Classification of Inflammatory Bowel Disease in Children

Fig. 15.1

Evaluation of child with suspected IBD. Atypical UC is a new IBD category consisting of 5 phenotypes and reflects a phenotype that should be treated as UC. IBD-U may be entertained as a tentative diagnosis after endoscopy and can be used as a final diagnosis after imaging and a full endoscopic workup. UC is divided into UC and atypical UC. CD Crohn’s disease, EGD esophagogastroduodenoscopy, FM fecal marker, IBD inflammatory bowel disease, MRE magnetic resonance enterography, UC ulcerative colitis, WCE wireless capsule endoscopy (Reprinted from Levine et al. [7], Fig. 1, page 797, with permission)

The typical features of classic CD are described elsewhere in this book, but there would be no dispute on diagnostic classification for patients with cobblestoning of the small bowel mucosa, skip lesions, perianal disease (fistulae, abscesses, or large inflamed skin tags), and the presence of complicated disease behaviors such as stricturing or penetrating disease. Involvement of the small intestine with reliable interpretation of imaging would also point to a diagnosis of CD. However, particularly in young children, colonic involvement in CD may be continuous, making it difficult to distinguish from UC, and consideration for the diagnosis of IBDU should be applied. Figure 15.2, from the ESPGHAN revised Porto criteria, provides criteria to assist the treating physician when assigning the label of CD, UC, or IBDU.

Fig. 15.2

Diagnostic characteristics associated with different types of IBD (Reprinted from Levine et al. [7], Table 3, page 799, with permission)

Phenotypic Classification of IBD: A Historical Perspective

For decades, CD was considered a relatively homogeneous condition, without attempts to further subclassify the phenotype. In 1975, Farmer et al. [36] were among the first to recognize that CD is not a homogeneous entity, hypothesizing that sites of inflammation influenced outcomes and disease behavior. The group categorized CD into (1) ileocolonic, (2) small intestinal, (3) isolated colonic, and (4) isolated anorectal disease. The authors attempted to correlate clinical symptoms at presentation with disease location in their clinical cohort and described the evolution of disease over time including the development of rectal and internal intestinal fistulae, growth impairment, intestinal obstruction, and need for surgery. By using categories of disease location, the authors provide some of the earliest data on the potential relationship between phenotype and clinical outcomes and recognized that such correlations might facilitate therapeutic decisions for these patients.

Further consideration toward a phenotypic classification of CD came from Greenstein et al. [37] who described two disease behavior patterns, perforating and non-perforating, using a cohort of 770 patients undergoing surgery. Site of inflammation (categorized as ileitis, colitis, or ileocolitis) was associated with type of surgery with patients with ileal disease more likely to require surgery for obstructive symptoms in comparison with ileocolonic disease where fistulizing disease was the main surgical indication.

These observations that disease location and behavior influence outcomes became the basis of the more rigorously developed Vienna, Montreal and Paris classifications for IBD [6, 38, 39].

The Vienna Classification

Between 1996 and 1998, an international working group was established to develop and validate a phenotypic classification for CD [38]. The final included categories were age at diagnosis (<40 or ≥40 years), disease location (terminal ileum, colonic, ileocolonic, or involvement of the upper GI tract), and disease behavior (non-stricturing non-penetrating, stricturing, or penetrating). While great efforts were made to develop a reproducible and validated phenotypic classification, there were some limitations. The Vienna classification cannot distinguish disease location when disease is present in both the upper GI tract and other intestinal regions or when it occurs in isolation. Likewise, perianal disease is not considered a separate category; rather it is categorized as “perforating” disease behavior making it impossible to distinguish whether a patient has perianal disease, internal intestinal fistulizing disease, or both.

The Montreal Classification

The Montreal classification was developed to provide a uniform system of designating subgroups of patients with IBD, with the aim of facilitating multicenter genotype-phenotype correlation studies. It is the most commonly used classification for adult patients. Unlike the Vienna classification and its predecessors, which focused on CD classification, the Montreal classification includes a phenotypic classification for UC and makes recommendations for assigning the diagnostic label of “inflammatory bowel disease-type unclassified” (IBDU) [39].

For CD, modifications to the Vienna classification included (1) an additional category to classify children diagnosed at ≤16 years of age, (2) allow for upper GI tract disease to be classified independently of ileocolonic and colonic involvement, and (3) classify perianal disease as a category independent of the “penetrating disease behavior” category.

The group proposed that the maximal disease extent prior to first resection in those undergoing surgery be used when considering the variable “disease location.” Given the propensity for disease behavior to evolve over time [40], the recommendation of the working group is to wait a minimum of 5 years before definitively assigning a disease behavior for the non-stricturing, non-penetrating category, particularly when data are used as part of research studies.

For UC, the group proposed that patients be classified according to maximal extent of inflammation at any time during follow-up. Maximal disease extent is E3, which denotes any disease extending proximal to the splenic flexure.

When a principal diagnosis of UC or CD cannot be established, the group recommends that the term colonic “IBD-type unclassified (IBDU)” be assigned and the term “indeterminate colitis” be reserved for use only after colectomy has been performed, when features of both CD and UC coexist.

The Paris Modification of the Montreal Classification for Pediatric IBD

In 2009 an international group of pediatric IBD experts took on the task of modifying the Montreal classification, to capture aspects of disease phenotype that are pertinent to pediatric patients. Following an extensive review of the literature, with attention focused on pediatric data, where available, and including recommendations from expert opinion and narrative review, the Paris classification of pediatric inflammatory bowel disease was published in 2011 [6]. The committee also reviewed, and was in agreement with, the 2007 paper put forward by NASPGHAN and the Crohn’s and Colitis Foundation of America (CCFA), which provided recommendations for differentiating UC from CD [2].

Novel Features of the Paris Classification

1.

A new age category, allowing differentiation between patients presenting prior to or after their 10th birthday, was introduced. The new classification proposed that children presenting prior to 10 years of age are designated to the age category A1a and those presenting from 10 to 17 years of age be assigned to the category A1b. Disease location in CD at diagnosis appears to be different in these two age categories with the younger group being more likely to have isolated colonic disease rather than ileal involvement. Ileal disease (whether isolated or in conjunction with disease in other locations) is more common in the older age group [5, 15, 41, 42]. For ulcerative colitis, the Pediatric Inflammatory Bowel Disease Collaborative Research Registry found the youngest patients, diagnosed aged 1–5 had greater use of mesalamine and thiopurines [43]. A Canadian population-based study found lower colectomy rates in children diagnosed under 10 years of age [44].

2.

The Paris group recognized that the Montreal classification did not optimally describe disease location, particularly regarding the upper GI tract category (L4), which is unable to distinguish between disease of the small intestine and disease proximal to the ligament of Treitz. The Paris classification recommends that the presence of upper GI tract disease only be assigned in the presence of macroscopic disease as there is no literature to suggest that histologic involvement alone influences disease progression or phenotypic classification over time. The presence of mucosal erythema or granularity is not sufficient to be considered as macroscopic disease. The Paris group subdivided the L4 Montreal category for upper GI tract disease into L4a (denoting disease proximal to the ligament of Treitz) and L4b (denoting disease distal to the ligament of Treitz).
3.

Disease behavior is inflammatory (B1) at diagnosis for the majority of patients, but may evolve into a more complicated phenotype, stricturing (B2), or penetrating (B3) over time. Some patients have a complicated disease phenotype at presentation. The Paris classification also allows capture of patients who have both concomitant stricturing and penetrating behavior (B2B3 category).
4.

Since the majority of pediatric patients with UC have extensive disease at presentation (in comparison to adult patients, many of whom have less extensive disease), the Paris classification includes an additional category for disease extension proximal to the hepatic flexure (E4) [14, 15, 42, 45].
5.

Regarding UC disease behavior, the Pediatric Ulcerative Colitis Activity Index (PUCAI) [46] is used to determine whether or not a patient has ever had severe disease (PUCAI ≥ 65) as studies have found that colectomy rates are higher in patients who have had severe disease [47].
6.

Pertinent to pediatric patients is the ability to capture growth impairment when classifying disease [48, 49]. A growth category was introduced which allows normal growth at diagnosis and over the course of follow-up (G0) or impaired linear growth, using height velocity Z-scores (G1) at any time point, to be captured. Z-scores should be adjusted for age (or bone age when delayed) and sex.

The Paris group described a list of clinical features, which when present do not support a diagnosis of UC:

1.

Perianal disease
2.

Microscopic skip lesions
3.

Stenosis, cobblestoned mucosa, and linear ileal ulcers (even in the setting of pancolitis)
4.

Macroscopic inflammation of the ileum in the absence of cecal inflammation
5.

Presence of a well-formed granuloma at a site that is not adjacent to a ruptured crypt
6.

Absolute rectal sparing (no macroscopic or histologic features of inflammation)

The group advised that the finding of a few small ulcers in the small intestine during capsule endoscopy should not preclude the diagnosis of UC (if other features point to this diagnosis) since these may be nonspecific and are sometimes seen in healthy people.

Evolution of Disease Phenotype

Disease phenotype in both adult and pediatric patents is not static [15, 40, 50, 51]. This represents a challenge when phenotyping both adult and pediatric patients. When the Montreal classification was proposed, the authors recommended waiting for 5 years, or until the time of surgery (whichever was sooner), before assigning a disease behavior. However, in reality, increasingly pediatric patients are being entered into prospective registries requiring a phenotypic classification at diagnosis be assigned. It is important that such registries have the ability to capture the evolution of disease phenotype such as disease extension in UC as well as CD, in addition to change in behavior to a more complicated phenotype (stricturing and/or penetrating) in CD. It is important to remember that IBD phenotype may evolve over time particularly in children [15, 42, 52, 53]. A population-based study determined that the diagnosis (CD, UC, or IBDU) of children <6 years changed in 16.2% of cases, compared with 18.1% of children diagnosed 6–10 years, and 13.8% of children diagnosed ≥10 years (P = 0.007) [44]. In addition, adult CD studies have demonstrated a change from non-stricturing to stricturing disease in 27% and penetrating disease in 28% [40]. In UC, other population-based studies demonstrated that more than half of the patients with initial proctosigmoiditis eventually extended more proximally, but 75% of extensive cases regressed [54].

A summary of the Paris classification of pediatric IBD is represented in Fig. 15.3.

Fig. 15.3

Paris classification of inflammatory bowel disease (Adapted from Levine et al. [6])

Future Directions

Classification of IBD has progressed well beyond the classic categories of CD, UC, or indeterminate colitis. The advent of the Montreal classification and the subsequent Paris modification for pediatric IBD have allowed for granularity in the description of IBD phenotype for both clinical and research purposes. While these classification systems have been extremely valuable, disease location and severity do not tell the full story of IBD. Some have suggested adding histology to the macroscopic description of the Paris modification to improve granularity and descriptive capability [55]. In addition, with advances in characterization of the role of gene, environment, and microbiome interactions and their effect on disease phenotype, we anticipate that future classification systems will incorporate new disease location categories, histology findings, protein expression characteristics, and other factors. In the era of personalized medicine, treatment choice will depend on many more factors than those described in current classification systems. We anticipate that definition of the inflammatory bowel diseases will consist of a continuum of categories resulting in even more power to describe the disease characteristics of a child with IBD.

Acknowledgments

We are grateful to Dr. Athos Bousvaros who authored the first edition of this chapter and allowed its use as the basis for this new edition. Eric Benchimol is supported by a New Investigator Award from the Canadian Institutes of Health Research, Crohn’s and Colitis Canada, and the Canadian Association of Gastroenterology.

References

Griffiths AM. Specificities of inflammatory bowel disease in childhood. Best Pract Res Clin Gastroenterol. 2004;18:509–23. doi:10.1016/j.bpg.2004.01.002.CrossRefPubMed

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