March 05, 2025
Author: Celine Hoyek, MD, Mayo Clinic, Arizona
The 2025 American Society of Clinical Oncology Gastrointestinal Cancers Symposium (ASCO GI) highlighted key advances in biomarker testing and targeted therapy, treatment sequencing, the role of circulating tumor DNA (ctDNA), and management of adverse effects. At a satellite event led by the Oncology Brothers—Rahul Gosain, MD, MBA and Rohit Gosain, MD—a panel of 6 GI oncology experts discussed how findings presented at ASCO GI could influence clinical practice and shape the evolving treatment landscape.
Upper GI cancers 
Mark Lewis, MD, from Intermountain Health, and Rutika Mehta, MD, from Weill Cornell Medicine/NewYork-Presbyterian, discussed the management of gastric and gastroesophageal cancers. Dr. Mehta mentioned that improved preoperative staging methods—endoscopic ultrasound, positron emission tomography (PET), and computed tomography (CT) imaging—have reduced the incidence of unexpected, advanced pathology in resected early-stage, node-negative patients. However, in cases where pathology reveals node-negative T2 disease, Dr. Mehta would not administer adjuvant therapy. Based on Asian studies, she recommends fluorouracil (5-FU)–based chemotherapy without radiation therapy for node-positive cases, given the low rate of local recurrence. In cases of positive margins, Dr. Mehta would proceed with postoperative chemoradiation, despite its challenging tolerability.
Dr. Lewis emphasized the systemic nature of upper GI cancers in cases of nodal involvement. In clinical stage T2 or higher node-positive disease, also known as locally advanced esophageal cancer, he compares preoperative chemoradiation from the CROSS trial with perioperative FLOT (5-FU, leucovorin [folinic acid], oxaliplatin, and docetaxel) from the 2024 ESOPEC trial. The CROSS trial’s improved survival at 3 years is attributed to better control of distant metastases and a more favorable toxicity profile, although the inclusion of 25% of patients with squamous cell carcinoma, which responds better to radiotherapy, is worth mentioning in the efficacy assessment. Additionally, according to the ESOPEC trial, patients who benefited less from FLOT were older than 50, with node-negative disease. Since the average age of patients with gastroesophageal cancer is in the mid-60s, some selection bias and demographic differences are worth considering before adopting this regimen. An up-front discussion of treatment intent, whether curative or palliative, and the decision to proceed with esophagectomy is also paramount.
Dr. Lewis then touched on the role of adjuvant nivolumab, evaluated in CheckMate 577, as an option for patients with residual disease after chemoradiation. According to ESOPEC, radiation may increase the probability that immunotherapy works. At the 2024 ASCO Plenary Session, Karyn Goodman, MD, discussed the possibility that patients treated with radiation therapy up front are more likely to benefit from radiation in a later setting. Dr. Lewis then offered 2 clinical scenarios: For robust patients under 50 with node-positive disease and fit for esophagectomy, he opts for aggressive FLOT; for frail or older patients, he prefers a CROSS-style regimen with adjuvant nivolumab in case of residual disease, in addition to recommending esophagectomy since the rates of clinical complete response and pathologic complete response (pCR) are not as high with chemoradiation alone. Looking ahead, Dr. Lewis believes we are on the cusp of nonoperative management strategies for gastroesophageal cancer, similar to the approach used in rectal cancer. Clear communication about treatment expectations, including the potential for watchful waiting, is essential to align with patient preferences and quality-of-life goals.
Dr. Gosain mentioned the importance of discussing CROSS versus ESOPEC to find the balance between adverse effects and the outcomes, with CROSS having a higher pCR (29% vs 19.3%). Postoperative FLOT is particularly difficult, with only two-thirds of patients receiving any postoperative treatment and just half completing the planned 4 cycles. Discussing treatment expectations with patients is crucial. Given the challenges of postoperative FLOT and recovery from extensive surgery, medical oncologists should manage patients supportively with hydration, growth factor support, and nutritional consultations, whereas surgeons should prepare patients for significant lifestyle changes after esophagectomy. To avoid chemotherapy after surgery, some oncologists prefer a total neoadjuvant therapy approach, whereas Dr. Mehta gives 6 cycles of chemotherapy before surgery and 2 after surgery.
For microsatellite instability high (MSI-H) tumors, either single-agent or doublet immunotherapy can be used, depending on the patient’s disease burden and age. In younger patients with aggressive and widespread disease, Dr. Mehta recommends doublet immunotherapy with a CTLA-4 inhibitor. For smaller, node-negative, MSI-H tumors, resection alone may suffice, based on subgroup analyses from the MAGIC trial. Adding immunotherapy to neoadjuvant FLOT may improve pCR rates, as seen in KEYNOTE-585 (pCR, 12.9% with pembrolizumab vs 2.0% with placebo) and the interim analysis of MATTERHORN (pCR, 19% with durvalumab plus FLOT vs 7% with FLOT alone). More mature MATTERHORN data are needed, but it is important to note that neither MATTERHORN nor KEYNOTE-585 study protocols mandated staging laparoscopy to diagnose locally advanced gastric cancer. An exception is the German subgroup within these studies—staging laparoscopy was consistently performed, so the results from this subgroup may provide more reliable insights into the benefits of combining immunotherapy with neoadjuvant FLOT for locally advanced gastric cancer patients.
In the advanced or metastatic stage of upper GI malignancies, treatment decisions are increasingly guided by biomarker status, including HER2, MSI, and claudin 18.2 (CLDN18.2) expression. Dr. Gosain emphasized the importance of identifying these biomarkers before initiating the recommended treatment.
1. HER2-positive
- First line: For PD-L1 of 1 or greater, chemotherapy plus trastuzumab plus pembrolizumab is recommended; For low/negative PD-L1, chemotherapy plus trastuzumab is recommended.
- Second line: trastuzumab deruxtecan (T-Dxd).
2. CLDN18.2-positive: when immunohistochemistry (IHC) staining is greater than 2 in at least 75% of cells, chemotherapy plus zolbetuximab is recommended.
3. PD-L1-positive: if combined positive score (CPS) is 1 or greater, either pembrolizumab or tislelizumab plus chemotherapy can be used. If CPS 5 or greater, nivolumab plus chemotherapy is recommended. Note that the National Comprehensive Cancer Network currently uses CPS thresholds of 5-10 for treatment decisions, although recent discussions at the Oncologic Drugs Advisory Committee meeting suggest a threshold of CPS of 1 or greater could soon become the standard.
4. Heavy visceral, symptomatic disease: up-front chemotherapy plus immunotherapy.
Managing patients with co-expression of HER2, CLDN18.2, or PD-L1 positivity remains to be investigated. For CLDN18.2-positive disease, zolbetuximab combined with chemotherapy has emerged as a promising option based on the Spotlight trial, though its rollout has been complicated by the need for companion diagnostics due to CLDN18.2 having a different cutoff for positivity than other assays. Managing CLDN18.2-positive patients who missed first-line zolbetuximab will be a significant clinical challenge in the next 1 to 2 years. Zolbetuximab shows minimal activity as a single agent and has improved efficacy when combined with FOLFOX (leucovorin, 5-FU, and oxaliplatin). However, combinations with FOLFIRI (leucovorin, 5-FU, and irinotecan). However, combinations with FOLFIRI raise toxicity concerns (gastritis, nausea, vomiting), whereas a 5-FU and zolbetuximab regimen is limited by the minimal activity of single-agent 5-FU (approximately 15% response rate). In patients co-expressing PD-L1 (5 or greater) and CLDN18.2, especially those with bulky tumors and significant disease burden, Dr. Mehta would administer chemotherapy with an immune checkpoint inhibitor (ICI) for rapid tumor control. In this case, zolbetuximab plus modified FOLFOX (mFOLFOX) might be suboptimal because, despite an improved progression-free survival (PFS) compared with mFOLFOX alone in Spotlight (10.61 months [95% CI, 8.90-12.48] vs 8.67 months [CI, 8.21-10.28]), the overall response rate (ORR) is moderate (48% in both arms). The primary adverse effect of zolbetuximab is nausea, which can be managed by slowing the infusion rate and using supportive medications such as aprepitant, ondansetron, and olanzapine. Oncologists prescribing zolbetuximab are advised to refer to the Delphi panel recommendations by Samuel Klempner, MD, which provide additional insight on managing adverse effects.
In HER2-positive cases, standard of care is trastuzumab combined with chemotherapy, with pertuzumab omitted when CPS is less than 1. Although initial approvals covered all patients, overall survival (OS) benefits have been demonstrated only in those with CPS greater than 1. T-DXd has shown impressive responses and is under evaluation in the frontline setting, but the risk of interstitial lung disease requires close monitoring for early symptoms.
In the second- and third-line settings, options include T-DXd for HER2-positive cases and FOLFIRI or ramucirumab for HER2-negative patients. Dr. Lewis prefers ramucirumab due to its mechanistic novelty. If a patient has residual neuropathy from prior platinum-based treatment, ramucirumab monotherapy can be used per REGARD, which showed an improved modified OS (mOS) from 3.8 months to 5.2 months compared with placebo.
Colorectal Cancer 
Cathy Eng, MD, from Vanderbilt-Ingram Cancer Center, and Richard Dunne, MD, from Wilmot Cancer Institute at the University of Rochester, discussed the rapidly evolving treatment landscape of colorectal cancer (CRC), starting with stage II.
Patients with stage IIA CRC can be divided into low and high risk, based on clinicopathological features. High-risk features include lymphovascular invasion, angiolymphatic invasion, and intestinal perforation and might be associated with a greater likelihood of recurrence. Dr. Dunne emphasized the importance of circulating tumor DNA (ctDNA) in detecting minimal residual disease after surgery, which constitutes an important technological advancement over more invasive diagnostic tests. At the University of Rochester, ctDNA analysis is systematically ordered 14 days post resection of stage II CRC, for treatment decision-making and dynamic studies. He mentioned that according to the BESPOKE CRC study, a majority of patients will have undetectable ctDNA and therefore a low risk of recurrence, so they may avoid chemotherapy. For patients with detectable ctDNA, recurrence risk is significantly higher. The decision to escalate therapy is individualized, acknowledging the limitations of current evidence; however, adjuvant chemotherapy can be recommended. Dr. Dunne typically uses oxaliplatin-based therapies similarly to the DYNAMIC trial, keeping in mind that historical studies (MOSAIC and later pooled analyses) question their benefit in high-risk stage II patients.
According to Dr. Eng, ordering ctDNA testing in stage II CRC is based on shared decision-making. In high-risk stage II CRC, she recommends ctDNA testing if the patient is agreeable. If a patient with no high-risk features tests positive for ctDNA, the decision to proceed with therapy is also made collaboratively, as a positive result does not guarantee recurrence. Dr. Eng advocates for serial ctDNA testing to monitor and guide adjuvant chemotherapy decisions. If the patient agrees to proceed with treatment and is a good candidate, she recommends oxaliplatin-based chemotherapy but also discusses the option of 5-FU–based regimens.
The experts then tackled the management of stage III CRC. For patients with T1-T3, N1, low-risk stage III disease, the current standard of care is based on findings from the IDEA trial, which support using CAPOX(capecitabine plus oxaliplatin) in the adjuvant setting for 3 months. This regimen has comparable efficacy to a 6-month regimen while sparing patients from long-term neuropathy. For patients unable to tolerate CAPOX due to factors such as poor creatinine clearance, the alternative is FOLFOX for 3 months followed by single-agent 5-FU to complete a 6-month therapy course. Since there is equipoise on ctDNA use and interpretation for clinical decision-making, Dr. Dunne enrolls his patients in the CIRCULATE-US trial, which seeks to understand how to integrate ctDNA into clinical practice. Off trial, Dr. Dunne uses ctDNA as a tool at this stage, along with carcinoembryonic antigen tests and CT scans for disease monitoring.
The discussion also addressed the ongoing debate regarding the necessity of bolus 5-FU and leucovorin. Dr. Eng emphasized the importance of omitting bolus 5-FU altogether due to its associated myelosuppression that leads to unnecessary dose delays, thereby compromising dose intensity. Dr. Dunne similarly supported the decision to eliminate bolus 5-FU in the metastatic setting, citing real-world data from Flatiron Health; however, he continues to implement it in the adjuvant setting.
One of the most challenging scenarios in stage III CRC involves managing patients with persistently positive ctDNA despite negative imaging findings after completing adjuvant chemotherapy. Dr. Eng mentioned that there is no role for adjuvant FOLFOXIRI (5-FU, leucovorin, oxaliplatin, and irinotecan) and recommends FOLFOX or CAPOX. Close monitoring is essential, with some oncologists opting for a CT scan after 3 months instead of the standard 6-month interval. To detect small liver metastases that might have been missed, Dr. Dunne alternates imaging modalities, such as MRI or PET, particularly for patients with rising ctDNA levels and negative CT scans. The ALTAIR study will provide more clarity on whether giving chemotherapy based only on ctDNA positivity will improve disease-free survival. Given potential false negatives, especially in detecting small lung or peritoneal metastases, oncologists must remain vigilant in balancing observation with timely intervention.
In the metastatic setting, biomarker testing is essential to guide therapy decisions. For patients with BRAF V600E–mutated metastatic CRC (mCRC), the BEACON trial established encorafenib combined with cetuximab as the standard of care for second-line treatment. The BREAKWATER trial, a phase III study evaluating first-line chemotherapy combined with encorafenib and cetuximab versus standard of care for BRAF V600E–mutated mCRC, reported an objective response rate (ORR) of 60.9% versus 40.0%. Despite pending overall survival, BREAKWATER is expected to be practice-changing for these patients with poor prognosis. However, skin toxicities are an adverse effect; Dr. Eng recommends proactive management, including sun protection and topical doxycycline.
For MSI-H disease, Dr. Eng favors single-agent pembrolizumab in certain cases, and combination therapy in patients with significant tumor burden. In CheckMate 8HW, presented at ASCO GI 2025, nivolumab-ipilimumab had significantly improved PFS (HR, 0.62; 95% CI, 0.48-0.81) and ORR (71% vs 58%) compared to nivolumab monotherapy, with pending OS data.
For left-sided RAS wild-type tumors, the standard treatment is 5-FU or capecitabine-based chemotherapy combined with an anti-EGFR agent or FOLFOXIRI. Dr. Eng mentioned that although data suggest an OS benefit with anti-EGFR therapy in certain subsets, patient preferences and disease characteristics frequently guide treatment choices, as some patients may prioritize quality of life over marginal survival benefits, particularly when faced with adverse effects such as extensive rash. For right-sided RAS wild-type tumors, chemotherapy and anti-VEGF therapies are used for first-line treatment. Dr. Eng mentioned a lack of great data for adding anti-EGFR therapies in second line. As for the role of next-generation sequencing, Dr. Dunne orders liquid biopsy at diagnosis and at first progression, especially in HER2-positive disease.
In the refractory setting, treatment options include targeted and nontargeted therapies. HER2-positive, RAS wild-type tumors are treated with tucatinib/trastuzumab, whereas HER2-positive, RAS mutant tumors are treated with T-DXd. On the other hand, nontargeted therapies such as TAS-102 with bevacizumab, regorafenib, and fruquintinib offer additional options for patients without actionable mutations. Dr. Eng emphasized the importance of tailoring therapy to patient needs and tumor characteristics.
Hepatobiliary and Neuroendocrine Tumors 
Tanios Bekaii-Saab, MD, from Mayo Clinic, and Arndt Vogel, MD, from Toronto General Hospital–Princess Margaret Cancer Centre, joined Dr. Rahul Gosain and Dr. Rohit Gosain for a comprehensive summary of treatment practices in hepatocellular carcinoma (HCC), pancreatic cancer, and neuroendocrine tumors (NETs). To begin, Dr. Bekaii-Saab laid down the range of treatment modalities used in managing early- and intermediate-stage HCC. For early-stage, Child-Pugh A-B HCC, options include surgical resection for a tumor less than 2 cm, ablation in the case of 2 or 3 nodules less than 2 cm, and liver transplantation for 1 to 3 nodules less than 3 cm. Localized radiation therapies like stereotactic body radiation therapy, as well as transarterial therapies such as transarterial radioembolization and transarterial chemoembolization, are reserved for intermediate-stage (B) and Child-Pugh A-B. With a multitude of treatments available, a multidisciplinary approach involving surgeons, radiation oncologists, interventional radiologists, pathologists, and medical oncologists is necessary. For advanced stages, systemic therapies include ICIs, tyrosine kinase inhibitors (TKIs), and combination regimens like atezolizumab-bevacizumab (atezo-bev).
Dr. Bekaii-Saab emphasized the importance of personalized treatment strategies: Individuals with solid organ transplants or uncontrolled autoimmune diseases are typically excluded from ICI-based treatments due to the risk of adverse immune reactions. Additionally, patients who live in rural areas or prefer oral therapies may resort to TKIs to avoid the logistical challenges of frequent intravenous infusions. However, for most patients with advanced HCC, ICI-based therapies remain the recommended treatments, with atezo-bev and dual checkpoint inhibitors durvalumab-tremelimumab (durva-treme) currently approved. Ipilimumab-nivolumab (ipi-nivo) is under investigation in CheckMate 9DW, a phase III trial comparing this combination to lenvatinib or sorafenib. Cross-study comparisons of ICI-based therapies are complicated by differences in trial designs and patient populations. For example, the phase III trial that led to the approval of atezo-bev for unresectable HCC included patients with main portal vein thrombosis, a subgroup with poorer prognosis and higher risk of bleeding. Typically, atezo-bev and ipi-nivo are recommended for patients with a heavy disease burden. However, despite the improved mOS in the ipi-nivo arm compared with the lenvatinib or sorafenib arm (23.7 months vs 20.6 months) in CheckMate 9DW, the rates of adverse events and treatment discontinuation were higher. As such, Dr. Bekaii-Saab limits the use of ipi-nivo in his practice due to tolerability concerns, along with historical data that do not show considerable survival advantage. However, long-term data from CheckMate 9DW are still needed. In patients with lower disease burden and a reduced risk of bleeding, Dr. Bekaii-Saab recommends durva-tremie, which has shown long-term survival rates in the HIMALAYA trial.
According to Dr. Vogel, therapy decisions are significantly influenced by factors such as liver function, performance status, and treatment intent rather than underlying liver disease, such ascirrhosis or metabolic dysfunction-associated steatohepatitis (formerly known as nonalcoholic steatohepatitis). Initiating patients on immuno-oncology (IO)-based therapies often preserves liver function, improves survival, and allows for subsequent therapy options. However, there is limited prospective data on the optimal sequencing after first-line ICI. A Mayo Clinic single-center study presented at ASCO GI 2025 suggested that TKIs do have post-IO benefit. In Dr. Bekaii-Saab’s practice, cabozantinib (preferentially) or regorafenib is used as second-line treatment after atezo-bev or durva-treme, whereas ramucirumab can be considered following progression on durva-treme. Though not as commonly prescribed, sorafenib may be another option in later lines at lower doses. Reintroducing IO therapies may be considered, particularly in patients who responded to first-line IO, but directly switching from one IO regimen to another is generally not recommended.
Managing TKI adverse effects, such as hypertension, diarrhea, and fatigue, remains a significant concern. Lenvatinib is often preferred over sorafenib due to its superior efficacy and improved quality of life. Close monitoring, early dose adjustments, and patient education are essential to relieve these adverse effects. Maintaining liver function is a priority because long-term systemic therapy can lead to hepatic decline. As the disease progresses, clinicians must balance aggressive treatment and palliative care for patient comfort by introducing these discussions and setting up expectations before initiating treatment.
Pancreatic Cancer
Dr. Bekaii-Saab mentioned that management of early-stage pancreatic cancer has largely shifted toward a fully neoadjuvant approach, driven by the disease’s systemic nature at diagnosis and the critical need for exposure to chemotherapy—ideally, before surgery, when patients are able to tolerate it. Among patients with clearly resectable disease (15%-20%), adjuvant therapy has been shown to significantly improve survival; however, many patients do not ultimately receive it. Modified FOLFIRINOX (mFOLFIRINOX) is the preferred regimen in both the adjuvant and neoadjuvant settings. For patients who may not tolerate mFOLFIRINOX due to performance status or adverse effects, neoadjuvant gemcitabine/nab-paclitaxel is an alternative; however, this combination failed to show benefit in the adjuvant setting.
In borderline resectable cases, neoadjuvant mFOLFIRINOX aims to achieve a significant or complete response prior to resection. If not reached, treatment is switched to gemcitabine/nab-paclitaxel. The role of radiation in pancreatic cancer remains controversial and is primarily reserved to improve resection outcomes in borderline resectable cases. Patients with BRCA1/2 or PALB2 mutations may benefit from gemcitabine-cisplatin, administered biweekly to balance efficacy and tolerability.
Dr. Vogel notes that in academic centers, younger patients often receive more intensive triplet chemotherapy regimens, such as NALIRIFOX (liposomal irinotecan, 5-FU, leucovorin, and oxaliplatin). Before the introduction of NALIRIFOX, mFOLFIRINOX and gemcitabine/nab-paclitaxel were standard-of-care options, though direct head-to-head comparisons are lacking. Although some cross-trial comparisons suggest that NALIRIFOX may be more effective than gemcitabine/nab-paclitaxel, its advantage over mFOLFIRINOX remains unclear due to limited data. As a result, both mFOLFIRINOX and NALIRIFOX constitute valid treatment options.
Management of treatment adverse effects is crucial. Gemcitabine/nab-paclitaxel is considered more tolerable than mFOLFIRINOX and is preferred for older or frail patients, with a recommended dosing schedule of every other week rather than weekly to reduce toxicity while maintaining efficacy. The most common adverse effects of nab-paclitaxel include neuropathy and hematologic toxicities. NALIRIFOX and mFOLFIRINOX have similar toxicity profiles, although GI adverse effects are more frequent with NALIRIFOX. These can be managed effectively through patient education and proactive symptom control.
For unresectable or metastatic pancreatic cancer, first-line systemic treatment options similarly include mFOLFIRINOX, NALIRIFOX, gemcitabine/nab-paclitaxel, and gemcitabine-cisplatin for patients with BRCA1/2 or PALB2 mutations. However, enrollment in clinical trials is typically recommended. Targeted mutations are rare, but therapies are available for NTRK fusions (entrectinib/larotrectinib), RET fusions (selpercatinib), MSI-H or tumor mutational burden (TMB)–high (TMB ≥10 mutations/megabase) tumors (pembrolizumab), and BRAF V600E mutations (dabrafenib plus trametinib).
Neuroendocrine Tumors
The management of NETs depends on tumor functionality. Nonfunctional, localized NETs are managed with observation, whereas functional or metastatic tumors often require intervention with local therapies such as liver-directed treatment, surgery, or systemic treatments. Strategies vary according to tumor grade, disease spread, and Ki-67 index. Managing metastatic somatostatin receptor–positive NET not amenable to liver-directed treatment or resection requires somatostatin analogues such as octreotide and lanreotide. Data from the STARTER-NET trial suggest that combining everolimus and lanreotide may improve PFS compared with everolimus alone, although OS benefits remain modest. Dr. Vogel noted that the PFS benefits seen in STARTER-NET align with previous studies, all of which failed to report an OS advantage. Regardless, the combination of everolimus and lanreotide can be a valid option for patients with high proliferation rates and poor prognostic factors. Capecitabine plus temozolomide (CAPTEM) can be considered in the first line, while streptozotocin remains an option for pancreatic NETs.
For patients with slowly progressing disease, close observation may be preferable; in more aggressive cases, a combination of everolimus and lanreotide or peptide receptor radionuclide therapy (PRRT) is recommended. At Mayo Clinic, PRRT has become a first-line option for somatostatin receptor–positive NETs, given its high response rates and durable outcomes. Patients can also be rechallenged with PRRT, depending on the treatment interval. If progression occurs on first-line PRRT, treatment options include everolimus, sunitinib (significant toxicity), cabozantinib (strong PFS data in the CABINET trial), and CAPTEM. However, temozolomide carries an increased risk of myelodysplastic syndrome and presents considerable toxicity concerns. Dr. Saab recommended considering cabozantinib post PRRT, except in highly symptomatic patients who may benefit more from the strong response associated with temozolomide/capecitabine.
Additionally, managing treatment-related toxicities, such as those associated with TKIs (diarrhea, fatigue, hypertension), remains essential, with supportive care measures (e.g., steroid mouthwash for mucositis) playing a key role. Throughout treatment, managing side effects, maintaining organ function, and considering patient preferences are central to improving outcomes across these challenging cancer types. Early integration of palliative care ensures a patient-centered approach, emphasizing quality of life alongside prolonged survival..