Laryngopharyngeal reflux in chronic obstructive pulmonary disease - a multi-centre study

Chronic obstructive pulmonary disease (COPD) is a chronic respiratory disease and one of the leading causes of mortality worldwide [1]. The clinical severity of COPD is determined by comorbidities, one of which is the gastroesophageal reflux disease (GERD) [2‐5]. GERD is a common cause of chronic cough [6] and a potential risk factor for exacerbations of COPD [7‐10]. Frequent exacerbators have a high prevalence of GERD, however approximately 58% of these patients lack typical GERD symptoms [11, 12].

Laryngopharyngeal reflux (LPR) represents an extra-esophageal manifestation of GERD. The reflux of gastric contents is fundamental in both LPR and GERD, but the mechanism and the symptoms of the disorders are distinct [13‐15]. LPR occurs when gastric contents pass the upper esophageal sphincter and usually occurs during daytime in the upright position, while GERD occurs when gastric contents pass the lower esophageal sphincter and takes place more often in the supine position at night-time or during sleep [16]. LPR may be a contributing factor in patients with symptomatic COPD however, there are only a few studies analyzing the impact of LPR in patients with COPD [13, 17, 18]. In a large longitudinal study of COPD patients, self-reported GERD or use of PPIs was associated with a 20–60% increased risk of moderate-severe exacerbations and hospitalized exacerbations during 3 years of follow up [19]. Yet, this study was based on a subjective, self-reported history of a physician’s diagnosis of GERD and studies based on objective evaluations by laryngeal-pharyngeal pH monitoring in a large COPD cohort are still missing. Here, we investigated the prevalence of LPR and explored its association with clinically relevant outcomes of COPD.

We included 193 patients with mild to severe COPD (GOLD I-IV). All patients completed the GERD questionnaire (GerdQ) and the Reflux symptom index (RSI), in order to assess reflux symptoms and the Leicester cough questionnaire, in order to assess life quality disturbance due to cough. Additionally, patients were evaluated for the prevalence of LPR and its association with lung function. The presence of LPR was assessed by trained and certified study nurses. Participants were fitted with a pharyngeal probe (Restech pH probe, Respiratory Technology Corp.) for 24 h. The pH was measured twice per second and was transmitted wirelessly to a data recorder. The thresholds for detecting LPR were for the upright position: RYAN score > 9.41 and for the supine position: RYAN score > 6.81 [20].

A subgroup of 107 patients belonged to a prospective, multicenter study [PREVENT, ISRCTN 45572998 [21, 22]] and was longitudinally assessed for 2 years (median follow-up period of 12 months) for an association between LPR and clinically relevant outcomes of COPD. In this pre-defined cohort, 41 patients had 67 mild acute exacerbations of COPD (AECOPD), defined as an acute worsening of respiratory symptoms leading to a change in medication and 37 patients had 62 severe AECOPD, requiring hospitalization. Among these 107 patients, 34 patients agreed to undergo a second evaluation of LPR after 1 month on PPI treatment (Supplementary Figure 1).

The descriptive characteristics of the patients are presented in Table 1A. Risk categories were defined for 106 of the 107 patients from the pre-defined cohort as follows: 8 in GOLD A, 63 in GOLD B, 4 in GOLD C and 31 in GOLD D.

The median (IQR) upright RYAN score was 37.01 (2.12–186.89) and the median (IQR) supine RYAN score was 2.17 (2.17–5.50). Totally, 65.8% (n = 127) of the patients had LPR, as detected by either a pathologic upright RYAN score (n = 85, 44.0%), or a pathologic supine RYAN score (n = 5, 2.6%), or both upright and supine pathologic RYAN scores (n = 37, 19.2%) (Fig. 1a).

There was no association between upright or supine RYAN score and lung function measurements (FEV₁% predicted, p = 0.076 and p = 0.488, respectively; RV % predicted, p = 0.282 and p = 0.800, respectively; TLC % predicted, p = 0.054 and p = 0.559, respectively; Fig. 2a and b).

In the pre-defined cohort with 2-years follow-up, linear regression analysis revealed that there was no association between the upright RYAN score and COPD severity, as revealed by the 6-min walking test, the BODE index, and lung function (Table 1B). Furthermore, there was no association between the upright RYAN and COPD outcome as revealed by the number of exacerbations during the study (Table 1B). Similar results were also obtained for the supine RYAN score (Fig. 2b). However, there was a significant positive association between the upright RYAN score and age (Beta = 0.239, p = 0.013), and a negative association with GerdQ (Beta = − 0.257, p = 0.008) (Table 1B).

There were no significant differences between patients with positive LPR and patients with negative LPR in GerdQ score (1.9 ± 3.3 vs 2.1 ± 3.3, p = 0.177), Leicester cough score (93.9 ± 40.0 vs 103.3 ± 32.8, p = 0.227) and RSI (9.9 ± 9.0 vs 9.4 ± 9.6, p = 0.905).

Using Cox regression and adjusting the model for age and PPI therapy, we found no effect of upright RYAN score neither on time to exacerbation (Exp(B) 1.325, p = 0.369) nor on time to severe exacerbation (Exp(B) 1.195, p = 0.722).

Within the pre-defined cohort with 2-year follow-up, 34 COPD patients were reevaluated for LPR, after 1 month on PPI treatment. There was a significant decrease in the upright RYAN score after treatment (p = 0.047) but not in the supine RYAN score (p = 0.285) (Fig. 2b). Comparing lung function before and after 1 month on PPI treatment, we found no significant difference in lung function parameters, or in any of the questionnaire scores.

To our knowledge, this is the largest study assessing LPR prevalence in a well-characterized COPD cohort. We assessed LPR by monitoring laryngopharyngeal pH [19] and we could demonstrate that the prevalence of LPR is high in COPD patients (65.8%). This is in agreement with the study of Hamdan et al. [13], where the RSI questionnaire was utilized to determine the presence of LPR in 27 COPD patients and 67% of them scored positive for LPR. In our study, there was no association between LPR and clinically relevant COPD outcomes within a 2-year follow-up period, contrary to the findings of Jung et al. [17], who found an association between the RSI score, the reflux finding score and severe exacerbations in 118 COPD patients. These discrepancies could be attributable to the fact that Jung et al. [17] refrained from objectively analyzing LPR prevalence and their diagnosis was based solely on symptoms. Additionally, as stated by Jung et al. [17], the respiratory symptoms of COPD, such as excess throat mucus, cough, throat clearing, and dysphonia, coincide with the measurements of the RSI questionnaire. Therefore, during an exacerbation of COPD, these symptoms would increase automatically, resulting in an increase in the RSI score, independently of LPR. Jung et al. [17] found no association between RSI, RFS, and COPD severity or any other lung function parameter, except between RFS and residual volume / total lung capacity (%).We found no association between the questionnaire results (GerdQ-Questionnaire, RSI to assess reflux symptoms and the Leicester cough questionnaire to assess lifetime quality disturbance due to cough) and RYAN score results which confirms findings in other studies [23, 24].

Currently, the main treatment for LPR is PPI. We found that in patients receiving PPI for 30 days, there was a significant decrease in the upright RYAN score but no improvement in lung function nor in symptoms, as assessed by the various questionnaires used. Further long-term investigations are required to clarify this finding though these results have been seen with GERD treatment [25] and LPR treatment for 2-months in COPD patients [18]. The findings are also in-line with the recent ERS guidelines on chronic cough, suggesting that PPI therapy is not beneficial for patients with reflux without dyspeptic symptoms [26].

Among the limitations of our study is the short follow-up period after PPI therapy. However, it was possible to determine a significant decrease in the upright RYAN score after 1 month on PPI. Within the strengths of our study is its multicentric design and the large well-characterised COPD cohort included.

LPR is not significantly associated with clinically relevant outcomes of COPD. Treatment with PPI significantly decreased the upright RYAN score without improving lung function. Our results further indicate that the presence or severity of LPR cannot be diagnosed based solely on symptoms and questionnaires.