Evaluating the Optimum Number of Biopsies to Assess Histological Inflammation in Ulcerative Colitis

A Retrospective Cohort Study

Robert Battat; Niels Vande Casteele; Rish K. Pai; Zhongya Wang; Guangyong Zou; John W. D. McDonald; Marjolijn Duijvestein; Jenny Jeyarajah; Claire E. Parker; Tanja Van Viegen; Sigrid A. Nelson; Brigid S. Boland; Siddharth Singh; Parambir S. Dulai; Mark A. Valasek; Brian G. Feagan; Vipul Jairath; William J. Sandborn


Aliment Pharmacol Ther. 2020;52(10):1574-1582. 

In This Article


Patient Characteristics and Number of Biopsies Sampled

A total of 299 biopsies from 50 patients were included. The median patient age was 45.8 years (30.3–61.9) and 52.0% (n = 26) were female (Table 1). A total of 60.0% (n = 30) of patients had extensive disease, 24.0% (n = 12) had left-sided disease and 12.0% (n = 6) had proctitis. The mean Mayo Clinic Endoscopic Subscore was 1.23 (SD 1.13). At the time of biopsy collection, 36.0% (n = 18) of patients were receiving immunosuppressants, and 64.0% (n = 29) were receiving biologics. The median number of biopsies collected was 5.0 (interquartile range 4.0–7.75, Table S1) and the mean RHI score was 4.22 (SD 7.68).

Minimum Number of Biopsies Required to Assess Histological Inflammation

A minimum of four rectosigmoid biopsies were available for 46 patients. The RHI score generated from three biopsies met the pre-specified definition of acceptable agreement (ie the tolerance interval is completely contained in the acceptance interval, which was defined as ±0.25 standard deviation of the maximum reference score of each index) compared to the 4-biopsy reference score (Table 2), with a tolerance interval of −4.86 to 3.46 and acceptance interval of −8.25 to 8.25. Similarly, the 2-biopsy RHI score also provided acceptable agreement to the 4-biopsy reference score (tolerance interval: −7.66 to 4.79). In contrast, the 1-biopsy RHI score did not meet the agreement threshold when compared to the 4-biopsy reference score (tolerance interval: −13.99 to 7.78). Analyses comparing the 4-biopsy reference score to scores based on fewer biopsies yielded the same conclusions for the VAS (Table 2).

For the GS, acceptable agreement between the 3-biopsy and 4-biopsy approaches was observed (tolerance interval: −1.20 to 0.90; acceptance interval: −1.25 to 1.25). In contrast to the RHI and VAS, agreement was not detected when the 4-biopsy reference score was compared to the 1- and 2-biopsy estimates (tolerance level: −3.11 to 1.63 and −1.75 to 1.10 respectively).

In subgroup analysis evaluating the effect of endoscopic disease activity, acceptable agreement was observed between the 4-biopsy reference score and 3-biopsy score in both the endoscopically active and inactive groups for both the RHI score and VAS (Table 3). Similarly, in patients with endoscopically inactive disease, there was acceptable agreement between the 4-biopsy and 2-biopsy RHI scores. However, in patients with active endoscopic disease, agreement between the 4-biopsy reference score and the 2-biopsy score was not observed for the RHI. Acceptable agreement was not detected for any of the GS-based comparisons.

The proportion of patients in RHI-defined histological remission was similar when the 4-biopsy reference score was compared to the 1-, 2- and 3-biopsy scores. Specifically, the rate of histological remission was the same (65.2%, n = 30, P = 0.99) when the 4-biopsy reference score was compared to the 2- and 3-biopsy scores. When comparing the 4-biopsy reference score to the 1-biopsy estimate, histological remission rates were numerically lower, but not statistically different (65.2% vs 71.7%, respectively; P = 0.65).

There was no statistically significant difference when the mean 4-biopsy RHI score (7.5, SD 11.2) was compared to the mean 2-biopsy (6.1, SD 9.6, P = 0.36) and 3-biopsy (6.8, SD 10.5, P = 0.71) RHI estimates (Figure S1). The 1-biopsy method resulted in a mean RHI estimate that was lower than that obtained with the 4-biopsy approach (4.4, SD 8.1 vs 7.5, SD 11.2, P = 0.06).

No linear relationship was observed between number of biopsies procured and disease activity when assessed by either the Mayo Clinic Endoscopic Subscore or the PCMS (Figure S2).

Effect of Biopsy Location on Estimates of Histological Inflammation

The total RHI reference score derived from rectosigmoid biopsies did not have acceptable agreement with the total RHI reference score derived from either the ascending colon (tolerance interval −28.9 to 25.0, acceptance interval −7.5 to 7.5) or transverse colon (tolerance interval −29.1 to 23.1) biopsies. Similar results were obtained for both the VAS and GS (Table 4).

The proportion of patients in histological remission as defined by RHI criteria was numerically lower when biopsies from all locations (ie the rectosigmoid, transverse and ascending colon), as opposed to biopsies from the rectosigmoid colon alone, were used to calculate RHI scores (33.3% vs 61.9%, respectively; P = 0.1). Mean RHI scores by location are provided in Table S2.

Construct Validity

A strong, statistically significant correlation was observed between the RHI and endoscopic scores (RHI vs Mayo Clinic Endoscopic Subscore r = 0.81, 0.71–0.92; P < 0.0001, and RHI vs Ulcerative Colitis Endoscopic Index of Severity 0.78 0.65–0.89, P < 0.0001; Table 5). Additionally, moderate, statistically significant correlations were observed between the RHI score and the stool frequency (0.45, 0.30–0.76, P = 0.002) and rectal bleeding (0.40, 0.15–0.75, P = 0.006) subscores. However, the Partial Mayo Clinic Score did not significantly correlate with the RHI score. Similar conclusions regarding the construct validity of the VAS or GS were made.

When RHI "total reference score" estimates were stratified by the Mayo Clinic Endoscopic Subscore, Partial Mayo Clinic Score and rectal bleeding score, a positive correlation between the RHI, Mayo Clinic Endoscopic Subscore and rectal bleeding subscore was observed. This relationship was not observed for the Partial Mayo Clinic Score (Table S3).