Adjuvant CAPOX in a Patient on Warfarin: A Dangerous INR Rise
64-year-old woman with pT3N1 sigmoid adenocarcinoma status post hemicolectomy presents on day 10 of adjuvant CAPOX. She has a mechanical mitral valve and takes warfarin 5 mg daily; baseline INR was 2.6. Other history includes type 2 diabetes and an eGFR of 52 mL/min/1.73 m². She reports three days of watery diarrhea (four to five stools/day), early hand-foot erythema, and no bleeding. Today's INR is 6.8, hemoglobin 10.9 g/dL, platelets 190,000/µL, and she is hemodynamically stable. She has no epistaxis, melena, or hematuria and has taken no new medicines apart from acetaminophen. She asks whether she can continue chemotherapy and why her warfarin dose suddenly became dangerous. The oncology fellow considers holding capecitabine, giving vitamin K, and switching to FOLFOX but is unsure which steps are urgent.
1. Explain the mechanism of the elevated INR and outline your immediate management, including whether she needs vitamin K, fresh frozen plasma, or platelet transfusion.
Read the worked answer
The rise is best explained by a pharmacokinetic interaction plus reduced vitamin K intake/absorption. Capecitabine is converted to 5-fluorouracil, which inhibits CYP2C9, the main enzyme metabolizing the more potent S-warfarin enantiomer. The INR can rise within days of starting capecitabine. Diarrhea adds to the problem by reducing vitamin K absorption and causing dehydration.
Immediate steps:
- Hold warfarin and hold capecitabine. Assess for bleeding, including fundoscopy and neurologic symptoms.
- She has an INR of 6.8 but no bleeding. For INR 4.5–10 without bleeding, guidelines recommend withholding warfarin and monitoring, not routine vitamin K. Do not give fresh frozen plasma or platelets; platelets are normal and there is no bleeding.
- Recheck INR in 24 hours, then every 1–2 days until it is near the therapeutic range. If she develops major bleeding, give vitamin K 5–10 mg IV plus prothrombin complex concentrate (or FFP if PCC unavailable).
- Treat diarrhea with oral rehydration and loperamide; check electrolytes and renal function.
- Involve cardiology because interrupting anticoagulation for a mechanical mitral valve carries thrombotic risk. Once the INR is near baseline, restart warfarin at a reduced dose with frequent INR checks, or use therapeutic LMWH bridging if warfarin cannot be safely managed.
2. After resolution of the acute problem, how would you modify the adjuvant regimen and monitoring? Address renal function, capecitabine dosing, and whether to switch to FOLFOX.
Read the worked answer
Her eGFR of 52 mL/min/1.73 m² is just above the threshold for routine capecitabine reduction; the label recommends a 75% starting dose only for CrCl 30–49 mL/min. I would not reduce the dose solely for eGFR 52. However, she has already had grade 2 diarrhea and grade 1 hand-foot syndrome, so toxicity-directed adjustment is needed.
Practical plan:
- Hold capecitabine until diarrhea is grade ≤1 and hand-foot syndrome has settled. Resume at the same dose if toxicity was grade 1–2 and resolves quickly; reduce by one dose level (for example, from 1000 to 750 mg/m² twice daily) if grade 2 toxicity recurs.
- Continue oxaliplatin unless neurotoxicity or other toxicity intervenes; stage III colon cancer benefits from the oxaliplatin backbone.
- Switching to FOLFOX does not remove the warfarin interaction because 5-fluorouracil also inhibits CYP2C9. It adds infusional logistics and oxaliplatin neuropathy without solving the interaction.
- Coordinate with anticoagulation: check INR twice weekly during the first cycle after restarting, then weekly; educate her to report diarrhea, bruising, or bleeding. Avoid trimethoprim, metronidazole, and fluconazole unless necessary because they add to the interaction.
3. What is the role of DPYD testing in this patient, and how would a heterozygous deleterious variant change your plan?
Read the worked answer
Dihydropyrimidine dehydrogenase (DPYD) deficiency increases the risk of severe fluoropyrimidine toxicity, including diarrhea, mucositis, myelosuppression, and hand-foot syndrome. Testing is recommended before starting a fluoropyrimidine when feasible, although it is not universally mandated and cannot replace clinical monitoring.
In this patient, testing is still useful because she developed early toxicity. However, her presentation is confounded by the warfarin interaction and diarrhea, so a normal DPYD result would not prove that capecitabine was safe at full dose.
If she is heterozygous for a deleterious variant such as DPYD*2A, I would reduce the capecitabine starting dose by 25–50% according to the variant and institutional/pharmacogenomic guidance, then titrate upward based on tolerance. If she is a compound heterozygote or homozygote with absent enzyme activity, I would avoid capecitabine and 5-fluorouracil entirely and use an oxaliplatin-based non-fluoropyrimidine regimen, accepting that evidence is weaker. A normal result would support re-exposure at full dose once the acute issues resolve, with close monitoring during the first cycle.