The ENSURE study will comprise two phases.

Phase 1: European multicenter survey of surveillance protocols after esophageal cancer surgery

ENSURE questionnaire will be circulated to representatives from participating European countries.

Phase 2: European multicenter retrospective observational study of the impact of postoperative surveillance protocols on oncologic outcome and HR-QL

Phase 2 will constitute a retrospective observational study of patients undergoing treatment with curative intent for esophageal cancer at participating Centers from June 2009 to June 2015.

Despite significant improvements in oncologic outcome for patients with esophageal cancer, with increased early detection and greater utilization of multimodal protocols,approximately 50% and 20% patients with locally advanced and localized disease treated with curative intent will develop disease recurrence within 5 years. After completion of treatment, there is little evidence to guide surveillance strategies, and whether routine clinical, biochemical, radiologic or endoscopic follow-up is associated with improved oncologic outcomes, or health-related quality of life (HR-QL), is unknown. This is in contrast to colorectal cancer, where meta-analyses demonstrate a significant overall survival (OS) benefit among patients undergoing intensive surveillance, likely driven by a proportion of patients achieving long-term survival following salvage treatments for oligometastatic disease. Importantly, intensive surveillance in colorectal cancer did not adversely impact HR-QL in survivorship.

In esophageal cancer, a lack of quality to indicate that aggressive treatment of recurrent disease provided an OS benefit had previously limited the rationale for intensive oncologic surveillance in this cohort. However, increasingly reports of long-term survival after salvage surgery for recurrent esophageal cancer among patients treated with esophagectomy as a component of initial treatment are emerging. For example, Depypere et al. reported that among the 15% of patients with isolated local recurrence or single solid organ metastasis who were treated surgically, 5-year survival from time of recurrence was 50% (median 55 months). This is in keeping with data from Schipper et al. who report 44% and 35% 3- and 5-year survival, respectively, in the minority of patients for whom R0 re-resection of local recurrence was possible. Furthermore, while meta-analysis now confirms the survival benefit of palliative chemotherapy (HR 0.81, 95% CI 0.71 to 0.92), and targeted therapy (HR 0.75, 95% CI 0.68 to 0.84), the independent impact of timing of intervention, and disease volume, on OS in the setting of recurrent disease is unclear.

Indeed, no randomized controlled trial has compared oncologic outcomes among cohorts undergoing different surveillance protocols, and only two retrospective observational studies have directly addressed this question. In a study from Canada, Peixoto et al. compared survival outcomes among 292 patients with gastroesophageal cancers (58% esophageal and junctional, with 23% treated with definitive chemoradiation only), undergoing four different surveillance modalities (discharge to General Practitioner, or Medical Oncology follow-up with either clinical, clinical and biochemical, or clinical and radiologic/endoscopic surveillance). There was no difference in recurrence-free or OS according to surveillance strategy on univariable or multivariable analysis. However, recent data from Sisic et al. advise caution in interpretation of the former study. Among 587 patients undergoing either high intensity radiologic or standard clinical surveillance after potentially curative treatment for gastric (47%) or junctional (53%) adenocarcinoma, intensive surveillance independently predicted OS on multivariable analysis and was associated with an approximate 5% increase in OS at 3 and 5 years. Notably, among patients who developed disease recurrence, OS was significantly greater (median 40 versus 19 months) for those with radiologically-detected versus clinical recurrence, and for those who underwent metastectomy (n = 29, 5% of total population) versus other treatments (median 66 versus 25 months), similar to previous series. Although this likely reflects, at least in part, inherent differences in disease biology and volume among those who recur symptomatically versus radiologically, patients undergoing high intensity surveillance were more likely to have treatment for recurrence, with a proportion of these achieving long-term survival (n = 11, 2% of total population). Though limited, these data indicate that intensive surveillance may provide benefit for the small proportion of patients who may be treated with curative intent for low volume recurrence.

Given the limited evidence to support postoperative surveillance protocols in esophageal cancer, it is unsurprising that international guidelines also fail to reach consensus on this topic. For example, the National Comprehensive Cancer Network (NCCN) recommend regular cross-sectional imaging with computed tomography (CT) for all patients with locally advanced disease (cT2-4 Nany) post multimodal therapy, and consideration of regular surveillance CT among patients with T1b Nany treated with upfront esophagectomy, up to 3 years postoperatively, with endoscopic surveillance only in the case of residual Barrett's esophagus. In contrast, European Society for Medical Oncology (ESMO) guidelines highlight a lack of evidence to support the role of postoperative surveillance protocols to improve oncologic outcomes, and determine that follow-up should concentrate on symptoms, nutrition and psychological support, and neither routine imaging nor endoscopic surveillance is advocated.

As indicated in the ESMO and ASGBI surveillance recommendations, there are potential additional benefits to postoperative surveillance protocols. Routine clinical surveillance may facilitate identification and management of late post treatment morbidity such as anastomotic stricture, post thoracotomy pain, malabsorption, altered appetite and dumping syndrome. Furthermore, data from the Netherlands indicate that patients exhibit a strong preference for routine imaging surveillance after esophagectomy, with 67% preferring imaging even if this approach would not provide a survival benefit.19 This is in keeping with UK colorectal cancer data demonstrating that although surveillance is associated with transient anxiety, negative test results provide significant reassurance for the majority. Interestingly, similar to the Dutch data, most patients (78%) stated that they would value finding out about the presence of recurrence, even if there would be no survival benefit. This would provide evidence to counter the concern that postoperative surveillance may negatively influence HR-QL, either through increased anxiety related to surveillance investigations, or earlier diagnosis of otherwise asymptomatic patients with recurrent disease.

Examining surveillance practices internationally, there is significant variability in the utilization of biochemical, radiologic and endoscopic surveillance. In Japan, a nationwide survey found that high intensity surveillance was common, and notably endoscopic surveillance was frequently utilized (85%), in contrast to Australia and New Zealand where only 6% of surgeons opted to undertake routine endoscopic surveillance of asymptomatic patients after potentially curative esophagectomy. The European Registration of Cancer Care (EURECCA) Upper Gastrointestinal group included a brief summary of postoperative surveillance protocols for esophagogastric adenocarcinoma among the 10 participating Centers in a recent publication on clinical pathways. Importantly, this demonstrated clear differences across European Centers in key elements of surveillance, such as the routine use of CT (in 40%) and tumor markers (in 40%), but did not describe endoscopic surveillance protocols or duration of surveillance, while oncologic outcomes according to surveillance intensity were not assessed.

Given these major differences in surveillance intensity and modality across Europe, with otherwise similar population demographics and treatment pathways, we in Europe are uniquely poised to further study the impact of surveillance with respect to oncologic outcome, and HR-QL, after potentially curative surgery for esophageal cancer. Therefore, the primary aims of this European collaborative multicenter study are:

1. Firstly, to characterize differences in postoperative oncologic surveillance protocols across European esophageal cancer centers, and
2. To determine the independent impact of high intensity surveillance (HIS) on disease-specific and overall survival.

Inclusion criteria

1. Age 18 years and above
2. Underwent surgery with curative intent for cTxNxM0 esophageal or esophagogastric junction (Siewert type I, II and III) cancer
3. Salvage surgery after failure of primary endoscopic or oncologic treatment will be included

Exclusion criteria

1. Endoscopic therapy or definitive oncological treatment as sole therapy for esophageal cancer
2. Missing follow-up data

Phase 2 will constitute a retrospective observational study of patients undergoing treatment with curative intent for esophageal cancer at participating Centers from June 2009 to June 2015.