Toxic megacolon

Authors

  • Daniel M. Autenrieth MD,

    1. Division of Gastroenterology and Hepatology, Department of Medicine, Virchow Hospital, Charité Medical School, Humboldt-University of Berlin, Germany
    Search for more papers by this author
  • Daniel C. Baumgart MD, PhD

    Corresponding author
    1. Division of Gastroenterology and Hepatology, Department of Medicine, Virchow Hospital, Charité Medical School, Humboldt-University of Berlin, Germany
    • Division of Gastroenterology and Hepatology, Department of Medicine, Charité Medical Center, Virchow Hospital, Medical School of the Humboldt-University, 13344 Berlin, Germany
    Search for more papers by this author

Abstract

Abstract: Toxic megacolon represents a dreaded complication of mainly inflammatory or infectious conditions of the colon. It is most commonly associated with inflammatory bowel disease (IBD), i.e., ulcerative colitis or ileocolonic Crohn's disease. Lately, the epidemiology has shifted toward infectious causes, specifically due to an increase of Clostridium difficile-associated colitis possibly due to the extensive (ab)use of broad-spectrum antibiotics. Other important infectious etiologies include Salmonella, Shigella, Campylobacter, Cytomegalovirus (CMV), rotavirus, Aspergillus, and Entameba. Less frequently, toxic megacolon has been attributed to ischemic colitis, collagenous colitis, or obstructive colorectal cancer. Toxic colonic dilatation may also occur in hemolytic-uremic syndrome (HUS) caused by enterohemorrhagic or enteroaggregative Escherichia coli O157 (EHEC, EAEC, or EAHEC). The pathophysiological mechanisms leading to toxic colonic dilatation are incompletely understood. The main characteristics of toxic megacolon are signs of systemic toxicity and severe colonic distension. Diagnosis is made by clinical evaluation for systemic toxicity and imaging studies depicting colonic dilatation. Plain abdominal imaging is still the most established radiological instrument. However, computed tomography scanning and transabdominal intestinal ultrasound are promising alternatives that add additional information. Management of toxic megacolon is an interdisciplinary task that requires close interaction of gastroenterologists and surgeons from the very beginning. The optimal timing of surgery for toxic megacolon can be challenging. Here we review the latest data on the pathogenesis, clinical presentation, laboratory, and imaging modalities and provide algorithms for an evidence-based diagnostic and therapeutic approach. (Inflamm Bowel Dis 2012;)

The term “toxic megacolon” denotes a rare but severe and potentially fatal complication of colonic inflammation. Its main characteristics are radiographic evidence of total or segmental colonic distension of >6 cm. In contrast to other types and causes of colonic dilatation like Ogilvie's syndrome or Hirschsprung's disease, toxic megacolon is defined by the additional presence of systemic toxicity and the inflammatory or infectious etiology of the underlying disease.1–3

EPIDEMIOLOGY

Precise data on the epidemiology of toxic megacolon is scarce and larger studies date back as far as the early 1980s. Moreover, the incidence of toxic megacolon varies depending on its etiology. Previous epidemiological studies mainly focused on the incidence of toxic megacolon in inflammatory bowel disease (IBD). In a retrospective study Greenstein et al4 reviewed 1236 patients with IBD admitted to hospital in a period from 1960 to 1979. In all, 75 cases of toxic megacolon were identified (6%). In 10% of the patients with ulcerative colitis (UC) and in 2.3% with Crohn's disease (CD) the condition was complicated by toxic megacolon. In a more recent prospective study toxic megacolon was reported in 7.9% of patients admitted with UC.5

Until the late 1990s the incidence of toxic megacolon in patients with pseudomembranous colitis (PMC) was reported between 0.4% and 3%.6–8 Over the last decade, however, major changes in the epidemiology of Clostridium difficile infections (CDI) have been observed.9–12 Zilberberg et al13 demonstrated an annual increase of 23% in the rate of hospitalization due to CDI in the USA over the time period from 2000 through 2005. Similar observations have been reported from a number of industrialized countries.14–16 Zilberberg et al13 documented an increase of CDI-related mortality rates from 1.2% in 2000 to 2.2% during an outbreak of CDI in Quebec in 2004. Several studies documented an overall increase in severity and mortality rates.17, 18 These changes are attributed to the more frequent use of broadband antibiotics and the emergence of a hypervirulent C. difficile strain (BI/NAP1/027)19 and account for an overall increase in C. difficile-associated toxic megacolon. Shifting demographics are an additional factor in epidemiological changes. Patients age 65 and older are at higher risk of both developing CDI and more severe courses of disease including toxic megacolon.9, 10 Cober and Malani20 retrospectively examined medical records of 70 CDI patients aged over 80 admitted to a tertiary care facility from January 2006 to December 2006. Complications in this population included three cases of toxic megacolon (4.3%). Notably, there has been a rise in the numbers of community-acquired CDI in recent years as well.21, 22 Moreover, increasing numbers of patients previously considered to be at low risk like peripartum women and pediatric patients develop community-acquired CDI.21 There is one case report of a pregnant woman with toxic megacolon complicating a severe CDI. No traditional risk factors of CDI were documented in that case.23 Toxic megacolon complicating PMC is associated with a mortality rate of 38%–80%.6, 24

There is a striking lack of current reliable data on the epidemiology of toxic megacolon. Therefore, further studies are required to account for therapeutic advances, etiological factors, and shifting demographics in recent years.

ETIOLOGY

Toxic megacolon is most commonly considered a complication of IBD, specifically UC and to a much lesser extent CD.25 Over the course of the last decades, however, the list of etiological factors was expanded by a vast array of inflammatory and infectious conditions: bacterial colitides such as C. difficile, Salmonella,26Shigella,27 and Campylobacter28 as well as viral (e.g., cytomegalovirus [CMV])29 and parasitic (Entameba)30 infections may be complicated by colonic dilatation. Further etiological factors include ischemic colitis,31 Behçet's disease,32 and malignancies like colonic lymphoma33 and Kaposi's sarcoma. There have been case reports of aspergillosis,34 collagenous colitis,35 and rotavirus36 involved in toxic megacolon. Toxic colonic dilatation may also accompany hemolytic-uremic syndrome (HUS) caused by E. coli O157.37

In a recent retrospective study by Ausch et al,38 70 patients treated surgically for toxic megacolon between 1985 and 2004 were examined. In 32 cases (46%) UC was identified as the main cause of disease, followed by infectious colitis (34%) and ischemic colitis (11%). One case (2%) of toxic megacolon complicating CD was documented. Remarkably, the percentage of toxic megacolon caused by IBD decreased in the latter half of the observation period from 30% of all cases to 17%, whereas other causes became more prevalent (14% vs. 39%). This finding may reflect the overall increase in toxic megacolon complicating CDI.

Toxic megacolon also plays a role in the immunocompromised patient. CMV and C. difficile infections are the predominant causes of toxic megacolon in patients with human immunodeficiency virus (HIV) or AIDS.39 Cytotoxic chemotherapy is also reported to be a potential cause of toxic megacolon.40 In the study by Ausch et al toxic megacolon caused by chemotherapy accounted for 2 out of 70 cases (3%). Hayes-Lattin et al41 described three cases of toxic megacolon following high-dose chemotherapy and autologous stem cell transplantation in patients with AL amyloidosis (Table 1).

Table 1. Etiological Factors of Toxic Megacolon
Inflammatory
 - Ulcerative colitis
 - Crohn's disease
 - Behçet's disease
Infectious
 - Clostridium difficile
 - Salmonella, Shigella, Yersinia, Campylobacter, E. coli
 - Cytomegalovirus, Rotavirus
 - Entameba
 - Aspergillosis
 - Cryptosporidium
Ischemia
Other causes
 - Collagenous colitis
 - Colonic Lymphoma
 - Kaposi's sarcoma
 - Chemotherapy

PATHOGENESIS

The precise pathophysiology of toxic megacolon is not fully understood. However, an association between inflammatory conditions of the colon and decreased smooth muscle contractility is well established.42, 43 The progression from uncomplicated UC to toxic megacolon is best studied to date. Being limited to the mucosa and superficial layers of submucosa in uncomplicated UC, the inflammatory processes penetrate into the muscularis propria in toxic megacolon. The depth of inflammation seems to be correlated with the extent of colonic dilatation.44 Moreover, a number of inflammatory mediators are known to have an inhibitory effect on colonic motility. Nitric oxide (NO) as the key nonadrenergic, noncholinergic neurotransmitter induces colonic smooth muscle relaxation.45 Increased expression of inducible nitric oxide synthase (iNOS) has been shown in animal models of colitis46 as well as in the colonic mucosa of UC patients.47 In a study by Mourelle et al48 high levels of iNOS were detected in the muscularis propria of patients with toxic megacolon. In a rat model, colonic diameter, motility, and intracolonic pressure in colitic animals could be improved by selective inhibition of NOS.49 Cao et al50 showed in an in vitro study that mucosal cells of patients with UC release significantly more H2O2, interleukin (IL)-1β, and NO into the submucosal side than mucosal cells of control patients. Treatment of normal sigmoid smooth muscle cells with the “undernatant” collected from the submucosal side of UC mucosa resulted in significantly decreased contractility. This effect could be partially prevented by pretreatment with the H2O2 scavenger catalase, an IL-1β, or the hemoglobin scavenger hemoglobin, suggesting an inhibitory effect of those mediators on colonic smooth muscle contractility. Changes in neuromuscular signaling due to inflammatory processes could also contribute to a decrease in colonic motility. Strong et al51 described the effects of inflammation on inhibitory purinergic neuromuscular transmission resulting in impeded motility in a guinea-pig colitis model. In contrast to previous studies, it was shown that this effect was not attributed to loss of neurons within the myenteric ganglia or to neuronal fiber deficits. These findings may represent distinguishing pathomorphological features of colonic dilatation due to colitic conditions as opposed to congenital megacolon in Hirschsprung's disease in which aganglionosis is the defining histological finding.23

CLINICAL PRESENTATION AND LABORATORY INVESTIGATIONS

Toxic megacolon is a diagnosis based on clinical signs of systemic toxicity in combination with radiographic evidence of colonic dilatation. There appears to be no predilection in any particular age group or gender. Patients with IBD are at highest risk of developing toxic megacolon at an early stage of disease: up to 30% of patients present within 3 months of diagnosis.2 In the study by Ausch et al38 toxic megacolon was the first clinical presentation of a previously undiagnosed UC in 13% of the patients. In a retrospective study by Benchimol et al,52 5 out of 10 pediatric patients with toxic megacolon were admitted with the first presentation of IBD. The history of present illness usually reveals signs and symptoms of acute colitis prior to the onset of colonic dilatation. A careful history of possible causes should include an inquiry about previous diagnosis of IBD, exposure to enteric pathogens and medication, especifically steroids, antibiotics, and antimotility agents (Table 2). Upon physical examination signs of systemic toxicity generally predominate. Abdominal tenderness and reduced bowel sounds are frequent findings; signs of peritonitis may indicate colonic perforation. It is important to account for precursory steroid treatment as it may mask the symptoms of toxicity. The most commonly used clinical criteria for the diagnosis of toxic megacolon were proposed by Jalan et al1 in 1969. Three of four of the following main criteria are mandatory for clinical diagnosis: fever, tachycardia, leukocytosis, or anemia. In addition, one of the following criteria should be met: dehydration, altered level of consciousness, electrolyte imbalances, or hypotension. One clinical study observed mental state alterations in up to 41% of adult toxic megacolon patients.53 However, clinical signs of toxic megacolon may differ in pediatric patients. Benchimol et al52 reported that hypotension and altered level of consciousness are rare clinical findings in children with toxic megacolon.

Table 2. Diagnostic Approach
 
History
Signs of Inflammatory bowel disease and acute colitis
 - Abdominal pain
 -  (Bloody) Diarrhea
 - Vomiting
 - Weight loss
 - History of previous exacerbations
 - Extraintestinal manifestations
Possible exposure to enteric pathogens
 - Family, environment
 - Recent travels
Medication, especially
 - Antibiotics
 - Steroids
 - Antidiarrheals
 - Anticholinergics
 - Opiates
Immune status
 - Chemotherapy, malignancy
 - HIV
Physical examination
 - Abdominal pain, tenderness, distension
 - Constipation, obstipation
 - Reduced bowel sounds
 - Fever
 - Tachycardia, hypotension
 - Mental changes
Laboratory investigations
Inflammation
 - Elevated white blood cell count
 - Elevated C-reactive protein
 - Elevated erythrocyte sedimentation rate
Anemia
Electrolyte imbalances
Blood culture
Fecal screening for pathogens
Stool samples for C. diff. culture and A/B toxin assay
Clinical Criteria for toxic megacolon (by Jalan et al.) [1]
Main criteria (at least three of the following)
 - Fever (> 38.6°C)
 - Tachycardia (>120/min)
 - Leukocytosis (>10.5 x 109/l)
 - Anemia
In addition at least one of the following
 - Dehydration
 - Altered level of consciousness
 - Electrolyte imbalances
Hypotension
Imaging studies
Plain abdominal imaging
 - Colonic dilation >6 cm
 - Disturbances or loss of colonic haustration
 - Air-fluid levels
 - Small bowel distension
Additional CT findings:
 - Colonic wall thickening
 -  ‘Accordion sign’; ‘target sign’
 - Pericolic stranding
 - Complications: abscesses, perforation, ascending pyelophlebitis
Abdominal Ultrasound

Laboratory studies in patients with toxic megacolon may present with several, nonspecific abnormalities and reflect system toxicity and the severity of colitis: Anemia and leukocytosis with a left shift is a common feature as well as an increased erythrocyte sedimentation rate and elevated C-reactive protein. Electrolyte imbalances are one of the main criteria for toxic megacolon proposed by Jalan et al. In particular, hypokalemia and hypoalbuminemia are associated with severe diarrhea, volume loss, and a poor prognosis in general. Fecal screening for pathogens should be performed in order to detect bacteremia blood cultures. CDI is diagnosed by bacterial culture or detection of toxin A and B in stool samples for example by enzyme immune assay (EIA).54

IMAGING STUDIES

Detection of colonic dilatation as a critical feature in the diagnosis of toxic megacolon is the domain of radiological examination. Typical features seen in plain abdominal radiographs include dilatation of the colon greater than 6 cm, not rarely up to 15 cm.55 Fazio56 documented a mean colonic diameter of 9,2 cm. Typically, ascending and transverse colon are the most dilated. Further radiographic features include colonic air-fluid levels and loss or disturbance of colonic haustration. In pediatric IBD patients over 11 years old a transverse colonic dilatation of more than 56 mm in addition to clinical signs is strongly indicative of toxic megacolon.52 However, in younger patients transverse colon diameter is unlikely to exceed 40 mm.57

In a case series by Maconi et al58 ultrasonographic findings in four patients with toxic megacolon in IBD were evaluated. Ultrasound revealed thin colonic walls with loss of haustra coli and increased diameter (>6 cm) with gaseous content similar to the features found in plain abdominal radiographs. Those features are nonspecific and in the individual context it may be difficult to differentiate between toxic megacolon and other causes of colonic dilatation. However, due to the wide availability of ultrasound it may be a valuable diagnostic tool in the recognition or suspicion of toxic megacolon. In a retrospective study Imbriaco and Balthazar59 examined computed tomography (CT) findings in 18 patients with toxic megacolon. CT scans were useful in detecting signs of severe colitis including diffuse colonic wall thickening, thickened haustra with alternating bands of high and low density (“accordion sign”), multilayered appearance caused by different densities of edematous submucosa, and hyperemic mucosa (“target sign”) and pericolic stranding. The additional detection of colonic dilatation (mean transverse diameter 7.7 cm, range 6–10 cm) was indicative of the development of toxic megacolon. CT was more reliable in evaluating both the length and severity of colitis and the presence of colonic dilatation than plain abdominal radiographs. Furthermore, CT scan represents an important tool in the diagnosis of abdominal complications of toxic megacolon like perforation or ascending pyelophlebitis. Imbriaco and Balthazar59 showed that in 4 out of 12 patients CT investigation found abdominal complications missed clinically and on plain abdominal films. However, CT imaging was not able to determine the etiology of the inflammatory process nor did it predict the clinical outcome.59 Larger clinical studies are warranted to assess the clinical significance of radiological studies in the assessment of toxic megacolon.

ENDOSCOPY

In the acute setting of toxic megacolon total colonoscopy bears the high risk of colonic perforation and is therefore generally contraindicated. Limited sigmoidoscopy is considered to be valuable in differentiating etiological causes, for example, to exclude CMV or by identification of pseudomembranes in acute C. difficile colitis. In a prospective study by Johal et al60 in 179 patients with diarrhea of unknown etiology the diagnostic significance of sigmoidoscopy was compared with stool assays for C. difficile toxin. Out of 56 patients who showed PMC on sigmoidoscopy, 29 had a negative cytotoxin assay for C. difficile. The authors therefore argued in favor of the use of limited sigmoidoscopy to confirm CDI. However, in a study by Hall et al61 in 36 patients undergoing total colectomy for C. difficile colitis, endoscopy (rigid sigmoidoscopy or colonoscopy) was only used in three cases to confirm diagnosis. In one patient perforation occurred. Therefore, in toxic megacolon the diagnostic use of flexible sigmoidoscopy should be carefully balanced with the high risk of perforation. If necessary the procedure should be performed solely with minimal air insufflation.

MANAGEMENT

The involvement of both medical and surgical expertise is critical to a favorable outcome of toxic megacolon. The main objectives of management of toxic megacolon are treatment of the underlying cause, attenuation of colitis, treatment of toxemia, and circumvention of further complications, specifically bowel perforation. Close medical monitoring and supportive care is imperative.

Abnormalities in electrolyte levels, especially hypokalemia, as well as dehydration and anemia may deteriorate atony or dilatation of the colon and therefore should be monitored and aggressively balanced. Medications impacting colonic motility, for instance, opiates, anticholinergics, and antidiarrheal, are considered to deteriorate the course of disease62 and should be discontinued immediately.

Bowel rest with total parenteral nutrition is not generally recommended in acute severe colitis. In a randomized trial by Gonzalez-Huix et al63 enteral and parenteral nutrition was compared in patients with acute severe UC. There were no significant differences with regard to remission rate and need for colectomy between both groups. Complications typically associated with total parenteral nutrition were less frequent in enterally fed patients. However, in the setting of toxic megacolon and colonic distension, bowel rest is usually necessary to attenuate symptoms of colonic atony

Nasogastric suction does not affect colonic distension and is considered unnecessary. It was observed that while the patient remains in prone position gas tends to accumulate in the transverse colon.64 Thus, some authors suggested repositioning of patients as a means of bowel decompression through redistribution of colonic gas to the distal colon and rectum. Repositioning techniques include rolling maneuvers55 and a prone knee-elbow position of the patient.65 However, performing those maneuvers in clinical routine may be impracticable.

IBD has been identified as an independent risk factor for venous thromboembolism. Miehsler et al66 showed a 3.6-fold increased risk of thromboembolism in patients with IBD compared to control subjects. Accordingly, prophylaxis for deep-venous thrombosis with low-dose heparin plays an important role in the basic medical management of toxic megacolon caused by IBD.

MEDICAL THERAPY

There is still a lack of controlled studies investigating medical management of toxic megacolon. Commonly accepted regimens are based on the treatment of the underlying cause, retrospective studies, and expert opinion. The diagnosis of toxic megacolon is not an absolute indication for a surgical course of treatment. However, in order to avoid colectomy medical treatment should be carried out aggressively and in a timely fashion.

The mainstay of medical therapy for patients with toxic megacolon caused by UC is high-dose intravenous steroids. Steroid treatment should be started immediately and not be delayed by pending microbiological results. Most authors recommend a daily dose of either 400 mg hydrocortisone (100 mg every 6 hours) or 60 mg methylprednisolone given intravenously for about 5 days.3, 67, 68 There is no benefit in much higher doses of steroids or extension of therapy.68, 69 The notion that the use of steroids increases the risk of perforation is not supported by clinical data.70 However, clinical signs of colonic perforation as well as symptoms of toxicity may be masked by the use of high-dose steroids. Moreover, the failure rate of intravenous steroids in patients with severe colitis is ≈20%–40%.71 In order to avoid delay in the timing of necessary surgery, steroid therapy should be reevaluated on a regular basis. Moreover, steroids should not be administered once an exclusively infectious cause of toxic megacolon has been established.

There are no data supporting a benefit of oral aminosalicylates. Therefore, sulfasalazine and 5-aminosalicylate (5-ASA) play no role in the treatment of toxic megacolon caused by IBD. Although used by some in the setting of severe UC, there are no controlled studies investigating the use of cyclosporin A or the monoclonal antibody infliximab in toxic megacolon. In a controlled double-blind trial, patients with steroid-refractory severe colitis were treated with cyclosporin in a dose of 4 mg/kg per day; 82% of the treated patients showed clinical improvement and colectomy was avoided.72 This favorable response rate has been reproduced by several other studies.73–75 In a study by D'Haens et al76 cyclosporin was shown to be an effective alternative to intravenous corticosteroids in severe UC. Tacrolimus, like cyclosporin, is a calcineurin inhibitor. In a randomized, placebo-controlled trial, patients with refractory UC treated with tacrolimus were more likely to achieve clinical response.77 There is a single case report of a successful bridging treatment of toxic megacolon caused by UC with tacrolimus.78 Infliximab (IFX) is also suggested (single dose of 5 mg/kg) as an effective rescue therapy in steroid-refractory UC.68 In a study by Jarnerot et al79 45 patients with steroid-refractory UC were treated with either IFX vs. placebo (with continued betamethasone). Out of 24 patients in the IFX group seven had to undergo colectomy within 3 months compared to 14 out of 21 patients in the placebo group. In a recent study by the same group the role of cyclosporine and infliximab as rescue therapies was examined.80 The results showed significantly lower colectomy frequencies in the cyclosporine group than with a single infusion of IFX. Moreover, the short half-life of cyclosporine may represent an advantage over infliximab once colectomy is indicated. Due to the lack of clinical studies in the setting of toxic megacolon and their strong immunosuppressive and toxic effects, both calcineurin inhibitors and IFX are not generally recommended in the treatment of toxic megacolon. Those substances may represent rescue options in steroid-refractory courses of disease. However, in order to avoid delay of necessary colectomy only a single attempt at rescue therapy should be considered. Due to the above-mentioned advantages we would favor a serum level controlled cyclosporine rescue therapy once a steroid-refractory course of disease is established. Empirical treatment with metronidazole 500 mg three times daily is indicated if an infectious cause is suspected.68 Due to the inherent risk of colonic perforation many authors advocate the use of antibiotics in toxic megacolon even without evidence for an infectious cause. In patients with toxic megacolon caused by PMC, antibiotics maintaining the infection with C. difficile should be withdrawn immediately. Antibiotics of choice are metronidazole or vancomycin.36 Single case reports describe improvement of toxic megacolon by administration of metronidazole in patients with IBD complicated by C. difficile infection,81 underlining the importance of early recognition and treatment of C. difficile in toxic megacolon. In toxic megacolon caused by CMV or a complicating superinfection with CMV, administration of gancyclovir is indicated. There is a single report of the use of leukocytapheresis in six patients with toxic megacolon complicating UC. Four out of six patients went into the remission stage of UC, circumventing colectomy.82 Yet due to lack of further studies this procedure does not play a role in the recommended management of toxic megacolon.

SURGERY

Previously proposed surgical methods in the treatment of toxic megacolon include resection by either subtotal colectomy with an end ileostomy, total proctocolectomy, or the Turnbull method. In the early 1970s, Turnbull et al. introduced a method of surgical colon decompression. In short, the procedure was intended to lower the perioperative risk of perforation and prepare patients for definitive colectomy at a later point by decompressing the colon at three different locations (blowholes) and performing an ileostomy. Ausch et al38 reported in a retrospective study that in six out of seven patients treated with this method alone, severe postoperative bleeding from the colon occurred; total mortality in those patients was 71%. In the same study 49 patients underwent subtotal colectomy with ileostomy. Mortality in this group was 6%; only four patients (8%) suffered from postoperative complications such as intraabdominal bleeding. In patients treated with total proctocolectomy the postoperative complication rate as well as mortality was 21%. These findings are consistent with previous studies suggesting that subtotal colectomy with an end ileostomy and a Hartmann closure of the rectum is associated with a lower mortality rate as compared to total proctocolectomy.83 In patients with UC as the underlying cause, elective resection of the remaining rectum with an ileal pouch is indicated once the acute condition has been resolved. The sole decompression of the colon as proposed by Turnbull is obsolete. The first-line surgical procedure in the acute phase of toxic megacolon is subtotal colectomy with an ileostomy and either a Hartmann's pouch or sigmoidostomy or rectostomy.38

MEDICAL THERAPY VS. SURGERY

Timing of surgery in toxic megacolon is still a matter of controversy. Circumventing the need for surgery constitutes the principal objective for all medical treatment. However, delay of surgical therapy bears the risk of complications such as bowel perforation or abdominal compartment syndrome entailing a poor prognosis.

Several investigators therefore argue that surgical management should be performed as soon as possible after diagnosis.1, 84 This opinion was underlined in a study by Grant and Dozois,85 which showed that 47% of patients with severe UC had to undergo surgical resection even after initially successful medical treatment. Grant and Dozois85 therefore argued that all medical treatment should be considered solely as a preparation for impending surgery. Goligher et al84 showed in a study in 1970 that mortality could be reduced from 20% to 7% by early surgery. On the other hand, in a study by Katzka et al86 19 patients treated conservatively with steroids and antibiotics improved clinically and showed no signs of perforation; 75% of those patients did not require colectomy in long-term follow-up.

These conflicting results mirror general differences in the attitude toward a medical vs. surgical approach. Absolute indications for colectomy include perforation, uncontrollable bleeding from colon or rectum, and general clinical deterioration. In a more recent study by D'Amico et al87 15 patients underwent surgery (primarily subtotal colectomy with terminal ileostomy) for toxic megacolon as an emergency procedure shortly after diagnosis. During postoperative follow-up, two cases of multiple organ failure occurred, ultimately leading to death and accounting for a mortality rate of 13%. The two cases of fatal complication occurred solely in patients over 65 years of age. The authors therefore argued that early surgery for toxic megacolon may result in a favorable outcome especially in patients under the age of 65. This study, however, is limited by the small number of patients examined and the lack of a control group. The controversy over the appropriate timing of surgery illustrates the importance of interdisciplinary management. Senior surgical consultation should be sought from the very beginning in the course of disease and both continuation of medical therapy and potential indication for colectomy should be evaluated on a daily basis. Signs of complications or deterioration should be an indication for surgical treatment. Toxic megacolon still bears a high morbidity and mortality. Close coordination of both medical and surgical expertise potentially fatal delays in correct treatment can be avoided.