The management of gastric and oesophageal cancer increasingly depends on histology, stage and molecular characteristics. Perioperative immunotherapy, HER2-directed treatments and biomarker-selected first-line regimens have expanded therapeutic options, while the appropriate sequencing of these approaches remains an important clinical consideration.
On October 15, 2026, Annals of Oncology published “Gastric and oesophageal cancer: ESMO Clinical Practice Guideline for diagnosis, treatment and follow-up,” authored by S.-E. Al-Batran and colleagues on behalf of the ESMO Guidelines Committee. The guideline covers diagnosis, staging, localized and metastatic disease, and survivorship. It distinguishes oesophageal squamous-cell carcinoma (OSCC) from gastroesophageal adenocarcinoma (GOAC), which includes adenocarcinomas of the oesophagus, oesophagogastric junction (OGJ) and stomach.
Diagnosis, Molecular Testing and Staging
Presenting symptoms may be non-specific, including dysphagia, weight loss, vomiting, early satiety and anaemia. Endoscopy with biopsy remains the diagnostic standard. The guideline advises that 6–10 biopsies can be obtained from advanced tumours to provide sufficient pathological material and account for tumour heterogeneity; 1–2 biopsies are generally adequate for early lesions [IV, B]. Histology can be reported using WHO and Laurén classifications.
For gastroesophageal adenocarcinoma, the guideline also describes four molecular groups—Epstein–Barr virus-positive, MSI-H, genomically stable and chromosomal instability tumours—although treatment algorithms principally depend on clinically actionable biomarkers.
For GOAC, testing for HER2, PD-L1, MSI-H/dMMR and CLDN18.2 informs treatment selection. HER2 expression is evaluated by immunohistochemistry (IHC), with in situ hybridisation when indicated. PD-L1 testing uses the combined positive score (CPS) or tumour area positivity (TAP), depending on the regimen. CLDN18.2 positivity for zolbetuximab requires moderate or strong membranous staining in at least 75% of tumour cells. In OSCC, PD-L1 may be measured using CPS, tumour proportion score (TPS) or TAP. The guideline assigns [I, A] to the relevant HER2, PD-L1 and CLDN18.2 assessments and [II, A] to MSI-H/dMMR testing for immunotherapy selection.
Initial staging includes clinical examination, endoscopy and CT of the thorax, abdomen and pelvis [V, A]. Endoscopic ultrasound can assist T and N staging. FDG-PET may be used in candidates for oesophagectomy, but is not routinely recommended in gastric cancer [III, D], particularly given limitations in diffuse or mucinous tumours. Diagnostic laparoscopy, with or without peritoneal washings for cytology, can be recommended for patients with resectable gastric cancer who are candidates for perioperative chemotherapy [III, B].
Nutritional status, weight loss and fitness for treatment should also be assessed. Positive peritoneal cytology without gross peritoneal dissemination carries an adverse prognosis, and the benefit of gastrectomy in this setting remains unclear. Such cases require individualized multidisciplinary decisions. Treatment decisions should be discussed by a multidisciplinary team [IV, A].
Localized Gastroesophageal Adenocarcinoma
For carefully selected stage I lesions, endoscopic or surgical resection alone may be curative [III, B]. Endoscopic submucosal dissection is preferred for appropriate early gastric tumours, with eligibility determined by invasion depth, differentiation, ulceration, size and lymphovascular involvement. More advanced gastric disease requiring surgery should undergo D2 gastrectomy in high-volume centres [II, A]; laparoscopic surgery is an option in expert hands [II, B].
For locally advanced, resectable GOAC (≥cT2 or cN+, M0), perioperative fluorouracil, leucovorin, oxaliplatin and docetaxel (FLOT) with radical surgery remains the standard chemotherapy-based approach [I, A]. In FLOT4, perioperative FLOT improved overall survival compared with older epirubicin-containing regimens (HR 0.77).
For resectable oesophageal and OGJ adenocarcinoma, ESOPEC favoured FLOT over CROSS-type preoperative chemoradiotherapy, with median overall survival of 66 versus 37 months (HR 0.70). The TOPGEAR trial did not show a survival benefit from adding preoperative chemoradiotherapy to perioperative chemotherapy. The guideline now recommends adding durvalumab to perioperative FLOT, followed by durvalumab maintenance, as the preferred option where available [I, A].
This recommendation follows the phase III MATTERHORN trial involving 948 patients with resectable stage II–IVA gastric or gastroesophageal adenocarcinoma. Two-year event-free survival was 67.4% with durvalumab–FLOT versus 58.5% with FLOT plus placebo (HR 0.71; P < 0.001); pathological complete response was 19.2% versus 7.2%.
Final analysis also showed an overall survival benefit (HR 0.78; P = 0.021), with grade 3–4 adverse events in 71.6% and 71.2%, respectively. The treatment pathway includes preoperative therapy, restaging to exclude metastases, surgery, postoperative treatment and maintenance. When durvalumab is unavailable, FLOT remains recommended.
For oesophageal or OGJ adenocarcinoma unsuitable for FLOT, preoperative chemoradiotherapy followed by surgery is an alternative. Adjuvant nivolumab can then be recommended in selected patients with PD-L1-positive tumours and residual disease after R0 resection [I, B]. Definitive chemoradiotherapy is recommended for patients with locally advanced, unresectable OGJ adenocarcinoma or those who are unwilling or unfit for surgery [III, B]. The benefit of perioperative chemotherapy for cT2N0 disease remains debated.
In localized MSI-H/dMMR gastric or oesophagogastric junction adenocarcinoma, the phase II NEONIPIGA and INFINITY studies reported pathological complete response rates of approximately 60% with neoadjuvant immunotherapy approaches. Although these findings have encouraged investigation of non-operative management, the guideline does not establish that strategy as routine care.
Localized Oesophageal Squamous-Cell Carcinoma
For fit patients with resectable, locally advanced OSCC, ESMO recommends neoadjuvant chemoradiotherapy using the CROSS regimen, restaging and surgery [I, A]. Surgery remains recommended even after an apparent complete clinical response. Following R0 resection after preoperative chemoradiotherapy, adjuvant nivolumab is recommended for patients with residual pathological disease (≥ypT1 and/or ≥ypN1) and no metastases [I, A].
Definitive chemoradiotherapy is preferred for unresectable OSCC or for patients who cannot or do not wish to undergo surgery [III, B], including selected cervical tumours. Salvage surgery may be considered after incomplete response or local recurrence, but only in selected patients at experienced centres [IV, C].
First-Line Treatment of Advanced GOAC
For unresectable or metastatic GOAC, the standard chemotherapy backbone is a platinum–fluoropyrimidine doublet [I, A]. Oxaliplatin is preferred particularly in older patients, while dose-reduced oxaliplatin–capecitabine is an option for those unable to tolerate full-dose chemotherapy. FOLFIRI may be used for platinum intolerance or progression during or soon after perioperative (D)-FLOT [I, B]. Additional treatment is selected according to biomarkers and prior therapy.
HER2-Positive Disease
The guideline recommends zanidatamab plus tislelizumab and chemotherapy, where available, for tumours that are HER2 IHC 3+ or IHC 2+/ISH-positive [I, A]. In HERIZON-GEA-01, both zanidatamab-based arms improved progression-free survival versus trastuzumab–chemotherapy (P < 0.0001), while the zanidatamab–tislelizumab triplet significantly improved overall survival (HR 0.72; 95% CI 0.57–0.90; P = 0.004).
Zanidatamab plus chemotherapy without tislelizumab can also be considered for HER2 IHC 3+ tumours [I, B]. If zanidatamab is unavailable, contraindicated or not tolerated, trastuzumab–chemotherapy is recommended [I, A], with pembrolizumab added when PD-L1 CPS is ≥1 [I, A]. The guideline records the zanidatamab-based regimens as FDA-approved, but not EMA-approved, at the time of writing. It also notes that the incremental benefit of adding tislelizumab to zanidatamab–chemotherapy in PD-L1-negative tumours remains uncertain.
HER2-Negative, PD-L1-Positive Disease
Immunotherapy–chemotherapy recommendations use drug-specific thresholds: nivolumab for CPS ≥5, pembrolizumab for CPS ≥1, and tislelizumab for TAP ≥5%, each with a [I, A] recommendation. The guideline notes that PD-L1 scoring methods employ different assessment methodologies.
CLDN18.2-Positive Disease
For patients with HER2-negative, CLDN18.2-positive advanced gastroesophageal adenocarcinoma, zolbetuximab combined with platinum–fluoropyrimidine chemotherapy is recommended when at least 75% of tumour cells have moderate or strong CLDN18.2 staining [I, A]. Nausea and vomiting, particularly during initial infusions, require attention to antiemetic prophylaxis and infusion management.
MSI-H/dMMR and Biomarker-Negative Disease
Nivolumab, pembrolizumab or tislelizumab combined with chemotherapy are recommended for untreated advanced MSI-H/dMMR GOAC [II, A]. The guideline specifies that no immune checkpoint inhibitor had an EMA- or FDA-approved indication specifically for this first-line MSI-H/dMMR setting at the time of writing. For selected patients with biomarker-negative tumours who are not candidates for immunotherapy and require rapid disease control, the docetaxel-based TFOX triplet can be considered [I, B].
Advanced Oesophageal Squamous-Cell Carcinoma
A platinum–fluoropyrimidine or platinum–taxane doublet is the chemotherapy standard for advanced OSCC [II, A]. For PD-L1-positive disease, first-line chemotherapy can be combined with nivolumab (TPS ≥1%), pembrolizumab (CPS ≥10), tislelizumab (TAP ≥5%) or toripalimab (CPS ≥1), each recommended [I, A]. Nivolumab–ipilimumab is another option for TPS ≥1% [I, B], but its lower radiological response rate and higher risk of early progression without chemotherapy must be considered.
For patients previously treated with platinum–fluoropyrimidine chemotherapy who have not received an immune checkpoint inhibitor, nivolumab or tislelizumab is recommended [I, A]. Taxane- or irinotecan-based chemotherapy remains an option according to previous treatment and fitness.
Second- and Later-Line Treatment of GOAC
For patients progressing after first-line therapy who can receive chemotherapy and antiangiogenic treatment, ramucirumab plus paclitaxel remains the recommended second-line regimen [I, A], based on the RAINBOW trial. Ramucirumab monotherapy is an option when chemotherapy is contraindicated [I, B]. For patients with contraindications to antiangiogenic therapy, irinotecan or taxane monotherapy is recommended [I, A]. When ramucirumab is unavailable, FOLFIRI may also be considered [III, B]. Pembrolizumab is recommended for MSI-H/dMMR tumours when anti-PD-1 therapy has not previously been given [II, A].
For HER2-positive GOAC previously treated with trastuzumab–chemotherapy, trastuzumab deruxtecan (T-DXd) is recommended in either the second- or third-line setting [I, A]. In a randomized phase III trial, T-DXd significantly improved overall survival compared with ramucirumab–paclitaxel in patients whose HER2 positivity was confirmed by tumour biopsy following progression on first-line trastuzumab-based treatment. Approximately 30% of patients screened for the trial had lost HER2 positivity after first-line therapy.
After at least two treatment lines, trifluridine–tipiracil is recommended if oral treatment is feasible [I, A]. Irinotecan or a taxane may be used if not previously administered [II, B]. Immunotherapy remains a consideration for MSI-H/dMMR tumours when not given earlier [III, A].
Local Treatment, Follow-Up and Supportive Care
Routine gastrectomy has not shown a survival advantage in metastatic gastric cancer in the REGATTA trial; the RENAISSANCE trial likewise did not establish an overall survival benefit from adding radical surgery to systemic treatment in limited metastatic GOAC.
Radiotherapy may alleviate dysphagia, bleeding or painful metastases, while early palliative care and nutritional support are recommended. After resection of OSCC or GOAC, structured surveillance with clinical assessment, endoscopy and imaging can be considered for at least five years, typically every three months during the first two years and every six months thereafter [IV, B].
After a complete response of OSCC to definitive chemoradiotherapy, follow-up every three months for at least two years with endoscopy, biopsies and CT is recommended as an option [IV, B]. In advanced disease, CT or MRI every 6–12 weeks can be considered for patients eligible for further treatment. Long-term care should address gastrointestinal symptoms, psychological needs and nutritional deficiencies, including vitamin B12, iron, folate, zinc and vitamin D after surgery.
Osteoporosis screening and management are also important after gastrectomy. Although surveillance can help detect potentially treatable recurrence and address treatment-related problems, the guideline notes that a survival benefit from routine follow-up has not been clearly established.
Conclusion
The 2026 ESMO guideline integrates stage-based treatment with molecular selection across gastric and oesophageal cancers. Durvalumab–FLOT is the preferred perioperative approach for eligible locally advanced GOAC where available, while HER2, PD-L1, CLDN18.2 and MSI/dMMR guide treatment in advanced adenocarcinoma. Separate OSCC pathways prioritize chemoradiotherapy and surgery for resectable disease and biomarker-selected immunotherapy for metastatic disease. Across settings, the guideline emphasizes multidisciplinary planning, appropriate sequencing, supportive care and individualized follow-up.
The full guideline is available in Annals of Oncology.
Find more videos on OncoDaily YouTube TV.
