High Diagnostic Yield of Abnormal Endoscopic Findings in the Evaluation of Laryngopharyngeal Reflux

Research Letter

Esophagogastroduodenoscopy (EGD) is recommended in patients with typical gastroesophageal reflux disease (GERD) symptoms (heartburn, regurgitation, chest pain) in the setting of proton pump inhibitor (PPI) non-response. EGD evaluates for erosive disease, assesses anti-reflux barrier integrity, excludes non-GERD conditions, and, in the absence of erosive findings, is followed by reflux testing.^1,2 The diagnostic utility of EGD is less clear in the evaluation for laryngopharyngeal reflux (LPR), and the current reference standard is ambulatory reflux monitoring.^1,3,4 In this study of patients referred for evaluation of chronic laryngeal symptoms, we aimed to 1) characterize endoscopic findings, 2) discern whether findings differ between patients with or without concomitant esophageal reflux symptoms, and 3) measure the association between endoscopic findings and objective GERD on ambulatory reflux monitoring.

This observational IRB approved multicenter (4 North American sites: University of California San Diego [San Diego, CA], Brigham and Women’s Hospital [Boston, MA], Northwestern University [Chicago, IL], and Washington University [St. Louis, MO]) and 1 Asian site: Tzu Chi University [Hualien County, Taiwan]) study analyzed data from adult patients who underwent an EGD for evaluation of laryngeal symptoms (cough, globus, dysphonia, throat clearing, sore throat) from 03/2018-05/2023. It is practice at all centers for patients with isolated laryngeal symptoms referred to Gastroenterology to undergo objective GERD evaluation. Primary analysis characterized prevalence of erosive disease or hiatal hernia on EGD.

Additional analyses compared prevalence of findings among patients with and without concomitant esophageal reflux symptoms and assessed the odds of conclusive GERD on reflux monitoring among patients with mild erosive disease on EGD versus those without. Conclusive GERD was defined as total distal acid exposure time (AET) ≥6% and/or ≥80 reflux episodes/24h on ambulatory reflux monitoring off PPI,² total AET ≥2.0% and/or ≥40 reflux episodes/24h on PPI,⁵ or Los Angeles (LA) grade C/D esophagitis or long segment Barrett’s esophagus (BE) (≥3cm) on EGD).²

Overall, 756 patients met inclusion criteria [275 (36%) males, mean age 53.3 (SD 15.7) years, mean body mass index 26.8 (6.4) kg/m², 548 (72%) reporting significant concomitant esophageal symptoms]. Notable endoscopic findings were present in 352/756 (47%) patients, including 131/756 (17%) with erosive esophagitis [LA grade A 75 (10%), LA B 39 (5%), LA C 11 (1%), LA D 6 (1%)], 25/754 (3%) with non-dysplastic BE (8 long segment (≥3cm), 17 short segment), and 277 (37%) with hiatal hernia ≥ 1cm (Table 1).

Table 1:

Baseline Characteristics

Demographics
Age (years)	53.3 (15.7)
Male Sex	275 (36%)
Body Mass Index (kg/m2)	26.8 (6.4)
Erosive Reflux Findings on Endoscopy
Erosive Esophagitis	131 (17%)
Los Angeles A	75 (10%)
Los Angeles B	39 (5%)
Los Angeles C	11 (1%)
Los Angeles D	6 (1%)
Long Segment Barrett’s Esophagus	8/754 (1%)
Short Segment Barrett’s Esophagus	17/754 (2%)
Esophageal Stricture	6/401 (1%)
Assessment of Antireflux Barrier on Endoscopy
Gastro-esophageal Valve Hill Grade
Grade 1	43/148 (29%)
Grade 2	93/148 (63%)
Grade 3	11/148 (7%)
Grade 4	1/148 (1%)
Hiatal Hernia on EGD or HRM ≥ 1cm	277 (37%)
Other Endoscopic Findings
Inlet Patch	15/401 (4%)
Eosinophilic Esophagitis	5/401 (1%)
Candida Esophagitis	7/401 (2%)
Questionnaires
GerdQ Score (n 571)	8.4 (3.1)
Reflux Symptom Index Score (n 560)	20.3 (8.8)
Esophageal Hypervigilance and Anxiety Scale (n 499)	33.6 (13.8)
Esophageal Symptoms (Heartburn, regurgitation, chest pain)	548 (72%)

Continuous variables presented mean ± SD and categorical variables presented as n(%).

The odds of BE [OR 0.67 (95% CI 0.30, 1.60); p=0.34] and odds of LA C/D esophagitis [OR 1.24 (95% CI 0.43, 4.44), p=0.71] were not significantly different between patients with or without concomitant esophageal symptoms (Supplemental Table).

Presence of mild to moderate erosive findings (LA A/B esophagitis or short segment BE) was significantly associated with conclusive GERD on ambulatory reflux monitoring when compared to absence of erosive findings [OR 3.4 (CI 2.28, 5.25); p<0.01]. Specifically, the positive predictive value for LA A esophagitis was 55% (38/69), for LA B esophagitis 78% (28/36) and for short segment BE 60% (9/15). Presence of hiatal hernia ≥ 1cm was significantly associated with conclusive GERD compared to those without hiatal hernia [OR 2.3 (CI 1.7, 3.2); p<0.01] (Supplemental Figure).

Recent guidelines and best practice updates suggest that patients with chronic laryngeal symptoms, in the absence of typical GERD symptoms, undergo objective testing prior to a PPI trial.^1,3,4 Objective testing can include wireless reflux monitoring 24h pH-impedance monitoring; however, the utility of EGD in the evaluation of these patients is unclear.^1,3,4 The key findings of this study of 756 patients referred for LPR evaluation undergoing EGD are 1) 47% of patients had notable endoscopic findings including 17% with erosive reflux disease and 37% with hiatal hernia, 2) patients with laryngeal and concomitant esophageal symptoms were no more likely to have severe (LA C/D or long segment BE) erosive disease on endoscopy compared to those with isolated laryngeal symptoms, and 3) presence of mild to moderate erosive reflux disease was significantly associated with conclusive GERD on ambulatory reflux monitoring

While well established that LA grade A esophagitis is insufficient to diagnose GERD and that LA C/D esophagitis are indicative of conclusive GERD; controversy exists over whether LA B esophagitis represent conclusive evidence of GERD.^1-3,6 In our study, 78% of patients with LA B esophagitis had conclusive GERD on ambulatory reflux monitoring, supporting the notion that LA grade B esophagitis is evidence of objective GERD in a population referred with chronic laryngeal symptoms, similar to recent studies supporting LA B esophagitis as evidence of GERD.⁶

The finding of comparable rates of BE and LA grade C/D esophagitis in patients with isolated laryngeal symptoms compared to those with concomitant significant esophageal symptoms suggests that EGD may be warranted across all patients with suspected LPR, regardless of presence of concomitant esophageal symptoms. Further, this patient population commonly has a hiatal hernia. Characterizing the integrity of the anti-reflux barrier, including Hill grade of the gastroesophageal flap valve and axial hiatus hernia length, is helpful to understand the mechanisms that are potentially contributing to chronic laryngeal symptoms beyond acid alone, which can help clinicians better understand symptoms and tailor treatment accordingly.^1,3

Strengths of this study include a large multicenter international cohort of well-characterized patients undergoing LPR evaluation. Despite inherent limitations of selection bias with retrospective study design, clinical practice at these sites is to perform objective testing for isolated laryngeal symptoms. Additional limitations include generalizability and a small subgroup with isolated laryngeal symptoms. Further, we were unable to uniformly capture features such as inlet patch or stricture across all sites, and whether patients were on PPI at time of endoscopy was unknown. Future directions include creating an endoscopic score for LPR, correlating EGD findings with treatment outcomes for LPR, and identifying EGD findings that may obviate the need for reflux monitoring.

In conclusion, 47% of patients referred for LPR had notable reflux-associated endoscopic findings. These data highlight the diagnostic yield of EGD in evaluation of chronic laryngeal symptoms, with implications on understanding underlying mechanisms of diseases and determining management. Further, these findings suggest that the decision to pursue EGD in evaluation of chronic laryngeal symptoms should not be based on the presence of esophageal symptoms.

Additional Laryngeal Symptoms Working Group Members on this Study:

Madeline Greytak, BA¹, Alexander M. Kaizer, PhD², Andrew Jenkins, MD^3-4, John E. Pandolfino, MD, MSCI, FACG, AGAF⁵, Vinathi Polamraju MD⁶, Ming-Wun Wong, MD, PhD⁷,

1. Department of Medicine, Division of Gastroenterology and Hepatology, University of California San Diego, La Jolla, California.

2. Department of Biostatistics and Informatics, Colorado School of Public Health, University of Colorado-Anschutz Medical Campus, Aurora, Colorado.

3. Department of Medicine, Division of Gastroenterology, Hepatology and Endoscopy, Brigham and Women’s Hospital, Boston, Massachusetts.

4. Harvard Medical School, Boston, Massachusetts

5. Department of Medicine, Division of Gastroenterology and Hepatology, Feinberg School of Medicine, Northwestern University, Chicago, Illinois.

6. Department of Medicine, Division of Gastroenterology, Washington University, St. Louis, Missouri.

7. Department of Medicine, Division of Gastroenterology and Hepatology, Tzu Chi University, Hualien County, Taiwan.