Effect of esomeprazole with and without a probiotic on fecal dysbiosis, intestinal inflammation, and fecal short-chain fatty acid concentrations in healthy dogs
Abstract
Background
Proton pump inhibitors can cause diarrhea and a transient increase in fecal dysbiosis index in dogs. It is unknown if concurrent probiotic administration mitigates these effects.
Objective/Hypothesis
To assess the fecal Canine Microbial Dysbiosis Index (CMDI), fecal short chain fatty acid (SCFA), and fecal calprotectin concentrations in dogs administered esomeprazole with and without a probiotic.
Animals
Eleven healthy dogs.
Methods
Prospective, within-subjects before and after study. All dogs received 7-day courses of esomeprazole (1 mg/kg PO q 24h) alone followed by esomeprazole with a probiotic (15 billion CFU/kg), separated by a 4-week washout period. Data were compared between phases using mixed effects ANOVA or generalized estimating equations with post-hoc Holm adjustment for 2-way comparisons.
Results
Compared to baseline (mean CMDI −2.66, SD 3.04), fecal CMDI was not different with esomeprazole administration alone (mean CMDI −1.48, SD 3.32, P = .08), but there was a significant increase (Diff 3.05, 95% CI [1.37, 4.74], P < .001, Effect size 2.02) when esomeprazole and a probiotic were administered concurrently (mean CMDI 0.39, SD 2.83). CMDI was significantly higher when esomeprazole was administered with a probiotic than alone (Diff 1.87, 95% CI [0.19, 1.87], P = .02, Effect size 1.24). Fecal calprotectin and SCFA concentrations did not differ between phases. The occurrence of vomiting and diarrhea was not different from baseline when esomeprazole was administered alone (36%/27%) or with a probiotic (46%/9%).
Conclusions and Clinical Importance
In healthy dogs, concurrent administration of a probiotic is unlikely to lessen adverse effects associated with esomeprazole administration.
Abbreviations
-
- CMDI
-
- Canine Microbial Dysbiosis Index
-
- GI
-
- gastrointestinal
-
- PPI
-
- proton pump Inhibitor
-
- SCFA
-
- short-chain fatty acid
1 INTRODUCTION
Proton pump inhibitors (PPIs) are the most effective drugs used in veterinary medicine to raise gastric pH for the treatment of gastroesophageal reflux disorder, esophagitis, and gastroduodenal ulcerations.1-4 However, PPIs are not without adverse adverse effects, the most common being diarrhea and altered intestinal microbial environment. The mechanism by which PPIs cause diarrhea has not been elucidated. After a 15-day course of omeprazole, dogs have an increased fecal Canine Microbiota Dysbiosis Index (CMDI).5 Alterations in the gastrointestinal microbiome can perpetuate mucosal inflammation and ulceration and result in changes in short-chain fatty acid production.6, 7 Short-chain fatty acids (SCFA), especially butyrate, are essential for intestinal health and their reduced production is associated with diarrhea.7 However, the influence of PPIs on SCFA has not been investigated.
The most commonly utilized PPI in veterinary medicine is omeprazole; however, esomeprazole, the s-isomer of omeprazole, results in a more rapid increase in gastric pH and a greater total area under the plasma concentration-time curve compared to the same dose of omeprazole in people.8, 9 Esomeprazole can be administered once daily in people and healthy Beagle dogs and still meet therapeutic criteria described in human medicine for proton pump inhibitors.10 With the advantage of once-daily dosing, esomeprazole is becoming more commonly utilized in veterinary medicine. While it is known that PPIs can alter the gastrointestinal microbiome, it remains unknown if these changes are associated with clinical manifestations such as diarrhea. Regardless, the widespread use of PPIs in veterinary medicine indicates that interventions to help attenuate their adverse effects are needed.11 Modulation of the microbiome with probiotics improves gut barrier function and reduces small intestinal bacterial overgrowth and intestinal inflammation in people receiving PPIs.12-14 In dogs, probiotics are beneficial in both acute and chronic diarrhea.15-20 However, there are no published studies evaluating the effect of probiotic administration on dogs concurrently receiving PPIs.
The purpose of the present study was to evaluate the effect of esomeprazole on the intestinal microbiota, SCFA production, and intestinal inflammation in healthy dogs, as well as the ability of probiotics to ameliorate such effects. We hypothesized that concurrent administration of probiotics in healthy dogs receiving esomeprazole would reduce PPI-induced gastrointestinal dysbiosis such that the fecal Canine Microbial Dysbiosis Index (CMDI) and fecal calprotectin concentrations would be lower, and SCFA production would be higher compared to when they were receiving esomeprazole alone.
2 MATERIALS AND METHODS
2.1 Study sample
Included dogs were determined to be healthy based on a lack of history of GI disease, physical exam performed by an author (RM) at the start of the study, and normal complete blood count and serum biochemistry panel. All included dogs had not received any antibiotics, probiotics, acid suppressant medications, anti-inflammatory drugs, or corticosteroids within the 6 months preceding study enrollment. Diet was not standardized, but all dogs were required to consistently eat their historical commercial diet without changes during the study period. Dogs being fed a raw or home-cooked diet were excluded. Owner consent was obtained, and this study protocol was reviewed and approved by the Institutional Animal Care and Use Committee at the University of Wisconsin-Madison (#V006524).
2.2 Study design
The study was a within-subjects, before and after study, whereby all dogs went through 5 phases (Figure 1). Each phase was 7 days long with the exception of phase III. Phase I consisted of 7 days of no treatment and allowed for the assessment of baseline values. No treatment was elected in place of a placebo (lactose) because it is unknown if lactose, a disaccharide, can change the microbiome through prebiotic effects.21 Phase II consisted of a 7-day course of esomeprazole (dosed at approximately 1 mg/kg by mouth every 24 hours). Phase III consisted of a 4-week washout period. This was chosen based on the knowledge that changes in the fecal CMDI induced by omeprazole resolved within 2 weeks of drug discontinuation and, although the duration of esomeprazole-induced dysbiosis is unknown, it was not expected to be substantially different given that esomeprazole is the s-isomer of omeprazole.22 To determine if 2 weeks was a sufficient washout period as has been shown with omeprazole, phase III was further divided into phase IIIa and phase IIIb. This allowed for fecal samples to be analyzed after 2 (phase IIIa) and 4 weeks (phase IIIb) of esomeprazole discontinuation. Phase IV consisted of a 7-day course of esomeprazole and a probiotic (Visbiome Vet, ExeGI Pharma at approximately 15 billion CFU/kg by mouth every 24 hours). Phase V consisted of 7 days with no treatment. A before and after study design was elected rather than a crossover design given that the duration of effect of Visbiome Vet on the studied variables is unknown.
Esomeprazole was dosed to the nearest capsule size to achieve an oral dose of approximately 1 mg/kg once daily. Esomeprazole is commercially available as 20 mg delayed-release capsules and was compounded by the UW Veterinary Care (UWVC) pharmacy into 10 and 30 mg capsules as needed. Participants were instructed to give the esomeprazole on an empty stomach 30 minutes before feeding. A multi-strain probiotic was dosed to the nearest capsule size to a goal of at least 15 billion CFU/kg based on previously published guidelines for the treatment of canine inflammatory bowel disease as well as manufacturer recommendations.20 The probiotic capsules contained 4 live strains of Lactobacilli (L. plantarum DSM24730TM, L. paracasei DSM24733TM, L. acidophillus DSM24735TM, and L. delbueckii subp. Bulgaricus DSM24734TM), 3 live strains of Bifidobacteria (B. longum DSM24736TM, B. infantis DSM24747TM, and B. breve DSM24732TM), 1 live strain of Streptococcus thermophiles DSM24731TM with microcrystalline cellulose, stearic acid, and magnesium sulfate in a vegetable capsule. This specific probiotic blend is sold under the brand name Visbiome Vet in the USA and Canada. The probiotic was administered simultaneously with esomeprazole as the Visbiome label recommends it be given on an empty stomach and a previous study showed that probiotics administered 30 minutes after a meal did not survive in as high of numbers as those administered 30 minutes before a meal.23
Fecal samples were collected daily on days 5 to 7 of phase I (no treatment), phase II (esomeprazole), phase VI (esomeprazole + probiotic), and phase V (no treatment). Fecal samples were also collected daily on days 12 to 14 of phases IIIa and IIIb during the 4-week wash-out period. All voided fecal samples were collected and refrigerated (4°C) by dog owners immediately after defecation. The samples were delivered to UWVC within 72 hours of collection. After arrival to UWVC, fecal samples were aliquoted and stored at −20°C until analysis. Storage of fecal samples at room temperature for up to 24 hours and 4°C for up to 3 days have minimal effects on the integrity of extracted nucleic acids and the composition of the microbial community in both human and canine stool samples.24-26 Similarly, fecal calprotectin concentrations are stable for up to 3 days at 4°C.27 The SCFA concentrations are stable frozen, and concentrations do not differ when stored at 4°C for up to 3 days compared to frozen samples.28 Finally, 3 mL of whole blood was collected from the jugular or lateral saphenous vein on day 7 of phases I, II, and IV to measure serum gastrin concentrations.
2.3 Canine microbiota dysbiosis index
The canine microbiota dysbiosis index (CDMI), a quantitative PCR-based assay validated for dogs with chronic enteropathy, was determined on fecal samples collected on day 5 of phases I, II, IV, and V as well as day 12 of phases IIIa and IIIb of the washout period.29 Approximately 1 g of feces was used for the CDMI performed at Texas A&M Gastrointestinal Laboratory by use of a PCR-based assay that quantifies 7 bacterial groups (Faecalibacterium, Turicibacter, Streptococcus, E. coli, Blautia, Fusobacterium, and C. hiranonis) and total bacteria into a single number. A dysbiosis index <0 is considered normal, 0 to 2 indicates mild dysbiosis, and >2 indicates severe dysbiosis.29
2.4 Calprotectin
Calprotectin was measured using previously described methods.30 Fecal calprotectin is shed intermittently such that dogs can have variable values within a 3-day period. Therefore, it was measured on samples from days 5 to 7 of phases I, II, IV, and V and days 12 to 14 of phases IIIa and IIIb. Results were recorded as either “normal” or “abnormal” during each phase with abnormal being defined as any value >961 ng/g, the upper limit of the current reference interval used at the Texas A&M Gastrointestinal Laboratory.31
2.5 Short-chain fatty acids
Short-chain fatty acids (acetate, butyrate, propionate, isobutyric, isovaleric acid, and valeric acid) were measured on fecal samples collected on day 5 of phases I, II, IV, and V as well as day 12 of phases IIIa and IIIb during the washout period. Analysis was performed at Texas A&M Gastrointestinal Laboratory using methods previously described.28
2.6 Gastrin
On day 7 of phases I, II, and IV, serum gastrin concentrations were measured to confirm participant compliance. All blood samples were collected after food was withheld from dogs for a minimum of 8 hours. After the blood clotted, it was centrifuged at 1100×g for 10 minutes, and 0.5 mL of serum was separated and stored at −20°C. All samples were shipped together after study completion on dry ice to the Texas A&M Gastrointestinal Laboratory for analysis. Serum gastrin concentrations were measured by an automated chemiluminescent, enzyme-labeled immunometric assay as previously described.32
2.7 Daily logs
Participant owners kept a daily log to record daily appetite, fecal score, and any occurrences of vomiting. Daily appetite was recorded by owners as a percentage estimate with normal baseline being recorded as 100%. Hyporexia was defined as eating ≤50% of daily caloric intake for 2 or more consecutive days. The number of vomiting episodes per day and daily fecal score (Fecal Scoring System, Nestle Purina PetCare Company, St. Louis, Minnesota) were also recorded. If any dog defecated multiple times in a day, their fecal scores were averaged for that day. A fecal score of >4 was considered abnormal.
2.8 Statistical analysis
The number of dogs enrolled was determined using an a priori power analysis. Based on published data evaluating the combined effect of carprofen and omeprazole administration on CDMI, we assumed a mean increase (SD) in CMDI of 5 (1.5) units from baseline to after 7 days of esomeprazole.33 We hypothesized that the mean increase in CDMI from baseline to after 7 days of esomeprazole with a probiotic would be 2 units less, indicating approximately 40% improvement. It was determined that complete data from 11 dogs was needed to have 80% power in a paired t-test to detect this hypothesized difference in treatment effects at a 5% significance level.
The paired differences of fecal CDMI from baseline to after treatment for both treatment phases were estimated and tested vs a null hypothesis of zero by mixed effects ANOVA with phase as a fixed effect and animal as a random effect. For washout phase comparisons, multiple comparisons vs a reference level were adjusted by Dunnett's adjustment. Two-way comparisons were performed between treatment phases and adjusted with Holm's adjustment for multiple testing. Significance level was set at 5%. Secondary analyses of numerical outcomes were analyzed in a similar fashion. Secondary analyses of binary outcomes utilized generalized estimating equations (GEE) with animals as a random effect. GEE analysis was not able to be fit when assessing adverse event rates because of zero instances at baseline. A Fisher's exact test was used instead and 95% confidence intervals were calculated.
3 RESULTS
3.1 Study sample
Twenty employee-owned dogs were prospectively screened for inclusion between January 19th, 2022 through May 11th, 2022. Of the 20 dogs initially screened, 16 dogs were enrolled (Figure 2). One owner failed to bring fecal samples to their screening exam and was excluded for lack of compliance. Three dogs were removed because of abnormalities on screening bloodwork or physical examination. Of the 16 dogs enrolled, 11 of the dogs completed all 5 study phases (Figure 2). One dog was withdrawn during phase I as the owner failed to administer esomeprazole. Three dogs were withdrawn for adverse effects suspected to be related to the study drugs administered. Finally, 1 dog was removed after the development of a cough thought to be unrelated to the study.
For the 11 dogs that completed the study, the median weight and age were 22.4 kg (10.5-37.0 kg) and 5 years (1-9 years), respectively. There were 3 mixed-breed dogs, 2 dachshunds, 2 greyhounds, 1 Rhodesian ridgeback, 2 golden retrievers, and 1 Brittany spaniel. Five dogs were neutered males, 2 were intact males, and 4 were spayed females. All dogs consistently ate their historical diet throughout the study period, and all dogs were eating a commercial diet with an AAFCO adequacy statement for their specific life stage. No dogs received any dietary supplements or medications other than flea, tick, and heartworm preventatives during the study period.
3.2 Adverse effects
Adverse events included diarrhea, vomiting, and self-resolving hyporexia. Adverse events by phase for each study dog are included in Table S1. Three dogs were removed from the study because of diarrhea (1 each during phases II, III, and IV) that required diet change or fiber supplementation. Two of these dogs had concurrent vomiting. For the 11 dogs that completed the study, vomiting was reported in 0 (95% CI 0-32)%, 36 (95% CI 12-68)%, 36 (95% CI 12-68)%, 45 (95% CI 18-75)%, and 9 (95% CI 0-43)% of dogs during phase I, II, III, IV, and V, respectively. Diarrhea was reported in 0 (95% CI 0-32)%, 27 (95% CI 7-61)%, 18 (95% CI 3-52)%, 9 (95% CI 0-43)%, and 18 (95% CI 3-52)% of dogs during phase I, II, III, IV, and V, respectively. Hyporexia was reported in 0 (95% CI 0-32)%, 0 (95% CI 0-32)%, 0 (95% CI 0-32)%, 9 (95% CI 0-43)%, 9 (95% CI 0-43)% of 11 dogs during phase I, II, III, IV, and V, respectively. The occurrence of vomiting, diarrhea, and hyporexia was not different from baseline during any phase.
3.3 Fecal canine microbiota dysbiosis index
For the 11 dogs that completed the study, the mean fecal CMDI for each phase and the number of dogs with a normal dysbiosis index, mild dysbiosis, and marked dysbiosis during each phase can be found in Tables 1 and 2, respectively. Figure 3 shows changes in fecal CMDI by phase for each dog. The mean fecal CMDI at baseline (phase I) was −2.66 (SD 3.04) with 1 dog showing mild dysbiosis with a fecal CMDI of 0.2 and another dog showing severe dysbiosis with a CMDI of 4.6. Administration of esomeprazole alone (phase II) did not result in a significant increase in the fecal CMDI compared to baseline (P = .08). However, an increased fecal CMDI compared to baseline was seen in 10 of 11 dogs (91%), with 2 dogs showing severe dysbiosis. In all dogs, the mean fecal CMDI quickly returned towards baseline after discontinuation of esomeprazole, with the mean fecal CMDI 2 weeks after discontinuation being no different from baseline (P = .84).
| Treatment phase comparisons | Washout phase comparisons | ||||||
|---|---|---|---|---|---|---|---|
| Phase | Mean (SD) | Contrast | P-value* | Phase | Mean (SD) | Diff (95% CI) | P-value** |
| I | −2.66 (3.04) | I vs II | .08 | I | −2.66 (3.04) | Reference | Reference |
| II | −1.48 (3.32) | II vs IV | .02 | IIIa | −299 (3.29) | −0.33 (−1.62-0.97) | .84 |
| IV | 0.39 (2.83) | I vs IV | <.001 | IIIb | −2.83 (3.19) | −0.16 (−1.46-1.13) | .96 |
| V | −2.18 (3.51) | 0.48 (−0.81-1.78) | .67 | ||||
- Note: CDMI values reported as mean (SD) or mean (95% CI). Significance was set at 5%, with significant values bolded. CDMI was calculated for each phase based on the abundances of 7 bacterial groups (Faecalibacterium, Turicibacter, Streptococcus, E. coli, Blautia, Fusobacterium, and C. hiranonis). Phase I: acclimation, Phase II: esomeprazole only, Phase III: washout, Phase IV: esomeprazole and a probiotic, Phase V: no treatment. CDMI was determined on fecal samples collected on day 5 of phases I, II, IV, and V. For the washout period (phase III), CMDI was determined on day 12 of phases IIIa and IIIb. n = 11.
- * P-value of 2-way comparisons with Holm adjustment.
- ** P-value of comparison to phase 1 with Dunnett adjustment.
| Phase | Dogs with CMDI < 0 | Dogs with CMDI 0-2 | Dogs with CMDI > 2 |
|---|---|---|---|
| I | 9 | 1 | 1 |
| II | 9 | 0 | 2 |
| IIIa | 10 | 0 | 1 |
| IIIb | 9 | 1 | 1 |
| IV | 5 | 4 | 2 |
| V | 9 | 1 | 1 |
- Note: For the 11 dogs that completed the study, the number of dogs with a canine microbiota dysbiosis index (CDMI) of <0, 0-2, or >2 during each phase is provided. CDMI was determined on fecal samples collected on day 5 of phases I, II, IV, and V as well as, day 12 of phases IIIa and IIIb. CDMI was calculated based on the abundances of 7 bacterial groups (Faecalibacterium, Turicibacter, Streptococcus, E. coli, Blautia, Fusobacterium, and C. hiranonis) as determined by qPCR methods. A CMDI < 0 indicated no dysbiosis. Dysbiosis was defined as mild if the CDMI was 0 to 2 or severe if the CDMI was >2. Phase I: acclimation, Phase II: esomeprazole only, Phase III: washout, Phase IV: esomeprazole and a probiotic, Phase V: no treatment.
Coadministration of esomeprazole with a probiotic (phase IV) resulted in a significant increase in the fecal CMDI compared to baseline (Diff 3.05, 95% CI [1.37, 4.74], P < .001, Effect size 2.02). When administered esomeprazole with a probiotic, 10 of 11 dogs had an increase in CMDI from baseline with 2 dogs showing mild dysbiosis and 4 dogs showing severe dysbiosis. Coadministration of a probiotic resulted in a significant increase in the fecal CMDI compared to administration of esomeprazole alone (Diff 1.87, 95% CI [0.19, 1.87], P = .02, Effect size 1.24). The fecal CMDI 1-week after discontinuation of esomeprazole was no different from baseline (P = .67).
3.4 Individual bacteria taxa
When individual bacteria taxa were considered, the abundance of Streptococcus and Blautia differed between treatments. Concurrent administration of esomeprazole and a probiotic (phase IV) resulted in an increase in the abundance of Streptococcus spp. (mean 7.34, SD 1.5) compared to baseline (mean 5.36, SD 1.79, P = .01) and esomeprazole administration alone (mean 5.75, SD 2.08, P = .01). The abundance of Streptococcus spp. in dogs receiving esomeprazole alone was not different from baseline (P = .45). The abundance of Blautia spp. was decreased when esomeprazole was administered alone or with a probiotic compared to baseline (P = .01 and P = .01, respectively).
3.5 Fecal calprotectin
Fecal calprotectin concentrations were not significantly different from baseline when esomeprazole was administered alone or with a probiotic (P = .32, P = .32, respectively). At baseline, 2 dogs (18%) had an abnormal fecal calprotectin. Four dogs (36%) in phase II and 4 dogs (36%) dogs in phase IV had an abnormal fecal calprotectin concentration.
3.6 Fecal short-chain fatty acids
Mean fecal acetate concentrations were 252 (SD 78) umol/g, 200 (SD 68) umol/g, and 203 (SD 67) umol/g during phase I, II, and IV, respectively, and did not differ between phase I and phase II (P = .32), phase I and IV (P = .32), or phase II and IV (P = .94). Mean fecal butyrate concentrations were 50.6 (SD 29.1) μmol/g, 43.6 (SD 17.7) μmol/g, and 39.1 (SD 18.6) during phase I, II, and IV, respectively, and did not differ between phase I and phase II (P = .90), phase I and IV (P = .66), or phase II and IV (P = .90). Mean fecal propionate concentrations were 135 (SD 71) μmol/g, 123 (SD 66) μmol/g, and 115 (SD 63) μmol/g during phase I, II, and IV, respectively, and did not differ between phase I and phase II (P = .99), phase I and IV (P = .97), or phase II and IV (P = .99). Mean fecal isobutyric acid concentrations were 4.15 (SD 1.59) μmol/g, 6.12 (SD 3.83) μmol/g, and 5.54 (SD 3.77) μmol/g during phase I, II, and IV, respectively, and did not differ between phase I and phase II (P = .23), phase I and IV (P = .40), or phase II and IV (P = .59). Mean fecal isovaleric acid concentrations were 5.31 (SD 2.21) μmol/g, 8.47 (SD 5.21) μmol/g, and 8.50 (SD 6.28) μmol/g during phase I, II, and IV, respectively, and did not differ between phase I and phase II (P = .18), phase I and IV (P = .18), or phase II and IV (P = .99). Mean fecal valeric acid concentrations were 0.71 (SD 1.45) μmol/g, 1.25 (SD 2.22) μmol/g, and 0.71 (SD 1.93) μmol/g during phase I, II, and IV respectively, and did not differ between phase I and phase II (P = .81), phase I and IV (P = .99), or phase II and IV (P = .81). There was no difference between fecal short chain fatty acid concentrations (acetate, butyrate, propionate, isobutyric, isovaleric acid, and valeric acid) between treatment phases.
3.7 Gastrin
Serum gastrin concentrations were significantly higher when dogs were administered esomeprazole either alone (phase II mean gastrin 54.5, SD 44.5; Diff 43.6, 95% CI [7.9, 79.3], P = .04, Effect size 1.09) or in combination with a probiotic (phase IV mean gastrin 79.1, SD 53.4; Diff 68.2, 95% CI [32.5, 103.9], P = .01, Effect size 1.70) compared to baseline. Serum gastrin concentrations were not different between phase II and phase IV (Diff 24.6, 95% CI [11.1, 60.3], P = .17, Effect size 0.62). Three dogs receiving esomeprazole did not exhibit the expected increase in serum gastrin concentration. Two dogs did not have an increased serum gastrin concentration in phase II, but had an increase in phase IV. One dog did not have an increased serum gastrin concentration in phase IV, but had an increase in phase II.
4 DISCUSSION
Diarrhea is the most common adverse effect associated with PPI administration in veterinary medicine, however, the mechanism is unknown. We evaluated the effect of esomeprazole on the CMDI, fecal calprotectin concentrations, and fecal SCFA concentrations in healthy dogs to look for changes that could contribute to development of adverse effects. Furthermore, we investigated whether coadministration of a probiotic could ameliorate these effects.
In this study, the fecal CMDI increased when esomeprazole was administered with a probiotic (phase IV) as compared to baseline (phase I). The fecal CMDI is a qPCR assay validated in dogs with chronic enteropathy.29 It quantifies the abundance of 7 different bacterial taxa within the feces and enumerates them as a single number. While most dogs showed an increase in fecal CMDI when administered esomeprazole alone, this was not statistically significant compared to baseline, and most dogs had a fecal CMDI that remained within the reference range (<0). Given that the fecal CMDI was highly variable in dogs at baseline, each dog was compared to itself, and despite 1 dog starting out with evidence of significant dysbiosis, this dog still showed the same pattern of an increasing fecal CMDI during both phase II and phase IV. In all dogs, the fecal CMDI returned to baseline within 2 weeks of esomeprazole discontinuation. Our findings are in contrast to a previous study where omeprazole induced a mild and reversible fecal dysbiosis in healthy dogs.22 It is likely that our study was underpowered to detect a significant difference. When all dogs that completed phase II (n = 15) were included in the analysis, a significant increase in the CMDI was reported, supporting a type 2 error. Furthermore, the effect of PPIs on the gastrointestinal microbiome is thought to result, at least in part, to their effect on gastric pH. Although once-daily esomeprazole dosing has been shown to meet the pH guidelines in healthy beagles, this might not be the case in a more heterogeneous sample of dogs being fed a nonstandard diet.10 Communications with those routinely using esomeprazole in patients state that while once-daily dosing appears to be sufficient for some dogs, others have an improved response when increased to twice daily dosing (personal communication, E.N. Gould). Genetic factors such as variability in CYP450 expression, could lead to differences in PPI metabolism between dogs. Consequently, the effects of esomeprazole on the microbiome in our study might have been more consistent among study dogs if twice daily dosing was used.
In the current study, concurrent administration of a probiotic with esomeprazole led to a significant increase in the fecal CMDI when compared to baseline and administration of esomeprazole alone. This increase in fecal CMDI was driven by a higher fecal abundance of Streptococcus species. Streptococcus spp. are 1 of the 7 bacteria species included in the canine fecal CMDI index because they have been shown to be significantly higher in dogs with inflammatory bowel disease and histologically confirmed chronic inflammatory enteropathy.19, 29, 34 Streptococcus spp. are also transiently increased in dogs with acute diarrhea.35 However, the probiotic (Visbiome Vet) administered to the dogs in the current study contains several bacteria including Streptococcus thermophilus, which could have contributed to the increased fecal CMDI seen in dogs receiving esomeprazole with a probiotic. This is supported by 2 recent studies. In 1 study, administration of Visbiome to dogs undergoing a hemilaminectomy resulted in an increase in the mean fecal CMDI, driven by Streptococcus, although there was no evidence of a significant time by treatment interaction.36 Similarly, healthy dogs given a symbiotic (containing Lactobacillus spp., Enterococcus faecium, and Streptococcus salivarius subsp. thermophilus) were found to have transient increases in Enterococcaceae and Streptococcus spp.37 The increase in Streptococcus spp. abundance seen with co-administration of esomeprazole and a probiotic was transient, with abundances returning to baseline within 5 days of discontinuing of treatment.
As our study design required dog owners to administer the study medications, we evaluated serum gastrin levels to confirm compliance of PPI administration. Although serum gastrin concentration was statistically different in our study when phase II and phase IV were compared to phase I, there were 3 dogs that did not have the expected increased serum gastrin concentrations while receiving esomeprazole. These dogs were still included given that gastrin concentrations do not always increase with administration of acid suppressants, as has been previously shown in healthy dogs receiving famotidine and omeprazole. In 1 study, 7 of 8 dogs receiving omeprazole had 1 day where their serum gastrin concentration was within the reference range, and 2 dogs had normal gastrin concentrations on all days gastrin was measured.38
Probiotics have been found to reduce the duration of acute diarrhea in shelter dogs and decrease the expression of enterotoxin genes in dogs with acute hemorrhagic diarrhea syndrome.15-18 Our results did not show a benefit to probiotic administration in this group of healthy dogs when administered concurrently with a PPI. There was no difference in the fecal CMDI of dogs undergoing hemilaminectomy that received intra-operative antibiotics with or without probiotic (Visbiome).36 Another recent study that evaluated dogs receiving an NSAID with and without a probiotic found that probiotic administration did not have an effect on the fecal CMDI or fecal calprotectin concentrations.31 Probiotic administration in dogs with chronic diarrhea can have beneficial effects, including upregulation of tight junction protein expression and normalization of dysbiosis, our results did not show a benefit of probiotic administration in dogs administrated PPI as measured by the fecal CMDI.19, 20
Calprotectin is a protein complex found primarily in neutrophils and is used as a marker of intestinal inflammation in people.39 Calprotectin can be increased in dogs with inflammatory bowel disease and chronic diarrhea.39, 40 Our results did not show a difference in fecal calprotectin among phase I (baseline), phase II (esomeprazole administration), and phase IV (esomeprazole and Visbiome Vet administration). This could indicate that PPI administration does not cause intestinal inflammation within healthy dogs or that the administration period was not long enough to induce a statistically significant change. In dogs administered omeprazole concurrently with carprofen, concentrations of fecal calprotectin were higher compared to dogs only given carprofen.33 It is likely that coadministration of omeprazole with carprofen has a synergistic effect that promotes inflammation. However, as the effect of omeprazole alone on fecal calprotectin concentrations in dogs has not been evaluated, it is possible that omeprazole could induce more intestinal inflammation than esomeprazole or our sample size was not large enough to detect a difference.
Changes to the microbiota can lead to changes in short chain fatty acid (SCFA) production which could lead to diarrhea. Synthesized by colonic microbiota from nonabsorbed macronutrients, largely carbohydrates, SCFAs are absorbed by colonic epithelial cells where they enhance colonic fluid and electrolyte absorption.41 Dogs administered omeprazole concurrently with carprofen had a lower abundance of SCFA-producing bacteria compared to dogs receiving carprofen alone.33 Our results also showed a mild, but significant decrease in the SCFA-producing bacteria Blautia spp. when esomeprazole was administered alone and concurrently with probiotics, although there was not a statistically significant difference in SCFA concentrations. This could reflect that although statically significant, the mild decrease in the abundance of Blautia spp. was not large enough to cause a change in SCFA production. In addition, it is possible that a concurrent increase in SCFA-producing bacteria, such as Lactobacillus spp., counteracted this decrease. This is supported by the finding that Lactobacillus spp. increase in the feces and duodenum of male dogs receiving omeprazole.5 Our ability to assess this possibility was limited by our use of qPCR methods that only targeted core bacteria of the canine gastrointestinal microbiome.
Although several bacteria included in the probiotic mixture including Streptococcus spp., Lactobacillus spp., and Bifidobacterium spp. are associated with the production of SCFA, an increase in fecal SCFA concentrations was not observed with probiotic administration in this study.32 In people receiving a Lactobacillus plantarum probiotic, fecal SCFA concentrations increase when measured after 2 weeks of probiotic administration.42 Fecal SCFA concentrations in the current study were assessed after 5 days of probiotic administration, which might not have allowed enough time for SCFA production. Furthermore, we cannot determine how well the bacteria administered in the probiotic colonized the gastrointestinal tract. They simply could have passed through in the feces, resulting in higher fecal abundances of the bacteria included in the probiotic mixture. An increase in Streptococcus spp. abundance was detected with probiotic administration in the current study; however, the qPCR methods utilized did not target Lactobacillus spp. or Bifidobacterium spp. to determine if there was also an in increase in their abundances.
Many dogs in our study vomited or had diarrhea during or after administration of esomeprazole with or without a probiotic. Most of these adverse effects were self-limiting. Despite 8 dogs developing vomiting or diarrhea on esomeprazole alone and 7 dogs developing vomiting or diarrhea with esomeprazole and probiotics, the clinical signs were only severe enough in 3 dogs (1 on esomeprazole alone, 1 during the washout phase, and 1 on esomeprazole and probiotic) to necessitate removal from the study. Hyporexia was only noted when dogs received probiotics concurrently with esomeprazole. The probiotic was administered PO and was not mixed into the food. These findings suggest, that although common, adverse effects associated with esomeprazole can be self-limiting, at least in healthy dogs.
Our study had several limitations. The current study was performed using a small group of healthy dogs that were administered relatively short courses of esomeprazole with and without a probiotic. To accommodate the number of fecal evaluations performed and need for 3 samples for calprotectin analysis, not all analyses could be done on the day 7 sample. Consequently, samples that were submitted for fecal CMDI were taken on day 5 during each study phase which could have affected results if there was not enough time given to see a larger change. Despite our group of dogs being screened to be healthy without evidence of GI disease before inclusion, there was 1 dog with severe dysbiosis present at baseline and another dog with abnormal calprotectin levels at baseline that ideally would have been excluded before enrollment. These dogs remained in our study because they had no clinical signs of GI disease at home and both dogs completed the study without complications.
It was elected to assess changes in the microbiome utilizing the Canine Microbial Dysbiosis Index. However, as this index was originally validated in dogs with chronic inflammatory enteropathy, it might not have been the best way to evaluate microbial changes in dogs suffering acute diarrhea secondary to PPI administration.29 The CMDI has been successfully used in tracking microbiota changes in acute diarrhea and after administration of antibiotics and fecal microbiota transplantation.35, 43 Furthermore, CMDI increases towards dysbiosis with administration of omeprazole alone or in combination with carprofen in healthy dogs, indicating that it can be used to document microbiota changes in dogs receiving PPIs.22, 33 However, the CMDI only accounts for 7 core bacterial groups and therefore did not allow for a detailed assessment of the entire fecal microbiota, which is a large limitation of the current study. With the unexpected finding of a higher CMDI in dogs receiving esomeprazole with a probiotic than esomeprazole alone, it would have been ideal if all bacteria included in the probiotic mixture were measured. The microbiota is a complex community and metagenomics, or deep DNA shotgun sequencing, would have allowed for assessment not only of the bacteria, but also of the archaea, fungi, and DNA viruses that make up the intestinal community. Additional studies are needed to evaluate the impact of probiotics on the gastrointestinal metagenome and metabolome.
ACKNOWLEDGMENT
Funding provided by the University of Wisconsin School of Veterinary Medicine Companion Animal Fund, CA-6, 2022. The probiotic used in this study (Visbiome Vet) was provided by ExeGi Pharma LLC, Rockville, MD. ExeGi Pharma LLC had no involvement in the design or performance of the study, writing the manuscript, or the decision to submit the manuscript for publication. [Correction added after first online publication on 14 October 2023. Acknowledgment section has been modified.]
CONFLICT OF INTEREST DECLARATION
Authors declare no conflict of interest.
OFF-LABEL ANTIMICROBIAL DECLARATION
Authors declare no off-label use of antimicrobials.
INSTITUTIONAL ANIMAL CARE AND USE COMMITTEE (IACUC) OR OTHER APPROVAL DECLARATION
Approved by the Animal Care and Use Committee of the University of Wisconsin-Madison (IACUC #V006524).
HUMAN ETHICS APPROVAL DECLARATION
Authors declare human ethics approval was not needed for this study.
