Abstract
Aim: to determine the physical and chemical characteristics of laryngopharyngeal reflux (LPR) and the quantitative thresholds associated with laryngopharyngeal symptoms (LPS) in patients with gastroesophageal reflux disease (GERD), and to introduce a novel integral diagnostic laryngopharyngeal index (LP-index) for interpreting 24-hour hypopharyngeal-esophageal multichannel intraluminal impedance-pH (HEMII-pH) monitoring data. Materials and methods. This study evaluated 80 patients with established GERD who underwent outpatient examination at the Motility Lab of the V.Kh. Vasilenko Clinic of Propaedeutics of Internal Diseases, Gastroenterology and Hepatology (Sechenov University) between September 2025 and March 2026. All patients presented with heartburn and regurgitation. Additionally, 40 patients reported laryngopharyngeal symptoms (LPS) — including throat pain, irritation, sour taste, cough and excess throat mucus—persisting for at least 8 weeks. At baseline, all participants completed the Reflux Symptom Index (RSI) questionnaire. Esophagogastroduodenoscopy (EGD) revealed non-erosive GERD in 73.3 % of patients and erosive esophagitis in 26.7 % of cases. Patients were divided into two equal groups based on the presence of LPS: Group 1 comprised patients with GERD without LPS ( n = 40), and Group 2 comprised patients with GERD and LPS ( n = 40). All participants underwent 24-hour HEMII-pH monitoring after manometric localization of the upper esophageal sphincter (UES), with the proximal pH sensor and impedance channel positioned 1 cm above the upper border of the UES. Results. Group 1 comprised patients with GERD without LPS ( n = 40, median age 45.1 [34.8; 53.3] years), and Group 2 comprised patients with GERD and LPS ( n = 40, median age 50.0 [37.0; 60.3] years; p = 0.108). The median Reflux Symptom Index score in Group 2 was 14 [3.5; 37.9]. High-resolution manometry ruled out diagnostically significant esophageal motility disorders. All participants underwent 24-hour HEMII-pH monitoring of the pharynx, esophagus, and stomach. The esophageal DeMeester score was 10.12 [3.51; 24.37] in Group 1 patients and 16.34 [8.55; 26.47] in Group 2 ( p = 0.42), allowing the conclusion that LPS development does not depend on the DeMeester score, and LPS may represent either extraesophageal manifestations of GERD or symptoms of laryngopharyngeal reflux disease. A total of 1,365 LPR episodes were detected in 80 patients: 71.9 % ( n = 981) were mixed and 28.1 % ( n = 384) were liquid. Gas reflux episodes were not recorded. Weakly acidic refluxes were recorded most frequently — 77.4 % ( n = 1057), weakly alkaline accounted for 14.2 % ( n = 194), and acidic episodes accounted for 8.4 % ( n = 114). In GERD patients without LPS, acidic LPR episodes were not recorded. Weakly acidic refluxes were recorded less frequently than in Group 2 (liquid, p < 0.01; mixed, p = 0.067). In GERD patients with LPS, unlike Group 1, liquid and mixed acidic LPR episodes were detected ( p < 0.001). The number of liquid weakly alkaline episodes showed no statistically significant difference between Groups 1 and 2 ( p = 0.905). Mixed weakly alkaline refluxes were registered more frequently in GERD patients without LPS compared to Group 2 ( p = 0.079). ROC analysis demonstrated the diagnostic value of the number of acidic liquid, acidic mixed, and non-acidic liquid LPR episodes for differentiating GERD patients with and without laryngopharyngeal symptoms (AUC = 0.650; p < 0.001). Meanwhile, non-acidic mixed LPR episodes did not demonstrate such diagnostic value (AUC = 0.518; p = 0.69). The percentage of pharyngeal bolus exposure time was significantly higher in GERD patients with LPS compared to GERD patients without LPS ( p < 0.001), confirming the effect of refluxate exposure time on LPS development. The LP-index was significantly higher in Group 2 (6.68 [4.01; 14.80] vs. 2.85 [1.80; 4.80]; p < 0.001), which can be regarded as evidence of the effect of LPR on the development of laryngopharyngeal symptoms. The area under the ROC curve for the LP-index was 0.912 (95% CI: 0.850–0.970), corresponding to excellent discriminatory power. The maximum Youden index value (0.725) was achieved at an LP-index threshold of 4.80 (sensitivity — 82.5 %, specificity — 90.0 %). A statistically significant association was identified between LPS development and a pathological LP-index (χ2 = 16.2; Fisher’s exact test p < 0.001). Identifying a pathological LP-index increases the odds of LPR-related laryngopharyngeal symptoms by 7-fold (OR = 7.00; 95% CI: 2.62–18.70; φ-coefficient = 0.451). Based on the data obtained, ROC analysis was performed to select a threshold number of LPR episodes contributing to the development of laryngopharyngeal symptoms in GERD patients. The area under the ROC curve for the number of LPR episodes was 0.680 (95% CI: 0.599–0.753; p < 0.001). The maximum Youden index value was achieved at a threshold of 3 LPR episodes. At this threshold, sensitivity was 76.25 % and specificity was 56.94 %. Conclusion. Based on 24-hour HEMII-pH monitoring data, ROC curve analysis demonstrated excellent discriminatory performance and high prognostic value of the novel integral laryngopharyngeal index (LP-index) in evaluating the association between laryngopharyngeal reflux episodes and the development of laryngopharyngeal symptoms in GERD patients. The threshold value for the LP-index was established at 4.80. Exceeding this cut-off level serves as objective evidence for the presence of LPR and its pathogenetic contribution to the development of LPS. Registration of 3 or more LPR episodes per 24 hours is associated with the possible development of laryngopharyngeal symptoms in patients with GERD. The numbers of acidic liquid, acidic mixed, and non-acidic liquid LPR episodes have the highest diagnostic value for differentiating GERD patients with and without laryngopharyngeal symptoms. The diagnostic and pathogenetic role of non-acidic mixed LPR episodes in the development of laryngopharyngeal symptoms requires further clinical investigation.

This publication has 31 references indexed in Scilit: