FOCUS4: lessons learnt from ten years of the first molecularly stratified adaptive trial platform in metastatic colorectal cancer


Session type:

Richard Adams1, Tim S Maughan2, Janet Graham3, Louise Brown4, David Fisher4, Jenny Seligmann5, Matthew Seymour5, Richard Kaplan4, Emma Yates4, Susan Richman5, Philip Quirke5, Rachel Butler6, Kai Keen Shiu4, Les Samuel7, Gary Middleton8, Harpreet Wasan9, Ewan Brown10, Fiona Collinson11, Robert Jones1, Michael Braun12, Richard Wilson13
1Cardiff University, 2University of Oxford, 3Beatson West of Scotland Cancer Centre, 4University College London (UCL), 5University of Leeds, 6North Bristol NHS Trust, 7NHS Grampian – Aberdeen Royal Infirmary, 8University of Birmingham, 9Imperial College London, 10Edinburgh Cancer Centre, 11Leeds Teaching Hospitals NHS Trust, 12Manchester University Hospitals NHS Foundation Trust, 13Beatson Institute for Cancer Research



FOCUS4 was the first molecularly stratified trial in metastatic colorectal cancer (mCRC) and one of the first umbrella trial designs launched globally. After 10 years of trial delivery, we report our molecular results and learning from all stakeholders.


FOCUS4 (phase II/III) tested safety and efficacy of targeted therapies in mCRC. It used adaptive statistical methodology to decide if sub-trials should close early and new therapies were added as protocol amendments. Patients with newly diagnosed mCRC were registered and central laboratory testing stratified tumours into molecular subtypes (MSI, BRAF, PIK3CA, TP53 and RAS). After 16 weeks therapy, those with stable/responding disease were eligible for randomisation into a molecularly stratified sub-trial (FOCUS4-B, C or D) or non-stratified FOCUS4-N.  The primary outcome was progression-free-survival (PFS) comparing intervention with active monitoring/placebo.  On trial completion we sent a detailed questionnaire to key study team members and recruiting sites via an online portal. 


From Jan 2014 - Oct 2020, 1,434 patients were registered (88 sites). Mutations were found in 1,291/1,382 samples: BRAF 10%; PIK3CA 14%; RAS 58%; TP53 69%; RAS+TP53 32%; BRAF-PIK3CA-RAS wildtype 26%.  908/1315 (69%) patients completing 16 weeks therapy were eligible for randomisation and 361 allocated into a sub-trial.  We activated 3 molecularly targeted sub-trials: FOCUS4-D evaluated AZD8931 in BRAF-PIK3CA-RAS wildtype subgroup (negative trial); FOCUS4-B evaluated aspirin in the PIK3CA mutant subgroup (unable to recruit) and FOCUS4-C evaluated adavosertib in the RAS+TP53 dual mutant subgroup (positive trial). FOCUS4-N evaluated capecitabine versus a treatment break (positive trial).  


Adaptive stratified medicine trials have many advantages and challenges.  Our stakeholder feedback has helped inform how such trial designs can succeed in answering multiple questions efficiently, providing resources are adequate. Robust biomarker analysis, correct positioning in the disease pathway, strong partnerships between participating organisations and nimble regulatory control are key to successful delivery of such complex studies.

Impact statement

The FOCUS4 trial was the first adaptive platform trial in metastatic colorectal cancer and learnings from this have helped in both the design and delivery of currently open and planned adaptive trials in multiple common solid tumours.