Epidemiology of CD

Incidence and prevalence —The identification of true increases in disease incidence, in conjunction with other population trends, may provide useful insights into disease pathogenesis, especially with regard to environmental influences. By contrast, trends in disease prevalence are invaluable for organizing current health care provisions and planning future service design. The rising incidence of IBD should be noted along with the general increase in incidence of complex immune diseases such as asthma and type 1 diabetes.

There is a large body of evidence with regard to incidence of pediatric-onset CD worldwide showing an increasing incidence in most areas surveyed, although many geographical areas are not represented and a significant proportion lack data on trends over time, as demonstrated in a recent exhaustive systematic review. Since the systematic review, new CD incidence data have emerged—of most note, the first Mediterranean PIBD data, with the robust longitudinal Spanish experience of 1996 to 2009. In Scotland, incidence data spanning over 40 years show a dramatic increase in PIBD, especially CD. The most recent Scottish data showed a minimum 66% increase in incidence of CD diagnosed in patients younger than 16 years of age from 2.9 per 100,000/year in 1990 to 1995 to 4.8 per 100,000/year in 2003 to 2008, with a 500% rise in pediatric CD incidence over just 40 years, and with further increases from 2009 to 2013. A clear north-south gradient exists in incidence of pediatric CD in the northern hemisphere, higher, for example, in the northern states of the United States than in the southern states, and in Scotland than the rest of the United Kingdom, within the north compared to the south of Scotland; the highest current worldwide incidence rates are in Scandinavia and Ontario, Canada.

With regard to worldwide rates of CD prevalence, there has been no specific systematic review for the pediatric-onset group, although that of adult CD has been recently published. As mentioned, disease prevalence rather than incidence is useful in the context of service provision. It is vital to have up-to-date data to inform service design, identify emerging cost and clinical pressures, and also allow affected families to be updated, as well as the charities and groups that support them. In particular, with the reported more extensive phenotype of PIBD, the rapid progression of PIBD early in the disease course, the need for timely endoscopic assessment of suspected PIBD, the need for endoscopic re-assessment in complicated PIBD cases, and the high use of immunosuppressants (thiopurines and methotrexate) and biological agents (infliximab and adalimumab) in PIBD, there are considerable cost pressures and cost-benefit issues that need to be addressed.

Classification of CD

The Montreal classification of IBD was developed in 2005 to accurately classify the phenotype (physical characteristics) of all subtypes of IBD, encompassing disease location (L; L1-4 for CD), behavior (B; B1-B3 for CD), severity (S), and age of onset (A; A1 for diagnosis <17 years of age) phenotypes, thus allowing coverage of features at diagno­sis and during evolution with time that had not been covered in the previous Vienna classification. However, the Montreal classification of IBD has limitations with respect to classification of PIBD; the dynamic features of pediatric disease phenotype (change in disease location and behavior over time, growth failure) were not sufficiently captured. An international group of pediatric IBD experts met in Paris, France, to develop evidence-based consensus recommendations for a pediatric modification of the Montreal criteria. The important modifications, termed the Paris classification and published in 2011(see Table 42-1 ) included classifying age at PIBD diagnosis as A1a (0 to <10 years) or A1b (10 to <17 years). For pediatric CD, important changes were the following: (1) distinguishing disease location above the distal ileum as L4a (proximal to ligament of Treitz) and L4b (ligament of Treitz to above distal ileum), (2) allowing both stenosing and penetrating CD to be classified in the same patient (B2B3), and (3) denoting the presence of growth failure in the patient at any time as G ₁ versus G ₀ (never growth failure). The Paris classification seamlessly integrates with the Montreal framework, allowing classification for adult-onset disease or by adult gastroenterologists.

Diagnosis of CD

The revised Porto criteria for diagnosis of IBD —An international group of experts in PIBD, mainly from the Porto IBD Working Group of ESPGHAN, have constructed a methodologically robust, consensus-based clinical guideline for the diagnosis of PIBD, including full facets of the assessment by all investigative modalities, and interpretation of these results. This updated and expanded the earlier work, which had become known as the “Porto Criteria,” using the best recent available evidence from the PIBD literature, relevant methodologically high quality data from the adult IBD literature, and clinical expertise and experience from PIBD specialist multidisciplinary teams. The Paris classification was a major reference point, highlighting those phenotypic characteristics that are either more common in or are unique to pediatric-onset rather than adult-onset IBD.

Diagnosing CD —Both diagnostic and nonspecific, but common, features of CD are presented in Table 42-2 . Typical CD features include noncontiguous aphthous or linear ulcers, primarily in the ileum or colon, although CD may involve any area of the gastrointestinal tract, and disease may be confluent. CD may present with extraintestinal manifestations initially; in this scenario, the definitive diagnosis requires evidence of gastrointestinal disease. Histologically the disease is usually characterized by chronic focal inflammation, with or without granulomas.

Diagnosis may become more difficult in cases of infantile-onset IBD (0 to 2 years of age) or cases with predominantly colonic disease, where confusion may arise with UC or IBDU or when children present with a rare CD phenotype (present in less than 5% of all CD presentations) for example, oral and/or genital and/or perianal CD without gut luminal involvement, or nondeforming arthritis.

Following the first well-recognized attempt to classify IBD by Lennard-Jones, the recognition of granulomatous inflammation in the diagnosis of CD continues to be emphasized. This must be remembered when dealing with rare phenotypic presentations of CD, defined as those occurring in less than 5% of cases, and which were omitted from the Paris classification for reasons of brevity (A Levine, personal communication). In a detailed report of the differences in phenotype at diagnosis and follow-up of 416 Scottish children and 1296 adults with IBD, Van Limbergen et al. reported that 5% (14 of 273) children at diagnosis of CD had no gastrointestinal luminal CD but oral and perianal, isolated perianal, or isolated oral CD alone; 70% developed luminal disease at 4 years or more of follow-up. Conventionally, the term orofacial granulomatosis (OFG) is used to describe patients with granulomatous oral lesions, but without evidence of CD elsewhere in the lumen of the gastrointestinal tract; patients with intestinal CD who have involvement of the mouth typically are described as having oral CD, although the two terms are often used interchangeably. OFG in childhood is most reported in Celtic populations, is usually associated with rapid development of luminal CD, and oral features can often be reduced as follow-up progresses; adult presentation of OFG is less likely to predict development of CD. Isolated perianal disease with granulomas on biopsy is also well recognized, again usually with either silent or rapidly developing luminal CD; in a large North American series, 10% of newly diagnosed pediatric patients with CD had perianal fistulas and/or abscesses at the time of diagnosis. Genital lymphedema with granulomas at skin biopsy is recognized as a metastatic form of CD; it has been the presenting feature with either silent or rapidly developing luminal CD in both boys and girls in 1.5% of a population-based cohort of pediatric CD in the United Kingdom (DC Wilson, unpublished observation).

The presence of granulomas on biopsy from an area of the gastrointestinal tract with nondiagnostic macroscopic CD ( Table 42-3 ) may define that location as having CD, or make the diagnosis of CD in a case that otherwise would be labeled as UC or IBDU, that is, once noncryptolytic granulomas are found, the diagnosis is CD. Granulomas are more common in childhood that in adult Crohn’s colitis at diagnosis, and tend to regress, often being absent in surgical specimens if surgery is later required. The finding of granulomas is also suggested to be associated with a worse prognosis.

Clinical presentation of CD —CD should be suspected when patients appear with appropriate symptoms, which may be very diverse. Bloody diarrhea is the most common presenting symptom in colonic CD as for UC, but CD may also present with vague abdominal pain, diarrhea, unexplained anemia, fever, weight loss, or growth retardation as frequently reported symptoms (see Table 42-3 ). However, the classic “triad” of symptoms—abdominal pain, diarrhea, and weight loss—occurs concurrently in a minority (25%) of CD patients. There is significant overlap for many of the symptoms. From 6% to 23% of children present with extraintestinal manifestations prior to diagnosis with a higher rate in those who are older than 6 years of age. Extraintestinal manifestations were present in 20% of 1178 cases of incident CD in EUROKIDS, a large prospective European (17 European countries and Israel) PIBD registry. The differential diagnosis of the presenting features of CD is shown in Table 42-4 .

Correct diagnosis of colonic CD —It is vital to differentiate colonic CD from IBDU and UC. Key features that point only to the diagnosis of CD include the presence of skip lesions; perianal and/or orofacial and/or genital locations of disease with granulomas on biopsy; the presence of well-formed noncaseating granulomas remote from ruptured crypts anywhere in the gut; the presence of macroscopic inflammation of the upper intestinal tract, in particular, deep serpentine ulcers and cob­blestoning; stenosis/stricturing of bowel (radiologic or surgical)—bowel wall thickening with luminal narrowing; stenosis, cobblestoning, and linear ulcerations in the ileum; or inflamed ileum with a normal cecum ( Table 42-2 ). Aphthous ulcerations and skip lesions are typical of CD but rarely may be seen in UC, especially in young children at diagnosis or during treatment. Both atypical phenotypes of UC and IBDU may be confused with colonic CD. Atypical UC phenotypes at endoscopy and histopathology include rectal sparing, short disease duration, cecal patch, upper gastrointestinal involvement, and acute severe colitis.

Diagnostic Work-Up of Children with Suspected CD

The revised Porto criteria provide recommendations and practice points that are designed to assist the confirmation of the diagnosis of PIBD, assess of disease location and the extent and severity, and recognize complications at diagnosis. Box 42-1 summarizes these with respect to the diagnostic workup for suspected pediatric CD.

Laboratory stool tests— Stool tests must be performed to exclude enteric infections, preferably prior to endoscopic assessment. Efforts to identify bacterial infections should include a stool culture to exclude Salmonella , Shigella , Yersinia , Campylobacter , and Clostridium difficile toxins in all children. Screening for enteric viruses can sometimes be helpful. Testing for Giardia lamblia is recommended in high-risk populations or endemic areas. A more extensive investigation for unusual infectious agents and parasites may be warranted in endemic areas or after travel.

Fecal surrogate markers for detection of inflammation at diagnosis include fecal calprotectin (FC), lactoferrin, S100A12, and lysozyme. A body of pediatric data exists primarily for FC, which is an excellent tool for identifying the presence of intestinal inflammation with high sensitivity. In a recent prospective study, FC was elevated at diagnosis in 95% of 60 unselected pediatric CD patients, whereas only 86% of the patients had an increased C-reactive protein (CRP) and 83% an elevated erythrocyte sedimentation rate (ESR). Although combining blood inflammatory markers may yield higher sensitivity, in one series, 15% of 48 incident IBD cases had no elevation of levels of any of five blood markers (hemoglobin, white cell count, platelets, ESR, and CRP) yet had abnormally raised FC level. FC levels at diagnosis of pediatric CD are superior to blood markers as a diagnostic marker for intestinal inflammation; in a recent large case-control study of FC, the area under the receiver operating characteristic (ROC) curve of FC to diagnose IBD was 0.93, considerably higher than the area under the curve (AUC) of blood inflammatory markers.

However, elevated FC levels cannot distinguish the type of IBD (CD vs. UC) or location of the disease (small vs. large bowel), and may occur in apparently healthy infants and toddlers. The most recent systematic review and meta-analysis of FC for suspected pediatric IBD contained 394 pediatric IBD patients and 321 non-IBD controls, and demonstrated pooled sensitivity and specificity for the diagnostic utility of FC during these investigations of 0.978 (95% confidence interval [CI] 0.947–0.996) and 0.682 (95% CI 0.502–0.863) respectively. Fecal markers may be most helpful in children with nonspecific symptoms (e.g., abdominal pain or nonbloody diarrhea) or signs (e.g., anemia, elevated CRP or ESR) to guide endoscopic workup. They are also useful in patients with extraintestinal manifestations without gastrointestinal symptoms in whom elevated CRP or ESR levels do not help to discriminate between a primary extraintestinal disorder (e.g., arthritis, erythema nodosum in rheumatologic diseases) and the same IBD-associated symptoms. Other fecal tests (e.g., for occult blood or fecal α1-antitrypsin) are not recommended for the routine initial diagnostic workup.

Laboratory blood tests —Results of multiple blood tests may be abnormal in IBD—complete blood count (decreased hemoglobin or elevated total white cell count and platelets), serum albumin (decreased), and inflammatory markers such as CRP and ESR, both of which are typically elevated in active disease. However, normal blood test results do not exclude the diagnosis of CD; data from PIBD registries indicate that at the time of diagnosis, 21% of children with mild CD have normal results for the combination of hemoglobin, albumin, CRP, and ESR. A low serum albumin level may indicate protein-losing enteropathy, and usually reflects disease activity as well as severity and not merely nutritional status. Examination of transaminases and γ-glutamyl transferase (GGT) and an ophthalmic examination should be undertaken to screen for IBD-associated extraintestinal disease such as uveitis and hepatobiliary disease; abnormal test results may, however, be elevated due to other causes. Other laboratory tests such as anti-tissue transglutaminase immunoglobulin A (IgA), which is used to exclude celiac disease, or tests for immunodeficiency disorders are reserved for use in appropriate circumstances and therefore are not recommended for routine use in all patients with suspected IBD.

Serologic tests —No serology pattern can make the diagnosis of CD (or preclude the diagnosis of UC), due to imperfect test performance of all existing antibodies. Although results are often negative, screening for serologic markers of IBD (e.g., anti- Saccharomyces cerevisiae antibody [ASCA] and anti-neutrophil cytoplasmic antibody [pANCA]) may, however, increase the likelihood for IBD in atypical cases if results are positive and may help to differentiate CD from UC. The presence of positive Crohn’s serology (e.g., ASCA+/pANCA-) does not necessarily preclude the diagnosis of UC but reduces the likelihood of UC. ASCA is found more often in CD (50% to 70%) than in UC (10% to 15%) and healthy controls (<5%); these antibodies increase with age (39) and are associated with a more severe disease course in CD. Conversely, pANCA is more common in UC (60% to 70%) than in CD (20% to 25%). The use of newer serologic markers (which include antibodies against Pseudomonas fluorescens –associated sequence [anti-I2], anti-outer membrane protein C of Escherichia coli [anti-OmpC], anti-outer membrane protein of Bacteroides caccae [anti-OmpW], and anti-flagellin antibodies [anti-CBir1]) has yet to enter routine clinical practice due to inconsistent data.

Upper and lower gastrointestinal endoscopy —In nonemergencies, the workup should start with an upper and lower gastrointestinal endoscopy. Ileocolonoscopy with biopsies is the most essential part of the diagnostic workup in PIBD; rectosigmoidoscopy and incomplete colonoscopy are insufficient. It is vital that endoscopy be performed by a pediatric gastroenterologist after age-appropriate preparation, under general anesthesia or deep sedation in a setting suited for children, and by personnel with training and expertise in PIBD. Failure to visualize the terminal ileum has been reported in approximately 20% of cases in experienced large pediatric centers. The added diagnostic yield of ileal intubation including histology is reported to be 13% in PIBD. The European Crohn’s and Colitis Organisation (ECCO) consensus statement on diagnosis and management of CD recommends that “multiple” biopsies from five sites around the colon (including the rectum) and the ileum should be obtained for a reliable diagnosis; a minimum of two samples from each of these six sites should be obtained.

In an audit of diagnostic workups in 1811 pediatric patients with IBD, 35% of the patients with CD had macroscopic abnormalities on esophagogastroduodenoscopy (EGD), and these abnormalities were specific for CD (aphthae, ulcerations, cobblestoning, and stenosis) in 24% of the patients. Microscopic abnormalities on EGD were crucial for the diagnosis of CD in 19 (4.5%) of 428 patients, including the isolated detection of granuloma on EGD in 13 (3%) of 428 patients. Detection of granulomas on EGD alone in pediatric patients with CD ranges from 2% to 21%; upper gastrointestinal endoscopy is particularly helpful in patients with otherwise nonspecific pancolitis.

Ultrasonography (US) —Abdominal US is useful for CD imaging, especially when screening for potential CD, as it is noninvasive, of low cost, and ubiquitously available. US can accurately detect inflammation of the bowel wall plus assess peri-intestinal abnormalities, with a good negative predictive value for CD; pathologic changes of inflamed bowel can essentially be divided into mural (bowel wall—altered echogenicity, loss of stratification, increased color Doppler signal of hyperemia, and decreased or lost peristalsis) and extramural (mesenteric, adipose tissue, and lymph node changes) findings. Although sensitive for detecting lesions of the terminal ileum, US has decreased sensitivity for proximal small bowel lesions and colonic lesions. Interobserver variability remains a major issue; small intestine contrast US (SICUS) may increase the overall sensitivity while reducing interobserver variability in adults; oral anechoic contrast solution (iso-osmolar polyethylene glycol) is used during SICUS, enhancing sensitivity and decreasing interobserver variability.

Magnetic Resonance Imaging (MRI) —MR enterography (MRE) is the preferred test for imaging the small bowel at diagnosis, as it detects characteristic CD changes while estimating both extent of intestinal inflammation and the degree of damage (stricturing and penetrating disease). It has, however, poor sensitivity for subtle luminal disease (no thickening of bowel wall or clearly increased intensity). In MRE, distension of the small-bowel loops is obtained by administering oral polyethylene glycol or sorbitol solution. MR enteroclysis (solution administered by nasoenteric intubation) is minimally superior to MRE yet more invasive; therefore, it is not used routinely in children. Evidence of mucosal inflammation on MRE includes bowel-wall thickening, intensity enhancement of the bowel, engorgement of mesenteric vessels (i.e., comb sign), enlarged lymph nodes, and fatty infiltration of the mesentery. Wall thickening with decreased luminal diameter may indicate stenotic disease, confirmed if prestenotic dilation is visible. A long-standing fibrotic stricture has thickened hypointense bowel wall without significant contrast enhancement; this is unlikely to benefit from medical treatment. Sinus tracts and fistulae appear as fluid-containing tracts with associated peripheral enhancement, and entero-enteric fistulae may form a complex network between closely adherent small bowel loops. MRE is limited by the large volume of solution required; this is so often associated with nausea and vomiting that prophylactic use of ondansetron and early recourse to instillation via nasogastric tube are useful. A helpful alternative protocol that uses only 150 mL total fluid (50 mL lactulose in 100 mL water) has demonstrated high diagnostic utility when compared prospectively with small-bowel follow through (SBFT) and endoscopy/histology. A systematic review of 11 relevant studies with 496 cases of suspected pediatric CD confirmed that MRE is both sensitive and specific for diagnosis, and should supersede conventional fluoroscopy for small bowel imaging in centers with appropriate expertise; meta-analysis of the six comparable studies gave a pooled sensitivity and specificity for MRE detection of active terminal ileal CD of 84% and 97%, respectively.

Pelvic MRI is recommended for evaluation of suspected or proven perianal involvement in CD, allowing definition of both extent and location of perianal fistulas and abscesses, guiding medical/surgical management, and assessing response to the necessary multidisciplinary therapy.

Wireless capsule endoscopy (WCE) —WCE is the best alternative to MRE for investigating mucosal abnormalities of the small bowel in CD, visualizing the entire small bowel with minimal discomfort, and detecting mucosal lesions with a higher sensitivity than MRE. WCE is limited by the need to assess patency of the small bowel prior to the test (imaging or patency capsule), its inability to detect complications, risk of capsule retention, inability to control capsule movement, and the high rate of incidental findings. Contraindications include intestinal strictures, severe disease with systemic features, and children younger than 1 year of age; caution is needed in children with previous abdominal surgery. A specifically designed device enables introduction of the capsule into the duodenum during upper endoscopy if children are unable to swallow the capsule.

In a prospective pediatric controlled study in 20 children with suspected small bowel CD with either normal (n = 15) or nonspecific findings (n = 5) on conventional imaging, WCE use confirmed the diagnosis of CD in 60%. Relative diagnostic yields in a meta-analysis of studies in PIBD ranged from 58% to 72% for WCE, 0 to 33% for SBFT, and 0 to 61% for ileocolonoscopy. If WCE is the only small bowel imaging available for CD, the high number of false positives (10% to 20%, especially with nonsteroidal anti-inflammatory drug [NSAID] use) and the lack of validated diagnostic criteria need to be borne in mind.

Balloon-assisted enteroscopy (BAE) —BAE is indicated only in special circumstances, for example, when all conventional endoscopy, WCE, and cross-sectional imaging do not allow a definite CD diagnosis in patients with high suspicion for small bowel disease. BAE (both double-balloon enteroscopy [DBE] and single-balloon enteroscopy [SBE]) has progressively replaced push and surgically assisted enteroscopy, yet its role in the initial diagnostic workup of suspected CD in childhood remains limited, but with advantages in visualizing lesions and taking biopsies. Successful DBE has been reported to be safe and effective in a review of five pediatric series. Recent studies have shown that SBE can confirm the diagnosis in some cases of suspected pediatric CD where there are only nonspecific findings on endoscopic and histopathologic assessment, and after use of MRE and WCE.

Very Early Onset CD

Both CD and IBD-like monogenic disorders can present with similar clinical, endoscopic, and histopathologic features in infancy and early childhood. It is vital to differentiate between these different diseases with similar, and sometimes near identical, intestinal inflammatory phenotypes but differing treatment pathways; the monogenic disorders are associated with higher mortality rates and morbidity loads, and merit detailed clinical genetic assessment and genetic counseling. Many of the monogenic disorders do not respond to conventional CD medical therapy and resectional surgery may be harmful, yet hematopoietic stem cell transplantation can be curative in some.

The starting point for the differential diagnosis of CD is a high index of suspicion—see Box 42-2 for “alarm signals.” The full diagnostic workup of upper gastrointestinal endoscopy and ileocolonoscopy with multiple biopsies plus small bowel imaging recommended in the revised Porto criteria needs to be supplemented in infantile and early childhood CD by consideration of allergic disorders, immunodeficiencies, and monogenic IBD-like disorders. We recommend acquiring immunoglobulin levels, neutrophil function (by nitroblue tetrazolium [NBT] or fluorescence-activated cell sorting [FACS]), lymphocyte subsets (by FACS), and considering food protein (especially cow’s milk protein) elimination and challenge initially. Furthermore, detailed tests of immunologic function, functional antibody testing, and autoimmunity complemented in some by candidate gene sequencing/next generation sequencing may then be indicated.

Natural History of CD

Although the natural history of a disease conventionally has referred to the progression of that disease in an individual over time in the absence of any specific treatment, we discuss the natural history of pediatric CD as the course of development of CD and its complications in the presence of treatment, and where complications may therefore be caused by CD itself, the therapies used for CD, or a mixture of both. Studies of natural history of CD should ideally be performed in population-based prospective cohorts of large size and with long duration of follow-up; very few studies fulfill these criteria, and therefore most have biases such as selection bias. Incidence and prevalence trends for CD have already been discussed in the epidemiology section, and comparison of phenotype at diagnosis and phenotypic progression with time between pediatric-onset CD and adult-onset CD in the introduction. CD in childhood has a male predominance, whereas adult CD has a female predominance.

Mortality and severe morbidities in CD —These are vital aspects of the natural history of CD. Mortality and development of cancer have only recently been explicitly studied in pediatric CD, although they have been featured in case series and cohort studies of natural history of CD. The natural history of PIBD has been described by Abraham et al. in a systematic review with the search of literature ending in August 2010. The mortality rate of PIBD cases in the population-based EPIMAD registry in Northern France was not any greater than that of the age-matched pediatric population, with six deaths (five CD). In contrast, there was a significantly increased rate of cancer; 9 (6 CD) cases in 698 PIBD cases yielded a standardized incidence ratio of 3.0 (1.3 to 5.9). A multinational study of 20 European countries and Israel by the Porto PIBD group reported 18 cancers (12 CD) and 31 deaths (10 CD); mortality was related primarily to infection, and then cancer and uncontrolled disease. In the earlier systematic review, there were 63 deaths directly related to IBD in 39,719 patient-years’ follow-up, and 6 cancers in 18,270 patient-years’ follow-up for CD. In a study looking at death in patients younger than age of 65, the presence of primary sclerosing cholangitis (PSC) in addition to the IBD was the most significant risk factor.

Cancer risk in CD and CD treatments: Cancer may occur in CD due to IBD itself (e.g., small bowel or colonic adenocarcinoma) or be associated with CD therapy. In pediatric CD, malignancies have been clearly associated with specific treatments; for example, thiopurines have been associated with lymphoproliferative disorders (Epstein-Barr virus (EBV)–associated lymphomas and hepatosplenic T-cell lymphomas). The relative contribution of thiopurines and anti-TNF (tumor necrosis factor) agents in hepatosplenic T-cell lymphoma (HSTCL) remains controversial, as do strategies adopted in response by some (avoidance of thiopurines in males adolescents with CD, using anti-TNF monotherapy or methotrexate for co-therapy with anti-TNF agents). In addition, non-melanoma skin cancers, cervical cancer, and acute myeloid leukemia have been recognized as an adverse event of immunosuppressive medications, including thiopurines and biologics.

Complications and associations of CD: The course of CD can involve both gastrointestinal and non-gastrointestinal complications of CD, as well as the development of extraintestinal manifestations of CD (if they were not present before or at the time of CD diagnosis). These gastrointestinal and nongastrointestinal complications, together with the extraintestinal manifestations, are summarized in Box 42-3 .

Measures of Disease Activity and Outcomes in CD

Both clinical and laboratory evaluations have traditionally been used to assess disease activity in CD. There has been increasing interest in measuring disease activity and outcomes in CD for many reasons, including attempts to induce and then maintain remission at all possible levels (clinical, laboratory, endoscopic, histopathologic, imaging, daily functioning, and quality of life [QoL]), the drive to have mucosal healing as a goal of CD therapies, and the consideration of which outcomes to use in CD clinical trials.

Control of inflammation in CD at the time of diagnosis results in induction of remission, and control of inflammation throughout the disease course prevents relapses and allows successful maintenance of remission without the development of disease-related complications (such as the change from inflammatory to stricturing or penetrating behavior phenotype). CD differs from UC because, as a transmural pan-intestinal disease, inflammation needs to be controlled not just in the mucosa, but also in all layers of the gut wall. The issue of which clinical index most accurately reflects complete control of inflammation in CD is therefore vital.

Disease activity scores —The PCDAI (pediatric CD activity score) has been used clinically and in research for more than two decades, having had recent prospective evaluation and reassessment of its cutoff scores for response, remission, and gradations of disease activity. The CDAI score is the best-known adult CD activity index; the PCDAI differs from the adult CDAI mainly in the addition of growth parameters and laboratory blood tests. The CDAI has been evaluated against the PCDAI for pediatric CD using physician global assessment as the gold standard and PCDAI was shown to more accurately classify CD activity in children. Furthermore, CDAI has recently been shown not to discriminate adult patients with symptoms due to active CD from adult patients with irritable bowel syndrome.

Attempts have been made to modify the PCDAI, with an abbreviated PCDAI (abbrPCDAI) that omits height measurement, presence of extraintestinal disease, and laboratory values. The modified PCDAI (modPCDAI) consists of the three laboratory components of the PCDAI plus CRP; short-form PCDAI (shPCDAI) was created by reweighting PCDAI components of the PCDAI best completed in clinical practice in an IBD registry. Most recently, a mathematically weighted version (wPCDAI) was developed and compared to PCDAI and the three newer versions using prospectively collected pediatric CD datasets. The wPCDAI displayed advantages over the other versions, and has been recommended for use in clinical trials in pediatric CD.

Mucosal healing (MH) —MH in CD is associated with improved medium and long-term outcomes in adult CD, and has been demonstrated in some intervention studies in pediatric CD. Given the nonfeasibility of frequent endoscopic reassessments in children, great hope lies with fecal surrogate markers of MH throughout the childhood CD course, from diagnosis via induction and maintenance of remission and up to transition to adult IBD services.

Endoscopic scores: There are currently no validated pediatric CD endoscopy activity scores available, unlike those used in adult CD.

Fecal surrogate markers —Fecal calprotectin (FC) is the only widely used fecal marker and, as discussed earlier in the diagnosis of CD section, is excellent for identifying the presence of intestinal inflammation with high sensitivity at diagnosis of CD, being superior to blood markers. Given that a normal FC has a high negative predictive value for mucosal inflammation at diagnosis of suspected CD, it would be hoped that data would also show that a normal FC reflects a lack of mucosal inflammation (or MH) in established pediatric CD. Use of FC as a marker of MH after induction therapy, however, reveals a major drawback, that of slow responsiveness. Utility during the maintenance phase of CD therapy can be divided into whether FC is a good marker of MH during the maintenance phase (i.e., no chronic inflammation present) and utility of FC for predicting relapse during maintenance therapy. However, there is a lack of evidence in pediatric CD; in contrast, FC has been shown to be useful in predicting relapse in adult IBD by systematic review and meta-analysis. In a population-based cohort of more than 200 pediatric CD cases, FC was a better marker than either PCDAI or blood markers of chronic CD activity sufficient to require change of thiopurine to methotrexate, anti-TNF agent, or surgery (Wilson DC, unpublished observations).

Blood inflammatory markers —The major problem with blood inflammatory markers such as ESR and CRP is that they do not increase at diagnosis or at relapse in a significant minority of pediatric patients with CD. FC is superior to blood markers as a diagnostic marker for suspected CD; however, in those children where elevated levels do occur, ESR and CRP (unlike FC) are both rapidly responsive, usually by 1 to 2 weeks after commencement of exclusive enteral nutrition (EEN) or corticosteroid therapy in responding patients.

Transmural disease and cumulative bowel damage assessment— MRE has been discussed earlier as the small bowel imaging choice for suspected CD at diagnosis, based on both the presence of strong diagnostic utility and the absence of radiation exposure. Furthermore, its ability to measure both intestinal inflammation and intestinal damage is valuable for the course of pediatric CD beyond initial diagnosis. The progression from inflammatory to stricturing and fistulizing CD behavior will damage the bowel, with inflammatory strictures becoming increasingly fibrous, and fibrous strictures causing poor bowel function more proximally; fibrous strictures and fistulas both can lead to permanent damage by loss of affected bowel at surgical resection.

There are currently no validated pediatric CD MRE scoring systems for inflammatory activity available, unlike those used in adult CD such as the Magnetic Resonance Index of Activity (MaRIA) score. A collaborative international research study, the ImageKids study, is currently developing the Pediatric MRE-Based Activity Index (P-MECAI).

The ability to measure cumulative intestinal damage in CD would allow the natural history of CD to be assessed, both during routine clinical care and as a vital medium to long-term outcome measure in trials of established and novel therapies. Although no MR-based damage score is currently available, the Lemann score is being developed for adult CD. The collaborators of the ImageKids study are also currently developing the Pediatric Crohn’s Disease Intestinal Damage Score (PECDID score).

Quality of life (QoL) scores —The evaluation of quality of life is an important aspect of management of chronic disease in children and adolescents. There are two basic types of health-related quality-of-life (HRQoL) measures: generic and disease-specific. The first PIBD disease-specific QoL instrument—the IMPACT questionnaire—was developed by a multidisciplinary research team in Toronto, and has since been evaluated and used widely, with development of the IMPACT II and then IMPACT III tools, plus widespread international validation, and subsequent use as a secondary outcome in landmark randomized controlled trials in pediatric CD.

Etiology

Etiology —What is driving this rise in incidence of pediatric CD? Lifestyle changes, such as the increased consumption of low-fiber, high-sugar diets, plus rising obesity rates have been suggested as possible etiologic contributors. Hypotheses highlighting dysbiosis of the gut microbiota include the cold chain hypothesis and the hygiene hypothesis. Such changes within well-defined populations in the last decades merit further exploration. Associations between antibiotic use in early childhood and development of CD in early childhood have been demonstrated in several studies. As with other diseases such as multiple sclerosis, sunlight exposure may also be pertinent, and work showing that vitamin D deficiency may play a role in immune disease development is certainly of interest. The latest IBD genetic research has highlighted the importance of the internal bacterial environment, and there is a current explosion of interest in the gut microbiota of humans. A possible mechanism to link genetics, immune system, and environment in a disease with rapidly rising incidence is that of epigenetics, well described in cancer but only recently being evaluated in common polygenic immune conditions such as CD. Future research to unravel the potential etiologies of CD is likely to focus closely on children, as this group is considered to be influenced less by exogenous factors such as diet, smoking, and medications, plus more accurate parental recall of environmental changes.

Serology

The presence or absence of immunologic markers has been used to categorize patients with IBD into different clinical subtypes but generally does not help in IBD diagnosis itself. The most commonly studied antibodies are Anti-neutrophil cytoplasmic antibody (ANCA), anti- Saccharomyces cerevisiae antibody (ASCA), an antibody to the Escherichia coli –related outer membrane porin C (anti-OmpC), and anti-Cbir1 antibody (antibody against flagellin). More recently, anti-glycan antibodies have been explored in pediatric patients with IBD, but their utility is only now being explored in further pediatric datasets.

pANCA associated with IBD is produced by mucosal B cells responding to various antigens. Although present in the serum of 60% to 70% of patients with ulcerative colitis, pANCA is also detected in 15% to 25% of patients with Crohn’s disease. Patients with Crohn’s disease who are pANCA-positive exhibit more of an ulcerative colitis–like phenotype with colonic disease and histopathologic expression, similar to that of ulcerative colitis.

Levels of ASCA (IgG or IgA) are detected more commonly in patients with CD than in patients with ulcerative colitis or healthy controls. The presence of high-titer ASCA in the absence of ANCA is highly predictive of Crohn’s disease. Higher levels of ASCA correlate with the presence of small bowel involvement, as well as with complicated disease and need for surgery. ASCA positivity has also been noted to be higher in unaffected IBD family members.

Antibodies to I2 (Crohn’s disease-associated bacterial sequence) and OmpC are associated with a greater number of strictures and internal perforations. Anti-Cbir1 is associated with Crohn’s disease, especially in patients with internal penetrating disease, fibrostenotic disease, and small bowel disease. Positivity to antibodies does seem to be important and manifests clinically as increasingly complicated CD increases in children in relation to escalating number and magnitude of these immune reactivity antibodies.

Serology has been used in attempts to characterize patients with IBDU to prognosticate about disease evolution to CD, ulcerative colitis, or staying as IBDU. Over time, most patients with IBDU evolve to a diagnosis of CD or UC, but some do not, particularly those who are negative to both pANCA and ASCA. In studies that then added anti-OmpC and anti-I2 serology, the predictive value of serology was not significantly improved.

Genetics

The genetics of IBD and CD, in particular, has undergone significant scientific advances in recent year thanks in part to changes in technology, which allows faster, less-expensive, and more detailed analysis together with the formation of consortia allowing vast numbers of patients and controls to be genotyped at any one time.

The cornerstone of the rapid genetic advances was the original family-based IBD genetics studies. The original twin-pair studies were followed by studies in first-degree relatives and wider family members, although it has been recently suggested that these original estimates of heritability were perhaps overestimates. It still remains true, however, that the single greatest risk factor for the development of CD is having a first-degree relative with the condition, with siblings of a patients with CD having an estimated 5% to 10% lifetime risk of CD. This increased risk continues into second- and third-degree relatives. Table 42-5 shows the relative risks of disease development for different family members.

Increased intestinal permeability and subclinical intestinal inflammation have been demonstrated in healthy relatives of patients with CD, suggesting a possible inherited defect with variable penetrance between individuals. This has been matched by demonstrating that relatives have increased numbers of disease-associated mutations, although most still do not develop signs of clinical CD.

The original genetic studies, which have now been largely superseded by more recent studies, used labor-intensive relative-pair linkage analysis to identify areas of disease susceptibility; this identified six chromosomal regions (loci) associated with an increased risk of IBD. This approach successfully identified the first and, to date, best-replicated gene for CD on chromosome 16—(nucleotide-binding oligomerization domain) NOD2 / (caspase activation recruitment domain) CARD15 gene. Three main polymorphisms of this gene increase susceptibility to CD (R702W, G908R, 1007fs). Having one copy of the risk allele confers a small increased risk (two- to fourfold), whereas having two copies increases the risk of developing CD 20- to 40-fold. There have been clear genotype–phenotype relationships described for patients carrying NOD2/CARD15 mutations, with increased likelihood of ileal disease, stricturing disease, and the need for surgery.

Following from these earlier studies, subsequent genome-wide association studies (GWAS) analyzed much larger numbers of individual patients (no need for relative pairs) and compared their genetic profiles with healthy controls. These have multiplied the number of susceptibility loci from the six described in the original genetic studies to well over 100. These studies have highlighted the key importance of the autophagy pathway in disease pathogenesis together with defects in both the adaptive and innate immune systems. These GWAS replicated the central role of NOD2/CARD15 in CD risk but also identified other genes strongly associated with CD, including an autophagy gene, ATG16L1 , the gene encoding the interleukin-23 (IL-23) receptor subunit, and IL-12B , IL-10 , and signal transducer and activator of transcription 3 ( STAT3 ) genes. ATG16L1, similar to NOD2/CARD15, affects the intracellular processing of bacterial components. A separate GWAS in children and young adults reproduced 23 of the 32 loci implicated in the large meta-analysis GWAS of adult-onset CD, suggesting that the majority of genes predisposing to the development of CD are shared between children and adults. The largest GWAS done to date in IBD involving more than 20,000 adult and pediatric patients with CD has interestingly highlighted the key role of genes in processing of the microbiome response as well as overlap with genes involved in primary immune deficiencies.

There has been a clear description of a significant crossover between earlier onset IBD (younger than 6 years of age) and genes associated with recognized and newly recognized immunodeficiencies. In particular, defects in the-IL10 and IL-10 receptors have been described in infants with IBD-like illness, often with systemic inflammatory features. Often these conditions are refractive to treatment but respond well to stem cell transplantation. These conditions, which are individually rare, are being discovered with increasing frequency. Whole exome sequencing is used to understand mutations at an individual pediatric patient level, thereby facilitating management of CD.

Microbiome

The current paradigm of IBD pathogenesis suggests an exaggerated immune response to the luminal microbiome in genetically susceptible individuals, with even the most recent genetic studies pointing to the central role that the microbiome plays in disease pathogenesis. The normal gastrointestinal tract is a complex system of microbes that are crucial in maintaining normal luminal health; within this community, the microbiome gene pool is several magnitudes larger than the whole human genome. A complex interaction of gut mucosal immune mechanisms with the intestinal flora is critical in the development and probably the persistence of inflammation in CD. The likely microbial biology of CD is a reduced bacterial diversity with consequent reduced microbiome gene diversity and loss of function. The microbiome of a patient with CD is characterized by a “dysbiosis,” with less diversity but an increased amount of the total number of bacteria compared to healthy controls. As part of this dysbiosis, there is a shift from the “friendly” balance of microbes to an increase in the number of pathogenic microbes. The exact proportions then of phyla that occur with these changes have not been consistently replicated between studies: Firmicutes have been shown to be reduced in some but not all studies, and similarly Bacteroidetes have been suggested to increase in some but not all. There does seem, however, to be an increase documented in the phylum Proteobacteria that contains populations of E. coli and Campylobacter concisus , two individual bacterial species of increasing potential importance in CD etiology (see below).

There are also documented changes in specific species that vary between studies, but the precise overall importance of each species remains a topic of ongoing investigation, for example, several studies have supported a beneficial effect of Faecalibacterium prausnitzii , with increased numbers associated with a reduction in the relapse rate after surgery, in addition to supportive animal models. This view is contrasted, however, with a reduction in total F. prausnitzii numbers in patients who respond to enteral nutrition as well as CD diagnosis in children.

Another species that has attracted interest for a number of reasons is E. coli , particularly strains with properties that enable adhesion to epithelial cells and subsequent cell invasion, termed adherent-invasive E. coli (AIEC). AIEC have been found in higher numbers of patients with CD than patients with ulcerative colitis or healthy controls. The ability of AIEC to penetrate epithelial cells and replicate in macrophages, driving a TNFα inflammatory response, has been demonstrated in patients with CD, including children, indicating that they may be a contributory factor in disease pathogenesis in some patients. Interesting translational work has shown that the normal passage of AIEC across M cells can be reduced by soluble fibers like plantains, providing a plausible link between differing fiber intakes and their role in CD.

C. concisus has also been suggested as a contributing risk species in patients with CD, with properties of adherence and epithelial cell invasion similar to that of AIEC. Initial studies demonstrated culture of the organism from a pediatric patient with CD; further studies went on to show higher numbers of C. concisus from pediatric patients with CD compared to controls. However, in a study that looked specifically at the presence of C. concisus at disease initiation in children, no differences were identified between patients with CD and healthy controls, suggesting no role in initiation of disease but that a role in the perturbation of inflammation in CD is still possible. An important similarity between AIEC and C. concisus is the pathogenic trait of cell invasion, survival, and proinflammatory effect. This may be exacerbated in CD patients with defective intracellular defense machinery, particularly with autophagy, a key process identified in CD genetic studies, as discussed in the previous section.

Immunology

In the normal gut mucosa, there is a dynamic interplay between appropriate response to mucosal immune stimulation and downregulatory forces that mitigate accompanying tissue injury by the gut mucosal immune system. This balanced relationship results in “physiologic” inflammation ( Figure 42-1 ). Defective barrier function (genetically induced or secondary to mucosal injury) may allow increased permeability to bacterial or other antigens, leading to unchecked stimulation of local immune cells. Defective downregulation of an appropriate response to luminal stimulation may also result in more exuberant inflammation, resulting in the tissue injury characteristic of CD.

Variable responses to environmental triggers (e.g., viral infection, nonsteroidal anti-inflammatory drugs) in genetically susceptible and normal hosts. In the normal host, injury leads to self-limited acute inflammation. In genetically predisposed hosts, the inflammatory response continues, leading to chronic inflammation and tissue damage.

(Adapted from Sartor (2004). )

Characterization of the mucosal “immunotype” in CD comprises changes in both the innate and adaptive immune response. Although these are discussed separately, in reality there is a dynamic connection between both arms of the immune system occurring concurrently during periods of inflammation in CD. A number of innate immune system changes have been described in CD including defective neutrophil function, macrophage processing differences, bacterial recognition and processing defects (see NOD2 and autophagy in the genetics section), increased numbers of abnormal Paneth cells, and defective defensin production.

The mucosal adaptive inflammatory immune response in the gut has been characterized by the designation of T helper cell (Th) 1, Th2, Th17, and T regulatory (Treg) pathways. The Th1 and Th17 pathways are important in the pathogenesis of CD. Th1 cytokines (TNFα, IL-12, and IL-18) are proinflammatory cytokines that induce differentiation of CD4+ T cells to the Th1 phenotype. TNFα, in particular, appears to have great importance, as it promotes chemokine secretion from intestinal epithelial cells, disrupts the epithelial barrier, and promotes apoptosis of intestinal epithelial cells. The clear success of anti-TNF therapies in the treatment of CD highlights the importance of this cytokine, although its complete mechanism of action is more complex than simply blocking TNFα and affects components of both the innate and adaptive immune systems. Animal models of CD have concentrated on CD4+ T-cell models, and interesting work has characterized CD8+ T-cell profiles that have been identified in other autoimmune diseases as also important in the prognosis of CD, although these results do require replication in a prospective clinical trial.

Th17 cells have derived the name by their ability to produce IL-17, but they are also involved in the production of several other cytokines. Th17 pathway cytokines also play a role in the protection of the host against extracellular bacteria and fungi by recruiting neutrophils (IL-17) together with inducing production of anti-microbial peptides and enhancing the epithelial barrier (IL-22); they may also interact with B cells. The Th17 pathway has been associated with IBD including CD (see Figure 42-2 ). IL-23R gene polymorphisms that influence IBD susceptibility have been identified in patients with CD and validated in multiple different populations including pediatric patients. Animal models show that suppression of the Th17 axis alters IBD expression in these diseases with translational studies of ustekinumab (a human IL-12/23 monoclonal antibody) showing effectiveness in a subset of patients with CD, although no published pediatric data are available.

Mucosal immunity and cytokine profile in normal mucosa and in patients with irritable bowel disease (IBD) with predominance of Th1- and Th17-mediated response.

(Adapted from Sartor (2008). )

Multiple other cytokines (IL-1, IL-6, IL-7, IL-11, and IL-15) have been implicated as possible participants in the mediation of chronic inflammation in CD.

Clear correlations exist between CD and other immunodeficiency states, for example, the recognition that genetic findings in patients with chronic granulomatous disease have also been demonstrated in patients with CD. This has been supported by the observation of Crohn’s-like disease in other immunodeficiency states such as glycogen storage disease type Ib, leukocyte adhesion deficiency, Chediak-Higashi syndrome, and a variety of neutropenias.

Therapy

The goals of therapy for children and young people with CD are changing. The recognition of the long-term evolution from inflammatory to stricturing and fistulizing disease originally described in adult studies but replicated in pediatric studies has been one reason for this, with current treatment strategies aimed to reduce or prevent these complications. There has also been a shift away from merely providing symptomatic relief for patients to a more critical and detailed assessment of response to therapy to promote intestinal healing in the hope of sustaining longer term disease remission, to mirror the effect of achieving mucosal healing from adult studies. The suggested key considerations when treating children and young people with CD are outlined in Figure 42-4 .

Individual treatment may be straightforward or relatively complex but to the extent possible should focus on the treatment needs of the individual patient. Before changing or escalating treatment, patients should be reassessed to ensure the presence of inflammatory activity to avoid overtreatment and to identify disease progression such as stricture formation, which may have a different therapeutic pathway. The overlap of irritable bowel syndrome in patients with CD is significant and clearly does not need a change of background anti-inflammatory treatment but merely symptomatic treatment of the irritable bowel syndrome (see Box 42-4 ). To highlight this point, retrospective analysis of the ACCENT 1 study (evaluating infliximab response in adults with CD) demonstrated that around one in three patients who entered the trial did not have evidence of active inflammation.

Maintenance Therapy

The use of partial enteral nutrition (PEN) to maintain longer-term remission in children has been supported by a few small observational studies. There are published studies in adults supporting the use of PEN including a randomized controlled trial comparing PEN to thiopurines showing equivalent outcomes at 2 years of follow-up.

Practical advice for delivering EEN is shown in Box 42-5 .

Box 42-5

Practical Advice for Using EEN

• •
  

  Can be used as treatment for active Crohn’s disease (CD)

  
  
• •
  

  6 to 8 weeks of EEN with a polymeric feed

  
  
• •
  

  EEN works best with involvement of the MDT

  
  
• •
  

  If not tolerated by the oral route, use a nasogastric tube

  
  
• •
  

  Supplementary partial enteral nutrition can be used as a maintenance option after a course of EEN for selected patients

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related posts:

Eating Disorders in Children and Adolescents Developmental Anatomy and Physiology of the Esophagus Maldigestion and Malabsorption Atresia and Stenosis of the Bowel Stomas of the Small and Large Intestine Total Pancreatectomy with Islet Autotransplantation and Pancreatic Allotransplantation

Stay updated, free articles. Join our Telegram channel

Join