P240 Accuracy of clinical staging for T2N0 oesophageal cancer: systematic review and meta-analysis

Background and Aims Oesophageal cancer is the 7th most common cause of cancer worldwide and is the 6th most common cause of overall cancer mortality. Clinical staging utilises multiple imaging modalities and is the most appropriate guide for treatment and prognostication. T2N0 oesophageal cancer is a treatment threshold for neo-adjuvant therapy, but data on the accuracy of current clinical staging tests for identifying this disease subgroup are conflicting. Methods We performed a meta-analysis of all primary studies evaluating the accuracy of clinical staging (index test) compared to histopathological staging of oesophagectomy specimens (reference standard) in T2N0 oesophageal cancer. Clinical staging used a combination of PET, EUS, and CT imaging modalities. Patients were excluded if they had neo-adjuvant therapy to allow direct comparison between clinical and pathological staging. Databases searched included: Ovid MEDLINE, Ovid Embase and The Cochrane Library up to September 2019. The primary outcome was diagnostic accuracy of combined T and N clinical staging. Secondary outcomes were accuracy of T stage; percentage T downstaged; percentage T upstaged and percentage N upstaged. We evaluated several sources of heterogeneity a priori including: publication date (before and after Jan 2015), date of first recruitment (before and after 2000), number of centre (single vs multicentre), sample size ( Results After removing duplicates, the search strategy identified 1,199 studies which were all screened by title and abstract. Eighteen studies met the inclusion criteria containing 5,115 patients. The combined TN T stage accuracy was 30% (95%CI, 24–37); percentage of patients with T down-staging was 40% (95%CI, 31–48); percentage of patients with T upstaging was 28% (95%CI, 23–33) and the percentage of patients with N upstaging was 35% (95%CI, 30–40)(table 1). Sources of heterogeneity in accuracy of TN p=0.01); sample size (n≤100 patients=15% vs n≥100 patients=27%; p=0.01); study region (USA=19% vs Asia=19% vs Europe=38%; p Conclusion Staging for T2N0 oesophageal cancer remains inaccurate with a significant proportion of patients having their disease downstaged (therefore, could have been potentially offered endotherapy instead of oesophagectomy) as well as upstaged (therefore, they could have been potentially offered neo-adjuvant therapy). These findings were largely unchanged over the past two decades suggesting that there is an urgent need for more accurate staging tests for this subgroup of patients.