Cancer of Oesophagus or Gastricus – New Assessment of Technology of Endosonography (COGNATE): report of pragmatic randomised trial

Background: Endoscopic ultrasonography is recommended for staging gastro-oesophageal cancers, but has never been evaluated. Objective: COGNATE (Cancer of Oesophagus or Gastricus – New Assessment of Technology of Endosonography) therefore aimed to evaluate whether adding ‘endoscopic ultrasound’ (EUS) to the usual staging algorithm changes treatment, improves (quality-adjusted) survival, and uses resources cost-effectively. Design: Pragmatic parallel-group trial. Patients with gastro-oesophageal cancer received standard staging algorithms. Multidisciplinary teams chose provisional management plans from endoscopic mucosal resection, immediate surgery, surgery after chemotherapy, or chemotherapy and radiotherapy. We used dynamic randomisation to allocate consenting patients remotely by telephone in equal proportions between EUS and not. Thereafter we recorded changes in management plan, use of health-care resources, and three aspects of participant-reported quality of life: generic [measured by European Quality of Life – 5 Dimensions (EQ-5D)], cancer related [Functional Assessment of Cancer Therapy – General scale (FACT-G)] and condition-specific [FACT – Additional Concerns scale (FACT-AC)]. We followed participants regularly until death or the end of the trial – for between 1 and 4.5 years. We devised a quality assurance programme to maintain standards of endosonographic reporting. Setting: Eight British hospitals, of which two – one Scottish teaching hospital and one English district general hospital – contributed 80% of participants; we combined the other six for analysis. Participants: Patients were eligible if they had a diagnosis of gastro-oesophageal cancer, had not started treatment, were free of metastatic disease, were fit for surgery (even if not planned) and had American Society of Anesthesiologists and World Health Organization grades of less than 3. Interventions: Intervention group: standard staging algorithm plus EUS; control group: standard staging algorithm. Main outcome measures: Primary: quality-adjusted survival. Secondary: survival; health-related quality of life (EQ-5D, FACT-G and FACT-AC scales); changes in management plan; and complete resection rate. Although blinding participants was neither possible nor desirable, those responsible for analysis remained blind until the Trial Steering Committee had reviewed the definitive analysis. Results: We randomised 223 patients, of whom 213 yielded enough data for primary analysis. EUS improved survival adjusted for generic quality of life with a hazard ratio of 0.705 [95% confidence interval (CI) 0.499 to 0.995], and crude survival with a hazard ratio of 0.706 (95% CI 0.501 to 0.996). The benefits of EUS were significantly greater for those with poor initial quality of life, but did not differ between centres. EUS reduced net use of health-care resources by £2860 (95% ‘bootstrapped’ CI from −£2200 to £8000). Combining benefits and savings shows that EUS is likely to be cost-effective, with 96% probability of achieving the National Institute for Health and Care Excellence criterion of costing of < £20,000 to gain a QALY. There were no serious adverse reactions attributable to EUS. EUS enhanced the management plan for many participants, increased the proportion of tumours completely resected from 80% (44 out of 55) to 91% (48 out of 53), and improved the survival of those who changed plan; although underpinning the significant differences in outcome, none of these process differences was itself significant. Conclusion: Endoscopic ultrasound significantly improves (quality-adjusted) survival, has the potential to reduce health-care resource use (not statistically significant) and is probably cost-effective (with 96% probability). We recommend research into the best time to evaluate new technologies. Trial registration: ISRCTN1444215. Funding: This project was funded by the NIHR Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 17, No. 39. See the HTA programme website for further project information