Central Neuromodulators in Irritable Bowel Syndrome: Why, How, and When

Abstract

Irritable bowel syndrome (IBS) is responsive to treatments using central neuromodulators. Central neuromodulators work by enhancing the synaptic transmission of 5-hydroxytryptamine, noradrenalin, and dopamine, achieving a slower regulation or desensitization of their postsynaptic receptors. Central neuromodulators act on receptors along the brain-gut axis, so they are useful in treating psychiatric comorbidities, modifying gut motility, improving central downregulation of visceral signals, and enhancing neurogenesis in patients with IBS. Choosing a central neuromodulator for treating IBS should be according to the pharmacological properties and predominant symptoms. The first-line treatment for pain management in IBS is using tricyclic antidepressants. An alternative for pain management is the serotonin and noradrenaline reuptake inhibitors. Selective serotonin reuptake inhibitors are useful when symptoms of anxiety and hypervigilance are dominant but are not helpful for treating abdominal pain. The predominant bowel habit is helpful when choosing a neuromodulator to treat IBS; selective serotonin reuptake inhibitors help constipation, not pain, but may cause diarrhea; tricyclic antidepressants help diarrhea but may cause constipation. A clinical response may occur in 6–8 weeks, but long-term treatment (usually 6–12 months) is required after the initial response to prevent relapse. Augmentation therapy may be beneficial when the therapeutic effect of the first agent is incomplete or associated with side effects. It is recommended to reduce the dose of the first agent and add a second complementary treatment. This may include an atypical antipsychotic or brain-gut behavioral treatment. When tapering central neuromodulators, the dose should be reduced slowly over 4 weeks but may take longer when discontinuation effects occur.

INTRODUCTION

Irritable bowel syndrome (IBS) is a disorder of gut-brain interaction (DGBI) characterized by abdominal pain related to defecation and changes in the frequency and form of stool (1,2). The predominant bowel habits are classified into subtypes, with a predominance of constipation (IBS-C), a predominance of diarrhea (IBS-D), mixed bowel habit (IBS-M), and unclassifiable (IBS-U) (2).

Like the other DGBI, IBS is frequently associated with neuropsychiatric disorders such as depression and anxiety, which are considered triggers for the onset of symptoms or occur in response to having them (3). In the Rome Foundation global study that included 54,127 participants, subjects with psychological distress or clinically relevant somatic symptoms were 4.45 times more likely to have 1 or more DGBI than those without psychological distress. The same study reported that those who met specific criteria for bowel disorders presented clinically relevant psychological distress or somatic symptoms in 55.5% of cases (4). In addition, in a meta-analysis that included 7,095 subjects with IBS exclusively, the global prevalence of depression was 36%: 38% in IBS-C, 37% in IBS-D, 34% in IBS-M, and 22% in IBS-U. Anxiety was present in 44% of patients with IBS, 47% in IBS-C, 37% in IBS-D, 37% in IBS-M, and 11% in IBS-U (5).

Based on this association, it would be logical to think that using neuromodulators could be of value in treating these psychological comorbidities. However, the value of neuromodulator treatment in IBS goes beyond its association with neuropsychiatric disorders (6). Neuromodulators act directly in regulating different pathophysiological mechanisms involved with visceral or central hypersensitivity, alterations of intestinal transit, or neurogenic effects, all of which are responsible for the clinical manifestations of IBS (6,7). This approach allows us to choose therapeutic alternatives based on symptom intensity and the nature of the symptoms (e.g., pain, diarrhea, constipation) treated (6). Neuromodulators are frequently prescribed after an unsatisfactory response to peripherally targeted treatments including diet, antispasmodics, microbiota modulation, laxatives or prokinetics, and antidiarrheals (depending on the subtype). We propose they can be prescribed initially based on the severity of the symptoms (e.g., predominant pain) or the coexistence of anxiety or depression. Because neuromodulator treatment is still considered off-label, many of the recommendations herein are based on expert consensus (6) and the experience of the senior author. Further studies are needed to confirm the empirically derived information.

In this article, we will review several relevant publications and provide our perspective on using neuromodulators in treating IBS to answer why, how they act, and when to use them. Although the focus is on treating IBS, this knowledge can also be applied to treating other DGBI. See Video 1 in the Supplementary Materials (Supplementary Digital Content 1, http://links.lww.com/AJG/D247) for a comprehensive review of the use of neuromodulators in IBS.

WHY USE NEUROMODULATORS IN IBS?

Neuromodulators (previously called antidepressants, antianxiety, or antipsychotic medications) act on receptors along the brain-gut axis (BGA) and affect brain-gut function relating to motility, secretion, and visceral and central neural signaling (6). Since IBS and other DGBI are caused by dysregulation of the BGA (see below), neuromodulators serve to normalize their action. Specifically, they have effects to (i) treat psychiatric comorbidity, (ii) modify gut motility, (iii) improve central downregulation of visceral signals, and (iv) enhance neurogenesis (6–8).

Some neuromodulators act peripherally in the enteric nervous system, including 5-hydroxytryptamine (5-HT) 4 receptor agonists, 5-HT3 receptor antagonists, guanylate cyclase C agonists, opioid receptor agonists/antagonists, delta ligand, and or muscarinic receptors antagonists. However, in this review, we will focus on centrally acting neuromodulators, particularly tricyclic antidepressants (TCAs), selective serotonin reuptake inhibitors (SSRIs), serotonin and noradrenaline reuptake inhibitors (SNRI), tetracyclics (TNAS), and atypical antipsychotics agents (AAPs), which act on various specific receptors as shown in Table 1 (6). Their actions help improve symptoms of pain or discomfort, bloating, nausea, and vomiting. They can also have peripheral effects on motility and secretion (6,8). Table 1 reviews the central neuromodulator receptor activity that affects these symptoms (6).

Table 1.

Activity of central neuromodulators antidepressants on specific receptors

Brain-gut axis

The BGA is a bidirectional complex connecting the central nervous and digestive systems. It comprises the brain, spinal cord, autonomic nervous system (which includes the sympathetic, parasympathetic, and enteric nervous systems), neuroendocrine, and neurohumoral systems (6,8,9).

The afferent nerve signals involved in the transmission of pain reach the brain through a chain of 3 orders of neurons from the primary spinal afferent arising from the gut to second-order neurons from the dorsal horn of the spinal cord to the thalamus and then to the midbrain (6,10).

The brain nuclei involved in regulating visceral pain include the nucleus of the solitary tract, parabrachial nucleus, locus coeruleus, rostral ventromedial medulla, anterior cingulate cortex, paraventricular nucleus, and the amygdala (11). Once these signals are registered centrally, the brain can modify incoming visceral signals through descending modulation via the gate control mechanism. This is related to the activity of descending brainstem fibers, which can affect the sensitivity of the dorsal horn neurons. This mechanism can alter visceral sensitivity and central control of pain perception through the involvement of 2 neurotransmitters, serotonin or 5-HT, and noradrenalin (NA), the main targets of drug treatment (6,10). See Video 2 in the Supplementary Materials (Supplementary Digital Content 1, http://links.lww.com/AJG/D247) for further information on how to explain the BGA to a patient.

The BGA also interacts with the intestinal microbiota's composition and the intestinal barrier's permeability, so much so that some authors use the term brain-gut-microbiota axis (9,12). The microbiome as part of the biopsychosocial conceptual model of the DGBI (Figure 1) can influence central nervous system function. The metabolome of an eubiotic microbiota contributes to the functioning of the healthy brain since some of its components can cross the blood-brain barrier (9,13,14). Tryptophan produced in the gut by the enterochromaffin cells, crosses the intestinal barrier, and is transformed into 5-HT after crossing the blood-brain barrier; 5-HT is crucial in the development and functioning of the microglia. Short-chain fatty acids, which are metabolic products of the microbiota, are directly involved in strengthening the tight junctions of intestinal barrier cells (9,13).

Figure 1.

Biopsychosocial conceptual model of disorders of gut brain interaction that integrates early-life factors with the central nervous system and enteric nervous system leading to the clinical expression of disorders of gut brain interaction. CNS, central nervous system; DGBI, disorder of gut-brain interaction; ENS, enteric nervous system. With permission from Drossman DA. Functional gastrointestinal disorders: History, pathophysiology, clinical features and Rome IV. Gastroenterology 2016;150(6):1262–79.e2.

Another aspect that supports the relationship between depression and the intestinal microbiota is the data showing that the microbiota composition differs significantly between healthy subjects and patients diagnosed with depression (15,16). In addition, a recent meta-analysis found changes in the composition of gut microbiota after the administration of antidepressants (17), and preliminary evidence suggests a possible effect of Bifidobacterium longum on improving psychological scores in patients with IBS (18); however, further research is needed.

A critical component of the BGA is the autonomic nervous system, which involves the enteric nervous system, acting as a mediator of the visceral response to central influences (19). There is evidence to suggest the presence of autonomic function disorders in patients with DGBI, which supports the presence of decreased or increased vagal outflow or sympathetic activity. Disturbance of autonomic balance can also alter visceral perception and be involved in pain perception; for these reasons, autonomic dysfunction could be a pathophysiological mechanism involved in many of the symptoms of IBS and contribute to associated factors such as sweating, cardiac arrhythmias, or alterations in the respiratory cycle (19,20), including their co-association with autonomically driven disorders like postural orthostatic tachycardia syndrome (21).

Figure 1 shows the biopsychosocial conceptual model demonstrating the influence of early-life factors on the BGA and its psychophysiological relationships leading to the clinical expression of IBS and other DGBI.

Neuromodulator effects on depression and pain: the monoamine hypothesis

The monoamine hypothesis proposes that depression results from a deficiency of 1 or more of 3 monoamines, which are 5-HT, NA, and dopamine. According to this hypothesis, antidepressants work by enhancing the synaptic transmission of these monoamines, achieving a slower regulation or desensitization of their postsynaptic receptors (Figure 2) (6).

Figure 2.

Hypothesized mechanism of action of antidepressants. Most currently available antidepressants work by blocking the presynaptic reuptake pump of 1 or more of the 3 main monoamine neurotransmitters (serotonin, noradrenaline, and dopamine), causing the respective neurotransmitter to accumulate in the synaptic cleft (a), which in turn leads to a delayed downregulation or desensitization of postsynaptic receptors for the respective neurotransmitter (b). From Drossman DA, Tack J, Ford AC, Szigethy E, Törnblom H, Van Oudenhove L. Neuromodulators for functional GI disorders (disorders of gut-brain interaction): A Rome Foundation Working Team Report. Gastroenterology 2018;154(4):1140–71.e1.

The same hypothesis can explain the activity of antidepressants on the perception of visceral pain through the BGA previously described since their monoaminergic actions interfere with the activity of brain circuits related to pain, emotions, anxiety, and cognitive ability (6,22). Antidepressants also interfere with nociceptive transmission mechanisms at the dorsal horn of the spinal cord, modulating the transmission of afferent pain. The increase of NA in the spinal cord by inhibiting reuptake directly inhibits pain through α₂-adrenergic receptors; in addition, NA acts on the locus coeruleus and improves the function of a descending noradrenergic inhibitory system (23). Brain regions, including the amygdala and anterior cingulate cortex, control the descending pathways from top to bottom. These projections are noradrenergic and serotonergic, so antidepressants can act by directly modulating these processes (6,22).

Neuroplasticity

Neuroplasticity refers to the brain's remarkable ability to reorganize itself by forming or losing new neural connections throughout life. It involves the brain's ability to adapt and change in response to experiences, learning, environmental influences, injury, and other factors. Neurodegeneration refers to the loss of cortical neurons with chronic pain, traumatic life events, and psychiatric disease, and neurogenesis of neurons refers to new growth and neural connections as may occur with clinical treatment. This new concept helps us understand how antidepressants can improve gastrointestinal (GI) symptoms (6). Central nervous system neurons are plastic and capable of new growth in regions such as the hippocampus and can die after severe psychological trauma, and this is associated with developing post-traumatic stress disorder or chronic pain as in IBS; reduced cortical density after trauma is seen in other brain regions involved in emotional and pain regulation and relevant here to pain control regions such as the cingulate cortex, in chronic and painful GI conditions like IBS (6,24).

Antidepressant treatments appear to increase precursor neuronal growth following reduced cortical density from traumatic experiences; brain-derived neurotrophic factor levels, a precursor of neuronal growth, increase with antidepressant treatment, which correlates with longer treatment periods and the degree of recovery from depression (6,25,26).

Furthermore, the longer patients are treated with antidepressants, the lower the frequency of relapse or recurrence of the depression. This may help explain why these treatments have more than immediate effects of symptom reduction; over time, they may help rewire the brain to approach a premorbid functioning state (6,27).

ACTION OF CENTRAL NEUROMODULATORS DETERMINED BY CLASS OF AGENT

Tricyclic antidepressants

The TCAs are central neuromodulators whose mechanism of action is by presynaptic 5-HT and NA reuptake inhibition in combination with additional antagonistic properties on postsynaptic 5-HT2A, 5-HT2C, 5-HT3, muscarinic 1, histamine receptor (H) 1, α-noradrenalin receptor (α) 1, and presynaptic α2 NA receptors (6). The antimuscarinic effect of TCAs is associated with constipation due to decreased intestinal motility. This condition can also be related to the inhibitory action of these agents on transient receptor potential channel canonical type 4 in colonic myocytes that disrupt colonic motility (6,28).

While historically, TCAs have been used in psychiatry to treat depression, newer agents such as SSRIs and SNRIs have largely supplanted their use (22). However, TCAs are also prescribed in low dosages to treat painful conditions like IBS and other DGBI (6,22). The hallmark feature of TCAs, believed to be primarily responsible for their antidepressant and analgesic properties, is a variable combination of 5-HT and NA reuptake inhibition properties (6,29). Based on this dual action, TCAs have more potential for analgesic effects than other antidepressant classes targeting only 1 monoamine system, such as SSRIs. However, this may also be associated with an increased risk of side effects (6,8).

Most TCAs have additional receptor affinities, some of which may be primarily responsible for their side effect profile (6,22,29). Thus, muscarinic-1 receptor antagonism may cause classic anticholinergic side effects, including dry mouth, constipation, drowsiness, and blurred vision; a1 adrenergic receptor antagonism may lead to dizziness, drowsiness, and orthostatic hypotension; and H1 receptor antagonism may lead to weight gain, especially in combination with 5-HT2C antagonism, as well as drowsiness. Sedation, fatigue, headache, nausea, and sexual dysfunction are additional adverse effects (6,10,30).

In addition, some TCAs have weak sodium channel-blocking properties, which leads to a risk of arrhythmias and coma or seizures upon overdosing; these side effects seem to be less common with the secondary amine TCAs, desipramine, and nortriptyline (6). Therefore, TCAs should be avoided in patients with bundle branch block or prolonged QT intervals (6,29). Some of the side effects of TCAs may be beneficial in patients with DGBI. For example, slowing of GI transit due to their anticholinergic properties, which should be helpful in patients with IBS-D, and increased appetite and weight gain will be useful in patients with functional dyspepsia with early satiation and weight loss (6,22).

TCA agents are subtyped into secondary amines like desipramine and nortriptyline and tertiary amines like amitriptyline and imipramine, the latter having greater antimuscarinic and antihistaminic actions (6,22). While both types can work for the pain of IBS, the secondary amines are preferred if constipation is a dominant symptom. The initial dose and titration schedule of the various TCAs are similar. However, tertiary amines have more side effects because of their greater antagonism of cholinergic, adrenergic, and histamine receptors (6,22). Dosing can start at 25 mg, going up to 50 mg after a week if well tolerated by the patient and, if needed, increased to 75 mg. If there is an insufficient benefit at this low dose after a month, the TCA dose can be further increased if tolerated and side effects are minimal (Table 2) (22).

Table 2.

TCA dose range

Doses used in psychiatry (150–300 mg) for depression are higher than are needed to target pain and other GI symptoms. One recent 6-month placebo-controlled trial of amitriptyline 10–30 mg dose range in primary care showed benefit with GI but not psychological symptoms although adverse anticholinergic effects were common. Unfortunately, the study did not provide information on the differential benefit or frequency of side effects based on dose. One small study was published using a 10 mg dose of amitriptyline for IBS-D that showed modest benefit in treating pain and diarrhea, but it had methodological limitations (31). While very low doses may be insufficient for the control of digestive symptoms, it could be a starting point for future research (6,22,32).

Many providers choose to start with low doses, presumably to avoid side effects (33) or ease patient anticipatory anxiety. However, if 10 mg is started, it needs to be increased to a higher dose target (Table 2). Suppose a patient reports side effects immediately and at low doses. In that case, it may be a nocebo effect rather than a medication effect, so rather than stop or switch the medication, we suggest the patient stay on the same dose for a week to adapt and then subsequently try to increase the dose further. Dose-related antimuscarinic side effects may be anticipated with higher doses up to 150 mg/d (34).

Selective serotonin reuptake inhibitors

SSRIs act by selective blockade of the presynaptic 5-HT transporter, boosting 5-HT neurotransmission and consequently stimulating intestinal transit; however, by not acting on NA receptors, SSRIs do not treat pain (6). Their primary serotonergic effect, without noradrenergic effect, leads to more significant expected benefits in treating anxiety, obsessive-compulsive disorder, and phobic-related behaviors. They can be added to a TCA or SNRI in low doses when anxiety is dominant (6,8). Similarly SSRIss may be used instead of a TCA to manage a patient with IBS-C when pain is not dominant. SSRI are more likely to cause diarrhea, while TCAs tend to be constipating due to their NA effect (29,30).

The SSRIs include fluoxetine, fluvoxamine, sertraline, paroxetine, citalopram, and escitalopram. Although all act on 5-HT reuptake inhibition, each of them has specific pharmacologic properties, like greater 5-HT2C antagonism of fluoxetine and greater anticholinergic action of paroxetine, so unlike the other SSRIs, paroxetine may produce constipation (22).

Sertraline, citalopram, and escitalopram tend to have the fewest pharmacokinetic drug interactions as they exhibit minimal effects on the cytochrome P450 enzyme system. Fluoxetine and paroxetine, however, have an increased risk of pharmacokinetic drug interactions through their strong inhibition of the P450 isoenzymes 1A2 and 2D6; therefore, they should be administered with caution when used in conjunction with TCAs and beta-blockers like metoprolol, opioids, lithium, tryptophan and monoamine oxidase inhibitors (22,35).

SSRIs are first-line pharmacologic agents for treating anxiety disorders, but they have the potential to induce restlessness and exacerbate anxiety when the drug is initiated. They are typically initiated at half of the usual starting dose to minimize these potential anxiogenic adverse effects. The dose may gradually increase to the regular starting dose after about 1 week (Table 3) (22). The higher effect of the SSRI is usually delayed 3–4 weeks, which may represent a problem for those patients with significant anxiety that is complicating treatment and causing significant functional impairment. A useful strategy for this situation is to schedule a long-acting benzodiazepine to temporarily bridge this lag time and provide symptomatic relief for the patient's anxiety symptoms. The benzodiazepine should then be tapered after about 4 weeks of SSRI treatment (22).

Table 3.

SSRI dose range

Other side effects of SSRIs are agitation, sleep disturbance, nausea, diarrhea, night sweats, headache, weight loss, and sexual dysfunction (6,22,30).

Serotonin and noradrenaline reuptake inhibitors

SNRIs block presynaptic 5-HT and NA reuptake, boosting 5-HT and NA neurotransmission and decreasing the perception of visceral pain but without antihistamine and anticholinergic effects, have less effect on intestinal motility than TCAs so that they can be used as first-line neuromodulator treatment in the presence of constipation, with lower risk of adverse effects (6). Their value has been demonstrated for somatic pain such as fibromyalgia, diabetic neuropathy, or migraine headaches. It has not been adequately studied for visceral pain. Still, they are commonly used for this purpose off-label, with empiric benefit, because of the lower side effect burden than the TCAs and similar pain reduction (22,36,37). It should be noted that in addition to showing benefits with depression and painful disorders, SNRIs have shown significant improvement in anxiety, so they could be considered for this purpose as monotherapy or as part of augmentation therapy (36).

SNRIs are prescribed as primary agents for treating IBS and other painful DGBI. They may be used in patients with pain who failed initial treatment with TCAs or experienced intolerable side effects from the TCAs that precluded them from reaching a potentially therapeutic dose (6,22).

Regarding adverse effects related to the use of SNRIs, nausea may occur using SNRIs, more frequently under duloxetine, but can be minimized when taken with meals (21,29). Other side effects related to SNRIs are hypertension, more frequently under venlafaxine, agitation, dizziness, sleep disturbance, fatigue, headache, especially when decreasing doses, and rarely liver dysfunction (6,22,30).

Inhibition of NA reuptake varies according to each neuromodulator, so dosage changes between agents (Table 4) (22). Venlafaxine, dosed for depression at 75–225 mg, acts primarily as an SSRI at lower doses, but higher doses (>225 mg) are needed to achieve NA inhibition and be effective as a pain treatment (22). Duloxetine has a strong and similar affinity for the 5-HT transporter and NA, acting as a true SNRI even at lower doses; in a comparative study, duloxetine activity was evaluated in patients with IBS-D, and the results showed a clinical remission of pain and diarrhea, as well as an increase in the threshold of visceral sensitivity, using the balloon dilation test (7,22). Milnacipran is more potent in inhibiting the reuptake of NA compared with its ability to inhibit the reuptake of 5-HT. For this reason, it is used for the treatment of chronic somatic pain, and although as an SNRI, it is not marketed for depression in the United States. This may help to facilitate acceptance by patients who are concerned about being prescribed a psychiatric medication (6,22).

Table 4.

SNRI dose range

Tetracyclics or noradrenergic and specific serotonergic antidepressants

The NA and specific serotonergic antidepressants, known as TNAS, have indirect effects resulting in increased NA and serotonergic activity through antagonism on α2 NA, 5-HT2A, 5-HT2C 5-HT3, H1, and muscarinic 1 receptors. The most representative agent of this class is mirtazapine (Table 5). However, their effects seem to be mainly on anxiety, early satiety, nausea, and other symptoms associated with esophageal and gastroduodenal disorders, so their use in IBS is limited (6,22,38,39). Because of its sedative effect, it is usually given at bedtime; however, it may be transient, and there is a paradoxical effect, with sedation decreasing at higher dosages of mirtazapine (40).

Table 5.

TNAS dose range

Mirtazapine boosts 5-HT and NA neurotransmission-blocking presynaptic α2 NA autoreceptors and heteroreceptors on NA and 5-HT neurons, which act as brakes on both NA and 5-HT release from these respective neurons (6,29). In addition, like some of the TCAs, it has 5-HT2A and 5-HT2C receptor antagonist properties, which may account for some additional antidepressant properties, as well as a more favorable side effect profile by blocking some of the unwanted receptor actions of boosting 5-HT transmission (6,29). The same applies to its 5-HT3 antagonist properties, which may explain its more favorable GI side effect profile, including reduced nausea, pain, and diarrhea, although this same mechanism could induce constipation (6,20,28). Likewise, through its H1 and 5-HT2C antagonist properties, mirtazapine may cause increased appetite, weight gain, sedation, fatigue, sweating, and dry mouth (6,30,39,41).

Atypical antipsychotic agents

AAP are a group of second-generation drugs, which differ from first-generation antipsychotics by presenting fewer adverse effects. This classification includes quetiapine, olanzapine, brexpiprazole, and aripiprazole (Table 6). However, the most used for DGBI are quetiapine and olanzapine, so in this review, we focus on these 2 agents (22).

Table 6.

AAP dose range

AAPs act through complex mechanisms, resulting in favorable clinical effects, like anxiety reduction and regulation of sleep patterns. AAPs also have effects as a norepinephrine transporter inhibitor, which is of theoretical advantage for analgesic effects; particularly, quetiapine has been used as an augmenting agent or second-line treatments in patients with abdominal pain related to IBS who do not respond to treatment with TCAs or SNRIs (6,22,42).

Dopamine receptor antagonist activity (D2) is the target of traditional antipsychotics, its modulation is responsible for the antipsychotic effect, and it is also the cause of unwanted side effects such as extrapyramidalism, dyskinesia, hyperprolactinemia, and affective or cognitive disorders (6). AAPs, on the other hand, have 5-HT2A receptor antagonist properties (olanzapine, quetiapine), rapid dissociation of the D2 receptor, partial D2 agonism, and/or partial 5-HT1A agonism (quetiapine); these additional mechanisms of action reduce the impact of its D2 receptor antagonist activity and thus reduce the incidence of side effects (6,8). However, related to their additional anticholinergic properties, different receptor antagonist effects (H1, 5-HT2C, α1-NA, and/or α2-NA), and other not-well-known mechanisms, AAPs can increase appetite and cause extrapyramidal syndrome, weight gain, fatigue, sweating, dizziness, cardiometabolic disease, and sedation; therefore, although in the treatment of IBS and DGBI low doses are prescribed, these agents should be used with sufficient care and monitored for side effects, especially when used chronically. However, most of these side effects are reported in the psychiatric literature with dosages 5–10 times greater than that prescribed for GI disorders. Thus, we would anticipate fewer side effects with lower dosages used for GI disorders (6,30,42–46).

The pharmacological properties by which quetiapine and olanzapine act to improve nausea, abdominal pain, and general symptoms of DGBI remain uncertain. It is known that olanzapine and quetiapine have combined D2/5-HT2A antagonist properties, characteristic of most atypical antipsychotics, with in addition H1, 5-HT2C, and α1-antagonist, as well as anticholinergic properties. Quetiapine has additional 5-HT1A partial agonist properties, as well as noradrenaline reuptake inhibitory effects, which may provide a rationale for its use in IBS and other DGBI (6,22).

CENTRAL NEUROMODULATOR SELECTION

Choosing a central neuromodulator for treating IBS should be taken according to the pharmacological properties of the different groups and the predominant symptoms of each patient (Table 7), as well as considering the potential side effects (Table 8).

Table 7.

Activity of central neuromodulators on abdominal pain, intestinal motility, anxiety, and depression

Table 8.

Potential side effects of central neuromodulators^a

HOW TO PRESCRIBE CENTRAL NEUROMODULATORS

General approach

Prescribing a central neuromodulator involves providing the rationale for their use, clarifying their targeted benefits and side effects, and paying attention to patient concerns. It is not uncommon for patients to believe a central neuromodulator is being prescribed for a psychiatric condition rather than a brain-gut disorder. It is important to explain to the patients, clearly and concisely, the meaning of the BGA and link their symptoms to dysregulation between the brain and the gut. So, neuromodulators are not necessarily used for the treatment of depression but are a therapeutic alternative in the management of DGBI. It helps to use the term “neuromodulator” instead of “antidepressant” (6,8) It also helps to clarify that these medications can treat pain and other GI symptoms independent of treating depression, and the dosages are often lower than those used for treating major depression. This will preclude any patient concerns that their symptoms are being underestimated or considered to be in their head (6,8). See Video 3 in Supplementary Materials (Supplementary Digital Content 1, http://links.lww.com/AJG/D247) for details how to best communicate the use of central neuromodulators to a patient.

This approach strengthens the patient's understanding and reduces the possibility of nonadherence. It also helps to note that chronic pain can lead to anxiety and depression so these medications can help both components. In this way, the patient will be more open to accepting a component of anxiety and depression in their illness (4–6).

Pharmacogenomic testing

Pharmacogenomic testing has gained ground in the selection of a neuromodulator for the treatment of psychiatric diseases, and there are limited studies to address this strategy in patients with DGBI (47,48). Pharmacogenomics is the study of the variability of the expression of individual genes relevant to disease susceptibility, as well as drug response. It is used to identify genetically determined interactions and neuromodulator profiles to optimize benefits and reduce toxicity. Practically, it can determine if a patient is a rapid (causing low blood levels and reduced benefit) or a slow (leading to high blood levels and toxicity) metabolizer, which helps determine the selection of medication or its dosage and helps when augmenting treatment using several medications where interaction effects are to be avoided (47,48). However, its clinical value has not yet been determined for GI disorders, so it is only a guide.

We propose that testing is not needed initially. The choice of neuromodulator, the starting dose, and possible changes over time should be made based on predominant symptoms and tolerance to the drug (as described in later sections of this article). However, we believe that pharmacogenomic testing is an option under 2 conditions. First, when the patient reports no symptom benefit to high doses of neuromodulators, to determine if the patient is a rapid metabolizer of them. If present, one can switch to another medication with normal metabolism. Second, testing may be of value for the patient who reports multiple side effects to several medications. Testing may determine if the patient is a slow metabolizer, thereby accumulating high blood levels and the provider can then search for a better option. If the study shows normal metabolism, this may help patient adherence. Then one could try to encourage the patient to stay on the medication a bit longer until treatment benefit occurs and side effects diminish.

TARGETS FOR TREATMENT BASED ON SYMPTOMS

The selection of a neuromodulator in patients with IBS should be made carefully, considering its pharmacological properties and side effects. In this section, we detail the main arguments for choosing and combining neuromodulators according to the predominant symptoms.

When the main symptom is abdominal pain

The first-line neuromodulator treatment for pain management in IBS is using a TCA (6,8,10,22,29). In a meta-analysis that evaluated the effects of TCAs and SSRIs in treating IBS, conducted from 18 RCTs with 1,127 patients, while both groups showed significant global improvement of symptoms, only the TCAs differed significantly in abdominal pain (49). Another first-line neuromodulator alternative for pain management is the SNRIs (6,8,10,22). A meta-analysis that included 13 studies with amitriptyline and 3 with duloxetine confirmed the efficacy of amitriptyline in terms of pain improvement in patients with IBS-D; the same study also showed benefit for duloxetine in the management of IBS pain. However, the duloxetine studies included uncontrolled trials with lower quality of evidence than that of TCAs (50). Nevertheless, SNRIs are increasingly used based on supportive data from treating other painful conditions and personal experience. The anticholinergic effect of TCAs could be reduced by using secondary amines (desipramine and nortriptyline), as well as with the alternative of using SNRIs, whose use in IBS-C with pain as the predominant symptom appears to be safer (6,22,29).

When the patient has anxiety

While not effective in treating abdominal pain, SSRIs are useful when symptoms of anxiety and hypervigilance, obsessive behaviors, social phobia, or agoraphobia are dominant (6,22). When the main symptom is abdominal pain in conjunction with significant anxiety, TCAs may be combined with low-dose SSRIs (6,8,22). In these cases, low dose escitalopram may be used, based on its tolerability and low frequency of drug interactions (22,30). Alternatively, an SNRI can be used as monotherapy for anxiety and pain.

When the patient has constipation

The predominant bowel habit is important when choosing a neuromodulator to treat IBS, both for the opportunity to improve it and to be aware of when the treatment may worsen it (6,8,10,22). When constipation predominates, avoid neuromodulators with strong anticholinergic action, such as tertiary amine TCAs like amitriptyline and imipramine. SNRIs or secondary amine TCAs are more acceptable due to less effect on reducing intestinal transit (6,22). Although SSRIs generally stimulate intestinal transit, they do not improve the pain of IBS-C (6,22,29,51).

When the patient has diarrhea

When diarrhea predominates, tertiary amines like amitriptyline and imipramine are a good choice due to their strong anticholinergic action; if constipation occurs, the option of adding treatment for the constipation may be considered, especially if the neuromodulator therapy helps reduce abdominal pain and other symptoms (6,22,29). Alternatively, a secondary amine having less anticholinergic effect or duloxetine may be prescribed (6,22). In a randomized study conducted with 61 patients with IBS-D, SNRI-duloxetine therapy showed a significant clinical remission of pain and diarrhea (7). In patients with diarrhea where the anxiety component is predominant over pain, paroxetine could be chosen since it has greater anticholinergic effects than the other SSRIs (6,22,51).

When the patient has mixed bowel habits

When there is a mixed bowel pattern, secondary amine TCAs such as desipramine and nortriptyline, or SNRIs like duloxetine, could be preferred because they do not have as strong an effect on intestinal transit. Note that mirtazapine, tertiary amine TCAs such as amitriptyline and imipramine, and paroxetine may cause constipation. By contrast, SSRIs may cause diarrhea, so they should be prescribed when anxiety cannot be effectively managed with neuromodulators of another class (6,8,22,29).

AUGMENTATION: WHAT TO DO WHEN MONOTHERAPY IS NOT SUFFICIENT

When to implement augmentation therapy

Central neuromodulators require between 4 and 8 weeks of use to reach their maximum level of effectiveness, although there is some evidence that a partial benefit may occur at 2–3 weeks and, when it occurs, predicts a favorable long-term response (49,50). It is also important to gradually increase the dose over several weeks to reach the targeted dose and minimize side effects. Therefore, allow proper time to determine when a treatment is not sufficient. In that case, consider augmentation treatment in several situations: (i) when a single treatment is ineffective in controlling symptoms, (ii) when increasing doses is considered risky or produces side effects, and (iii) when comorbidities also need additional treatment. Under these circumstances, a second central acting agent, a peripheral neuromodulator, or a brain-gut behavioral treatment should be added (6,8,10,22,29).

How to implement augmentation therapy

Adding another first-line neuromodulator agent may be beneficial when the therapeutic effect of the first agent is partial. Benefits can occur with medications that have complementary mechanisms of action; thus, one would add an SSRI if a patient with IBS shows pain relief using a TCA but has insufficient control of coexisting anxiety; it may also be that patients are using SSRIs prescribed by a psychiatrist; in these cases, we consider adding a TCA or SNRI agent to control the symptoms of IBS, starting with low doses, ideally in collaboration with the psychiatrist. It is important to remember that the usual dose of TCAs may not be enough to treat the psychiatric condition or to produce serotonin-related side effects (6).

Another option is adding a second-line neuromodulator agent like an AAP, which, as previously explained, has less risk for side effects than first-generation antipsychotics. There is some experience from using quetiapine in treating chronic pain; quetiapine can also improve pain when used to augment the effects of a TCA or SNRI. This combination has added clinical effects, like anxiety reduction and establishing a normal sleep pattern; its main metabolite also has effects as an NA transporter inhibitor, which is a theoretical advantage for analgesic effects (6,22,43).

Using quetiapine above 200 mg/d may lead to poor tolerability because of excessive sedation and dizziness and metabolic side effects such as weight gain, hyperlipidemia, and diabetes. For this reason, the recommended dose range would be 25–200 mg/d for patients with IBS or DGBI. Sometimes we use higher dosing when there are coexisting psychological comorbidities such as post-traumatic stress disorder or anxiety. Figure 3 demonstrates the selection of neuromodulators for monotherapy or augmentation when treating DGBI (6,22,42).

Figure 3.

Clinical characteristics to considerate when selecting central neuromodulators in DGBIs. Drugs in the upper part can be considered as first-line options. In the lower part of the figure, the pharmacologic options most often used to augment treatment effects are depicted, as well as some nonpharmacologic treatment alternatives. CBT, cognitive behavioral therapy; DBT, dialectical behavior therapy; DGBI, disorder of gut-brain interaction; EMDR, eye movement desensitization and reprocessing; FGID, functional gastrointestinal disorder; GI, gastrointestinal; PTSD, post-traumatic stress disorder; SNRI, serotonin and noradrenaline reuptake inhibitor; SSRI, selective serotonin reuptake inhibitor; TCA, tricyclic antidepressant. From Drossman DA, Tack J, Ford AC, Szigethy E, Törnblom H, Van Oudenhove L. Neuromodulators for functional gastrointestinal disorders (disorders of gut-brain interaction): A Rome Foundation Working Team Report. Gastroenterology 2018;154(4):1140–71.e1.

It should be noted that although we recommend the use of AAP as previously described, some gastroenterologists and primary care providers may feel uncomfortable because they are not familiar with using AAP (33). It should also be considered that the regulations for prescribing these drugs differ according to different countries. In this situation, we recommend prescribing AAP in conjunction with a psychiatrist.

IMPORTANT SIDE EFFECTS AND SITUATIONS TO ADDRESS

Central neuromodulators are generally safe, and treating patients with DGBI are frequently prescribed in lower doses than used in psychiatry (6,22). However, all agents have well-recognized side effect profiles that providers must consider (30,52). This is especially true when combining medications for augmentation or when patients have underlying medical conditions, such as heart disease or pregnancy (30,52,53). The following section will review some of the frequent side effects or situations requiring medication adjustment.

Serotonin syndrome

Many of the central neuromodulators activate serotonin receptors. Therefore, a possible complication is serotonin syndrome, which in its more severe form is characterized by fever, hyperreflexia, spontaneous clonus, muscle stiffness, tremors, confusion, tachycardia, seizures, pupillary dilation, and increased risk of death if not treated immediately (6,22,30). However, providers should be more alert to milder episodes of increased anxiety and tachycardia.

Serotonin syndrome is more likely to occur when using high doses of neuromodulators that strongly inhibit serotonin reuptake such as SSRIs or when several agents with serotonergic effects are combined. Clinical manifestations usually appear shortly after implementing augmentation therapy when serum serotonin levels are highest (22).

All drugs with serotonergic properties should be temporarily discontinued when the syndrome occurs. Reimplementation occurs gradually, starting with low doses that are slowly increased over several days to a week. In addition, other classes of medications such as tramadol, ondansetron, or triptans may also increase serotonin levels and trigger the syndrome (22,30).

Cardiac side effects

QT prolongation.

Although initial publications suggest that SSRIs and SNRIs have a lower risk of cardiovascular disease when compared with TCAs, many of the newer antidepressants are not exempt from the onset of heart disease (30,53,54). In a comprehensive review that analyzed the antidepressant-induced QT prolongation in people with psychiatric disorders, the authors found an increased mortality odds ratio of 2.11 with high doses of TCA and 2.78 in SSRI users compared with controls matching for age and sex (55,56).

The QT interval is the period between the onset of a Q wave and the end of a T wave on the electrocardiogram (EKG). By contrast, the corrected QT interval (QTc) refers to the QT after adjustment to baseline heart rate. An increased QT interval can be associated with a malignant ventricular tachyarrhythmia known as Torsade de Pointes (55). The risk of QT prolongation should be considered during the use of TCAs and in patients when using SSRIs (30,53,54). A meta-analysis found that among SSRIs, citalopram appears to be the agent most significantly associated with QTc prolongation (57).

A clinically consistent association with cardiac conduction defects or arrhythmias has not been identified with the SNRIs. However, venlafaxine (in doses over 200 mg daily) and duloxetine have been associated with increases in diastolic blood pressure (54,58,59).

Although the risk of cardiovascular disease is not high in central neuromodulator users, they should be prescribed in the lowest doses required to be effective. An EKG should be considered to evaluate for QT prolongation and cardiac arrhythmias and must be done in the elderly population and patients with heart disease or concomitant treatment with drugs known to cause prolongation of the QTc interval. Examples include antiarrhythmics such as amiodarone, sotalol, quinidine, procainamide, verapamil, and diltiazem. In addition, noncardiovascular drugs such as ondansetron, macrolide, fluoroquinolone antibiotics, and first-generation antipsychotic agents such as haloperidol, thioridazine, and sertindole should be considered. If the EKG confirms QT prolongation, stop the high-risk medication and restart treatment at a lower dose after a cardiac consult and then repeat the EKG (55,60).

Orthostatic hypotension.

The risk of orthostatic hypotension, secondary to using central neuromodulators, has been well-established with TCAs due to their well-known antagonistic α1-adrenergic receptor activity (30,53). Paroxetine seems to be the SSRI most frequently linked with orthostatic hypotension because of its anticholinergic effects, especially in the elderly population; however, the main mechanisms associated with other SSRI-induced orthostatic hypotension remain unknown (52,60,61). Mirtazapine may also cause orthostatic hypotension in up to 7% of patients. However, venlafaxine, due to its strong noradrenergic action, can cause this side effect in more than 50% of patients older than 60 years (53,61–63).

Weight gain

Weight gain related to the use of central neuromodulators may be beneficial for patients with DGBI with early satiation and low weight. Still, it may be inappropriate when there are overweight and metabolic disorders risks. This side effect may occur during the acute and maintenance phases of treatment with central neuromodulators (6,30,53).

Weight gain induced by central neuromodulators may occur due to the interaction of several mechanisms, including (i) action on specific neuroreceptors, (ii) decreased caloric expenditure due to sedative effects, (iii) change in food preference, and (iv) dry mouth/throat may induce increased intake of caloric beverages (53).

As mentioned in previous sections, the affinity of neuromodulators for the H1 receptor seems to be linked to weight gain. Based on this hypothesis, one study evaluated weight gain as part of metabolic syndrome during antidepressant treatment and showed that patients who used amitriptyline, trimipramine, mirtazapine, and nortriptyline showed high affinity for the H1 receptor and experienced significantly greater weight gain than those who used duloxetine, venlafaxine, citalopram, escitalopram, sertraline, paroxetine, and fluoxetine, which showed a low affinity for the H1 receptor (64).

In a meta-analysis that evaluated the adverse effects of antidepressants, amitriptyline (relative risk [RR] = 8.74), mirtazapine (RR = 6), and nortriptyline (RR = 2.9) were significantly associated with weight gain compared with placebo (54). The magnitude of the effect has been quantified in patients receiving low-modest doses of tricyclic antidepressants given for an average of 6 months. There was a mean weight increase of 0.59–1.32 kg/mo, which led to an average total weight gain of 1.36–7.26 kg, depending on drug, dose, and duration of the treatment (65,66). Another meta-analysis conducted from studies with antipsychotics found that the use of olanzapine and, to a lesser extent, quetiapine may be associated with significant weight gain compared with placebo (67). The magnitude of weight gain appears to be dose-related, being more likely at higher doses (68). It is also more common when combining an antidepressant with an AAP.

Although SSRIs are commonly associated with weight loss at least initially, evidence suggests that paroxetine may induce weight gain (69). Regarding SNRIs, weight gain appears to be an uncommon adverse effect in patients treated with venlafaxine and duloxetine (70).

If weight gain occurs during treatment with neuromodulators, reduce the dose of the neuromodulator. Additional options include recommending dietary changes and possibly concomitantly using interventions for treating obesity, such as metformin or glucagon-like peptide 1 agents (65,71).

Complications related to pregnancy

Central neuromodulators are considered safe during pregnancy, especially TCAs, SNRs, mirtazapine, and AAPs (72). Although SSRIs appear to be less safe, they are the most prescribed antidepressants during pregnancy; however, publications questioning the safety profile of SSRIs appear to have confounding factors that may produce questionable results, such as maternal age, smoking, and/or concomitant use of other medications like anticonvulsants (53,72,73).

While SSRIs are the most used neuromodulators during pregnancy (72,73), evidence shows that the use of SSRIs is associated with reduced fetal head growth and increased risk of preterm birth (72). Avoid paroxetine as it is related to congenital heart defects compared with other SSRIs. Other birth defects may be related to the use of SSRIs. Thus, sertraline seems to increase the risk of heart defects and craniosynostosis, citalopram with cardiac malformations, and escitalopram with musculoskeletal malformations (53).

SNRIs during pregnancy do not appear to be associated with an increased risk of birth defects; however, although the evidence is not extensive, SNRIs have been associated with an increased risk of postpartum hemorrhage and particularly venlafaxine with hypertension during pregnancy (53).

Treating with central neuromodulators during pregnancy should balance with its associated risks. If the patient is stabilized on a specific drug, it is preferable to maintain the same treatment, except for paroxetine. If it is a woman who has not received treatment with neuromodulators, the use of TCAs, mirtazapine, AAPs, and SNRIs are quite safe alternatives; if SSRIs are required, sertraline and citalopram appear to be the most appropriate alternatives (54,73).

PRESCRIBING AND DISCONTINUING CENTRAL NEUROMODULATORS

How to prescribe and taper

The general practice is to prescribe central neuromodulators to reduce GI symptoms of pain, nausea, bloating, bowel dysmotility, and, at times, psychological symptoms of anxiety or depression and to titrate the dose up until an optimal dose achieves benefit. We recommend that the medication be prescribed at half dose for 1–2 weeks to assess side effects and if no side effects occur the full dose can be prescribed. If side effects occur and are tolerated, the patient should be encouraged to stay with the treatment for 1–2 weeks more because it is likely that the effects will diminish and can then be increased to the targeted dose. In general, if side effects do not occur with the initial dose, they are much less likely to occur when increasing the dose.

It should be kept in mind that nocebo effects may occur where the patient experiences symptoms related more to anticipatory anxiety or conditioning from earlier treatment experiences rather than to the pharmacological effect of the medication. This may occur when the side effects are not usual for the particular medication or occur after 1 pill is taken, before achieving adequate blood levels. In one National Institutes of Health study of patients taking desipramine, most of the side effects were reported as symptoms existing before beginning treatment or were not the anticipated side effects. Furthermore, the severity of the side effect symptoms correlated with anxiety scores and was not related to drug blood levels (74). When this occurs, it is important to help the patient understand the importance of staying on a course with a medication with a gradual increase in dose to achieve anticipated expectations instead of switching from 1 medication to another.

Using central neuromodulators for IBS requires long-term treatment. From our experience, 6–12 months of treatment or more are needed to increase the likelihood of remission. In some cases, a good clinical response may occur in a shorter period. The decision to discontinue is based on the clinical response. If the patient achieves meaningful improvement or best symptom resolution over this period, remission may have occurred, and a slow reduction in the dose can begin. If there is a partial response, the patient should be treated longer. Treatment may also continue when there are ongoing psychosocial stressors, a history of multiple previous episodes, or psychiatric comorbidities (6,8,10,22,29).

The provider can consider reducing or tapering off the central neuromodulators when the patient has achieved reasonable symptom control for at least 6 months. The dose should be tapered slowly over 4 weeks (25% per week) but may take longer when discontinuation effects occur. This commonly occurs with SNRIs. If the patient has been on the central neuromodulator for less than 4–6 weeks, it can usually be discontinued rapidly with minimal if any side effects occurring (6,8). However, in our experience, many, if not most, patients' symptoms tend to wax and wane over weeks or months. This often requires titrating the medications up and down to accommodate the fluctuations in the clinical conditions. It is essential to wait for 3–4 weeks after a dose adjustment to determine if the treatment was effective.

If the treatment is not effective after this period, one can then switch to another neuromodulator. When switching from 1 medication to another, for example, a TCA to an SNRI, because of overlapping receptor effects, it is possible to switch over a relatively short time when compared with tapering off monotherapy. For example, the TCA could be reduced to half dose while adding the SNRI at half dose, and then 2 weeks later, the TCA is stopped while raising the SNRI to full dose. Changes in dosage may take longer or shorter depending on the patient's therapeutic response and tolerance.

Antidepressant discontinuation syndrome

If the decision is to withdraw treatment, abrupt discontinuation of central serotonergic neuromodulators (mainly SSRIs and SNRIs) may be associated with antidepressant discontinuation syndrome (ADS), which consists of various symptoms, including nausea, headache, flulike symptoms, imbalance, insomnia, anxiety, agitation, and sensory disturbance like brain zaps and paresthesia (6,22,75).

The ADS is more common using agents with a shorter half-life. So, for the SSRIs, paroxetine, with a half-life of less than a day, is more likely to produce ADS when stopped abruptly. By contrast, fluoxetine has a half-life of 3–4 days, and ADS is less common. Citalopram, escitalopram, and sertraline are intermediate and exhibit half-lives of longer than a day but may still require a gradual taper to avoid the uncomfortable symptoms of ADS (22,75).

There is no clearly defined strategy for preventing and treating ADS symptoms, and patient attributions to the withdrawal must be considered. The patient who is hypervigilant to withdrawal effects is more likely to have difficulty withdrawing. In addition, when comorbidities such as major depression or anxiety disorders are present, cognitive behavioral therapy or other brain-gut behavior therapies should be used concurrently to supplement the loss of the use of the medication. In addition, self-care behavior interventions like mindfulness, relaxation, and supportive relationships explain up to 20%–30% of the variance in predicting successful discontinuation of the antidepressants (75).

Some strategies may be helpful. It is better to discontinue the morning dose for a week and then discontinue the evening dose. Another alternative is to switch to an agent with a longer half-life, like fluoxetine when planning to stop another antidepressant. However, in cases where symptoms are severe, sometimes it is necessary to reintroduce the neuromodulator and retry using a slower discontinuation process (22,76). Since the lowest dose of duloxetine is 20 mg, we sometimes advocate patients opening the capsule and taking half the dose (10 mg) when continuing the tapering protocol.

CONCLUSION

In this review article, we have discussed how to use central neuromodulators when treating IBS. Their action on the reuptake of neurotransmitters such as 5-HT, dopamine, or NA demonstrates that its mechanism of action goes beyond improving psychiatric comorbidity for some of these patients via its action on visceral sensitivity, intestinal transit, and neurogenesis as seen with IBS.

Central neuromodulators are an essential treatment in managing IBS when symptoms, particularly pain, are dominant or when there are psychological comorbidities. TCAs seem to have the best evidence as to their benefit in pain management, SNRIs are also a useful alternative to treat pain with fewer adverse effects, but further research is needed to confirm their value. SSRIs should be considered when a significant component of anxiety without pain is present.

Augmentation therapy is a strategy to improve therapeutic effects when the initial treatment is insufficient, and it is impossible to increase doses due to the risk of side effects. This should be done by adding to the initial neuromodulator chosen to treat pain a second centrally acting agent, one that acts peripherally or a behavioral treatment.

Gastroenterologists managing patients with IBS and other DGBI should be familiar with central neuromodulators and know their characteristics, mechanism of action, indications, and possible adverse effects because they are a beneficial alternative in the treatment of these disorders.

CONFLICTS OF INTEREST

Guarantor of the article: Douglas A. Drossman and Ignacio Hanna-Jairala.

Specific author contributions: Medical education.

Financial support: Nothing to declare.

Potential competing interests: Internal medicine—gastroenterology—clinical practice.

Supplementary Material

acg-119-1272-s001.pdf ^{(51.1KB, pdf)}