A multicentre cohort study to redefine and validate pathological assessment of response to neoadjuvant therapy in treated oesophagogastric adenocarcinoma


Session type:


Megan A Lloyd1,Fergus Noble1,Rob Walker1,Damien McManus2,Richard Turkington2,Matt Bedford3,Ewen Griffiths3,John Whiting3,Barbara Nutzinger4,Nicola Grehan4,J Robert O'Neill5,Richard JE Skipworth5,Vicki Save5,Sutaria Rupesh6,Stuart Mercer6,John H Saunders7,Irshad N Soomro7,Simon L Parsons7,Rebecca C Fitzgerald8,Timothy J Underwood9
1Cancer Sciences Unit, University of Southampton,2Centre for Cancer Research and Cell Biology, Queen’s University Belfast,3University Hospitals Birmingham NHS Foundation Trust,4on behalf of OCCAMS, Hutchison/MRC Cancer Unit, University of Cambridge,5Royal Infirmary of Edinburgh,6Portsmouth NHS Trust,7Nottingham University Hospitals NHS Trust,8on behalf of OCCAMS, Hutchison/MRC Cancer Unit, University of Cambridge, Cambridge, UK,9on behalf of OCCAMS, Cancer Sciences Unit, University of Southampton



A universally accepted measure of significant response to neoadjuvant therapy in oesophageal adenocarcinoma (OAC) is required to stratify patients for precision treatment. Our study aimed to define the utility of Mandard tumour regression grade (TRG) in UK clinical practice, assess its reliability, and validate a clinically meaningful endpoint of significant pathological response.


A questionnaire assessed current clinical use of neoadjuvant response assessment in 11 UK centres. 7 centres provided prospectively collected clinicopathological data from patients with treated OAC (2001-2016). Pathological primary tumour response to neoadjuvant therapy was assessed using TRG, with blinded validation of scoring.   Lymph node downstaging was assessed by comparing clinical and pathological staging (cN+ to ypN0).


Of 11 centres 64% used response to neoadjuvant therapy to guide adjuvant treatment. TRG was recorded by 73% of centres with 63% using TRG 1-3 to define a significant response to therapy. Of 1258 patients studied, a significant overall survival advantage was seen with TRG 1-2 (responders, n=189; 15%) compared to TRG 3-5 (non-responders, n=1069; 85%) (mean overall survival; TRG 1-2: 10.4 years, 95% CI: 9.4-11.3 vs TRG 3-5: 4.8 years, 95% CI: 4.3-5.2, p<0.0001). A subset of non-responders (TRG 3-5) demonstrated lymph node (LN) downstaging with a subsequent survival advantage (mean overall survival; TRG 3-5 with LN downstaged, n=155: 7.9 years, 95% CI: 7.0-8.7 vs TRG 3-5 LN not downstaged, n=581: 3.2 years, 95% CI: 2.9-3.6, p<0.0001).


Variability in the application of TRG to guide treatment in OAC exists in the UK. This multi-centre study validates responders to neoadjuvant therapy with overall survival double that of non-responders. Assessment of response should not be confined to the primary tumour as a cohort of apparent non-responders gain a survival advantage by lymph node down staging. These robust criteria will aid the development of biomarkers to predict response.