When Pancreatic Cancer Comes Back: Recurrence Questions

Most people watching for recurrence are watching a number. It is worth knowing what NCI says about that number.

No tumor-specific markers exist for pancreatic cancer. CA 19-9 has low specificity. Most people have a raised level at diagnosis anyway. A rise during or after definitive treatment may point to growing disease. But a normal level does not rule recurrence out.

So a rising CA 19-9 is a prompt to look, not a diagnosis. The scan still decides. It also means a reassuring number is not proof of anything, which is worth saying to anyone using it as a monthly verdict on their own future.

This is the hardest sentence in NCI's summary, and it is worth quoting rather than softening. Patients with unresectable, metastatic or recurrent disease are unlikely to benefit from surgical resection.

Surgery is described as the mainstay of curative treatment, but only for small localized tumors, and only alongside drug treatment. NCI reports that complete removal can produce 5-year survival of 18 to 24 percent. It adds that control remains poor, because both local and distant recurrence are common.

That reframes the question. At recurrence the decision is almost always about which drugs, not which operation. See local recurrence and distant recurrence for why the distinction still matters for symptoms.

The regimen given after the original surgery shapes everything that follows.

NCI names gemcitabine plus capecitabine as the standard of care after resection. In the ESPAC-4 trial, median overall survival was 28.0 months with the pair and 25.5 months with gemcitabine alone. Estimated 5-year survival rose from 16.3 percent to 28.8 percent.

For people in good shape, adjuvant FOLFIRINOX is the other option. In PRODIGE-24, median disease-free survival was 21.4 months against 12.8 for gemcitabine. Median overall survival was 53.5 months against 35.5. Five-year survival was 43.2 percent against 31.4 percent.

The cost is on the same page. Grade 3 or 4 toxicity hit 75.9 percent of the FOLFIRINOX group, against 52.9 percent on gemcitabine. And 33 percent stopped treatment early, against 21 percent.

Nerve damage is the part that lingers. In the trial of gemcitabine with nab-paclitaxel, grade 3 neuropathy took a median of 29 days to fall back to grade 1 or clear. Of those affected, 44 percent restarted at a lower dose, after a median of 23 days. Anyone still carrying numbness should raise it before a platinum drug is proposed again. See peripheral neuropathy.

NCI names three standard first-line options, each with a trial behind it.

NALIRIFOX. Fluorouracil, irinotecan sucrosofate and oxaliplatin. In NAPOLI 3, with 770 untreated patients, median overall survival was 11.1 months against 9.2 for gemcitabine with nab-paclitaxel. Median progression-free survival was 7.4 months against 5.6. Peripheral neuropathy was less common, at 3 percent against 6 percent.

FOLFIRINOX. NCI calls it a standard treatment option for advanced disease. It is the same backbone as the adjuvant regimen above, which is why the timing of the earlier course matters so much.

Gemcitabine with nab-paclitaxel. Median overall survival was 8.5 months against 6.7 for gemcitabine alone. Median progression-free survival was 5.5 months against 3.7. Grade 3 neutropenia hit 38 percent, against 27 percent on gemcitabine alone.

One older combination is worth knowing about mainly to set expectations. Adding erlotinib to gemcitabine lengthened median survival from 5.9 months to 6.2. NCI describes that as a modest gain. It is a useful yardstick for how hard this disease has been to move.

The second-line evidence is thinner, and NCI is candid about it.

In NAPOLI-1, irinotecan sucrosofate with fluorouracil and leucovorin gave median overall survival of 6.1 months, against 4.2 months for fluorouracil and leucovorin. Irinotecan sucrosofate alone did not beat the control.

Two trials came out flat. PANCREOX added oxaliplatin after gemcitabine and found no gain: median progression-free survival of 3.1 months against 2.9. Grade 3 or 4 events were 63 percent against 11 percent. GEMPAX added paclitaxel to gemcitabine after FOLFIRINOX and found no survival gain either, at 6.4 months against 5.9. Progression-free survival and response rate did improve, and two subgroups did better: people aged 65 or under, and people with very high CA 19-9 at the start.

Germline BRCA1 and BRCA2. NCI reports these variants in 4 to 8 percent of pancreatic adenocarcinoma. Those tumors respond better to platinum drugs. Olaparib maintenance can be considered for carriers whose disease responded to first-line platinum treatment for more than four months. In the POLO trial, median progression-free survival was 7.4 months on olaparib against 3.8 on placebo. Overall survival was not different, at 18.9 months against 18.1. Grade 3 or 4 side effects were 40 percent against 23 percent.

DPYD. This gene encodes the enzyme that clears fluorouracil and capecitabine. NCI estimates 1 to 2 percent of people carry a harmful variant. Those with the DPYD*2A variant may have severe, life-threatening, sometimes fatal toxicity. Avoiding the drug or halving the dose may be advised, depending on genotype. NCI notes testing costs under $200, that insurance coverage varies, and that it can delay treatment by two weeks — which is why it calls the issue controversial rather than settled.

• Is the recurrence confirmed on imaging, or is CA 19-9 the only signal so far?
• Which adjuvant regimen did I have, and how long since it ended?
• How much neuropathy remains, and does it rule out oxaliplatin now?
• Was germline BRCA1 and BRCA2 testing ever sent?
• Should DPYD genotyping be done before another fluorouracil regimen?
• Is a trial open that would be a better first move than a repeat regimen?

For the wider picture, see pancreatic cancer and when cancer comes back.

• Call 911 or go to an emergency department if you have a fever with shaking chills together with yellow skin or eyes, or pain under the right ribs. An infected bile duct can turn into sepsis quickly, and recurrence near the bile duct or an old stent makes that more likely.
• Call 911 or go to an emergency department if you have new breathlessness, or chest pain that is worse when you breathe in or cough. Pancreatic cancer carries one of the highest clot risks of any cancer.
• Phone your oncology team the moment your temperature reaches 100.4°F (38°C) or higher on chemotherapy. CDC ranks this as a medical emergency rather than a same-day question, because infection can move quickly when counts are low. Call day or night. If no one answers soon, go to an emergency department and say you are on chemotherapy.
• Call your care team the same day if you have mouth sores, or severe diarrhea. On a fluorouracil regimen, severe early toxicity can be the first sign of a DPYD variant.
• Call your care team the same day if your skin or eyes yellow, your urine darkens, your stools go pale, or you cannot keep food or fluids down.
• Call your care team within a day or two if belly or back pain is climbing past what your current medicine covers, one leg swells or turns painful and warm, or numbness and tingling in your hands or feet worsen on oxaliplatin.

• https://www.cancer.gov/types/pancreatic/hp/pancreatic-treatment-pdq
• https://www.cancer.gov/about-cancer/treatment/side-effects/infection
• https://medlineplus.gov/ency/article/000290.htm
• https://www.cdc.gov/blood-clots/risk-factors/cancer.html
• https://www.cdc.gov/cancer-preventing-infections/patients/fever.html