Several studies have shown the existence of family aggregations with the disease, implying that genetic factors also must play a role.^{5,119,148,351,354,384,769} For example, there is a high degree of concordance in monozygotic twins.⁵⁹⁹ Although ulcerative colitis and Crohn’s disease are not classic genetic disorders, the occurrence of IBD in family members born in widely separated areas or living apart for long periods, along with the increased incidence among Jews and the tendency toward familial aggregation of cases with ankylosing spondylitis in Crohn’s disease, suggest a genetically mediated mechanism in the causation of the conditions.^148,353,566 Roth and colleagues, in a study of Ashkenazi Jews with IBD, calculated the true lifetime risk for the development of IBD in relatives to be 8.9% for offspring, 8.8% for siblings, and 3.5% for parents.⁶³⁴ Although the possibility of a common environment may contribute to the increased risk of IBD, the shared genetic pool is much more likely to be the primary factor. A familial occurrence has been noted in 17.5% of more than 600 patients with IBD.⁶⁸⁸ In a report from the Cleveland Clinic, Farmer and colleagues evaluated the family histories of more than 800 patients with an onset of IBD before the age of 21.¹⁴⁸ Twenty-nine percent of those with ulcerative
colitis had a positive history, and 35% of those with Crohn’s disease had a positive family history for IBD.

The first IBD susceptibility locus was found on chromosome 16 and identified as IBD1. This is apparently a purely Crohn’s locus, however. Most of the current studies seem to demonstrate a consistent association with various genetic mutations and Crohn’s disease, but not that of ulcerative colitis.

Satsangi and colleagues propose that the ethnic, familial, twin, disease association, and genetic marker studies of IBD are best explained by the concept that the conditions represent multifactorial diseases.⁶⁶⁰ Environmental and genetic factors contribute to disease susceptibility. They conclude that it is possible for different susceptibility genes to underlie phenotypic differences between Crohn’s disease and ulcerative colitis, leading to diverse manifestations of the two conditions as well as varied degrees of severity. Mendeloff believes that there is a deficiency in the investigation of these diseases because the mortality is low and the genetic determinants are multiple.⁴⁸⁷ He suggests that in the future it will be necessary to develop a better fundamental means of acquiring and recording data, and that at least for the immediate present a concerted effort should be made to identify those families in which multiple cases of IBD exist and to investigate these people thoroughly. This would include genetic, psychological, and metabolic studies.

The idea of circulating antiepithelial antibodies combining with antigens on the intestinal cell surface and damaging the cells seems a reasonable theory to explain the etiology of IBD. This is the concept of autoimmunity. Snook clearly defines the criteria for the classification of a disease as autoimmune.⁶⁹² According to him, such a condition “… requires the demonstration of autoreactive lymphocytes or autoantibodies, or both, which are specific for the disease concerned, present in all cases, and most important, capable of reproducing the disease on syngeneic transfer.”⁶⁹² However, despite the demonstration of anticolon antibodies in both blood and tissue of patients with IBD, current evidence seems to militate against the likelihood that these play a primary pathogenetic role in the two conditions.⁶⁴²

Immune complex mediation of IBD has been thought by some to be a responsible factor, but studies have failed to corroborate a significantly increased frequency or concentration of these complexes regardless of disease activity.⁶⁴² Other immunologic mechanisms that have been investigated include abnormality and variability of circulating lymphocytes, lymphocyte cytotoxicity, defective cell-mediated immunity, immediate hypersensitivity, leukocyte chemotaxic impairment, and immunoregulatory cellular imbalance.^353,642 At the level of the immune response, a genetic influence is suggested by the association of ulcerative colitis with HLADR2 and by the occurrence of various autoantibodies in the unaffected relatives of patients with ulcerative colitis.⁶⁷⁷ Furthermore, studies of monozygotic twins have indicated that the altered mucosal production of immunoglobulin G1 (IgG1) and IgG2 in ulcerative colitis may be genetically determined.⁶⁷⁷ James and colleagues suggest that the presence of circulating antigen-nonspecific suppressor T cells in patients with Crohn’s disease in its early stages is the result of an immunoregulatory abnormality of antigen-specific helper and suppressor T cells.³⁰⁵ A useful concept of the pathogenesis for IBD may involve an interaction between host responses, immunologic genetic influences, and external agents, but no definitive proof has yet been forthcoming.

With respect to infectious agents, ulcerative colitis in particular has been attributed to bacterial causes for more than 60 years. In 1928, Bargen reported that the condition was caused by a transmissible diplococcus.³⁰ He prepared a vaccine from this organism, an approach that was believed to be a valuable part of the treatment of the condition. In the article that introduced the technique of skin-grafted ileostomy for the surgical management of ulcerative colitis (see Chapter 31), Dragstedt and colleagues were persuaded that “bacterium necrophorum, together with other factors … plays an etiologic role in the disease.”¹⁴⁰ Crohn’s disease specifically was often confused with tuberculosis. This observation led Crohn and coworkers to suggest that the disease might be caused by a mycobacterial agent.¹¹⁴ Subsequent studies have failed to demonstrate conclusively an association with an infective agent, and, in fact, the incidence of IBD correlates inversely
with that of the infectious dysenteries.³⁵³ An exception is that of cytomegalovirus infection, which has been shown to complicate ulcerative colitis in immunocompromised patients, such as individuals with AIDS.⁷⁶¹

The concept of a microbial infection as the offending agent has been resurrected with the recognition of newer bacterial causes of enteritis and colitis (especially Campylobacter jejuni and Clostridium difficile).^{63,380,534,739} Although several studies suggest the possibility of these two organisms contributing to relapse of IBD, Gurian and colleagues, in examination of stool specimens from 32 patients who had exacerbation of IBD, revealed no Clostridium difficile cytotoxin and negative cultures for Campylobacter jejuni.²³⁵ Other bacteriologic agents (Shigella, Salmonella, Streptococcus faecalis, Pseudomonas variant, Chlamydia, Mycobacterium, and many others) have been proposed, but their role has not been confirmed.^{89,203,242,730} More recently, it has been suggested that probiotics play an important role in preventing overgrowth of potentially pathogenic bacteria and in maintaining the integrity of the gut mucosal barrier.⁷⁵⁶ Therefore, there may be a role for such agents in the treatment of IBD (see Medical Management).

There has been considerable interest in a possible viral etiology of ulcerative colitis and Crohn’s disease. Transmission of granulomatous lesions has been successfully carried out in experimental animals.^98,135,722 Tissue culture and electron microscopic investigation have also suggested that a viral agent is present in tissue from patients with IBD.^14,204,774 However, considerable controversy continues concerning the specificity of these findings. Whether the evidence is sufficiently compelling to permit accurate identification of viral particles on electron microscopic examination of affected tissue remains unresolved.^642,787

Dietary factors, especially cow’s milk, have been implicated as possible causative agents for the development of IBD. Early studies seemed to demonstrate an elevated milk protein antibody level in patients with ulcerative colitis in comparison with a control population. Subsequent studies in which milk was excluded from the diet failed to demonstrate an improvement in the clinical response, and later studies with milk and milk products failed to demonstrate any correlation. Other factors that have been under investigation include chemical food additives, mercury ingestion, inadequate fiber, excess intake of refined sugar, and even the increased consumption of corn flakes.^471,642 Generally, articles dealing with diet and IBD are conflicting, confusing, and frequently unreliable. One can safely state that there is no clear consensus to suggest that dietary factors play a role in the etiology of either ulcerative colitis or Crohn’s disease.

Recent evidence suggests that abnormal oxidative metabolism may be of significance in the activity of IBD. Increased attention has been placed on the role of free radicals in both normal metabolism and defense against disease.⁶⁸⁷ A free radical is defined as any species capable of independent existence that contains one or more unpaired electrons, an unpaired electron being defined as one that is alone in an orbital.⁶⁸⁷ It appears that reactive oxygen metabolites are produced in excess in active IBD. The effects of specific anti-inflammatory antioxidants, such as aminosalicylates, are compatible with the proposition that free radicals play a major role in the pathogenesis of IBD.⁶⁸⁷

Studies of the role of the fecal stream in causing an exacerbation of symptoms in Crohn’s colitis have yielded confusing results. For example, Harper and colleagues evaluated the effects of introducing small bowel effluent and a sterile ultrafiltrate of the effluent into the defunctionalized colon after a loop ileostomy had been created.²⁵⁴ There was little response to the ultrafiltrate challenge, but there was a definite clinical exacerbation following introduction of the effluent. Conversely, Korelitz and colleagues reported that diverting the fecal stream had an adverse effect on the clinical course in four patients.³⁶⁸ Following reestablishment of intestinal continuity, the bowel returned to normal.

Two distinct patterns of cigarette smoking seem to be relevant in patients with IBD; those with ulcerative colitis are much less likely to smoke than those with Crohn’s disease. Additionally, cigarette smoking has been found to have a negative correlation with ulcerative colitis.^{77,122,255,310,737,759} In some cases, complete remission of symptoms was obtained through the use of nicotine-laced chewing gum. In other cases, exacerbation of the disease was noted when patients ceased smoking. Conversely, Crohn’s disease is more common in smokers than in those who have never smoked.^330,737,759 The increased risk seems to be more apparent in women and may also be associated with a greater likelihood of recurrence.⁷¹⁴

Both ulcerative colitis and Crohn’s disease have been found to be more common among women using oral contraceptives than in those who do not.⁷⁵⁹ A possible vascular (ischemic) basis for this observation has been suggested. There is no information to date on the effect of stopping the contraceptive pill on the activity of IBD.

The psychological aspects and the possible psychosomatic factors contributing to the onset and exacerbation of IBD, and of ulcerative colitis in particular, have been a subject of considerable debate since publication of the original article by Murray in 1930.⁵²² Karush and colleagues, in a book on Psychotherapy in Chronic Ulcerative Colitis, state that susceptibility to the problem develops in colitic patients through “disruption and distortion of the relationship to the parents and to other significant persons as early as the second year of life.”³²⁸ They further contend that this results in exaggerated emotional manifestations, egocentricity, dependency conflicts, and poor mechanisms for coping with the stresses of life.³²⁸ Frequent anxiety or depression, in the opinion of the authors, is also characteristic, and this predisposition, they believe, explains the relatively high incidence of schizophrenia in patients with ulcerative colitis. In their study of precipitating emotional factors, the authors reported that a well-defined event of powerful emotional impact preceded the onset of the disease by a few days or weeks. Subsequent recurrences were also heralded by such events.

Although many articles have been published to support this description of the susceptible personality, opponents of the psychosomatic theory point out that the concept is based on either anecdotal or uncontrolled studies.⁶⁴² Several publications have compared patients with ulcerative colitis with normal subjects and those with other illnesses and have found no evidence of an increased frequency of psychiatric illness.^{49,141,267,488} Furthermore, those with ulcerative colitis who had a psychiatric illness did not appear to have more serious gastrointestinal involvement, nor did severity of the ulcerative colitis predict a more frequent or more severe psychiatric disorder.²⁶⁷ Bercovitz noted that there are no long-term, prospective psychological studies available of any large group of people.⁵¹ North and colleagues reviewed all known English-language articles at the time (138 studies) on the association between psychiatric factors and ulcerative colitis and found that most contained serious flaws in research design (e.g., absence of control subjects and lack of diagnostic criteria).⁵⁴⁷ In the seven publications that truly represented meaningful, systematic, investigative efforts, no association was identified. Cunnien concluded that patients with ulcerative colitis have no universal unique personality characteristics and no documented increase in psychiatric illness when compared with medically ill and population controls.¹¹⁷ He further observed that no demonstrable emotional precipitating factors have been uniformly identified, no symbolic emotional conflicts have been documented in controlled studies, and psychotherapy has not been effective in altering the disease course.

As has been suggested, the current concept of the etiology of IBD in most patients is that the condition may be triggered by genetic predisposition to a variety of environmental factors. The association of one of these factors, appendectomy, has been the subject of numerous investigations. Koutroubakis and colleagues undertook a meta-analysis of 17 case- controlled studies and found that appendectomy was inversely associated with the development of ulcerative colitis (P < .0001).³⁷² The role of the appendix in the development of mucosal immunity is currently the subject of intensive study.

Ulcerative colitis and Crohn’s disease occur at any age but are most commonly seen in persons younger than the age of 30 years. The incidence is highest among teenagers, but a small secondary peak in the incidence of the two conditions occurs late in the sixth decade.³⁵³ In most series, both sexes are equally affected. Farmer and colleagues noted, in a study covering 20 years ending in 1974, that 838 patients were 20 years old or younger at the time of diagnosis at the Cleveland Clinic.¹⁴⁸ Thirteen percent with ulcerative colitis and 5% with Crohn’s disease were younger than 11 years of age. Approximately one-third of the patients in each group were between the ages of 11 and 15. There were 316 patients with ulcerative colitis, with a male-to-female ratio of almost 1:1, and 522 patients with Crohn’s disease, 57% of whom were male. In Goligher’s experience, approximately one-half of the patients were between 20 and 39 years of age when the disease was diagnosed, with a 4:3 predominance of female over male patients.²¹² With Crohn’s disease, Goligher reported that in his own personal series, female patients predominated in a 3:2 ratio.²¹⁴ There seemed to be a tendency for the disease to develop later than ulcerative colitis does, 70% of patients being between the ages of 20 and 49 years.

In the Lahey Clinic experience of 151 patients who underwent proctocolectomy, the mean age at surgery for both ulcerative colitis and Crohn’s disease was 36 years.¹⁰⁸ Fifty-six percent of patients with ulcerative colitis were men, whereas 57% who had Crohn’s colitis were women.

Goldman and colleagues suggest that Crohn’s disease in black patients may be more common than is generally appreciated.²⁰⁷ This group represented 11% of the patients in their experience during a 10-year period. The authors noted that extraintestinal manifestations developed in all patients. Furthermore, the disease generally appeared to be associated with more severe complications than had been observed in whites.

Physical examination is usually unrewarding in patients with ulcerative colitis who do not have fulminant disease. Abdominal tenderness is usually absent, and there is no abdominal distension. No masses are palpable. However, in the acutely ill person, abdominal distension may be associated with toxic megacolon. Diffuse tenderness may be apparent, and if perforation has ensued, all the usual signs and symptoms of an intra-abdominal catastrophe may be noted.

Proctosigmoidoscopy, flexible sigmoidoscopy, and colonoscopy are important tools for evaluating the bowel and for confirming the presence or absence of IBD. Proctosigmoidoscopic examination is particularly useful in differentiating Crohn’s disease from ulcerative colitis; the rectum is always diseased during attacks of ulcerative colitis. The earliest manifestation of inflammation is the loss of a normal vessel pattern, the result of edema of the bowel wall. Contact bleeding, granularity, and ulceration are more obvious signs of inflammatory disease. The rectum is spared in 40% of patients with Crohn’s colitis, irrespective of anal or perianal involvement. However, when the rectum is involved by this condition, differentiation between the two diseases may be quite difficult.

In individuals with distal disease (i.e., ulcerative proctitis or proctosigmoiditis), complete endoscopic examination by means of the colonoscope is unnecessary at the time of presentation. The extent of disease can usually be determined with the rigid instrument or the flexible sigmoidoscope. The presence of diffuse, confluent, symmetric disease from the
dentate line cephalad to the limit of the inflammatory reaction is consistent with ulcerative proctitis or proctosigmoiditis, depending on the extent of involvement (Figure 29-1). In individuals with treated ulcerative colitis, the finding of rectal sparing or patchiness should not necessarily alter the diagnosis to Crohn’s disease.⁵⁴ If the patient’s symptoms are appropriate to the endoscopic findings, treatment can be initiated without further contrast study or colonoscopy. Conversely, if the patient’s disease extends beyond the limit of the endoscopic procedure performed, it will be necessary at some point to undertake total colonic evaluation.

Colonoscopy has replaced barium enema examination for the evaluation of IBD. Generally, endoscopic examination will identify more proximal inflammatory changes than will the radiologic study. Furthermore, histologic examination of random biopsy specimens will often reveal more proximal disease than was suspected by endoscopic examination.^146,783 Das and colleagues reported 31 patients with idiopathic proctitis who underwent colonoscopy while they were asymptomatic.¹¹⁸ Multiple biopsy samples were taken from throughout the colon and rectum. Although the obvious disease appeared limited to the distal 20 cm, microscopic abnormalities were seen commonly in more proximal locations. The authors postulated that the clinical course would be more consistent with distal disease when this indeed could be confirmed by biopsy, whereas more proximal involvement usually implied that the disease would be refractory to conventional management (e.g., topical steroids). A corollary to this observation is to perform biopsies distal to obvious inflammatory changes if the rectum appears to be spared because one may discover that the rectum is not truly normal. This may cause the physician to reassess the accuracy of a diagnosis of Crohn’s disease based on what was initially thought to be a lack of rectal involvement.

The place of colonoscopy in the evaluation and followup of IBD has been extensively reviewed by many authors. Teague and Waye recommend colonoscopy for five indications: differential diagnosis, resolution of radiographic abnormalities (e.g., filling defects and strictures), preoperative and postoperative evaluation in Crohn’s disease, examination of stomas, and screening for premalignant and malignant changes.⁷²⁷

The development of IBD in the older adult population has been a source of some confusion. Many older patients who have signs and symptoms suggestive of IBD are thought to have ischemic colitis. Conversely, patients thought to have IBD subsequently have been proved to have ischemia as the cause of their symptoms. Brandt and colleagues reviewed 81 patients with colitis whose symptoms began after the age of 50 years.⁶⁸ In this retrospective review, one-half of patients classified as having nonspecific IBD were really thought to have had ischemic colitis. In older persons, ulcerative colitis may have a sudden and fulminating onset progressing to a fatal outcome.

Data from Edinburgh suggest that the rising incidence of IBD in young people is entirely a consequence of Crohn’s disease, a condition now more common than ulcerative colitis in this age group.^670,671 When the condition occurs in children, there may be a more rapid onset and progression than when the disease occurs in young adults. Symptoms are the same as in adults, but toxic megacolon, bowel perforation, and massive hemorrhage are not uncommon sequelae. These youngsters often become chronically ill, have impaired growth and decreased mental acuity, and are less developed physically than their healthy peers (Figure 29-26).^{120,121,499,728,747} Growth failure, especially, is the result of prolonged inadequate caloric intake.³⁴³ It is because of these concerns that implementation of an elemental diet and parenteral nutrition are often part of the management of patients in this age group (see Medical Management).^343,658 However, to be maximally effective, therapy must be initiated before puberty.³⁴³ Furthermore, unless medical treatment can achieve a sustained remission, operative intervention may be the only appropriate method for addressing the problem of retarded development (see Surgical Management).¹³⁷ Particular emphasis should be placed on the assessment of growth and development as well as psychological support for both patient and family.^65,633,658 Close cooperation between the physician and the surgeon is perhaps even more than usually appropriate in the management of these vulnerable individuals.⁶²⁶

Because IBD is common in patients of childbearing age, the possibility of becoming pregnant is often an issue in medical and surgical care. However, pregnancy is not that frequent an event in patients with IBD. The reason probably is related to the fact that these patients may suffer any number of hormonal imbalances as a result of acute and chronic illness, often severely impairing their ability to become pregnant.

However, ulcerative colitis and Crohn’s disease do not adversely affect fertility, nor do they necessarily impede the progress of a pregnancy or the delivery of a normal, term infant. According to Zetzel, pregnancy in association with a preexisting colitis or complicated by the development of IBD is attended by the same prospect of a full-term delivery of a healthy child as is pregnancy in a healthy woman.⁷⁹² However, Schade and colleagues found a significantly greater incidence of low birth weight (less than 2,500 g) in infants of mothers with ulcerative colitis than in a control population.⁶⁶² Baird and colleagues found no evidence of an increased risk for pregnancy loss, but the likelihood of preterm birth was significantly greater.²¹

Levy and colleagues reviewed the clinical course of ulcerative colitis with respect to 60 pregnancies in 31 patients.⁴¹¹ Twenty percent were improved, 18% deteriorated, and 62% demonstrated no change during the course of pregnancy. Fourteen percent of the pregnancies were ended by spontaneous abortion and two by artificial abortions. One premature birth was noted in 50 full-term deliveries. All the births produced healthy children. The authors concluded that pregnancy does not seem to exacerbate preexisting ulcerative colitis, nor does colitis interfere with the outcome of the pregnancy.

Crohn and colleagues reported 74 pregnancies in 47 women whose colitis was inactive at the time of conception.¹¹⁵ All subsequent therapeutic and spontaneous abortions (including one stillbirth) occurred in patients in whom the colitis became activated. There was no difference in the incidence of abortion in patients who conceived during an active phase of colitis in comparison with those who conceived during an inactive phase. These observations have been confirmed by others.¹²⁴

What happens to the colitis in patients who are pregnant? Zetzel reviewed a number of reported series and found it helpful to group the patients into several categories.⁷⁹² Only 30% of patients who became pregnant during a quiescent phase of the colitis had an exacerbation of their disease, but a recrudescence developed in 60% when the pregnancy occurred during an active phase of the illness. In patients whose colitis developed initially during pregnancy or in the postpartum period, a particularly severe result was noted, with more than 60% having worsening of their symptoms. Khosla and colleagues noted that the infertility rate of patients with Crohn’s disease was similar to that seen in the
general population (12%).³⁴⁶ Those whose condition was in remission at the time of conception had a normal pregnancy, with the disease remaining quiescent in most individuals. However, active disease at the time of conception tended to inhibit remission despite therapy.³⁴⁶ In studies specifically in women with Crohn’s disease, pregnancy entailed no increased risk for exacerbation of the bowel inflammation.^540,784 However, an increased risk for premature delivery and spontaneous abortion has been observed in those with active disease or in whom resection is required.

In the counseling of a colitis patient who is contemplating pregnancy, there is no justification for suggesting that attempts at conception be avoided, except when the possibility of teratogenic effects of a medication exists (e.g., the use of metronidazole [Flagyl]). Certainly, immunosuppressive treatment should be avoided in a patient wishing to conceive.

Concerns are also expressed about the safety of drug therapy in men, which could damage sperm and theoretically be associated with teratogenicity. Infertility in men is commonly associated with sulfasalazine administration. Sperm analysis may be helpful in determining whether a problem in conception can be attributed to this cause.

As mentioned, women who have a quiescent form of the disease are unlikely to experience problems with pregnancy and delivery. Conversely, if the patient is experiencing an exacerbation of the colitis, the illness itself may preclude the possibility of pregnancy. If the disease is more than moderately active, Zetzel counsels a temporary waiting period and introduction of appropriate medical therapy to secure a remission.⁷⁹² However, even in this situation, the chances of a normal pregnancy and delivery approximate 50%.⁷⁹² Contraception need be considered only in those women whose disease is so severe that surgery is imminent.⁷⁹¹ Certainly, if the prospective parents wish to have a child, no benefit may be expected from a therapeutic abortion. Even in the severely ill pregnant woman, there is no evidence to suggest that the pregnancy cannot be brought to a successful conclusion with the birth of a healthy child.

If surgery becomes necessary during pregnancy, the method of treatment should be identical to that of a patient who is not pregnant. In other words, drug management (steroids, sulfasalazine) is not contraindicated. It has also been determined that azathioprine is safe and that termination of the pregnancy is not mandatory for those who conceive while taking the drug.⁶ Similarly, if an operation becomes necessary, the procedure should be performed as if the patient were not pregnant, although one should probably defer the implementation of a major reconstructive procedure (see Surgical Management). Closer to term, the enlarged uterus may preclude the possibility of performing even a conventional proctectomy; a staged operation, sparing the rectum, is therefore appropriate.

A high fetal and maternal mortality has been reported if surgical intervention becomes necessary for fulminant colitis.^476,575 Bohe and colleagues noted two cases of fulminant disease during pregnancy that required subtotal colectomy and ileostomy in one, and proctocolectomy in the other.⁶¹ These operations were undertaken in the 32nd and 33rd week of pregnancy, respectively. I have performed surgery in two women during pregnancy: one patient underwent a proctocolectomy at 3 months, and the other underwent a total abdominal colectomy and ileostomy at 5 months. Both proceeded to uneventful conclusions of their pregnancy and delivered normal, healthy infants. In my opinion, it is not in the interest of the mother or the fetus to delay surgery until the pregnancy can be terminated with a viable child. In other words, I do not advocate waiting until the 32nd to perform a cesarean section while the mother is forced to postpone needed surgery.

Lindhagen and colleagues assessed the fertility and outcome of pregnancy in 78 women who had previously undergone resection for Crohn’s disease.⁴²⁶ Neither the number of live births nor the frequency of abortions differed from that which would be expected in the general population. The major factor, again, appeared to be that the disease was in remission or, as in the aforementioned reference, removed.

Successful childbirth has been reported following restorative proctocolectomy with pelvic ileal reservoir (see later discussion in that section).^490,532,590 Twenty-eight patients carried 37 pregnancies to term after continent (Kock) ileostomy, according to a report from Göteborg, Sweden.⁵⁶² Problems encountered were an increased urge to empty the reservoir, especially in the last trimester, and some difficulties with intubation. In most patients, a vaginal delivery was successful, with cesarean section reserved for obstetric indications.

If pregnancy develops following proctocolectomy and ileostomy, the question arises whether the prospective mother should undergo a cesarean section or deliver vaginally. My own feeling is that if the pregnant woman has an adequate pelvis for a normal vaginal delivery, this should be attempted. A cesarean section is not mandatory simply because the patient has an ileostomy, but if an episiotomy is performed, there may be a delay in healing of the perineal wound. However, it has been my experience that the obstetrician almost invariably will opt for a cesarean section. A national registry of ostomates is being maintained; a 1985 United States published survey indicated that about 1,000 had become pregnant.²¹⁷ Parenthetically, prolapse of a loop or end colostomy is sometimes seen, especially in the last trimester. This usually will resolve spontaneously following delivery.

Extraintestinal manifestations were at one time thought to be primarily associated with Crohn’s disease, but with several exceptions, they can be found in both conditions. These are discussed in Chapter 30.

Various factors predispose a colitic patient to colon cancer. These include total colonic or pancolonic disease; prolonged duration of the illness (the earliest reported case is in a patient with the disease of 7 years’ duration); continuous active disease, as opposed to intermittent symptoms; and possibly the severity of disease. An early age of onset probably poses no increased cancer risk save for the fact that cancer risk often parallels duration. The cumulative risk for cancer increases with the duration of colitis, reaching 25% to 30% at 25 years, 35% at 30 years, 45% at 35 years, and 65% at 40 years.⁵⁶⁰ The data also indicates a greater risk in patients with a family history of colon cancer and in those with primary sclerosing cholangitis. Sugita and colleagues observed a strong correlation between the age of onset of ulcerative colitis and the age of onset of cancer, a correlation found both in patients with extensive disease and in those with illness affecting the left side.⁷¹² However, colitis and cancers developed in patients with left-sided colitis about a decade later than in those with extensive disease, although the mean duration of the colitis before the development of cancer was virtually the same in both groups (approximately 21 years), irrespective of the age at onset of the disease.⁷¹² The incidence of cancer in patients with ulcerative colitis has been variously reported to be between 2% and 5%. Öhman reported the overall incidence to be 2.7%.⁵⁶⁰ Johnson and coworkers noted long-term findings in more than 1,400 patients with ulcerative colitis and found that colorectal cancer developed in 63 (4.4%).³¹³ No statistically significant difference was observed in the probability of carcinoma of the colon and rectum developing following proctitis in comparison with total colitis. Patients with left-sided disease seemed to fare considerably better with respect to risk for the development of colorectal cancer. Greenstein reported that neoplasms developed in 11.2% of 267 patients with ulcerative colitis.²²⁶ Colorectal cancers developed in 13% of patients with universal colitis, and malignant change developed in 5% with left-sided colitis. Cancer tended to develop in patients with left-sided disease at least a decade later than in those with universal colitis. The median duration from onset of colitis to diagnosis of cancer was 20 years for those with universal colitis and 32 years for individuals with left-sided colitis.²²⁶ Cancer did not develop in any patient with left-sided colitis before the 23rd year of disease. In a review of more than 1,200 patients seen at the Cleveland Clinic, the cumulative risk for colorectal cancer was significantly higher in those with extensive colitis t han with left-sided disease.⁵⁰⁴ Ekbom and colleagues reviewed more than 3,000 patients with ulcerative colitis and noted that the absolute risk for the development of colorectal cancer 35 years after diagnosis was 30%.¹⁴⁵

In patients who undergo resective surgery for ulcerative colitis, the incidence of associated cancer is considerably higher. Van Heerden and Beart reported the Mayo Clinic experience with 726 patients who underwent surgical exploration for chronic ulcerative colitis between the years 1961 and 1975.⁷⁵⁴ Seventy patients (9.6%) were found to have a carcinoma of the colon. These individuals represented 1.4% of all patients in whom chronic ulcerative colitis was diagnosed during the period of study.

Generally, the incidence of carcinoma is the same in both sexes. This is not surprising in light of the fact that ulcerative colitis affects each in approximately equal numbers. Welch and Hedberg reported a bimodal distribution of the incidence—one peak in the fourth decade and the other in the seventh.⁷⁶⁸

An increased risk of the development of colon carcinoma also appears to be evident in Crohn’s disease, especially of the small bowel.⁵³ This is discussed in Chapter 30.

The distribution of tumors in patients having ulcerative colitis and carcinoma was reported by Öhman to demonstrate multicentricity much more commonly than in those having colorectal cancer without IBD.⁵⁶⁰ He also noted that 28% of patients had lesions of the transverse colon. Similarly, only 28% of patients had cancers involving the rectum and rectosigmoid, whereas patients with lesions of the sigmoid colon comprised 17% of the series. Conversely, Riddell and
colleagues reported that the rectum was the most common site of involvement and that this was more noticeable in men than in women.⁶²³ In their study, more than 40% of the cancers were found in the rectum, and approximately 25% of patients had multiple tumors. It is certainly clear that multicentricity of the cancers is a frequently reported phenomenon (Figure 29-27).

Another characteristic of colorectal cancer with ulcerative colitis is that very often the cancer tends to be infiltrative and scirrhous. Visible tumor involving the mucosa may not be observed even by careful endoscopic examination (Figure 29-28). Although such is the most common clinical presentation of cancer in this disease, it can also appear as a typical ulcerating or polypoid appearance (Figure 29-29).

Another pathologic feature of carcinoma arising in ulcerative colitis is the tendency of the lesion to be highly aggressive and poorly differentiated. More than half of young patients with ulcerative colitis and colorectal cancers have colloid carcinomas with histologically apparent mucus-secreting tumors of the signet-ring cell type⁶⁹⁶ (Figure 29-30; see also Figure 23-22). The fact that there may be few or no symptoms tends to lull both patient and physician into a false sense of security. Witness the situation illustrated in Figure 29-17, in which the mucosa has been completely denuded. When no mucosa is present, bleeding does not occur and mucous discharge is no longer evident. This may cause the physician and the patient to believe the medical measures that have been implemented are effectively controlling the disease. It is in just this kind of circumstance, a patient with long-standing ulcerative colitis, that a carcinoma can supervene. As suggested, physical examination, barium enema, and even endoscopic evaluation may fail to identify the lesion. But when stricture occurs, the patient
must be presumed to have a carcinoma until it can be proved otherwise (Figures 29-31,29-32 and 29-33). The presence of a stricture in a patient with ulcerative colitis is an indication for operative intervention. Lashner and colleagues demonstrated that of 15 patients with strictures, 11 had dysplasia and 2 were found to have cancer on colonoscopy/biopsy.³⁸⁵ An additional 4 patients were found to have carcinomas at the stricture site at the time of colectomy.

Carpeting with “pseudopolyps” makes interpretation of biopsies very difficult. One can make a good case for definitive proctocolectomy and ileo-pouch anal surgery under these circumstances.

Ritchie and colleagues reviewed the St. Mark’s Hospital (Harrow, United Kingdom) experience of carcinoma complicating ulcerative colitis between the years 1947 and 1980; 67 patients with carcinoma were identified.⁶²⁷ In comparison with those who underwent surgery for carcinoma of the colon and rectum in the same time period, it was felt that the colitic group had a higher proportion of inoperable and high-grade tumors, but the prognosis was found to be very similar in patients with and without colitis for the same stage of lesion. In the Mayo Clinic series, 40% of patients with carcinoma in chronic ulcerative colitis had a Dukes’ A or B growth, in comparison with a 63% incidence if carcinoma arose in the absence of the disease.⁷⁵⁴ Conversely, 60% with carcinoma and ulcerative colitis had Dukes’ C and D lesions, in comparison with 37% who had carcinoma alone. The mean age of their patients at the onset of the colitis was 26 years, with a duration of disease of 17 years before the development of malignancy; 23% exhibited multicentric tumors. Those whose carcinoma was identified incidentally during prophylactic colectomy had a 5-year survival of 72%, whereas those with clinical or radiographic evidence suggestive of cancer had a much poorer survival rate (35%).

The advanced nature of the cancerous change is attested to by the report of Johnson and colleagues.³¹³ Only 57% of their patients underwent curative resection, and the overall survival rate in this group was 61%. In Öhman’s experience, two-thirds of those operated on for cure survived 5 years, a percentage virtually identical to the 69% of noncolitic patients.⁵⁶⁰ All with Dukes’ A lesions survived 5 years. Lavery and colleagues reviewed the Cleveland Clinic experience with 79 patients found to have carcinoma arising in ulcerative colitis.³⁸⁸ In comparing their survival s tatistics with those of patients with noncolitic cancer, they too noted no statistically significant difference in survival rates for the same stage of invasion. The poorer results were a consequence of the fact that a higher percentage of patients presented with more advanced or incurable disease at the time of surgery. All reports confirm that the prognosis for colitis-associated colorectal cancers, as for noncolitic cancers, is directly related to the degree of invasion (i.e., Dukes’ stage).^102,711

In 1967, Morson and Pang described a phenomenon they called dysplasia, a frequent and widespread histologic change in patients with carcinoma complicating ulcerative colitis.⁵¹² They suggested that biopsy of the rectum could detect this premalignant situation and possibly dictate the requirement for surgical intervention.

This appeared to be a singular advance in the management of patients with long-standing disease. Before the introduction of the concept of dysplasia, “prophylactic” proctocolectomy was advocated to protect patients from the development of malignancy. This was justified on the basis that although the patient might be asymptomatic, the cancer could be far advanced when discovered. Many individuals with few or no complaints related to the gastrointestinal tract were submitted to surgery because of this understandable concern. With the promulgation of this concept, one ideally can seek to identify those patients who on biopsy and histologic examination are found to harbor this dysplastic phenomenon.

Another method for identifying precancerous changes is flow cytometry. Several studies have demonstrated that DNA aneuploidy correlates with the presence of dysplasia and, therefore, a high risk for developing cancer.^{286,425,432,485,637,715} Suzuki and colleagues showed that 77% of dysplastic tissue demonstrated aneuploidy or polyploidy, whereas 94% of specimens of nondysplastic tissue exhibited diploidy.⁷¹⁵ Löfberg and colleagues found that 20% of 59 patients with long-standing total ulcerative colitis harbored an aneuploid DNA pattern on colonoscopy/biopsy, and that this correlated with the presence of dysplasia.⁴³¹ The frequency of aneuploidy is higher in patients with disease longer than 10 years and with a greater extent of involvement.²⁸⁶

As mentioned, one of the problems with the concept of dysplasia alone is interobserver and intraobserver variability.⁴³² Sampling error as well as total reliance on histologic information has its own inherent limitations. In the experience of Rubin and coworkers, a significant correlation between aneuploidy and severity of histologic abnormality was found in patients at high risk for cancer (negative, indefinite, dysplasia, or cancer).⁶³⁷ In a prospective study from their institution of 25 high-risk individuals without dysplasia, 20% were found to have aneuploidy, and all of these patients progressed to dysplasia within 2.5 years. Conversely, all 19 individuals who failed to demonstrate aneuploidy did not progress to either aneuploidy or dysplasia within the limits of the study (up to 9 years). The authors concluded that patients who demonstrate aneuploidy should be submitted to more extensive and frequent colonoscopic surveillance, whereas those who do not require less frequent investigations.⁶³⁷ Löfberg and colleagues confirmed that nuclear DNA content appears to be an earlier phenomenon than dysplasia in the malignant transformation of the colorectal mucosa, and that the use of flow cytometry in surveillance programs might be of particular value for selecting individuals at an increased risk for the development of cancer.⁴³²

In summary, flow cytometry may be usefully applied to complement histologic examination when dysplasia is suspected.^408,485 However, there is no evidence to support the use of DNA aneuploidy as the sole indication for prophylactic cancer surgery in patients with ulcerative colitis.⁴²⁵

If a carcinoma is identified in the rectum of a patient with ulcerative colitis, proctocolectomy is the treatment of choice. No attempt should be made to preserve the rectal mucosa. Alternative operations may, however, include the continent ileostomy (Kock) and the ileoanal anastomosis with intervening pouch, provided that sphincter preservation does not compromise adequate tumor margins, and the risk for the requirement of postoperative radiation therapy is remote (see Surgical Management).

Approximately 4% to 9% of UC patients will require colectomy within the first year of diagnosis^381,517, whereas the risk of colectomy following that is 1% per year afterward.³⁸² The vast majority of UC patients will require medical therapy throughout their life. Therefore, understanding of the appropriate use of these agents and their adverse effects is important for the physician caring for these individuals. In recent years, the medical management of UC has changed significantly. In particular, the advent and approval of anti-tumor necrosis factor-α (TNF-α) agents like infliximab in the management of moderate-to-severe UC has expanded the role of medical therapy in this condition.^381,382,517

The general principles in medical therapy for UC patients are improvement of quality of life, induction and maintenance of remission and mucosal healing, avoidance of complications and colectomy, and decreasing the likelihood of the development of cancer. Estimated lifetime risk of a severe exacerbation of UC requiring hospitalization
is approximately 15%.¹⁴⁴ Patients with extensive disease (macroscopic disease proximal to the splenic flexure) are more likely to develop acute severe colitis.

A number of agents have been shown to have clinical benefit for induction and maintenance therapy for mild-to-moderate UC. However, the challenge lies with the management of severe UC, which does not respond to corticosteroid therapy. A recent study showed that there was a 7% absolute reduction in the risk of colectomy in the infliximab versus placebo group (10% vs. 17%, respectively) over a 54-week follow-up period.⁶⁵⁶ Whether the use of infliximab decreases the risk of colectomy in the long run is not known. This section will address various medications used in the treatment of UC.

Sulfasalazine and 5-aminosalicylate (5-ASA) remain the first-line therapy for the induction of remission in patients with mild-to-moderate active UC.^82,713 Oral 5-ASAs come in a wide range of formulations with different release characteristics.^418,420 Sulfasalazine, a 5-ASA bound to sulfapyridine by an azo bond, is the initial form found to be effective in the treatment of UC.^132,603 Because the 5-ASA portion of the agent is the therapeutically active compound, several oral preparations of 5-ASA have subsequently been developed. However, sulfasalazine appears to have comparable efficacy with the newer ASA formulations.⁵⁴¹ The type and dosage of 5-ASA therapy are determined by location, severity of disease, cost and insurance coverage, as well as patient preference. Most ASA agents have comparable pharmacokinetics in terms of systemic absorption, urinary excretion, and fecal excretion of active ingredient. Meta-analyses showed that topical 5-ASA agents delivered rectally appeared to be superior to placebo or topical corticosteroids for the induction of remission in distal UC.^392,463,611 However, concomitant topical application of 5-ASA and corticosteroid agents was shown to be superior to topical 5-ASA alone. Topical 5-ASA appears to be at least as effective as oral 5-ASA in maintaining remission for distal UC. 5-ASA appears to be more effective than placebo across all dosage ranges with a trend toward a dose-response effect. Patients with active proctitis or distal colitis can be treated with either topical (enemas or suppositories) and/or oral 5-ASAs. However, controlled trials have shown that rectal therapies have a more rapid effect than oral treatment. Combination therapy with oral and topical 5-ASAs may achieve a higher remission rate than either rectal 5-ASA or oral 5-ASA alone for distal UC. In one study, patients treated with both topical and oral 5-ASAs had a remission rate of 89%, compa red with 69% for topical 5-ASA alone and 46% for oral 5-ASA alone.⁶⁴⁴ In patients with left-sided disease or extensive mild-to-moderate active UC, oral 5-ASAs may be used along with topical 5-ASAs.

The available oral 5-ASA agents appear to be equally effective in producing response rates of 40% to 75% after 4 to 8 weeks of treatment.⁵¹⁵ In those with active UC, delayedrelease oral mesalamine (Asacol HD; Proctor and Gamble Pharmaceutical, Cincinnati, OH) in doses of 2.4 g/day demonstrated a comparable efficacy (51% vs. 56%) versus 4.8 g/day. However, a dose of 4.8 g/day was more effective in moderate disease (57% vs. 72%).²⁴⁵ A formulation of ASA using a MultiMatrix (MMX) release system (Lialda; Shire US, Wayne, PA), which, in addition to being pH dependent (breaking down at pH >7, normally in the terminal ileum, slowly releases 5-ASA throughout the entire colon), has been studied. Clinical remission rates in active mild-to-moderate UC were 37.2% and 35.1% in the 2.4 and 4.8 g/day groups, respectively, after 8 weeks of treatment, compared with 17.5% in the placebo group.^323,419,653 These once daily doses provide an opportunity to improve patient adherence. Of note, combination therapy with oral and rectal mesalamines may be superior to oral or rectal therapy alone in patients with extensive colitis.⁴⁶⁴ With regard to maintenance, oral mesalamine has been shown to decrease relapse rates to 23% to 37% at 12 months compared with 50% to 65% in patients receiving placebo.^515,713

In a subsequent trial, the efficacy of maintenance therapy of MMX was evaluated over a 12-month period (1.2 g/day vs. 2.4 g/day). Remission rates were similar—64.4% with 1.2 g and 68.5% with 2.4 g/day.³²² Another oral mesalamine formulation was developed with the trade name Apriso (Salofalk Granu-Stix, Dr. Falk Pharma, Buckinghamshire, United Kingdom), which has mesalamine granules that have both a gastric acid-resistant enteric coating (which dissolves at pH >6) that delays release. It also has a retarding polymer matrix in the granule core that extends release throughout the colon. This is similar to Lialda. Clinical trials showed that Apriso can be administered once daily and was shown to be as effective and safe in mild-to-moderate UC, with a three times a day dosing schedule.^373,374,595 Finally, several observational studies and a meta-analysis have shown a potential protective effect of 5-ASA therapy on the development of UC-associated colorectal cancer and dysplasia at doses >1.2 g/day.^409,757

Oral corticosteroids can be used for both left-sided and extensive colitis. In patients with proctitis or left-sided disease, topical corticosteroids in the form of foams and enemas are commonly used. In a meta-analysis of topical therapy for active distal UC, topical hydrocortisone appears to be less effective than topical 5-ASAs in inducing remission, whereas rectal application of hydrocortisone or budesonide may be superior to placebo. Topical hydrocortisone may be considered as an alterative agent in patients who fail topical 5-ASA therapy.^304,399

In patients with extensive colitis, moderate-to-severe colitis, or refractory disease, oral and topical 5-ASA therapy and oral corticosteroids may be used as induction therapy. However, parenteral corticosteroids are often required in patients with severe colitis.¹⁵⁴ Response to oral corticosteroid therapy is expected to be within 10 to 14 days, but oral corticosteroid agents should be tapered and should not be used for maintenance therapy because of side effects.⁷⁴⁰ In fact, one of the common mistakes in managing UC is the use of corticosteroids for long-term maintenance. On the other hand, the requirement for oral or parenteral corticosteroids has been considered a prognostic factor. In UC patients requiring corticosteroids, approximately one-third underwent colectomy within 12 months.⁵³⁵

To minimize systemic toxicity, topically active corticosteroid formulations have been tried in UC. Budesonide is an oral glucocorticoid with high first-pass metabolism and, thus, has limited systemic toxicity. In a study evaluating oral budesonide in patients with extensive and left-sided mild-to-moderate
UC, comparable efficacy to that of prednisolone in inducing remission was observed. However, improvement in endoscopic and histologic scores was superior in the prednisolone group.⁴³³ Therefore, oral budesonide has not been routinely used in treating UC.

6-Mercaptopurine (6-MP) and its prodrug, azathioprine (AZA), are antimetabolites that inhibit purine synthesis. Therapeutically, effective doses of AZA are 2.0 to 3.0 mg/kg/day and of 6-MP are 1.0 to 1.5 mg/kg/day. Their therapeutic effect may take up to 17 weeks.⁵⁷⁸ The dosage of 6-MP and AZA may be directed by measuring thiopurine S-methyltransferase (TPMT) activity. Low to intermediate levels of TPMT are associated with the risk of leukopenia.⁵⁷ Thus, patients with normal TPMT activity may receive standard doses of AZA or 6-MP. Patients with intermediate activity can receive 50% of the standard dose. Those who have no TPMT activity should not be treated with these agents.⁶⁹³

6-MP and AZA are mainly used as steroid-sparing drugs for maintenance therapy in UC. It has been shown that patients with UC in remission on AZA >6 months, but later crossed over to placebo, had a higher rate of relapse at 1 year (59%) than those who continued AZA (36%).²⁶¹ 6-MP/AZA typically works in UC patients whose disease activity responds to corticosteroid therapy.^2,13,198,430 6-MP/AZA should not be used as an induction agent and perhaps not for maintenance therapy in patients who fail induction therapy with corticosteroids.

6-MP/AZA use may impact the disease course of UC. For example, there appears to be fewer colectomies undertaken in those maintained on AZA.^13,430 A prospective study evaluating the efficacy of AZA in maintaining remission in patients with acute severe UC after successful induction with corticosteroids found decreased rates of relapse (10% vs. 55%) and severe relapse (0% vs. 36%) when compared with a historical cohort that did not receive AZA after steroid-induced remission.³²⁷ Controlled trials also have demonstrated therapeutic benefit of AZA plus sulfasalazine versus sulfasalazine alone in individuals with newly diagnosed UC.⁷³⁵ A similar ability to maintain remission with less toxicity in AZA alone than AZA plus olsalazine was demonstrated in patients with steroid-dependent UC.⁴⁵⁵ In a meta-analysis of 6-MP/AZA compared with either placebo or 5-ASAs, the mean efficacy (pooled data) was 60% (95% confidence interval [CI], 51%-69%) in the 6-MP/AZA group and 37% (95% CI, 28%-47%) in the control group. When only compared to placebo, 6-MP/AZA was found to be beneficial in maintaining remission (odds ratio [OR] = 2.59; 95% CI, 1.26-5.30).²⁰² The results of the meta-analysis support the efficacy of thiopurines in the maintenance of remission. Accordingly, the American Gastroenterological Association’s guidelines recommended that patients with steroid-dependent UC should be treated with 6-MP/AZA, based on the grade A evidence (homogeneous randomized controlled trials or well-designed cohort studies).⁴¹⁵

Cyclosporine A (CSA), an immunosuppressant that inhibits T-lymphocyte function, has been used in severe corticoste-roid-refractory UC.⁴²¹ In a double-blind, placebo-controlled trial of 20 patients who had failed intravenous corticosteroid therapy, 82% responded to CSA, whereas 0% responded to placebo. Of the 9 patients in the placebo group who did not respond initially, 55% did well after crossed over to openlabel CSA. However, 3/11 in the cyclosporine group versus 4/9 in the placebo group required colectomy within 1 month.⁴²¹ Subsequently, four additional controlled trials of CSA in patients with severe UC were reported. The studies identified a response rate of approximately 80%.^{131,514,716,753} In a study of 30 patients who were randomized to either CSA or methylprednisolone, 64% who received CSA and 53% who received corticosteroids achieved clinical remission within 8 days. At 1 year, 78% of patients in the CSA group were in remission as opposed to only 37% in the corticosteroid group. After 1 year, 7/9 responders in the CSA group were still in remission. This compared with 4/8 in the corticosteroid group (P > .05); colectomy rates were similar.^131,685

Long-term response rates in patients treated with CSA have also been evaluated. In a study of 42 patients with severe steroid-refractory UC treated with intravenous CSA who were followed over a 5-year period, 45% were able to avoid colectomy. This represented a higher rate in those who initially responded to CSA and in those concomitantly on AZA or 6-MP (49% vs. 17%).⁶⁸² A study comparing quality of life (QOL) in patients who underwent colectomy versus those who were managed with CSA found that the latter individuals scored well or better than their surgical counterparts.⁹⁶ In a retrospective single-center study of 86 patients treated with CSA, 25% of initial responders required colectomy at a mean interval of 178 days, and analysis showed that of all patients treated with CSA, 55% would avoid colectomy at 3 years.¹⁵ The dose used in many of the studies is 4 mg/kg per 24 h, with a goal trough serum level of 300 to 400 ng/mL. However, some studies suggest that a lower dose of CSA may be as effective.^514,608 Intravenous CSA is effective in the induction of remission in patients with severe, steroid-refractory UC. However, the efficacy of durable treatment response is limited to those who were “naive” to AZA/6-MP before treatment. All patients require AZA/6-MP for maintenance of remission after induction with CSA.

TNF inhibitors have increasingly been used for the treatment of moderate-to-severe UC. Infliximab (IFX), a chimeric monoclonal immunoglobulin G1 antibody to TNF-α, was the first drug of the category approved by the U.S. Food and Drug Administration (FDA) for the treatment of UC. IFX has been studied in several small open-labeled trials in steroid-refractory and steroid-dependent UC.^{87,88,222,329,710} Subsequently, a double-blind, placebo-controlled study of IFX was terminated prematurely due to slow enrollment, but it showed a 50% response rate at 2 weeks in those previously refractory to intravenous corticosteroids.⁶⁵⁹ Two additional placebo-controlled studies of steroid-refractory disease found favorable results that led to large subsequent studies.^307,596

The Active Ulcerative Colitis Trials (ACT 1 and 2), which enrolled patients with moderate-to-severe active UC treated
with corticosteroids and/or 6-MP/AZA (ACT 1) or with UC refractory to at least one standard therapy (ACT 2), are the landmark studies evaluating the efficacy of IFX.⁶⁴⁰ In ACT 1, the clinical response to IFX at 8 weeks was 69% for 5 mg/kg dosing and 61% for 10 mg/kg dosing versus 37% in the placebo group (P < .001), and similar response rates were found at 8 weeks in ACT 2. In both studies, patients who received IFX were more likely to have a clinical response at week 30, and in ACT 1, more patients who received IFX had a clinical response at week 54. Endoscopic remission, which has been of interest as an end point, was also seen in more than 50% of the IFX-treated groups in ACT 1 at 30 and 54 weeks. Subsequently, a follow-up study was done to evaluate health-related QOL in patients treated with IFX to see whether improved clinical response translated to improvement in QOL. Substantially improved QOL was sustained through 1 year with maintenance therapy.¹⁶² However, the influence on the risk of colectomy was not evaluated until recently. In a published follow-up of ACT 1 and 2 studies evaluating colectomy rates, the cumulative incidence of colectomy in patients treated with IFX through 54 weeks was 10%.⁶⁵⁶ This compared with 17% for the patients in the placebo group. However, those enrolled in the study had moderate-to-severe UC who had not received intravenous corticosteroid within 2 weeks and were judged unlikely to require colectomy within 12 weeks. Hence, the reduction in risk cannot be entirely attributed to IFX.⁶⁵⁶

Previous studies have addressed the risk of colectomy in patients with severe UC. In a small pilot study of 11 patients hospitalized with severe steroid-refractory disease, 50% receiving IFX responded.⁶⁵⁹ This compared with no response in patients in the placebo group, but the numbers were too small to detect a statistically significant benefit. In a study of 45 severe UC in-patients at risk for colectomy, a decreased rate of colectomy at 3 months was demonstrated in those who received one dose of IFX (29% vs. 67%, P = .017).³⁰⁷

The risk of colectomy in the long run was evaluated in a study of 314 UC patients from Italy.¹² Fifty-two (16.5%) had severe UC. Fifteen of the 52 patients (29%) did not respond to a median of 7 days of intravenous corticosteroids. Of these, 4 underwent urgent colectomy, and 11 received IFX. A clinical response was observed in all IFX-treated patients. In the long term, another 6 patients underwent elective colectomy. The overall colectomy rate following acute flare-up was 19%. The long-term colectomy risk was comparable in IFX-treated patients (18%) and in steroid-responsive patients (11%). However, those treated with IFX had a shorter colectomy-free disease course than the individuals who responded to intravenous corticosteroids. The authors speculated that steroid-refractory patients who achieve remission with IFX have a more severe disease than steroid-responsive patients.¹²

Adalimumab (ADA) is a fully humanized IgG1 monoclonal antibody to TNF. Although adalimumab was initially studied in CD, a pilot study subsequently reported its efficacy in UC. More recently, it was studied for the induction of clinical remission in anti-TNF naive patients with moderately to severely active UC.⁶¹³ In the multicenter, randomized, double-blind, placebo-controlled study in North America and Europe, patients with a Mayo score of ≥6 points and endoscopic subscore of ≥2 points despite treatment with corticosteroids and/or immunosuppressants were enrolled. These were randomized to subcutaneous treatment with ADA 160/80 (160 mg at week 0, 80 mg at week 2, 40 mg at weeks 4 and 6) or placebo group. At week 8, 18.5% of patients in the ADA 160/80 group were in remission. This compared with 9.2% in the placebo group (P = .031). Serious adverse events occurred in 7.6% and 4.0% of patients in the placebo and ADA 160/80 groups, respectively. ADA 160/80 appeared to be safe and modestly effective for induction of clinical remission in patients with moderately to severely active UC who failed treatment with corticosteroids and/or immunosuppressants.⁶¹³

All aminosalicylates (sulfasalazine, mesalamines, balsalazide) are pregnancy category B, except olsalazine, which is pregnancy category C. Multiple studies have demonstrated the safety of aminosalicylates in pregnancy. However, because of the antifolate effects, pregnant women are advocates to take folic acid, 2 mg daily, in the prenatal period and throughout pregnancy. Breastfeeding is also considered low risk with sulfasalazine, except for rare diarrhea in infants.

Corticosteroids are pregnancy category C drugs. A case-control study of corticosteroid use during the first trimester of pregnancy noted an increased risk of oral clefts in the newborn.⁶³⁰ This was confirmed by a large case-control study and a meta-analysis that reported a summary for case-control studies that examined the risk of oral clefts (3.35; 95% CI, 1.97-5.69).⁵⁷⁴ However, a subsequent study did not confirm this finding.²³⁴ Overall, the use of corticosteroids poses a small risk to the developing infant, but the mother needs to be informed of both the benefits and the risks of therapy.

Methotrexate, a pregnancy category × drug, is clearly teratogenic and should not be used in women considering conception. Methotrexate is a folic acid antagonist, and its use during the critical period of organogenesis (6 to 8 weeks postconception) is associated with multiple congenital anomalies.⁷⁰ Methotrexate may persist in tissues for long periods, and it is suggested that patients wait at least 6 months from the discontinuation of the drug before attempting conception.

6-MP and AZA are pregnancy category D drugs. Animal studies have demonstrated teratogenicity with increased frequencies of cleft palate, open-eye, and skeletal anomalies seen in mice exposed to AZA. The largest evidence on safety comes from transplantation studies where rates of anomalies ranged from 0.0% to 11.8%.⁵⁹³ In IBD, multiple clinical case series have not noted an increase in congenital anomalies.^344,513 One epidemiologic study did report a higher incidence of fetal loss in women with IBD with prior treatment on 6-MP compared with those who never had 6-MP exposure.⁷⁹⁵ However, a nationwide cohort study
found that women with CD exposed to corticosteroids and AZA/6-MP were more likely to have preterm birth (12.3% and 25.0%, respectively), compared with non-IBD controls (6.5%).⁵⁴⁶ Congenital anomalies were also more prevalent among AZA/6-MP-exposed cases compared with the reference group (15.4% vs. 5.7%), with an odds ratio of 2.9 (95% CI, 0.9-8.9).¹⁰³ Finally, the largest single-center study to date studied 189 women who were exposed to AZA during pregnancy and compared them with 230 women who did not take any teratogenic medications during pregnancy.⁵⁷⁴ The rate of major malformations did not differ between groups.²⁰⁸ The rate was 3.5% for AZA and 3.0% for the control group (P = .775; OR = 1.17). Thus, the conflicting reports in the literature suggest that the decision on the use of 6-MP/AZA during pregnancy should be based on a case-by-case basis.

Cyclosporine is a pregnancy category C drug. A metaanalysis of 15 studies of pregnancy outcomes after cyclosporine therapy reported a total of 410 patients with data on major malformations.²⁷ The rate was 4.1%, which is not different from the general population. In the setting of severe, corticosteroid-refractory UC, cyclosporine may be an option rather than surgery if operation poses substantial risk to the mother and fetus.

IFX, a pregnancy category B drug, is an IgG1 antibody, which does not cross the placenta in the first trimester, but very efficiently crosses in the second and third trimesters.⁶⁸¹ Although this protects the infant from exposure during the crucial period of organogenesis, it is present in the infant for several months from birth.

Current evidence suggests that INF is low risk in pregnancy. The two largest studies are from the TREAT Registry⁴¹⁶ and the INF Safety Database.³³² The TREAT Registry is a prospective registry of patients with CD. Of more than 6,200 patients enrolled, 168 pregnancies were reported, 117 with IFX exposure. The rates of spontaneous abortion (10.0% vs. 6.7%) and neonatal complications (6.9% vs. 10.0%) were not significantly different between IFX-treated and IFX-untreated patients, respectively.⁴¹⁶ The INF Safety Database is a retrospective data collection instrument. The expected versus observed outcomes among women exposed to INF were not different from those of the general population.³³²

IFX crosses the placenta and is detectable in the infant for several months after birth. However, IFX has not been detected in breast milk.⁴⁴⁹ Anti-TNFs should be continued through conception and the first and second trimesters on schedule. If the patient is in remission, the last dose of IFX needs to be given at week 30 of gestation and then immediately after delivery.⁴⁴⁷ The last dose of ADA needs to be given at approximately week 32 of gestation and then immediately after delivery. If the mother flares during this time period, options include giving a dose of anti-TNF or using steroids to manage the patient until delivery.

Adalimumab (ADA), a pregnancy category B drug, is FDA approved for induction and maintenance of remission in CD. The safety of it in UC patients has not been studied as yet.

Patients with acute, fulminant, “toxic” megacolon may present with very minimal symptoms or may be critically ill. High fever, tachycardia, and abdominal pain are frequently noted. However, clinical signs and symptoms may be masked by the patient’s medications, especially steroids. One must keep in mind the possibility of perforation, even in the absence of colonic dilatation. In the experience of Greenstein and colleagues, classic physical signs of peritonitis were absent in six of seven patients with free perforation.²²⁸

The usual supportive measures—intravenous fluid replacement, and blood, colloid, and steroid therapy—should be supplemented with broad-spectrum antibiotic coverage. The single most important guide in the management of a patient with acute toxic dilatation is the assessment obtained with plain abdominal x-ray studies. With serial abdominal films, the effectiveness of medical management can be evaluated (see Figure 29-2). If the dilatation decreases, one may be reasonably assured that surgery can be deferred. Conversely, if colonic dilatation progresses or fails to improve during the period of maximum therapy, surgical intervention is advised.

Any medications that “slow” gastrointestinal activity, such as anticholinergics or opiates, are discontinued. A nasogastric tube is suggested, although some physicians prefer a long tube (e.g., Miller-Abbott) in the expectation of decompressing the colon. Placing the patient on the abdomen for a few minutes every 2 or 3 hours may help to distribute the gas, moving it into the rectum. Rectal tubes have also been advocated, but these are potentially dangerous in that they can cause a perforation of the sigmoid colon. Barium enema examination and colonoscopy are contraindicated; in fact, barium enema study has been reported to precipitate toxic megacolon. A case of successful decompression by means of colonoscopy has been reported in a patient who refused surgical intervention.²⁶

The clinical course and ultimate outcome of toxic megacolon has been well documented by numerous investigators. A high incidence of recurrent toxic dilatation and perforation and the requirement for emergency or urgent operation have been reported.^224,270 This is in contrast to the group of patients with severe, acute colitis without dilatation, who can usually be effectively managed by nonsurgical means.⁵¹⁰ In the series reported by our group from the Cleveland Clinic, only 7 of 115 patients (6%) were successfully managed medically, and 5 of these came to colectomy in later years.¹⁵⁵

For those who fail intravenous corticosteroid therapy, intravenous cyclosporine may be a reasonable approach (see earlier discussion), especially in those who refuse surgery. These usually are “first-episode” patients, those lacking a history of chronic debilitation or individuals whose comorbid conditions create a formidable surgical risk.

The cyclosporine experience has had a mixed response, with an initial 82% success rate, but falling to 59% at 6 months.⁴²¹ A review of more than 20 uncontrolled studies in 1998 demonstrated a 68% response in avoiding colectomy, but the long-term response was only 42%.⁴⁶⁰ An international survey of practitioners (flawed perhaps in that two-thirds of responders had experience with fewer than five patients) reported “good” results in 29.5%, “acceptable” with recurrence in 58.6%, and “poor” in 14%.⁴⁹⁵

The University of Chicago study was most favorable, with 72% of initial responders avoiding colectomy after 5 years.⁹⁷ The important observation here was the value of concomitant 6-MP or AZA therapy. Of the patients receiving cyclosporine and given these drugs, 80% avoided colectomy and maintained their initial response. Further analysis of a 5-year follow-up of 42 cyclosporine-treated patients at the University of Chicago initially receiving from one to four courses of intravenous cyclosporine revealed that 18 (43%) retained their colons after a median of 6.7 years.⁹² It is interesting to note that only 1 of 8 patients receiving more than one course of cyclosporine avoided colectomy. These investigators concluded that short-term cyclosporine followed by 6-MP/AZA permits more than 50% of steroid-resistant patients to avoid colectomy. However, re-treatment with cyclosporine is rarely successful.⁹²

Some reports confirm the efficacy of cyclosporine in the management of severe colitis.^298,474 Although many initial responders subsequently relapse, a substantial minority remain in long-term remission.⁴⁷⁴ Moreover, there appears to be no increased incidence of perioperative complications associated with its use, provided the treatment is for a defined period, and needed surgery is not delayed.²⁹⁸ It is, however, doubtful if cyclosporine can be considered in the realm of a truly long-term, effective therapy.⁶⁰⁸

The following are the guidelines of the American College of Gastroenterology with respect to managing severe ulcerative colitis:

Neither an elemental diet nor total parenteral nutrition decreases the inflammation associated with ulcerative colitis.²⁴⁴ However, evidence suggests that patients frequently are hospitalized with varying states of malnutrition. As a consequence, hyperalimentation, either parenteral or oral, has been recommended in a supportive role for patients with IBD. Specifically, elemental diets and total parenteral nutrition with bowel rest improved the symptoms, inflammatory sequelae, and nutritional status in individuals with Crohn’s disease more readily than in those with ulcerative colitis (see Chapter 30). It has been demonstrated by some authors that patients who have lost more than 20% of their usual weight before undergoing abdominal surgery have higher rates of morbidity and mortality than those who have not exhibited weight loss. Conversely, in a study by Higgens and colleagues, preoperative weight loss did not adversely affect the postoperative outcome in those undergoing elective resection.²⁸⁰ There is extensive, often confusing literature on nutritional data, diet, and intravenous hyperalimentation.^{94,192,263,317,370,405,406,618}

With IBD, the rationale for implementing intravenous hyperalimentation is that the bowel is “put to rest.” If this were attempted without supplementary intravenous caloric intake, the patient’s nutritional status would rapidly deteriorate. Intravenous hyperalimentation, therefore, permits the patient with IBD to be managed with bowel rest while simultaneously providing adequate amino acids and calories for anabolism.^134,256 If surgery is believed to be inevitable, however, the Veterans Administration Cooperative Study of 395 malnourished patients revealed that total parenteral nutrition should be limited to those who are severely malnourished unless there are other specific indications for this treatment.⁷⁶⁰

Therapy must be individualized to the patient’s dose-response experience, the extent of the disease, prior relapse history, and whether therapy or “no therapy” has been of value previously. Clearly, most patients will benefit from some form of maintenance treatment. The concept of inducing a complete remission must be emphasized before considering changing or reducing an effective treatment program to a maintenance schedule. Prematurely tapering steroids prior to achieving complete remission is a frequent error, but of
equal importance is the fact that steroids are ineffective in maintaining remission. Their value is in short-term use or in induction therapy.²⁴⁴

The mainstay in this effort is 5-ASA. If left untreated, 80% of patients will relapse. Numerous studies have shown that treated patients will remain in remission longer than those given a placebo.^144,496,505 A multicenter trial comparing oral, topical, and oral with topical mesalamine in maintaining remission in distal ulcerative colitis showed equal efficacy of all three therapies in the prevention of relapse.²⁸¹

Sachar, at the Mount Sinai School of Medicine, presented a personal essay on common errors in the management of IBD which is worthy of reproducing here. They are as follows:

Preparation of the patient for elective surgery is not significantly different whether the procedure is resection for IBD or surgery for cancer. It is a wise idea, however, to limit the amount of laxative administered. In fact, if a patient is troubled by diarrhea, a preoperative cathartic should be avoided. On the morning of surgery, an enema may be carefully administered until the returns are clear. This is the only mechanical preparation advised for patients with severe bowel frequency problems. Those who are to undergo small bowel resection for Crohn’s disease do not require a mechanical preparation unless the possibility of colonic resection also exists. Intravenous antibiotics are administered as described in Chapter 23.