Inflammatory bowel disease (IBD) includes Crohn disease and ulcerative colitis, and is often diagnosed in late childhood and early adulthood. What determines the age of onset remains unexplained. Early onset may represent the “pure” form of the disease process and hence may hold secrets of the initiating events of IBD pathogenesis. Clinical scientists continue to focus on pediatric IBD because it may shed light on the cause and prevention of this lifelong disease. Over the last decade, data in pediatric IBD studies have demonstrated many similarities and differences between pediatric and adult onset, which continue to add pieces to an increasingly complex IBD puzzle. The mechanism responsible for these similarities and differences remains unanswered. This article discusses clinically relevant epidemiology and treatment aspects of pediatric IBD, with special focus on similarities and differences in pediatric and adult IBD. Evidence-based treatment algorithms, with special focus on pediatric studies and care for children, are also highlighted.
Inflammatory bowel disease (IBD) includes Crohn disease (CD) and ulcerative colitis (UC), and is often diagnosed in late childhood and early adulthood. IBD is thought to develop as a result of dysregulation of the immune response to normal gut flora in a genetically susceptible host. Approximately 25% of incident cases of IBD occur during childhood and the rest occur throughout adulthood, peaking in the second and third decades of life. What determines the age of onset remains unexplained. Studying early-onset presentation and epidemiology of complex predominately adult diseases such as IBD is particularly necessary, as the early onset may represent the “pure” form of the disease process and hence may hold secrets of the initiating events of IBD pathogenesis. Basic, translational, and clinical scientists continue to focus on pediatric IBD, because it may shed light not only on the cause but also the prevention of this lifelong disease. Over the last decade, data from pediatric IBD studies have demonstrated many similarities and differences between pediatric and adult onset, which continue to add pieces to an increasingly complex IBD puzzle. The mechanism responsible for these similarities and differences remains unanswered.
The purpose of this article is to discuss clinically relevant epidemiology and treatment aspects of pediatric IBD, with a special focus on similarities and differences in pediatric and adult IBD. Epidemiologic similarities and differences may ultimately provide the link to a better understanding of the pathogenesis of IBD. This article also highlights evidence-based treatment algorithms, with special focus on pediatric studies and care for children.
Epidemiology
Gender Differences
The male to female ratio of IBD differs in multiple studies when comparing pediatric IBD to adult IBD. Whereas in adult IBD there is an equal ratio of male to female disease or perhaps more women with disease, prepubertal males seem to be more affected by pediatric CD. Van Limbergen and colleagues demonstrated that in pediatric CD there is a strong trend toward males, with a male to female ratio of 1.5:1. Vernier-Massouille and colleagues also confirmed a male predilection, with a similar ratio of 1.4:1 in children younger than 15 years. This figure directly compares with a ratio nearing 1:1 in patients older than 15 years in the same population. Other adult epidemiologic studies similarly have shown a gender ratio of approximately 1:1. A summary of recent epidemiologic studies demonstrating a male preponderance in pediatric CD is provided in Table 1 .
| CD | ||||
|---|---|---|---|---|
| Male | Female | Male:Female Ratio | ||
| Van Limbergen et al | 164 | 112 | 1.464285714 | Childhood |
| 216 | 380 | 0.568421053 | Adult | |
| Kugathasan et al | 80 | 49 | 1.632653061 | Childhood |
| Vernier-Massouille et al | 256 | 216 | 1.185185185 | Childhood |
| Kappelman et al (per 100,000) | 47 | 38 | 1.236842105 | Childhood |
| 183 | 216 | 0.847222222 | Adult | |
| Herrinton et al (per 100,000) | 40 | 27 | 1.481481481 | Childhood |
| 184 | 188 | 0.978723404 | Adult | |
| Sawczenko et al (per 100 diagnoses) | 62 | 38 | 1.631578947 | Childhood |
| Newby et al | 83 | 33 | 2.515151515 | Childhood |
Pediatric UC does not demonstrate the male predominance seen in pediatric CD. In fact, similar to adult UC, males and females are equally affected in pediatric UC in multiple studies. Table 2 demonstrates the equal distribution of males and females in pediatric UC.
| UC | ||||
|---|---|---|---|---|
| Male | Female | Male:Female Ratio | ||
| Van Limbergen et al | 48 | 51 | 0.941176471 | Childhood |
| 342 | 359 | 0.95264624 | Adult | |
| Kugathasan et al | 33 | 27 | 1.222222222 | Childhood |
| Kappelman et al (per 100,000) | 29 | 26 | 1.115384615 | Childhood |
| 238 | 237 | 1.004219409 | Adult | |
| Herrinton et al (per 100,000) | 28 | 29 | 0.965517241 | Childhood |
| 312 | 283 | 1.102473498 | Adult | |
| Sawczenko et al (per 100 diagnoses) | 51 | 49 | 1.040816327 | Childhood |
| — | — | — | — | |
| Newby et al | 45 | 29 | 1.551724138 | Childhood |
Very early onset IBD (age <5–8 years) has been recently suggested as perhaps a different spectrum of IBD. Multiple epidemiologic studies have demonstrated a male preponderance of very early onset IBD. It is unclear whether this male preponderance is only seen in CD or in both CD and UC, as many of the studies do not provide specific gender information based on specific diagnosis.
Together, these gender differences generate more questions than they answer. The effect of puberty and sex hormones on disease pathogenesis continues to be unanswered. Further study will continue to explore these interesting epidemiologic findings and may ultimately provide important information as to the cause of pediatric IBD.
Crohn Disease:Ulcerative Colitis Ratio
The ratio of CD:UC significantly differs in children and adults. Van Limbergen and colleagues demonstrated a significant predilection for CD in children, with a ratio of 2.8:1. Adult IBD demonstrated a ratio of 0.85:1. Other epidemiologic studies that have been performed using various methods (ie, population based, insurance claims, and so forth) have suggested a similar higher ratio of CD than UC in children. This directly differs from adult studies that have demonstrated more UC than CD diagnoses. Table 3 illustrates recent studies that reveal this significant difference in CD:UC ratios in children and adults.
| CD | UC | CD:UC Ratio | ||
|---|---|---|---|---|
| Van Limbergen et al | 276 | 99 | 2.79 | Childhood |
| 596 | 701 | 0.85 | Adult | |
| Kugathasan et al | 129 | 60 | 2.15 | Childhood |
| Vernier-Massouille et al | 472 | 151 | 3.13 | Childhood |
| Kappelman et al (per 100,000 prevalence) | 1118 | 722 | 1.55 | Childhood |
| 12800 | 15151 | 0.84 | Adult | |
| Herrinton et al (per 100,000 period prevalence) | 67 | 57 | 1.18 | Childhood |
| 186 | 299 | 0.62 | Adult | |
| Sawczenko et al | 431 | 211 | 2.04 | Childhood |
| Auvin et al | 367 | 122 | 3.01 | Childhood |
| Newby et al | 116 | 74 | 1.57 | Childhood |
| Heyman et al | 798 | 393 | 2.03 | Childhood |
There is clearly a significantly higher CD:UC ratio in children than in adults. Similar to gender differences, this observation clearly requires further investigation, with few data available to suggest a mechanism for this significant difference between pediatric and adult IBD.
Disease Location
Disease location at presentation differs in pediatric IBD compared with adult IBD. In pediatric CD a majority of patients have ileocolonic disease or colonic disease, whereas adults more often present with terminal ileal disease without colonic involvement. During follow-up of pediatric CD, Vernier-Massouille and colleagues and Van Limbergen and colleagues both demonstrated a progression of disease location with increasing ileocolonic disease at follow-up. Meanwhile, Van Limbergen and colleagues showed that more than one-third of adults had terminal ileal disease only. In multiple pediatric studies, up to 80% to 90% of children experience colonic disease (colon only or ileocolonic) whereas only approximately 50% of adults experience colonic CD. It is unclear why there is an increased ileocolonic/colonic disease in children and no accepted theory has been offered as to this observation. No treatment studies have suggested that children with ileocolonic disease require different treatment from terminal ileal disease alone. Table 4 demonstrates the high rate of ileocolonic disease in children at both diagnosis and follow-up in 2 studies.
| CD Location | |||||
|---|---|---|---|---|---|
| Terminal Ileum (L1) | Colon Only (L2) | Ileocolonic (L3) | Upper Gastrointestinal Disease (L4) | ||
| Vernier-Massouille et al | 14% | 17% | 69% | 34% | Childhood diagnosis |
| 9% | 9% | 82% | 48% | Childhood follow-up (10 y) | |
| Van Limbergen et al | 6% | 36% | 51% | 51% | Childhood diagnosis |
| 5% | 36% | 54% | 61% | Childhood follow-up (4 y) | |
| 36% | 38% | 23% | 12% | Adult | |
| Kugathasan et al | 25% | 32% | 29% | 14% | Childhood diagnosis |
| Sawczenko et al | 9% | 7% | 84% | 50% | Childhood diagnosis |
| Auvin et al | 19% | 10% | 71% | — | Childhood diagnosis |
Pediatric UC location also differs from that of adult disease. Pediatric UC presents more often with pancolitis versus left-sided colitis/proctitis. In fact, most pediatric UC studies demonstrate up to 80% to 90% of children present with pancolitis, and a recent study by Van Limbergen and colleagues suggested that pediatric UC progresses with increasing percentage of pancolitis at follow-up. The significance of the high percentage of pancolitis at presentation is unknown. Table 5 demonstrates an increased pancolitis presentation in pediatric UC.
| UC Location | ||||
|---|---|---|---|---|
| Proctitis (E1) | Distal Disease (Left-sided) (E2) | Pancolitis (E3) | ||
| Van Limbergen et al | 4% | 21% | 75% | Childhood diagnosis |
| 1% | 16% | 82% | Childhood follow-up | |
| 17% | 35% | 48% | Adult follow-up | |
| Hyams et al | 22% | 39% | 43% | Childhood diagnosis |
| Kugathasan et al | — | 10% | 90% | Childhood diagnosis |
| Hyams et al | — | — | 80% | — |
| Sawczenko et al | 4% | — | 81% | Childhood diagnosis |
| Auvin et al | 11% | 57% | 32% | Childhood diagnosis |
In summary, pediatric CD more often involves the ileocolonic/colonic regions whereas adult CD does not demonstrate a high proportion of colonic disease. Meanwhile, pediatric UC more often presents with pancolitis whereas adult UC more often presents with left-sided colitis. The mechanism behind these observations is not well understood and no data are available to support any concrete hypotheses.
Disease Phenotype
Disease phenotype in both CD and UC differs when comparing children with adults. Pediatric CD presents predominantly with inflammatory or nonstricturing, nonpenetrating disease. Stricturing and penetrating disease is relatively uncommon at presentation in pediatric CD. However, even with treatment, multiple studies have shown that CD progresses to stricturing and penetrating disease in many children. Adult disease presents more often with stricturing and penetrating disease. Two recent natural history articles reveal a significant progression of pediatric CD from inflammatory disease to sticturing disease, as illustrated in Table 6 .
| CD Phenotype | ||||
|---|---|---|---|---|
| B1: Inflammatory (NS, NP) | B2: Stricturing | B3: Penetrating | ||
| Van Limbergen et al | 91% | 4% | 5% | Childhood diagnosis |
| 76% | 13% | 11% | Childhood 4 y follow-up | |
| 66% | 14% | 20% | Adult follow-up | |
| Vernier-Massouille et al | 71% | 25% | 4% | Childhood diagnosis |
| 41% | 44% | 15% | Childhood 10 y follow-up | |
Pediatric UC more often presents with pancolitis, and has been suggested to be a more severe phenotype in children than in adults. Recent epidemiologic data demonstrate that indeed, time from diagnosis to first surgery in UC is significantly shorter in children than in adults. By 10 years after diagnosis more than 40% of children had undergone colectomy, whereas only 20% of adult-onset UC patients had undergone colectomy.
In summary, pediatric CD frequently displays an inflammatory phenotype at diagnosis that progresses to fistulizing/stricturing disease in some patients, whereas adult CD more often presents with fistulizing/stricturing disease. Although some investigators have suggested disease duration and a delay in diagnosis may be the reason for this difference in CD, no data have been published to support this hypothesis. Pediatric UC frequently displays an aggressive phenotype, with pancolitis and early time to first surgery compared with adult UC, which is more often limited to the left colon.
Genetics
When studying early-onset presentations of disease, there is an assumption that these represent a more severe, more genetically influenced group of patients. It is appealing to geneticists to study these patients because of the increased chance of finding novel risk variants. One of the most compelling hypotheses is that pediatric-onset IBD is more likely to be influenced by genetics compared with late- or adult-onset IBD, as there is less time for environmental modifiers to have influenced the disease.
IBD is highly heritable. This concept is strongly supported by family, twin, and phenotype concordance studies, and now is confirmed by the discoveries of many susceptibility genes. Initial family-based linkage studies of IBD implicated the NOD2 gene in CD and the MHC region on chromosome 6p in UC for increased susceptibility. Genome-wide association scanning (GWAS), which employs high-density single nucleotide polymorphism (SNP) array technology, has recently increased the possible genetic factors linked to IBD pathogenesis. This method of broad, unbiased screening for the contribution of common genetic variation for disease susceptibility has provided strong evidence for many CD and UC susceptibility loci. GWAS has identified loci in both UC and CD that are already known to be involved in adaptive immunity genes such as IL23R, IL12B, STAT3, loci on 3p21 (MST1), and 10q24 (NKX2-3). Variants in innate immunity genes, particularly those mediating autophagy and bacterial sensing (ATG16L1, IRGM, and NOD2) have also been discovered through these methods in CD. To date, the majority of this genetic analysis in IBD has been done in adult cohorts with adult-onset disease as the primary phenotype, therefore even less is known about early-onset variants.
Several CD susceptibility alleles have been confirmed in both pediatric and adult populations. However, most of the genetic variation seen in adults has not been studied in children in a large cohort with adequate power. Two pediatric studies attempting to replicate the effect of adult-onset IBD loci in children have been performed recently. These studies have demonstrated that autophagy genes play a role in pediatric CD but also that known genetic risk factors found in adults may not distinguish early- and late-onset IBD.
The first pediatric GWAS IBD scan was performed recently, revealing 2 risk variants not previously reported in adults, in addition to confirming the most significant adult risk variants. Two novel loci, the TNFRSF6B and PSMG1 genes, were discovered using more than 1000 cases of pediatric IBD. The gene TNFRS6B, which encodes a decoy receptor for the FasL pathway (DCR3), was found to increase the risk for pediatric-onset CD and UC. On comparison with adult GWAS scans these same loci were identified, but were below the expected threshold when correcting for multiple tests. However, until a GWAS is performed in an exclusively pediatric-onset IBD cohort, it is very difficult to deny that additional pediatric-onset IBD susceptibility genes do not exist. As such, GWAS studies involving larger pediatric-onset CD cohorts and early-onset UC are presently underway.
Adult and pediatric GWAS studies have yet to discover risk variants that are specific to pediatric or adult IBD. Instead, all risk variants that have been discovered are present in both adult and pediatric scans, although not necessarily in the statistically significant range. This observation, if confirmed in additional larger GWAS studies, may further suggest that pediatric and adult IBD have similar genetics and thus are the same disease with different age of presentation.
More detailed functional exploration of genes associated with susceptibility loci reported in GWAS will be instrumental in shedding light on their role in IBD pathogenesis. Taken together, recent pediatric GWAS results substantially advance the current understanding of pediatric-onset IBD by highlighting key pathogenetic mechanisms, and allowing for the first time a comparison between genetic susceptibility in an exclusively pediatric cohort and the previously described populations with predominantly adult-onset disease.
Clinical Presentation
Clinical presentation is similar in adult and pediatric IBD, and for the most part correlates with disease location. Pediatric CD presents with more ileal and colonic disease than adult CD, and therefore more often presents with hematochezia. Small bowel disease presents with diarrhea regardless of childhood or adult onset. Pediatric and adult IBD share many of the same gastrointestinal symptoms which, as one would expect, are associated more with mucosal disease than with age.
Extraintestinal manifestations similarly are present in both children and adults in similar numbers. Extraintestinal manifestations are present in 6% of children prior to diagnosis in one recent study, and cumulative incidence approaches 25%, similar to adult data.
Growth is the most significant difference in presentation between adult and pediatric IBD. Poor growth prior to diagnosis has been documented in multiple studies examining growth in pediatric IBD. Furthermore, puberty has been shown to be delayed and some patients have decreased final adult height. In addition, a recent study has demonstrated that despite new treatments, catch-up growth does not occur in patients diagnosed with IBD, although it remains to be seen whether these patients have delayed puberty and ultimately achieve their expected adult height. Persistent poor growth may also be one of the only signs of increased disease activity, thus it is important not only in presentation but also in disease activity during treatment.
Treatment
Pediatric IBD treatment employs many of the same treatment paradigms as adult IBD. Most medication clinical trials have largely been performed only on adults, and therefore much of the evidence given here is based on adult data. Only a few well-designed clinical trials have been performed in children, and most of those show similar efficacy to adult trials.
The authors have chosen to separate the Treatment section into Induction and Maintenance of remission. What follows is not meant to be an exhaustive review of current literature; rather a review some of the current data and a report on any additional data specific to children.
Corticosteroids
Crohn disease
Induction of remission
Two recent Cochrane review articles examined budesonide and conventional corticosteroids (prednisone) as therapy for the induction of remission in CD. A majority of these were adult studies, although in some studies children older than 16 years were included. The reviews clearly show efficacy for induction of remission with both budesonide and conventional steroids, and show slightly less efficacy of budesonide compared with conventional steroids, at least in severe disease.
Maintenance of remission
Corticosteroids, including both conventional steroids and budesonide, are not recommended for maintenance of remission. Any benefits in maintenance of remission are offset by treatment-related adverse events and thus these medications should be avoided for maintenance of remission, especially in children in whom corticosteroids can significantly affect growth.
Ulcerative colitis
Induction of remission
The use of corticosteroids for induction of remission in UC was first described in 1974, and has been the mainstay for induction of remission in moderate to severe UC. More recently, approximately 84% of adults with UC demonstrated a complete or partial improvement of disease activity with corticosteroids, and approximately 49% had prolonged response at 1 year. Budesonide, which has significant first-pass metabolism, is delivered to the distal ileum and proximal colon and thus likely has little efficacy in UC, although a randomized controlled trial has never been published.
Maintenance of remission
As with CD, corticosteroids are not recommended for the maintenance of remission in UC as any benefits are more than offset by side effects, especially in growing children.
Summary
In practice, corticosteroids are often used in induction of remission in CD and UC but are often avoided if possible after induction, due to growth side effects and other morbidity associated with persistent corticosteroid usage. Studies have clearly shown that morbidity in CD is associated with corticosteroid use. Therefore, children with new diagnosis CD and UC are often started on corticosteroids with a taper over 2 months. Corticosteroids are avoided, if possible at all other times other than induction of remission.
Nutritional Therapy
Crohn disease
Induction of remission
A recent Cochrane review compared nutritional therapy (liquid formula by mouth or via tube) to corticosteroid therapy for induction of remission. Although sole nutritional therapy has been shown to be effective in induction of remission in CD, it remains inferior to corticosteroids in induction of remission. In addition, the same review concluded that protein composition (ie, elemental or nonelemental protein) has no effect on efficacy of nutritional therapy. The only study that favors enteral nutrition over corticosteroids was a pediatric study.
However, practically speaking, the induction of remission with sole nutritional therapy remains difficult due to adherence in children. Most parents are reluctant to commit total enteral nutrition for their children for 6 to 8 weeks as required. In addition, few children are able to consume adequate formula volume by mouth, and thus would require insertion of nasogastric tubes or possibly a gastrostomy tube.
Maintenance of remission
Only 2 randomized studies have been published regarding maintenance of remission in CD with nutrition in adults, and no studies have examined this in children. Takagi and colleagues conducted a randomized controlled trial of 51 adults in remission, assigning one group a half-elemental diet and another group a regular diet with no instructions or limitations, with all patients taking mesalamine. The study demonstrated a significantly lower relapse rate in the half-elemental diet group (34.6% vs 64.0%). A recent Cochrane review suggested that there may be some efficacy in enteral nutrition for maintenance therapy, although larger studies are needed to confirm this possibility.
Ulcerative colitis
No studies have been reported for the induction or maintenance of remission with enteral therapy for UC.
Summary
Although both the European Society for Pediatric Gastroenterology, Hepatology, and Nutrition and the Japanese Society for Pediatric Gastroenterology, Hepatology, and Nutrition recommend nutritional therapy as sole primary therapy for CD, it is often a difficult treatment to initiate in the United States. The most significant deterrent continues to be the resistance by many parents and children to commit to 8 weeks of specialized formula alone (either by mouth or through a feeding tube such as a nasogastric tube) without taking any other food by mouth. Furthermore, it is unclear whether a child can be maintained on nutritional therapy alone and therefore, other maintenance medications need to be initiated. In addition, the short-term side effects with corticosteroid induction are debatably minimal.
5-Aminosalicylate
Crohn disease
Induction of remission
No randomized controlled studies have been performed examining the induction of remission by aminosalicylates in pediatric CD. Hanauer and colleagues published a meta-analysis examining aminosalicylates in active CD and showed a modest effect (if any) on improvement of Crohn Disease Activity Index (CDAI).
Although some pediatric gastroenterologists continue to use aminosalicylates for the induction of remission in CD, there are no good data to support the use for induction of remission, although there may be modest beneficial effects.
Maintenance of remission
No randomized studies have been published for maintenance of remission in pediatric CD. A recent Cochrane review examined aminosalicylates in the maintenance of remission in CD. The results do not show any benefit of aminosalicylates compared with placebo.
Ulcerative colitis
Induction of remission
A recent Cochrane review that includes only adult studies demonstrated a benefit from high-dose aminosalicylates (>3 g) in the induction of remission in UC, although remission rates remain significantly lower than in those using corticosteroids.
Most importantly in pediatric UC, a majority of patients present with moderate to severe disease, as demonstrated by the high percentage of pancolitis. Whereas aminosalicylates can be used for induction of remission, due to severity of disease in pediatric UC they are rarely used as sole induction, but rather in conjunction with corticosteroids for induction of remission.
Maintenance of remission
As with many of the medications mentioned here, no randomized trials have been conducted in children examining the use of aminosalicylates in maintenance of remission in children. A Cochrane review has demonstrated efficacy of aminosalicylates in maintaining remission in UC compared with placebo.
Summary
There are no randomized controlled trials for the induction or maintenance of remission of aminosalicylates in pediatric IBD. Aminosalicylates are still often used in pediatric IBD. These agents are well tolerated in children, have few side effects, and thus have continued to be used by pediatric gastroenterologists despite the lack of evidence in CD and efficacy only in mild to moderate UC. Even though aminosalicylates are well tolerated, the most significant “side effect” is decreased quality of life due to the number of pills or capsules ingested each day, which often is considered in treatment of children with IBD.
However, it should be noted that aminosalicylates likely have most efficacy in colonic disease regardless of whether the diagnosis is CD or UC. Given the data demonstrating more colonic disease in children (up to 80%), the use of aminosalicylates in colonic CD may be of some benefit, although there are no data examining this question. Whether aminosalicylates could be used as sole therapy in colonic CD has never been examined.
Based on the current data, the authors cannot support the usage of aminosalicylates in ileal CD, as there seem to be no compelling data to support their use. However, there are data to support the use of high-dose aminosalicylates in the management of UC, and there may be some efficacy in colonic CD, although there are no data for this possibility.
Immunomodulators (6-Mercaptopurine, Azathioprine, Methotrexate)
Crohn disease
Induction of remission
Because of its delay in efficacy, 6-mercaptopurine (6-MP) and azathioprine (AZA) are not used for induction of remission; however, they are often used in conjunction with corticosteroids or other therapy used to induce remission with the knowledge that by the time corticosteroids are weaned, 6-MP and AZA will be effective in maintaining remission ( Table 7 ).
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
