Response of Eosinophilic Oesophagitis to Proton Pump Inhibitors Is Associated With Impedance-pH Parameters Implying Anti-reflux Mechanism of Action

Marzio Frazzoni; Leonardo Frazzoni; Nicola De Bortoli; Salvatore Russo; Salvatore Tolone; Elena Arsiè; Rita Conigliaro; Roberto Penagini; Edoardo Savarino

Disclosures

Aliment Pharmacol Ther. 2021;53(11):1183-1189. 

In This Article

Discussion

In the present study, we found considerable PSPW-associated pH increments (ΔpH) in EoE demonstrating the occurrence of oesophago-salivary reflex in this disease. Moreover, PSPW-associated ΔpH was significantly higher in PPI-responsive than in PPI-refractory EoE, similar to what happens in GERD, and was an independent predictor of PPI response, implying anti-reflux mechanism of action of these drugs. These results show that oesophago-salivary reflex represents a defence reaction against reflux in EoE as well as in GERD and that reflux plays a pathogenic role in EoE.

Chemical clearance restores oesophageal pH by means of swallowed saliva delivered by primary peristalsis as elicited by a vagal oesophago-salivary reflex, in turn incrementing salivary flow.[12] In recent years, the role of chemical clearance as a defence mechanism against reflux has been acknowledged in studies assessing the PSPW index,[13] a novel impedance parameter which increases the diagnostic yield of impedance-pH monitoring in GERD,[14] efficiently separating reflux-related from reflux-unrelated, that is, functional heartburn both off[23] and on[22] PPI therapy. Off-PPI, PSPW index does not predict PPI response in GERD[16] or in EoE,[15] whereas on-PPI significant improvement is independently associated with PPI response both in GERD[16] and in EoE.[15] PSPWs stand for swallows following reflux events within 30 seconds, the time interval associated with the lowest probability of random association with reflux episodes.[24] However, PSPWs reveal post-reflux swallows but do not necessarily imply arrival of saliva with its bicarbonate content into the distal oesophagus; indeed, only substantial pH increments can confirm that saliva has been swallowed in response to activation of the oesophago-salivary reflex. Very recently, we showed that PSPWs are associated with considerable pH increments in patients with GERD, quite similar to those detected in healthy controls.[17] In the present study, we found a 1.4 median PSPW-associated ΔpH both in 80 EoE patients and in 80 gender- and age-matched GERD cases; such pH increments can be explained only by the arrival of bicarbonate-containing saliva into the distal oesophagus, confirming that PSPWs reveal occurrence of the oesophago-salivary reflex. PSPWs are assessed after the end of reflux episodes, that is, following volume clearance which removes the bulk of refluxate and increments oesophageal pH in most instances. PSPW-associated pH increments are calculated starting from the onset of PSPWs and are not influenced by nadir pH of reflux events; indeed, nadir pH of reflux events did not differ between PPI-responsive and PPI-refractory cases in both EoE and GERD group. On the other hand, PSPW-associated ΔpH was significantly higher in PPI-responsive than in PPI-refractory cases both in EoE and in GERD, showing that the oesophago-salivary reflex is a key factor of PPI response in GERD and in EoE as well.

In EoE, eosinophil migration is due to eotaxin-3, a chemokine produced as a consequence of genetically determined aberrant Th2 response.[3] In vitro studies carried out in an acidic environment showed that omeprazole blocks eotaxin-3 expression by oesophageal squamous cells from patients with EoE as well as from patients with GERD.[11] Since a pro-inflammatory response due to innate immunity with over-expression of cytokines occurs in GERD,[25] direct anti-inflammatory activity of PPIs has been postulated.[11] PPIs have relatively short (about 1 hour) plasma half-lives and require an acidic environment like the parietal cell canaliculus to be activated. Actually, according to previous[15] and present findings, oesophageal mucosa is rarely acidic during the daytime in patients with EoE, making direct PPI activation within the oesophageal mucosa unlikely. Alternatively, it has been suggested that proton release from eosinophils could acidify oesophageal mucosa; however, this mechanism cannot explain the ongoing efficacy of PPIs when oesophageal eosinophilia is totally or nearly totally abolished.[4–10,15] Last but not least, in EoE, similar high response rates with PPIs and vonoprazan have recently been reported.[26] Vonoprazan is a potassium-competitive acid blocker (P-CAB) rapidly inhibiting gastric acid secretion but with no demonstrated direct anti-inflammatory properties, its efficacy in EoE therefore being attributable to anti-reflux mechanism of action.

Low baseline impedance reflects impairment of mucosal integrity in GERD[27] and in EoE[28] and may be the consequence of repeated reflux episodes with slow recovery.[29] MNBI represents a simple and accurate[20] method for impedance measurement at impedance-pH monitoring and is an efficient predictor of PPI response in patients with reflux symptoms.[30,31] Impairment of the oesophago-salivary reflex as measured with PSPW-associated ΔpH may favour oesophageal mucosal damage by prolonging exposure to acidic reflux episodes; permeability of epithelial barrier to food antigens may then increase, as shown by MNBI values lower than those detected in GERD, determining in genetically susceptible individuals the cascade leading to oesophageal eosinophilia. The role of reflux in PPI-responsive EoE is further highlighted by ΔMNBI values higher than those found in PPI-refractory EoE, showing that oesophageal mucosal damage is more severe in the distal oesophagus where acidity of refluxate is higher and therefore more influenced by PPI or P-CAB anti-secretory action.

Currently, PPIs represent first-line treatment for EoE in clinical practice, being preferred to topical steroids in several countries[8] and inducing clinical and histological response in approximately half of cases[4–6,8,15] as we confirm in the present study. Maintenance treatment is warranted in EoE since the disease recurs after discontinuation of therapy;[32] however, results of 5-year continuous treatment with topical steroids have been somewhat disappointing[33] while PPIs maintained a high efficacy profile at 2-year follow-up.[7] In recent years, concerns about PPI safety have been raised, but current evidence suggests that long-term PPI therapy rarely produces relevant harms;[34,35] nevertheless, prescription is recommended to be limited to persons with defined indications.[34,35] Since PPI treatment for EoE is off label, many general practitioners are reluctant to long-term prescription owing to overall restrictions from regulatory agencies and fearing of possible legal controversies in case of untoward side effects related to a non-approved drug use. Evidence that PPI efficacy in EoE relies on anti-reflux mechanism of action could increase confidence with long-term prescription in PPI-responsive EoE cases.

Limitations of clinical applicability of PSPW-associated ΔpH and ΔMNBI should be acknowledged. Combined assessment of these two new parameters allowed a very efficient separation of PPI-responsive from PPI-refractory EoE as shown by great AUC at ROC analysis, suggesting that impedance-pH monitoring could be regarded as a useful tool to predict PPI response. However, assessment of these two new parameters is time-consuming, requiring careful visual review of impedance-pH tracings and manual calculation, and is far from being automatised by software analysis. For such reasons, and given the costs of impedance-pH monitoring and of unavoidable previous oesophageal manometry,[36] an empiric PPI trial followed by histological assessment to ascertain remission would be preferred in routine clinical practice.

In conclusion, higher efficacy of oesophago-salivary reflex and more severe mucosal damage in the distal oesophagus are associated with histological response of EoE to PPIs, implying that an anti-reflux mechanism of action is most likely.

processing....