Purpose: In previous patient-based studies our group has shown that proton pump inhibitors (PPIs) induce a transepithelial leak in gastric mucosa (Mullin, et al. 2008). This leak seemingly occurs without the onset of tissue inflammation or damage as assessed endoscopically. Using rat gastric corpus mucosa in Ussing chamber studies, we observed that PPI-induced leak is bidirectional and has a size limit of 10 kDa for solutes than can permeate through it (Murray, et al. 2009). We also observed that whereas the leak allows for charged as well as uncharged molecules to permeate, it exhibits a surprising degree of structural specificity in the types of molecules that can pass (Gabello, et al. 2009). Two small molecule drugs, digoxin and phenytoin, showed surprisingly different patterns, with digoxin able to permeate but phenytoin unable to permeate. We likewise observed that the small peptide bradykinin, can permeate, but the small peptide, oxytocin, cannot (Gabello, et al. 2010). Concerning kinetics, a (standard) 40 mg dose of esomeprazole is required for five days before a statistically significant leak is observed clinically. However, in rat gastric corpus Ussing chamber studies, exposure to 25 μM omeprazole results in instantaneous leak. The current study aimed to assess the long term effect of PPI use on transepithelial gastric leak.
Methods: Subjects were enrolled with up to 20 years duration of PPI usage. To quantify transepithelial leak, subjects drank a concentrated sucrose solution followed by overnight collection of urine with subsequent measurement of sucrose concentration.
Results: We observed that the PPI-induced leak effect decreases spontaneously in the first two years of PPI usage. Most of the decrease in leak is occurring in the first 10 months of PPI use. The phenomenon therefore exhibits tachyphyllaxis, self-correcting in the continued presence of the PPI. A linear regression of the sucrose leak in mg of sucrose vs. months on PPIs for 41 patients shows a declining slope (m=-0.8) with a y intercept of 220 mg and an r2 of 0.07. If the data are grouped as PPI use less than 3 months (n=11) and greater than 3 months (n=30) the mean values for leak are 330 mg and 130 mg respectively (P < 0.0003, Student's t).
Conclusion: While PPI-induced trans-epithelial sucrose leak is present, it appears to decrease as a function of time. Most of this decrease occurs in the first 10 months of usage. Statistical significance of this decrease in leak was demonstrated based on whether patients had been taking PPIs for less than 3 months or greater than 3 months.