Crohn disease (CD) is an idiopathic inflammatory bowel disease (IBD). Inflammation is mucosal and may extend transmurally and affect the entire digestive tract in a generalized or patchy manner. At the time of the diagnosis, the disease distribution is more extensive in young patients with CD compared with adults (1). Between 9.5% and 40% children with CD experience growth failure during disease course (2,3), which can already be present at the time of diagnosis (3,4). Within 10 years, 20% to 30% of children undergo surgery (5,6). Indications for surgery are variable: unsuccessful medical therapy, growth failure, or complications such as stricture, perforation, abscess, fistula, or severe perianal disease (7). The use of infliximab or 5-aminosalicylic acid may decrease the need for an operation (8,9), but further studies are needed. Studies with a follow-up ranging from 2.5 to 9.7 years (5,10–15) show that relapses are common, but their frequency can vary a great deal depending on the definition (clinical, radiological, endoscopic, surgical) and the duration of the follow-up (10). It is well known that surgery cannot cure the disease, but it may provide a disease-free interval for normal growth and development, if done before late puberty (2,10,16). Regardless of the high relapse rate, nearly all children reach catch-up growth after surgery (5,10–12,17,18).
It is clear that patients who undergo surgery in childhood experience a more aggressive disease, and presently it is impossible to predict the patient's future with the disease. In children, CD often presents as pancolitis (19), which makes a differential diagnosis of ulcerative colitis (UC) difficult and misdiagnoses are common (19–24). Of patients with a primary diagnosis of UC who have undergone proctocolectomy with ileal pouch anal anastomosis formation, 13% eventually were diagnosed as having CD a mean of 2 years after the primary operation (25). In patients with pediatric-onset CD, the effects of small bowel resection or colectomy on quality of life (QoL), later development, and health have not been systematically studied. The functional outcome and degree of bowel symptoms after surgery are unclear. Therefore, we conducted a study on long-term health outcomes and QoL in children diagnosed as having CD and with a median follow-up of 10 years after bowel surgery.
The hospital records of patients with pediatric-onset (16 years or younger) CD who had undergone surgery (bowel resection only) during childhood (18 years or younger) between 1985 and 2008 were reviewed in 2 major tertiary care hospitals in Helsinki and Tampere (covering 56% of all children in Finland). Patients with less than 2 years of follow-up were excluded. During this period, 36 such children (24 girls) had undergone bowel resection (82% in Helsinki).
Questionnaires were mailed out in spring 2010. The postal addresses of all eligible patients included in the database of the Population Register Centre were traced and recovered for the study. During the first stage, questionnaires were mailed to the patients. The nonrespondents received a second mailing. The questions included present health status, bowel function (stool frequency, soiling, stool consistency, urgency, discrimination of flatus, evacuation, medications to control stool frequency), occurrence and management of pouchitis (proctocolectomized patients with an ileal pouch anal anastomosis), satisfaction with surgery, and any restrictions resulting from the surgery. Soiling was defined as any smearing or staining of the underwear during the last 3 months before the questionnaire. QoL was assessed in the questionnaire using a visual analogue scale (scores 1–7), with 4 generic questions assessing 3 dimensions of QoL (physical, emotional, and social functioning) and overall QoL. High scores indicated better QoL, as described by Turunen et al and Pakarinen et al (6,24).
Previously, we collected a control population with no IBD (n = 1650) from the Population Register Centre of Finland to serve as controls in our recent study concerning long-term follow-up of young patients with UC (24). From this control population, we selected 2 controls per patient with CD matched for age and sex (n = 82, with no IBD) and compared their life situation with that of patients with CD.
Preoperative Disease Characteristics and Follow-up
The medication before surgery, postoperative use of biologicals, indications for surgery, all endoscopies postoperatively, and possible surgical complications and re-resections were recorded from case notes. We defined proctocolectomy as the resection of both the colon and the rectum with ileoanal anastomosis, colectomy as the resection of the colon with ileorectal anastomosis, and colon resection as the resection of only part of the colon. Small bowel resections were limited to the diseased segments and included the ileocecal region when necessary. Stricturoplasties were not performed. Disease re-activation was defined as the inflammation seen in endoscopy or wireless capsule endoscopy that indicated an adjustment of medical therapy or surgery. Only re-resections were counted as reoperations.
Primary outcome was postoperative bowel function. Frequencies of relapses, reoperations, complications during follow-up, and postoperative long-term QoL were also studied.
We used the Mann-Whitney U[r] test (2-sample rank-sum test), Kruskal-Wallis test, and Dunn multiple comparison test to compare continuous variables between the groups. Fisher exact test was used to compare frequencies between different groups due to the small number of cases. Kaplan-Meier curves were used to analyze survival. The level of significance was set at P value <0.05. The data are shown in median and range or mean ± standard deviation.
The ethical committees of Tampere and Helsinki University hospitals approved the study protocol.
The majority of patients underwent surgery because of active disease despite optimal medical therapy (3 with infliximab) or steroid dependency (56%). Table 1 presents the background data of the patients. The median follow-up after primary resection was 10 years (range 2–21 years).
Respondents to Questionnaire
The response rate to the questionnaire was 67% (24/36). On average, there were <4% missing values for each question. To assess possible selection bias between respondents and nonrespondents, we performed a dropout analysis. The median disease duration before the operation was 3.0 years (range 0.6–10.0) versus 2.3 years (range 1.7–5.2), and the amount of time that had elapsed after the operation was 9.7 years (range 2.4–21.1) versus 9.4 years (range 2.1–18.0), which was similar between the respondents and nonrespondents, respectively. Those patients who completed the questionnaire were more often women (71% vs 50%; P = 0.28). Of the patients who responded, 5 had small bowel or ileocecal resection, 13 had both ileal and colon resection, and 6 had colon resection, colectomy, or proctocolectomy.
Outcome of Bowel Function
The dimensions (stool frequency, soiling, stool consistency, urgency, discrimination of flatus, evacuation, medications to control stool frequency) of the bowel function were comparable between the operation types (P ≥ 0.05, Table 2). The median daytime stool frequency was 3 (range 0–23), whereas it was zero at nighttime (range 0–10). The daytime (P = 0.05) and maximal stool frequency (P = 0.04) were higher in patients whose both ileum and colon were resected (n = 7) compared with patients who only had an ileum resection (n = 5). In the whole series, any degree of daytime soiling (39%) or nighttime soiling (33%) was reported and considered significant; however, soiling occurred <3 times per week in all of the patients. Overall, 33% of the patients were totally continent and only 11% of them reported watery stools. Most of them (73%) had an urgency period >15 minutes and most were able to discriminate between flatus and feces. Patients with permanent ostomy (n = 6) were excluded from these analyses.
QoL of Patients Compared With That of Controls
Surgery-related bowel problems had caused an absence from school or work in 30% of patients. Seventy-one percent of patients reported school absences less often than once per month or never; however, surgery appeared to have delayed educational milestones for 46% of the patients. Overall, 96% were completely or moderately satisfied with the outcomes of the surgery. Only 2 patients would not recommend surgery to other patients with CD.
Soiling or permanent ostomy had no effect on QoL. The different dimensions and overall QoL scores were similar between the patients and their matched controls (Fig. 1), and the type of surgery or the number of complications had no effect on the results (data not shown). The overall QoL (P = 0.03), physical (P = 0.002) and social (P = 0.003) QoL, at follow-up was significantly lower among patients who had missed school or work because of surgery compared with patients who could normally attend school or work (Fig. 1). The different dimensions of QoL were similar between re-resected patients and those without redo surgery during the length of the follow-up. At the last follow-up, both patients, younger and older than 20 years, reported comparable QoLs. In healthy control children, the mean scores were in line with previously published data (6,24).
Disease Reactivation After Surgery
By the end of the follow-up, the disease had reactivated in 34 (verified in endoscopy n = 32, or wireless capsule endoscopy n = 2) of 36 patients (94%). Postoperatively, before the first relapse, 2 patients had received infliximab due to active disease (introduction of therapy 2 months and 3 years postoperatively, respectively).
All relapses occurred during the first 13 years after the primary operation (median 1.8 years, range 0.2–12.7). Figure 2 shows the Kaplan-Meier estimate curve for the disease-free period as a function of time. The type of primary surgery or none of these disease-related characteristics, for example, disease duration for 1 year, perianal disease, or failed medical therapy as the only indication for operation, had no effect on the length of the disease-free period (data not shown). Patients had no uniform rescue therapy after surgery to diminish the relapse rate.
In total, 46 re-resections were performed in 20 patients (54%), of whom 14 (38%) had several of them (median 2, range 1–8, Table 3). There was no statistical difference in the number of re-resections between primary resection types (P ≥ 0.05). The length of time without re-resection was comparable between the different primary resection types (data not shown). Figure 3 shows the Kaplan-Meier estimate curve for the period without re-resection as a function of time. The median time from first surgery to re-resection was 2 years (range 0–16), and median period without re-resection was 7 years. Indication for primary surgery did not associate with re-resection rate.
Six of 18 patients receiving postoperative infliximab for active disease (a median of 5.5 years [range 0.0–12.5] to first postoperative infusion) underwent re-resection. We found no significant difference in the number of relapses or re-resections between patients who underwent surgery before the era of tumor necrosis factor (TNF)-α-antagonist therapy (starting by 1999) or after it.
By the end of the follow-up, 7 patients (19%) had permanent ostomy. Three of these patients previously had a J-pouch procedure that was converted to ileostomy because of pouch complications.
Complications Related to Surgery
There was no operation-related mortality. At least 1 surgical complication (median 2; range 0–12) occurred in 28 of 36 patients (77%, Table 4). Because of complications, a total of 16 re-resections and 5 fistulectomies were performed on 11 patients during follow-up. Anastomotic stenosis was common and 85% of them were dilated by endoscopy. One patient developed short-bowel syndrome (after 9 resections). Patients who had ongoing steroid therapy (n = 20) before the operation had an equal number of complications compared with patients without corticosteroids (n = 16).
Of the proctocolectomized patients, 8 of 9 of them had undergone ileal pouch formation. For these 9 patients, the disease had been diagnosed as UC before primary surgery. The diagnosis of CD was confirmed later a median of 1.6 (range 0.1–11.0) years after the proctocolectomy. All patients had experienced pouchitis, and 5 patients were receiving ongoing medical therapy for pouchitis. Four patients needed bowel diversion, which was permanent in 3 of the patients.
We studied the long-term bowel function after surgery in pediatric-onset CD median 10 years after primary bowel resection. Disease relapses, re-resection complications, and long-term QoL were also studied. Nearly all of the patients (94%) had an endoscopically verified and medically or surgically treated disease relapse by 13-year follow-up and every other patient already by 1.8-year follow-up. More than half of the patients had undergone re-resection; however, bowel function was acceptable and the QoL was reassuringly similar in the patients and their matched, population-based controls.
In most patients, the stool consistency and defecation were nearly normal, and 33% of the patients were totally continent. Although, compared with the healthy Finnish population (ages 18–26), soiling was more common and stool frequency higher in patients with CD with intestinal resection as in healthy individuals, stool frequency varies from once every other day to twice daily (26). There were no major differences in the overall bowel function between different types of surgery; however, the patients with longer ileocolonic resections experienced greater stool frequency during the day compared with those patients who had ileal resection only. To our knowledge, this is the first study that assesses long-term bowel function in pediatric-onset patients with CD who have undergone bowel resections.
For the first time in a study involving pediatric patients with CD, the long-term QoL of surgically treated patients was compared with the QoL of population-based controls. We did not find any significant differences in the overall QoL between patients and controls; however, the subgroup of patients who had missed school or work because of the disease had a reduced QoL. Barrena et al (11) noted that after a median of 6.2 years of follow-up, the QoL of 12 patients with CD having undergone surgery was considered normal; however, they used an unstructured questionnaire and did not have a control group. El-Baba et al (12) reported that all 22 of their patients with CD felt better after surgery, although 22% had relapsed after a mean of 4.7 years follow-up. It is possible that a high relapse rate may reduce QoL because patients who were absent from school or work most likely experienced clinical disease relapses; however, patients who had undergone re-resections had a surprisingly good QoL, and poor bowel function did not significantly worsen the QoL. According to Scott and Hughes (27), many adult patients stated that they would have preferred to have been operated on earlier. Accordingly, nearly all of our patients were satisfied with the results of surgery.
Pacilli et al (5) reported that 55% of pediatric patients had a clinical relapse within 2.5 years postoperatively. This is approximately the same figure as reported by Besnard et al (10) (50% within 2 years) and Baldassano et al (13) (60% within 5 years). Although the definition of a relapse varies and most reports define relapse according to clinical indices (Harvey-Bradshaw (5), Crohn's Disease Activity Index, Pediatric Crohn's Disease Activity Index (10,13)), results from Besnard et al (10) and Baldassano et al (13) are in line with our result of approximately 50% of patients relapsing during 1.8 years. In the recent study of Boualit et al (15) including 130 operated-on pediatric patients with CD, recurrence was defined as the need of immunosuppressant or biological medication postoperatively in patients naïve to these medicines preoperatively. In their study, the risk of recurrence was 18% at 2 years, 34% at 5 years, and 47% at 10 years. They also found that that risk for recurrence was 2.5 times less frequent if the resection was performed during the 3 first years after CD diagnosis. In our study, median time to first surgery from diagnosis was 3. The ages of patients also varied between studies; in our study, all patients had their first operation by the median age of 14, whereas in the study of Boualit et al, the patients were somewhat older. These above-mentioned facts can explain the differences seen in relapse rates between reports.
In adults, the postoperative recurrence rates at 5 years, defined by clinical symptoms, have been between 17% and 55% (28,29). The need for reoperation has been reported to range from 11% to 32% (28). These figures are lower than in the pediatric population; however, the endoscopic relapse rate at 5 years (58%–89%), depending on the population studied (29), in adults is comparable with our data.
The risk factors for early recurrence are not widely studied in children. According to the literature, the failure of medical treatment as the only indication for surgery (5,13), the preoperative use of 6-mercaptopurine (13), and the presence of active disease for more than a year before the operation (13,14) are related to early recurrence. The disease site also affects early recurrence. Diffuse ileocolonic disease (5,14), upper gastrointestinal involvement, perianal disease (10), and segmental colonic disease (14) are all related to early recurrences. We could not identify any specific risk factors for an early relapse, but the small number of patients precludes reliable statistical analysis.
To our knowledge, only 3 studies exist on pediatric patients with CD where the reoperation rate has been recorded. These show variable reoperation rates of 18% after a median follow-up of 2.5 years (5) and 27% after a median follow-up of 13 months (30) and a probability of reoperation of 8% at 2 years, 17% at 5 years, and 29% at 10 years (15). Our reoperation rate of approximately 50% by 10 years appears to be the highest figure. In our population, the median age at diagnosis was 12 years compared with 14.2 in the population of Boualit et al (15). This can explain our higher reoperation rate because the young age at diagnosis seems to be associated with complicated disease behavior (15). According to the literature, patients with rectum sparing colon resection have more reoperations than patients undergoing proctocolectomy (24,28–30). In our study, the reoperation rates for patients with colon resection, colectomy, proctocolectomy, or small bowel resection were statistically similar.
Nearly all of the patients had at least 1 surgical complication and the majority of the complications were strictures in the anastomosis that were managed by endoscopic dilatation. Compared with the adult data, our number of complications is higher, but in adult study the follow-up period was shorter (median 42 months) (31). In addition, all of our patients with ileal pouches experienced pouchitis, confirming the previous finding of a high incidence of pouch problems in patients with CD (32); however, the preoperative disease characteristics were indistinguishable from UC. Therefore, restorative proctocolectomies were performed (19). By the end of the follow-up, 19% of all patients had a permanent ostomy. In the long term, only 1 patient developed short-bowel syndrome, which is a serious complication that may require long-term parenteral nutrition (33) or even intestinal transplantation.
The timing of surgery is crucial because surgery can offer a disease-free interval to ensure normal growth and development when performed before late puberty (2,10,15,16). In the present study, it was not possible to assess the effect of surgery on growth. The prevention of relapses is an important aim because immediate relapses reduce the benefits of surgery (2,10,16).
Guidelines for the postoperative treatment of pediatric CD are lacking. Promisingly, for adults, the TNF-α-antagonists given immediately after surgery cut the relapse rate (34,35) without increasing the complication rate (36). Of our patients, nearly half underwent primary surgery before 1999, the landmark for the start of the clinical use of TNF-α-antagonist therapy; however, one-third of our patients treated postoperatively with biologics underwent re-resections despite the therapy. Some evidence indicates that in pediatric patients, biological treatment may delay surgery (8,9), but the effect of postoperative treatment with biological medication on preventing relapses needs further investigation.
Because the relapses and reoperations are common, we have adopted a policy to follow fecal calprotectin levels postoperatively for early detection of the possible disease reactivation. Detected high values schedule the patient earlier to the first postoperative endoscopy that is routinely performed at 3 to 6 months after surgery.
These findings show that after surgery, patients have a nearly normal bowel function and a good QoL. This supports the view that the effect of surgery on patients’ lives is, in most cases, positive, even though surgery does not cure CD; however, the small number of patients, the retrospective nature of the study, and the change in the practice of surgical and medical treatments during the lengthy follow-up times for patients are clear limitations of the present study.
In conclusion, surgery for pediatric-onset CD carries high risks, even under expert care, and nearly all patients experienced at least 1 surgical complication. Both endoscopic relapses and re-resections were common following surgery, indicating the need for a careful postoperative follow-up. The QoL of our patients, however, was comparable with that of healthy controls, and bowel function in the long term was, in most patients, nearly normal.
The authors thank Ms Anne Nikkonen for assistance in collecting the patient data.
1. Vernier-Massouille G, Balde M, Salleron J, et al. Natural history of pediatric Crohn's disease: a population-based cohort study. Gastroenterology 2008; 135:1038–1041.
2. Newby EA, Sawczenko A, Thomas AG, et al. Interventions for growth failure in childhood Crohn's disease (review). Cochrane Database Syst Rev 2005; 20:CD003873.
3. Vasseur F, Gower-Rousseau C, Vernier-Massouille G, et al. Nutritional status and growth in pediatric Crohn's disease: a population-based study. Am J Gastroenterol 2010; 105:1893–1900.
4. Motil KJ, Grand RJ, Davis-Kraft L, et al. Growth failure in children with inflammatory bowel disease: a prospective study. Gastroenterology 1993; 105:681–691.
5. Pacilli M, Eaton S, Fell JM, et al. Surgery in children with Crohn's disease refractory to medical therapy. J Pediatr Gastroenterol Nut 2011; 52:286–290.
6. Turunen P, Ashorn M, Auvinen A, et al. Long-term health outcomes in pediatric inflammatory bowel disease: a population-based study. Inflamm Bowel Dis 2009; 15:56–62.
7. Podolsky D. Inflammatory bowel disease. N Engl J Med 2002; 347:417–429.
8. Gupta N, Cohen SA, Bostrom AG, et al. Risk factors for initial surgery in pediatric patients with Crohn's disease. Gastroenterology 2006; 130:1069–1077.
9. Afzal NA, Ozzard A, Keady S, et al. Infliximab delays but does not avoid the need for surgery in treatment-resistant pediatric Crohn's disease. Dig Dis Sci 2007; 52:3329–3333.
10. Besnard M, Jady O, Mougenot JF, et al. Postoperative outcome of Crohn's disease in 30 children. Gut 1998; 43:634–638.
11. Barrena S, Matinez L, Hernandez F, et al. Surgical treatment of chronic inflammatory bowel disease in children. Pediatr Surg Int 2011; 27:385–390.
12. El-Baba M, Chuan-Hao L, Klein M, et al. Outcome after surgical intervention in children with chronic inflammatory bowel disease. Am Surg 1996; 62:1014–1017.
13. Baldassano RN, Jeshion WC, Piccoli DA, et al. Pediatric Crohn's disease: risk factors for postoperative recurrence. Am J Gastroenterol 2001; 96:2169–2176.
14. Griffths AM, Wesson DE, Shandling B, et al. Factors influencing postoperative recurrence of Crohn's disease in childhood. Gut 1991; 32:491–495.
15. Boualit M, Salleron J, Turck D, et al. Long-term outcome after first intestinal resection in pediatric-onset Crohn's disease: a population-based study. Inflamm Bowel Dis 2013; 19:7–14.
16. Ranger GS, Lamparelli MJ, Aldridge A, et al. Surgery results in significant improvement in growth in children with Crohn's disease refractory to medical therapy. Pediat Surg Int 2006; 22:347–352.
17. Puntis J, McNeish AS, Allan RN. Long term prognosis of Crohn's disease with onset in childhood and adolescence. Gut 1984; 25:329–336.
18. Langholz E, Munkholm P, Krasilnikoff PA, et al. Inflammatory bowel diseases with onset in childhood. Clinical features, morbidity, and mortality in a regional cohort. Scand J Gastroenterol 1997; 32:139–147.
19. Auvin S, Moliniè F, Gower-Rousseau C, et al. Incidence, clinical presentation and location at diagnosis of pediatric inflammatory bowel disease: a prospective population-based study in northern France (1988–1999). J Pediatr Gastroenterol Nutr 2005; 41:49–55.
20. Kugathasan S, Judd RH, Hoffman RG, et al. Epidemiologic and clinical characteristics of children with newly diagnosed inflammatory bowel disease in Wisconsin: a statewide population-based study. J Pediatr 2003; 143:525–531.
21. Heyman MB, Kirschner BS, Gold BD, et al. Children with early-onset inflammatory bowel disease (IBD) analysis of a pediatric IBD consortium registry. J Pediatr 2005; 146:35–40.
22. Bousvaros A, Antonioli DA, Dubinsky MC, et al. Differentiating ulcerative colitis from Crohn disease in children and young adults: report of a working group of the North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition and the Crohn's and Colitis Foundation of America. J Pediatr Gastroenterol Nutr 2007; 44:653–664.
23. Griffiths AM, Buller HB. Walker WA, Durie PR. Inflammatory bowel disease. Pediatric Gastrointestinal Diseases 3rd ed.New York:Marcel Decker; 2000. 613–652.
24. Pakarinen M, Natunen J, Ashorn M, et al. Long-term outcomes of restorative proctocolectomy in children with ulcerative colitis. Pediatrics 2009; 5:1377–1382.
25. Mortellaro VE, Green J, Islam S, et al. Occurance of Crohn's disease in children alter total colectomy for ulcerative colitis. J Surg Res 2011; 170:38–40.
26. Kyrklund K, Koivusalo A, Rintala RJ, et al. Evaluation of bowel function and fecal continence in 594 Finnish individuals aged 4 to 26 years. Dis Colon Rectum 2012; 55:671–676.
27. Scott NA, Hughes LE. Timing of ileocolonic resection for symptomatic Crohn's disease—the patient's view. Gut 1994; 35:656–657.
28. Yamato T. Factors affecting recurrence after surgery for Crohn's disease. World J Gastroenterol 2005; 11:3971–3979.
29. Buisson A, Chevaux JB, Allen PB, et al. Review article: the natural history of postoperative Crohn's disease recurrence. Aliment Pharmacol Ther 2012; 35:625–633.
30. Patel HI, Leichtner AM, Colodny AH, et al. Surgery for Crohn's disease in infants and children. J Pediatr Surg 1997; 32:1063–1068.
31. Kasparek MS, Glatzle J, Mueller MH, et al. Postoperative complications have little influence on long-term quality of life in Crohn's patients. J Gastrointest Surg 2008; 12:569–576.
32. Alexander F, Sarigol S, DiFiore J, et al. Fate of the pouch in 151 pediatric patients after ileal pouch anal anastomosis. J Pediatr Surg 2003; 38:78–82.
33. Schalamon J, Mayr JM, Höllwarth ME. Mortality and economics in short bowel syndrome. Best Pract Res Clin Gastroenterol 2003; 17:931–942.
34. Sorrentino D, Paviotti A, Terrosu G, et al. Low-dose maintenance therapy with infliximab prevents postsurgical recurrence of Crohn's disease. Clin Gastroenterol Hepatol 2010; 8:591–599.
35. Regueiro M, Schraut W, Baidoo L, et al. Infliximab prevents Crohn's disease recurrence after ileal resection. Gastroenterology 2009; 136:441–450.
36. Regueiro M, El-Hachem S, Kip KE, et al. Postoperative infliximab is not associated with an increase in adverse events in Crohn's disease. Dig Dis Sci 2011; 56:3610–3615.