01Core principlesThe concepts and mechanisms needed to understand the subject.
An upper-GI cancer report must allow another clinician to reconstruct the lesion. Distance from incisors distinguishes cervical, thoracic and junctional oesophagus; cardia and junction landmarks help allocate a tumour; gastric site determines the likely resection. Photographs and diagrams complement, but do not replace, clear text. A statement such as “mass biopsied” is insufficient for planning.
Sample viable tumour at its edge and surface. Current BSG/AUGIS guidance specifies at least eight samples for an advanced oesophageal or gastric cancer, one or two targeted samples for a potentially resectable superficial lesion, and at least ten bite-on-bite samples for suspected linitis. Separate jars prevent a second lesion or background precursor field from being merged. When diffuse infiltration remains suspicious despite negative mucosa, repeat expert deep sampling or an alternative tissue route rather than accepting a false-negative surface result.
Diagnosis and stage are separate questions. Whole-body CT maps histologically confirmed cancer; PET-CT is added before radical oesophageal or junctional treatment except T1a; EUS is selective and not used solely to distinguish T2 from T3. Endoscopic resection itself stages suspected T1N0 oesophageal cancer, and MAPS III advises against routine CT, EUS, MRI or PET-CT before resection of an appropriate early gastric lesion. Staging laparoscopy then detects occult peritoneal spread in the radical gastric-surgery pathway.
Key points
- Describe tumour site relative to incisors, gastro-oesophageal junction, cardia, pylorus and other fixed landmarks because classification changes staging and operation.
- Record length, circumference, morphology, ulceration, bleeding, obstruction and whether the scope passes without force.
- Match sample burden to the lesion: at least eight viable biopsies for advanced oesophageal or gastric cancer, only one or two targeted biopsies for potentially endoscopically resectable neoplasia, and at least ten bite-on-bite samples for suspected gastric linitis.
- Biopsy establishes tumour type; whole-body CT maps established cancer, while suspected T1N0 oesophageal adenocarcinoma is first staged by endoscopic resection and an early gastric lesion suitable for resection needs no routine pre-resection CT, EUS or PET-CT.
- Endoscopic resection is both diagnostic and therapeutic for selected superficial lesions because the intact specimen reveals depth, margins and lymphovascular invasion.
- Request biomarker tests on adequate tissue when advanced disease is possible, preserving material and repeating biopsy if results will determine systemic therapy.
02Mechanisms and patternsImportant relationships and how to distinguish them.
Contact bleeding, ulceration, shouldering and progressive narrowing require multiple viable biopsies.
Smooth mucosa over a firm bulge can represent stromal, lymphoid or infiltrative disease and may need EUS-guided tissue.
Barrett dysplasia or gastric metaplasia surrounding cancer must be mapped separately because residual mucosa changes follow-up.
Inability to pass safely changes nutrition and staging technique; forceful traversal risks perforation.
Benign surface tissue with a rigid stomach or malignant imaging pattern mandates multidisciplinary re-evaluation.
03Interpreting evidenceInformation, measurements and their limitations.
Consider the information, its meaning and its limitations before deciding what follows.
- 01
High-definition endoscopy - Why
- Map the lesion and identify the safest viable biopsy targets.
- Interpretation and limitations
- Report fixed landmarks, dimensions, circumference, morphology and traversability with representative images.
- 02
Forceps biopsy matched to lesion intent - Why
- Establish histology while preserving the next diagnostic or therapeutic step.
- Interpretation and limitations
- Use at least eight viable biopsies for advanced cancer, one or two targeted biopsies for a potentially endoscopically resectable lesion, and at least ten bite-on-bite biopsies for suspected linitis; stop unsafe traversal.
- 03
Endoscopic resection specimen - Why
- Provide en-bloc pathological staging for selected superficial neoplasia.
- Interpretation and limitations
- Depth, differentiation, ulceration, margins and lymphovascular invasion determine whether resection was curative. For T1b oesophageal adenocarcinoma in a person fit for surgery, NICE identifies deep submucosal invasion over 500 micrometres, lymphovascular invasion or incomplete resection as high-risk findings that support oesophagectomy.
- 04
Pathology review - Why
- Resolve histological uncertainty and confirm dysplasia or early cancer.
- Interpretation and limitations
- Expert upper-GI review is especially important before ablation, organ-removing surgery or when endoscopy and tissue disagree.
- 05
Validated tumour biomarker testing - Why
- Identify treatment-relevant biology in advanced disease.
- Interpretation and limitations
- Offer HER2 testing in advanced oesophago-gastric adenocarcinoma, including locally unresectable or metastatic disease. The specific trastuzumab combination with cisplatin plus capecitabine or fluorouracil is an option for untreated HER2-positive metastatic gastric or junctional adenocarcinoma. This regimen restriction does not narrow the broader testing indication; PD-L1 and MMR/MSI answer different treatment questions.
04Applied reasoningWorked examples connecting principles to decisions.
01Worked case: obstructing lesionObtain decisive tissue without forcing passageA 68-year-old with progressive dysphagia has a friable distal oesophageal lesion whose lumen will not admit the standard gastroscope without resistance.+
- 1Record the visible tumour from 34 to 39 cm, its circumferential narrowing and non-traversability, and decide against diagnostic dilation because staging can continue without forcing the scope.
- 2Obtain eight viable biopsies from different non-necrotic edges in one correctly labelled jar; pathology reports poorly differentiated adenocarcinoma with enough tumour for HER2 and other indicated biomarker testing.
- 3Staging CT identifies two liver lesions, and an image-guided liver biopsy confirms metastatic adenocarcinoma, making detailed local EUS staging unnecessary.
- 4Verify that pathology, HER2 testing, CT and nutritional assessment reach the MDT together; the documented outcome is a systemic-treatment consultation plus rapid dysphagia palliation rather than oesophagectomy.
02Worked case: linitis suspicionAct on discordanceCT and poor gastric distension suggest diffuse cancer but initial mucosal biopsies show inflammation only.+
- 1Review image quality, sampling depth and pathology with the specialist MDT.
- 2Repeat expert endoscopy with at least ten bite-on-bite samples from the stiffened gastric area and consider EUS-guided or image-guided tissue if mucosal yield remains negative.
- 3Do not label the stomach benign until the infiltrative imaging pattern has a credible explanation.
- 4Verify the final diagnosis and staging route in a documented MDT decision.
03Worked case: superficial lesionPreserve staging tissueA small well-demarcated mucosal lesion appears suitable for endoscopic treatment.+
- 1Use enhanced imaging to define the superficial lesion and synchronous abnormalities; take only one or two targeted biopsies before planned en-bloc resection so fibrosis does not compromise treatment.
- 2Choose en-bloc endoscopic resection rather than repeated piecemeal forceps destruction when criteria are met.
- 3Review the whole specimen for depth, margins, differentiation and lymphovascular invasion.
- 4Confirm whether resection is curative or whether surgery and nodal treatment are required.
05Checking understandingVerify the reasoning, revisit uncertainties and apply feedback.
- Track every biopsy specimen from endoscopy to final authorised pathology and MDT review.
- Document post-biopsy bleeding, pain, fever or respiratory deterioration promptly.
- Record whether sufficient tumour remains for biomarker testing before using tissue on nonessential studies.
- Reconcile endoscopic, radiological and pathological stage when any result conflicts.
- Audit reports for landmarks, dimensions, morphology, obstruction and image documentation.
06Special situationsVariants, exceptions and circumstances that change the usual approach.
Necrosis is poor tissue
The dramatic centre of an ulcer may contain no viable tumour; edges often provide better diagnostic material.
An intact specimen stages depth
Endoscopic resection supplies information that fragmented forceps biopsies cannot provide.
Failure to traverse is a finding
It quantifies obstruction and should guide safe alternative staging, not provoke force.
Jar labels preserve anatomy
Antrum, corpus, lesion and Barrett background lose meaning when combined.
07Common pitfallsFrequent interpretation and management errors.
- 01
Writing “lesion biopsied” without landmarks or size.
- 02
Dilating a suspected cancer simply to pass the scope.
- 03
Accepting one superficial benign biopsy despite rigid infiltrative imaging.
- 04
Consuming limited tumour tissue before essential biomarker testing.