Assessing the Value of Histology and Anatomic Segment Evaluation Among Patients Undergoing Pouchoscopy

Abstract

Background

The value of histologic assessment after ileal pouch-anal anastomosis (IPAA) has not been definitively determined. We evaluated the correlation between histology and endoscopic findings, as well as the proportion of patients with inflammation in areas beyond the pouch body on their initial pouchoscopy after IPAA.

Methods

In a retrospective cohort study, we evaluated patients who underwent IPAA for UC between 2012 and 2020 and subsequently underwent a pouchoscopy with routine biopsies of the pouch body, pre-pouch ileum, and rectal cuff. We compared endoscopic and histologic assessments in each location using χ² testing and Spearman correlation, as well as the development of pouchitis and Crohn’s-like disease of the pouch (CLDP) in longitudinal follow-up.

Results

Among 126 patients, the median time to pouchoscopy after IPAA was 384 days, with 82 patients (65%) having inflammation of the pouch body. Significantly more patients with pouch body inflammation had histologic inflammation compared with patients without pouch body inflammation (96% vs 22%, P < .001, r = 0.769). Additionally, 16 patients (13%) were found to have endoscopic inflammation of the pre-pouch ileum with corresponding histologic inflammation in 88%; of these, 31% later developed CLDP. In contrast, 13% of patients with no endoscopic inflammation displayed histologic inflammation, with none later developing CLDP. Forty-six percent of patients had rectal cuff inflammation (correlation with histologic inflammation r = 0.580).

Conclusions

In our evaluation, the added benefit of histology in the presence of visible endoscopic inflammation for disease activity assessment scores is unclear. The prognostic value of histologic inflammation without endoscopic inflammation warrants a longitudinal study.

Key Messages.

What is already known?

• Multiple disease activity indices have been used in the evaluation of patients after ileal pouch-anal anastomosis.

What is new here?

• We demonstrated a strong correlation between visible and histologic inflammation.

• The added benefit of histologic inflammation in the setting of visible endoscopic inflammation is unclear.

• Over one-half of patients evaluated demonstrated inflammation in the rectal cuff or pre-pouch ileum.

How can this study help patient care?

• These findings suggest that routine evaluation beyond the pouch body is informative.

• Whether histologic inflammation without endoscopic inflammation offers prognostic value warrants a future clinical study.

Introduction

Although pouchitis is the most common complication after restorative proctocolectomy with ileal pouch-anal anastomosis (IPAA), questions remain regarding appropriate and standardized methods of assessment of patients with pouchitis and other inflammatory conditions of the pouch. Specifically, the role of histologic assessment in patients with pouchitis (and those who have no clinical symptoms and/or an endoscopically-normal appearing pouch) remain questions of interest to both clinicians and researchers.^1-3 Given that 10% of patients will require colectomy within the first 10 years of diagnosis,⁴ understanding the value of histology and early identification of at risk patients after IPAA in the current era is increasingly valuable.

Multiple scoring systems have been proposed for the assessment of patients after IPAA, with the Pouchitis Disease Activity Index (PDAI)⁵ or the modified PDAI (mPDAI)⁶ being among the most common. In the mPDAI, endoscopic biopsy and the resultant histology were notably removed from the original 18-pount scoring system of the PDAI, citing a simplified pouchitis diagnostic criteria and reduced cost of diagnosis, with equivalent sensitivity and specificity.⁶ Both the PDAI and the mPDAI rely on evaluations of the pouch body alone, however more recent endoscopic assessments such as the Endoscopic Pouch Score⁷ and the Chicago Classification of Pouchitis⁸ evaluate other segments beyond the pouch body in a more granular fashion. Other scoring systems have evaluated the pouch body alone (Atlantic Pouchitis Index), also employing more granular assessments, while also indicating the need for histology in scoring algorithms.⁹ Additionally, prior authors have identified a lack of correlation between endoscopic, histologic, and clinical findings in patients with pouchitis and other inflammatory conditions of the pouch.^1,10

To aid in clarifying the potential role of histology in the assessment of patients after IPAA and the impact of inflammation in areas outside the pouch body, we performed a retrospective study evaluating the correlation between histology and endoscopic findings as well as the proportion of patients with inflammation in areas beyond the pouch body on their initial pouchoscopy after IPAA. Specifically, we aimed to evaluate the correlation between endoscopic and histologic inflammation as well as the occurrence of clinical pouchitis and Crohn’s-like disease of the pouch (CLDP) during follow up among patients with histologic but no endoscopic inflammation.

Methods

Study Design and Objectives

We performed a retrospective cohort study evaluating patients who underwent restorative proctocolectomy with IPAA for UC in the UNC Health system between January 1, 2012, and December 31, 2020, and subsequently underwent a pouchoscopy with routine biopsies of the pouch body, pre-pouch ileum, and rectal cuff. The study had 3 main objectives: (1) to determine whether histologic inflammation correlates with endoscopic findings on pouchoscopy, (2) to evaluate whether histologic inflammation without endoscopic inflammation can predict future development of pouchitis and CLDP, and (3) to assess whether biopsy of anatomic segments outside the pouch body has clinical prognostic value in predicting future development of pouchitis and CLDP.

Study Population

All patients underwent restorative proctocolectomy with IPAA for UC between January 1, 2012, and December 31, 2020, and subsequently underwent a pouchoscopy with routine biopsies of the pouch body, pre-pouch ileum, and rectal cuff (and routine image capture of these areas using the protocol of the UNC Multidisciplinary Pouch Clinic). Patients were identified utilizing the Carolina Data Warehouse for Health (CDW-H), a central data repository that contains clinical and administrative data from electronic health records in the UNC Health system.^11,12 All patients were required to be over the age of 18 at the time of proctocolectomy with IPAA. The preoperative diagnosis of UC was based on available clinical, endoscopic, and pathologic data prior to the time of colectomy and confirmed on review of the medical record. Any patient with a preoperative diagnosis of CD or IBD-unclassified (IBD-U) at the time of colectomy was excluded.

Endoscopic and Histologic Assessment

We evaluated the first available recorded pouchoscopy since IPAA, reviewing the endoscopy report and pathology report for descriptors of endoscopic and histologic inflammation. The endoscopic subscore of the PDAI was also assessed if available, with active pouchitis defined as an endoscopic subscore of the PDAI greater than or equal to 2. Inflammation of the pre-pouch ileum was defined as distal (affecting any area 10 cm or less from the afferent limb proximal to the pouch inlet) and/or proximal (affecting the any area more than 10 cm proximal to the pouch inlet).

Longitudinal Analyses

In addition to evaluating the relationship between endoscopic and histologic inflammation, we also assessed the incidence of clinical pouchitis symptoms and Crohn’s-like disease of the pouch (CLDP) in follow-up in patients with histologic inflammation but no endoscopic inflammation. Clinical symptoms of pouchitis included frequency, urgency, abdominal pain, or pelvic discomfort.^11,13 Crohn’s-like disease of the pouch was diagnosed according to a commonly employed clinical definition: presence of a fistula or fistulae at least 12 months after IPAA, presence of a stricture of the pre-pouch ileum or pouch body, or the presence of pre-pouch ileitis.^13,14

Statistical Analysis

Continuous variables were summarized using means and standard deviations and compared using the Wilcoxon rank-sum test. Categorical variables were expressed as proportions and compared using Fisher’s exact and χ² testing as appropriate. Correlation between endoscopic and histologic inflammation was assessed using Spearman’s rank correlation. All statistical analyses were performed using SAS (version 9.4) statistical software (SAS Institute, Cary, NC, USA). The study protocol was approved by the institutional review board at UNC.

Results

Within the UNC Multidisciplinary Pouch Center Database, we identified 126 patients who underwent an initial pouchoscopy with routine biopsies during the study period. The median time from the final stage of IPAA surgery to the initial pouchoscopy was 384 days (interquartile range [IQR], 163-1064). Baseline demographics and clinical characteristics are demonstrated in Table 1.

Table 1.

Demographics and clinical characteristics of patients undergoing pouchoscopy with biopsy.

	Patients undergoing pouchoscopy (n = 126)
	Median	IQR
Age at surgery, in years	44.0	29-55
Disease duration prior to surgery	8.1	2.5-15
Days to pouchoscopy	384	163-1064

	n	%
Male Sex	60	48
Racea
White	76	88
Non-White	9	12
Hispanic	5	4
Family history of CD or UC	28	22
Indication for surgery
Medically refractory colitis	94	75
Dysplasia or cancer	11	9
Other indication	21	16
Disease extent prior to surgery
Proctitis	3	2
Left-sided colitis	32	25
Extensive colitis	84	67
Primary Sclerosing Cholangitis	1	1
Number of Stages IPAA Surgery
1	13	10
2	35	28
Modified 2b	44	35
3	34	27
Ileitis on colectomy specimen	22	18

^aRace missing for 41 patients.

^bModified 2-stage defined as a total abdominal colectomy with end ileostomy followed by completion proctectomy and ileal pouch anal anastomosis without diverting ileostomy.

Abbreviations: CD, Crohn’s disease; IPAA, ileal pouch-anal anastomosis; IQR, interquartile range; UC, ulcerative colitis.

Pouch Body

Of the 126 patients, 76 (60%) had inflammation of the pouch body (Figure 1). Significantly more patients with pouch body inflammation had histologic inflammation compared with patients without pouch body inflammation (96% vs 22%, P < .001, r = 0.769, Table 2). Additionally, those patients who demonstrated histologic inflammation without endoscopic inflammation were more likely to develop pouchitis when monitored longitudinally. Among 11 patients without pouch body inflammation but active histologic inflammation, 10 (91%) later developed symptoms of pouchitis. Of note, all 11 had experienced clinical symptoms of pouchitis prior to the procedure. Additionally, patients who had received antibiotics between the time of the final stage of IPAA surgery and the index pouchoscopy were more likely to have pouch body inflammation compared with patients who had no antibiotics prior to the index pouchoscopy (72% vs 42%, P = .001, Supplemental Table 1). Among the 76 patients with pouch body inflammation, 70 (92%) had documented antibiotic treatment in the immediate postprocedure period, with fluoroquinolone monotherapy or fluoroquinolone combination with metronidazole being the chosen antibiotic therapy in 50 of 70 (71%) patients (Supplemental Table 2).

Figure 1.

Endoscopic and histologic inflammation by anatomic segment. A, Pouch body. B, Pre-pouch ileum. C, Rectal cuff.

Table 2.

Endoscopic and histologic inflammation by anatomic segment on initial pouchoscopy after ileal pouch-anal anastomosis.

	Endoscopic Inflammation n = 126		Histologic Inflammation n = 126
	n	%	n	%	Spearman correlation (r)
Rectal Cuff	57	46	51	40	0.580
Pouch Body	76	60	84	67	0.769
Pre-Pouch Ileum	16	13	28	22	0.600

Rectal Cuff

Among all 126 patients, 57 patients (46%) had inflammation of the rectal cuff at the time of initial pouchoscopy (correlation with histologic inflammation r = 0.580). Among 69 patients without endoscopic inflammation of the rectal cuff, 10 (14%) had histologic inflammation on biopsies. Among these 10 patients with histologic inflammation without endoscopic inflammation, 8 (80%) had endoscopically visible inflammation of the pouch body. Among all 126 patients, rectal cuff and pouch body endoscopic inflammation did not demonstrate a strong correlation (r = 0.280). Among the 57 patients with rectal cuff inflammation, 50 (88%) had documented therapy in the immediate postprocedure period, with mesalamine suppository being the preferred therapy (60% of treated patients, Supplemental Table 3).

Pre-pouch Ileum

On pouchoscopy, 16 of 126 (13%) patients were found to have endoscopic inflammation of the pre-pouch ileum. Among these patients, 9 of 16 (56%) had isolated inflammation of the distal pre-pouch ileum, while the remainder had inflammation of the distal and proximal pre-pouch ileum. When comparing endoscopic inflammation and histologic inflammation, corresponding histologic inflammation was identified in 14 of 16 patients (88%). Of the 16 patients with endoscopic inflammation of the pre-pouch ileum, 5 (31%) developed CLDP during follow-up. When assessing patients without endoscopic inflammation, 14 of 110 patients (13%) without endoscopic inflammation of the pre-pouch ileum displayed histologic inflammation. None of these patients later developed CLDP, however 9 of 14 (64%) had endoscopically active inflammation of the pouch body at the time of initial pouchoscopy. In evaluating treatment patterns for patients with pre-pouch inflammation, 12 of 16 (75%) had documented treatment, with 8 patients receiving antibiotics alone and 4 patients initiating budesonide with or without an advanced therapy (Supplemental Table 4).

Discussion

In a cohort of patients undergoing an initial pouchoscopy after IPAA, we demonstrated a strong correlation between endoscopic findings and histologic inflammation while also suggesting the importance of evaluating anatomic segments beyond the pouch body. Although pouchoscopy is a key objective measure in the assessment of patients with suspected pouchitis and other pouch-related disorders, significant variability exists in proposed scoring systems and objective measures used in the evaluation of patients after IPAA. In particular, the impact of histologic inflammation (and the need for histologic assessment) in the pouch body and other anatomic segments including pre-pouch ileum and rectal cuff remains an important clinical and practical research discussion. Our findings suggest that standardized, objective evaluation of the rectal cuff and pre-pouch ileum are important in the evaluation of patients after IPAA. The added benefit of histologic evaluation to the endoscopic examination remains undefined based on these data and existing practice patterns.

In our study, we demonstrated a strong positive correlation between endoscopic findings in the pouch body and histologic inflammation. These findings would suggest that the added benefit of histologic assessment in scoring systems of pouch inflammation may be limited, as has been described in prior modifications of classic scoring systems for pouchitis.⁶ Among the 11 patients with histologic inflammation without endoscopic pouch body inflammation, 91% later went on to develop symptomatic pouchitis. Given that pouchoscopy was not performed in our population on a standardized timetable, each of these 11 patients had previously demonstrated clinical symptoms of pouchitis prior to the index pouchoscopy, which may indicate a selection bias for future development of pouchitis. In other analyses examining histologic inflammation in endoscopically normal ileal pouches, histologic inflammation of the pouch was significantly associated with a 4-fold increased risk of acute symptomatic pouchitis over the subsequent 2 years.¹ In this analysis by Gupta and colleagues, pouchoscopy was not performed on a standardized interval after IPAA, with a median interval of 48 months between stoma reversal and pouchoscopy. Additionally, at the time of pouchoscopy, 18% of the study population in the Gupta study was being treated with antibiotics, although a secondary analysis excluding these patients demonstrated a similar association between histologic activity and later development of pouchitis. These findings suggest the possibility of histologic inflammation functioning as a precursor to future development of pouchitis. However, to best assess the impact of histologic inflammation in asymptomatic patients, a prospective study using routine biopsies in a standardized manner (such as 6 or 12 months after IPAA) is necessary. Whether the standardized use of noninvasive biomarkers such as calprotectin could further add to the prospective evaluation of patients after IPAA (or serve as an alternative method for monitoring) remains an important question.

In the evaluation of anatomic segments beyond the pouch body (pre-pouch ileum and rectal cuff), our results show a moderately positive correlation between histologic inflammation and endoscopic findings. Given that classic scoring systems would not include the rectal cuff or the pre-pouch ileum in the assessment,^5,6 these results do suggest that there may be an advantage to including these anatomic segments in standardized assessments of patients with pouch-related disorders. Additionally, the prognostic impact of pre-pouch ileal inflammation should be considered, particularly when contemplating standardized or objective scoring assessments after IPAA. Over 30% of patients with endoscopic and corresponding histologic inflammation of the pre-pouch ileum later developed CLDP. In contrast, none of the 14 patients who demonstrated histologic inflammation without endoscopic inflammation later developed CLDP, indicating that histologic inflammation in the pre-pouch ileum may not be indicative of an increased risk for future development of CLDP.

The majority of prior evaluations of the association between histologic and endoscopic inflammation have been limited to the pouch body itself. The current study expands to evaluate the impact of endoscopic and histologic inflammation in anatomic locations beyond the pouch body. One of the earliest studies on differing histology by anatomic location was conducted by Shepherd and colleagues.¹⁵ In this investigation using standardized biopsy protocols and pathologic scoring systems, a significant increase in inflammatory scores of posterior wall biopsy specimens was demonstrated compared with anterior wall specimens, highlighting the need for standardized endoscopic examinations and biopsy protocols. A subsequent study explored patterns in the distribution of both endoscopic and histological inflammation after IPAA.¹⁶ In this evaluation, endoscopy correlated with histological scores for biopsies of the distal ileal reservoir, but tended to overestimate underlying histological changes in cases deemed to be normal. In our current study, we confirm the strong positive correlation between endoscopic findings and histologic inflammation of the pouch body, as well as a moderate positive correlation for anatomic segments beyond it. These findings suggest a possible clinical advantage to the inclusion of these segments in the objective assessment of patients with suspected pouchitis and other pouch-related disorders, given that over half of our patients demonstrated pre-pouch or rectal cuff inflammation.

Several key findings of our study warrant further investigation. Firstly, our data demonstrate that over half of our study population did show inflammation of the rectal cuff or pre-pouch ileum. Given these anatomic locations are not included in traditional scoring systems, there may be clinical advantage to assessment of these locations on pouchoscopy evaluation post-IPAA to predict future pouchitis and ensure therapies target the full scope of disease. Further longitudinal studies can leverage these findings to elucidate the combined value of pouch, rectal cuff, and pre-pouch ileum assessment to improve the accuracy of scoring systems and better track responses to therapy outside the pouch body. Our analysis demonstrated a strong correlation between endoscopic and histologic inflammation within the pouch body, thus questioning the need for histology in standardized scoring systems of the pouch body (or the rectal cuff or pre-pouch ileum). However, the prognostic value of histologic inflammation in patients with endoscopically normal appearing mucosa remains an area of future research. Notably over half of patients with histologic inflammation in the absence of endoscopic inflammation of the rectal cuff or pre-pouch ileum had endoscopically visible pouch body inflammation. Standardized, prospectively collected data may help determine the clinical prognostic value of histology not only to predict future risk of pouchitis and other inflammatory conditions of the pouch but also to effectively identify patients for primary or secondary prevention of pouchitis and other pouch-related disorders. Additionally, standardized prospectively collected data would allow for uniform evaluations of endoscopic inflammation which are limited in retrospective evaluations and a more dedicated assessment of the correlation between the severity of inflammation seen on endoscopy and inflammation on histology. This particular limitation is noted in the current study given our inability to accurately assess the severity of inflammation in anatomic segments due to nonstandardized reporting and a lack of uniform terminology across pouchoscopy reports. Similarly, prospective evaluation would allow for standardized histologic evaluation. Given that all biopsies were performed in the standard of care fashion, we did not have access to the original slides for re-review of pathology and utilization of standardized histology scoring systems.

In our evaluation, a significantly higher number of patients who had pouch body inflammation had received antibiotics (for any indication) prior to undergoing pouchoscopy compared with those with no pouch body inflammation. This may indicate early treatment for pouchitis symptoms or perhaps an increased risk for development of pouchitis after antibiotic exposure, as has been suggested previously.¹⁷ Unfortunately, we are unable to adequately evaluate this relationship in a retrospective fashion, and thus this strengthens the need for well-conducted prospective studies.

Our study evaluated a large population of patients undergoing an initial pouchoscopy using a routine biopsy protocol. However, our study does have limitations. Given that patients did not undergo pouchoscopy at pre-specified intervals after IPAA, there remains a potential selection bias in these results, particularly as pouchoscopies could have been prompted by clinical symptoms rather than routine surveillance. Although we ensured follow-up for an initial 2 years after IPAA, the potential remains that patients may have changed locations of care as their follow-up continued, potentially decreasing the proportion of patients who went on to develop chronic outcomes such as CLDP. Although an important limitation when considering the evaluation of patients with pre-pouch ileal inflammation in particular, this limitation would bias towards the null. Finally, although a routine biopsy protocol and pouchoscopy evaluation was utilized in our multidisciplinary pouch clinic, the individual biopsy sites, clinical decision-making, and pathology reads were conducted in the standard of care fashion and not as part of a prospective study, and thus the inherent biases of a retrospective study remain.

In conclusion, in an evaluation of 126 patients with an IPAA for UC, histologic inflammation correlated with endoscopic findings in the majority of cases. Although feasible in the context of clinical care, the potential added benefit of histology in the presence of visible endoscopic inflammation for disease activity assessment scores is unclear. In our evaluation, assessments for endoscopic inflammation beyond the pouch body appeared to add significant value to current and future decision-making, and thus we believe the pre-pouch ileum and rectal cuff should be included in objective evaluations and disease activity assessments after IPAA. It is unclear whether histologic inflammation without endoscopic inflammation increases the risk for an individual patient for a complicated course after IPAA, and therefore, this likely warrants a prospective standardized study.

Supplementary Data

Supplementary data is available at Inflammatory Bowel Diseases online.

Funding

This research was supported by grants from the National Institutes of Health (K23DK127157). The funding sources had no role in the study design, collection, analysis and interpretation of the data or in the drafting of the manuscript.

Conflicts of Interest

M.D. and S.P. have no disclosures or conflicts of interest.

H.H. has served as a consultant for Alivio, AMAG, BMS, Boehringer, ExeGi Pharma, Finch, Gilead, Janssen, Lycera, Merck, Otsuka, Pfizer, PureTech, Seres and research support from Allakos, Artizan, NovoNordisk, Pfizer

E.B. has served as a consultant for AbbVie, Boomerang, Direct Biologics, Eli Lilly, Pfizer, Target RWE.