CA 19-9 in pancreatic cancer — what the number can and cannot tell you

Most people arrive here having typed something close to “ca 19 9 pancreatic cancer” into a search box, usually with a printed result in the other hand and a figure on it that means nothing to them yet. So, plainly: CA 19-9 is a carbohydrate antigen — a sugar-and-protein structure carried on the surface of the cells that line the pancreas and the bile ducts. A little of it leaks into the bloodstream normally. When those cells are inflamed, obstructed or malignant, more of it reaches the blood, and the test measures how much.

That is the entire mechanism, and it explains both why the test is useful and why it disappoints people. It reports on a tissue, not on a diagnosis. Anything that irritates or blocks the pancreatic and biliary lining pushes the level up. A tumour does that. So does a gallstone wedged in the bile duct, an attack of pancreatitis, or a liver that is scarred. The blood test cannot tell you which of those it is looking at, and no laboratory report will claim otherwise.

Searches for “ca19-9 levels” and for the ca 19 9 tumour marker almost always come down to a single question: is this figure bad? The honest answer is that a level only becomes readable once you know three things — whether you were jaundiced when the blood was taken, what the value was last time, and what the scan shows. Take any one of those away and the number on its own says very little.

Used properly, CA 19-9 is a tracking tool rather than a detector. A baseline is established before treatment starts, then repeated at planned intervals so the direction of travel can be seen. Read next to imaging, an examination and how you actually feel, that direction is genuinely informative. Read alone, on a single day, it is one of the easiest results in oncology to misread. It also never travels by itself on the request form; the blood and liver function tests used in pancreatic cancer sit on the same report and change how the marker is interpreted.

1. 1

   The free consultation comes first

   A 45-minute consultation goes through your symptoms, your scans and every blood result you already hold before anything new is ordered. Nothing is repeated simply because it sits on a standard list.

   In-house at CION
2. 2

   Jaundice is settled before a baseline is taken

   If the bile duct is blocked, the marker is not yet worth measuring. Drainage with an endoscopic stent is arranged first and is carried out by our specialist endoscopy partners at their centre, where it may also be billed. The baseline is drawn once the bilirubin has fallen.

   Stenting coordinated with endoscopy partners
3. 3

   A baseline goes off with the rest of the bloods

   CA 19-9 is drawn alongside bilirubin, liver function, blood counts and nutritional markers, all ordered and reported by CION across 35+ centres, so the marker is never read in isolation.

   Ordered and reported by CION
4. 4

   The trend is followed, not the reading

   Repeat tests are timed to the treatment plan and run at the same laboratory on the same assay wherever possible, so successive values can be compared with each other rather than argued about.

   Ordered and reported by CION
5. 5

   It is read next to the scan, never instead of it

   The tumour board weighs the marker trend against imaging response, weight, appetite and how you actually feel. Where the marker and the scan disagree, the scan and the clinical picture decide the plan.

   In-house at CION
6. 6

   It carries on into follow-up

   After treatment the marker joins the surveillance schedule set out in follow-up and surveillance after pancreatic cancer, where a steady rise is a reason to bring a scan forward rather than a finding in itself.

   In-house at CION

Different laboratories use different assays, and their reference ranges are not interchangeable. A value from one hospital and a value from another are not strictly comparable, even when both look like ordinary figures on ordinary paper. This is the commonest reason people believe their marker has jumped when in fact only the laboratory has changed. Wherever the plan allows it, repeat tests are sent to the same laboratory so that like is compared with like.

A one-off rise is also not the same thing as progression. An episode of cholangitis, a stent that has silted up, a flare of pancreatitis or a bout of biliary infection can each lift a marker sharply and temporarily. The right response to a single unexpected jump is usually a check for one of those causes and a repeat test, not a change of treatment. Equally, a marker that falls does not mean the cancer has gone. It means the tissue producing the antigen is producing less of it, which is encouraging, and encouraging is not the same as cured.

Some people never make CA 19-9 in measurable quantities at all, and in them the test has no role. Their disease is followed on imaging, symptoms and examination instead, which is a perfectly adequate way to run a treatment plan — the marker is a convenience, not a requirement. If your team has stopped ordering it, that is usually why, and it is worth asking so you are not left wondering. How the tests, the stage and the plan fit together as a whole is set out in the complete pancreatic cancer guide.

Finally, a figure on a page is not a prognosis. Markers are one input among several, and none of them describes an individual person's outcome. What changes an outcome is the decision that follows — whether the disease can be removed, whether systemic therapy can shrink it into operability, and how well you are supported through the months that follow. That decision is made at a tumour board, with your scans in front of it, not by a blood test.

• Was I jaundiced when this blood was taken? If the bile duct was obstructed, the value describes the blockage as much as anything else, and the repeat taken after drainage is the one that counts.
• What was the previous value, and when? A marker with no date and no predecessor is a single dot. Two dots make a line, and the line is what your team is actually reading.
• Was it the same laboratory and the same assay? If not, the comparison is shakier than it looks, and the safest step is to repeat both at one laboratory before drawing any conclusion.
• What do my other bloods say? Bilirubin, liver enzymes, albumin and blood counts all change how the marker is read, and the blood and liver function tests in pancreatic cancer explains what each of them adds.
• Am I someone who makes this marker at all? A persistently low level in the face of obvious disease is a recognised pattern, covered in can you have pancreatic cancer with a normal CA 19-9.
• What happens next if it rises? Ask for the plan in advance — usually a repeat test, a check for infection or a blocked stent, and a scan brought forward, exactly as described in follow-up and surveillance.

Medical disclaimer: This page explains what the CA 19-9 tumour marker measures and how it is interpreted in pancreatic cancer, and is reviewed by a CION medical oncologist with reference to NCCN guidance on pancreatic adenocarcinoma. It is general information and deliberately states no reference range, cut-off value or accuracy figure, because a marker result is only meaningful alongside your own bilirubin, your previous values, your imaging and your symptoms; your own result should be discussed with your treating team. CA 19-9, bilirubin, liver function tests and routine bloods, pancreatic-protocol CT and MRI/MRCP, tumour-board review, chemotherapy, radiation, chemoradiation and SBRT, genetic counselling, nutrition and pancreatic enzyme (PERT) support, pain, psycho-oncology and survivorship care are delivered by CION. ERCP and biliary or duodenal stenting, endoscopic ultrasound and biopsy, staging laparoscopy, coeliac plexus block, all pancreatic surgery, PET-CT and DOTATATE PET, and peptide receptor radionuclide therapy are coordinated with specialist hepatobiliary, gastroenterology, endoscopy and nuclear medicine partner centres and may be billed there.