|Home | About | Journals | Submit | Contact Us | Français|
A 20 year old woman with a history of ulcerative colitis is referred for symptoms of frequent bloody stools, tenesmus and the inability to taper off of corticosteroids. She was initially diagnosed with left sided ulcerative colitis at the age of 8 and had an average of one moderate flare per year that responded to an oral steroid taper. She had a normal terminal ileum on previous colonoscopy and a normal small bowel follow-through. Over the past 2 years, her disease has become refractory to medical therapies. During this time period, she has been on multiple medications including oral mesalamine, topical rectal corticosteroid and mesalamine preparations, and oral corticosteroids (10-60 mg prednisone). Infliximab was initiated 2 years prior at 5 mg/kg and titrated to 10 mg/kg over the past 18 months. Her last dose of infliximab was 4 months before this evaluation when it was discontinued due to lack of response.
Her current symptoms consist of 10-15 stools per day associated with mucous and blood. She has intermittent moderate left lower quadrant abdominal pain. She has significant tenesmus. Over the past 6 months, she has lost approximately 15 lbs. She denies any skin rashes or joint symptoms. She has not had any fevers or night sweats. Her current medications consist of prednisone 60 mg daily (for the past 12 days), mesalamine suppository 1 g per rectum at bedtime, oral mesalamine 3.6 g daily and hydrocortisone 10% rectal foam. The patient is a non-smoker, she is married and would like to start a family in the next 2-3 years.
She is 5 feet 4 inches tall and weighs 66 kg. Her heart rate is 98 with a blood pressure of 118/76. She is moderately tender to palpation in the left lower quadrant. There is no perianal disease. She has an elevated ESR of 45 and a hemoglobin of 10.2. Stool studies are negative for enteric pathogens, Clostridium difficile toxin A and B, and ova and parasites. A plain abdominal film shows no evidence of colonic dilatation. On flexible sigmoidoscopy, there is a contiguous area of severe erythema, granularity, friability and contact bleeding from the rectum to the mid-transverse colon. After the mid-transverse colon, the mucosa appears normal. Biopsies from the involved area showed moderate to severe chronic active colitis without dysplasia. Biopsies were negative for cytomegalovirus (CMV). She is hospitalized and initiated on a continuous infusion of intravenous corticosteroids at 2 mg/hour (48 mg/day). Over 5 days, there is no change in her bowel movement frequency or bleeding.
The lifetime risk of a severe exacerbation of ulcerative colitis requiring hospitalization is 15%.1 Patients with extensive disease (macroscopic disease proximal to the splenic flexure) are more likely to develop acute severe colitis, however it also occurs in patients with left-sided disease. Determination of disease activity in ulcerative colitis is based on clinical findings (both objective and subjective) and visual assessments of inflammation (endoscopy, imaging). Activity is described in terms of both severity and extent of the disease 2, 3 Instruments for measuring disease activity include the Truelove and Witts Severity Index (Table 1), the Mayo Score (also called the Disease Activity Index) and the Lichtiger Index which was developed specifically for severe ulcerative colitis.4
In a patient with hospitalized severe ulcerative colitis that is refractory to maximal therapy with intravenous corticosteroids, oral aminosalicylates and topical medications, the question of “salvage” medical versus surgical therapy is paramount. Medical therapy in a hospitalized patient requires close monitoring and continual reassessment, as in a severely active patient surgery may in fact be life-saving. In patients with refractory hospitalized ulcerative colitis, it is difficult to know with certainty which patients will respond to intravenous corticosteroids. Certain characteristics have correlated with corticosteroid failure in previous studies. These include frequent bowel movements, low albumin and colonic dilatation on imaging in spite of treatment with intravenous steroids. After three days of intravenous steroids, 85% of patients with >8 stools per day or a stool frequency between 3 and 8 with a CRP >45 mg/l will require colectomy.5 Other prognostic indicators for corticosteroid failure include elevated maximum daily temperature, pulse, stool frequency >3 and low albumin.6 These criteria are often used clinically to expedite escalation to more aggressive medical or surgical treatment earlier in the hospital course.
To decide on the appropriate course of therapy for this patient, the outcomes and complications of medical and surgical therapy for severe ulcerative colitis in a young woman need to be considered. Despite the introduction of new medications for the treatment of ulcerative colitis, the overall colectomy rate has remained stable at approximately 27% over the past 3 decades.7 However, these data do not consider the introduction of infliximab for UC which shows promise for reducing colectomy rates. Issues particular to young women, including fecundity (the likelihood of conceiving as a function of duration of time), should also be considered in the discussion of medical versus surgical therapy.
IBD results from a genetic susceptibility to immune mediated bowel injury triggered by environmental stimuli. Intestinal inflammation is likely due to an overly aggressive immune response to luminal bacteria. Recent genetic findings highlight the important role of innate immunity against enteric microbes in disease pathogenesis. Ultimately, aberrant innate immune responses to enteric microbes initiate and perpetuate an exaggerated T cell-driven adaptive immune response. Increased T cell migration, retention, activation and proliferation in the colon contribute to uncontrolled inflammation. Mechanisms of tissue injury are related to the secretion of inflammatory cytokines such as TNF, IL-12/23, interferon-γ, IL-5, IL-13, and IL-17. In contemplating current and future therapeutic interventions in UC, clinically validated targets include inhibition of the activated T cell response (for example, cyclosporine), blockade of trafficking of inflammatory cells to the intestine (for example, natalizumab), and inhibition of inflammatory cytokines released from activated cells like TNF (for example, infliximab).
In UC, there are no validated serologic or genetic markers that predict who will develop a severe exacerbation. Toll-like receptor-2 (TLR2)8 genetic polymorphisms seem to be more common among patients with severe steroid resistant UC. The discovery and validation of biomarkers that predict disease course and response to therapy in UC is an important priority for future research.
An important medical treatment for severe UC that has failed outpatient management is intravenous corticosteroids. Adjunctive therapies may include oral mesalamine, topical mesalamine and corticosteroids, antibiotics, TPN, and immunosuppresants such as 6-mercatopurine (6-MP) and azathioprine (AZA). There are few randomized placebo controlled studies of these agents in severe UC. Intravenous corticosteroids have been used for the treatment of hospitalized UC patients since Truelove's description in the 1950's.3 Truelove demonstrated a 73% remission rate (36/49) with a 5 day course of intravenous corticosteroids in patients with severe UC.9 There is no evidence that intermittent versus continuous intravenous steroid infusion changes outcomes.10 Antibiotics demonstrated no additional benefit when combined with intravenous steroid administration.11-13 Likewise, total parenteral nutrition and complete bowel rest in combination with corticosteroids showed no additional benefit.14-16 Due to delayed onset of action, 6-MP and AZA are not used in the acute management of severe UC. These medications can be used for long-term maintenance if remission can be achieved by other means. When patient hospitalized with severe UC do not respond to 3-5 days of intravenous corticosteroids, salvage therapies such as cyclosporine, infliximab, and surgery become necessary. Clinical and laboratory criteria are often used to make decisions about defining early corticosteroid failure at day 3.5, 17
Cyclosporine is an inhibitor of the transcription factor nuclear factor of activated T cells (NFAT) and acts in part by attenuating production of interleukin-2 by activated T cells. Cyclosporine is effective in severe UC to induce rapid response during hospitalization. Initial response rates range from 62-86%, defined as improvement in clinical symptoms and initial avoidance of colectomy.18-23 Early studies used doses of cyclosporine of 4 mg/kg/day, but 2 mg/kg/day dosing is equally effective with a better side effect profile.24 Long term clinical success with cyclosporine depends upon initiation of 6-MP or AZA as a maintenance agent. Among those who initially responded to cyclosporine and were newly initiated on 6-MP, 20% required colectomy over a 5 year follow up period. Those who were not initiated on 6-MP or previously failed 6-MP had a 45% colectomy rate.19 Likewise, in patients treated with cyclosporine and AZA, 80% were free of colectomy at 1 year and 60% were colectomy-free at 7 years.20 In patients on cyclosporine alone, 47% were free of colectomy at 1 year, but only 15% were colectomy-free at 7 years. These data demonstrate the importance of an exit strategy after initiation of cyclosporine due to short and long term toxicities. Major adverse events that have been associated with cyclosporine use include: nephrotoxicity (5.4%), serious infection (6.3%), seizures (3.6%), anaphylaxis (0.9%), death (1.8%).25 Other minor adverse events include paresthesias, hypomagnesmia, hypertension, hypertrichosis, headache, minor infections, hyperkalemia and gingival swelling.25
Cyclosporine is generally used as salvage treatment prior to colectomy. There is no increased operative mortality associated with pre-surgical cyclosporine in patients undergoing restorative proctocolectomy. Overall, a 36% post-operative complication rate has been reported with cyclosporine26, similar to the historical complication rate for patients undergoing colectomy not treated with cyclosporine.
Practical considerations with cyclosporine include close monitoring of laboratory values and adjustment of the infusion rate. The initial dose is 2 mg/kg/day. The target cyclosporine level is between 150 and 250 ng/ml (by monoclonal assay, target levels will vary with polyclonal assays) for the intravenous dose.
Conversion to oral dosing is usually twice the intravenous dose (given q 12 hours), with a trough level adjusted between 100 and 200 ng/ml. An exit strategy with the initiation of 6MP or AZA is necessary. Prophylaxis for pneumocystis carinii (PCP) should be initiated during the hospitalization and continued as an outpatient. In summary, for severe UC, cyclosporine induces a prompt response and may result in prolonged response if 6-MP or AZA can be initiated.
Infliximab is a chimeric IgG1 monoclonal antibody that neutralizes the biologic activity of the inflammatory cytokine TNF. Infliximab is FDA approved to induce and maintain remission in patients with moderate to severe UC. Sands et al demonstrated a 50% clinical response rate to infliximab in a trial of hospitalized patients with UC (8 patients randomized to infliximab, 3 patients randomized to placebo) failing intravenous steroids.27 Jarnerot et al randomized a cohort of 45 patients with severe hospitalized UC to rescue therapy with infliximab versus placebo after 3-8 days of intravenous steroids. The primary end point was colectomy within 3 months. The colectomy rate in the placebo group at 3 months was 67% versus 29% in the infliximab group (OR 4.9 (1.4 - 17); p=0.017)17. Upon entry, patients were classified as “severe” and “fulminant”. Fulminant patients were randomized on day 4, and severe patients on day 6 to 8. When the fulminant patients were separately analyzed, 69% (9/13) in the placebo group required colectomy compared to 47% (7/15) in the infliximab group. (p = 0.3). Thus, there was a trend toward decreased colectomy rates in the most severe patients randomized to infliximab, but this was not statistically significant, possibly due to a small sample size and a type II error.
In patients who fail infliximab and progress to surgery, there are mixed data on associated complication rates. Kunitake also found no increase in postoperative complications associated with infliximab use.28 However, there may be an increased rate of postoperative complications among infliximab treated patients undergoing restorative proctocolectomy with ileal pouch anal anastomosis (IPAA) compared to controls (OR 3.54, 95% CI 1.51-8.31 for early complications).29 In a prospective series, although overall morbidity was similar between patients who received pre-IPAA infliximab and those who did not, anastomotic leaks (p = 0.02), pouch-specific (p = 0.01) and infectious (p < 0.01) complications were more common in the infliximab group. 30
There is no role for the combination of cyclosporine and infliximab in steroid refractory UC. The combination of these medications is associated with a higher morbidity. There is an 80% rate of surgical and infectious complications associated with concomitant use of both medications.31 Additionally, when cyclosporine and infliximab are used as rescue therapy for patients failing the other modality, the risks likely outweigh the benefits. Maser et al showed approximately one third of patients receiving salvage therapy by crossing over to the alternate medication were in remission at one year, however serious side effects (including 1 death) occurred in 16% of patients.32 An overview of the differences between cyclosporine and infliximab in the treatment of severe UC is demonstrated in Table 2.
Emergency indications for surgical intervention in severe UC include free perforation, hemorrhage or systemic instability. An urgent indication for colectomy is a severe attack that is unresponsive to medical therapy. “Unresponsive to medical therapy” is a subjective term, not precisely delineated in the literature for severe UC. However, the important point is that delaying surgery in an unstable or clinically deteriorating patient can lead to dire consequences. In our practice, the development of medical instability, worsening symptoms, or the lack of partial improvement over the first 3 to 5 days of salvage therapy should lead to serious consideration of immediate surgical intervention.
In the setting of severe UC, the procedure of choice is subtotal colectomy and ileostomy. The residual rectal disease is controllable in most patients. In general, there are advantages to the subtotal colectomy approach, including a lower morbidity if pelvic dissection is not performed, preservation of the rectum so that reconstructive procedures can be performed later, and allowing the definitive procedure to be deferred to an optimal situation when the patient is off immunosuppressive medications and has improved nutritional status. Usually, the staged reconstruction with IPAA or definitive total proctocolectomy is performed several months later.
There can be surgical morbidity associated with colectomy, although these risks are minimal compared to the risks associated with continuing futile medical therapy in an unstable patient. Loftus et al33 evaluated outcomes post-colectomy in a large population of patients with UC (n=25,586). In the 180 days post-colectomy, 54% of patients required a second operative procedure (including planned staged operations, 15.3% of patients had unplanned operations). Complications occurring post-colectomy included: abscess (11.6% early; 16.3% late), fistulas (4.2% early; 6.0% late) and bacteremia/sepsis (7.9% early; 9.3% late). In a retrospective cohort study (n=184), Ferrante et al34 found a 27% early postoperative complication rate for patients with UC undergoing colectomy. Over a median follow-up of 6.5 years, 35% of patients developed septic and/or obstructive complications.
Generally, patients have good outcomes after restorative IPAA. Median bowel frequency is 6 stools per day. Post-procedure symptoms can include: intermittent urgency (23% of patients), evacuation difficulties (12% of patients), and intermittent soiling at night (52% of males and 32% of females). Overall, greater than 90% of patients are satisfied with their pouch.35 There is a small rate (6.2%) of pouch failure over time.36 For women, there is an additional concern of reduced fecundity (likelihood of conceiving as a function of duration of time) post-IPAA. After colectomy and restorative IPAA, the fecundity ratio is significantly impaired (fecundity ratio 0.25 (0.17-0.37)) compared to the reference population.37 These effects are likely related to the surgery itself, and are seen in other populations undergoing IPAA.38 For this reason, in women of childbearing age, it may be reasonable to delay restorative IPAA until after childbearing years.
There are currently no reliable clinical markers or biomarkers that can be used to determine the likelihood of success for cyclosporine versus infliximab for an individual patient. Reliable predictors of response to these medications are needed to better avoid complications associated with delaying colectomy. There are two controlled trials in progress in Europe comparing cyclosporine to infliximab as rescue therapy in severe ulcerative colitis. These data will help to guide our choice of therapies in the future.
As this case illustrates, the decision of whether to use infliximab or cyclosporine for severe UC in the hospitalized patient is perhaps no longer the most relevant clinical scenario. Often, these patients have already been exposed to infliximab in the outpatient setting. The most appropriate use of infliximab may in fact be to avoid hospitalization, intravenous corticosteroids, and surgery in the medically refractory outpatient. There are major gaps in clinical information on how to proceed in the setting of prior infliximab exposure as an outpatient in hospitalized UC patients. There are no data about response to intravenous corticosteroids or cyclosporine in UC patients who are exposed (and didn't respond) to infliximab as an outpatient. In the authors’ practice, intravenous steroids and cyclosporine are contemplated in patients if they had not been previously refractory to 6-MP/AZA. In patients who have previously failed 6-MP/AZA, surgery must be considered before intravenous steroids and cyclosporine as these patients are less likely to maintain remission long-term.
The only published United States guidelines for the management of severe UC are from the American College of Gastroenterology. The Ulcerative Colitis Practice Guidelines in Adults (ACG 2004) recommend hospitalization for a course of intravenous steroids in those refractory to oral treatment with prednisone, mesalamine and topical medications, or in the patient who presents with toxicity. These guidelines were formulated prior to the published experiences with infliximab in severe UC, and need to be updated to reflect this medical option. Stool samples should be evaluated for superimposed infection with enteric pathogens and C. difficile. There is no role for treatment with antibiotics, although these are used in patients with signs of toxicity. There is no benefit to total parenteral nutrition as a primary therapy for UC. Improved nutrition with total parenteral nutrition may be beneficial as an adjunct to enteral nutrition in patients with significant nutritional depletion. A flexible sigmoidoscopy with biopsies to exclude superimposed CMV should be considered in patients not responding to maximal immunosuppressive therapy. Failure to demonstrate significant improvement within 3-5 days is an indication for colectomy or intravenous cyclosporine. Long term remission is enhanced in patients initiated on cyclosporine with 6-MP or AZA as a maintenance strategy.
Recently, the Association of Coloproctology of Great Britain and Ireland published a position statement on the management of acute severe colitis 39. Recommendations about diagnosis, monitoring, and treatment are consistent with the 2004 ACG recommendations, but also account for the use of infliximab. The necessity of joint care between an experienced gastroenterologist and a colorectal surgeon to assure optimal outcome is emphasized. Intravenous corticosteroids remain the first line medical approach in acute severe colitis, and 3 to 5 days of treatment is recommended prior to initiation of other options. A single attempt at rescue therapy with cyclosporine or infliximab could be considered in selected patients prior to colectomy, but the choice of agent is limited by the absence of adequately powered randomized controlled trials. For patients with toxic megacolon, 24 hours of intensive medical therapy is recommended and continued only if there is radiological and clinical evidence of improvement. Unresponsiveness to medical therapy is the most common indication for surgery, but is difficult to precisely define. As a general guideline for patients with severe colitis without toxic megacolon, after 5-7 days of rescue therapy, surgical treatment is indicated if there is no clinical improvement. The most problematic patients are those who respond partially to medical therapy as there is no data to adequately weigh the risks of continued medical therapy versus surgical intervention. Based on these guidelines and the present literature review, an outline for management of poorly responsive ulcerative colitis in the hospital is depicted in Figure 1.
The patient initially failed to respond to high dose oral prednisone during a 1-2 week period. Appropriate medical decisions at this juncture would be an outpatient trial of infliximab or admission to the hospital for intravenous steroids and other treatment modalities. There is no clear medical evidence to guide when within this 2 week window a patient should be considered to an oral corticosteroid failure, but severe symptoms, systemic symptoms, progression of disease activity may prompt an earlier decision to escalate therapy; while a partial response to oral prednisone may be an indication to exercise patience clinically.
The patient then did not respond to intravenous steroids over a course of 5 days. Serial abdominal exams remained benign and abdominal x-rays did not show colonic dilatation. She was initiated on cyclosporine and underwent a 5 day course with dramatic improvement in her stool frequency, pain and ESR. Thiopurine s-methyltransferase activity was normal. She was initiated on 6-MP at a dose of 1.5 mg/kg at discharge from the hospital. Prior to discharge, she was transitioned to oral cyclosporine at a dose of 150 mg twice daily (approximately twice the intravenous dose she was receiving). At discharge, she had 2-3 bowel movements per day without blood. She was placed on an oral prednisone taper and maintained on topical and oral mesalamine medications. She was also started on sulfamethoxazole and trimethoprim three times weekly for Pneumocystis pneumonia prophylaxis. At follow up 8 weeks post discharge, the patient was still doing well, in remission with 2-3 formed bowel movements daily. She had completely tapered off of oral prednisone. After 12 weeks, she was tapered off her cyclosporine and is currently doing well 10 months later on 6-MP, oral and topical mesalamine.
FINANCIAL DISCLOSURES: Scott Plevy discloses financial relationships with the following companies: Centocor Inc. (Advisory Boards, Speakers Bureau, Grant support), Abbott Laboratories (Advisory Boards, Speakers Bureau, Grant support), UCB Biopharma (Advisory Boards, Speakers Bureau), Elan Pharmaceuticals ((Advisory Boards, Speakers Bureau).
Millie Long – No conflicts of interest exist.