Gemcitabine with or without cisplatin in patients (pts) with advanced or metastatic biliary tract cancer (ABC): results of a multicentre, randomised phase III trial (the UK ABC-02 trial)


Session type:

Juan Valle1, Harpreet S Wasan2, Daniel D Palmer3, David Cunningham4, Allan Anthoney5, Anthony Maraveyas6, Sharon Hughes7, Michael Roughton7, John Bridgewater7

1Christie Hospital NHS Trust, Manchester, UK, 2Imperial College Healthcare Trust, London, UK, 3University of Birmingham, UK, 4Royal Marsden Hospital, London, UK, 5St. James' University Hospital, Leeds, UK, 6Castle Hill Hospital, Hull, UK, 7University College London, UK


Proffered paper presentation

There is no established standard chemotherapy for pts with inoperable ABC. We previously reported an improvement in progression-free survival (PFS) in a randomised phase II trial of 86 pts (ABC-01) using gemcitabine/cisplatin (GemCis) vs. gemcitabine (Gem) (Valle ASCO-GI 2006, abstr. 98). This study was extended into ABC-02, a phase III trial, to recruit a further 314 pts with overall survival (OS) as the primary end-point.

Consenting pts with histologically/cytologically-confirmed ABC, aged ≥18 years, ECOG performance status 0 - 2, and adequate haematological, hepatic and renal function were randomised to receive either Cis (25 mg/m2) followed by Gem (1000 mg/m2 D1, 8 q21d) for 8 cycles, or Gem alone (1000 mg/m2 on D1, 8, 15 q28d) for 6 cycles, stratified by extent of disease, site of primary tumour, ECOG score and centre. The trial had an 80% power to detect an OS hazard ratio of 0.73.

From May 2005 to October 2008, 324 pts were randomised to ABC-02 from 34 UK centres. We report the pre-planned combined analysis of ABC-01 and ABC-02 based on 410 pts (GemCis=206/Gem=204). Patient characteristics: median age 64 yrs (range 23-85); male (47%); metastatic disease (75%), locally advanced (25%); gallbladder (36%), bile duct (59%), ampulla (5%); and ECOG 0-1 (87%), 2 (12%). With a median follow-up of 6.1 months and 263 deaths, the median OS was greater with GemCis than Gem, 11.7 vs. 8.2 months (log rank p=0.002), with hazard ratio 0.68 (95%-CI 0.53, 0.86). The median PFS was greater with GemCis than Gem, 8.5 vs. 6.5 months (log rank p=0.003), with hazard ratio 0.70 (95%-CI 0.56, 0.88). Toxicity was similar between the arms (by week 12, 57% had a grade 3/4 toxicity in each arm), though there was a slight excess of neutropenia using GemCis.

This study demonstrates a clear survival advantage for GemCis in ABC without added clinically significant toxicity. We will present updated and final survival, CA19.9 and QoL data.