Upper GI and Pancreatic Cancer: Gastric, Oesophageal, Liver and Pancreas
Four cancers of the upper gut share a single chemotherapy spine — platinum plus a fluoropyrimidine — yet each hides its own twist: a HER2 test that unlocks a smart antibody in stomach cancer, a brutal three-drug regimen for the pancreas, and an immunotherapy-plus-antibody pair that, after a decade of stalemate, finally moved the needle in liver cancer. Learn the spine once, then learn the twists.
Picture a tumour board on a Tuesday morning: three patients, three difficult cancers of the upper gut, one shared slide. The oncologist points to a common spine that runs through all of them — a platinum drug paired with a fluoropyrimidine. Then she flips to what makes each unique. For the stomach cancer, a single lab test — HER2 — decides whether a designer antibody joins the fight. For the pancreas, there is no clever antibody; just a hard, three-drug chemotherapy regimen and a search for a rare BRCA mutation. For the liver — riding on years of hepatitis and cirrhosis — the answer that finally worked was not chemo at all, but immunotherapy paired with an anti-angiogenic antibody. Same neighbourhood, three very different battle plans.
The shared spine: platinum + a fluoropyrimidine
Start with what these tumours have in common. The chemotherapy backbone for gastric and oesophageal cancer — and, in modified form, for the pancreas — pairs a platinum agent (oxaliplatin or cisplatin) with a fluoropyrimidine (5-fluorouracil, 5-FU, or its oral prodrug capecitabine). The platinum drug crosslinks DNA so the strands can no longer separate to copy themselves; the fluoropyrimidine starves the cell of the building blocks (thymidine) it needs to make DNA. Two hits on the same pathway — replication — from two directions. The fluoropyrimidines belong to the antimetabolites, and the platinums to the DNA-crosslinking alkylating-like agents; both are covered in depth in their own class chapters.
Whenever you see "FOLFOX" or "CapeOx" in an upper-GI regimen, read it as the spine in disguise: FOL = folinic acid (leucovorin, which boosts 5-FU), F = 5-FU, OX = oxaliplatin. "Cape" simply swaps IV 5-FU for oral capecitabine. Learn the four letters and half the regimens across the whole GI tract stop looking intimidating.
Gastric & oesophageal cancer: test before you treat
Over the platinum-plus-fluoropyrimidine spine, gastric and oesophageal (gastro-oesophageal) cancer add targeted layers — but only after biomarker testing tells you the tumour will respond. Three tests matter. First, HER2 (human epidermal growth factor receptor 2): if the tumour over-expresses it, adding trastuzumab — a monoclonal antibody that locks onto HER2 — improves survival. Second, PD-L1 and MSI/MMR status: a tumour that displays PD-L1 or is mismatch-repair-deficient (MSI-high) responds to checkpoint immunotherapy (pembrolizumab or nivolumab), which releases the brakes on the patient's own T-cells. Third, in later lines, ramucirumab — an antibody against VEGFR2 — starves the tumour of new blood vessels. Trastuzumab, ramucirumab and the anti-angiogenics live in the monoclonal-antibodies chapter; the checkpoint inhibitors have their own chapter.
Two patients arrive with the same advanced gastric cancer and both start FOLFOX. The pathology comes back: patient A is HER2-positive, so trastuzumab is added to the chemotherapy; patient B is HER2-negative but PD-L1-high, so a checkpoint inhibitor is added instead. Identical starting chemo, two different add-ons — and the deciding factor was a stain on a slide, not the size of the tumour.
One more thread ties gastric cancer to the antimicrobials story: Helicobacter pylori, the stomach bacterium behind most peptic ulcers, is a recognised cause of gastric cancer. Chronic H. pylori infection drives inflammation, atrophy and, over years, malignant change — which is why eradicating it (the triple/quadruple regimens covered in the antimicrobials and GI sections) is not only ulcer treatment but a genuine cancer-prevention measure.
- Backbone: a platinum (oxaliplatin/cisplatin) + a fluoropyrimidine (5-FU/capecitabine).
- HER2-positive → add trastuzumab (a monoclonal antibody against HER2).
- PD-L1-high or MSI-high → add checkpoint immunotherapy (pembrolizumab/nivolumab).
- Later lines: ramucirumab (anti-VEGFR2) blocks tumour angiogenesis.
- H. pylori causes gastric cancer — eradication is prevention (see Antimicrobials/GI).
Pancreatic cancer: hard chemistry, few targets
The pancreas is the hardest room in this house. Pancreatic ductal adenocarcinoma is notoriously aggressive and, because early symptoms are vague, it is usually diagnosed late — often after it has already spread. There is no HER2-style switch and no reliable immunotherapy target for most patients; treatment stays largely chemotherapy-driven. Two combinations dominate. FOLFIRINOX stacks four agents — 5-FU (plus leucovorin), oxaliplatin and irinotecan — a potent but toxic regimen reserved for fitter patients. The gentler alternative is gemcitabine plus nab-paclitaxel (albumin-bound paclitaxel). Gemcitabine, like 5-FU, is an antimetabolite; oxaliplatin is a platinum; irinotecan is a topoisomerase inhibitor — each detailed in its own class chapter.
There is one bright targeted exception. A minority of pancreatic cancers carry an inherited BRCA1/2 mutation, leaving the tumour unable to repair DNA the usual way. In these patients, after platinum-based chemotherapy has controlled the disease, a PARP inhibitor (olaparib) given as maintenance blocks the cell's back-up repair route — the two defects together are lethal to the cancer cell (synthetic lethality). It only works if you looked for the BRCA mutation, which is exactly why guidelines urge germline testing in pancreatic cancer.
Pancreatic cancer is as much a supportive-care problem as a chemotherapy one. A tumour in the pancreatic head can block the bile duct, causing jaundice that may need a stent; loss of exocrine pancreas function leaves patients unable to digest fat, so pancreatic enzyme replacement (with meals) treats the resulting weight loss and steatorrhoea. Managing pain, nutrition and clots is not an afterthought here — it is central to keeping the patient well enough to receive any chemotherapy at all.
- Ductal adenocarcinoma is aggressive and usually caught late; treatment is chemo-led.
- FOLFIRINOX (5-FU/oxaliplatin/irinotecan) for fit patients; gemcitabine + nab-paclitaxel otherwise.
- BRCA-mutant tumours → olaparib (a PARP inhibitor) as maintenance after platinum.
- Test the germline for BRCA — you can't offer PARP maintenance you never looked for.
- Supportive care is central: biliary stenting, pancreatic enzymes, nutrition, pain, clots.
Hepatocellular carcinoma: immunotherapy moves the needle
Liver cancer (hepatocellular carcinoma, HCC) almost always grows on a diseased background — years of cirrhosis, usually from chronic hepatitis B or C, or from fatty liver and alcohol. That underlying liver damage limits how much chemotherapy the organ can tolerate, and classic cytotoxic chemo works poorly here. For a decade the only systemic option was a single multikinase inhibitor, sorafenib, with modest benefit. The breakthrough came from immunotherapy: the combination of atezolizumab (a checkpoint inhibitor) plus bevacizumab (an anti-VEGF antibody) is now first-line for advanced HCC and outperformed sorafenib. Where immunotherapy is unsuitable, multikinase inhibitors — sorafenib or lenvatinib — remain the alternative. These kinase inhibitors are covered in the kinase-inhibitors chapter, the checkpoint drugs in the checkpoint-inhibitors chapter, and bevacizumab among the monoclonal antibodies.
Not every liver tumour goes straight to systemic drugs. When disease is confined to the liver but too extensive for surgery, locoregional therapies treat it directly — most commonly transarterial chemoembolization (TACE), which threads a catheter into the artery feeding the tumour and delivers chemotherapy while blocking its blood supply. And because so much HCC is preventable, the most powerful drug here may be one given years earlier: the hepatitis B vaccine, and antiviral treatment of chronic hepatitis B and C, dramatically cut the risk of ever developing liver cancer — the prevention thread that runs back to the antimicrobials section.
- HCC grows on cirrhosis — usually hepatitis B/C — which limits chemotherapy tolerance.
- First-line systemic therapy: atezolizumab (checkpoint) + bevacizumab (anti-VEGF).
- Alternatives: multikinase inhibitors sorafenib or lenvatinib.
- Liver-confined disease may get locoregional therapy such as TACE.
- Prevention is potent: hepatitis B vaccination and antiviral therapy cut HCC risk.
- Skipping HER2 and PD-L1/MSI testing in gastro-oesophageal cancer — you miss the patients who'd benefit from trastuzumab or immunotherapy.
- Not sending BRCA/germline testing in pancreatic cancer — without it you never identify the patients eligible for olaparib maintenance.
- Ignoring the hepatitis background in liver cancer — the underlying cirrhosis shapes treatment tolerance, and its prevention (vaccine/antivirals) is part of the plan.
- Treating "upper-GI chemo" as one thing — the shared spine is real, but the twist (antibody, three-drug regimen, or immunotherapy pair) is what changes the outcome.
A patient with advanced gastric cancer is starting a platinum + fluoropyrimidine backbone. Which single result would most directly justify ADDING trastuzumab?
- A platinum + fluoropyrimidine spine runs through gastric, oesophageal and (modified) pancreatic chemo.
- Gastro-oesophageal: test HER2 (→ trastuzumab), PD-L1/MSI (→ immunotherapy); ramucirumab later. H. pylori is a cause.
- Pancreas: FOLFIRINOX or gemcitabine + nab-paclitaxel; olaparib maintenance if BRCA-mutant; few other targets.
- Liver (HCC): atezolizumab + bevacizumab first-line; sorafenib/lenvatinib alternatives; TACE locally; hepatitis B vaccine/antivirals prevent.
- NCCN Clinical Practice Guidelines in Oncology — Gastric Cancer; Esophageal and Esophagogastric Junction Cancers.
- NCCN Clinical Practice Guidelines in Oncology — Pancreatic Adenocarcinoma; Hepatobiliary Cancers.
- ESMO Clinical Practice Guidelines — Gastric Cancer; Oesophageal Cancer; Pancreatic Cancer; Hepatocellular Carcinoma.
- DeVita, Hellman & Rosenberg's Cancer: Principles & Practice of Oncology — Cancers of the upper gastrointestinal tract & pancreas.
- Brunton LL, et al. Goodman & Gilman's The Pharmacological Basis of Therapeutics — Antineoplastic agents: antimetabolites, platinum compounds, targeted antibodies & checkpoint inhibitors.
