Abstract
Background/Aim: This study aimed to validate the inter-rater agreement of the assessment of crypt branching (also called crypt fission) in patients with ulcerative colitis (UC) and to elucidate its potential diagnostic and prognostic impact. Patients and Methods: A total of 100 biopsies from patients with UC were analyzed. Two cohorts of 50 patients each served as controls [infectious controls/controls with low grade intraepithelial neoplasia (LGIN) in UC alio loco]. Three pathologists scored the number of total crypts in fission as well as the number of crypts in symmetric (CSF) and asymmetric fission (CAF). Results: Inter-rater agreement ranged from good to excellent in the study cohort. The number of crypts in fission correlated significantly with UC-activity (p=0.001; p<0.001). The study cohort showed higher mean counts of crypts in fission and significant more total and CAF than the infectious controls (p=0.007 and p=0.008), especially in male patients (p<0.001) The LGIN-control cohort showed significant more CSF (p=0.012). Conclusion: Inter-rater agreement for crypt branching was reliable in the study cohort. Crypt branching proved as an additional histologic feature to distinguish active UC against infectious colitis. Higher amounts of CSF may help identifying patients at higher risk for neoplasia.
The colon mucosa of humans is histologically built of numerous mucosal inwarded folds, called crypts. On perfect vertical cuts, the crypts are standing in a 90 degree angle towards the lumen, in line and parallel next to each other, resulting in a “test-tube”-like appearance. These crypts are lined by a heterogenous population of epithelial cells with different tasks. The epithelial precursor cells with the ability to divide are mostly confined to the basal half of the crypts (1). During the postnatal period, new crypts are created by a process called crypt fission or crypt branching. Crypt branching normally starts at the base of the crypt and progresses symmetrically upwards resulting in two identical individual crypts (1-3). Symmetrical crypt branching, however, is a rare event in the normal mucosa of adults and accepted as normal only at innominate grooves (1, 4-9).
In patients with inflammatory bowel disease (IBD), crypt architectural distortion is one of the diagnostic hallmarks against other inflammatory conditions and includes crypt branching and loss of parallelism (10). Occasional isolated crypt branching is thought acceptable but should prompt a search for other evidence of chronic inflammation. Crypt branching has been subclassified as vertical and horizontal on properly vertical orientated cuts (11). However, on randomly sampled routine biopsies, crypts are often cut transversally or diagonally resulting in round to oval rings of different size and shape. Until recently, little was known about the morphological spectrum of crypt branching in patients with IBD. In 2020, Rubio et al. found in colectomy specimens of patients with ulcerative colitis (UC) not only an increased number of crypts in symmetric fission (CSF) but additionally crypts in asymmetric fission (CAF) (12). Those findings turned out to be reproducible in two additional biopsy cohorts from Swedish and German patients with UC (13, 14). The definition of symmetric and asymmetric crypt fission was dependent on the cutting plane.
Symmetric fission: Transected (cut across), back-to-back, ring-shaped, symmetric twin crypts of similar diameter, separated from each other either by a thin layer of epithelial cells or a rim of lamina propria in transected (cut-across) crypts (15, 16). Two identical colonic crypts, sharing a single luminal opening on top in vertically-oriented crypts.
Asymmetric fission: Two or more back-to-back “ringshaped” crypts varying in diameter and/or shape, joined by a thin epithelial rim in transected (cut across) crypts and two or more crypts of different diameters, lengths and/or shapes, sharing a single luminal opening on top in vertically-oriented sections (12-14).
However, nothing is known about the biological significance of CSF and CAF in patients with IBD yet. The aims of this study were: 1) to validate crypt branching regarding its interobserver reliability on a consecutive cohort of 100 patients with UC, 2) to elucidate its diagnostic impact and, 3) to prove a potential prognostic impact of crypt branching.
Patients and Methods
Study cohort. The study cohort included endoscopic biopsies from patients with UC without dysplasia or carcinoma, under surveillance. A minimum of two endoscopic biopsies were taken from terminal ileum and six different colon-levels (cecum, ascending, transverse and descending colon, sigmoid, and rectum) for each patient. Eight histological step sections (4 μm thick) were cut for each location per slide and stained with Hematoxylin and Eosin (H&E). From the electronic archive (DC Pathos database), Institute of Pathology, Klinikum Bayreuth GmbH (DC Systeme, Heiligenhaus, Germany), cases with H&E stained slides of biopsies from one hundred patients with UC (one slide per patient) were selected in a consecutive manner. Biopsies from sigmoid colon were excluded as they have been reported previously (14).
Selection criteria were consecutively chosen: 50 male and 50 female patients to compare gender differences, 50 cases initially graded as Nancy Histological Index (NHI) 1 and IBD-DCA D1 or 2, C1, A0 (chronic inflammation, no active inflammation), and 50 cases NHI 3 and IBD-DCA D1 or 2, C 1 or 2 and A2 (moderately active disease) to compare differences between different grades of activity (17, 18).
Thus, 25 male patients, NHI 1 and IBD-DCA D1 or 2, C1, A0, 25 male patients NHI 3 and IBD-DCA D1 or 2, C 1 or 2 and A2, 25 female patients NHI 1 and IBD-DCA D1 or 2, C1, A0 and 25 female patients NHI 3 and IBD-DCA D1 or 2, C 1 or 2 and A2 were included in this study.
Control cohorts. To gain information about the potential prognostic and diagnostic value of crypt branching, two additional patient cohorts were selected from the same database: 1) biopsies of UC-affected, non-neoplastic colon tissue of 50 patients with UC, who were diagnosed with UC-associated low grade intraepithelial neoplasia alio loco (preceding, simultaneously or after the current analysis) from a formerly well characterized cohort, hereinafter called „LGIN controls” (19). 2) Biopsies from 50 patients, diagnosed as infectious colitis (no chronic inflammatory bowel disease), hereinafter called „infectious controls”. The slides (study cohort and control cohorts) were randomly pseudo-anonymized by numbers from 1 to 200 to achieve blinding.
All pseudo-anonymized slides were digitalized at the Institute of Pathology, Klinikum Bayreuth, Germany, using a NanoZoomer S360 scanner (Hamamatsu, Herrsching am Ammersee, Germany). The observers were given access to the slides online via a password protected platform using NDPView 2 (Hamamatsu). To test reliability and to validate if acceptable agreement could be consistently reached among pathologists, three pathologists with special interest in gastrointestinal pathology (CR, CLS, MV) evaluated the slides independently and blinded to patients’ and each others results.
Definition and assessment of crypts in fission. In this study, crypt branching was defined as following:
CSF: twin (2) back-to-back amalgamating isometrics crypt-rings in horizontal sections or dichotomous branching of a crypt in vertical sections, the two new crypts having the same architecture allowing for mirroring them on a thought axis of symmetry;
CAF: ≥2 anisometric crypt-rings in horizontal sections or dichotomous branching crypts in vertical sections with differences in size or shape of the two new crypts or crypt branching with >2 new crypts.
Histological example images of symmetric and asymmetric branching crypts are shown in Figure 1. Figure 2 shows example schemes of symmetric and asymmetric branching for better visualization.
Histological example images of crypts in asymmetric and symmetric fission. (A) Crypt in asymmetric fission, vertically cut (magnification 23.7×). (B) Crypt in asymmetric fission, horizontally cut (magnification 22.9×). (C) Crypt in symmetric fission, horizontally cut (magnification 26.8×). (D) Crypt in symmetric fission, vertically cut (magnification 15.9×, staining H&E each).
Schematic images of crypts in asymmetric and symmetric fission. (A) Crypt in asymmetric fission, vertically cut. (B) and (C) crypts in asymmetric fission, horizontally cut. (D) Crypt in symmetric fission, vertically cut. (E) Crypt in symmetric fission, horizontally cut.
Each pathologist recorded on an excel-sheet: the total number of crypts in fission, the number of CSF, and the number of CAF. The ethics commission of Friedrich-Alexander-University Erlangen-Nuremberg approved the study (study number 272_21 Bc).
Statistics. Statistical analyses were performed using the statistics program SPSS 21, (IBM Corp. Released 2012, IBM SPSS Statistics for Windows, Armonk, NY, USA). Interrater variability was tested by Intraclass Correlation Coefficients (ICCs) with 95% Confidence Intervals (CI). Strength of agreement was graded as follows: <0.5 poor, 0.5-0.75 moderate, 0.75-0.9 good, and >0.9 excellent. The non-parametric Mann-Whitney U-test was applied to compare differences between independent samples. The t-test for paired samples was applied to compare differences between crypt branching values (total, asymmetric, symmetric). Pearson’s Chi Square test was used to compare differences between equally defined variables. Statistical significance was defined as p<0.05.
Results
Number of crypts in fission. In the study cohort, the average (mean) number of crypts in fission for all three observers (pooled) was 5.87 for the total number of crypts in fission (symmetric and asymmetric; min 0, max 22, standard deviation (SD): 4.429), 1.26 for symmetric (min 0, max 6, SD: 1.366) and 4.63 for asymmetric fission (min 0, max 18, SD: 3.784), respectively. The differences between the three groups were statistically significant (p-values: symmetric vs. asymmetric: 0.018; symmetric vs. total: <0.001; asymmetric vs. total: <0.001, respectively).
In the LGIN controls group, the average (mean) number of total crypts in fission for all three observers was 5.99 (min 0, max 25, SD: 6.049), 1.68 for symmetric (min 0, max 7, SD: 1.897) and 4.35 for asymmetric fission (min 0, max 19, SD: 4.591). Differences between number of total crypts in fission, number of CSF, and number of CAF were statistically significant (p<0.001 each).
In the infectious controls group, the average (mean) number of crypts in fission in all three observers was 3.17 for total (min 0, max 27, SD: 4.199), 0.71 for symmetric (min 0, max 4, SD: 0.938) and 2.52 for asymmetric fission (min 0, max 23, SD: 3.585). Differences between groups were statistically significant (p<0.001 each).
Interrater agreement. In the 100 cases of the study cohort, interrater agreement for the total number of crypts in fission was excellent [intraclass correlation coefficient (ICC): 0.905, 95% confidence interval (CI)=0.871-0.932] and good for asymmetric branching (ICC: 0.892, 95%CI=0.853-0.922) and symmetric branching (ICC: 0.814, 95%CI=0.752-0.864). Regarding the two control groups, interrater agreement was at best good for the LGIN control group but poor for the infectious controls. However, in pairwise analysis, ICCs differed between different observers from poor to excellent. In sum, agreement was excellent in three, good in eight, moderate in eight, and poor in eight out of 27 measures, resulting in 70.4% with at least good interobserver agreement.
Correlation with gender and activity. For the study cohort, correlation analyses between independent samples revealed a highly significant correlation between the total number of crypts in fission for each case of all 3 three observers and the two activity scores (NHI, grade 1 versus 3 and IBD-DCA D1/2, C1, A0 versus D1/2, C1/2, A2 p<0.001 for observer 1, 2 and 3 each, respectively) as well as between the number of CAF and the activity scores (p<0.001 for observer 1 and 2 and p=0.001 for observer 3). Cases with moderately active disease showed higher mean counts of crypts in fission in all three measures (total, asymmetric, symmetric) in all three observers compared to cases in remission. Table I shows the number of crypts in fission in correlation to NHI and IBD-DCA.
Number of crypts in fission for all three observers in correlation to Nancy Histological Index and inflammatory bowel disease-distribution, chronicity, activity (IBD-DCA)-score.
No correlation was found between the number of CSF and NHI or IBD-DCA (p=0.121 for observer 1, p=0.440 for observer 2 and p=0.143 for observer 3), or between the number of crypts in fission and sex (data not shown).
Correlation between study cohort and control cohorts. Study cohort versus infectious controls. A significant difference between both cohorts for all three observers (pooled) was found for the total number of crypts in fission and the number of CAF (p=0.002 for total and p=0.001 for asymmetric). Cases in the study cohort showed significant more crypts in fission (total and asymmetric) than cases in the infectious controls cohort. No significant difference was found between the study cohort and the infectious controls in CSF (0.322).
When stratified into NHI 1, IBD DCA D1/2, C1, A0 and NHI 3, IBD-DCA D1/2, C1/2, A2, the difference remained significant between the study cohort, moderate activity and the infectious controls for the total number of crypts in fission and the number of crypts in asymmetric fission (p=0.007 for total and p=0.008 for asymmetric), whereas between the study cohort in remission and the infectious controls, only a trend was seen in asymmetric fission (p=0.070), with no significant differences. Interestingly, the difference between both cohorts regarding the total number of crypts in fission as well as the number of CAF highly correlated with male sex, whereas no differences were observed in females (male: p<0.001 for total and asymmetric fission, p=0.126 for symmetric fission; female: p=0.448 for total, p=0.354 for asymmetric and p=0.081 for symmetric fission). These significant differences between both cohorts in male patients were independent from NHI and IBD-DCA. Detailed results between the study cohort and the infectious controls are shown in Table II.
Comparison of the number of crypts in fission between the study cohort and the infectious controls.
Study cohort versus LGIN controls. In contrast to the results between study cohort and infectious controls, correlation analysis between the study cohort and the LGIN controls revealed a significant difference for symmetric fission between both cohorts (all observers, pooled, p=0.012). Patients who were diagnosed with LGIN alio loco (et tempo) showed significant more CSF compared to the study cohort. Male patients showed a trend to have more CSF in the LGIN controls (p=0.057), whereas no differences were seen in female patients.
No differences were found for the total number of crypts in fission and the number of CAF (p=0.917 and p=0.407, respectively). Stratification into NHI 1 and 3 or IBD-DCA D1/2, C1, A0 and D1/2, C1/2, A2 revealed a trend for symmetric fission and remission (NH1, D1/2, C1, A0, p=0.052) and a weaker trend between symmetric fission and moderately active UC (NHI3, D1/2, C1/2, A2, p=0.086) but no differences to total and asymmetric fission. Detailed results between the study cohort and the LGIN controls are shown in Table III.
Comparison of the number of crypts in fission between the study cohort and the LGIN controls.
Discussion
Distortions of crypt architecture including crypt branching are one of the diagnostic hallmarks of UC (8, 10). Architectural distortions often persist even if inflammation has vanished during treatment (11). Among architectural distortions in UC, the morphological spectrum of crypt branching has recently been further described by a member of our group (12, 13). Here, GI-pathologists from two different centers in Europe aimed to validate crypt branching in UC and to elucidate its diagnostic and prognostic potential.
In this study, crypts in fission (total and CAF) were significantly more often seen in patients with UC compared to infectious controls. This was especially true for the moderately active UC-cases. This may lead to the conclusion that crypt branching might be more related to the degree of active inflammation than to a specific inflammation, even if the diagnosis “infectious colitis” is not very strictly defined and comprises a large spectrum of activity grades. However, interestingly, a trend towards more crypts in fission in UC against infection was also seen in remission, allowing for the conclusion that CAF might be an additional helpful histopathological criterium in the diagnosis of UC. The difference between the study cohort and the infectious controls was limited to male sex, which we cannot explain yet and which needs to be further elucidated in larger cohorts.
Most interestingly, patients who were already diagnosed with LGIN alio loco showed in biopsies from their non-neoplastic, UC-affected mucosa significantly more CSF compared to the IN-naïve study cohort. This could be a helpful morphological tool to identify patients at higher risk to develop IN based on their nonneoplastic mucosa, however, morphological criteria are lacking yet. Thus, it can be speculated whether the number of CSF may indicate a mucosal point of no return for the risk of developing neoplasia and rather identifies a time point for more intense surveillance. Molecular alterations in non-neoplastic or inflamed mucosa of UC and CD-patients like TP53-mutations or copy number changes have already been shown in several studies, indicating an occult tumor evolution via neoplastic clones in histopathological non-neoplastic mucosa (20-22).
Conclusion
Even if there is still space upwards concerning interrater agreement and final definitions of crypt branching, our study shows that crypt branching in UC has a diagnostic and prognostic impact. CSF might be helpful in identifying a subgroup of UC-patients at higher risk to develop IN with the need for shorter follow up intervals (as it is routine in established risk factors like - for example - primary sclerosing cholangitis). Hopefully, this study will fuel the discussion around crypt branching in larger cohorts of patients with IBD and identify further correlations of the diagnostic and prognostic impact.
Acknowledgements
The participants wish to thank Dr. Tilman Schulz for providing the schemes of Figure 2.
Footnotes
Authors’ Contributions
Conception and study design: Carlos A. Rubio, Corinna Lang-Schwarz. Acquisition of data: All Authors. Article drafting: Corinna Lang-Schwarz. Analysis and interpretation of data: Corinna Lang-Schwarz. Critical revision of the article for important intellectual content: All Authors. Final approval of the submitted version: All Authors.
Conflicts of Interest
The Authors declare that they have no conflicts of interest in relation to this study.
- Received October 31, 2021.
- Revision received November 18, 2021.
- Accepted November 19, 2021.
- Copyright © 2022 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.