OBJECTIVE. Long-term outcomes of restorative proctocolectomy for pediatric-onset ulcerative colitis are unclear.
METHODS. Questionnaires on health outcomes and quality of life were mailed to patients with childhood-onset ulcerative colitis who had undergone proctocolectomy with ileoanal anastomosis in 2 university hospitals between 1985 and 2005. Investigators not involved in the surgical management of the patients approached participants. Matched control children were randomly chosen from the Population Register Centre of Finland.
RESULTS. Fifty-two (66%) patients and 117 (37%) controls responded. After a mean follow-up of 10 years, at least 1 surgical complication had occurred in 39 (75%) patients, and 28 (54%) had undergone reoperation. Only 1 failure of ileoanal anastomosis occurred. Ulcerative colitis had been reclassified as Crohn disease in 6 (12%) patients. Pouchitis occurred in 37 (73%) patients. The median stool frequency was 5 for day and 1 for night, but 46% used medication to control stool frequency. Nighttime soiling was reported by 56% of the patients. The mean overall quality-of-life score, the mean BMI (22 kg/m2 for both), and the number of subjects (aged >20 years) with offspring (14% vs 15%) was similar to the population-based controls.
CONCLUSIONS. Stool frequency after restorative proctocolectomy in children with ulcerative colitis is stable and comparable to those of adult patients. Although nighttime incontinence is common, general health status and overall quality of life are comparable to the normal population.
An increasing amount of evidence suggests that ulcerative colitis (UC) in children may have a more aggressive disease course than in adults. In children with UC, total colonic involvement is more common than in adults.1 Most pediatric patients with UC need systemic corticosteroids, and rate of proctocolectomy is up to 24% after a median disease duration of 8 years.2 The majority of operations are performed early in the disease course,3 and often more urgently than in adults.4,5 The more aggressive preoperative disease course could affect outcomes and general health after proctocolectomy in patients with childhood-onset UC.
There have bee few studies describing long-term outcomes of proctocolectomy with ileoanal anastomosis in children.4,6–14 The majority of these reports described single-center experiences with relatively short follow-up times primarily concentrating on outcomes of bowel function and surgical complications.7–11,15 Many of the previous studies also included young adults and patients undergoing proctocolectomy for indications other than UC.4,7,10 Therefore, we conducted a questionnaire survey on the long-term outcomes and health status after proctocolectomy in children with UC.
PATIENTS AND METHODS
The hospital records of patients with pediatric-onset (aged <16 years) UC were completely reviewed in 2 major hospitals covering 56% of the total child population in Finland from 1985 to 2005. The diagnosis of UC was based on endoscopic findings and histology in all cases. During this period, 81 children with pediatric-onset UC had undergone proctocolectomy with ileoanal anastomosis at our hospitals. The mean age at the time of proctocolectomy was 11.6 ± 3.8 years. One of the patients had died of an unrelated cause, and another had emigrated. Postal addresses of the remaining 79 patients were traced from the database of the Population Register Centre and were eligible for this study. Data abstracted from the patient charts included age at the time of diagnosis, indication for surgery, preoperative medication, operative details, age at the time of surgery, surgical complications and frequency, diagnostic modalities, and treatment of pouchitis.
Recently, our research group for pediatric inflammatory bowel disease randomly collected a population of 646 subjects from the Population Register Centre of Finland to serve as controls for patients with pediatric-onset inflammatory bowel disease.2 From this control population, a total of 117 eligible controls matched for age, gender, and same municipalities of residence were stratified for this study.
Questionnaires on current health status and bowel function were mailed during 2006. Investigators not involved in the surgical management of the patients mailed the questionnaires and the accompanying letter with information on the study purpose. Information was requested from patients and controls on height, weight, number of offspring, chronic diseases (asthma and liver, kidney, heart, joint, bile duct, and thyroid diseases), abdominal operations, and current medication. Quality of life was assessed in the questionnaires with a visual analog scale (scores 1–7), with 4 generic questions assessing 3 different dimensions of life quality (physical, emotional, and social functioning), and overall quality of life was presented in detail recently.2 In children considered healthy, the mean scores ± SD for these domains were 5.9 ± 1.0, 5.7 ± 1.0, 6.0 ± 0.9, and 6.0 ± 1.0, respectively.2
In the patient group, we also asked about the initial and current diagnosis, type of surgery, complications, reoperations, medical therapy, satisfaction with surgery, and restrictions of life resulting from proctocolectomy. Questions concerning bowel function included stool frequency, soiling, stool consistency, urgency, discrimination of flatus, evacuation, medications to control stool frequency, and occurrence and treatment of pouchitis. Soiling was defined as any smearing or staining of underwear during the last 3 months.
Proctocolectomy included transanal mucosectomy and a hand-sewn ileoanal anastomosis in every case.
Unless otherwise stated, the data are presented as mean ± SD. The Mann-Whitney U[r] test (2-sample rank-sum test) was used to compare continuous variables between groups. Frequencies were compared with χ2 tests. Frequencies of chronic diseases were compared with Fisher's exact test because of the small number of cases. Correlations were tested with a simple regression analysis.
The ethical committees of Tampere and Helsinki University hospitals approved the study protocol.
The response rate was 66% (52 of 79) among the patients and 37% among the controls.2 A total of 58% of proctocolectomies were performed in Helsinki and 42% in Tampere. Demographic data of the patients and the controls are displayed in Table 1. Gender distribution, age, weight, height, and BMI were similar between the patients and the controls.
The item response rate was high in both groups. On average there was <2% missing values for each question in both groups. To assess possible selection bias between responders and nonresponders, a drop-out analysis was performed. Disease duration before proctocolectomy, age at operation, and time elapsed after proctocolectomy were similar between responders and nonresponders (2.3 ± 1.9 vs 2.5 ± 1.9, 12 ± 3.8 vs 12 ± 3.3 years, and 10 ± 5.7 vs 11 ± 4.5 years, respectively). Those patients who completed the questionnaires were more frequently female (60% vs 38%, respectively; P = .11).
Preoperative Disease Characteristics
Preoperatively, all patients had been treated with systemic glucocorticoids and 5-aminosalicylic acid (ASA) preparations. Immunosuppressive medication including cyclosporine, methotrexate, or infliximab had been prescribed to 11 patients (21%). Indications for surgery were persisting disease activity and symptoms despite optimal medical therapy, cortisone dependence, or adverse effects of medical therapy in 39 (75%) patients. Fulminant colitis was an indication for surgery in 13 (25%) patients.
Of the 52 patients, 42 underwent J-pouch ileoanal anastomosis, 1 had H-pouch ileoanal anastomosis, and 9 patients had straight ileoanal anastomosis. A covering ileostomy was used on 42 (81%) occasions.
There was no operation-related mortality. The incidence of complications is shown in Table 2. Three fourths of the patients had at least 1 complication. A total of 48 episodes of adhesive bowel obstructions in 26 patients were recorded. Of these, 17 (35%) required operative intervention. Anastomotic stenosis responded to a single dilatation in most cases. Only 2 patients with an anastomotic stricture required surgical intervention. Anastomotic leakage occurred in 7 (13%) patients, and it was often associated with fistula formation or abscess (Table 2). Septic pelvic complications necessitated temporary fecal diversion with ileostomy in 7 (13%) patients. Anastomotic leakage tended to occur more often when ileoanal anostomosis was performed without (3 of 10) than with (4 of 42) a covering ileostomy (P = .088). Overall, >50% of the patients had a reoperation because of complications (Table 2).
Inflammatory bowel disease had been reclassified from UC to Crohn disease in 6 (12%) patients. The diagnosis of Crohn disease was based on histology in 3 patients and capsule endoscopy findings in 2 patients. The median time period between proctocolectomy and the diagnosis of Crohn disease was 20 months (range: 2–133 months).
Only 1 pouch failure occurred. This patient was converted to permanent ileostomy because of Crohn disease and enterovaginal fistula. The rest of the patients (98%) had a functional ileoanal anastomosis.
Overall, 73% (37 of 51) of the patients had at least 1 episode of pouchitis, and 63% (31 of 51) of the patients reported more than 1 episode. The diagnosis of pouchitis was confirmed with endoscopy and biopsies at least once in 84% (31 of 37) of the patients. For the rest of the patients the diagnosis was based on clinical symptoms of increased frequency of bowel movements and abdominal discomfort with or without fever. Pouchitis had been treated with antibiotics in all 37 patients most commonly with metronidazole or ciprofloxacine. The incidence of pouchitis was more common after pouch (33 of 42) than after straight (4 of 9) ileoanal anstomosis (P < .05).
Outcome of Bowel Function
The data on stool frequency and fecal continence are summarized in Table 3. Stool frequencies and incidence of nighttime soiling were higher after straight than after pouch ileoanal anastomosis, although the difference reached a statistically significant level only for daytime stool frequency. In the whole series, any degree of daytime or nighttime soiling was reported by 22% and 56% of the patients, respectively. Daytime and nighttime soiling occurred less often than 3 times per week in 96% and 80% of the patients, respectively. Overall, 44% of the patients were totally continent without any degree of soiling ever. Nearly half (46%) of the patients reported use of medication to control frequency of bowel movements, loperamide being the most commonly used preparation.
The data on stool consistency and defecation are shown in Table 4. The results were similar for patients with straight and pouch ileoanal anastomosis. Overall, only 8% of the patients reported watery stools. Nearly 95% had an urgency period longer than 15 minutes, and most were able to discriminate flatus from feces at least partly.
Among the patients with functional pouch ileoanal anastomosis (n = 41), daytime stool frequency correlated positively (R = .33; P = .04) with the age at the proctocolectomy but was unrelated to postoperative follow-up time (R = .20; P = .21) or the current age of the patients (R = .02; P = .89). Thus, young age at proctocolectomy was associated with a decreased frequency of bowel movements.
Quality of Life
The different dimensions and overall quality-of-life scores were similar between the patients and the controls (Table 5). Overall, 82% of the patients were completely or moderately satisfied with the outcome of the surgery. Only 1 patient would have refused to recommend the operation to others with UC. Surgery had delayed education in 29% of the patients and interfered with sport activities in 25% of the patients. Proctocolectomy-related bowel problems had caused an absence from school or work for 54% of the patients at any time point after the operation. However, 90% of the patients reported absences less often than once per month (or never).
Overall quality of life was inversely related to stool frequency (R = .42; P = .002). Thus, increased frequency of bowel motions was associated with decreased quality of life. The overall quality-of-life score was significantly decreased among patients with any surgical complication (6.3 ± 0.6 vs 5.4 ± 1.2, respectively; P < .05). In addition, the patients who had missed school or work because of proctocolectomy-related bowel problems reported decreased overall quality of life (6.1 ± 0.7 vs 5.3 ± 1.3, respectively; P < .05). The overall quality-of-life score was 6.1 ± 0.8 among totally continent patients in relation to 5.4 ± 1.3 among patients with any degree of daytime or nighttime soiling (P = .065).
There were no differences in the number of children between patients and controls. Of the 29 (56%) patients and 68 (58%) controls born before 1986 (aged >20 years), 4 (14%) and 10 (15%) had children, respectively. Among the patients a total of 15% (3 of 20) of the women born before 1986 had children. The respective figure was 16% (7 of 44) in the control group.
Joint and biliary tract diseases occurred more frequently among patients than in the controls. The frequency of both joint and biliary tract disease was 5.9% (3 of 51) among patients, whereas no cases were reported by the controls (P < .05, Fisher's exact test). Of the 3 patients with biliary tract disease, 2 had sclerosing cholangitis and 1 had biliary stone disease. Asthma was equally common among the patients (7.8% [4 of 51]) and controls (6.8% [8 of 117]).
The disease course of UC may be more severe in children than in adults.1–4 Although this may modify outcomes and general health after proctocolectomy, only a few long-term follow-up studies are available regarding children.4,6,8,11,12,15 In this study we assessed, in a population-based manner, long-term outcomes of proctocolectomy in children with a confirmed diagnosis of UC, including general health status and quality of life. Investigators not involved with surgical management approached the patients to reassure truthful description of outcome variables essentially related to surgery. At the time of the study patients were in early adulthood, and they had undergone proctocolectomy with ileanal anastomosis an average of 10 years earlier. Our findings suggest that eminent preoperative disease activity in childhood-onset UC is associated with a high complication rate after proctocolectomy. Despite frequent complications, outcome of bowel function and continence is comparable with that in adult series.5,16,17 Excluding increased occurrence of joint and biliary tract diseases, general health status and quality of life of patients who have undergone proctocolectomy in their childhood is comparable with those of the normal population in adulthood.
Our findings support the concept of aggressive disease course in childhood-onset UC.2 Preoperatively, all children had been on systemic corticosteroids with or without immunosuppressive therapy. Fulminant colitis necessitated urgent surgery in 25% of the children. The proportion of adults undergoing emergency surgery for UC is several times lower.5
The overall rate of surgical complications after proctocolectomy and ileoanal anastomosis was high (75%) in the present series in relation to figures reported previously ranging from 30% to 53% after follow-up times up to 5.8 years.4,6,11–13 Extending the follow-up time to 10 years resulted in an increased rate of adhesive bowel obstruction, which accounted for more than half of all complications and reoperations. The risk of bowel obstruction increases together with postoperative follow-up time after proctocolectomy also in adults.16 The frequency of anastomotic leakage (13%) and other related pelvic septic complications were well in the range of the previous reports in children6,11 but slightly higher than in adult series.5 A tendency toward more frequent anastomotic leakage among patients operated on without an ileostomy may support the use of a covering ostomy. However, a high rate of pouch salvage was achieved after pelvic septic complications reflected by a low (2%) overall rate of pouch failure.
Inflammatory bowel disease was reclassified from UC to Crohn disease for 12% of the patients. The respective figure ranges from 2% to 15% in previous studies in children4,6,11–13 as opposed to a figure of 0.02% in a large adult population after follow-up of 20 years.17 In children, Crohn disease often presents as colitis without small intestinal involvement, which makes differential diagnosis between Crohn disease and UC challenging. Whether novel serologic markers18 and routine use of preoperative capsule endoscopy of the small intestine will improve the diagnostic accuracy among children remains to be proven.
Overall, 73% of the patients had been treated for pouchitis, this rate being higher than previously reported in children4,6,8–14 but not in adults.17 In the majority (84%), pouchitis was confirmed with endoscopy and histology at least once, and it occurred more often after pouch than early straight ileoanal anastomosis. The high incidence of pouchitis was most likely associated with extended follow-up time relative to previous studies in children. Cumulative frequency of pouchitis is positively related to length of the follow-up in adult patients with UC.16
Our data show that despite the high incidence of surgical complications, outcome of bowel function and stool frequency remain constant in the first decade after proctocolectomy in children. The mean daytime stool frequency was 5.9 (median: 5), and the mean nighttime stool frequency was 1.4 (median: 1). These results are comparable with previous findings in children4,7,9,10–13,15 and adults.5,16,19 Young age at proctocolectomy was associated with decreased frequency of bowel motions. This may reflect the better adaptive ability of younger children and supports prompt surgical treatment as soon as medical management fails irrespective of the age of a patient. Although a majority (78%) of the patients were continent during the days, the proportion of totally continent patients was relatively low (44%). Previous studies in children have reported similar continence rates of 56% to 87% during daytime and 44% to 72% during nighttime.7,9,12,13 In this study any smearing or staining of underwear was considered soiling. In line with previous studies,10,12,13 the main shortcoming related to defecation was frequently impaired ability to distinguish flatus from stools. A vast majority of the patients were able to postpone defecation sufficiently to meet social needs.
Our data suggest that proctocolectomy in children with UC provides good quality of life. Unexpectedly, their mean adult height and weight were not statistically significantly reduced. The patients graded their emotional, physical, and overall well-being as well as social functioning to levels similar to those of their matched population-based controls. The quality-of-life dimensions were assessed with a visual analog scale that focused on physical, emotional, and social functioning and quality of life during the past week as recently described by us in detail.2 This is a simple method including 1 item per domain but discriminated those with a more complicated postoperative outcome. Despite good self-perception scores and overall satisfaction, surgery had restricted education, work, and sport activities in many in line with a previous study.12 Although there have been several recent studies on quality of life in pediatric patients with irritable bowel disease,20,21 we are not aware of such reports focusing on those with proctocolectomy. Our recent study showed decreased overall quality of life among patients with childhood-onset inflammatory bowel disease in general, but the patients with proctocolectomy were not stratified.2
Proctocolectomy may cause subfertility.22 Because of the young age of the patients, only a few had children. The proportions of girls and boys with offspring were similar to those among matched controls. Although the sample size was relatively small, these findings do not support subfertility in early adulthood. Of the extraintestinal manifestations of UC, joint and biliary tract diseases were more common in the patients than in the controls, as expected.2
The response rate for this study was 66% for the patients. Other questionnaire surveys have reported similar proportions of responders.23 On the basis of a drop-out analysis, nonresponders and responders had equal disease duration, operation age, and postoperative follow-up period, which makes significant selection bias among patients very unlikely. Girls responded more often both among patients and controls, as expected.24 It is remarked that the population-based controls were stratified from a large group included in our previous study.2 Although the response rate of the controls was disappointingly low (37%), the health status of the controls assessed by asthma frequency was similar to published figures in Finland, which justifies the representativeness of the control group.2
This work was supported by a Helsinki University Central Hospital grant, the Finnish Pediatric Research Foundation, the Päivikki and Sakari Sohlberg Foundation, and the Tampere University Research Fund.
We thank Ms Sari Honkanen, Ms Minna Kuurne-Koivisto, Ms Anne Nikkonen, and Ms Leila Uusimaa for excellent assistance.
- Accepted September 9, 2008.
- Address correspondence to Mikko P. Pakarinen, MD, PhD, Hospital for Children and Adolescents, University of Helsinki, PO Box 281, FIN-00029 HUS, Helsinki, Finland. E-mail
The authors have indicated they have no financial relationships relevant to this article to disclose.
What's Known on This Subject
Restorative proctocolectomy with ileoanal anastomosis is generally accepted treatment for pediatric-onset UC when medical management fails. Long-term outcomes of restorative proctocolectomy for pediatric-onset UC are unclear.
What This Study Adds
Stool frequency after restorative proctocolectomy in children with UC is stable and comparable to those of adult patients. Although nighttime incontinence and surgical complications are common, general health status and overall quality of life are comparable to the normal population.
- ↵Griffiths AM, Buller HB. Inflammatory bowel disease. In: Walker WA, Durie PR, Hamilton JR, Walker-Smith JA, Watkins JB, eds. Pediatric Gastrointestinal Diseases. 3rd ed. Ontario, British Columbia, Canada: Decker Inc; 2000:613–652
- ↵Hommel KA, Davis CM, Baldassano RN. Medication adherence and quality of life in pediatric inflammatory bowel disease. J Pediatr Psychol.2008;33 (8):867– 874
- ↵Eaker S, Bergström B, Bergström A, Adami HO, Nyren O. Response rate to mailed epidemiologic questionnaires: a population-based randomized trial of variations in design and mailing routines. Am J Epidemiol.1998;147 (1):74– 78
- Copyright © 2009 by the American Academy of Pediatrics