25. 1 Acute pancreatitis

Вид материала

Документы

Содержание Ulcerative colitis Clinical manifestations Medical management Indications for surgical treatment Surgical management

Подобный материал:

• Acute biliary pancreatitis. Treatment tactics, 134.96kb.
• The problem of acute pain and its substantial relief is common for every medical field., 344.88kb.

ULCERATIVE COLITIS

James M. Becker, M.D., and Frank G. Moody, M.D.

Ulcerative colitis, a diffuse inflammatory disease of the mucosal lining of the colon and rectum, is characterized by bloody diarrhea that exacerbates and abates without apparent cause. It is difficult to realize that a disease so devastating remains without an identified etiology or specific medical therapy. Total removal of the affected organs—the colon and rectum—provides a complete cure, but at a sacrifice, since patients so treated must learn to live with an external abdominal stoma (an ileostomy) for the remainder of their lives. Since the disease has its peak onset in early and middle adulthood, this represents a long time span for most patients. Fortunately, new surgical alternatives have eliminated the need for a permanent ileostomy without sacrificing definitive treatment of the disease.

ETIOLOGY

The etiology of ulcerative colitis remains unknown despite intensive work by many investigators. The examination of bacterial and viral agents continues to be an area of great activity. Whether the infectious agents are more likely to be triggers of disease or perpetuators of disease is of great controversy. To be a trigger, an infectious agent would have to act by initiation or reactivation. Agents could initiate an autoimmune response by altering antigens, affecting molecular immunity, or increasing immune responsiveness. The microbial agent might also trigger the pathologic response by increasing mucosal permeability or stimulating epithelial injury or localized ischemia. The microbial agent could reactivate the inflammatory process directly, by secondary infection, or by the release of toxins. Evidence for microbial agents as triggers in inflammatory bowel disease is only indirect.

Other investigators have suggested that infectious agents may perpetuate the disease. The full clinical expression of ulcerative colitis requires an intact mucosal immune system and depends on normal intestinal flora and their products. Thus, alterations in the disease may result from changes in intestinal flora. In addition, treatment interventions may affect disease activity by altering the flora and therefore the energetic or immunologic environment. As discussed later in this chapter, short-chain fatty acids are effective in treating diversion colitis and are natural products of the intestinal flora. Metronidazole may have an effect on ulcerative colitis by altering the flora. Finally, remissions of inflammatory bowel disease have been anecdotally observed in patients with acquired immunodeficiency syndrome. Despite scattered reports suggesting that Chlamydia, cytomegalovirus, or Yersinia is involved in the pathogenesis of ulcerative colitis, these reports have not been substantiated by further work. Clostridium difficile toxin activity has been associated with relapses of ulcerative colitis but appears to be better correlated with prior antibiotic administration than with disease activity. Ljungh and Wadstrom 60 have isolated Escherichia coli with unique binding characteristics from patients with ulcerative colitis. The immune response to this strain is being investigated, in an effort to examine factors that might lead to chronic infections with a resultant chronic autoimmune inflammatory disease of the colon. A viral cause also appears unlikely, since the disease cannot be transmitted and viral particles have not been identified. Specifically, rotavirus and Norwalk agents cannot be identified serologically as important factors in recurrences. A cytopathic agent with physical and chemical characteristics suggestive of a 16-nm. RNA virus has been identified in patients with ulcerative colitis, but this finding has not been duplicated. Although serum lysozymes are elevated in patients with Crohn's disease, they are normal in those with ulcerative colitis.

Genetic factors may have a role, since most studies have suggested that ulcerative colitis is two to four times more common in Jewish than in non-Jewish white populations and is probably about 50% less frequent in nonwhite than in white populations. Gilat and colleagues, however, in a study of Jews in Tel Aviv, reported a remarkably decreased incidence of ulcerative colitis in that city (3.7 per 100,000 population), compared with the incidences reported from Copenhagen (7.3 per 100,000), Oxford, England (7.3 per 100,000), and Rochester, Minnesota (7.2 per 100,000). In addition, the female to male ratio was only 0.8, as compared with 1.3 for the other studies. A greater frequency (10% to 15%) of ulcerative colitis has been identified in family members of patients with confirmed ulcerative colitis. Some families have reportedly had up to six members affected. The disease occurs with greater frequency in monozygotic twins. Finally, HLA phenotypes AW24 and BW35 are associated with ulcerative colitis in those Israeli Jews of European origin mentioned previously. The AW24 phenotype also occurs with increased frequency in patients with an early onset of chronic ulcerative colitis and moderate to severe disease. The genetic mechanisms involved are not known, although multiple gene alterations are likely. Genetic possibilities in ulcerative colitis include a polygenetic mode of inheritance, a specific form of somatic gene mutation in mesenchymal stem cells, the growth of a forbidden clone of cells whose mutant humoral products attack the colonic mucosa, and a rare additive major gene. The identification of specific biologic markers of chronic ulcerative colitis would greatly facilitate genetic epidemiologic studies and further clarify the nature of the disorder. It is hoped that the study of the genetically modified transgenic rat model and the genetic “knockout” animal model will provide important clues to the genetic nature of ulcerative colitis. Obviously, geographic as well as racial differences influence the occurrence of the disease. The incidence of ulcerative colitis is highest in developed or urban regions of the world and lowest in developing regions, although there are recent signs that the incidence rates of inflammatory bowel disease may be leveling off in the developed countries and starting to increase in the developing world. Ulcerative colitis is being reported with increasing frequency in Japan, India, Thailand, and other countries in Asia.

Psychologic factors have long been thought to have a critical role in exacerbations of the disease. It is now clear that patients with ulcerative colitis have no unusual predisposing factors when compared with matched controls. Moreover, colectomy is usually followed by a marked improvement in pre-existing morbid psychologic states such as depression or social estrangement. Psychosomatic factors most likely only facilitate the colonic mucosal reaction to another as yet unidentified causative agent.

There has been considerable speculation that ulcerative colitis is an autoimmune disease. For example, many patients have circulating antibodies to normal colonic epithelium that cross-react with specific enterobacterial lipopolysaccharide antigens. In addition, lymphocytes may be rendered cytotoxic to colonic epithelium by incubation with serum from patients with ulcerative colitis. These patients have also been found to have alterations of their T- and B-cell lymphocyte activation and homing properties. Whereas total lymphocyte and T-cell lymphocyte counts are normal in patients with ulcerative colitis, thymosine-dependent T-lymphocyte response may be abnormal, suggesting an immune-deficient state. 11 These interesting aberrations have been reviewed by members of Work Group VIII on Progress in Digestive Disease, who point out that these changes may not necessarily contribute to the pathogenesis of the disease but may indeed be a consequence of its activity. 15 In fact, Brandtzaeg and colleagues 8 demonstrated quite clearly that rather than a defect occurring in immunoglobulin activity at the tissue level in the remaining glands of patients with ulcerative colitis, IgA transport is normal, whereas IgG immunocyte response is five times that of control patients. It is possible, therefore, that IgG antibodies have a role in the chronicity of the disease but may not be involved in its onset.

Another area of great interest has been that of cytokines and immunoregulatory molecules involved in the control of the immune response. The production of interferon during inflammation could have a significant role in the differentiation of mature memory and effector cells within the intestine. Specific activities of interleukins that are potentially relevant to inflammatory bowel disease have been identified. Most important of these may be interleukin-1 (IL-1), which activates T and B lymphocytes as well as macrophages and neutrophils. IL-1 stimulates production of eicosanoids, cytokines, growth factors, and destructive enzymes; increases adhesion of neutrophils and monocytes to endothelial cells; induces acute-phase response as well as fever, anorexia, and sleep; and stimulates collagen production and thus fibrosis. IL-1 has been shown to be elevated in ulcerative colitis as well as in experimental models of colitis. The increase in IL-1 levels seems to correlate with severity of disease. Alterations in IL-2, IL-6, IL-8, and interferon-gamma have been identified in tissues from patients with ulcerative colitis. The production of interferon during inflammation could play a significant role in the differentiation of mature memory and effector cells within the intestine. Tumor necrosis factor may also be particularly important in the activation of mesenchymal cells but has not been fully evaluated in ulcerative colitis. Thus, it appears that cytokines are integrally involved in the pathogenesis of inflammatory bowel disease with both immunoregulatory and proinflammatory properties.

Abnormalities in complement function or metabolism have also been reported. The finding of deficient activity of IL-2 in patients with ulcerative colitis suggests abnormal T-cell proliferation and clonal expansion response, leading to a chronic inflammatory reaction. Low levels of IL-2, however, have not correlated with duration, activity, or anatomic location of disease or response to corticosteroid therapy.

Other investigators have explored the role of helper T cells in controlling intestinal immune responses. In conclusion, although abnormalities in immune regulation have been implicated in ulcerative colitis, definitive proof of basic immunologic defects or autoimmune phenomena in this disorder is still lacking.

Several exciting alternative approaches to the pathogenesis of ulcerative colitis have been taken. Roediger suggested that ulcerative colitis represents an energy-deficient disease of the colonic epithelium. Colonocytes from patients with ulcerative colitis demonstrated lower oxidation of butyrate to carbon dioxide and decreased free coenzyme A (CoA). In pigs, it was found that the induction of low colonic mucosal CoA content produced a colitis that resembled human ulcerative colitis. In this model, resolution of the colitis occurred when CoA deficiency was corrected with pantothenic acid. The precise cause of the decreased fatty acid oxidation>
A further clue comes from the work of Harig and associates. Patients with diversion colitis were found to have reduced short-chain fatty acid (SCFA) levels within the bypassed segments. Treatment with intraluminal instillation of an isotonic SCFA solution caused complete endoscopic healing in all patients and recurrence when saline was substituted for the SCFA solution. Luminal SCFAs have also been shown to accelerate healing of surgical anastomoses, increase regional blood flow and oxygen intake, and have contrasting effects on epithelial proliferation in vitro as compared with in vivo. Thus, whereas short-chain fatty acids may have a role in the pathogenesis and treatment of inflammatory bowel disease, this requires further study in patients with ulcerative colitis.

Podolsky and Isselbacher suggested that there might be alterations in colonic mucosal glycoprotein composition in patients with ulcerative colitis. Mucin profiles were performed on mucosa from patients with ulcerative colitis, and a selective decrease in mucin species IV was identified. Normal mucin profiles were found in patients with Crohn's disease, ischemic colitis, infectious colitis, and radiation colitis. Patients with ulcerative colitis were found to have this abnormality even in the absence of active disease. It was thought, therefore, that this alteration might be permissive for injury by an additional factor or factors. It is unclear, however, whether this defect is a cause or an effect of the disease.

A relationship between cigarette smoking and ulcerative colitis has been suggested. Several studies have found that current and former smoking has opposite effects on the risks of this disease. Boyko critically reviewed the available literature on this topic and found that current smokers had a 60% reduction in risk of ulcerative colitis compared with those who never smoked, whereas former smokers had a twofold increase in risk compared with those who never smoked. Plausible biologic mechanisms for these relationships have yet to be defined.

Further experimental and clinical work is necessary to evaluate the etiologic possibilities in ulcerative colitis. Over the last decade, animal models of intestinal inflammation have substantially augmented our understanding of the pathogenesis of ulcerative colitis, particularly in the areas of inflammatory mediators and cytokine regulation, genetic susceptibility, and the influence of ubiquitous luminal bacterial constituents. Inducible models, such as administration of acetic acid, trinitrobenzene sulfonic acid–ethanol, and indomethacin to rats and feeding dextran sodium sulfate to mice, are cheap, easily accomplished, and reproducible, making these models the preferred routes for testing novel pharmaceutical agents. Submucosal injection of the bacterial cell wall polymer peptidoglycan-polysaccharide 64 and intravenous administration of preformed immune complexes after rectal installation of formalin elicit more immunologically and environmentally relevant inflammatory responses than the toxin-induced models, permitting more in-depth dissection of immunoregulatory mechanisms of acute and chronic intestinal inflammation. The cotton-top tamarin monkey is unique in that it exhibits spontaneous colitis with associated adenocarcinoma of the colon.

Unprecedented advances in molecular biology now provide techniques to routinely overexpress or delete selected genes in rodents. In vivo overexpression (transgenic) or deletion (knockout) of genes encoding targeted cytokines, T-cell receptors, HLA molecules, and intracellular messengers by basic scientists outside of the inflammatory bowel disease field have unexpectedly created a whole new class of animal models. Spontaneous intestinal inflammation in these genetically engineered rodents, in addition to colitis that follows a spontaneous genetic mutation in C3H/HeJ mice and restoration of T-lymphocyte subsets in immunocompromised hosts, now permits exciting new approaches to explore mechanisms of chronic, spontaneous gastrointestinal inflammation.

PATHOLOGY

Ulcerative colitis is, for the most part, a disease confined to the mucosal and submucosal layers of the colonic wall, progressing from mucosal edema and lipemia to vascular congestion, superficial ulcers, increased cellular infiltration of the lamina propria, and cyst abscesses beginning in the rectum and advancing proximally to involve the entire colon. In 10% of patients, the terminal ileum may show mild inflammation and dilation, a process that has been called backwash ileitis. On gross inspection, the colonic mucosa demonstrates healed granular superficial ulcers superimposed on a friable and thickened mucosa with increased vascularity. Patients may also demonstrate superficial fissures and small and regular pseudopolyps. This is in contradistinction to the transmural inflammatory changes found in Crohn's disease of the colon, in which all layers may be involved in a granulomatous inflammatory process. The pathologic changes observed in ulcerative colitis, however, are nonspecific and can be seen in shigellosis, amebiasis, and gonorrheal colitis.

In its earliest stage, the typical lesion consists of infiltration of round cells and polymorphonuclear leukocytes into the crypt abscesses Light microscopy reveals poor staining and vacuolization of overlying epithelial cells. There is swelling of mitochondria, widening of intercellular spaces, and broadening of the endoplasmic reticulum observed by transmission electron microscopy. As the lesions progress, there is a coalescence of crypt abscesses and desquamation of overlying cells to form an ulcer. This is associated with undermining of adjacent, relatively normal mucosa, which becomes edematous and assumes a polypoid configuration as it becomes isolated between adjacent ulcers. Collagen and a luxurious growth of granulation tissue occupy the areas of ulceration, which extend down to, but rarely through, the muscularis. In fulminating ulcerative colitis and toxic megacolon, such lesions may penetrate through the full thickness of the bowel wall and lead to perforation into the peritoneal cavity. Fortunately, these forms of the disease are infrequent (15% and 3%, respectively). The pathologic changes described offer a clear explanation of the clinical manifestations of the disease. The denuded, remarkably distorted mucosal lining provides little opportunity for absorption of sodium or water. Each bowel action milks large volumes of blood from the exposed hillocks of granulation tissue. Loss of haustral markings, an early roentgenographic finding, is thought to be due to paralysis of the muscularis mucosa. The foreshortening of the colon and its rigid stovepipe appearance on barium roentgenograms are consequences of repeated injury and the scar that forms with repair of these injuries.

Little is known about why some patients have involvement of only the rectum and others develop changes throughout the colon. Moreover, the factors determining the severity and time course of the disease are poorly understood. Possibly these factors relate to the extent of immunologic disturbance engendered by the initial attack. There is also some evidence that prostaglandins may have a role in acute episodes of the disease. Unfortunately, a positive response to prostaglandin synthetase inhibitors, such as indomethacin, has not yet been reported. More recent evidence suggest that acute episodes of colitis may in fact be associated with prostaglandin deficiencies.

CLINICAL MANIFESTATIONS

The initial presentation of ulcerative colitis can take many forms. Bloody diarrhea is the most common early symptom. Occasionally, extraintestinal manifestations, including arthritis, iritis, hepatic dysfunction, and skin lesions, may be paramount. The disease presents as a chronic, relatively low-grade illness in most patients. In a small number of patients (15%), it has an acute and catastrophic fulminating course. Such patients present with frequent bloody bowel movements (up to 30 per day), high fever, and abdominal pain. The disease therefore has a wide spectrum of clinical manifestations, ranging from a mild diarrheal illness to an overwhelming life-threatening event of short duration that demands immediate medical attention.

Onset of the disease occurs in patients less than 15 years of age in approximately 15% of cases, and presentation in patients over 40 years of age is not uncommon. The incidence of ulcerative colitis is 3.5 to 6.5 per 100,000 population, and the prevalence is 60 per 100,000. A slight female predominance has been reported. 11

Physical findings are directly related to the duration and presentation of the disease. Weight loss and pallor are usually present. In the active phase, the abdomen, in the region of the colon, is usually tender to palpation. There may be signs of an acute abdomen accompanied by fever and decreased bowel sounds. This is especially true during acute attacks or in the fulminating form of the disease. Abdominal distention is unusual, except in patients who have toxic megacolon, in which case the patient is usually febrile and has signs of an acute abdomen. The perianal area may be excoriated from the numerous wipings associated with bowel movements. There may be evidence of perianal inflammation in the form of a fissure, abscess, or fistula in ano, although the last is more common in Crohn's disease. Rectal examination is almost always painful and, in the presence of perianal inflammation, should be done with gentle care. Examination of the integument, tongue, joints, and eyes is important, since the presence of disease in these areas suggests ulcerative colitis as a likely cause of the diarrheal illness.

Proctosigmoidoscopy is a helpful and specific diagnostic aid, since ulcerative colitis involves the distal colon and rectum in 90% to 95% of cases. In fact, the mucosa of both the rectum and the sigmoid colon is usually erythematous and granular and bleeds easily when touched by the endoscope or rubbed with a cotton swab. Normal colonic vascular markings may be absent, or the mucosa may be hyperemic; in the disease-bearing mucosa, superficial (less than 2 mm.) mucosal alterations are seen. The intercolonic haustra are thick and blunted. Cobblestoning and deep linear ulceration, which are common endoscopic findings in Crohn's disease, are unusual in ulcerative colitis. In advanced disease, ulcers may be present, surrounded by hyperplastic areas of granulation tissue and edematous mucosa, which may assume a polypoid appearance (pseudopolyps). Mucosal bridging is also commonly found. In chronic advanced disease, the lumen of the rectosigmoid may be remarkably contracted. The use of flexible sigmoidoscopy has improved diagnostic accuracy and patient acceptance. Colonoscopic examination is of value in determining the extent and activity of the disease. Unless a distinct granuloma is identified, endoscopic biopsies are of little value in differentiating ulcerative colitis from Crohn's colitis.

Although recent studies suggest that previous reports may have overestimated the risk of cancer in the adult population with ulcerative colitis, patients with this disease still appear to be confronted with at least a 10% to 20% likelihood of developing carcinoma within 20 years of the diagnosis of ulcerative colitis. 100 Adenocarcinoma in association with ulcerative colitis is multicentric in 15% of patients. In addition, the cancers tend to be flatter and perhaps more infiltrating. These tumors are more evenly distributed throughout the colon, with approximately 50% being found proximal to the splenic flexure. Carcinoma in association with ulcerative colitis is more difficult to diagnose by history and physical examination, stool guaiac testing, and radiographic studies. The likelihood of carcinoma in patients with ulcerative colitis appears to relate to both the extent of colonic involvement and the duration of disease. Although it is generally accepted that patients with extensive total ulcerative colitis are at increased risk of developing carcinoma, the question of what constitutes extensive colitis is still not fully resolved. In addition, the assessment is variable if judged radiographically or colonoscopically. The evidence that patients with left-sided ulcerative colitis, by any criteria, are at increased risk when compared with the general population—which carries a 4% to 6% likelihood of developing colorectal carcinoma, with three fourths of these cancers occurring on the left side—is far from overwhelming. The likelihood of cancer may be related to duration of activity and age of onset, although this has not been clearly established. Although it was held for some time that the carcinoma associated with ulcerative colitis was more aggressive than that in the general population, recent studies have demonstrated that the natural evolution of the cancer is likely the same in both groups.

Rectal biopsies have also been advocated to assess the presence or absence of dysplasia. Morson and Pang 71 advocate a surveillance program of rectal biopsy to assess the point at which a patient becomes at high risk for colonic cancer. When dysplasia of the rectal mucosa is identified, colectomy has been advocated. Other investigators in this field have found the test less useful, with false-negative results of 20% to 40% and false-positive results of 30% to 40%. Colonoscopy may improve the accuracy of surveillance; however, random biopsies have a very low yield because of the immense sampling problem. Between 20 and 25 equally spaced biopsies are required on a 10-cm. length of colon to reasonably detect a patch of dysplasia 2 cm. in diameter. Moreover, the endoscopic appearances of both dysplasia and carcinoma in ulcerative colitis remain nearly undocumented. The biopsy of target lesions, that is, any lesion that cannot be reasonably accepted as part of the chronic disease state, is recommended. In addition, the end-point of surveillance remains controversial. Many gastroenterologists recommend colectomy only in the presence of high-grade dysplasia, a dysplasia-associated mass lesion, or a frank carcinoma. Unfortunately, the presence of dysplasia, whether low-grade or high-grade, can give rise directly to an invasive carcinoma, and all large centers have had patients under surveillance who developed and died of colorectal carcinoma. Some evidence suggests that even low-grade dysplasia unassociated with severe inflammation, if it is unequivocal, should prompt colectomy. To date, no prospective study has clearly demonstrated that surveillance lowers the mortality from colorectal cancer in association with ulcerative colitis, 14 although a large review from St. Marks suggested an overall 5-year survival of 87% in those patients who underwent surveillance and 55% in those who did not.

A plain abdominal film may reveal a variant of the disease called toxic megacolon, in which there may be free air within the peritoneal cavity from perforation of the colon. A more common sign is a remarkable dilation of the transverse colon.

Barium enema examination, usually with air contrast, can be performed safely in most patients and is extremely helpful in identifying the extent and severity of the disease. Barium roentgenographic signs include loss of haustral markings and irregularities of the colon wall, which represent small ulcerations. As the disease progresses, pseudopolyps become a prominent roentgenographic sign. In advanced disease, the colon assumes the appearance of a rigid contracted tube. The barium roentgenogram, although useful, should be avoided in the presence of toxic megacolon, since it may exacerbate the colitis. When diarrhea is not present, a liquid diet for 3 days prior to examination is recommended. Barium roentgenogram should be omitted when the clinical signs of toxic megacolon are present. Upper gastrointestinal contrast studies are also indicated in most patients to exclude Crohn's disease.

The aforementioned clinical manifestations and simple diagnostic tests usually help identify the presence of ulcerative colitis. It is necessary, however, to obtain stool smears and cultures to exclude colitis due to viruses, Chlamydia, bacterial pathogens, and parasites. Particularly important and difficult to exclude are pseudomembranous colitis, the proctocolitis seen increasingly in homosexual males, and traveler's diarrhea. Cello and Meyer 12 provided a useful schema for distinguishing ulcerative colitis from granulomatous colitis. Note, however, the low frequency of discriminating clinical characteristics, except for associated small bowel disease or skip areas within the colon, when the etiology is Crohn's disease.

The strong association of cancer of the colon with ulcerative colitis bears further emphasis. For example, 40% of patients with total colonic involvement may die of cancer if they survive their disease and the colon is left in place. Three percent of children with ulcerative colitis have cancer of the colon at 10 years; 20% develop cancer during each ensuing decade. With the availability of far more acceptable surgical alternatives to proctocolectomy and ileostomy, it is hoped that patients will obtain definitive treatment for the disease well before they enter the phase of accelerating cancer risk. These data support close medical management for such patients and surgical intervention, on this basis alone, when chronicity is well established.

The extracolonic manifestations of ulcerative colitis can be categorized as the colitis group, the pathophysiologic group, and the miscellaneous group of disorders. The colitis group of extracolonic manifestations generally parallels the activity of the underlying bowel disease, being present and most active when the colitis is active and usually subsiding when the colitis goes into remission induced by medical therapy, by surgical intervention, or spontaneously. It appears most likely that these extracolonic disorders represent antigen-antibody immune complex disorders. Ocular manifestations are common in ulcerative colitis and include conjunctivitis, iritis, and choroiditis. These are closely related to disease activity and respond to steroid therapy. More severe and rare eye diseases, including ulcerative panophthalmitis, are more difficult to treat, even with high-dose steroid suppression. Articular disorders, including peripheral joint disease, arthralgias, swelling, pain, and redness with migratory involvement, usually parallel the intensity of the colitis and respond to medical or surgical treatment. The joints of the lower extremities are most frequently involved. Fortunately, permanent deformity of these joints is very uncommon. A certain percentage of patients go on to develop clear evidence of rheumatoid arthritis even after colectomy. Ankylosing spondylitis and sacroiliitis, in contrast, can cause permanent fixation of the spine and need to be treated aggressively. Bone involvement specific to the axial skeleton is less closely related to the severity of the inflammatory state of the colon and, in fact, may precede frank evidence of ulcerative colitis. Patients with ulcerative colitis frequently experience dermatologic disorders, including erythema nodosum and pyoderma gangrenosum. Although these difficult problems resolve after colectomy in most patients, in others, they may precede the colonic disease or may not become manifest until after proctocolectomy has been performed.

Pathophysiologic disorders are more often seen in Crohn's disease than in ulcerative colitis, since in ulcerative colitis, the normal physiology of the terminal ileum is not disturbed. Liver disease is common in patients with both ulcerative colitis and Crohn's disease. Nonspecific inflammation and fatty metamorphosis manifested by mild increases in the serum transaminase values are common in ulcerative colitis. Pruritus and elevation of the alkaline phosphatase are commonly associated with the pericholangitis that occasionally accompanies ulcerative colitis. The most dreaded complication, sclerosing cholangitis, presents with pruritus, alkaline phosphatase elevation, right upper quadrant pain and tenderness, and jaundice. The diagnosis is most often made by endoscopic retrograde cholangiopancreatography or transhepatic cholangiography. It has been estimated that 50% of patients who present with sclerosing cholangitis already have or will develop frank ulcerative colitis. Controversy surrounds the treatment of this disorder. Whereas some patients respond to colectomy, many others show progression of their disease even after colon resection. Surgical drainage, internal stent placement, antibiotics, and ultimately liver transplantation have all been reported to be of value in the treatment of symptomatic sclerosing cholangitis. Cholangiocarcinomas have also been reported in patients with ulcerative colitis, usually after many years of sclerosing cholangitis.

MEDICAL MANAGEMENT

The outcome of an acute episode of ulcerative colitis relates to the severity of the disease as manifested by systemic symptoms. Duration of the disease and extent of involvement of the colon do not appear to be determinants of survival if ulcerative proctitis is excluded from consideration. Those who present with advanced signs of acute illness require hospitalization and supportive, as well as specific, therapy for associated metabolic and hematologic derangements. Because of the massive fluid and electrolyte loss per rectum, such patients often present with metabolic acidosis, contracted extravascular volume, and prerenal azotemia. The serum potassium level is usually low because of excessive loss in stool and urine. Intravenous administration of balanced salt solutions in amounts sufficient to replace these losses is an initial step in management. Patients with long-standing disease may have lost considerable protein and probably are in a depleted nutritional state. The precise role of specialized nutritional support in ulcerative colitis and, in particular, of total parenteral nutrition is unclear. Despite early enthusiasm, total parenteral nutrition does not appear to have a specialized therapeutic role in this disease. Total parenteral nutrition improves the overall nutritional state of patients with ulcerative colitis and may reverse growth retardation in children, but it certainly does not replace conventional medical treatment or prevent or delay colectomy. In fact, in patients with severe acute colitis, it may be impossible to attain a positive nitrogen balance while the colon is still in place.

Corticosteroids and immunosuppressive agents have both been demonstrated to be effective in the management of ulcerative colitis. Both agents, however, are capable of producing significant side effects. In general, corticosteroids have been more readily accepted by the medical community as therapeutic agents and remain the mainstay of therapy in acute attacks. Between 40 and 60 mg. of prednisone in a single daily dose is effective in inducing remission. Rectal steroids have been shown to be effective in left colon disease or proctitis and may have therapeutic efficacy in universal colitis as well, perhaps because approximately 30% of the steroid given rectally is absorbed into the systemic circulation. In an attempt to avoid systemic effects of steroid enemas, tixocortol pivalate was synthesized by adding a thiol ester group at position on the hydrocortisone molecule. In trials, this agent has been useful for treating patients with left-sided colitis and has resulted in a reduction in systemic steroid side effects. The controversy over intravenous steroids versus intravenous adrenocorticotropic hormone (ACTH) has now been resolved by a randomized trial that revealed a similar response to equipotent doses of either hormone. A recent study suggests that ACTH may be more effective in patients not previously treated with corticosteroids, whereas corticosteroids appear to be preferable for patients already receiving steroid therapy. A steroid-induced remission is not more likely to exacerbate than a spontaneous remission, and an ACTH-induced remission is not more likely to exacerbate than a corticosteroid-induced remission. The usual recommended doses are 300 mg. of hydrocortisone or 40 units of ACTH per day. Occasionally, massive doses of steroids (over 1 gm. per day) are required. The usual response is rapid, and acute signs of inflammation subside within a few days. The optimal duration of intravenous steroid therapy is 5 to 7 days, although this may be extended in patients supported nutritionally with total parenteral nutrition. Proctoscopic examination is useful in following response to therapy. There is still controversy as to whether maintenance steroid thereby reduces recurrence of the disease. Although maintenance steroids may be useful in controlling symptoms of patients with continuing activity, maintenance therapy with low-dose corticosteroids for patients with inactive disease has not been demonstrated to prevent relapse. Patients must be monitored carefully for the long-term adverse sequelae of corticosteroid use, including hypertension, hyperglycemia, cataracts, osteoporosis, and osteomalacia.

Sulfasalazine has enjoyed widespread use in the chronic phases of ulcerative colitis. Its mode of action is unknown. Sulfasalazine may exert this prophylactic effect by inhibiting mucosal prostaglandin synthesis, although not all studies have supported this mechanism. Whatever the mechanism of action, sulfasalazine appears to be associated with fewer exacerbations as assessed by controlled randomized trials. The drug appears to be of lesser value in severe ulcerative colitis. Sulfasalazine is metabolized by bacteria to 5-aminosalicylic acid (5-ASA) and sulfapyridine. Dose-related side effects of sulfasalazine include nausea, vomiting, headache, and abdominal discomfort. Reversible hypospermia and infertility are observed in males. Hypersensitivity effects include fever, skin rash, agranulocytosis, and hemolytic anemia. Studies have indicated that the sulfapyridine produced by bacterial degradation of sulfasalazine is responsible for the majority of the side effects, whereas the 5-ASA component appears to be the effective moiety of the drug. Five-aminosalicylic acid is now available in this country for clinical use. In all studies to date, these compounds have been shown to be as efficacious as sulfasalazine in treating acute ulcerative colitis as well as in preventing relapse.

A third approach has been the use of immunosuppressive agents. Rosenberg and colleagues concluded, based on a well-controlled study, that azathioprine allows reduction of the use of steroids in chronic cases but does not, in itself, control exacerbation of the disease. In a more recent controlled trial, however, Kirk and Lennard-Jones 54 demonstrated that clinical improvement may occur in about 25% of patients treated with a dose of azathioprine at 2 to 2.5 mg. per kg. Uncontrolled trials have demonstrated a favorable response to 6-mercaptopurine (6-MP) in 64% to 70% of patients with refractory ulcerative colitis. Because these drugs do not produce a clinical response for several months, they have no role in the treatment of acute flares of ulcerative colitis. Cyclosporine (CS), which has a more rapid onset of action, has been advocated for the treatment of severe, refractory acute ulcerative colitis. Both uncontrolled trials and one controlled study suggest that high-dose CS is efficacious for severe ulcerative colitis. There is, however, significant theoretical risk of irreversible CS-associated nephropathy following treatment with high-dose CS. Severe infectious complications may also occur. A trial of 6-MP or CS may be warranted when steroids and sulfasalazine have failed, when the disease is confined to the left side of the colon or rectum, when the patient is compliant, and when there is no absolute indication for immediate surgical therapy. However, before prescribing these immunosuppressive agents, one must be fully familiar with the dosing, monitoring, toxicity, and possible induction of lymphoma or other malignancies associated with these drugs.

Although widely prescribed for both ulcerative colitis and Crohn's disease, metronidazole and other antibiotics have no proven value in the treatment of inflammatory bowel disease.

The major therapeutic problem between acute episodes is control of diarrhea and maintenance of nutrition. Diet therapy is no longer recommended, and patients are encouraged to eat a substantial diet of their choice. Milk products are to be avoided only if they cause problems such as increasing diarrhea or cramps (as they may in about half of patients with ulcerative colitis). The reason for this is not clear but relates to something specific in cow's milk rather than to the lactase deficiency that exists in many patients with ulcerative colitis. Opiates such as codeine or paregoric should be avoided. Nocturnal diarrhea can be controlled by anticholinergics or diphenoxylate with atropine. The synthetic peripheral-acting opioid loperamide may be more effective than diphenoxylate in this situation and avoids the atropine side effect associated with this drug. Stool bulk formers, such as psyllium, are also helpful. Finally, the importance of rest and peace of mind cannot be overemphasized. Patients are advised to remain at rest during episodes of exacerbation.

INDICATIONS FOR SURGICAL TREATMENT

Since total removal of the colon and rectum (proctocolectomy) cures ulcerative colitis, one might reasonably ask why all patients with established chronicity are not so treated. The incidence of surgical intervention appears to be related to the availability of skilled and knowledgeable gastrointestinal surgeons and enlightened physicians. For example, the clinic at Leeds offers surgical care to approximately half of its patient population, whereas in another center the reported operative rate is below 10%. There are several well-identified complications that require urgent operation for survival. 36 These include massive, unrelenting hemorrhage; toxic megacolon with impending or frank perforation; fulminating acute ulcerative colitis that is unresponsive to steroid therapy; obstruction from stricture; and suspicion or demonstration of colonic cancer. Surgical therapy is also recommended in children who fail to mature at an acceptable rate. The largest number of colectomies for ulcerative colitis are performed for less dramatic indications, as the disease enters an intractable chronic phase and becomes both a physical and a social burden to the patient.

Acute perforation occurs infrequently, with the incidence directly related to both the severity of the initial episode and the extent of the disease in the bowel. Although the overall incidence of perforation during a first attack is less than 4%, if it is severe, the incidence rises to 9.7%. If the total colon is involved, the perforation rate is 14.6%, and if the attack is both severe and involves the total colon, it increases to 19.2%.

Obstruction caused by benign stricture formation occurs in 11% of patients, 34% of these occurring in the rectum. They usually follow submucosal fibrosis and occasionally mucosal hyperplasia. Although they do not usually cause acute obstruction, the lesions must be differentiated from carcinoma by biopsy or excision, and particular attention should be given to excluding Crohn's disease. Strictures caused by carcinoma are less common than those caused by benign disease and are more prone to perforate.

Massive hemorrhage secondary to ulcerative colitis is rare, occurring in less than 1% of patients. Prompt surgical intervention is indicated after hemodynamic stabilization. More than 50% of patients with acute colonic bleeding have toxic megacolon, so one should be suspicious of the coexistence of the two complications. Uncontrollable hemorrhage from the entire colorectal mucosa may be the one clear indication for emergency proctocolectomy. If possible, the rectum should be spared for later mucosal proctectomy with ileoanal anastomosis.

Acute toxic megacolon can occur in both ulcerative colitis and Crohn's disease. Its incidence is between 6% and 13% in patients with ulcerative colitis. Patients usually present clinically with the onset of abdominal pain and severe diarrhea (greater than 10 stools per day), followed by abdominal distention and generalized tenderness. Once megacolon and toxicity develop, fever, leukocytosis, tachycardia, pallor, lethargy, and shock ensue. It is important to note that any of these manifestations can be masked by chronic steroid use and the generally poor nutritional condition of the patient. An abdominal radiograph usually shows dilation of the transverse and occasionally the sigmoid colon that is greater than 5 cm. and averages 9.2 cm. Thickening and nodularity of the bowel wall due to mucosal inflammation are also noted. Caprilli and colleagues reported abnormal gaseous distention of the small bowel in association with toxic megacolon; thus, this finding may be a useful predictor of its development in patients with severe colitis.

The morbidity and mortality for acute toxic megacolon remain high. Soyer and Aldrete reported a series of 12 patients in which the incidence of postoperative sepsis was 50%, wound infection 58%, abscess of fistula 33%, and delayed wound healing 25%. Postoperative mortality ranges from 11% to 16%; for the subset of patients with perforation, mortality is 27% to 44%. These data support the use of combined aggressive medical and surgical treatment of this disease.

Initial treatment for toxic megacolon includes intravenous fluid and electrolyte resuscitation, nasogastric suction, broad-spectrum antibiotics to include anaerobic and aerobic gram-negative coverage, and total parenteral nutrition to improve nutritional status. Proctoscopy may be helpful in determining the etiology of the attack, as may culture of the stool. Although the efficacy of steroids is still in question, most patients presenting with toxic megacolon are already on steroid therapy and thus need stress doses of corticosteroids to prevent adrenal crisis. Most clinicians think that steroids help reduce the inflammation and may cool down an acute toxic episode in up to 50% of patients, although long-term remissions are not achieved. Moreover, the short-term use of corticosteroids does not appear to increase surgical morbidity. Long-term use of larger doses, however, does increase the incidence of wound and septic complications. The authors agree with Fazio that, provided the patient is stable, initial medical trial is warranted in order to make the operation elective rather than urgent. If no clear response is obtained within 24 to 48 hours, surgical therapy is warranted. Larger doses of steroids after initial medical failure probably will not benefit the patient and, as noted, may be deleterious. During medical therapy, serial blood counts, serum electrolyte levels, and abdominal roentgenograms should be closely monitored.

In the presence of acute toxic megacolon caused by ulcerative colitis, surgical therapy can be associated with a high operative morbidity and mortality. Block and colleagues noted an overall mortality following emergency operation of 8.7%; 6.1% after total abdominal colectomy, and 14.7% after proctocolectomy. This suggests that more conservative surgical intervention is appropriate in the acute setting. Also, with the recent popularity of anal sphincter–sparing procedures, when operating for acute ulcerative colitis, one should weigh the possibility of subsequent surgical correction for continence. Specifically, leaving the rectum intact allows its use for subsequent surgical mucosal proctectomy and ileoanal anastomosis. When urgent colectomy is required, total abdominal colectomy, Brooke ileostomy, and Hartmann's pouch are appropriate. Although ileostomy alone for acute complications has been abandoned, it has been used in the recent past with good success by Turnbull and co-workers, in combination with skin-level transverse and sigmoid colostomies, for toxic megacolon. This is a relatively simple procedure that spares such desperately ill patients a major operative intervention until their acute illness has subsided. Because the procedure involves only decompression of the colon and does not remove the acutely inflamed tissue, most surgeons prefer colon resection.

SURGICAL MANAGEMENT

Total proctocolectomy with permanent Brooke ileostomy offers definitive treatment for ulcerative colitis by eliminating diseased mucosa and the risk of malignant transformation. Nevertheless, it remains controversial and is poorly accepted by patients and their physicians. Patients with a permanent ileostomy are incontinent of gas and stool and must wear a collecting bag day and night. As many as 40% to 50% of patients with Brooke ileostomies have appliance-related problems, and the psychologic and social implications, particularly for young patients, are tremendous. Therefore, the search has continued for adequate alternatives to proctocolectomy and ileostomy.

Until recently, single-stage total proctocolectomy was the procedure of choice when complications of the disease were treated electively. This procedure is performed through a midline incision. The rectum is excised from the abdomen after mobilization and circumferential incision from the perineum. When cancer is not suspected, excision is performed rapidly, with division of the mesentery close to the bowel wall. This principle is especially important in the pelvic colon and rectum, where injury to the sacral parasympathetic nerves may lead to bladder and sexual dysfunction. Endorectal mucosal resection, as described later, appears to offer the best way to avoid such serious complications. 34 After standard proctocolectomy, management of the perineal wound is a problem, since chronic infection and poor healing may cause a lingering sinus tract between the buttocks. The authors' preference is to apply active closed drainage to this area for 3 to 5 days following operation. Gauze packing of the perineum should be reserved for pelvic hemorrhage that cannot otherwise be controlled. Perineal wound problems may be reduced by performing an intersphincteric proctectomy, which entails dissecting between the internal and external anal sphincter when removing the rectum, thus preserving the levator ani and external anal sphincter muscles. These muscles can then be included in the closure of the perineum. Using this technique, complete healing of the perineum approaches 95% at 6 months.

The importance of providing patients with a well-functioning, trouble-free ileostomy cannot be overemphasized. Most surgeons have accepted the technique of Brooke. The principles include passing the end of the ileum through an opening in the midaspect of the right rectus muscle at a point below the umbilicus that allows convenient placement of the forepiece of an ileostomy bag. Placement that is too low or too lateral may lead to serious problems in ileostomy care and function. The length of the stoma is important. Approximately 5 cm. should be withdrawn above the skin so that when the tip is folded back upon itself, 2 to 3 cm. protrude from the surface. The folding back or maturing prevents the development of an inflammatory response in the serosa and provides more substance to the protruding ileal nipple. Easily applied receptacles are now available. In the final analysis, it is the need for an external stoma that limits the more general use of colectomy for patients with established ulcerative colitis. Proctocolectomy and ileostomy continue to have a role in the surgical management of ulcerative colitis, being used most commonly in elderly patients, those with poor sphincter function, or those with carcinomas in the lower rectum.

At present, there are several alternatives to proctocolectomy and Brooke ileostomy. Subtotal colectomy with ileorectal anastomosis has been employed as a compromise operation for ulcerative colitis for decades. One advantage of the operation is that it eliminates an abdominal ileostomy and can be offered to patients who adamantly refuse ileostomy. In addition, since the pelvic autonomic nerves are not disturbed, impotence or bladder dysfunction is not encountered. The disadvantages of ileorectal anastomosis, however, are considerable, and with the availability of newer alternatives, these disadvantages may outweigh any advantages. The operation does not eliminate the proctitis. At least 10% of patients require proctectomy for control of the inflammatory disease alone. Patients with ileorectal anastomosis have a considerable risk of developing carcinoma in the rectal remnant: 15% at 30 years in the series of Johnson and colleagues. Functional results also vary, with a high stool frequency necessitating subsequent proctectomy in another 10% of patients. The authors therefore believe that subtotal colectomy with ileorectal anastomosis should be considered only in patients who are not candidates for ileoanal anastomosis and who refuse proctectomy.

In 1969, Kock described a new type of ileostomy, a continent ileostomy, made entirely of terminal ileum and consisting of a pouch that would hold intestinal contents and an ileal conduit that led from the pouch to a cutaneous stoma. This was modified in 1973 to include an intestinal valve between the pouch and the stoma, the valve being constructed by intussuscepting the terminal ileum in a retrograde manner into the pouch for 3 to 4 cm.. Patients would then empty the pouch via the stoma. This technique offered the patient a new lifestyle by making the ileostomy continent, thereby avoiding the need for an external appliance. Although results with the Kock pouch have improved, technical and anatomic complications still necessitate reoperation in up to 40% to 50% of patients. The majority of the problems revolve around the nipple valve, including valve disintussusception and stenosis. Currently, there are few indications for the continent ileostomy. It is an option for patients who have undergone proctocolectomy or who present with carcinoma in the lower rectum. Patients who have a poor anal sphincter mechanism and those who have a failed ileoanal anastomosis are also potential candidates for a continent ileostomy. The latter is currently the most frequent indication for a continent ileostomy. Crohn's disease is an absolute contraindication to the continent ileostomy.

In 1947, Ravitch and Sabiston proposed an anal sphincter–sparing operation that consisted of abdominal colectomy, mucosal proctectomy, and endorectal ileoanal pull-through and anastomosis. As initially proposed, the operation was performed by first resecting the colon in the standard manner. Rather than removing the entire rectum and anus, the disease-bearing mucosa of the rectum was dissected free and resected, preserving an intact rectal muscular cuff and anal sphincter mechanism. Continuity of the intestinal tract was re-established by extending the terminal ileum into the pelvis within the muscular tube and circumferentially suturing it to the anus. The potential advantages of this approach are elimination of all diseased mucosa; preservation of parasympathetic innervation to the bladder and genitalia, and thus avoidance of impotence; avoidance of a permanent abdominal ileostomy; and preservation of the anorectal sphincter apparatus, which is responsible for fecal continence. Despite these theoretical advantages, the operation was associated initially with a high complication rate and an unpredictable functional result, and enthusiasm for ileoanal anastomosis among surgeons declined. During the late 1970s, there was a resurgence of interest in the ileoanal pull-through operation, in part because of disillusionment with the Kock pouch, but mainly because of the improved success of the operation. This improvement was in part a result of the generalized advancement in perioperative and intraoperative surgical care, but specifically, it was the result of several technical alterations in the operation.

Perhaps the most important modification of the operation was the creation of an ileal pouch or reservoir proximal to the ileoanal anastomosis. This addition was partially prompted by the physiologic studies of Heppel and co-workers, which showed an inverse correlation between ileal compliance or capacity and stool frequencies in patients following straight ileoanal pull-through. Several pouch configurations exist: J, S, and W (in increasing size). Studies comparing the functional result following ileoanal anastomosis with and without an ileal reservoir found that 24-hour stool frequency was significantly reduced in patients with ileal pouches, particularly in the early postoperative period.

For most patients, the operation is performed in two stages. The first stage consists of abdominal colectomy, mucosal proctectomy, endorectal ileal pouch–anal anastomosis, and diverting loop ileostomy. During the second stage, performed at least 8 weeks after the initial operation, the loop ileostomy is closed. In patients in whom an emergency colectomy was required, the operation is staged. The first stage consists of abdominal colectomy, ileostomy, and Hartmann closure of the rectum. During the second stage, the rectal mucosa is dissected free and the ileoanal anastomosis is performed with loop ileostomy. Finally, the loop ileostomy is closed. Patients who required prior abdominal colectomy followed by a staged mucosal proctectomy with ileal pouch–anal anastomosis were compared with matched patients who had undergone colectomy with ileoanal anastomosis at a single operation. Previous abdominal colectomy was associated with a higher cumulative operative morbidity, prolonged hospital stay, increased costs, and a less optimal functional result. Aggressive and overly extended medical therapy, including cyclosporine, has been associated with an increased incidence of patients requiring staged subtotal colectomy with delayed ileoanal anastomosis. Therefore, patients with acute ulcerative colitis should be managed initially with a clearly defined course of medical therapy. However, if they have not responded within a reasonable period (i.e., 7 to 10 days), they should undergo prompt surgical intervention, preferably to include single-stage colectomy and ileal pouch–anal anastomosis.

The postoperative morbidity and functional results in most large series after ileoanal pull-through have been encouraging. Eighty-two percent of the patients in a series were operated on for ulcerative colitis and 18% for familial polyposis coli. The mean age was 35 years, with a range of 11 to 67 years. Sixty-two percent of the patients were male. Experience with ileal pouch–anal anastomosis supports the absence of mortality and low morbidity that can be achieved with this operation if it is performed frequently, carefully, and with a standard operative technique. No operative deaths occurred in the series, and the overall operative morbidity after the ileal pouch–anal anastomosis portion of the operation was about 10%. The major operative morbidity was bowel obstruction, both after the initial operation and after loop ileostomy closure. The bowel obstruction rate requiring reoperation compares favorably with the 7% to 13% incidence of reoperation reported after proctocolectomy and ileostomy. An obstruction rate of 10% to 25% is reported in most series of patients undergoing ileal pouch–anal anastomosis. Pelvic and wound infections have been reported to occur in 10% to 20% of patients undergoing ileoanal anastomosis, although the overall infection rate was reduced to about 5% in several more recent large series. A 5% to 10% failure rate necessitating conversion to permanent ileostomy has been reported in several series.

These studies demonstrated that the number of bowel movements during the day is in the range of 1.5 to 6.2, with an average of 5. Nocturnal bowel movements occur 0.2 to 1.8 times nightly, with a mean of slightly more than 1. Of greater importance were control and urgency of bowel movements, which were variable, depending on the time after the operation. Daytime incontinence was extremely uncommon, although nocturnal seepage occurred in 0.6% to 52% of patients. Improvement in these functions continued for more than 2 years postoperatively.

Although results with mucosal proctectomy and ileal pouch–anal anastomosis have been excellent, divergent points of view have arisen regarding the operative technique and its effect on anal physiology and functional result. In recent years, a number of surgeons have advocated an alternative approach to conventional endoanal rectal mucosal resection that eliminates distal mucosal proctectomy altogether. Instead, the distal rectum is divided near the pelvic floor, leaving the anal canal largely intact. The ileal pouch is then stapled to the top of the anal canal. The rationale for this approach is that, by preserving the mucosa of the anal transition zone, the anatomic integrity of the anal canal is preserved and the rate of fecal incontinence improved. Although several studies have suggested that patients have improved sensation and better functional results following preservation of the anal transition zone, this has not been documented by prospective study. The obvious concern is that, by leaving disease-bearing mucosa in the anal canal, the patients are exposed to a lifelong risk of persistent or recurrent inflammatory disease as well as the potential for malignant transformation. Until this technique is further evaluated, patients require careful lifetime surveillance. Mucosectomy must be recommended in patients with rectal dysplasia, proximal rectal cancer, diffuse colonic dysplasia, and familial polyposis. Several recent reports have questioned the need for a proximal diverting ileostomy at the time of ileal pouch–anal anastomosis for ulcerative colitis. The avoidance of a diverting loop ileostomy has several theoretical advantages: it eliminates the additional surgery needed to close the ileostomy, it eliminates the complications of ileostomy and ileostomy closure, and it may reduce diversion enteritis. A diverting ileostomy, however, reduces the risk of leakage from the ileal pouch or ileoanal anastomosis, a serious complication associated with significant morbidity and the potential for total loss of the ileal pouch. Only prospective, randomized, controlled trials will answer the question.

The most frequent late complication in patients undergoing ileoanal anastomosis is ileal pouch dysfunction or pouchitis, which has been reported to occur in 10% to 50% of patients undergoing this procedure for ulcerative colitis. Pouchitis is an incompletely defined and poorly understood clinical syndrome consisting of increased stool frequency, watery stools, cramping, urgency, nocturnal leakage of stool, arthralgias, malaise, and fever. The syndrome is similar to that found in patients with Kock continent ileostomy pouches. The etiology of this condition is unknown; speculations have included early Crohn's disease, bacterial overgrowth or bacterial dysbiosis, either primary or secondary malabsorption, stasis, ischemia, and nutritional or immune deficiencies.

Fortunately, a short course of metronidazole is successful in treating approximately two thirds of patients with pouchitis. The remaining patients have recurrent pouchitis, which responds to repeat metronidazole therapy, or a chronic unresponsive form. It has been argued that, with the high incidence of pouchitis observed in some patients following ileoanal anastomosis, one disease (chronic ulcerative colitis) is simply being replaced by another (pouchitis). Analysis of data from the major large clinical series, however, suggests an overall incidence of pouchitis of approximately 15%. Of these patients, only 10% appear to develop a recurrent or unresponsive variant of pouchitis.

Mortality for elective surgical therapy is in the range of 0% to 2%; for emergency operation, it is about 4% to 5%; and for toxic megacolon, it rises to 17%. These are remarkable statistics when one considers the debilitating nature of the disease and the fact that many patients have had long-term steroid therapy. The major complication in all reported series is sepsis, either in the wound or in the intra-abdominal cavity. There is little evidence that the development of more potent and specific antibiotics has significantly reduced the incidence of this complication; attention to the details of operative management continues to be the best way to ensure a smooth postoperative course. The most common late complication of resectional therapy with ileostomy or ileoanal anastomosis is intestinal obstruction, which occurs in about 10% of patients. Other bothersome but nonlethal complications following proctocolectomy include delay in perineal closure (25%), sexual dysfunction (5% to 10%), and renal stones (10%). Ileostomy dysfunction as a consequence of stenosis has been reduced to 2% by the Brooke-Turnbull ileostomy. Additional uncommon complications include prolapse, herniation, and ulceration of the stoma, which is usually a sign of the development of Crohn's disease within the ileal stoma. Whether the outcome following surgical therapy for Crohn's disease of the colon is as favorable as for ulcerative colitis continues to be a source of controversy.

Results with ileoanal anastomosis in patients with ulcerative colitis are such that this operation is preferable in most patients. The authors and others have found that more than 90% of patients are satisfied with their results, would not consider another alternative, and have fewer restrictions in their daily activities than do patients with Brooke ileostomies or Kock pouches.

The formation of social groups (ileostomy clubs) has provided an important mechanism for the education of patients by those who have already mastered the technique of ileostomy management. Some hospitals have enterostomal therapists. These professionals are highly skilled in dealing with the physical and emotional problems of stomal management. In institutions that perform a large number of ileal pouch operations, specialized patient-oriented support groups are essential.

These are extraordinary advances for patients who suffer from this poorly understood disease. New advances in surgical therapy have made the operative approach even more attractive to patients with ulcerative colitis. The cumulative mortality, as reported by Goligher and associates in their excellent monograph on the subject, suggests that operative therapy should be administered quite liberally in the chronic or acutely fatal forms of the disease. In most patients, fortunately, ulcerative colitis is episodic and mild. Unfortunately, the high incidence of cancer in the presence of persistent disease (especially after 10 years' duration) does not allow these patients a life free of concern.