The importance of subgrouping refractory NERD patients according to esophageal pH-impedance testing.

To The Editor: We read with great interest the paper by Frazzoni et al. (1) on the role of robot-assisted laparoscopic fundoplication in the treatment of PPI-refractory GERD patients as diagnosed by on-PPI impedance-pH monitoring. We must congratulate with the authors as they clearly demonstrated the efficacy of anti-reflux surgery in PPI-refractory GERD patients, including a long-term (3years) evaluation to reinforce their findings. They clearly showed that the major therapeutic gain of surgery relies on the reduction of weakly acidic reflux episodes, but we argue that some points of the study need to be discussed more in depth. The authors included in their analysis patients who underwent preoperative on-PPI impedance-pH testing revealing positive symptom association probability (SAP)/symptom index (SI), and/or abnormal oesophageal acid exposure time (AET), and/or abnormal number of total refluxes. Thanks to the clinical application of 24-hour oesophageal impedance-pH, we have previously proposed a subclassification of patients with typical reflux symptoms and normal upper gastrointestinal endoscopy into three different groups: (a) non-erosive reflux disease (NERD) pH-POS patients with normal endoscopy and abnormal distal AET; (b) hypersensitive oesophagus – patients with normal endoscopy, normal distal AET and positive symptom association for either acid or non-acid reflux; and (c) functional heartburn (FH) – patients with normal endoscopy, normal distal AET and negative symptom association for acid and non-acid reflux (2,3). It is not clear whether the authors have totally excluded from their analysis the subgroup of patients with FH, who represent about 25% of those having typical reflux symptoms without any kind of reflux underlying them. They cannot anymore considered within the realm of GERD according to Rome III criteria (4,5) and frequently do not respond to PPIs. It is also obvious that they cannot benefit from surgical anti-reflux therapy. Since the introduction of impedance-pH monitoring in clinical practice, different normal values have been proposed in order to diagnose GERD. The most common used [i.e. the United States (US), Belgian-French (BF) and Italian (ITA) normal values] have different upper limits of normality for distal AET (USA=6.3%, BF=6.2% and ITA=4.2%, respectively) and total number of reflux episodes (USA=73/daily, BF=72/daily and ITA=54/daily, respectively), the two main parameters used to distinguish normal from abnormal GER (6-8). Moreover, a recent paper by Zerbib et al (9) modified the French impedance normal values reducing their upper limit of normality for total reflux episodes to 53 that was almost the same value we proposed several years before (8). In their study, Frazzoni et al. applied normal values that are significantly lower than those mentioned before (AET=3.3%; total number of reflux episodes=45). Given these lower limits of normality, one can hypothesize that some patients have been enrolled and treated despite the presence of a mild or border-line disease, with the risk of weaken the very good outcomes obtained among their PPI-refractory GERD patients. Therefore, we believe that indicating the number and intensity of symptoms reported by their patients during the testing day could be useful in order to justify the choice of a surgical approach and to corroborate the strength of their findings in terms of post-surgical outcome. Whereas the diagnostic utility of impedance-pH monitoring in diagnosing GERD in both patients on- and off-PPI therapy have been extensively demonstrated in several recent studies (10-14), very scant data are available on the real clinical impact of this novel technique in GERD management (15). This is particular true for patients with NERD and normal esophageal acid exposure with positive symptom association to acid and/or non-acid reflux who are characterized by an “hypersensitive esophagus” and represent about half of the NERD patients and 1/3 of the entire GERD population (16). To date, these patients are considered a very difficult task for both gastroenterologists and surgeons in terms of management and treatment, since no effective drugs are available in our pharmacological armamentarium (i.e. PPI and H2 antagonists as well as antacids are not effective in these patients). Thus, we think that the Authors, demonstrating the very good efficacy of anti-reflux surgery in patients with positive symptoms association (64%) and without history of esophagitis (50%), had the remarkable opportunity to emphasize this concept and highlight that an excellent therapeutic chance for this large group of patients actually exists and should be strongly considered, although further studies are necessary to confirm these findings. Moreover, the Authors compared the results of impedance-pH testing performed before and after robot-assisted laparoscopic fundoplication without discussing the fact that the patients were on PPI drugs during the first examination and their results could have been different if the preoperative impedance-pH was done in patients not taking any antisecretory compounds. In fact, it is well known that PPI therapy does not reduce the number of total reflux episodes, but changes the chemical nature of refluxate which becomes predominantly weakly acidic from acid. We are aware that surgical therapy enables to block both acid and weakly acidic refluxes and this certainly affected the positive results obtained by Frazzoni et al (1) on the prevalent control of weakly acidic reflux episodes. We think that it is not fair to compare functional tests on PPIs before and off PPIs after surgery. In conclusion, we believe that this is a very interesting study showing the benefit of surgical therapy in controlling mainly weakly acidic refluxes of patients not responding to PPIs; however, a better characterization of the study population and a more in depth discussion of the main findings would have been useful for the reader.