Gallbladder Cancer: Adjuvant Capecitabine After an Incidental Diagnosis
A cancer found only after the gallbladder was already out raises a genuinely uncomfortable question: does the trial behind adjuvant chemotherapy here actually prove what the recommendation implies it proves?
D.O., a 67-year-old retired postal worker, had his gallbladder removed three weeks ago for what everyone, including him, assumed was routine symptomatic cholelithiasis — recurring right-upper- quadrant pain after fatty meals, an ultrasound full of stones, a same-day laparoscopic cholecystectomy he expected to recover from in a week. The pathology report changed that: a 2.8cm adenocarcinoma in the gallbladder wall, invading the perimuscular connective tissue without reaching the serosa, staged T2N0 after a completion resection confirmed clear margins and no nodal disease. He has moderate chronic kidney disease from longstanding hypertension — creatinine 1.6, estimated CrCl 42 — stable for years and never previously relevant to anything.
Adjuvant capecitabine after resected biliary tract cancer, including gallbladder cancer, rests on BILCAP (Primrose et al., Lancet Oncology 2019), and the trial is genuinely more equivocal than the practice it generated: the intention-to-treat analysis — the pre-specified primary comparison — showed a median overall survival advantage that did not reach statistical significance (51 vs 36 months, p=0.097). What did reach significance, in a pre-planned sensitivity analysis adjusting for known prognostic factors (nodal status, grade, resection margin), was a real overall survival benefit, and progression-free survival was significantly improved in the intention-to-treat population as well. Most guideline bodies now recommend capecitabine on the strength of that adjusted analysis and the PFS signal, not the primary ITT overall-survival comparison — a distinction worth stating plainly to a patient making his own decision, particularly one whose renal function will need the drug dose-adjusted before he even starts. Capecitabine itself is not renally cleared — it is a prodrug, and it is the downstream catabolite FBAL — alpha-fluoro-beta-alanine — that accumulates as clearance falls, which is why the label calls for a 25% starting-dose reduction below a creatinine clearance of 51 and contraindicates the drug outright below 30 — D.O.'s CrCl of 42 sits squarely inside the reduced-dose range, not the exclusion range, but it means his own numbers will need to be part of the conversation about benefit and risk, not an afterthought handled by the pharmacy after he's already decided.
Post-operative oncology clinic, discussing adjuvant therapy
I'd recommend adjuvant capecitabine here. BILCAP is the only randomized trial we have for resected biliary tract cancer, and it showed a significant progression-free survival benefit in the intention-to-treat population, with the pre-planned analysis adjusting for nodal status, grade, and margin status also showing a significant overall survival benefit. For a disease this short on evidence, that's a real, guideline-supported reason to treat a fit patient like D.O.
You're right that the adjusted analysis and the PFS result are real findings, and I'm not arguing against treating him. But BILCAP's own pre-specified primary endpoint — intention-to- treat overall survival — came in at p=0.097, and I don't think it's fair to D.O. to describe this as unambiguous "proven survival benefit" without naming that the headline comparison technically missed significance. His CrCl is 42, which means starting dose needs real reduction, and he deserves to weigh a genuinely meaningful but not iron-clad benefit against real hand-foot syndrome and diarrhea risk on a drug his kidneys will clear more slowly.
None of this changes the dosing math — renal adjustment is required regardless of which trial result we lead with when we talk to him.
Listening to both of you, I don't think this is really a disagreement about whether to treat — it's about how honestly we describe the evidence while doing it. Say both numbers to him directly: the primary intention-to-treat comparison didn't reach significance, and the adjusted analysis accounting for his own actual risk factors did, alongside a real PFS benefit. He's fit, recovered, and this is his decision to make once he has the real distinction in front of him — not ours to resolve by choosing which number to lead with.
Agreed: adjuvant capecitabine at a renally-adjusted starting dose, with dose reassessment at cycle one based on tolerability, and both the ITT and adjusted BILCAP results explained to D.O. directly before he consented.
The team's own internal disagreement resolved once framing was separated from the treatment decision — all three voices ultimately supported treatment, with the pharmacologist's insistence on disclosing the ITT result's own p-value treated not as a reason to withhold therapy but as part of what informed consent actually requires here.