PD02.07. Delayed Gastric Conduit Emptying: Validation of a Consensus Symptom Grading Score

Abstract

Esophageal Cancer: Surgical Treatment of Esophageal Cancer – early outcomes and complications Delayed gastric conduit emptying (DGCE) following esophagectomy with gastric reconstruction is frequent. Although the Konradsson score was developed to standardize the diagnosis of DGCE, its validity and clinical utility remain insufficiently established. This study aimed to evaluate the Konradsson score’s diagnostic performance and to assess its relationship with therapeutic interventions. A retrospective, single-center, observational study included patients who underwent esophagectomy with gastric conduit reconstruction between March 2023 and December 2024. DGCE symptoms were assessed using the Konradsson questionnaire at postoperative day 14, day 22 and at three months. The five key symptoms were early satiety, vomiting, nausea, regurgitation, and inability to meet caloric needs; each graded from 0 to 3 (total score from 0 to 15). A cut-off score of ≥4 could be derived for the diagnosis and severity grading of late DGCE (at least 2 out of 5 symptoms must reach a score ≥ 2). Early DGCE was defined based on nasogastric tube output and chest X-ray while late DGCE was defined as a positive Konradsson score combined with an objective need for treatment. Associations between Konradsson scores, DGCE diagnoses and therapeutic interventions (prokinetics and pyloric dilatations) were analyzed using correlation analyses, chi-square tests, logistic regression and ROC analyses. Out of 163 cancer patients who underwent esophagectomy with gastric reconstruction, 102 patients completed at least one questionnaire. Early DGCE occurred in 31.4% of patients, while late DGCE was diagnosed in 27.5% at least once during follow-up. Patients with an intrathoracic anastomosis (OR = 0.051; 95%CI 0.017-0.154) and those undergoing open surgery (OR = 0.051; 95%CI 0.017-0.154) showed a lower likelihood of early DGCE compared with those with a cervical anastomosis and those undergoing minimal invasive surgery. Conversely, there was no association between anastomotic location or surgical approach and late DGCE. Early DGCE did not predict late DGCE at any time point. ROC analyses demonstrated a reasonable discriminative performance for late DGCE (AUC 0.739–0.890), but with an optimal cut-off value closer to 5.5 rather than the proposed threshold of 4. Konradsson scores showed no significant association with the use of prokinetic medication or pyloric dilatation at any time point. The Konradsson score demonstrated moderate diagnostic discrimination and showed limited clinical utility as a guide for therapeutic decision-making. Early DGCE did not reliably predict late DGCE. Refinement of the scoring system together with the integration of a pathophysiological classification and a standardized manner of imaging may be required to improve the diagnosis and management of DGCE.