Mack C. Mitchell Jr.
Constipation
Definition
Constipation is often defined as the infrequent, difficult passage of stool. However, it may mean different things to different people: that the stools are too infrequent, too difficult to expel, too hard, or too small, or that there is a sensation of incomplete evacuation. Of these, frequency of bowel movements is the most easily measured. Several studies have identified a wide variation in the frequency of bowel movements among normal subjects of both sexes and of all ages, ranging from three per day to three per week (1). Therefore, someone who has fewer than three bowel movements per week is constipated, by definition. On the other hand, a change in frequency of movements, for example, from two per day to three per week, may also signify constipation. However, it is probably not necessary to evaluate or to treat people merely because they report fewer than three bowel movements per week. Even one movement per week is acceptable if it does not represent a recent change in bowel frequency and is not associated with symptoms such as pain on defecation or bloating. Constipation is among the most prevalent of gastrointestinal (GI) complaints, accounting for more than 2.5 million visits to health care providers in the United States annually and more than half a billion dollars in sales of laxatives (2). The prevalence of constipation increases with age and is more common in women and in lower socioeconomic groups (2,3).
Almost always, constipation is caused by a delay in transit within the colon or through the pelvic outlet (4). A wide variety of conditions can affect colonic transit (Table 45.1). There may be structural abnormalities that obstruct the passage of intraluminal contents, or there may be conditions that alter colonic motility. Evaluation of patients with constipation must therefore include consideration of a wide variety of possible etiologies. Although it is difficult to be precise about the relative frequencies of the various causes, chronic constipation (months or longer) is most commonly the result of a motility disorder, the use of constipating drugs, local anorectal problems (fissures, hemorrhoids, tumors), or pelvic outlet delay.
Evaluation
History
The history may reveal a gross misconception about normal bowel habits or a neurotic preoccupation with bowel function. Reassurance that there is a broad range of normal bowel frequencies may be all that is needed in many
P.688
cases of self-defined constipation. It is important to determine the onset and duration of constipation and whether there has been a recent change in frequency of bowel movements. Associated symptoms such as straining, abdominal pain, bloating, and distention may shed light on the severity of symptoms. A detailed dietary history should be obtained and the amount of dietary fiber intake should be estimated. Patients should be questioned about symptoms that suggest a systemic process (e.g., hypothyroidism, hyperparathyroidism, scleroderma) or a neurologic disorder (e.g., cerebrovascular disease, Parkinson disease). They should also be questioned about the use of medications (e.g., anticholinergics, calcium channel blockers, opiates, antidepressants) that are known to impair colonic motility. Most systemic and neurologic diseases affect organs outside the GI tract so that, in addition to constipation, patients have symptoms that reflect extra-intestinal dysfunction. On the other hand, even local processes (e.g., strictures, tumors) produce other GI symptoms in addition to constipation, such as abdominal pain or rectal bleeding. Rectal bleeding should always be thoroughly evaluated (see later discussion), even though in constipated patients it often is caused by perianal disease (e.g., fissures, hemorrhoids). Abdominal pain with constipation is also a prominent feature of the irritable bowel syndrome (see Chapter 44). Most patients with idiopathic diet-related or drug-induced constipation are otherwise asymptomatic, although a complaint of a bloated sensation is common if constipation is prolonged. A history of the type, frequency, and duration of laxative use may highlight a dependence on laxatives or may even suggest a psychological disorder if the laxative use is inappropriate.
|
TABLE 45.1 Causes of Constipation |
|
|
Physical Examination
The physical examination should attempt to identify underlying causes of constipation. Rectal examination can identify fissures and hemorrhoids as well as anal stenosis or stricture secondary to surgery or inflammation. The anal sphincter is normally closed. A gaping anal opening or asymmetry of the anal opening may indicate a neurologic disorder (spinal cord trauma, peripheral neuropathy) that impairs sphincteric function. After inspection, a careful digital examination should be performed to evaluate the strength of the anal sphincters, the presence of masses, the consistency of the stool, and the presence of any painful or tender areas. The patient should be asked to strain and the amount of perineal descent should be noted. In general there should be 2 to 3 cm of perineal descent with straining. In patients with pelvic floor disorders, there may be no descent or paradoxical perineal ascent. The stool should also be tested for occult blood.
Laboratory Tests
Before any endoscopic evaluation, basic laboratory testing is sometimes helpful. Serum creatinine, urea nitrogen and electrolytes, especially Na+, K+, Ca2+, and Mg2+, should be checked. Thyroid function tests should be performed if hypothyroidism is suspected.
Endoscopy
Although most gastroenterologists agree that endoscopy should be performed for newly constipated patients in whom the cause is not obvious, there is no firm evidence on which to base this recommendation (5). Patients over the age of 50 years should have colonoscopy because of their increased risk of colon cancer. Air-contrast barium enema is not a substitute for endoscopy, because subtle mucosal abnormalities cannot be detected and because the most distal 15 to 20 cm of the colon are difficult to evaluate
P.689
radiographically. Anoscopy also may be needed to search properly for hemorrhoids or fissures (see Chapter 98).
In patients younger than 40, flexible sigmoidoscopy is probably a reasonable substitute for colonoscopy unless there are worrisome symptoms such as GI bleeding or weight loss.
Flexible sigmoidoscopy can be performed in the practitioner's office with the patient in the knee–chest position on a routine examining table or in the left lateral position. Sigmoidoscopy should be performed to the most proximal level possible, with limitations imposed by the patient's tolerance of the procedure, the presence of stool, and the length of the instrument. Chapter 46 describes the patient's experience during flexible sigmoidoscopy and colonoscopy.
Inflamed hemorrhoids and fissures found during anoscopy or flexible sigmoidoscopy may be secondary to constipation, but they may also cause pain on defecation and thereby promote constipation. These lesions are also common causes of bleeding in the chronically constipated patient. A spotty or diffuse brown pigmentation of the mucosa, known as melanosis coli, is indicative of chronic laxative abuse, particularly of the anthraquinone family (e.g., cascara, senna, aloe). A mass lesion, such as a carcinoma or polyp, may also be identified by flexible sigmoidoscopy or colonoscopy.
A rectal biopsy may diagnose amyloidosis, ulcerative colitis, or Crohn disease, and a deeper, suction biopsy of a rectal valve may diagnose Hirschsprung disease. Biopsies of the rectal mucosa can be taken safely below the peritoneal reflection (approximately 12 cm proximal to the anus in men, 8 cm in women). Punch biopsies can be performed by any clinician who has had appropriate training and experience. Rectal biopsy above the dentate line should be painless and the risks of bleeding or perforation are low.
Radiographic Studies
Radiographic examination is helpful primarily in the detection of obstructing lesions. The presence of a megacolon or a volvulus, of either the sigmoid colon or the cecum, may be easily diagnosed by a plain film. Some practitioners obtain a barium enema after flexible sigmoidoscopy in all patients with new-onset constipation, rather than simply performing a colonoscopy. As stated earlier, colonoscopy is the preferable procedure, although the two approaches have not been compared prospectively. Obstructing neoplasms and strictures can be identified by barium studies, although more subtle mucosal lesions may be missed.
Other Studies
Colonic motility tests and transit studies are additional procedures that may provide insight into the pathophysiology of constipation. These studies are generally available only in specialized centers. They should be performed for the few patients who are severely impaired by constipation and who are refractory to conventional therapy.
Colonic motility studies are performed by placing catheters to monitor intracolonic pressures in the rectal and sigmoid regions. The studies can identify various patterns of colonic activity in patients with constipation (6). In some, high-amplitude phasic contractions are seen spontaneously, as well as in response to stimulation. This type of segmental activity is sometimes associated with pain and is believed to cause constipation by impeding the distal flow of luminal contents. In other patients, an atonic pattern is found, characterized by a decreased response to stimulation and by a loss of resistance to distention. Anorectal manometry is helpful in excluding Hirschsprung disease if it demonstrates a normal rectoanal inhibitory reflex. It may also provide useful information regarding pelvic floor function (6).
Colonic transit time can be measured by plotting the expulsion of radiopaque markers (Sitzmarks, commercially available) after daily ingestion of a capsule containing 24 markers. Once a steady state is reached, the number of markers entering the GI tract must equal the number being excreted, and the transit time in hours is equal to the number of markers still in the colon on plain film (7). Moreover, the distribution of these markers may have a relationship to the underlying motility disturbance. The retention of markers only in the rectum suggests a failure of expulsion, whereas retention throughout the colon suggests generalized colonic inertia.
Treatment
The treatment for constipation should, if possible, be based on correction of the underlying abnormality; for example, a systemic disease that can be treated (e.g., hypothyroidism) or a constipating drug that can be stopped. The routine use of laxatives for treatment of all complaints of constipation should be discouraged. Successful therapy should include an effort to educate the patient about the broad limits of normal bowel function and to understand the patient's own concepts of normal bowel activity.
Bowel Retraining
Bowel retraining is an important initial aspect of therapy for patients whose constipation does not have an identifiable and remediable cause. The patient should be encouraged to have a regular daily routine with time set aside for having a bowel movement, preferably within 5 to 10 minutes after a meal, to take advantage of the strong stimulus of the gastrocolic reflex. This behavior modification program allows the patient to become more aware of, and responsive to, the normal urges to defecate. Patients should
P.690
be advised always to respond to such urges. In severely constipated patients, a bowel retraining program may be initiated, with enemas or suppositories used to enhance bowel activity at the desired time. For enemas, lukewarm tap water should be used, because all other solutions may be irritating if used repetitively. The enema should also be done soon after eating a meal to take advantage of the gastrocolic reflex. For suppositories, bisacodyl (Dulcolax) should be used, again just after eating; however, these suppositories should not be used for more than a few days, because they may eventually be irritating. Glycerin suppositories are commonly used but are not very effective.
Diet
Dietary fiber has long been considered an important factor in bowel function although there is little evidence to support this conclusion (5). However, patients with constipation often respond to an increase in fiber in their diet, and there is evidence that oroanal transit times are decreased when intake of dietary fiber is increased (8). As an initial step, the patient with constipation should be advised to follow a diet rich in fiber. As listed in Table 45.2, a variety of foods are high in fiber. As a practical matter, it may be reasonable to add a commercial fiber preparation to a high-fiber diet. Maintaining an adequate intake of fluids (1.5 to 2 L/day), although particularly difficult for some older patients with constipation, is also helpful (9).
Laxatives
Laxatives are popular for the treatment of constipation. The presence of 700 or more commercially available laxatives and enema preparations attests to the widespread use of these agents. The mechanism of action for most laxatives is poorly understood, and the potential for toxicity is often underestimated (Table 45.3). Few data are available for comparison among the various laxatives, and the decision to use a particular laxative often is determined by individual preference rather than by objective evidence of efficacy or safety.
There are a number of different mechanisms by which a laxative effect may be achieved. Bulk-forming agents are natural or synthetic polysaccharides or cellulose derivatives that exert their laxative effect primarily by retaining water within the lumen of the colon and increasing fecal mass. Psyllium seed and bran are examples of such laxatives. In addition to their hydrophilic properties, these agents are metabolized by colonic bacteria, resulting in accumulation of osmotically active metabolites. These laxatives are effective in increasing the frequency and in softening the consistency of stool. There have been isolated reports of obstruction secondary to use of hydrophilic agents in patients with esophageal or small-bowel strictures. In most cases, however, this type of laxative is highly effective and the potential for adverse effects appears to be low. However, in patients with an atonic form of constipation, particularly with megacolon, these agents are often ineffective and produce an uncomfortable sensation of bloating and gaseousness at high dosages. Bloating is common in all patients when they begin to take bulk agents, but it is usually transient and can be minimized by increasing the dosage gradually over a period of weeks. A meta-analysis found that laxatives and fiber increased the frequency of bowel movements by a mean of 1.4 times per week (10). Fiber and bulk laxatives may also reduce pain related to constipation.
|
TABLE 45.2 Fiber Content of Various Foods |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Polyethylene glycol (PEG) is a nonabsorbable saccharide that increases the osmotic content of the stool and helps to retain water within the stool. It is available both in a powdered form (Miralax) and in solutions (Go-lytely,
P.691
Nu-lytely, Colyte) that are frequently used as purgatives before colonoscopy or radiographic procedures. There is good evidence from a systematic review to support the efficacy of PEG in the treatment of chronic constipation (11).
|
TABLE 45.3 Laxatives |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Lactulose (Cephulac or Chronulac syrup) is a nonabsorbable semisynthetic disaccharide. As a result, water and electrolytes are retained within the intestinal lumen by the osmotic effect of this undigested sugar. In addition, this agent is converted by colonic bacteria to organic acids, which may further alter electrolyte transport or affect colonic motility, or both. Lactulose is commonly used in patients with hepatic encephalopathy (see Chapter 47). It has also been shown to be an effective laxative in patients with chronic constipation. There is little information on the relative merits of lactulose and bulk laxatives except that lactulose is expensive and requires 24 to 48 hours to achieve its effect.
Docusate sodium (Colace) is often labeled as a stool softener or a wetting agent. It works by lowering surface tension, allowing water to enter the stool more easily. It is generally well tolerated, but may not be effective in many patients with severe constipation or irritable bowel syndrome (IBS).
P.692
The saline laxatives are magnesium or sodium salts (e.g., milk of magnesia, sodium phosphate) that are poorly absorbed and therefore act as hyperosmolar solutions. Complications include hypermagnesemia in patients with renal failure and hypocalcemia from phosphate overdoses.
Stimulant laxatives, such as anthraquinone derivatives (senna, aloe, cascara) and diphenylmethane compounds (phenolphthalein, bisacodyl), exert their effects primarily by altering electrolyte transport by the intestinal mucosa and increasing intestinal motor activity. Phenolphthalein, an ingredient found in many over-the-counter (OTC) preparations, has been associated with severe allergic dermatitis and the Stevens–Johnson syndrome. The chronic use of the anthraquinone derivatives has been reported to induce damage to the myenteric plexus and therefore may eventually impair bowel motility. Agents such as these that affect electrolyte transport may result in significant hypokalemia, factitious diarrhea, protein-losing enteropathy, and salt overload.
Castor oil exerts its cathartic effect by alteration of intestinal fluid and electrolyte secretion. Ricinoleic acid, the active ingredient of castor oil, has effects on the small and large intestine similar to those of bile acids: It inhibits absorption of sodium and glucose and stimulates fluid and electrolyte secretion. Its use is not recommended because of the potential for fluid–electrolyte disturbances.
Tegaserod (Zelnorm) is a serotonin-4 (5-HT4) receptor agonist that is effective for the treatment of either constipation-predominant IBS or chronic constipation. Tegaserod improves the frequency of bowel movements in women with IBS and in those with chronic constipation in the absence of associated abdominal bloating or discomfort. Tegaserod works by acting on the 5-HT4 receptor, increasing gut motility.
A systematic review showed that the evidence of efficacy in the treatment of constipation was good for tegaserod as well as for polyethylene glycol and moderate for psyllium and lactulose (11). There was very little evidence showing efficacy of milk of magnesia, senna, bisacodyl, or stool softeners.
Surgery
Surgery is rarely necessary in the treatment of constipated patients. However, it is required for resection of an obstructing lesion, and myectomy may be needed for treatment of Hirschsprung disease. Various procedures have been recommended for patients with megacolon with recurrent volvulus, ranging from simple tacking down of the loose mesentery to resection of bowel. Finally, in severe cases of intractable constipation, extensive surgery has been advocated, varying from resection of redundant sigmoid loops to subtotal colectomy with ileal proctostomy. The exact role and precise indication for this type of surgery remain to be defined more clearly, but there is evidence that subtotal colectomy is effective and well tolerated in patients with severe atonic constipation (12).
Summary of Recommendations
The approach to management of constipation should be directed first at identifying and treating any underlying disorder (Table 45.1). If constipation is drug-induced, the drug should be discontinued (if possible) or an alternative that is less constipating should be substituted. When constipation is idiopathic, caused by an irreversible underlying disorder (e.g., diabetes mellitus), or secondary to a necessary drug, dietary changes and some form of laxative therapy may be necessary.
In many patients with constipation, a high-fiber diet or a bulk laxative along with an increase in dietary fluid intake is helpful. The amount of dietary fiber or bulk laxative that is needed varies from patient to patient and must be determined individually. Because the only significant side effect from this form of therapy is excessive gas and a bloated sensation, the dosage can be gradually increased until either the constipation is resolved or the side effects become too uncomfortable. For most patients, this form of therapy is successful and appears to be very safe on a chronic, defined (i.e., not as-needed) basis. Otherwise, tegaserod, polyethylene glycol, psyllium, or lactulose should be prescribed. No other laxatives should be recommended except under special circumstances.
In patients with partial bowel obstruction (as recognized on radiography) and in those with an atonic form of constipation (e.g., institutional megacolon), the high fiber–bulk laxative approach usually is not effective. A patient with constipation caused by partial obstruction usually must be treated surgically. A patient with an atonic colon may require a stimulant laxative, if there is an inadequate response to tegaserod or PEG. Senna compounds, bisacodyl (either as a tablet or as a suppository), or enemas may be effective in such cases. Combinations of a stimulant laxative with a bulk agent (e.g., Perdiem) or with a softener (e.g., Peri-Colace) are another option. Bulk agents alone in such cases simply distend the already distended bowel further without improving bowel function.
Special consideration must be given to the bedridden or chair-bound patient. In such patients, use of laxatives may result in incontinence because the patient may not be able to recognize or respond quickly enough to the sudden urge to defecate. In these circumstances, bulk agents are useful to keep the stool soft (Table 45.3), but suppositories or enemas should also be used (simple tap water enemas are usually sufficient), with the patient already positioned on the commode. In this fashion, the embarrassment and soilage of fecal incontinence can be avoided and fecal impactions can be prevented. Enemas or suppositories may be used regularly (daily to every third day) to prevent fecal impactions and overflow diarrhea and incontinence. A
P.693
pamphlet for patients, entitled, What Is Constipation?, is available from the National Institute of Arthritis, Diabetes, Digestive and Kidney Diseases. (This pamphlet may be ordered [and then duplicated] from Clearinghouse, 1555 Wilson Boulevard, Suite 600, Rosslyn, VA 22209-2461.)
Diarrhea
Diarrhea is a common and troublesome problem. In infants and children, diarrhea can rapidly lead to death if untreated. The frail elderly also are at increased risk for strokes or other major ischemic events, if they become volume depleted. In most adults, diarrheal illnesses begin abruptly, last only a day or two, and resolve without serious sequelae (see Chapter 35). The illness seldom continues for longer than 1 week. The task facing the clinician is to identify the few patients with a significant underlying disorder who may require a specific therapeutic approach.
Definition
Patients who complain of diarrhea usually have an increase in both the frequency and fluid volume of the bowel movement. Stool weight is the best objective measurement of diarrhea, with mean weights in the United States ranging normally between 100 and 200 g/day. In the patient with chronic diarrhea, objective documentation of daily fecal output is sometimes necessary (see later discussion). Most patients with significant diarrhea produce more than 200 g of stool per day. However, patients with inflammatory conditions or space-occupying lesions of the rectum may present with the frequent passage of small volumes of liquid stool. Patients with the IBS have stool volumes either within or slightly above the normal range. Patients with secretory forms of diarrhea or small-bowel disorders with malabsorption often pass very large volumes of stool, in the range of 500 to 1,000 g/day or more.
Pathophysiology
There are four basic mechanisms of diarrhea: increased osmotic load within the intestine, excessive secretion of electrolytes and water into the intestinal lumen; exudation of protein and fluid from the intestinal mucosa; and altered intestinal motility resulting in rapid transit through the colon.
Osmotic diarrhea results from retention of orally ingested substances that are poorly absorbed causing an influx of water into the lumen of the bowel. This mechanism operates in patients with malabsorption or with lactose intolerance, in which undigested sugars accumulate within the intestinal lumen and exert a considerable osmotic load. Magnesium-containing laxatives and some magnesium-containing antacids (e.g., Maalox) can produce diarrhea through a similar mechanism.
Secretory diarrhea results from increased mucosal secretion of water and electrolytes under the stimulation of a variety of substances. The profuse watery diarrhea in patients with cholera results from increased secretion stimulated by cholera toxin, but a number of other enterotoxin-producing organisms (e.g., enterotoxigenic Escherichia coli) produce diarrhea in the same way (see Chapter 35). Other substances that induce secretory diarrhea include bile acids and long-chain fatty acids (e.g., after ileal resection, in Crohn disease, with a malabsorption syndrome), certain GI hormones, and anthraquinone laxatives.
Exudative diarrhea results from the outpouring of protein, blood, or mucus from an inflamed or ulcerated mucosa. Ulcerative colitis, Crohn disease, invasive infections (see Chapter 35), and infiltrative disorders such as Whipple disease or lymphoma are examples of this mechanism.
Motility disorders may lead to diarrhea, although the exact correlation between the abnormal motility and the diarrhea is not completely understood. The IBS (see Chapter 44) is generally believed to be a motor disorder that causes abdominal pain and altered bowel habits, with diarrhea predominating in some patients. Diabetes mellitus may also lead to diarrhea caused by neurogenic dysfunction. Other conditions, such as scleroderma, can lead to stasis of the bowel with resultant bacterial overgrowth, steatorrhea, and diarrhea.
In some patients more than one mechanism may be responsible for diarrhea. An appreciation of pathophysiology enables the clinician to understand the clinical features of a diarrheal illness better and to select appropriate therapy.
Evaluation of Acute Diarrhea
Acute diarrhea caused by infectious agents or by ingested toxins is discussed in greater detail in Chapter 35. Most patients who present to the clinician with a sudden onset of diarrhea have a benign, self-limited illness. These patients do not require extensive evaluation and can simply be reassured. However, a small percentage of such patients have a significant underlying illness for which specific therapy is needed.
If diarrhea persists for longer than 72 hours or if there is gross blood in the stool, an evaluation is indicated. In any case, the patient should always be evaluated before medicine is prescribed, because in certain situations even nonspecific antidiarrheal therapy may be harmful. In particular, it has been shown that opiate-containing antidiarrheal drugs such as diphenoxylate (Lomotil) or loperamide (Imodium) can prolong the course of acute infectious diarrhea by hindering the natural mechanisms that clear the body of the organism (13) and can potentiate the development of the hemolytic-uremic syndrome in patients infected with E. coli O157:H7 (14).
P.694
History
The history may provide clues to the underlying cause. The sudden onset of loose, watery stool is most commonly caused by an infectious process, and much less often by ingestion of drugs or poisons. Often, no specific bacterial agent is identified and the syndrome is assumed to be a viral gastroenteritis. Bacteria can cause diarrhea by direct mucosal injury or by elaboration of a toxin that produces intestinal dysfunction. Toxin-induced diarrhea, often associated with vomiting, begins within 6 hours after ingestion of contaminated food, whereas diarrhea because of direct mucosal invasion does not begin for 12 to 24 hours. Examples of toxin-induced diarrhea are enterotoxigenic E. coli infection and staphylococcal, clostridial, and Bacillus cereus food poisoning. Bloody diarrhea should never be ascribed to viral or toxin-mediated diarrhea; it is more likely to be caused by bacterial infection (Shigella, Campylobacter, Yersinia, Salmonella, enterohemorrhagic and enteroinvasive E. coli), ulcerative colitis, diverticulosis, ischemic bowel disease, or radiation colitis. Information about recent travel (seeChapter 41) should include not only trips out of the country but also camping or fishing trips within the United States. Giardia may contaminate water supplies and cause both epidemic outbreaks and individual cases of acute diarrhea among campers and hikers. In cases of giardiasis, lactose intolerance commonly occurs and may confuse the diagnosis. Acute onset of lactose intolerance should prompt consideration of giardia as a cause. Recent use of drugs is a common cause of new-onset diarrhea (Table 45.4) that may not be recognized by the clinician or the patient. Magnesium-containing antacids, broad-spectrum antimicrobials, hydralazine, and quinidine are commonly used drugs that can lead to diarrhea (see later discussion). A history of recent antibiotic therapy, multiple affected friends or family members, associated symptoms such as abdominal or rectal pain, tenesmus, or a history of anal intercourse may provide clues to the etiology of diarrhea.
A number of intestinal infections are seen more commonly in immunocompromised individuals (e.g., those with human immunodeficiency virus [HIV] infection, lymphoma, or immunosuppressive therapy after organ transplantation). Appropriate questions should be asked to identify patients who are at risk for HIV infection (see Chapter 39). Patients with CD4-positive T lymphocyte counts lower than 50 per cubic millimeter or viral burdens greater than 70,000 copies per milliliter are at particular risk for acquiring infections. In addition to being more susceptible to conventional bacterial and parasitic infections, immunocompromised individuals are at risk for diarrhea from unusual organisms such as Microsporidium, Cyclospora, Isospora belli, Cryptosporidium, Mycobacterium avium-intracellulare, Cytomegalovirus, Herpes simplex, and Candida.
|
TABLE 45.4 Common Drugs That May Induce Diarrhea |
||
|
Physical Examination
The physical examination in acute diarrhea is generally unremarkable. The patient's state of hydration should be estimated because it is an important measure of the severity of the diarrhea and of the need for hospitalization. Some inflammatory and infectious diarrheas are associated with intermittent fever. Abdominal examination usually reveals no more than mild diffuse tenderness. The bowel sounds usually are active or hyperactive. Rectal examination is essential because diarrhea may be the initial manifestation of obstructing rectal carcinoma; furthermore, in the geriatric population, fecal impactions may result in overflow diarrhea, and constipating agents may be mistakenly recommended.
Stool Examination
If diarrhea continues for longer than 3 to 4 days, the stool should be examined for the presence of blood, leukocytes, and enteric pathogens. Blood in the stool suggests mucosal disruption and is not a feature of osmotic, secretory, or motor causes of diarrhea. Inflammatory conditions such as ulcerative colitis, pseudomembranous colitis (see Chapter 35), Shigella and invasive strains of E. coli, often cause bloody diarrhea.
Fecal leukocytes are best seen by microscopic examination of the liquid portion of the stool after staining with methylene blue or Gram stain (see Chapter 35, where the technique is described). The latex agglutination test for the
P.695
neutrophil product lactoferrin is more sensitive and more specific and can be performed rapidly (15). Fecal leukocytes are not seen in infectious processes that do not invade the mucosa such as viral enteritis, in toxin-mediated diarrhea such as cholera, or in infection with noninvasive E. coli (see Chapter 35).
Invasive organisms such as Salmonella, Shigella, Amoeba, and Campylobacter, and idiopathic inflammatory bowel disease typically lead to exudation of fecal leukocytes. In these conditions, in which the mucosal barrier is broken, the course of the diarrheal illness is unpredictable and may even become life-threatening. The absence of fecal leukocytes or blood, even in a single specimen, suggests a viral or drug-induced cause, and in these cases the disease is usually transient and no therapy is required.
Stool cultures for bacterial pathogens should be obtained in all patients who have fecal leukocytes. Specific isolation techniques are needed to diagnose Yersinia and Campylobacter. Acute infectious diarrhea caused by Salmonella or Shigella or other invasive organisms is impossible to differentiate from acute inflammatory bowel disease without a stool culture. Gram staining of the stool for bacteria is not helpful.
Examination of the stool for parasites such as Entamoeba histolytica and Giardia lamblia is important if diarrhea persists for more than a few days, even in the absence of a history of travel, because the organism can be passed by contact with a carrier. Microscopic examination of the inflammatory exudate of a patient with acute amebic colitis almost always demonstrates motile trophozoites, but only if the slide is prewarmed (e.g., over a light bulb) and then examined immediately. Rectal biopsy may identify the organisms in the exudate. Serologic tests for amebae can be useful but do not distinguish recent from remote infection. Measurement of acute and convalescent titers is more specific for recent infection but gives the diagnosis only in retrospect. Giardia can be detected by microscopic examination of fresh stool in only about half of affected patients, but the enzyme-linked immunosorbent assay (ELISA) test for Giardia antigen in stool has a sensitivity of 96% and a specificity of 100% (16). This test becomes negative after eradication of the infection and should replace microscopic examination of stool for diagnosing Giardia infection. If there is a history of recent antibiotic use, stool specimens should be evaluated forClostridium difficile toxin. Several rapid tests for C. difficile toxins A and B have high sensitivities and specificities (17).
Endoscopy
Flexible sigmoidoscopy may be helpful in patients who have acute diarrhea associated with fecal leukocytes or bloody diarrhea if the cause is not obvious. The examination should be performed without a prior enema, because enemas may alter the appearance of the mucosa and reduce the chance of detecting intestinal pathogens such as amebae. The marginal diagnostic return in examining the entire colon is quite small, so flexible sigmoidoscopy is preferable to colonoscopy in this situation.
Many acute diarrheal illnesses produce a similar abnormal but nonspecific mucosal appearance on sigmoidoscopy. However, certain findings suggest specific diseases. In viral enteritis, giardiasis, toxin-mediated diarrhea, drug-induced diarrhea, and other conditions not accompanied by fecal leukocytes or blood loss, the sigmoidoscopy is normal. In ulcerative colitis, the rectal mucosa is involved in at least 95% of cases, and the mucosa is uniformly abnormal with bleeding and a granular, friable appearance. Crohn disease uncommonly affects the rectum, but it is often apparent in the sigmoid or more proximal portions of the colon. It may appear as discrete aphthoid ulcers or patches of grossly abnormal mucosa with normal intervening tissue. Amebiasis occasionally produces flask-shaped ulcers that may be single or multiple, with normal intervening mucosa; more often, however, it produces a pattern very similar to that of ulcerative colitis. Inshigellosis, multiple small, superficial ulcers may be seen, but the appearance may also be indistinguishable from that of ulcerative colitis.Pseudomembranous colitis is identified by the presence of numerous raised yellow plaques covering an inflamed mucosa. Occasionally a carcinoma or large villous adenoma may be detected by sigmoidoscopy. Sigmoidoscopy also provides an opportune time to obtain samples of stool and exudate for culture and microscopic examination. Chapter 46 describes the patient's experience with sigmoidoscopy.
Radiographic Studies
Radiography is of limited use in the evaluation of acute diarrhea and may be confusing. In patients with suspected inflammatory bowel disease or ischemic colitis, plain films of the abdomen may demonstrate an irregular appearance of the bowel wall secondary to mucosal edema, often described as “thumbprinting.” In the gravely ill patient with fulminant colitis, the radiograph may confirm the presence of toxic megacolon. In most cases of acute diarrhea, a plain film is not needed. Barium studies during the acute phase of the diarrhea are likewise not needed and in certain conditions may even be hazardous, as in patients with severe colitis or ischemic bowel disease. A small bowel series in acute viral enteritis may be frighteningly abnormal, resembling sprue, and yet may rapidly return to normal after resolution of the acute illness. CT scanning may show thickening of the colon, although this finding is nonspecific.
Evaluation of Chronic Diarrhea
The approach to patients with either acute diarrhea that lasts longer than 3 to 4 days or chronic diarrhea (lasting
P.696
longer than 4 weeks) is similar. Although the clinician may be reassured by knowing that the majority of such patients do not have a serious progressive or disabling disease, the patient requires a specific diagnosis and effective therapy. The differential diagnosis is so varied, and the available tests are so numerous, that the diagnostic workup of chronic diarrhea poses a difficult problem.
History
The history may be helpful in identifying patients with functional diarrhea (e.g., irritable bowel syndrome; see Chapter 44). It is important to determine whether the pathogenic mechanism is osmotic, secretory, motor, or exudative (see earlier discussion). Patients should be asked to describe the characteristics of the stools, whether they are too frequent, loose, excessive in volume, oily, or associated with fever, blood, mucus, tenesmus, bloating, or excessive flatus.
In patients with irritable bowel syndrome (see Chapter 44) or functional diarrhea, the history of diarrhea often dates back many months or years, although occasionally it can be traced to a specific acute diarrheal illness. Despite the chronicity, weight loss, anemia, or hypoalbuminemia do not occur. The patient typically complains of several watery, at times explosive bowel movements early in the morning, and then no subsequent movements the rest of the day. Nocturnal bowel movements are rare. The total stool output is usually small—often less than 200 g per day and rarely, if ever, more than 500 g per day. Mucus is often present. There is no blood in the stool unless secondary conditions (e.g., anal fissures) have developed. Abdominal pain or discomfort that is usually related to bowel movements is a feature of IBS. The condition often waxes and wanes in severity, and emotional stress often exacerbates the symptoms.
The most common cause of chronic secretory diarrhea is laxative use. It should be suspected in apparently healthy patients with large-volume diarrhea, especially if they have melanosis coli (spotty or diffuse brownish mucosal pigmentation) on endoscopic examination. Although such patients may have emotional problems (see Chapter 21), often laxative abuse results from a misconception about the frequency of normal bowel movements and an attempt to adhere to that standard.
In patients with organic disease, the history may indicate the part of the intestinal tract that is involved. The passage of a large volume of frothy, malodorous stool without blood suggests small-bowel diarrhea, often secondary to malabsorption. The frequent passage of small volumes of poorly formed, bloody stools suggests inflammatory, exudative disorders of the colon such as ulcerative colitis. The presence of recognizable fat droplets (oil) suggests malabsorption, often secondary to pancreatic insufficiency. (Floating stools and undigested food in the stools are not helpful observations because they may be seen in both organic and functional diarrheal states.) The association of diarrhea after ingestion of certain dietary products (milk, fruit, hyperosmolar solutions, sorbitol-containing gum or candy) may not be recognized by the patient unless he or she is specifically asked. A detailed drug history is also important, because many drugs can cause diarrhea (Table 45.4; see later discussion).
Other symptoms may help the clinician to arrive at a specific diagnosis. Arthritis and arthralgias suggest the presence of one of several uncommon bowel diseases, such as inflammatory bowel disease or Whipple disease; conversely, diarrhea may be an important feature of reactive arthritis (see Chapter 78). Weight loss, in the absence of anorexia, should suggest malabsorption, hyperthyroidism, or a malignant tumor. Abdominal pain may reflect the irritable bowel syndrome (see Chapter 44), in which case it is generally in the left lower quadrant or the suprapubic region; or a disease of the small bowel (e.g., Crohn disease), in which case it is often periumbilical or in the right lower quadrant; or a gastrinoma (Zollinger–Ellison syndrome), in which case peptic ulcers are responsible for the associated upper abdominal pain.
A history of previous GI or biliary surgery, intravenous drug abuse, or anal intercourse, or a family history of inflammatory bowel disease or celiac sprue, may sometimes provide clues to the etiology of diarrhea. Some patients may report fecal incontinence as diarrhea; incontinence may not be reported voluntarily and should be specifically inquired about.
Physical Examination
The physical examination may reveal additional information about the cause of the diarrhea. Patients with malabsorption may have evidence of weight loss, peripheral neuropathy (secondary to vitamin B deficiency), and carpopedal spasm (secondary to hypocalcemia). Erythema nodosum and pyoderma gangrenosum are seen in some cases of inflammatory bowel disease. Hyperpigmentation is a feature of both Whipple disease and Addison disease. Diabetic diarrhea is often associated with other evidence of autonomic dysfunction, such as postural hypotension. Nondeforming arthritis is a feature of Whipple disease and of inflammatory bowel disease. Hepatosplenomegaly and lymphadenopathy suggest lymphoma or Whipple disease. The abdominal examination may reveal an arterial bruit or an aortic aneurysm, which suggests ischemic bowel disease. A rectal examination may disclose perianal disease (e.g., abscesses or fistulas secondary to Crohn disease), a rectal tumor, or a fecal impaction.
Endoscopy
Flexible sigmoidoscopy, done primarily if disease is thought to be limited to the rectosigmoid region (e.g.,
P.697
ulcerative proctitis), may be performed during the initial visit without a prior cleansing enema. Examination and biopsy of the rectosigmoid mucosa may suggest specific etiologies (ulcerative colitis, Crohn disease, amebiasis, pseudomembranous colitis, Whipple disease, or amyloidosis). Mucosal biopsy should be performed because collagenous colitis or microscopic colitis, for example, may be present in patients with chronic watery diarrhea. At the time of sigmoidoscopy, stool specimens should be obtained for microscopic evaluation and culture. Gonococcal proctitis appears similar to ulcerative proctitis (discussed later) and requires direct plating on a warm special culture medium (Thayer–Martin) with prompt incubation. The stool should be examined for leukocytes or lactoferrin, blood, fat (Sudan stain), and parasites (see Chapter 35) and routine stool cultures should be taken as well.
Colonoscopy requires bowel preparation, but may be preferable, especially in patients 40 years of age or older, an age when the incidence of colon cancer begins to rise. In approximately one third of patients with chronic diarrhea, a histologic diagnosis is made after colonoscopy and biopsy (18). In younger patients, colonoscopy is less likely to be useful unless a specific disorder is suspected. The presence of blood in the stool, at any age, is a clear indication for a colonoscopy, regardless of the presence of hemorrhoids or fissures. Neoplastic and inflammatory conditions may be diagnosed in this way, and ulcerative colitis and Crohn disease of the colon can be differentiated in most cases.
After this initial evaluation, the cause of the chronic diarrhea in most patients is either evident or strongly suspected. It is usually possible to distinguish functional from organic diarrhea, to detect evidence of inflammatory or infiltrative disease, to suspect the presence of malabsorption, to characterize the diarrhea as small bowel or large bowel, and even to suggest the underlying pathophysiology. Further evaluation is dictated by the results of this initial workup.
Laboratory Studies
Laboratory studies should be selected to support the clinical impression, but they rarely establish or exclude a specific diagnosis. In general, the possibility of blood loss (stool Hemoccult and blood test for hemoglobin and hematocrit), malnutrition (serum albumin concentration), and fluid-electrolyte imbalance are assessed. Patients with chronic diarrhea should be tested for serum tissue transglutaminase antibody or anti-endomysial antibody to exclude celiac sprue.
Radiographic Studies
A plain film or CT scan of the abdomen may reveal pancreatic calcifications (indicative of chronic pancreatitis), a dilated small bowel, or an abnormal bowel contour (as in inflammatory bowel disease or lymphoma). A CT scan of the abdomen may show thickening of the wall of the colon or terminal ileum in patients with inflammatory disease, but this finding is not specific. Hypoalbuminemia may also cause similar findings.
A small bowel series is helpful in distinguishing mucosal disease (celiac sprue), inflammatory conditions (Crohn disease), and infiltrative processes (Whipple disease or amyloidosis). In addition, a small bowel series is indicated in postsurgical patients to clarify the anatomy (e.g., blind loop, fistula) and to detect localized areas of dilation and stasis.
Other Studies
A quantitative 72-hour stool collection, although regarded as unpleasant by both the patient and laboratory personnel, may be very informative. The test can be performed in the ambulatory setting by having the patient collect all stool in a preweighed container supplied by the clinical laboratory. A log of daily food intake and bowel activity is kept, and all nonessential medications are avoided. The normal stool weight is less than 200 g/day (or volume less than 200 mL/day). Most patients with IBS or other functional forms of chronic diarrhea have stool weights and volumes within this range. A stool volume of more than 1,000 mL/day suggests a secretory diarrhea or malabsorption.
Fecal fat should also be measured during this collection. Normally less than 9% of ingested fat is secreted in the stool per day (4 to 7 g on an average American diet containing 60 to 100 g of fat). Fecal fat excretion of more than 14 g/day is considered steatorrhea. It is important to remember that severe diarrhea by itself can cause fat malabsorption with fecal fat values between 7 and 14 g/day. Although it is less reliable, a qualitative test for stool fat by Sudan stain can be useful as a rapid screening test for steatorrhea. The presence of steatorrhea dictates a different approach to the remainder of the workup (see later discussion). In the absence of excessive fat excretion or evidence of an exudative process (e.g., inflammatory bowel disease), high-volume diarrhea suggests a secretory process.
The osmotic gap of the stool can help differentiate secretory from osmotic diarrhea (19). The osmotic gap can be calculated as 290 - 2 ([Na+] + [K+]). The concentration of Na+ and K+ in homogenized stool water can be directly measured. The sum of the concentrations of these ions is multiplied by 2 to account for unmeasured anions. The final product is then subtracted from 290 because studies have shown that in the distal intestine, the stool osmolality equilibrates with the osmolality of plasma. In pure osmotic diarrhea, the stool osmotic gap is greater than 125. In pure secretory diarrhea, the osmotic gap is less than 50. In mixed osmotic and secretory diarrhea, the osmotic gap
P.698
may be between 50 and 125. A fecal pH of less than 5.3 is suggestive of carbohydrate malabsorption. However, fecal pH is higher if there is associated malabsorption of fat and protein.
In patients with suspected secretory diarrhea, further evaluation usually requires hospitalization and consultation with a gastroenterologist. In patients with secretory diarrhea, having the patient ingest nothing by mouth for 48 hours does not alter the volume of diarrhea, whereas in osmotic diarrhea the stool volume significantly decreases. Urine and stool specimens should be sent for spectrophotometry or thin-layer chromatography to detect anthraquinones, bisacodyl, or phenolphthalein in patients with suspected laxative abuse. If laxative abuse has been excluded, measuring serum gastrin (Zollinger–Ellison syndrome) or serum vasoactive intestinal peptide may help to identify a tumor as the cause of secretory diarrhea.
Evaluation of Malabsorption
If the quantitative or qualitative fecal fat determination indicates steatorrhea, the evaluation of the diarrhea should focus on the cause of malabsorption or maldigestion because of small-intestinal disease, pancreatic disease, hepatobiliary disease, or gastric disease.
A CT scan of the abdomen may show calcifications within the pancreas or evidence of diseases of the small intestine. A small bowel series may also show dilated loops of bowel or other signs of mucosal disease. If radiographs suggest mucosal disease, a small-bowel biopsyshould be performed next. Endoscopic biopsy of the small intestine is often useful in the detection of various disorders that can cause malabsorption or diarrhea, or both. Disorders that may be diagnosed by small-bowel biopsy include sprue, Whipple disease, intestinal lymphoma, amyloidosis, lymphangiectasia, and eosinophilic gastroenteritis. Giardiasis can also be diagnosed by examination of the intestinal mucosa and the intestinal mucus or fluid.
Pancreatic insufficiency can be confirmed by measuring pancreatic secretions or by measuring intraluminal contents after a test meal or administration of secretin or cholecystokinin. These cumbersome tests involve intubation of the duodenum, with collection of intraluminal contents. Other methods do not require intubation but simply entail urine collection; for example, the bentiromide test uses a nonabsorbable synthetic peptide (500 mg of bentiromide) that is cleaved by the pancreatic enzyme chymotrypsin to release para-aminobenzoic acid, which is then absorbed and measured when it is excreted in the urine. Urinary excretion of less than 85 mg of para-aminobenzoic acid in 6 hours is considered a positive test. This test has a sensitivity of 85% and specificity of 95% (20). Measurement of fecal concentrations of the pancreatic enzymes chymotrypsin and elastase also has a sensitivity of 85% for severe pancreatic insufficiency, but this is relatively insensitive for mild insufficiency. The test is easier to perform than the bentiromide test (21). Alternatively if a presumptive diagnosis of pancreatic insufficiency is made, the patient may be treated empirically with pancreatic enzymes (e.g., Viokase, three to six tablets with meals, or Pancrease, two to three tablets with meals). If the diagnosis is correct, symptoms may improve although possibly not be entirely eliminated. Diagnostic and therapeutic decisions should be made in consultation with a gastroenterologist.
Disorders of the terminal ileum (Crohn disease, ileal resection) may also lead to diarrhea and malabsorption. The bile salt breath testmeasures bile salt (acid) absorption, which is also abnormal in disorders of the terminal ileum, the site of bile salt absorption. The patient is given orally a radiolabeled (14C) bile salt. In the presence of disease in the terminal ileum, the bile salt is not absorbed and reaches the colon, where bacteria deconjugate it and release 14CO2, which is excreted in the breath. Therefore, in ileal disease, the level of 14CO2 in the patient's expired air is abnormally high. The same abnormality can be seen when there is bacterial overgrowth in the small bowel, so that the bile acid is deconjugated and metabolized there instead of in the colon. Bile acid malabsorption caused by bacterial overgrowth is corrected when the patient is given antibiotics.
Specific Causes of Chronic Diarrhea
Lactose Intolerance
Lactose is by far the most commonly malabsorbed carbohydrate. Lactose intolerance results from a deficiency of the enzyme lactase in the brush border of the intestinal mucosa, which causes maldigestion and therefore malabsorption of lactose. The unabsorbed carbohydrate exerts an osmotic effect that draws water into the intestinal lumen. In the colon, the lactose is metabolized by bacteria to organic acid, CO2, and hydrogen. Diarrhea results from both an osmotic effect and from an irritant effect on the colonic mucosa. Patients may experience gaseousness, bloating, and abdominal cramps even in the absence of diarrhea.
Lactose intolerance may be present either as an inherited condition, “lactase nonpersistence” after childhood, or one that is acquired because of damage to the intestinal epithelium (e.g., infectious enteritis, giardiasis, or sprue). Even in patients with a genetic disorder, the onset of the disease is unpredictable and may not occur until adulthood. Isolated lactase deficiency is most common in African Americans (70% to 100% prevalence) and in Asians (more than 90%) but may also be found in 12% of the Caucasian population in the United States (22). Lactose intolerance because of other causes may be reversible if the underlying disease is successfully treated.
P.699
The severity of the clinical symptoms is highly variable. In some patients even small amounts of lactose produce severe symptoms, but in others large quantities may be consumed with no or only minimal symptoms. The condition is more pronounced in certain clinical settings: when superimposed on another diarrheal disorder, most commonly IBS; after gastric surgery, which permits rapid delivery of lactose to the small bowel; and when a patient consumes a large amount of milk.
The diagnosis of lactose intolerance is suggested by the history and the response to a lactose-free diet. However, almost one third of patients with symptomatic lactose intolerance may not have made the correlation between dietary intake and the resulting symptoms, because a wide variety of foods contain lactose, ranging from bread to instant coffee.
Ordinarily, a 3-week trial of a diet that is free of milk and milk products is a satisfactory therapeutic trial to test the diagnosis of lactose intolerance. Other tests are indicated only in equivocal cases or when the patient's nutritional status would be compromised by eliminating milk products unnecessarily.
Specific tests for the diagnosis of lactose intolerance include the lactose tolerance test and the hydrogen breath test. The lactose tolerance test measures changes in the concentration of serum glucose at 1 and 2 hours after ingestion of 50 g of lactose. A rise in glucose of 20 mg per 100 mL above the fasting level is normal. The test has approximately a 30% false-positive rate, and its validity depends on a variety of factors besides simply the presence of the lactase enzyme. The hydrogen breath test is easier to perform and is more accurate. Unabsorbed lactose is fermented by colonic bacteria, and the resultant hydrogen is absorbed and expired in the breath. In normal subjects, after a 25-g lactose load, there is only a trace amount of hydrogen in the expired air, whereas in lactase-deficient patients substantial levels are recorded. An increase in breath hydrogen of 20 parts per million (ppm) above baseline within 4 hours is considered positive.
Fecal Impaction
Although it is the result of chronic constipation, fecal impaction (23) can cause diarrhea. Elderly sedentary people, many of whom are bedridden, are at highest risk for developing fecal impaction. Impaction usually occurs in the rectum or the rectosigmoid region but occasionally may extend proximally into the colon (rarely, even to the cecum). The leaking of colonic fluid around the impaction results in the passage of frequent, small-volume, watery bowel movements. Other symptoms are usually nonspecific: a sense of fullness in the rectum, vague lower abdominal pain, nausea, and headache. On physical examination, firm stool is palpable in the left lower quadrant of the abdomen, which is best examined bimanually (a finger of one hand in the rectum and the other hand on the abdomen). The impaction is best removed manually if it is low enough. Repeated enemas (e.g., a Fleet enema) may be helpful once some of the hard stool is removed. Complications of impaction include recurrent urinary tract infection (because of compression of the ureters, more common in women), urinary incontinence, intestinal obstruction, perforation of the colon, and local ulceration (stercoral ulcer). Prevention of fecal impaction is an important goal in the sedentary elderly population and is best done by increasing dietary fiber, adding a bulk laxative if necessary, and urging that at least 2 quarts of liquid be ingested each day, in addition to that consumed in the course of meals. Patients should also be sure to respond promptly to the urge to defecate.
Ulcerative Colitis
Ulcerative colitis is a chronic inflammatory disorder of the colonic and rectal mucosa; its cause is unknown. It is recommended that patients with this condition be monitored by their primary care provider in consultation with a gastroenterologist. The disorder may affect patients of any age, with a peak incidence in the third decade and a second peak in the seventh decade.
The clinical picture of ulcerative colitis is highly variable (Table 45.5). The disorder may be limited to the rectum (ulcerative proctitis) or may involve the entire colon. Symptoms may range from occasional rectal bleeding, even without diarrhea, to profuse purulent and bloody diarrhea. The severity of the initial presentation and the extent of disease at the time of the initial attack have been shown to be useful predictors of the eventual course of the disease (24). Most patients (approximately 60%) have mild disease, that is, fewer than four bowel movements a day without fever, weight loss, or hypoalbuminemia. The vast majority of these patients have colitis limited to the rectosigmoid region or descending colon. Approximately 10%
P.700
to 15% of patients with ulcerative colitis develop severe pancolitis with accompanying deterioration in their general health. Another 25% have moderate disease with more troublesome diarrhea, often containing blood, accompanied by crampy lower abdominal pain. Patients with moderate or severe disease may also have systemic symptoms of fever, fatigue, and weight loss. The clinical course is characterized by periodic exacerbations that generally respond well to adjustments in medical therapy. The smallest group of patients with ulcerative colitis consists of those with severe disease. This group includes the 1% of patients who present initially with fulminant colitis. In patients with severe colitis, symptoms may suddenly worsen, with profuse diarrhea, rectal bleeding, and high fevers. Plain films of the abdomen may demonstrate a dilated bowel (toxic megacolon). Mortality is high in this group of patients.
|
TABLE 45.5 Severity of Ulcerative Colitis |
|||||||||||||||
|
|||||||||||||||
Because there is no specific test for, or histopathology of, ulcerative colitis, the diagnosis depends on the constellation of symptoms, the appropriate endoscopic and histologic appearance of the colonic mucosa, the exclusion of other inflammatory conditions, and the natural history of the disorder. Patients with acute presentations, depending on the circumstances, must be differentiated from patients with infectious diarrheas caused by Salmonella, Shigella, or Campylobacter (see Chapter 35); amebiasis; Crohn disease (Table 45.6); and ischemic colitis. (Patients with ischemic colitis present with acute abdominal pain and the passage of bloody stool; this is a disease of middle-aged or older people who usually have evidence of generalized atherosclerosis.)
Treatment
Medical therapy for ulcerative colitis is determined by the severity of the attack and the extent of disease, and it should be individualized. (The use of antidiarrheal drugs in patients with chronic diarrhea is discussed later in this chapter.) Mild attacks may respond to sulfasalazine (Azulfidine, 4 to 6 grams per day) or to 5-aminosalicylic acid (ASA) agents (mesalamine, 2 to 4.8 g/day, or olsalazine, 1.5 to 3.0 grams per day), whereas moderate or severe attacks require treatment with oral or intravenous corticosteroids (equivalent to prednisone 40 to 60 mg/day). Corticosteroids are very useful for acute exacerbations, especially in patients with severe disease or toxic symptoms, but they are not helpful in preventing relapses. Conversely, sulfasalazine is of limited value in the treatment of acute attacks but has been shown to reduce the frequency of exacerbations and may allow reduction of the dosage of corticosteroids. Topical mesalamine or corticosteroids in the form of enemas or suppositories may be sufficient therapy for mild to moderate distal colitis. Patients with severe disease who do not improve after 7 to 10 days of high-dose corticosteroid therapy may require either intravenous cyclosporine, 4 mg/kg/day, or surgery. A minority of patients with ulcerative colitis become steroid dependent and require 6-mercaptopurine (6-MP) or azathioprine for 3 to 4 months before steroids can be successfully tapered, followed by long-term therapy with these agents. For the maintenance of remission in ulcerative colitis, oral mesalamine or sulfasalazine is recommended. Chronic treatment with oral corticosteroids has not been shown to prevent recurrences and is not recommended. Azathioprine or 6-MP, 1.5 to 2.5 mg/kg/day, may be used in patients who are steroid-dependent or refractory to corticosteroids or 5-ASA agents, preferably in consultation with a gastroenterologist.
It has been discovered that nicotine alleviates symptoms in patients with ulcerative colitis (25), but whether it should be recommended treatment remains to be clarified.
Surgery in ulcerative colitis is curative. Patients should be counseled early in their course about the role of surgery in the treatment of this disorder. Patients should be
P.701
informed about the indications for surgery and the types of operations that are available. Early attention to this issue enables the patient to accept an operation more readily if it is needed. Surgery for ulcerative colitis involves a proctocolectomy with an ileostomy to which a stomal appliance is attached to ensure continence. Construction of a continent ileostomy (Kock pouch) avoids the need for a stomal appliance, but the procedure is technically difficult and often requires revision. Another alternative is construction of an internal pouch from a loop of small bowel anastomosed to the anus (Park procedure). These last two procedures are most successful in motivated young patients who are undergoing elective colectomy.
|
TABLE 45.6 Differentiating Features of Crohn Disease and Ulcerative Colitis |
|||||||||||||||||||||
|
Cancer of the Colon
The risk of colorectal cancer is increased 5 to 10 times in patients with ulcerative colitis (26). The major risk factors are duration of disease (risk increases significantly after 8 to 10 years of disease), extent of colonic involvement (pancolitis carries the highest risk, whereas the risk in patients with ulcerative proctitis is similar to that of the general population), and age at onset of disease (patients younger than 25 years of age at the time of onset have the highest risk, independent of the extent of disease). The cancer may be found anywhere in the colon, although most commonly it is within the rectum or rectosigmoid. It may be multicentric and does not arise in adenomatous polyps. It is important that the patient know about the risk of cancer because it may influence a decision to undergo colectomy. After they have had the disease for 8 to 10 years, high-risk patients should have yearly evaluations of the colon by colonoscopy. Because dysplastic changes of the colonic mucosa have been shown to correlate closely with the development of cancer elsewhere in the colon, serial colonic and rectal biopsies should be obtained during this yearly colonoscopy in high-risk patients.
Crohn Disease
Crohn disease, or regional enteritis, is a chronic inflammatory condition of unknown cause involving all layers of the intestine, as opposed to involvement of only the mucosa (as in ulcerative colitis; see Table 45.6). The condition most commonly affects the terminal ileum, but any area from the esophagus to the anus can be involved. The onset of the disease most commonly occurs in adolescence or young adulthood. The incidence of Crohn disease has been rising in recent years, particularly Crohn disease of the colon.
Presentation
Crohn disease may be localized initially to the small bowel, involve small bowel and colon, or be confined to the colon only. The inflammatory process often remains confined to the initial site of involvement unless surgery is performed. Recurrence is the rule after surgery, and the condition may then involve additional segments of bowel. Spontaneous progression of the disease tends to occur in a proximal (orad) direction. As the inflammatory process persists, the bowel wall becomes thickened and stenotic, leading to bowel obstruction. Fistula formation is characteristic and may involve any contiguous structure. As a result, abscess formation and infection can complicate the clinical course. Diarrhea, abdominal pain, and weight loss are the most common symptoms. Unlike ulcerative colitis, rectal bleeding is not a prominent feature unless the colon is the major site of involvement.
The differential diagnosis includes disorders of both the small and the large bowel. Occasionally the patient presents with fever and acute right lower quadrant pain resembling acute appendicitis. When there is involvement of the terminal ileum, Crohn disease must be distinguished from lymphoma and tuberculosis. Colonic involvement may suggest ulcerative colitis, ischemic colitis, or carcinoma. Involvement of the distal small bowel and the right colon, the presence of characteristic skip areas, stricturing of the bowel, perianal disease, and fistula formation are helpful diagnostic features that suggest Crohn disease.
Treatment
Medical therapy for Crohn disease is similar to that for ulcerative colitis, depending heavily on 5-ASA preparations or sulfasalazine and immune suppression. For mild to moderate active Crohn disease, 5-ASA agents or sulfasalazine (see earlier discussion for dosing schedule) should be used as first-line agents. Metronidazole (1 to 2 g/day) or ciprofloxacin (1 g/day) is as effective as mesalamine or sulfasalazine and may be used in patients who are allergic or intolerant to these latter two drugs or in combination with the first-line agents. Long-term metronidazole therapy is, however, associated with development of peripheral neuropathy in some individuals. Patients with severe disease usually require an initial course of prednisone, 40 to 60 mg/day until the symptoms resolve, followed by a slow taper. A monoclonal antibody to tumor necrosis factor alpha (Infliximab), given in a single intravenous dose of 5 mg/kg, may be used as adjunctive therapy in individuals not responding to corticosteroids. Patients with severe or fulminant disease require aggressive treatment, which often includes hospitalization, parenteral hyperalimentation, and either intravenous cyclosporine or oral tacrolimus. Surgery may become necessary if response to medical therapy is not optimal. As in ulcerative colitis, corticosteroids are not effective in preventing relapses and should not be used for maintenance of remission. After the control of acute disease, mesalamine, sulfasalazine, metronidazole, ciprofloxacin, 6-MP, or azathioprine may be used for maintenance therapy. Infliximab is effective in closing fistulas in one third or more of the patients (27). However, some patients require periodic infusions every 8 to 12 weeks to prevent recurrence. Although they
P.702
are not as effective, antibiotics, 6-MP, and azathioprine may also be useful in the treatment of fistulizing Crohn disease.
Because of the complexity of the disease, management should be accomplished by, or in consultation with, a gastroenterologist.
Surgery for resection of fibrotic obstructing lesions, drainage of abscesses, and resection of complicated fistulas is sometimes necessary. Occasionally the disease is refractory to medical therapy and the diseased bowel must be resected. It must be recognized that surgery is not intended to be curative, so removal of normal bowel to achieve wide, disease-free margins is not indicated. In the rare patient whose disease is extensive and unresponsive to medical and surgical intervention or in whom a short-bowel syndrome has developed secondary to the disease and to repeated surgery, long-term home parenteral hyperalimentation can be beneficial to provide good nutritional support and ameliorate symptoms.
Course
Despite its chronicity and tendency for recurrence, Crohn disease takes a highly variable course. Prolonged asymptomatic periods occur, even after years of disease activity and multiple operations. There is a poor correlation between clinical severity and the radiographic appearance of the disease, so followup radiographs are not indicated unless there is a suspicion of a new development in the disease (e.g., a fistula). The risk of colorectal carcinoma is elevated significantly only in Crohn disease involving the colon. Mortality from the disease is low, but morbidity is high. Because the natural history is so variable, the clinician should approach the patient with Crohn disease in a positive and hopeful fashion, yet be aware of the potential for significant morbidity. All patients should be monitored in close consultation with a gastroenterologist.
Drug-Induced Diarrhea
A variety of commonly used medications can cause diarrhea (Table 45.4). The diarrhea may be a direct result of the pharmacologic activity of the drug (e.g., magnesium-containing antacids, colchicine), or the mechanism for the induction of diarrhea may be unknown (e.g., hydralazine, propranolol). The diarrhea may also signify drug toxicity (e.g., digitalis). Certain drugs have repeatedly been associated with diarrhea (antibiotics, antacids, quinidine, digitalis, and alcohol), whereas in other cases (hydralazine, propranolol), the relationship is rare and not well defined.
Diarrhea Associated with Antibiotics
Diarrhea associated with antibiotics may range from a mild increase in the frequency and volume of stools to a toxic, life-threatening condition. Diarrhea may develop during the course of antibiotic therapy, or after parenteral or oral use, but it may also occur up to months after discontinuation of the drugs. The antibiotics most commonly associated with diarrhea are ampicillin, tetracycline, clindamycin, and the cephalosporins.
In the more severe forms of antibiotic-associated diarrhea, the diarrhea is bloody and is accompanied by abdominal cramps and fever. Endoscopy may reveal pseudomembranes, which appear as raised yellowish plaques on edematous, friable mucosa. Histologically, these pseudomembranes are collections of fibrin, mucin, and leukocytes. Proliferation of C. difficile and the elaboration of its toxin cause pseudomembranous colitis. This organism accounts for the vast majority of cases of pseudomembranous colitis and for 20% to 30% of antibiotic-associated diarrhea in general. The toxin elaborated by C. difficile can be assayed in stool. The organism can also be cultured, but it is difficult to grow. Because culture is less sensitive than the toxin titer and does not correlate as well with symptoms, it is not recommended.
Therapy involves discontinuation of the antibiotics and, in cases of pseudomembranous colitis, administration of metronidazole (Flagyl), 500 mg four times a day for 7 days (see Chapter 35). This antimicrobial agent is effective against clostridial organisms, and the response is fairly rapid. Relapses after discontinuation of metronidazole are not uncommon. In such cases, a second course of metronidazole should be administered. Vancomycin may be prescribed (125 to 500 mg orally four times a day for 7 days) for cases repeatedly resistant or relapsing after metronidazole. Oral cholestyramine has also been used effectively to bind the toxin. Antidiarrheal medications are contraindicated because they may actually prolong the duration of the disease. It is probably unnecessary to treat patients who are found to have C. difficiletoxin–positive stools, as is common in nursing homes, unless there are accompanying symptoms or an outbreak is in progress. Symptomatic patients can be treated with bismuth subsalicylate (Pepto-Bismol), 30 mL or 2 tablets every 4 hours, while results of the assay for the toxin are awaited.
Postsurgical Diarrhea
A variety of surgical procedures may result in diarrhea. Predictably, extensive small-bowel resections (e.g., for mesenteric vascular occlusions) result in severe diarrhea and steatorrhea (short-bowel syndrome). Management of such cases requires careful attention to nutritional factors and use of narcotics to control the diarrhea. Long-term home hyperalimentation has allowed patients to overcome the severe malabsorption that would accompany massive small-bowel resection.
Resection of the ileum is less well tolerated than resection of the jejunum, because the ileum serves as the only site for absorption of bile acids. When the ileal resection is limited (less than 100 cm), the total bile acid pool remains sufficient to prevent significant steatorrhea. However, there is still an excessive loading of bile acids into
P.703
the colon, where they stimulate mucosal secretion and result in diarrhea. Therapy for this form of diarrhea is aimed at binding the fecal bile acids with an agent such as cholestyramine. The dosage is 4 g given before meals and at bedtime. When ileal resection is more extensive (more than 100 cm), the total bile acid pool becomes diminished below the critical level needed for proper digestion and absorption of fat, and steatorrhea develops. The use of cholestyramine in this situation further depletes the bile acid pool and worsens the steatorrhea and diarrhea. Therefore, dietary fat should be supplied in the form of medium-chain triglycerides, which do not require bile acids for absorption. Commercial preparations are available (e.g., Portagen), and consultation with a nutritionist as well as a gastroenterologist is recommended.
Diarrhea may also occur after gastric surgery with vagotomy. At times the vagotomy causes diarrhea by altering intestinal motility and, for unclear reasons, by increasing the concentration of fecal bile acids. Therapy with cholestyramine has been successful in this postvagotomy syndrome. Gastric surgery may also unmask latent lactase deficiency or, rarely, latent celiac disease. The blind loop syndrome with resultant bacterial overgrowth, dumping syndrome, inadvertent gastroileal anastomosis, and gastrocolic fistula are all complications that can result in diarrhea in patients who have undergone gastrectomy (see Chapter 43).
Diarrhea may occur after cholecystectomy, in association with an increased concentration of fecal bile acids. Therapy with cholestyramine is effective. Subtotal colectomy, with an ileal-rectal anastomosis (e.g., for multiple polyposis), often results in diarrhea that is usually easily controlled by antidiarrheal medication and diminishes with time. Segmental colonic resection usually does not result in diarrhea because of the large functional reserve of the normal colon.
Symptomatic Antidiarrheal Therapy
Diarrhea is merely a symptom, and therapy, if possible, should be directed at the underlying process. However, a wide variety of agents are available for symptomatic control of diarrhea. The efficacy of these agents is highly variable, and the mechanism of action of many is poorly understood. Symptomatic treatment should be avoided in patients with suspected acute infectious diarrhea because of bacteria likeSalmonella and Shigella, because suppression of bowel movements in these conditions may prolong the diarrhea.
Hydrophilic bulk-forming agents, such as psyllium (Metamucil, Konsyl), have been shown to improve the consistency of ileostomy and colostomy effluent (see earlier discussion). These agents, which paradoxically are also used in treating constipation, are particularly useful in patients with irritable bowel syndrome (Chapter 44).
Another group of antidiarrheal medications consists of those classified as absorbents, on the premise that these agents bind factors within the intestinal lumen that cause diarrhea. Medications of this group include kaolin and pectin (Kaopectate), bismuth salts (Pepto-Bismol), aluminum hydroxide (Amphojel), and cholestyramine (Questran). Most are available over the counter, but their value is not well established. However, Pepto-Bismol has been shown to be effective in controlling the symptoms of traveler's diarrhea (see Chapter 35). Cholestyramine is effective in treating bile acid–induced diarrhea, as occurs in patients after ileal resection, vagotomy, or cholecystectomy. This drug also may bind other compounds, such as digoxin and warfarin, and thereby decrease their absorption.
Opioid derivatives are probably the most effective antidiarrheal medications. Opiate drugs delay the transit of intraluminal contents through the small and large intestines. A central effect is also likely. In patients with extensive small-bowel resection, codeine may be the only effective form of therapy. The synthetic agents diphenoxylate–atropine (Lomotil) and loperamide (Imodium) are also effective and are generally well tolerated. The atropine in Lomotil contributes little to its antidiarrheal effect and may cause significant toxicity. Imodium has the theoretical advantages of a more favorable ratio of GI effects to central nervous system effects and a longer duration of action. An OTC formulation is now available. Imodium has the practical disadvantage of being expensive. The development of megacolon, prolongation of symptoms, and worsening of pseudomembranous colitis have all been linked to the injudicious use of these agents in patients with bacterial diarrhea. The potential risk for abuse is theoretically less for Imodium.
Another group of drugs that is under investigation is classified as antisecretory. Some of these drugs inhibit the synthesis of prostaglandins, which increase intestinal secretion by stimulating adenylate cyclase activity within intestinal cells. (Adenylate cyclase is the enzyme that catalyzes the formation of cyclic adenosine monophosphate [(cAMP)], the concentration of which influences certain transport systems in cell membranes.) Other drugs of this class inhibit adenylate cyclase directly. For example, indomethacin inhibits prostaglandin synthesis and has been shown experimentally to inhibit the effect of enterotoxin. Propranolol, an inhibitor of adenylate cyclase, suppresses bile acid–induced fluid accumulation in intestinal loops. Certain diuretics (e.g., ethacrynic acid) that act on electrolyte transport have also been shown to be effective enterotoxin antagonists. Endorphin-like peptides are also under study as antidiarrheal agents. Investigation into their mechanism of action may lead to the development of new effective forms of therapy against diarrhea. A somatostatin analog (octreotide) is available for the treatment of GI endocrine tumors (e.g., vasoactive intestinal peptide–secreting tumor, gastrinoma, carcinoid). It is not recommended for use in other forms of diarrhea.
Because there are few data to allow an objective comparison of the various antidiarrheal medications, the
P.704
choice of drug must be based on efficacy, safety, and cost. For acute, self-limited illnesses, drugs such as Kaopectate and bismuth salts are often tried by patients, even before a health care provider is consulted. For such patients, diphenoxylate–atropine (Lomotil) or loperamide (Imodium) is highly effective. Patients should be instructed to use medication after a diarrheal movement and not to exceed 8 tablets per day. Loperamide may provide longer diarrhea-free intervals with fewer side effects (28). Oral rehydration therapy (see Chapters 35 and 41) is important, especially with voluminous diarrhea in the frail elderly and in children. Both glucose-based and rice-based electrolyte solutions are available for rehydration. Commercial rehydration solutions designed for athletes (e.g., Gatorade) have inadequate concentrations of electrolytes and therefore are not satisfactory in the treatment of diarrhea.
For patients with chronic diarrhea, the choice of medication is based on the severity and cause of the diarrhea. In patients with the diarrhea-predominant form of the irritable bowel syndrome (Chapter 44), hydrophilic agents may be useful. The dosage should be titrated to the desired bowel habits, with dosages ranging from 1 teaspoon to 2 tablespoons per day mixed in 8 ounces of juice or water per dose. In patients with diarrhea from other causes, diphenoxylate–atropine or loperamide should be tried. These medications can be given in divided doses throughout the day. Diarrhea can also be prevented by taking one or two tablets before engaging in an event associated with diarrhea (e.g., meals, examinations).
In more severe cases of diarrhea, narcotics are necessary. Tincture of opium is convenient because it can easily be titrated (by the drop) to control diarrhea at the lowest possible dosage. A recommended starting dosage is six drops every 4 to 6 hours, to be adjusted by one or two drops per dose depending on the patient's response. Codeine, at a dose of 15 to 30 mg, may also be used with the same dosage schedule.
Specific References*
For annotated General References and resources related to this chapter, visit http://www.hopkinsbayview.org/PAMreferences.