Tryciclic Antidepressants in Refractory GERD: Poorly Effective Drugs or Wrong Patients?

pressants in the management of refractory sastro-esophageal refl ux disease (GERD) patients. However, few limitations of this study need to be underlined. First of all, they enrolled patients with no or partial response to a single standard dose of PPI and we know that symptom resolution can be achieved in part of them by simply doubling the PPI dose, as suggested by current guidelines, and therefore they should not be considered as true refractory patients ( 3,4 ). Second, the authors used the positive results to one of the two traditional indices of symptom association analysis, that is SI and SAP, to identify EH. Recent studies emphasized the limitations of this method and suggested to apply with caution these refl ux monitoring features to distinguish patients with GERD from those without ( 2,4 ). Th erefore, it has been suggested that a more reliable diagnosis could be obtained by considering only cases positive to both indices ( 5 ). Further, an alternative and more pathophysiological approach has been recently suggested, always with the use of impedance-pH monitoring, which includes the use of modern impedance variables, such as baseline impedance and postrefl ux swallow-induced peristaltic wave ( 6 ), which were found to be more accurate in distinguishing the various forms of GERD ( 7 ). Th is new diagnostic approach has been recently supported by data on the positive correlation between mucosal integrity abnormalities and histo logical refl ux-related changes with BI levels ( 8 ). In conclusion, we believe that studies aiming to evaluate the eff ectiveness of novel therapeutic approaches in refractory GERD should be performed in better selected populations.

pressants in the management of refractory sastro-esophageal refl ux disease (GERD) patients. However, few limitations of this study need to be underlined.
First of all, they enrolled patients with no or partial response to a single standard dose of PPI and we know that symptom resolution can be achieved in part of them by simply doubling the PPI dose, as suggested by current guidelines, and therefore they should not be considered as true refractory patients ( 3,4 ).
Second, the authors used the positive results to one of the two traditional indices of symptom association analysis, that is SI and SAP, to identify EH. Recent studies emphasized the limitations of this method and suggested to apply with caution these refl ux monitoring features to distinguish patients with GERD from those without ( 2,4 ). Th erefore, it has been suggested that a more reliable diagnosis could be obtained by considering only cases positive to both indices ( 5 ). Further, an alternative and more pathophysiological approach has been recently suggested, always with the use of impedance-pH monitoring, which includes the use of modern impedance variables, such as baseline impedance and postrefl ux swallow-induced peristaltic wave ( 6 ), which were found to be more accurate in distinguishing the various forms of GERD ( 7 ). Th is new diagnostic approach has been recently supported by data on the positive correlation between mucosal integrity abnormalities and histo logical refl ux-related changes with BI levels ( 8 ).
In conclusion, we believe that studies aiming to evaluate the eff ectiveness of novel therapeutic approaches in refractory GERD should be performed in better selected populations.
To the Editor: We read with great interest the paper by Limsrivilai et al.
( 1 ) carried out to assess the benefi t of low dosage imipramine (25 mg daily) compared with placebo for 8 weeks in patients with esophageal hypersensitivity (EH) and functional heartburn (FH). Th e authors enrolled non-erosive refl ux disease patients with no or partial response to once-daily proton pump inhibitors (PPIs) and classifi ed them as EH or FH by impedance-pH offtherapy ( 2 ). Imipramine was not more eff ective than placebo in achieving symptomatic relief (37.2 vs. 37.5%, respectively) in the above combined subpopulations, thus reducing the role of tryciclic antide-healthy volunteers ( 9 ), but there have been no studies focusing on the eff ect of CYP2D6 and CYP2C19 on TCAs in treatment of functional gastrointestinal disorders. In our study, we elected to use imipramine based on its profi le to increase threshold of esophageal pain and low dose was chosen given the expected low prevalence of CYP2D6 EM, but we failed to demonstrate the positive eff ects on symptoms improvement. Nonetheless, the current study was limited by lack of CYP2D6 and CYP2C19 genotype assessment and imipramine drug level measurement; thus, the benefi t of dose increment is yet to be determined. Future research with genotypic/ phenotypic polymorphism assessment and drug level monitoring in various ethnic backgrounds should provide an opportunity to optimize the use of TCAs in treatment of esophageal functional disorder. We also agree that other TCAs with better pharmacokinetic profi les and documented benefi ts in other functional gastrointestinal disorders such as amitriptyline and nortriptyline should be further studied ( 10 ).

CONFLICT OF INTEREST
Th e authors declare no confl ict of interest.
our recent study using imipramine in treatment of esophageal hypersensitivity (EH) and functional heartburn (FH) ( 2 ). Dr Savarino's notion on proper patient selection for tricyclic antidepressant (TCA) treatment in our study is very well taken.
First, Fass et al. ( 3,4 ) has demonstrated that patients with gastroesophageal refl ux disease (GERD) who did not respond to once daily dosing of proton pump inhibitor (PPI) had an improvement in heartburn symptoms aft er doubling the dose to twice daily. However, the impedance-pH monitoring to stratify patient population was not performed, and thus patients with true acid refl ux may have been responsible for the treatment response in their studies. It should also be noted that the clinical response to double-dose PPI among patients failing standard dose who have negative refl ux monitoring has not been reported, but we expect the response rate is theoretically low. Th erefore, we only enrolled those who failed once daily dosing PPI as all of our study participants were documented to have no acid refl ux. Nonetheless, we do agree that recruiting patients who failed doubledose PPI may better represent true PPI refractory population noting that the definition of refractory GERD is still not well established.
Second, we concur that the accuracy of either symptom-associated probability index or symptom index alone to identify EH and FH is imperfect especially when there are low refl ux rates ( 5 ). Esophageal baseline impedance (BI) and post-refl ux swallow-induced peristaltic wave (PSPW) have been shown to have better diagnostic accuracy than esophageal acid exposure time in the diagnosis of erosive refl ux disease and nonerosive refl ux disease. Moreover, recently published data from Frazzoni et al. ( 6 ) demonstrated that the PSPW index and the mean nocturnal BI increased the diagnostic yield of impedance pH monitoring of patients with refl ux disease. Th us, the combination of traditional indices of symptom-associated analysis and the new diagnostic tools should permit future studies to better identify proper patient population for each treatment strategy.