A Mass in the Pancreaticoduodenal Groove: Medical Trial First, or Straight to Resection?
A groove mass that looks inflammatory on every clinical clue available still can't be told apart from cancer without tissue — the disagreement is whether a medical trial is a reasonable first step, or a reasonable delay of the one test that would actually settle it.
Victor D. has spent most weekends for the last decade restoring an old sailboat docked a short drive from his house, a project he's been slower to get to lately because of a dull, gnawing epigastric discomfort and intermittent vomiting that started about two months ago. His history includes heavy alcohol use for over twenty years, ongoing rather than resolved, along with heartburn he's self-treated with over-the-counter antacids for years without ever mentioning it to a doctor. CT and MRCP both show a mass centered specifically in the groove between the pancreatic head, duodenum, and common bile duct — not the pancreatic parenchyma itself — with cystic changes in the duodenal wall and thickening consistent with chronic inflammation rather than a discrete, solid tumor mass.
Everything about his clinical picture points toward groove pancreatitis: the anatomic location, the alcohol history, the cystic duodenal-wall changes rather than a clean mass margin, and the absence of the biliary obstruction or weight loss that more often accompanies pancreatic head cancer. But groove pancreatitis and pancreatic head adenocarcinoma occupy genuinely overlapping territory on imaging — Triantopoulou and colleagues' review (European Radiology, 2009) found the distinction extraordinarily difficult even prospectively, which is why most cases like this end up resected rather than confidently called — and the diagnosis that actually matters most to get right the first time is the one where a delay has the highest cost. He has not yet had tissue sampling — every read so far is radiographic and clinical, not histologic.
He admitted, when asked directly, that he'd been putting off seeing anyone for the vomiting because he assumed it was "just his stomach acting up again," the same explanation he'd given himself for the heartburn he never mentioned to a physician either. It was his son, visiting for a weekend of boat work and noticing how little his father had eaten over two days, who finally pushed him to schedule the appointment that led to this imaging. His weight has been stable so far despite the vomiting, and he has no family history of pancreatic or gastrointestinal cancer his team is aware of — neither of which reassures anyone in the room as much as it sounds like it should, since most pancreatic head cancers arise with no family history at all, and two months is early enough that stable weight is a statement about timing rather than about biology.
Multidisciplinary review
His overall picture strongly favors groove pancreatitis — the anatomic location, the alcohol history, cystic duodenal-wall change rather than a solid mass, and no biliary obstruction or weight loss. A bounded PPI-and-steroid trial with imaging reassessment in six weeks is reasonable here, and a mass that resolves on that trial would spare him a Whipple procedure's real morbidity for a condition that was never malignant to begin with.
I understand the clinical pattern favors an inflammatory read, but groove pancreatitis and pancreatic head cancer are genuinely, documentedly hard to tell apart on imaging alone — that overlap is exactly the problem, not a detail this case happens to avoid. Treating a mass presumptively as inflammatory without tissue confirmation risks losing time in exactly the window when a cancer, if that's what this turns out to be, is most resectable.
I'd get EUS-FNA tissue sampling before starting any trial, not after. If it comes back genuinely inflammatory, the medical trial becomes a well-supported next step rather than a bet made on imaging pattern alone.
Agreed: EUS-guided tissue sampling first; the PPI-and-steroid trial proceeds only if biopsy confirms a benign inflammatory process.
Not agreed: what to do if the biopsy is non-diagnostic rather than clearly benign or malignant — the surgical oncologist would move toward resection given the diagnostic stakes; the gastroenterologist would prefer a repeat, deeper sampling attempt before escalating to surgery on an inconclusive result.