Gastro-esophageal reflux disease (GERD) refers to “a condition that develops when the reflux of stomach contents causes troublesome symptoms and/or complications” in the Montreal Classification [1]. As one of the most common gastrointestinal diseases, its prevalence has increased in the last few decades [2, 3], leading to a considerable healthcare burden and low quality of life. Current methods employed in the diagnosis of GERD include GERD questionnaires, “PPI test”, endoscopy and ambulatory esophageal reflux monitoring. However, the sensitivity and specificity of such methods have been questioned. Recent studies showed that the PPI test and structured questionnaires did not obtain ideal sensitivity and specificity for diagnosis of GERD [4–6]. Endoscopy is not adequate since non-erosive reflux disease (NERD) is more prevalent than erosive reflux disease in the GERD population [7]. Reflux monitoring, including esophageal pH metry and impedance-pH metry, although currently used as the available gold standard, has a lower sensitivity in patients with NERD compared to those with reflux esophagitis and lacks reproducibility [8]. After all, none of the current approaches has been proven to be a reliable tool for GERD, and more specific, non-invasive and cost effective diagnostic methods are warranted.

Pepsin, a potential factor contributing to the mucosal tissues when gastro-esophageal reflux (GER) occurs, is a protease originating from pepsinogen synthesized by the gastric chief cells. Pepsin has been found in many different tissue samples such as laryngeal mucosa, paranasal sinus mucosa, saliva, middle ear effusion, tracheal secretions and bronchoalveolar lavage fluid [9–13]. The presence of pepsin in esophagus or more proximal sites is indicative of reflux, suggesting that pepsin may be used as a biomarker for the objective assessment of GERD. Some studies have shown that pepsin detection in the sputum and/or saliva can be regarded as a sensitive, non-invasive method for the diagnosis of the proximal reflux of gastric contents or laryngopharyngeal reflux (LPR) [14–17]. Thus, in the present study, we aimed to determine whether the use of pepsin determination in saliva could be useful for diagnosing GERD in the adult cohort from China with Peptest (RD Biomed Ltd., Hull, UK).

GERD is a common but often misdiagnosed disease in clinical practice. Studies [21–28] showed that GERD was the underlying factor of many pulmonary and otolaryngology diseases and conditions, such as asthma, chronic cough, pneumonia, laryngitis, pharyngitis, hoarseness, and even malignant tumor of head and neck. Because of the nonspecific symptoms of reflux disease, the definitive diagnosis of GERD is still challenging. Currently, the most reliable test for GERD diagnosis is ambulatory 24-h MII-pH monitoring. But it has inherent limitations. First, a considerable proportion of GERD patients could not be detected via the MII-pH metry [29], and dietary modifications and activity restrictions during reflux monitoring may lead to false negative results. Second, there are some asymptomatic cases of reflux who will escape reflux-related monitoring. Third, it is an invasive, expensive technique and a proportion of patients could not tolerate this test. Therefore, the development of a noninvasive, convenient, effective and sensitive method is warranted in a clinical setting.

Pepsin has been proposed as a promising biomarker for GER [24, 30–32]. Several studies [33–37] have indicated that pepsin is a major cause of GER, particularly in weakly acidic and non-acidic reflux. Pepsin may be present in the saliva/sputum of patients during episodes of GER and can keep stable below pH 8.0 [38]. Its presence in the oropharynx suggests GER.

The Peptest test is a convenient, office-based, noninvasive and quick technique for pepsin determination with the use of two unique human monoclonal antibodies to capture pepsin-3b independent of pH, which is superior to the digestion assay [39]. Several studies have shown that the measurement of pepsin in saliva/sputum may be used as a promising tool for diagnosing LPR using the Peptest test [40, 41]. This article is the largest-scale, prospective study in which we assessed the diagnostic value of the Peptest test for objective GERD confirmed by 24-h MII pH monitoring combined with endoscopy. Our study showed that (1) patients with GERD had a higher prevalence of pepsin in saliva and higher pepsin concentration than patients with non-GERD and healthy controls, (2) Postmeal saliva samples collected at the time of symptomatic episodes had a higher positive rate for pepsin and level of pepsin than overall postprandial samples in GERD patients, (3) About 40% of healthy asymptomatic subjects had salivary pepsin with a low concentration of less than 130 ng/mL, (4) Using the cut-off value of 76 ng/mL, the sensitivity and specificity of the Peptest test were determined to be 73.0% and 88.3% respectively.

Kim et al. [14] used the western blot analysis to detect the pepsin in the sputum/saliva in patients with clinically suspected atypical GERD symptoms. The results found that the sensitivity and negative predictive value of pepsin determination were excellent in most of atypical symptom groups (100%), whereas its specificity and positive predictive value were relatively low (76%). Similar researches are few, so the significance of pepsin detection for adult patients with atypical symptoms should be further studied.

To date, three studies have assessed the value of salivary pepsin for GERD with heartburn in adult cohorts using the Peptest test [42–44]. In these three studies, MII-pH metry [43, 44] and pH metry combined with endoscopy [42] were utilized as diagnostic criteria for GERD or reflux-related symptoms. AET with or without reflux number was the only parameter that was used to determine whether reflux monitoring was abnormal, which would result in false negative results. By contrast, we used pH monitoring data (Demeester scores and AET), impedance study (BET and number of all reflux) and endoscopy as the basis for test characteristics, which had a higher sensitivity for detecting GERD.

In the study by Saritas Yuksel [42], salivary pepsin was positive in 12% of controls and 50% of objective GERD (abnormal AET and/or esophagitis). Pepsin assay had a low sensitivity of 50% and a high specificity of 92% using the cut-off of 50 ng/mL with only one saliva sample collected at a random time point. In the published study by Bortoli [43], the Peptest test was positive in 94% of NERD, similar to the results obtained by us, and the pepsin analysis was found to have a sensitivity of 100% and a specificity of 80%, superior to our results. So the authors proposed that salivary pepsin determination was a convenient, economic, accurate and highly specific test to detect GERD without esophagitis.

More recently, Hayat et al. [44] took three salivary samples for each participant on waking, 1 h after lunch and dinner during reflux monitoring period. They found that pepsin could be found saliva in healthy subjects and patients, and the postprandial saliva samples were more likely to be positive for pepsin and have a higher level of pepsin compared to the morning saliva samples. Both the prevalence and concentrations of salivary pepsin were significantly higher in patients with GERD and hypersensitive esophagus (HE) compared to those with functional heartburn (FH) and controls, with higher concentrations predicting a greater probability of reflux. These results were consistent with those obtained by us. However, there were some differences needed to be issued between the two studies. Compared to the study by Hayat et al., the prevalence and median concentrations of pepsin in saliva of GERD patients seemed to be higher in our study (prevalence, 78% vs. 91%; concentrations, 126 (49.7–246.3) ng/mL vs. 153.3 (70–365.7) ng/mL). Our optimal cut-off value of pepsin was lower than that obtained by Hayat et al. (207 ng/mL vs. 76 ng/mL), and the area under the ROC and the Youden index were bigger in the study by us. Namely, the results about the diagnostic value of the Peptest test in our study were superior to those obtained by Hayat. The differences might be explained by some reasons. First, we used MII-pH combined with endoscopy as gold standard to define GERD, which can low the false negative results. Such more “true” GERD patients could be detected with more significant differences on the pepsin concentrations between GERD patients and non-GERD patients. Second, postprandial saliva samples tend to have a higher positive rate of pepsin and level of pepsin compared to the morning waking samples by increased gastric pepsin concentration and volume of reflux after meals [24, 44]. In addition, Kim et al. [14] found that pepsin was significantly more likely to be detected at the time of symptoms than on the waking. In the present study, 27 samples taken at the symptoms after meals from GERD patients had a higher prevalence of pepsin and level of pepsin compared to the overall postprandial samples. Based on the evidences above, we think that postprandial salivary samples when the symptoms occur may have a more powerful ability of differentiating GERD patients from non-GERD patients and we recommend postprandial saliva sampling during the symptomatic phase in the clinical application. Third, the study protocol in our study might be different from that of Hayat.

Pepsin was found in saliva of 40% of healthy control subjects with a low concentration of pepsin as a result of physiological reflux. A higher concentration of pepsin and positive rate is more likely to capture GERD. In our study, all three negative saliva samples suggested a 76% probability of diagnosing non-GERD, whereas a positive samples ≥200 ng/mL suggested GERD with a probability of 94%.

Additionally, our study demonstrated that there were correlations between the level of pepsin and reflux parameters and esophageal motility (LES), suggesting that the pepsin test can be used as an indicator of degree of reflux. The Peptest analysis achieved the accepted sensitivity and specificity for diagnosing GERD. Though far from perfect, its intrigue is that it provides a non-invasive, easy to perform, and inexpensive technique different from the currently available tools.

There were some limitations in our study. First, a small number of healthy asymptomatic subjects were enrolled into this prospective study. Further large-scale population-based study is required to establish an appropriate normal value for healthy subjects. Second, our study did not measure reflux-symptom association analysis and did not classify non-GERD into HE and FH via this parameter because of the lack of definite consensus about the diagnostic criteria for HE and FH. Third, there was no follow-up data to observe pepsin change before and after antireflux therapy and we could not assess its predictive value for clinical outcome.

In summary, as a rapid, convenient, cost effective and non-invasive method, the detection of salivary pepsin had moderate diagnostic value for GERD and may be a promising tool to replace the use of currently invasive tools.