Efficacy and safety of cisplatin and gemcitabine (CG) chemotherapy for advanced biliary tract cancer (ABC) in jaundiced patients (pts).

Abstract

294 Background: The ABC-02 study established cisplatin/gemcitabine (CG) as a 1st-line regimen for pts with advanced/metastatic biliary tract cancer (ABC). Biliary tract obstruction (BTO) is common in ABC and pts with bilirubin (bili) ≥1.5xULN were excluded from ABC-02. We assessed retrospectively the safety/efficacy of CG in ABC pts with high bili despite optimal stenting. Methods: Eligible pts had biopsy-proven ABC; received 1st-line CG; baseline bili levels ≥1.5xULN and had follow-up, response and toxicity data available. Survival analysis (Kaplan-Meier), long-rank test, Cox regression and multiple linear regression analyses were performed. Results: Thirty-three pts (of 545 pts screened) met inclusion criteria: median age 58.8 years (range 23-79); male 58%; cholangiocarcinoma 76%; gallbladder 18% or ampullary cancer 6%; locally advanced 42%; metastatic 58% (79% in liver); 68% had a biliary stent; and 97% of pts had good baseline performance status (PS) 0/1/2: 33%/64%/3%. Median baseline bili was 55umol/l (range 32-286); due to BTO in 76% or liver metastases (LM) in 24%. A median of 6 (range 1-8) cycles were given, 70% of pts started full dose CG; 40% completed 8 planned cycles; 40% needed a dose reduction. Early discontinuation (61% of all pts) was due to toxicity (25%), clinical (60%) or radiological progression (15%). Bili normalised in 64% of pts after CG; no unexpected side effects occurred. Most pts achieved stable disease (61%); with one partial response (3%); the median PFS was 6.9 mo (95%-CI: 4.4-9.0) and median OS 9.5 mo (95%-CI: 5.7-12.8). BTO pts had a trend for longer PFS than LM (7.0 vs. 2.6 mo; p=0.1633). Normalisation of bili and completion of CG were associated with longer OS (11.4 vs. 2.9 mo, HR 0.49; p=0.08 and 15.2 vs. 5.4 mo, HR 0.12 p<0.001, respectively). Completion of CG was an independent prognostic factor for OS (multivariate analysis, p=0.001); no difference in OS was shown between the percentiles of baseline bili (p=0.57). Conclusions: For fit ABC patients with high bili, CG can be employed safely with encouraging results. Cause of jaundice (BTO, rather than bilirubin level), completion of chemotherapy and normalisation of bili are associated with improved OS.