ESPAC-4: An international randomized phase-3 trial of adjuvant combination chemotherapy using gemcitabine (GEM) and capecitabine (CAP) versus mono therapy gemcitabine in resected pancreatic cancer


Session type:

John Neoptolemos1,Daniel Palmer2,Paula Ghaneh2,Juan Valle3,David Cunningham4,Jon Wadsley5,Tim Meyer6,Alan Anthoney7,Bengt Glimelius8,Stephen Falk9,Pehr Lind10,Jakob Izbicki11,Gary Middleton12,Paul Ross13,Harpreet Wasan14,Alec McDonald15,Tom Crosby16,Eftychia Psarelli2,Pascal Hammel17,Markus Buchler18
1Cancer Research UK Liverpool Cancer Trials Unit,2University of Liverpool,3The Christie Hospital,4Royal Marsden Hospital,5Weston Park Hospital,6Royal Free Hospital,7St James's University Hospital,8University of Uppsala,9Bristol Haematology and Oncology Centre,10Karolinska University Hospital,11University of Hamburg Medical Institutions UKE,12Royal Surrey County Hospital,13Guy's Hospital, London,14Hammersmith Hospital,15The Beatson West of Scotland Cancer Centre,16Velindre Hospital,17Hopital Beaujon, Clichy,18University of Heidelberg



The ESPAC-3 trial compared adjuvant GEM with 5-fluorouracil/folinic acid for resected pancreatic cancer. GEM is the standard of care based on similar survival and less toxicity. ESPAC-4 aimed to determine whether combination chemotherapy with GEM/CAP improved survival compared to GEM monotherapy.


Patients with pancreatic cancer were randomized to have either six 4wk cycles of IV GEM alone or GEM with oral CAP. The primary endpoint was overall survival; secondary endpoints were toxicity, relapse free survival, 2 and 5 year survival and quality of life. The trial was designed to detect a hazard ratio of 0.74. Using a two-sided alpha level of 0.05, 480 events were required to obtain 90% power.


Between 732 patients were randomized with 730 included in the full analysis set (366 GEM, 364 GEM/CAP). Median age was 65 years, 57% were men. WHO performance status was 0, 1 or 2 in 42% 55% and 3% respectively. Postoperative median CA19-9 was 19 kU/L. Median maximum tumor size was 30 mm, 60% were R1 resections, 80% were node positive and 40% were poorly differentiated. On Dec 11 2015 the Independent Trial Steering Committee requested that the trial proceed to full analysis. Stratified log-rank for overall survival produced an HR=0.82 [95% CI, 0.68 – 0.98]; χ2 (1) = 4.61, P=0.032. Median survival (months) for patients treated with GEM/CAP was 28.0 (95% CI, 23.5 – 31.5) and 25.5 (22.7 – 27.9) for GEM. The 5 year survival rate for the GEM/CAP was 28.8 (22.9 – 35.2) % and 16.3 (10.2 – 23.7) % for GEM. 196 out of 366 GEM patients in the safety set reported 481 grade 3/4 adverse events, while 226 out of 359 GEM/CAP patients reported 608 grade 3/4 adverse events (P=0.242).


Adjuvant GEM/CAP is the standard of care for resected pancreatic cancer.