Laryngopharyngeal Reflux (LPR) refers to a disease characterized by symptoms, signs, and tissue changes in the aerodigestive upper tract attributed to the retrograde movement of gastric contents.1 LPR has raised mounting concerns due to the persistence of symptoms and its impact on the quality of life of patients.2,3 Moreover, accumulating studies have demonstrated that LPR contributes to the development of numerous pharyngeal voice diseases and even respiratory diseases. In addition, a plethora of LPR patients may develop noticeable pharyngeal symptoms, such as pharyngeal sensation and dysphagia, and laryngeal symptoms, including hoarseness, sore throat, and chronic cough, without any symptoms of esophageal injury (such as acid reflux and heartburn).4 As a survey reported, 10% of outpatients presenting to the otolaryngology department suffered from LPR5 with diverse non-specific clinical manifestations. Currently, LPR patients are clinically screened through clinical manifestations and laryngoscopic findings as per the Reflux Finding Scores (RFS) or the Reflux Symptom Index (RSI). Moreover, the diagnosis of patients with suspected LPR on the scale may be confirmed or excluded by a Proton Pump Inhibitor (PPI) test or 24 h laryngopharyngeal pH monitoring. However, the RSI and RFS scores are poorly specific, highly subjective, and less reliable.6 Additionally, the PPI test causes various side effects, including chronic kidney disease, acute interstitial nephritis, drug interaction with hepatic drug metabolites, Clostridium difficile infection, collagenous colitis, and osteopenia.7 Meanwhile, 24 h laryngopharyngeal pH monitoring, the gold standard for LPR diagnosis, has not been extensively applied given its low sensitivity, high false-negative rate, invasiveness, and high cost.8 As a result, a necessity has arisen in the last few years to identify a sensitive and noninvasive diagnostic approach for LPR, which is expected to be fulfilled by the salivary pepsin test.

The salivary pepsin test is considered the most promising approach for the diagnosis of LPR because of its extremely sensitive, noninvasive, and pragmatic characteristics.9 Nonetheless, a wide range of procedures (time and number) and methodologies (cutoff value and pepsin testing) have been used for saliva sampling, which results in a wide variation in diagnostic findings. Thus far, a consensus has not been reached regarding normal values for salivary pepsin testing. Therefore, a systematic review was conducted by Wang et al. to analyze the diagnostic efficacy of salivary pepsin for LPR.10 This review, however, only included studies until 2017 and lacked explorations of various diagnostic cutoff values.

In this study, a novel systematic review was developed by updating and performing the meta-analysis to identify different cutoff values of salivary pepsin, thus evaluating the diagnostic accuracy of salivary pepsin.

LPR is diagnosed by the results of 24 h Multichannel Intraluminal Impedance and pH (MII-pH) monitoring (a current gold standard), RFS, and RSI. Nevertheless, the symptoms reported in RSI (such as globus pharyngeus, cough, and hoarseness of voice) do not exhibit specificity to LPR and may indicate other laryngeal pathologies, as did endoscopic laryngopharyngeal findings in RFS. In addition, patient-reported symptomatology is inadequate to independently diagnose LPR.29 Although pharyngeal pH analysis alone is insufficiently sensitive, the 24 h hypopharyngeal-esophageal MII-pH probe is a highly specific test. This probe, however, is costly, invasive, and poorly tolerated by patients.30

LPR and Gastroesophageal Reflux Disease (GERD) are considered to be two different diseases, even though they can co-exist. In particular, LPR patients mainly present with laryngopharyngeal symptoms such as pharyngeal itching, burning, dryness, and pain sensation, while GERD patients are manifested by heartburn, nausea, dysphagia, cough, and atypical chest pain. In addition, GERD results from lower esophageal sphincter dysfunction and usually occurs in the supine position at night. Conversely, LPR, attributable to the upper gastroesophageal sphincter, frequently occurs in the upright position during the day, especially during intense physical activity. In terms of diagnosis, laryngoscopy is used for LPR and gastroscopy for GERD.

Clinical preferences comprise empirical treatment with a 3-month trial of medications and changes in diet and lifestyle, in which retrospective positive responses suggest LPR as a diagnosis based on patient symptoms, without a clear, reliable, and less invasive diagnosis. Nonetheless, little is known about the effect of PPI therapy on LPR.31,32 Moreover, long-term administration of PPIs triggers the following risks: osteopenia, chronic kidney disease, acute interstitial nephritis, drug interactions with hepatic drug metabolites, and collagenous colitis.7,33 Growing attention has been attracted to the formal diagnosis of LPR with less invasive and more cost-effective means before treatment.

In addition to the aforementioned experiments, LPR also can be diagnosed by detecting pepsin in saliva. Pepsin, an enzyme discovered in gastric juice, is activated by hydrochloric acid within pepsinogen. Of note, its presence in the upper digestive tract is merely interpreted by the occurrence of reflux.34 Pepsin is an active form of pepsinogen that is a peptidase enzyme secreted by the glandular cells (chief cells) of the stomach, which can digest proteins by hydrolyzing peptide bonds. Moreover, pepsin can damage the mucosal membranes of the structures it contacts when it leaks out of the stomach together with other gastric contents. Additionally, pepsin also causes damage to the epithelial barrier by digesting intercellular junctions (gap junctions). Hence, pepsin is a major cause of laryngopharyngeal mucosal damage caused by reflux35–37 and can be regarded as direct evidence of reflux.

The salivary pepsin test is expected to be a first-line screening test for the simplicity of saliva collection and the low-cost and noninvasive nature of salivary pepsin detection.29

The present study revealed that salivary pepsin had an AUC of 0.67 (95% CI 0.63‒0.71), a pooled specificity of 0.67 (95% CI 0.48‒0.81), and a pooled sensitivity of 0.61 (95% CI 0.50‒0.71), indicating that the diagnostic value of salivary pepsin is not as high as suggested in the available meta-analyses and may be concurrent with a structured questionnaire which demonstrated a sensitivity of 63% and a specificity of 67%.38 Although our results illustrated the low diagnostic accuracy of salivary pepsin, different cutoff values of pepsin in the recruited literature led to inconsistent diagnostic data. Consequently, the predictive value, specificity, likelihood ratio, and sensitivity of different cutoff values were analyzed in the current study. The results of the subgroup analysis elucidated that the 50 ng/mL cutoff value had superior diagnostic data were superior to 16 ng/mL. Thus, setting a higher cutoff value may contribute to higher specificity and then enhance diagnostic data.

The salivary pepsin concentration varies depending on the time of saliva collection during the day.39 Our data identified that the test on saliva collected at multiple time points might elevate diagnostic sensitivity as compared to the test on only a single pepsin sample. Young et al. confirmed that waking hours were the optimal time to measure the presence of salivary pepsin.39 Klimara et al. found that pepsin was detected most frequently in the morning samples.15 Hayat et al. reported that positive saliva samples were most likely to be obtained 1 h after dinner in both the control and patient groups.24 Overall, the sensitivity of the salivary pepsin test can be improved by increasing the frequency of sampling and sampling in the presence of symptoms. Further research should be performed to probe the role of repeated saliva samplings and the optimal cutoff point.

This meta-analysis, nevertheless, has several limitations. First, some of the literature published online may have been overlooked despite our maximum efforts to retrieve relevant literature. Second, the heterogeneity was high (sensitivity: I² = 86.09%, p = 0.00; specificity: I² = 86.84%, p = 0.00) in our study. More notably, although meta-regression analyses were carried out in the present study, we only found that the size of the study was a source of heterogeneity and did not identify other sources of heterogeneity. Third, the degree of blinding was insufficient in the included studies. Although blinding of therapists is difficult, blinding of participants and outcome evaluation are required to eliminate trial performance and evaluation bias. Fourth, the small sample size of the trials may have caused inefficient trials. Last, the included studies had variable cutoff values for pepsin, which may result in inconsistent diagnostic data. Therefore, our results need to be verified by additional high-quality studies with different cutoff values and larger sample sizes.

In summary, our meta-analysis unraveled that salivary pepsin had a low diagnostic value for LPR. Furthermore, the optimal cutoff point and repeated saliva sampling were capable of increasing the diagnostic value of salivary pepsin. Nevertheless, the salivary pepsin test has rarely been applied. In this context, in-depth research is merited to determine the optimal approach to detecting salivary pepsin for LPR diagnosis, thus providing a consensus regarding the best timing and minimal threshold for salivary pepsin.

The present review unveils that salivary pepsin has a low diagnostic value for LPR and that the cutoff value of pepsin at 50 ng/mL exhibits superior diagnostic accuracy. The diagnostic value of salivary pepsin might be affected by the established diagnostic criteria. Therefore, the effectiveness of salivary pepsin needs to be further evaluated by longer-term and more rigorous randomized controlled trials. Likewise, better salivary pepsin tests warrant further investigation for the diagnosis of LPR.

The authors declare no conflicts of interest.