In this study, the case-control method was used to collect the adult patients who were diagnosed as LPRD according to the “Experts consensus on diagnosis and treatment of laryngopharyngeal reflux disease (2015)”15 from January 2021 to June 2022. The diagnostic criteria for LPRD were reflux symptom index > 13, or reflux finding score > 7, corroborated by either an esophagogastroduodenoscopy (erosive esophagitis or Barrett esophagus) or positive prolonged esophageal pH monitoring. The enrolled LPRD patients were designed as the LPRD group. At the same time, healthy volunteers were selected as the control group. All healthy volunteers were between 18 and 70 years old, with no LPRD symptoms, such as hoarseness, dysphagia, globus, regurgitation, heartburn, cough/throat clearing, or excessive throat mucus, within 2 months of the study. Individuals with a history of severe systemic diseases, laryngopharyngeal surgery, or hiatal hernia, were excluded from the control group. Those who had received any treatments for OSAHS or LPR were also excluded.The subjects were fully informed of the relevant information and signed the informed consent form. This study was designed in compliance with the Declaration of Helsinki and reviewed by ethical review committee of The Fourth Affiliated Hospital, Zhejiang University School of Medicine. Informed consent was obtained from all individual participants included in the study.

Inclusion criteria for PDRD patients: 1) The Reflux Symptom Index (RSI) was more than 13 and/or the Reflux Finding Score (RFS) was more than 7; 2) More than 3 laryngopharyngeal reflux events within 24h (24h multichannel intraluminal impedance-pH monitor); 3) No history of taking proton pump inhibitors or other drugs that inhibit gastric acid secretion within 1 month before the first visit; 4) No history of taking preparations of antibiotics, micro-ecology-related products, non-steroidal anti-inflammatory drugs, immunosuppressants, hormones and other drugs that affect the microbiota.

Exclusion criteria for all subjects: 1) Those with other organic diseases at the throat (such as benign or malignant tumors of the throat); 2) Those who have chronic systemic diseases and need long-term medication (such as hypertension, diabetes, kidney disease, etc.); 3) Patients who were previously diagnosed as gastroesophageal reflux disease; 4) Patients previously diagnosed as chronic sinusitis, chronic tonsillitis, allergic rhinitis or other allergic diseases; 5) Those who have upper respiratory tract infection; 6) Those who take drugs for pharyngitis in the past month; 7) Those with mental disorders or pregnancy.

All enrolled patients were given Esomeprazole Magnesium Enteric-coated Tablets (AstraZeneca AB) 20mg, twice a day, 30–60min before meals, for 8 weeks. Follow-up was conducted at the 8th week of treatment.

The curative effect was evaluated after treatment, and the RSI/RFS scale was collected by the researcher the same as before medication.

Effective: symptoms improved by more than 50%, RSI less than 13 or decreased by more than 4.

Invalid: no improvement in symptoms, and no reduction in RSI.

Under the electronic laryngoscope, the secretion from the laryngopharynx (piriform sinus) were taken from patients in fasting state, into sterile culture tubes and sterile Eppendorf tubes. Sterile culture tubes were sent to the laboratory within 1h for microbial culture. The sterile Eppendorf tubes were quickly frozen by liquid nitrogen and stored at −80°.

Pharyngeal swabs were placed on blood agar plates, chocolate agar plates, and China Blue agar plates and incubated at 37°C±1 in atmosphere of air or 5% CO2. After incubation, bacterial identification was performed on an automatic Phoenix system (Becton Dickinson, USA) or an automatic Vitek2 system (bioMérieux Inc., France).

Total nucleic acid isolation was conducted in duplicate using the MagMAX Microbiome Ultra Nucleic Acid Isolation Kit (Applied Biosystems, USA). Yields were evaluated ed on a NanoDrop™ 2000 Spectrophotometer (Thermo Scientific, USA).

The diluted DNA sample (1ng/μL) was used as the template to complete PCR amplification targeting the hypervariable region of the 16S rRNA region of bacteria using the primers.16 PCR products were purified and recovered using Poly-Gel DNA Extraction Kit (Solarbio, China).

PCR products (10μL) were analyzed by DGGE. Under the condition of 150V and 60°C, the PCR product was denatured by polyacrylamide gel at a gradient of 35%‒55% and a mass fraction of 8% in 1×TAE buffer for 5h. After DGGE, deionized water was poured out. The gels were fixed and stained with s silver dye solution. The differentiation analysis of the DGGE gel atlas was carried out with Quantity One software, and the allowable error was 2%, to collect the number of bands and gray values. The Dice index calculated by the Quantity One software was used to construct the system tree for cluster analysis between different bacterial groups. Shannon-Wiener index and evenness index represented the diversity and evenness.

PCR products was reversely transcribed into cDNA with OneScript Hot cDNA Synthesis Kit (Abcam) using random primers. For total bacterial DNA quantification, droplet digital PCR reaction was prepared with sample DNA, EvaGreen Supermix (BioRad, USA), and primers. For quantification of Klebsiella oxytoca, Klebsiella oxytoca PCR kit was used. Samples were placed into a QX200 Automated Droplet Generator (BioRad, USA) and subjected to PCR amplification. Amplification product was placed in a QX200 Droplet Reader (BioRad, USA), to space out positive and negative droplets based on their individual fluorescence.

According to the previous literature, we estimated that the RSI score of the LPRD group would be 4.5 points higher than that of the control group after treatment.17 A sample size of 29 patients per group was calculated with an assumption of alpha value of 0.05% and 80% power using a two-tailed t-test. Considering a dropout rate of 10%, 33 patients will be recruited for each group. In actual recruitment, we will plan to recruit 35 participants for the intervention and control groups. We will plan to recruit sufficient participants by posting posters in ENT clinics for 4-months.

Statistical analyses were completed in GraphPad Prism 9 and SPSS 22.0 statistical software. Shapiro-Wilk test was used to check normal distribution for all data sets. For non-normally distributed data, log-transformed or non-parametric tests were used to determine the statistical significances. For normally or lognormally distributed data, unpaired two-tailed Student’s t-test or one-way analysis of variance was used to determine the statistical significances. The Pearson Chi-Square test was used for comparing categorical data between treatment effective and invalid groups. The predictive performance of Klebsiella oxytoca was assessed by Receiver Operating Characteristic (ROC) curve analysis. A p-value < 0.05 refers to a statistical significance.

A total of 57 subjects were finally included in the study, of which 29 were in the LPRD group and 28 in healthy group (with one dropped out). LPRD group included 18 males and 11 females, with an average age of 49.52±7.90 years. The healthy group included 16 males and 12 females, with an average age of 48.64±8.76 years. There was no significant difference in age, gender, Body Mass Index (BMI), smoking history, drinking history, and history of gastropathy between the case group and the control group (Table 1). After treatment, the effective cases were 18 and invalid ones were 11 among patients with LPRD. There was no significant difference in age, gender, Body Mass Index (BMI), smoking history, drinking history, and history of gastropathy between the treatment-effective group and the treatment-invalid group (Table 2).

A total of 86 specimens of throat secretions were cultured in this study. As shown in Fig. 1A, the bacteria in healthy participants were 26 kinds (60 strains). Among the 29 cases of LPRD pretreatment, 28 kinds of bacteria and 77 strains were cultivated, including 14 kinds of gram-positive bacteria and 12 kinds of gram-negative bacteria. The top three bacteria were Klebsiella pneumoniae, Viridans Streptococci, and Neisseria flavescens. After treatment, the effective cases were cultured to 26 kinds of bacteria, while the invalid ones were cultured to 27 kinds of bacteria. The detection rate of Klebsiella oxytoca and Klebsiella pneumoniae was different among the four groups (Fig. 1B).

Figure 1.

Composition of microbiome by bacterial culture method. (A) Pie charts of microbiome microbiome composition. (B) The distribution of Gram-positive and negative bacteria.

(1.12MB).

According to the intensity and migration degree of the bands in DGGE, the similarity between samples is calculated according to the dice coefficient. The cluster analysis was conducted and the generated system tree. As shown in Fig. 2A, the cluster similarity between the health group and the treatment-effective group was high, as well as the similarity between the LPRD-pretreatment and treatment-invalid group. Droplet digital PCR reaction quantified the total bacteria (Fig. 2B), showing that total bacteria number in treatment-effective group was significantly higher than that in the pre-treatment group and treatment-invalid group. This increase in bacteria abundance was accompanied by a rise in bacterial alpha diversity as measured by their Shannon Index (Fig. 2C) and evenness index (Fig. 2D).

Figure 2.

Comparisons of alpha diversity indices among the groups. (A) Dendrogram constructed based on Dice coefficients derived from DGGE fingerprints. (B) Total bacterial DNA copies were quantified by droplet digital PCR. The microbial alpha diversity within each sample was analyzed based on the Shannon-Wienner diversity index (C) and the Pielou's evenness (D).

(0.28MB).

Next, Klebsiella oxytoca aroused our interest. By using droplet digital PCR, we quantified Klebsiella oxytoca in the treatment-effective and -invalid groups. As shown in Fig. 3A, compared with the treatment-effective group, the treatment-invalid group showed lower Klebsiella oxytoca abundance. Then, the ROC curve was plotted to assess the discriminatory ability of Klebsiella oxytoca abundance in treatment outcomes. As shown in Fig. 3B, the abundance of Klebsiella oxytoca was an effective biomarker to distinguish treatment-effective and -invalid groups (AUC=0.859) with a sensitivity of 77.78% and specificity of 90.91%.

Figure 3.

Klebsiella oxytoca had potential to discriminate the different outcomes of LPRD patients. (A) Relative copies of Klebsiella oxytoca were quantified by droplet digital PCR. (B) The ROC curve was constructed based on the relative copies of Klebsiella oxytoca.

(0.17MB).