Address for Correspondence Pieter Janssen, Translational Research Center for Gastrointestinal Disorders, University of Leuven, Herestraat 49, bus 701, BE-3000 Leuven, Belgium. Tel: +32 (0)16 330147; fax: +32 (0)16 345939; e-mail: firstname.lastname@example.org
Background Antidepressants such as mianserin can improve symptoms in some functional dyspeptic patients but their mechanism of action remains unclear. We aimed to assess the effects of mianserin on gastric sensorimotor function in man.
Methods In this randomized, placebo-controlled, double-blind, crossover study 12 healthy subjects (six men) underwent a gastric barostat study and a gastric emptying breath test after 7 days pretreatment with placebo or mianserin (20 mg; p.o.). Graded isobaric and isovolumetric distentions were performed to determine gastric compliance and sensitivity. Subsequently, intrabag pressure was held constant and the volume increase after administration of a liquid meal (200 mL; 300 kcal) was studied. Breath was sampled before and after ingestion of a test meal and half-emptying times for solids and liquids were determined from the breath samples. Mianserin was compared to placebo using t-tests and mixed model analysis (mean ± SD).
Key Results Mianserin did not affect pressures or volumes needed to induce first perception or discomfort. During isovolumetric distensions compliance was decreased after mianserin treatment (1.8 ± 0.4 vs 2.0 ± 0.3 mmHg 100 mL−1; P < 0.05). Premeal volumes were comparable in both treatment arms (221 ± 99 vs 220 ± 88 mL), but meal-induced relaxation during the first 30 min was significantly inhibited after mianserin treatment (F6,40 = 2.58, P < 0.05). Mianserin did not affect either solid or liquid gastric emptying.
Conclusions & Inferences Mianserin does not alter gastric emptying rate or sensitivity to gastric distension, but inhibits gastric accommodation to a meal in its early phase. These observations provide no explanation for the effects of mianserin in functional dyspeptic patients.
Functional dyspepsia (FD) is a complex disorder with a heterogeneous pathophysiology: visceral hypersensitivity in response to gastric distension,1 impaired gastric accommodation to a meal,2 delayed gastric emptying,3 helicobacter pylori infection,4 intestinal motility disorders5 have all been reported to play a role. Moreover, FD symptoms are associated with psychosocial factors, and symptom severity has been correlated with anxiety, anger-hostility, depression, and general psychological distress.6,7
Based on the comorbidity of psychiatric and psychological conditions and given the central processing of visceral pain stimuli antidepressive agents have been proposed as a treatment option in FD.8 Indeed, in controlled trials, tri- and tetracyclic antidepressants such as mianserin and amitriptyline have been shown to alleviate symptoms in FD patients.9–11 Moreover the selective serotonin reuptake inhibitor antidepressant fluoxetine has been demonstrated to be effective in the treatment of FD symptoms.12,13 Despite these reports, the use of antidepressants in the treatment of FD remains controversial: many of the above-mentioned studies are underpowered13,14 while other studies are negative: the selective serotonin and norepinephrine reuptake inhibitor venlafaxine was not more effective than placebo in multi-center, randomized, double-blind, placebo-controlled trial in FD patients.15
There is no general mechanism of action postulated for the antidepressive agents in the treatment of FD. Besides the antidepressive properties of the agents, peripheral working mechanisms were also proposed: in a follow-up study in healthy volunteers amitriptyline was shown to delay gastric emptying and reduce nausea to a nutrient-rich meal but did not affect gastric accommodation or satiation.10 Moreover, for mianserin it was suggested that the clinical efficacy is most likely not solely attributed to the antidepressive properties. Indeed, mianserin is an adrenergic, serotoninergic and histaminergic receptor antagonist.9 Alpha2 and 5-hydroxytryptamine (5-HT) receptors are well-known to mediate gastric motility, gastric emptying, and sensitivity16–19 and might provide a possible working mechanism for mianserin in the treatment of FD.
We hypothesized that mianserin would decrease sensitivity to distension, increase compliance, gastric accommodation to a meal or gastric emptying. The aim of this study was therefore to assess the effects of mianserin treatment on gastric sensorimotor function in healthy volunteers. More specifically, the effect of mianserin was tested on gastric compliance and gastric sensitivity during stepwise distension, on gastric accommodation after nutrient drink ingestion and on gastric emptying of the liquid and solid fraction of a test meal.
All study procedures were performed in the Leuven University Hospital (Belgium) and were approved by the Ethics Committee of the hospital before the start of the study. Written, informed consent was obtained from all subjects, and studies conformed to the Declaration of Helsinki.
General study design and study subjects
This study used a randomized, placebo-controlled, double-blind, crossover design in which remunerated healthy volunteers (HVs) participated in a separate barostat and gastric emptying study. A collaborator that was not involved in the experimental procedure designed a randomization scheme (http://www.randomization.com) so that per volunteer a treatment (placebo or mianserin) was randomly assigned to a visit. This collaborator filled gelatin capsules with either 20 mg mianserin or placebo (lactose) and handed them over to the investigator that was unaware what the content of the pills was. Before each study the HVs were pretreated for 7 days with either mianserin or placebo. The investigator analyzed the results and only at the completion of the study after all the analysis was performed the original randomization scheme was released.
A washout period of 1 week was incorporated between each of the four study visits. Twelve HVs [six men, age: 25 ± 4 years, body mass index (BMI): 21.8 ± 1.9 kg m−2] were recruited and presented themselves without symptoms or a history of gastrointestinal disease, other significant diseases, psychological disorders or drug allergies; none were taking any medication or had any drug history. All volunteers completed the gastric emptying study however one volunteer chose to leave the study after the first barostat experiment. Data from this volunteer were excluded from the barostat analysis.
Barostat study protocol A double lumen polyvinyl tube (Salem sump tube 14 Ch.; Sherwood Medical, Petit Rechain, Belgium) with an adherent plastic bag (1200 mL capacity; 17 cm maximal diameter) finely folded, was introduced through the mouth and secured to the patient’s chin with adhesive tape. The position of the bag in the gastric fundus was checked fluoroscopically.
The polyvinyl tube was then connected to a computer-driven programmable volume-displacement barostat device (Synectics Visceral Stimulator, Stockholm, Sweden). To initially unfold the gastric balloon, the HVs were asked to lay down on a bed while the balloon was inflated with a fixed-volume of 300 mL of air for 2 min, and then the balloon was deflated completely. After a 10-min equilibration period, the patients were positioned in a comfortable sitting position with the knees slightly bent (80°) in a bed, specifically designed for that purpose. The barostat protocol is divided in three consecutive parts:
1Isobaric distention sequence: After a 30-min equilibration period the minimal intra-balloon distending pressure (MDP) was determined as the lowest pressure level that provided an intra-balloon volume of 30 mL or more during a protocol in which the balloon pressure was increased stepwise by 1 mmHg every minute.20,21 The MDP represents the pressure level that equilibrates the intra-abdominal pressure. Subsequently, sequential ramp distentions were performed in stepwise increments of 2 mmHg starting from MDP, each lasting for 2 min, while the corresponding intra-balloon volume was being recorded. Patients were instructed to score their perception of upper abdominal sensations induced by each distending stimulus at the end of every distending step, using a graphic rating scale that combined verbal descriptors on a scale graded from 0 to 6.20,21 The end point of each sequence of distentions was established at an intra-balloon volume of 1000 mL, or when the patients reported discomfort (score 5) or pain (score 6).
2Isovolumetric distension sequence: after a 30-min equilibration period (during which the balloon is deflated), isovolumetric distentions were performed in stepwise increments of 100 mL, each lasting 2 min, while the corresponding pressure was recorded. Similar to the isobaric distension sequence perception scores are recorded. The end point of each sequence of distentions was established at an intra-balloon volume of 1000 mL, or when the patients reported discomfort (score 5) or pain (score 6).
3Meal-induced accommodation: after another 30-min equilibration period, the intra-balloon pressure level was set at MDP + 2 mmHg. Thirty minutes hereafter, the volunteers were asked to drink a nutrient drink (200 mL; 300 kcal; 13% proteins, 48% carbohydrates, 39% lipids; Nutridrink®, Nutricia, Belgium). Intra-balloon volume measurements continued for 60 min.
Gastric emptying study protocol Gastric emptying for solids and liquids was simultaneously assessed using the previously validated 14C-octanoic acid/13C-glycine breath test.3,22 The test meal consisted of 60 g of white bread and one egg, the yolk of which was mixed with 74 kBq of 14C-octanoic acid sodium salt. The meal was ingested within 5 min, followed immediately by 150 mL of water dosed with 100 mg of 13C-glycine. The total caloric value of the test meal was 250 kcal. Breath samples were taken before the meal and at 15-min intervals for a period of 240 min postprandially. Gastric half-emptying time (T1/2) was calculated as previously described.22
All analyses were performed using SAS statistical software version 9.1.3. (SAS Institute, Cary, NC, USA). Data are presented as the mean ± SD. Differences were considered to be significant at the 5% level. Based on several studies performed by our group in which a similar protocol was applied we estimated that with the number of volunteers that were recruited in this study it would be possible to detect a 30% change in pre- and postprandial balloon volumes (with a power of 80%, α = 0.05) and compliance and sensitivity thresholds (with a power or 85%, α = 0.05). For solid and liquid gastric emptying the study had a power of 85 and 90% to detect a 30% change (α = 0.05) respectively.21,23–25
Barostat study During the isobaric and isovolumetric distension sequences, pressure and volume was measured by averaging the recordings during every 2 min distending step. Gastric compliance was calculated as the slope of the pressure–volume or volume–pressure curve obtained from stepwise distensions using linear regression analysis and compared between placebo and mianserin treatment using the Student’s t-test. Perception threshold was defined as the first level of pressure or volume that evoked a perception score of 1 or more. Discomfort threshold was defined as the first level of pressure or volume that provoked a perception score of 5 or more. Perception and discomfort thresholds and compliances were compared between treatments using the paired Student’s t-test.
To evaluate gastric tone, the mean intra-balloon volume was calculated over consecutive 5-min intervals. The first six intervals (30 min) constitute the baseline or preprandial period and the final 12 intervals (60 min) constitute the postprandial period (i.e. after nutrient drink ingestion) (Fig. 1). Meal-induced gastric relaxation at each time interval of the postprandial period was quantified by subtracting the average preprandial volume from the postprandial volume. In order to compare pre- and postprandial relaxation over time between treatments, delta values (relaxationplacebo−relaxationmianserin) were calculated at each time point; these delta values were used as the dependent variable in all analyses described below. Mixed model analysis (using ‘proc mixed’ in SAS) with time as a categorical fixed effect was used to test the hypothesis that this delta-relaxation curve was significantly different from zero over time, this corresponds to a treatment effect over time. Random and/or repeated effects were added to the model in order to account for the between- and within-subject variability structure of the data in the best possible way. A model with random effects intercept and time (as a continuous variable and random slope) and with repeated effect time (as categorical variable) fitted the data best (based on Akaike’s Information Criterion) and was reported.
Gastric emptying: Gastric half-emptying times for both solid and liquid emptying were compared using paired Student’s t-tests.
Influence of mianserin on gastric compliance and sensitivity
The mean MDP after placebo and mianserin treatment was not significantly different (9.0 ± 1.2 vs 9.5 ± 1.4 mmHg; NS). During isobaric distensions, gastric compliance remained unaltered after placebo and mianserin treatment (56 ± 22 vs 52 ± 20 mL mmHg−1 respectively; NS; Fig. 2A). During isovolumetric distensions stomach compliance was marginally but significantly decreased after mianserin treatment (1.8 ± 0.4 vs 2.0 ± 0.3 mmHg 100mL−1 after placebo and mianserin treatment respectively; P < 0.05; Fig. 2B, C).
Perception scores during isobaric or isovolumetric distentions were not affected by mianserin (Fig. 2D, E). The pressure thresholds for first perception (2.4 ± 1.2 vs 2.6 ± 2.2 mmHg above MDP) or discomfort (8.9 ± 2.3 vs 10.0 ± 4.3 mmHg above MDP) during isobaric distentions did not differ significantly after placebo and mianserin treatment respectively. Moreover, the volume thresholds for first perception (142 ± 77 vs 170 ± 82 mmHg above MDP) or discomfort (472 ± 172 vs 540 ± 189 mmHg above MDP) during isovolumetric distentions did not differ significantly after placebo and mianserin treatment respectively.
Influence of mianserin on gastric accommodation to a meal
Preprandial volumes were not different between both groups (221 ± 99 vs 220 ± 88 mL for placebo and mianserin respectively). Overall, the postprandial volume increase was not significantly different after placebo or mianserin treatment. However, the postprandial volume increase in the first 30 min after meal intake was significantly inhibited after mianserin treatment (F6,40 = 2.58, P < 0.05; Fig. 1). After mianserin treatment the maximal postprandial increase in intra-balloon volume was significantly decreased (275 ± 150 vs 210 ± 133 mL for placebo and mianserin respectively; P < 0.05). The time to maximal postprandial volume (26 ± 18 vs 25 ± 21 min for placebo and mianserin respectively) was not affected by mianserin.
Both solid emptying (79 ± 17 vs 96 ± 35 min for placebo and mianserin respectively) and liquid emptying (59 ± 13 vs 56 ± 20 min for placebo and mianserin respectively) were not significantly affected by mianserin pretreatment (Fig. 3).
Although tri- and tetracyclic antidepressants such as mianserin have been shown to alleviate symptoms in FD patients9 their working mechanism remains unclear. In this study we investigated whether effects on gastric sensorimotor functions could help to explain the effect of mianserin in the clinic. Seven days treatment with mianserin did not affect gastric emptying or gastric sensitivity, gastric compliance and accommodation to a meal on the other hand were significantly decreased. The dose of mianserin that was used was lower than the one used in a randomized controlled trial with this drug in FD, but it is in agreement with the lower dose ranges used in clinical practice when managing FD with tri-and tetracyclic agents.9–14
Functional dyspepsia is a heterogeneous clinical syndrome defined by chronic or recurrent upper abdominal symptoms without identifiable cause by conventional diagnostic means.26 Symptoms are related to abnormalities of gastric motility,3,27 visceral hypersensitivity1,28 but also influenced by psychiatric comorbidities, especially depression and anxiety.7,13,29–33 In this perspective, different studies have suggested that antianxiety or antidepressive agents could be beneficial for the treatment of functional gastrointestinal disorders such as FD.9,13,15 Although the beneficial effects of these medications are mainly attributed to their central nervous system activity, most of these agents also have peripheral effects that could help to explain why these agents are effective in FD.10,13,34 The same might be true for mianserin which is successfully used in the clinic to treat FD patients, a practice based on a study from 1996 in which mianserin was shown to be effective in patients with non-ulcer dyspepsia.9 A peripheral mechanism of action of mianserin cannot be excluded especially given its complex pharmacology: mianserin possesses antihistaminic properties and is an antagonist of different serotoninergic and adrenergic receptors (5-HT1D, 5-HT2A, 5-HT2C, 5-HT3, 5-HT6, 5-HT7, α1 and α2 receptors).35,36
Our results indicated that mianserin has no significant effect on liquid or solid gastric emptying in HVs. This finding was in contrast to what we anticipated as 5-HT1, 5-HT2, 5-HT3 and α2 receptors are involved in the regulation of gastric emptying.37–43 We hypothesize that no overall effect of mianserin on gastric emptying could be observed because mianserin blocks receptors that might both accelerate and delay gastric emptying. Our findings are in concordance with other studies that could not show any effect of acute or chronic treatment with tricyclic and tetracyclic antidepressant agents on gastric emptying and motor functions.34,44,45 It has to be remarked though that although the 14C-octanoic acid/13C-glycine breath test has been proposed as a reliable indirect method of measuring gastric emptying of solids and liquids,3,22 not all studies have found a good correlation with gastric scintigraphy, which is regarded as the gold standard for measuring gastric emptying.46,47 An alternative explanation for the lack of effect of mianserin on gastric emptying might be the limited sample size. Although this study was powered to detect a 30% difference and we have previously used similar sample sizes to detect meaningful effects on gastric sensorimotor function and gastric emptying 20,21,23–25 it cannot be excluded that we missed effects because of the limited sample size, and therefore the non-significant differences have to be taken with caution. Further studies that are powered to detect smaller differences are warranted.
Mianserin did not affect gastric sensitivity during balloon distension. Even though sample size might be a limitation, as discussed above, this finding is in concordance with previous studies in which antidepressant agents such as paroxetine and sertraline were shown not to affect gastric sensations.48,49 On the other hand mianserin decreased compliance and impaired gastric accommodation to a meal. Given the wide range of receptors that are influenced by mianserin that have been shown to mediate gastric tone it is difficult to speculate through which receptor this effect is mediated: activation of 5-HT1 and 5-HT7 receptors has been shown to relax the stomach,21,50,51 while also activation of α2 receptors has been shown to relax the stomach.16 Antagonism of either of these receptors might be responsible for the observed effect on gastric compliance and gastric accommodation. It has to be noted that the observed effects were only moderate: gastric compliance during isobaric distentions was not affected by mianserin and, although significant, the question remains whether a decrease of 0.2 mmHg per 100 mL distension is clinically relevant. Moreover gastric accommodation was only significantly affected in the first 30 min after food intake, while no significant effect could be observed when comparing the volume in the 60 min after food intake. And here opposite effects could be responsible: activation of 5-HT2A and 5-HT2B receptors has been shown to contract gastric smooth muscles in vitro52–54 and antagonism of these receptors might oppose the effect induced by antagonism of 5-HT1, 5-HT7 and/or α2 receptors.
Taken together, these findings in HVs do not explain why mianserin alleviates symptoms in FD patients. These results could indicate that the effect of mianserin in patients is mediated through its central nervous system actions. However, our findings in HVs do not exclude that mianserin affects gastric sensorimotor functions and gastric emptying in FD patients. Indeed, symptoms in FD patients are, among others, related to hypersensitivity to gastric distension,1 impaired gastric accommodation to a meal2 and delayed gastric emptying3 and it is not excluded that mianserin alleviates hypersensitivity to gastric distension, improves impaired gastric accommodation or delayed gastric emptying, even though no effect is seen in health. As can be expected from a drug with antihistaminic properties, mianserin has been shown to inhibit gastric secretion.55 Excessive gastric secretion can induce dyspeptic symptoms and inhibition of this secretion might provide another alternative explanation for the effect seen with mianserin in dyspeptic patients.
In conclusion, in this study in HVs with limited sample size we have not found indications that mianserin affects gastric emptying or gastric sensitivity, while gastric accommodation to a meal and gastric compliance were significantly decreased, even though the clinical relevance of the latter is unclear. These observations provide no explanation for the effects of mianserin in FD patients. Research is warranted in which the effect of mianserin on gastric sensorimotor functions and gastric emptying is investigated in FD patients, because the effect of mianserin on impaired gastric sensorimotor function and gastric emptying might be different.
Acknowledgments and Disclosures
Pieter Janssen and Lukas Van Oudenhove are postdoctoral research fellows of the FWO Flanders. This work was supported by an FWO grant and a Methusalem grant to Jan Tack, MD, PhD.
PJ interpreted and analyzed the data and wrote the paper; RV performed the research; LVO and JT analyzed the data, designed the study and revised the manuscript critically for important intellectual content.