The term “recurrence” in oncology denotes that cancer has recurred after some time during which it could not be detected. Postoperative recurrence (POR) in Crohn’s disease (CD) is not truly recurrence as the original disease persists microscopically because most of the resections are limited and not radical as in oncological practice. This later manifests as endoscopic/clinical/surgical recurrence. So, recurrence in postoperative CD is primarily “relapse” or “persistent CD” rather than true recurrence. Among the histological features predicting recurrence in postoperative CD, the most commonly studied factor is positive resection margins. Microscopic positive resection margin increases the risk of clinical, surgical, and early recurrence (<12 months) by 2-, 9-, and 6-fold, respectively (Table 1) [1]. Similarly, a retrospective study has shown that the inclusion of mesentery at ileocolonic resection in CD reduces the surgical recurrence rate substantially (40% to 2.9%) [2]. However, the study results are confounded by different follow-up times and concerns regarding a higher risk of postoperative complications due to vascular compromise [3]. Hence, this is currently studied in randomized controlled trials (RCTs) [4]. Myenteric plexitis (MP) is associated with a 30% higher risk of clinical but not surgical recurrence. Granulomas in resected mesenteric lymph nodes and in the intestine were associated with a higher risk of both surgical and endoscopic recurrence [1]. Apart from these factors, depth of infiltration (e.g., transmural involvement) can also be predictive of POR. We incorporated all these factors in our proposed Crohn’s primary site, nodes, mesentery (CNM) staging to systematically evaluate and histopathological report resected surgical specimens in CD. This could potentially help to risk stratify postoperative CD and plan controlled trials studying histological factors for predicting POR in CD along with other clinical predictors.

For ease of classification, under the heading “C” we have included predictive factors which are within the intestine wall like resection margin, inflammation in the neural plexus (plexitis), granuloma and depth of infiltration. Under heading “N,” we have included factors within lymph nodes (e.g., granuloma) and in “M” we have included mesenteric involvement (Table 2).

The authors realized the need for a uniform classification in postoperative CD based on histopathological examination of the resection specimen which could be helpful to predict recurrence. This was based on the realization that recurrence in CD might not be true recurrence as in oncological resections as it was increasingly evident that microscopic persistence of disease at the margins of resection increase the possibility of endoscopic, clinical, and surgical recurrence in CD [1]. Recurrence in CD is multifaceted and occur in the sequence of endoscopic, clinical and surgical recurrence with a strong correlation between each of them (Fig. 2) [35]. Recent studies have shown that radiological evidence of recurrence could precede endoscopic recurrence as endoscopy only evaluate the mucosal aspect of transmural CD [36]. It became increasingly evident that factors other than active microscopic disease margins like depth of infiltration (transmural and extramural), granuloma and plexitis at Crohn’s primary resection site could predict recurrence [7,12,15,19,25]. Apart from the primary site of involvement, involvement of the adjacent structures like nodes (specially granuloma) and mesentery predict recurrence. This could be due to the fact mesenteric adipose tissue and lymph nodes help in mediating immunological, inflammatory and fibrotic signals in the intestine [2]. This is substantiated by the fact that circulating lymphocytes are incapable of producing recurrent lesions after a small bowel transplant in CD if mesenteric lymph nodes were removed fully along with the whole bowel [37]. As conceptualized, the authors reviewed the literature for all the relevant literature on the histological prediction of recurrence in CD. Previous pathology reports of postoperative patients with CD were reviewed and most of the above -mentioned factors were described except for plexitis. With the help of an expert pathologist and based on published literature, we developed the various components of the classification. The CNM staging was developed keeping in mind all these factors with emphasis on simplicity and practical applicability without putting an additional burden on the reporting pathologist.

The effectiveness of postoperative prophylactic therapy in CD in the future would depend on risk stratification based on clinical as well as histopathological predictors. The POR in CD is a complex phenomenon affected by a multitude of factors other than clinical and histologic predictors such as luminal microbiota (depletion of butyrate-producing Lachnospiraceae and abundance of Enterobacteriaceae) and cytokine profile (interleukin 6 and interferon-gamma) [38,39]. A complete “pathophysiological excision of Crohn’s” removing microscopically involved bowel along with mesentery and draining lymph nodes can reduce the risk of POR. Systematic reporting of such factors in the pathological report of surgical resection based on the proposed CNM classification would enable the study of the histopathological factors systematically.

Examples of staging according to the CNM system are given in Table 2. The capital letters signify Crohn’s primary site, draining mesenteric lymph nodes, and mesenteric involvement. In “C” category (Crohn’s primary site), there are 4 subcategories designated by capital letters (GRIP): G, granuloma; R, resection margin; I, infiltration depth; P, plexitis. In these subcategories, increasing involvement is designated by the addition of a number (1, 2, 3, 4) and various types within the subcategory are designated by capital letters (A, B, C) following the number (Table 2). Final C staging is based on the presence increasing number of the following factors: granuloma, positive resection margin, moderate to severe plexitis or transmural involvement (Table 2). Granuloma (G) is kept at the beginning so that the “GRIP” mnemonic is easy to remember. In the “N” (node) category, Nx indicates that regional lymph nodes cannot be assessed, N0 indicates no granulomas in resected regional lymph nodes, N1 indicates granulomas in the resected lymph nodes. In “M” category, M0 denotes no involvement of adjacent mesentery in the form of thickening of the surface mesothelium/submesothelial/interlobular connective tissue with or without the presence of connective tissue septations. M1 indicates mesenteric involvement microscopically without grossly visible fat wrapping or mesenteric thickening. M2 indicates grossly visible fat wrapping or mesenteric thickening. M2 can be further divided into A, B, C according to the mesenteric disease activity index [2].

For example, a 25-year-old female with 8 years history of CD, who underwent right hemicolectomy 9 months back presented with diarrhea and abdominal pain. Hence, she had an early (<12 visible months) clinical and endoscopic recurrence as colonoscopy revealed deep ulceration at the anastomotic site as well as neoterminal ileum (along with wall thickening in mid ileum on computed tomography: computed tomography enterography). She was started on azathioprine immediately postoperatively. Later biologics were started due to early recurrence along with oral steroids. A review of her operative specimen showed no granuloma in the resection specimen (G0), positive resection margins (R1), transmural infiltration (I3), no plexitis (P0), no nodal granuloma (N0), and gross mesenteric involvement (M2). So, the final staging would is C2 N0 M2.

Apart from the above mentioned histological factors predicting POR of CD, other factors such as microbiota recolonization after surgery, gene expression profile of ileal mucosa and various blood-based biomarkers are shown to predict POR in CD [40-42]. Fusobacteria is the most important microbial driver or POR, whereas expression of JAK-STAT at the ileal margin and mitochondrial dysfunction inside foci of chronic inflammation is associated with a higher probability of POR [40,41]. Markers of memory T cell response like CXCR3 ligands are also associated with higher POR in CD [42]. Future staging systems may incorporate these factors for prediction of POR in CD.

There is growing evidence that many histological factors in the resected intestinal specimen in CD (C), mesenteric lymph nodes (N) and mesentery involvement (M) predict POR in CD. Most of the evidence is based mainly on retrospective studies without adjusting for confounding factors like disease severity and postoperative prophylaxis. Hence there is a need for future prospective studies evaluating all these factors. However, such a study is challenging given the various factors that need to be studied. We have attempted to incorporate all the histologic factors in the proposed CNM classification to enable systematic reporting which will help design such studies. However, the individual weightage of each component of the CNM classification needs to be evaluated based on validation using multivariate analysis in future prospective studies. Until such studies are published, further refinements of the classification are warranted based on expert consensus development before its routine clinical use [43,44]. Nevertheless, the CNM classification is the first attempt to systemically report and study these predictive factors for POR. The CNM classification along with existing clinical predictors (e.g., smoking, history of intestinal resection) could help to predict recurrence effectively and more accurately. Moreover, like the TNM classification for malignancy, the CNM classification for CD has the potential to not only predict “recurrence” systematically but also influence postoperative surveillance and prophylactic therapy.