA was a sensitive gauge that provided more convincing evidence to support when to continue BRAF/MEK inhibition versus when to further investigate for suspected recurrence.

Differentiation of tumor progression from pseudoprogression

A 59-year-old man with diabetes, prostate cancer, and a BRCA2 germline mutation developed worsening back pain. Imaging revealed an infiltrative tumor in the pancreatic body/tail as well as several liver lesions. Liver biopsy showed moderately differentiated adenocarcinoma compatible with pancreaticobiliary/upper GI tract origin (with intact nuclear expression of MMR proteins). He was started on FOLFIRINOX and had a good radiographic response after four cycles of treatment, which was confirmed by a dramatic reduction in ctDNA from 158 to 3 MTM/mL, with a further decrease down to 1 MTM/mL after cycle six. He completed a total of 12 cycles of FOLFIRINOX, followed by restaging scans, which showed disease progression, and was then started on a clinical trial (TAPUR) with ipilimumab/nivolumab. A baseline ctDNA test prior to starting immunotherapy was not performed. After two cycles of immunotherapy, reimaging showed an increasing size of liver lesions with transaminitis, which was originally interpreted as possible disease progression. However, the patient was adamant that he had experienced an improvement in pain and overall symptom burden while on immunotherapy, calling into question whether the imaging findings were indicative of pseudoprogression. Though a pre-immunotherapy ctDNA baseline was not assessed, ctDNA at the time of imaging and three weeks later showed a reduction of ctDNA from 1,218 to 396 MTM/mL, which was interpreted as a possible indication of disease response. Furthermore, biopsy of non-tumor liver parenchyma showed signs of sinusoidal obstruction syndrome, consistent with a drug/toxin-induced liver injury, suggesting the ctDNA decline and clinical improvement were signs of disease response, and the radiographic findings were more consistent with pseudoprogression than POD. However, given immunotherapy was held due to a presumed immune-mediated adverse event, the patient’s ctDNA ultimately began to rise, so he came off study and switched to next-line chemotherapy.

Toward clinical integration of ctDNA

This section contains two cases of metastatic disease, one treated with targeted therapy and one case with possible pseudoprogression, in which their clinical management was informed by ctDNA testing. Cytotoxic chemotherapy in pancreatic cancer has modest efficacy with response rates in the roughly 30% range (25) and limited duration, given five-year relative survival with distant disease is only 3% (4). Treatment in the metastatic setting is for palliative intent, and while combination cytotoxic chemotherapy is the mainstay of treatment and improves survival, it is also associated with many toxicities. Determining quickly whether a chemotherapy regimen is working is crucial to ensure that patients are only being exposed to such highly toxic regimens if they are working. Standard assessment of disease response is conducted using CT scans at two months, however earlier assessment of disease response using tumor markers may have clinical utility. For example, deep responses to chemotherapy may permit earlier de-escalation of toxic chemotherapy, whereas early identification of regimen futility may enable discontinuation of ineffective therapy in favor of a better alternative prior to clinical decompensation. If validated data confirm that early ctDNA kinetics (e.g., after one month of treatment) are predictive of key efficacy outcomes such as response rate and progression-free survival, these strategies merit formal evaluation in prospective clinical trials.

In the case of radiographically enlarging lesions, ctDNA was used to help distinguish POD vs. pseudoprogression. Understanding whether a selected treatment is working, especially given the highly toxic nature of the treatments in the PDAC arsenal, is crucial. Pseudoprogression can be a concern in the management of patients treated with immunotherapy, in which the treatment induces an influx of immune cells and inflammation, causing a tumor to appear larger. In the aforementioned pancreatic cancer case, immunotherapy was implemented as part of a clinical trial. Generally considered an immunologically cold tumor, and outside the context of the rare case of MSI-H pancreatic cancer, immunotherapy is not commonly used. Nevertheless, evidence from the management of other cancers shows that ctDNA has prognostic value to measure immunotherapy response (26). As vaccines and other immune-based therapies are more widely studied in pancreatic cancer, ctDNA may prove useful as an additional biomarker to help evaluate for pseudoprogression.

Maintenance

Standard monitoring approach

Patients with metastatic disease who achieve a response or stable disease after induction chemotherapy can be considered for maintenance therapy (8), with the goal of preserving disease control and maximizing progression-free intervals while supporting quality of life. De-escalation of therapy, either by reducing the toxicity of multi-agent regimens or via the introduction of new agents, often involves mono- or dual-agent therapy combinations, sometimes on modified schedules, to improve tolerability. Multiple de-escalation strategies are available using 5-FU alone (PANOPTIMOX) (27), gemcitabine with reduced nab-paclitaxel frequency (ALPACA) (28), PARP inhibitors for BRCA-mutated patients (POLO) (29,30), and chemotherapy holidays. Besides ongoing serial imaging at the physician’s discretion, NCCN guidelines do not have further explicit recommendations about how to conduct disease monitoring in this maintenance setting. ctDNA monitoring may be particularly useful in this setting to assist physicians with decisions to escalate and de-escalate treatment as illustrated by the following clinical cases.

Clinical vignettes

De-escalation & resumption of maintenance therapy

A 79-year-old woman developed left upper quadrant abdominal pain, weight loss, and loose stools. Imaging showed a pancreatic mass, which was biopsied and revealed poorly differentiated adenocarcinoma. She was started on FOLFIRINOX with dose reduction due to intolerance, and was then continued on FOLFOX for a cycle, followed by 5-FU alone due to adverse effects of chemotherapy. For her fourth cycle, she received one more dose of FOLFOX, followed by imaging, which showed her pancreatic mass had decreased in size from 2.8 to 1.8 cm. She underwent Whipple surgery, followed by repeat imaging, which showed enlarging pulmonary nodules compatible with metastatic disease. She was then started on GnP, with improvement in lung nodules after just two cycles. After another two cycles, repeat imaging showed stable disease. CtDNA, which had been as high as 0.90 MTM/mL and had turned negative for the first time approximately three weeks prior, was repeated one month later and remained negative. The patient elected to take a chemotherapy holiday with close surveillance. A repeat ctDNA test one month later remained negative, and the patient was doing well except for active anxiety and depression. After another month, ctDNA turned positive. She remained asymptomatic; however, repeat imaging showed disease progression with increased pulmonary nodules, one of which was biopsied and showed adenocarcinoma. She resumed chemotherapy with GnP, and after another two cycles of chemotherapy, a repeat CT showed stable disease in her chest, abdomen, and pelvis. Repeat ctDNA at this time showed an ongoing but mild increase, so she was continued on chemotherapy, which she tolerated well.

Maintenance escalation and de-escalation of therapy

A 61-year-old man developed abdominal discomfort and weight loss. Imaging revealed a pancreatic mass with suspected peritoneal, adrenal, and lung involvement. Biopsy of peritoneal disease revealed moderately differentiated adenocarcinoma, and he was started on FOLFIRINOX. After four cycles, irinotecan was held due to diarrhea, and he continued with improved tolerance on FOLFOX for six cycles, after which he transitioned to single-agent 5-FU as maintenance therapy. CT imaging after two cycles of 5-FU showed an overall decrease in tumor burden. He was continued on 5-FU monotherapy for five months, with ongoing disease stability on imaging, at which point ctDNA testing was negative. Given disease stability was supported by the patient’s imaging and CA 19-9 testing, and he desired to minimize therapy while traveling, a treatment break with close ctDNA monitoring was discussed with the patient, who chose this option. He felt well off treatment; however, ctDNA approximately two months later showed an increase from 0.00 to 0.12 MTM/mL. Repeat imaging showed ongoing stable but significant disease, and though he continued to feel well off treatment, given his significant disease burden as well as rising ctDNA, he was advised to resume 5-FU or capecitabine and opted for the latter. He developed mild hand-foot syndrome but has otherwise tolerated single-agent capecitabine well, and his ctDNA level down-trended and remains <1 MTM/mL.

Toward clinical integration of ctDNA

These cases highlight how ctDNA changes may help inform treatment decisions in the maintenance setting. The prolonged nature of maintenance therapy, with no clearly defined duration, can appear daunting to patients. The option of chemotherapy holidays and de-escalation can diminish this burden and lower its associated toxicities. The above cases illustrate the potential for ctDNA to help clinicians judge how to balance the risks and benefits of continuing vs. holding chemotherapy, as well as which regimens to use. However, future trials are needed to help strengthen and formalize the growing application of ctDNA into more robust clinical guidelines. Our field would benefit from large, well-designed studies comparing management of maintenance with and without ctDNA data, to determine whether ctDNA testing makes a meaningful impact on cancer outcomes and quality of life. Various research efforts have already looked at using ctDNA as a prognostic marker to guide which patients may qualify for de-escalation of chemotherapy, though trials such as AGITG DYNAMIC-Pancreas (12) are often limited to other settings.

Strengths and limitations

A key strength of this manuscript is its integrated, real-world perspective on ctDNA use across the full pancreatic cancer continuum, encompassing neoadjuvant, adjuvant, surveillance, and advanced disease settings. By contextualizing illustrative clinical cases alongside current standards of care and proposed trial designs, this work provides a pragmatic framework for how ctDNA may inform decision-making in scenarios where conventional imaging and serum biomarkers are limited. Nonetheless, our findings are hypothesis-generating and derive from anecdotal clinical experience rather than prospective validation. The absence of systematic outcome data, the use of a single commercially available tumor-informed assay, and the evolving evidence base for ctDNA in pancreatic cancer limit the ability to draw definitive conclusions regarding clinical benefit, timing, or standardization of implementation.

Conclusions

As illustrated by our clinical cases, the emergence of ctDNA as a minimally invasive, personalized biomarker offers a promising avenue, not only for earlier detection of molecular recurrence or progression but also for clarifying ambiguous clinical scenarios. The evolving landscape of PDAC management underscores the critical need for more refined surveillance and therapeutic strategies. Current disease monitoring, largely dependent on imaging and informed by CA 19-9 testing, has demonstrated limited capacity to improve survival, highlighting the inadequacy of existing tools in detecting early or micrometastatic recurrence. Substantial work remains to determine the precise role and standardization of ctDNA assays, as well as the establishment of validated clinical protocols for its application in both localized and advanced disease settings. Moving forward, rigorously designed prospective studies are needed to validate ctDNA as a predictive and prognostic tool, define its impact on clinical outcomes, and identify optimal combinatorial strategies that leverage both molecular surveillance and novel therapeutic targets. In contrast to retrospective, real-world evidence studies, where usage patterns of ctDNA between providers may vary widely, the advantage of prospective design is to improve the consistency of ctDNA draw collections and facilitate data aggregation by timepoint. Emphasis should be placed not only on technical refinement but also on demonstrating clear clinical utility, ensuring that advances translate into meaningful benefits for all patients with cancer.

Acknowledgments

None.

Footnote

Peer Review File: Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2026-1-0153/prf

Funding: None.

Conflicts of Interest: Both authors have completed the ICMJE uniform disclosure form (available at https://jgo.amegroups.com/article/view/10.21037/jgo-2026-1-0153/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. Nakamura Y, Watanabe J, Akazawa N, et al. ctDNA-based molecular residual disease and survival in resectable colorectal cancer. Nat Med 2024;30:3272-83. [Crossref] [PubMed]
2. Tie J, Cohen JD, Lahouel K, et al. Circulating Tumor DNA Analysis Guiding Adjuvant Therapy in Stage II Colon Cancer. N Engl J Med 2022;386:2261-72. [Crossref] [PubMed]
3. Bando H, Watanabe J, Kotaka M, et al. A randomized, double-blind, phase III study comparing trifluridine/tipiracil (FTD/TPI) versus placebo in patients with molecular residual disease following curative resection of colorectal cancer (CRC): The ALTAIR study. J Clin Oncol 2025;43:LBA22.
4. American Cancer Society. Cancer Facts & Figures 2025. Atlanta: American Cancer Society; 2025. Available online: https://www.cancer.org/research/cancer-facts-statistics/all-cancer-facts-figures/2025-cancer-facts-figures.html
5. Sugawara T, Rodriguez Franco S, Sherman S, et al. Characteristics and prognosis of patients with pancreatic adenocarcinoma not expressing CA19-9: Analysis of the National Cancer Database. Pancreatology 2024;24:1340-7. [Crossref] [PubMed]
6. Eckhoff AM, Kanu E, Fletcher A, et al. Initial Report: Personalized Circulating Tumor DNA and Survival in Patients with Resectable Pancreatic Cancer. Ann Surg Oncol 2024;31:1444-6. [Crossref] [PubMed]
7. Nassour I, Parrish A, Baptist L, et al. National adoption of neoadjuvant chemotherapy: paradigm shift in the treatment of pancreatic cancer. HPB (Oxford) 2023;25:1323-8. [Crossref] [PubMed]
8. NCCN Clinical Practice Guidelines in Oncology. Pancreatic Adenocarcinoma. Version 2.2025. National Comprehensive Cancer Network. Published February 3, 2025. Available online: https://www.nccn.org/professionals/physician_gls/pdf/pancreatic.pdf. Accessed June 8, 2025.
9. Altman AM, Wirth K, Marmor S, et al. Completion of Adjuvant Chemotherapy After Upfront Surgical Resection for Pancreatic Cancer Is Uncommon Yet Associated With Improved Survival. Ann Surg Oncol 2019;26:4108-16. [Crossref] [PubMed]
10. Botta GP, Abdelrahim M, Drengler RL, et al. Association of personalized and tumor-informed ctDNA with patient survival outcomes in pancreatic adenocarcinoma. Oncologist 2024;29:859-69. [Crossref] [PubMed]
11. Habib JR, Kinny-Köster B, Javed AA, et al. Impact of Adjuvant Chemotherapy on Resected Intraductal Papillary Mucinous Neoplasm-Derived Pancreatic Cancer: Results From an International Multicenter Study. J Clin Oncol 2024;42:4317-26. [Crossref] [PubMed]
12. Lee B, Tie J, Wang Y, et al. The potential role of serial circulating tumor DNA (ctDNA) testing after upfront surgery to guide adjuvant chemotherapy for early stage pancreatic cancer: The AGITG DYNAMIC-Pancreas trial. J Clin Oncol 2024;42:107.
13. Witkowski ER, Smith JK, Ragulin-Coyne E, et al. Is it worth looking? Abdominal imaging after pancreatic cancer resection: a national study. J Gastrointest Surg 2012;16:121-8.
14. Kasi PM, Rivero S, Aushev VN, et al. Patient-reported outcomes from the BESPOKE CRC study. J Clin Oncol 2024;42:54.
15. Woof VG, Lee RJ, Lorigan P, et al. Circulating tumour DNA monitoring and early treatment for relapse: views from patients with early-stage melanoma. Br J Cancer 2022;126:1450-6. [Crossref] [PubMed]
16. Rustin GJ, van der Burg ME, Griffin CL, et al. Early versus delayed treatment of relapsed ovarian cancer (MRC OV05/EORTC 55955): a randomised trial. Lancet 2010;376:1155-63. [Crossref] [PubMed]
17. Lepage C, Phelip JM, Cany L, et al. Effect of 5 years of CT-scan and CEA follow-up on survival endpoints in patients with colorectal cancer: the PRODIGE-13 FFCD phase III trial. Ann Oncol 2025;36:1468-79. [Crossref] [PubMed]
18. Janjigian YY, Foote MB, Lumish M, et al. Abstract CT002: Circulating tumor DNA status to direct adjuvant immunotherapy for mismatch repair deficient tumors. Cancer Res 2025;85:CT002.
19. Nixon AB, Navarro FCP, Zhou KI, et al. Monitoring treatment response using an ultra-sensitive ctDNA assay in advanced esophagogastric cancer patients. Sci Rep 2026;16:14766. [Crossref] [PubMed]
20. Pietrasz D, Sereni E, Lancelotti F, et al. Circulating tumour DNA: a challenging innovation to develop “precision onco-surgery” in pancreatic adenocarcinoma. Br J Cancer 2022;126:1676-83. [Crossref] [PubMed]
21. U.S. Food and Drug Administration. Use of Circulating Tumor DNA for Curative-Intent Solid Tumor Drug Development: Guidance for Industry. Silver Spring, MD: U.S. Department of Health and Human Services, Food and Drug Administration; November 2024. Available online: https://www.fda.gov/media/183874/download
22. Kolbeinsson H, Hoppe A, Bayat A, et al. Recurrence patterns and postrecurrence survival after curative intent resection for pancreatic ductal adenocarcinoma. Surgery 2021;169:649-54. [Crossref] [PubMed]
23. Tanaka M, Mihaljevic AL, Probst P, et al. Meta-analysis of recurrence pattern after resection for pancreatic cancer. Br J Surg 2019;106:1590-601. [Crossref] [PubMed]
24. Chawla A, Qadan M, Castillo CF, et al. Prospective Phase II Trials Validate the Effect of Neoadjuvant Chemotherapy on Pattern of Recurrence in Pancreatic Adenocarcinoma. Ann Surg 2022;276:e502-9. [Crossref] [PubMed]
25. Meneses-Medina MI, Gervaso L, Cella CA, et al. Chemotherapy in pancreatic ductal adenocarcinoma: When cytoreduction is the aim. A systematic review and meta-analysis. Cancer Treat Rev 2022;104:102338.
26. Al-Showbaki L, Wilson B, Tamimi F, et al. Changes in circulating tumor DNA and outcomes in solid tumors treated with immune checkpoint inhibitors: a systematic review. J Immunother Cancer 2023;11:e005854. [Crossref] [PubMed]
27. Dahan L, Williet N, Le Malicot K, et al. Randomized Phase II Trial Evaluating Two Sequential Treatments in First Line of Metastatic Pancreatic Cancer: Results of the PANOPTIMOX-PRODIGE 35 Trial. J Clin Oncol 2021;39:3242-50. [Crossref] [PubMed]
28. Dorman K, Boeck S, Caca K, et al. Alternating gemcitabine plus nab-paclitaxel and gemcitabine alone versus continuous gemcitabine plus nab-paclitaxel after induction treatment of metastatic pancreatic cancer (ALPACA): a multicentre, randomised, open-label, phase 2 trial. Lancet Gastroenterol Hepatol 2024;9:935-43. [Crossref] [PubMed]
29. Kindler HL, Hammel P, Reni M, et al. Overall Survival Results From the POLO Trial: A Phase III Study of Active Maintenance Olaparib Versus Placebo for Germline BRCA-Mutated Metastatic Pancreatic Cancer. J Clin Oncol 2022;40:3929-39. [Crossref] [PubMed]
30. Reiss KA, Mick R, O’Hara MH, et al. Phase II Study of Maintenance Rucaparib in Patients With Platinum-Sensitive Advanced Pancreatic Cancer and a Pathogenic Germline or Somatic Variant in BRCA1, BRCA2, or PALB2. J Clin Oncol 2021;39:2497-505. [Crossref] [PubMed]