Submucosal Saline Injection Followed by Endoscopic Ultrasound versus Endoscopic Ultrasound Only for Distinguishing between T1a and T1b Esophageal Cancer: A Single-Blind, Superiority, Phase 3, Randomized Controlled Trial

Background: Traditional endoscopic ultrasound's (EUS) accuracy in distinguishing between T1a and T1b stage esophageal cancer (EC) is inadequate. As such, we examined whether submucosal saline injection (SSI) can improve its diagnostic accuracy. Methods: We conducted a randomized controlled phase 3 trial at Sun Yat-sen University Cancer Center. A total of 180 patients aged 18-85 years old with T1N0M0 stage EC indicated by prior EUS and radiologic imaging were enrolled and randomly assigned (1:1) to receive either EUS examination after SSI (EUS examination after 3-5 ml submucosal saline injection) or EUS only examination. The randomization sequence was computer-generated and no stratification factor was used. All the patients were referred to thoracic surgeons to receive endoscopic submucosal dissection (ESD) or esophagectomy within one week after EUS examination. Patients, research assistants, thoracic surgeons, pathologists, and statisticians were masked to the allocated examination. Standard EUS criteria were used to preoperatively stage the EC cases, while surgical pathology reports after referral were used to postoperatively stage the cases. The primary endpoint was the diagnostic accuracy of T1a staging (defined as the sum of the true positive [T1a] and true negative [T1b] cases divided by the total number of cases) analyzed in the per-protocol population. Findings: Between February 14, 2012 to October 23, 2017, 180 patients were enrolled and randomly assigned to receive SSI EUS (n=90) or EUS only (n=90) examination. Among the per-protocol population, the SSI EUS group (n=81) was superior to the EUS-only group (n=87) in terms of the diagnostic accuracy for T1a staging (86.4% [95% CI, 79.0-93.9] vs. 66.7% [95% CI, 56.8-76.6]; p=0.003). The specificities with SSI EUS and EUS-only were similar (92.3% [95% CI, 82.1-100.0] vs. 90.5% [95% CI, 81.6-99.4]; p>0.999), whereas the sensitivity of SSI EUS was significantly higher than EUS-only (83.6% [95% CI, 73.9-93.4] vs. 44.4% [95% CI, 29.9-59.0]; p<0.001). Adverse events during examination were recorded in one patient in the SSI EUS group and one in the EUS-only group (grade 1 esophageal hemorrhage for both). Compared with those in the EUS-only group, patients with postoperative T1a disease in the SSI EUS group had more commonly received ESD (38 [69.1%] of 55 patients vs. 16 [35.6%] of 45, p<0.001) and had a lower median total medical cost (4,913.8 vs. 11,717.7 USD, p<0.001) and shorter median hospital stays (13.0 vs. 23.0 days, p=0.003). With a median follow-up time of 40.0 months (range, 8.4-71.2 months) in the SSI EUS group and 35.7 months (range, 8.3-76.0 months) in the EUSonly group, the 3-year overall survival rates were comparable between the SSI EUS and EUS-only groups (91.5% vs. 84.7%, p=0.510) for postoperative T1a stage EC patients. Interpretation: SSI significantly improves the diagnostic accuracy of EUS for T1a staging and might be of great value for optimizing treatment strategies for T1 stage EC patients. Clinical Trial Number: This trial is registered at ClinicalTrials.gov, number NCT01555801. The trial is ongoing but no longer recruiting participants. Funding Statement: The study is supported by the National High Technology Research and Development Program of China (863 Program, 2015AA020103); Science and Technology Plan Projects of Guangdong Province, China (2011B080701015, 2012B061700076, and 2014A020212146); Sun Yat-sen University Cancer Center Clinical Research 308 Program, and Plan Project of Guangdong Esophageal Cancer Institute (2014). The funding agencies will not affect the study results. Declaration of Interests: We declare no competing interests. Ethics Approval Statement: This study was approved by the institutional review board of SYSUCC. This trial was conducted in accordance with the Declaration of Helsinki, the guidelines for Good Clinical Practice, the European Union Clinical Trial Directive, and local regulations. All participants provided written informed consent.