Many common symptoms relate to dysfunction or disease of the gastrointestinal tract. Digestive diseases afflict 12% of all American adults and account for 16% of all absences from work. In a study of 25,000 illnesses in a group of Cleveland families, acute diarrheal illness was one of the most common illnesses reported, second only to the common cold. Therefore, some of the most commonly used drugs are those directed at symptoms and diseases of the alimentary tract.
Special features of the clinical pharmacology of the digestive tract
The therapy of diseases of the alimentary system presents a number of special challenges that arise from the several special characteristics of the digestive tract. The digestive tract is the usual route of administration for most drugs; they may have a direct action within the gastrointestinal tract or on its mucosal lining at relatively high concentrations. For example, salicylates and other nonsteroidal anti-inflammatory drugs may produce bleeding or erosive gastritis by a direct toxic effect on gastric mucosa.
A therapeutic benefit from certain drugs can be achieved without the drug’s entering the circulatory system. Some drugs exert their action solely within the lumen of the gastrointestinal tract. For example, the poorly absorbed antacids function only within the lumen of the stomach and duodenum to neutralize gastric acid. Some antibiotics (e.g., neomycin) or disaccharides (e.g., lactulose), useful for their action on the gastrointestinal flora, are poorly absorbed and exert their effects locally.
The presence of an enterohepatic cycle dependent on intestinal absorption, hepatic uptake, and biliary excretion may be exploited in the design of therapeutic regimens. Poisoning or overdoses with drugs that have an extensive enterohepatic circulation, such as theophylline, may be successfully treated with oral activated charcoal even after the initial dose has been completely absorbed.
The gastrointestinal tract harbors a rich microbial flora, and some drugs require metabolism by bacteria to attain their full therapeutic activity. The action of sulfasalazine in the treatment of ulcerative colitis depends on the cleavage of its azo bond by bacteria of the colon. Bacteria may lead to the production of toxic metabolites from otherwise safe drugs. If cyclamates represent a carcinogenic hazard (the magnitude of this risk is uncertain in humans), they may do so because of bacterial conversion to a known carcinogen. Bacterial enzymes are further complemented by metabolizing enzymes of the intestinal mucosa that may modify drug pharmacokinetics, as in the case of cyclosporine A.
Because the alimentary tract is the route of administration of most drugs, there are potentially many drug-drug and drug-food interactions that could influence the absorption and effectiveness of various therapeutic agents. Examples of these include the adsorption of tetracycline by aluminum, calcium, or magnesium antacids; and the binding and inactivation of digitalis glycosides by the bile acid sequestrant cholestyramine. Because food may have a marked effect on the rate of absorption of drugs and vitamins and on their unwanted effects, it is very important to consider the impact of meals on the rate of absorption and effectiveness of orally administered drugs.
The gut is one of the richest endocrine organs of the body, and endogenous control or exogenous manipulation of gastrointestinal hormones may be critical in control of symptoms.
Finally, it is worth noting that disease and dysfunction of the gastrointestinal tract may influence the effectiveness of any therapeutic regimen that uses orally administered drugs. For example, patients with acquired immunodeficiency syndrome and Candida-related esophagitis frequently fail to respond to oral ketoconazole. The bioavailability of ketoconazole is reduced as a result of gastric achlorhydria, commonly seen in patients with acquired immunodeficiency syndrome. This effect can be reversed by concurrent administration of dilute hydrochloric acid. Similarly, it is unsound to extrapolate expected drug effects from one population to another when the rates of coexisting diseases (e.g., parasitic infestations, diarrheal illnesses) are different in the population tested from those in the target population.
The effects of disease of one organ on the diagnosis and treatment of disease in another organ should make the “specialist” aware that focus on the disorders of a single organ system cannot be allowed to result in underappreciation of the complex relationships between organ systems.
A general principle, shared with other organ systems, is that an understanding of the pathophysiology of a symptom or symptom complex in an individual patient usually will result in more selective and more effective therapy. For example, diarrhea as a symptom of enteric infection is more effectively and more definitively treated by specific therapy directed at the enteric pathogen than by nonspecific antidiarrheal agents. Sometimes therapy must be empirically directed at symptoms without a full understanding of their genesis; usually this practice is reserved for temporary control of symptoms during diagnostic evaluation or for those instances in which diagnostic analysis has failed to yield a full understanding of the cause of symptoms.
Continued use of drugs to control symptoms, without a comprehensive effort to understand the disease process that underlies these symptoms, is incomplete and dangerous medicine, delaying correct diagnosis and specific therapy.
Some gastrointestinal symptoms may be controlled or eliminated by withdrawal of an offending agent. Many drugs produce gastrointestinal symptoms; when such drugs are withdrawn or decreased in dose, symptoms usually abate or disappear. Orally administered drugs often exert a direct toxic effect on the gastrointestinal mucosa. Salicylates are believed to interfere with the integrity of the gastric epithelial membrane, allowing back diffusion of hydrogen ions. Neomycin causes malabsorption partly because of its direct toxicity on the small intestinal epithelium.
Drug dependency is not as frequent in patients with gastrointestinal disease as it is in patients with disease of other organ systems. However, the individual who abuses laxatives may so interrupt normal reflex function as to become functionally dependent on laxatives. Once again, the withdrawal of laxatives and appropriate changes in diet, with reeducation of and dependence on gastrointestinal reflex function, will result in a successful return to normal physiology.
In most societies, addiction is not considered important unless the drug predominantly affects the central nervous system. Such an attitude is unrealistic. Furthermore, education of the patient during withdrawal may be the critical determinant of success of such withdrawal.
Patients commonly attribute gastrointestinal symptoms to specific foods. Although many of these cause-and-effect relationships fade under controlled scrutiny, others, such as diarrhea and distension after ingestion of milk by individuals with deficiency of intestinal lactase, have a firm basis in the pathophysiology of the gastrointestinal system. In such patients the undigested lactose is fermented by colonic bacteria, with generation of hydrogen gas and organic acids. Much more rarely, however, foods are the cause of intestinal disease rather than of intestinal symptoms; in these instances, there usually is a genetic predisposition.
An example is celiac sprue, in which the injury to the intestinal mucosa is caused by a genetically determined sensitivity to dietary gluten (wheat protein). Both the mucosal injury and the malabsorption syndrome are reversed by fastidious withdrawal of gluten from the diet. Far less common than celiac disease, but greatly overdiagnosed, is the problem of food allergy. Milk-protein allergy is the best studied and most clearly documented. Withdrawal of the offending protein, such as cow’s milk, may result in dramatic improvement in diarrhea or extraintestinal signs, and reversal of the morphologic abnormality of the intestinal mucosa.
When gastrointestinal disease impairs digestive or absorptive function, resulting nutritional deficiencies are managed by replacement of missing factors or supplemental administration of deficient nutrients, or both. Intuitively one might think that loss by disease or surgical resection of the acid- and pepsin-producing cells of the stomach might adversely affect digestion and could, therefore, call for replacement therapy. In practice, the peptic activity of the stomach is unnecessary for the individual eating a diet of cooked proteins.
Such is not the case with disease or resection of the exocrine pancreas. Although the reserve is very great (90% of the pancreas must be destroyed before maldigestion occurs), such dysfunction can have very serious effects on intestinal digestion and nutrition. Under these circumstances, oral administration of pancreatic enzymes may improve the digestion and absorption of food.
Replacement therapy may also be required for restoration of nutrients lost because of malabsorption associated with gastrointestinal disease. The patient with pernicious anemia, who lacks gastric intrinsic factor, requires vitamin B12 by injection. The patient with steatorrhea due to pancreatic or intestinal disease may well require supplemental replacement of vitamins by oral or parenteral routes.
Gastrointestinal therapeutics may require interruption of normal or abnormal physiological processes. Often a process that is qualitatively normal but quantitatively excessive can result in gastrointestinal disease. Thus, peptic ulcer of the duodenum is associated with normal or excessive gastric acid secretion and may be treated by pharmacological inhibition or neutralization of acid secretion.
Understanding the pathogenesis of a disease and the pharmacology of a drug clearly allows the imaginative and effective design of new uses for old drugs. For most informed physicians, the main challenge in therapy is in construction of a new and logical hypothesis, in testing it with sound design, and in drawing valid conclusions related to the cause-and-effect relationship of drug-patient response.
Therapy is often directed at decreasing the functional stimulus to actively inflamed or diseased organs. An example of this principle is the effort to decrease the secretory responses of the pancreas in the presence of acute pancreatitis. By decreasing the flow of acid and food into the duodenum, one may decrease neural and hormonal stimuli to pancreatic secretion. In severe pancreatitis, this may require nasogastric suction; under milder conditions, small-volume feedings might minimize the pancreatic and gastric secretory response.
Similarly, the inflamed small intestine or colon in Crohn’s disease or ulcerative colitis is “put at rest” to diminish diarrhea, abdominal pain, and cramping by decreasing dietary intake or, more drastically, by changing to an elemental diet resulting in a decreased residue and decreased stimulus to bowel function. In the most symptomatic individuals, bowel rest is not achieved without placing the patient on a nothing-by-mouth regimen and substituting intravenous administration of fluids and medications. Total parenteral nutrition is used to prevent the worsening catabolic state associated with inadequate intake of calories and nitrogen when standard intravenous therapy is prolonged.
In the absence of defined etiology or specific therapy, empirical therapy of proven utility should be used, even if the mechanism of its benefit is not understood. This empirical approach to therapy is of particular importance in inflammatory diseases such as regional enteritis and ulcerative colitis; controlled clinical trials have shown the benefit of anti-inflammatory drugs such as corticosteroids and sulfasalazine, although the pathophysiological basis of this therapeutic benefit is not completely known.
When a drug whose pharmacology is reasonably well understood is empirically found to have efficacy in a disease whose pathogenesis is poorly understood, the finding of efficacy ultimately may shed light on the mechanism of disease.
Drug-induced gastrointestinal disorders
Diarrhea can be defined as an abnormal increase in stool frequency, weight, or liquidity. The former two are relatively straightforward to measure; the wet weight of stools is cumbersome to quantify but likely to be the factor that most easily correlates with the patient’s complaint. The average person eating three meals each day is likely to have 9000 mL of fluid traversing the duodenum, 1000 mL traversing the ileocecal valve, and 100 mL exiting as stool. Because the fluid balance of the gut must be exceedingly well controlled in order to have a normal water content in stools, minor changes in absorptive capacity of the bowel can likely play a major role in determining stool water content despite the ability of the bowel to adjust to changes in the delivery of abnormal quantities of fluid.
Diarrhea has many diverse causes but can be classified mechanistically into malabsorptive, maldigestive, or secretory processes; inflammatory states; and deranged intestinal motility. The differential diagnosis of diarrhea is lengthy and has been previously reviewed. As with other gastrointestinal symptoms, it is important to judge its severity and prognosis so that the need for treatment can be established, and to identify and specifically treat the underlying condition if possible and necessary.
Drug-induced diarrhea may be caused by any of the mechanisms mentioned above. Most commonly, diarrhea is caused by the use of laxatives or stool softeners. These may be poorly absorbed sugars such as lactulose and sorbitol that are fermented by intestinal bacteria in the colon or poorly absorbed salts of magnesium (sulfate, oxide, or hydroxide) or sodium (sulfate or citrate) ions. The diarrhea that ensues is characterized by a stool osmolality higher than that of plasma. Other commonly used laxatives (ricinoleic acid, phenolphthalein, dioctyl sodium sulfosuccinate, and senna) cause diarrhea characterized by its continuance even during fasting and a stool osmolality gap of less than 50 mOsm/kg.
The most common nosocomial diarrhea, and the most serious drug-induced diarrhea, is pseudomembranous colitis due to Clostridium difficile, a condition facilitated by antibiotic therapy. Most patients with antibiotic-related diarrhea have a benign illness that begins during administration of the drug and lasts less than 1 week following discontinuation of the offending agent. A small percentage of patients will develop severe diarrhea with evidence of invasive colitis (fever, tenesmus, mucus, or bloody stools) that persists after the antibiotic is discontinued.
Patients in this group are usually elderly and in the hospital or a skilled nursing facility. The antibiotics most frequently implicated are ampicillin or amoxicillin, clindamycin, and cephalosporins. In patients with pseudo-membranous colitis, laboratory studies generally reveal hypoalbuminemia and fecal leukocytes, flexible sigmoidoscopy generally demonstrates the characteristic 3- to 20-mm pseudomembranes bordered by normal or hyperemic colonic mucosa, and analysis generally reveals the presence of C. difficile toxin in the stool.
Therapy consists of discontinuing the implicated antibiotic, maintaining an adequate state of hydration and nutrition, and instituting enteric isolation procedures to limit person-to-person spread. Specific antibiotic therapy directed against C. difficile should be given to patients with signs and symptoms of moderate to severe colitis, or to those who fail to improve following nonspecific measures and discontinuation of antibiotics. Oral vancomycin and metronidazole appear to be equally efficacious; bacitracin may have equal activity, although it has not been as extensively studied. Parenteral metronidazole should be administered only to patients who cannot take oral medications.
Therapy is generally accompanied by rapid defervescence and loss of diarrhea within 7 days; toxin generally is still present in the stool at the end of successful therapy. Relapses occur in 20–25% of patients treated with any of the three antibiotics and are heralded by recurrent symptoms within 7 days of the end of therapy. Relapses may be treated with another course of antibiotics if the seriousness of the colitis warrants specific treatment. Prevention of C. difficile colitis can be attempted by limiting the use of unnecessary broad-spectrum antibiotics and tailoring antibiotic therapy to culture results with drugs that have a narrow spectrum.
An alternative treatment for antibiotic-related pseudomembranous colitis involves administration of cholestyramine, which binds the toxin, leading to symptomatic improvement. However, this approach is not as efficacious as specific antibiotic therapy, should not be used in seriously ill patients, and may also bind oral vancomycin, rendering it inactive.
Understanding the details of the pharmacology of a drug and the pathogenesis of a disease occasionally leads to otherwise nonobvious use of the drug in the disease with gratifying rewards for having considered the hypothesis.
Treatment of gastrointestinal disorders
Nausea and Vomiting
Vomiting is a complex clinical behavior that results in the evacuation of stomach contents and involves coordinated activity of the gastrointestinal tract and the nervous system. It is frequently preceded by nausea (an unpleasant sensation that has been felt by most people but is difficult to describe) and retching, or contractions of the diaphragm and chest wall against a closed glottis.
The neurophysiology of vomiting has been studied in detail in cats since the 1950s. Findings in studies of cats have generally been very relevant to humans. In brief, a vomiting center, which is located in the dorsal portion of the lateral reticular formation of the medulla, does not itself carry out the act of vomiting but coordinates the many organs involved in the intricate act of vomiting. The vomiting center can be stimulated by a chemoreceptor trigger zone located in the area postrema of the medulla on the floor of the fourth ventricle.
The chemoreceptor trigger zone, in turn, is sensitive to chemical stimulation, including direct application of drugs such as apomorphine or toxins such as uremic plasma. Dopamine receptors in the chemoreceptor trigger zone likely play a role in the act of vomiting; dopamine agonists such as apomorphine or levodopa initiate vomiting, and dopamine antagonists such as metoclopramide or domperidone diminish vomiting. The vomiting center can also be stimulated by afferent nerve stimuli originating in the gut, the pharynx, the vestibular system, and probably other sites in the body; these stimuli travel via vagal fibers stimulated by mechanoreceptors (sensitive to distension or abnormal gut motility) or chemoreceptors (enterochromaffin cells).
Serotonin, acting on 5-HT3 and to a lesser extent 5-HT4 receptors, mediate the vomiting induced by chemotherapeutic drugs but not that caused by pregnancy or motion sickness. The latter appears to be modulated by drugs that affect the histamine H1 receptor and the muscarinic cholinergic receptor. Vomiting induced by stimulation of abdominal visceral afferent nerves by ingested toxins or chemoreceptor trigger zone stimulation by apomorphine is blocked by selective 5-HT3 receptor antagonists. Higher central nervous system influences may also affect the vomiting center.
The vomiting act is preceded by a prodrome consisting of nausea, cold sweats, pupillary dilatation, tachycardia, and salivation. Vagal efferent nerve fibers relax the Proximal stomach, contract the esophagus longitudinally and pull the proximal stomach into the thorax, evacuate the upper small intestine into the stomach via a retrograde giant contraction, and empty the lower intestine into the colon. The vomiting of gastric contents is caused by the compression of the stomach through a strong simultaneous contraction of the diaphragm and abdominal muscles.
Numerous conditions can cause nausea and vomiting, and most can be explained from these neurophysiological relationships. Structural abnormalities of the gastrointestinal tract may lead to nausea and vomiting by either of two general mechanisms: mechanical obstruction of a hollow viscus (such as congenital pyloric stenosis, achalasia, or Crohn’s disease of the small intestine) or by a nonobstructing lesion affecting any or all components of the wall of gastrointestinal organs (such as an antral peptic ulcer, erosive gastritis, acute cholecystitis).
Systemic conditions may have local or distant effects on the gut or on any neuromuscular component of the vomiting reflex; examples are drugs, diabetes mellitus (either through ketoacidosis or local effects on the innervation of the stomach leading to gastroparesis), uremia, pregnancy, adrenal insufficiency, or infiltration by mass lesions (tumors, amyloid) in critical areas of the nervous system or the intrinsic musculature of the stomach wall. Finally, psychiatric illness may manifest itself as predominantly a vomiting disorder, most vividly seen in bulemia. These conditions and others that can lead to nausea and vomiting have been discussed in detail.
The therapy of nausea and vomiting must first be directed at identifying the underlying condition that has precipitated the problem. Healing a pyloric channel ulcer with antisecretory therapy or the surgical removal of an acutely inflamed, calculous gallbladder is the appropriate therapy to manage the vomiting that frequently accompanies these disorders. However, there are many conditions accompanied by vomiting that have no specific treatment (e.g., viral gastroenteritis or hepatitis) or have vomiting as a predictable effect of therapy (e.g., cisplatin chemotherapy, total nodal radiation). In these cases, the vomiting must be managed without being able to address its cause directly.
Drugs for nausea and vomiting have typically been tested in patients given highly emetogenic chemotherapy. Lessons from these experiments may not apply to nausea and vomiting caused by vestibular disorders or other central nervous system causes. Management of vomiting in situations for which therapy is indicated must take into account the neurophysiological correlates of vomiting as well as learned responses resulting from prior associations. This has been best studied in situations for which vomiting is predictable and severe, such as that caused by the intravenous infusion of high doses of cisplatin during the therapy of various solid tumors. Even if there has been no prior experience with emetogenic chemotherapy, one must expect the patient to develop purely anticipatory nausea and vomiting. Such nausea and vomiting can be triggered by sights, odors, or any other memory associated with prior episodes of vomiting.
These symptoms have been successfully treated with behavioral modification techniques or anxiolytic–amnesic agents such as lorazepam, or both.
Gastric distension can be avoided by beginning therapy following an overnight fast or, if there is an impediment to emptying the stomach, by evacuating its contents using a nasogastric tube. Maintaining adequate hydration is an important goal during potentially dehydrating therapy, and a large-bore intravenous line with adequate fluid replacement must be used. Prevention of nausea and vomiting, rather than rescue therapy, should be the goal of treatment.
Drug therapy of vomiting is aimed at the interruption of the vomiting reflex at any and all levels. Phenothiazines such as prochlorperazine were first used for this purpose in the 1950s, and since that time, agents with similar, substantial antidopaminergic effects have been used. These include the butyrophenones such as droperidol and substituted benzamides such as metoclopramide. Corticosteroids such as dexamethasone or methylprednisolone have been found to be useful, although the nature of their antiemetic action is unclear. Similarly, natural or synthetic cannabinoids have been used in antiemetic therapy following the empirical observation of an antiemetic effect in habitual users who were undergoing chemotherapy. However, their mechanism of action is unknown, and their efficacy has not yet been adequately demonstrated in well designed studies.
Ondansetron, a selective inhibitor of 5-HT3 receptors, was observed to prevent the vomiting induced by cisplatin in laboratory animals and patients. Studies also suggested that cisplatin treatment increased the release of serotonin from enterochromaffin cells, as measured by urinary excretion of 5-hydroxyindolacetic acid and plasma chromogranin A levels, thus providing a possible mechanism for cisplatin-induced emesis and the beneficial effect of the drug. In addition, 5-HT3 receptors are concentrated in the area postrema — the location of the chemoreceptor trigger zone and the entry site for most vagal afferents. Thus, there is evidence for both a peripheral and central action of serotonin in the pathophysiology of vomiting.
Because single agents lack universal effectiveness in the prevention of vomiting, regimens using combinations of drugs are sensible and useful. The goal of combining drugs for this indication is the same as for any indication for which combinations are used. The combination should result in a nausea-free patient and should minimize or abolish the potential adverse effects of the high doses of agents that would be necessary if they were used alone. One should choose combination therapy using agents whose efficacy is proved, whose modes of action are different, and whose potential toxicities are nonoverlapping.
Drug combinations can be useful when single agents fail to provide the desired efficacy or freedom from toxicity. Knowledge of the mode of action, pharmacology, and toxicity of drugs as single agents is useful in the rational planning of combination therapy. Drugs may have additive or even unique toxicities when used together, and the therapist must be alert to make observations that may not be predicted by studies on single agents.
The information learned from drug studies of patients receiving highly emetogenic drugs is applicable to other clinical situations and should be used to plan therapy of patients who vomit for other reasons. For example, the routine care of patients with severe vomiting from an exacerbation of chronic pancreatitis should include nasogastric suction to empty the stomach, fluid and caloric support by the parenteral route, and prevention or treatment of nausea and vomiting as necessary to maintain patient comfort. This does not exclude therapy that may be specific to the underlying disease, such as surgery to drain pseudocysts or endoscopic sphincterotomy to remove a common bile duct stone. Therapy that is specific to the underlying disease should always be a part of the patient’s management.
Peptic Ulcer Disease
Failure of the gastroesophageal junction to provide a barrier to gastric and duodenal contents can result in reflux esophagitis. Although heartburn is an extremely common complaint, and is both recurrent and generally not progressive, up to 20% of patients referred to a specialist with symptomatic reflux esophagitis may develop complications such as esophageal stricture or Barrett’s esophagus, a premalignant condition. Although antacids and lifestyle changes (low-fat diet, elevation of the head of the bed, no recumbent position after meals, no smoking) can improve the symptoms of patients with reflux esophagitis, nificant minority of patients have chronic and disabling symptoms that require more aggressive management.
The success of therapy for reflux esophagitis, whether measured by symptomatic relief or by improvement in endoscopic and histological evidence of inflammation of the esophagus, is directly related to two factors: the degree of acid suppression and the improvement in emptying of esophageal contents. Therapy with proton pump inhibitors, either used singly or in combination with prokinetic agents such as cisapride, is superior to therapy with H2 blockers in both healing reflux esophagitis and in preventing recurrence.
However, therapy is not curative because the underlying motor disorder that results in an incompetent gastroesophageal barrier is not fixed by treatment. Also, not all patients respond to therapy: reflux of alkaline stomach contents or bile salts may perpetuate the esophageal injury, and eradication of H. pylori infection may heal the gastric mucosa and increase stomach acid production. This results in a minority of patients requiring a surgical antireflux procedure because of complications (stricture, nonhealing ulcers, bleeding), or persistent symptoms, particularly in young patients.
Diarrhea is generally defined as the passing of watery stools or an increased frequency of relatively loose stools. Acute diarrhea is a common condition. It affects adults in developed countries once a year on the average. This illness is not likely to lead to consultation with a healthcare worker unless the patient is an infant or small child, or unless the illness is particularly severe. Chronic diarrhea, lasting longer than 3 weeks, is an uncommon condition. It requires a diagnostic workup that can be extensive and often requires a specialist. After illnesses that are managed with specific therapy have been identified and treated, the management of either acute or chronic diarrhea may be similar.
Acute diarrhea is generally regarded as an attempt by the gastrointestinal tract to get rid of disease-causing microorganisms and toxins. This is believed to be adaptive, and therapy meant merely to decrease the number and volume of stools is generally not recommended. More likely, however, diarrhea provides an efficient way to disseminate and to propagate the organisms that cause it and is an adaptive mechanism of the parasite and not the host.
The mechanisms by which diarrhea occurs fall under four general categories: increased osmolality of intestinal contents, decreased fluid absorption, increased intestinal secretion, and abnormal intestinal motility. Any of these mechanisms may be responsible for diarrhea from any given cause. For example, loss of mature intestinal cells in villus tips due to an acute rotavirus infection is likely to lead to decreased mucosal absorptive surface, decreased fluid absorption, and a self-limited lactose intolerance and osmotic diarrhea with ingestion of milk. Escherichia coli or Vibrio cholera infections cause diarrhea through an enterotoxin that causes net excretion of chloride by the enterocyte.
The ability of acute diarrhea to cause severe dehydration in children, which often leads to death in areas of the world where poverty and malnutrition are common, makes it an important worldwide health-care problem. In industrialized countries, acute diarrhea is likely to be due to a viral agent (rotavirus, Norwalk virus, or similar viruses) and requires no specific treatment. Travelers to less developed nations are exposed to diarrheal illnesses not common in their native countries (cholera, enterotoxigenic E. coli infections, Entamoeba histolytica) and must be aware of the symptoms and correct management of these illnesses.
When diarrhea occurs in a traveler, is accompanied by signs of dysentery (temperature greater than 103°F, systemic symptoms, bloody stools, or severe abdominal or rectal pain), or lasts longer than 14 days, one must consider a bacterial or protozoan cause. Further evaluation is necessary to determine whether specific antimicrobial therapy will be necessary.
For the vast majority of people with diarrhea who do not have an invasive infection and will have a self-limited condition, the goal of therapy is to maintain an adequate state of hydration. Simple measures such as avoiding substances that may increase intestinal secretion and motility (caffeinated beverages, alcoholic beverages, spicy foods, milk products) and adequate intake (2–3 L or more per day) of fruit juices and noncarbonated beverages generally are sufficient.
Oral solutions of rehydration represent a major advance in the therapy of severe diarrhea. They take advantage of glucose-coupled sodium uptake and solvent drag in the small intestine. Both are processes that result in absorption of sodium and free water even in the face of bacterial toxin-induced secretory diarrhea. When moderate or severe dehydration is already present and the potential for further dehydration is high (such as during cholera), a solution high in sodium is necessary in order to prevent hyponatremia.
The World Health Organization has recommended an oral rehydration solution containing 90 mEq/L of sodium, 20 mEq/L of potassium, 80 mEq/L of chloride, 30 mEq/L of bicarbonate, and 20 g/L of glucose. For a less serious degree of dehydration or to prevent it from occurring, solutions containing 45 to 50 mEq/L of sodium are commercially available (e.g., Infalyte powder, Pedialyte liquid). If dehydration is very severe (>10% of body weight loss) or if the diarrheal illness is accompanied by vomiting or inability to comply with oral fluid therapy, then intravenous fluids will be necessary.
If diarrhea is not accompanied by signs suggestive of an invasive infection, then symptomatic therapy with antidiarrheals should be considered. Two different classes of agents have been proved to be useful: opiates and nonsteroidal anti-inflammatory drugs. Adsorptive compounds such as kaolin or pectin have been used to treat diarrhea for centuries. They alter stool composition, turning loose stool into lumpy stool. They do not decrease stool volume or frequency and are not recommended.
Opiates (e.g., diphenoxylate with atropine, loperamide, deodorized tincture of opium) decrease intestinal motility, increase mucosal absorption, and decrease fluid and electrolyte secretion. Presumably they act on intestinal mu opioid receptors. Their net effect is to reduce stool volume and to alleviate tenesmus and abdominal cramps. Loperamide (4 mg at the onset of diarrhea, and 2 mg after each bowel movement not to exceed 16 mg in 24 hours) is useful for treating “traveler’s diarrhea” from multiple causes. A theoretical advantage of this drug is that it crosses the blood–brain barrier poorly and is likely to have few central nervous system effects. Opiates are not recommended for children under 2 years of age because they can blunt alertness and interfere with oral rehydration therapy.
Although the nonsteroidal anti-inflammatory drugs such as aspirin or indomethacin can decrease stool volume in the setting of acute infectious diarrhea, the effect is not to the degree that would make them clinically useful. Bismuth subsalicylate in large doses (30 to 60 mL every 30 minutes for eight doses following the onset of diarrhea) can decrease stool frequency and abdominal pain in mild-to-moderate acute, self-limited diarrhea. However, this amount of salicylate may lead to toxic salicylate concentrations in the blood. It is not recommended for patients with renal failure or with concomitant use of other salicylates. Loperamide is more effective for treating severe diarrhea.
Specific antimicrobial treatment is currently recommended for symptomatic cases of diarrhea caused by Shigella spp., C. difficile, Salmonella typhi with typhoid fever, Giardia lamblia, Entamoeba histolytica, and Vibrio cholerae. Certain patients with E. coli (enterohemorrhagic or enterotoxigenic E. coli, infants with enteropathogenic or enteroadherent E. coli), Salmonella infections without typhoid fever, or prolonged Campylobacter jejuni diarrhea may also benefit from specific antimicrobial treatment. Details of specific antimicrobial therapy change frequently and are best obtained from a frequently revised guide. However, most bacterial causes of diarrhea can be treated effectively with 500 mg ciprofloxacin twice daily for 5 days. This empirical therapy may be used while awaiting the results of specific cultures.
Therapy for chronic diarrhea is generally directed at the underlying disease. If the underlying disease cannot be identified or cured, then the general principles of management of acute diarrheal states are implemented. Other drugs may be considered to manage difficult cases of chronic diarrhea. Clonidine and lithium carbonate increase sodium chloride absorption in the gut; they have been used in chronic secretory diarrheas due to tumors elaborating vasoactive intestinal peptide (VIPomas) with limited success. Somatostatin or its analog octreotide decreases fluid and electrolyte secretion, decreases intestinal motility, and may decrease the release of a secretagogue from tumors such as VIPomas or nonmalignant tissue. Because of its reduction of release of growth hormone, somatostatin should not be used in children.