Challenges and limitations in assessing mucosal healing in Crohn’s disease: Discrepancies between endoscopic and histologic evaluations

Abstract

The recent study published by Lee et al examined the discrepancies between endoscopic and histologic evaluations of mucosal inflammation in active ileal Crohn’s disease (CD). While this research contributes to our understanding of the limitations of current biopsy protocols, it raises several concerns about the generalizability of its findings, sample size, and methodology. One major limitation was the exclusion of patients with ileal strictures, ileostomies, or J-pouches, which reduced the applicability of the results to the wider CD population. Furthermore, the biopsy protocol's focus on single biopsies from specific locations may inadequately capture the patchy inflammation characteristic of CD. The study also uses histologic indices primarily developed for ulcerative colitis, which may not be suitable for assessing CD. It is recommended that multi-center studies be conducted and histologic indices specific to CD be developed to improve the relevance of future research. Additionally, researchers should consider the influence of treatment regimens on the findings. Addressing these limitations would enhance the clinical significance of the study and inform better diagnostic and therapeutic approaches for CD.

Core Tip: The recent study by Lee et al showed a significant discrepancy between endoscopic and histologic assessments in patients with ileal Crohn’s disease. However, the generalizability of these findings may be limited due to certain factors, such as the exclusion of specific patient groups, the single-center design of the study, and the potential under-sampling of inflammation. Future research should prioritize multi-center studies, creating Crohn’s specific histologic indices, and considering treatment regimen effects to improve our understanding and improve biopsy protocols for Crohn’s disease.

TO THE EDITOR

We read the article published recently by Lee et al[1], with great interest. This topic is particularly intriguing as it examines whether a protocolized approach to biopsy collection can help understand the discordance observed in patients with ileal Crohn’s disease (CD). The findings indicated a significant discordance between endoscopic and histological assessments of mucosal inflammation in patients with active ileal CD, even in biopsies taken from the edge of ulcers. The authors have addressed a significant clinical and research question and contributed to understanding the limitations of current biopsy protocols. While the study was designed and inspiring, we believe that certain aspects deserve further discussion, and we would like to offer constructive suggestions for additional refinement and to improve the validity and applicability of the findings.

Patients with ileal strictures, ileostomy, or J-pouch have been excluded from this study, yet these subgroups are commonly seen in CD[2]. They often exhibit unique inflammatory patterns and disease trajectories, complicating the generalization of findings to the broader CD population, particularly those with severe disease phenotypes. Since the data is derived from a single academic hospital, it may introduce institutional or geographic biases[3]. It is also necessary to highlight the limitations associated with the biopsy protocol. The protocol specifies one biopsy per location (ulcer edge, 7 mm, and 14 mm), which may lead to under-sampling, given the patchy nature of CD inflammation. Conducting multiple biopsies at each site could provide a more accurate representation of histologic changes, improving the reliability of histopathologic assessments[4]. While using colored dyes is innovative, randomization alone may not fully eliminate human error in labeling or dye application. Thus, the reported failures in dye application highlight the need for standardized protocols to reduce human error. Further, variability in dye adherence and interpretation may affect the accuracy of location-based comparisons.

Although the authors used validated histologic indices (GHAS, RHI, Picasso), these indices were primarily developed for ulcerative colitis (UC), which features a more uniform pattern of inflammation[5]. Whether these indices are appropriate for the patchy inflammation seen in CD remains unclear, which differs markedly from UC’s more uniform mucosal involvement. The authors should address whether these indices are optimal for CD or propose modifications to tailor them for CD-specific pathological evaluations. The authors thoughtfully discussed the broader implications for histological scoring and clinical trial design. However, a more detailed discussion on the limitations of these indices in the context of CD would provide valuable context.

When evaluating the findings of a single-center study, it is important to consider how these findings may limit their generalizability[6]. Only 36 patients were in the final analysis. This may have resulted in underpowered conclusions that could miss subtle differences, especially when examining factors like ulcer size and distance from the ulcers. As a result, the conclusions drawn from such studies may only apply to some of the population or clinical contexts. The authors should address this limitation in their discussion. Additionally, variations in endoscopic techniques and pathological evaluations across different healthcare facilities can further affect the generalizability of results in multicenter studies. Future research in CD should prioritize multicenter cohort studies to improve representation across diverse populations. Creating specific histologic indices tailored to CD can significantly improve the accuracy of assessments related to mucosal healing. Additionally, longitudinal study designs would be beneficial for predicting disease progression over time. Notably, integrating artificial intelligence could offer potential advancements in the reliability of histopathologic evaluations. Furthermore, close attention must be given to pediatric cases of CD and the application of capsule endoscopy as a diagnostic tool.

When considering the findings of a single-center study, it is important to be mindful of the limitations. Conclusions drawn from such studies may not be universally applicable, as they may not reflect the broader population or diverse clinical contexts[7]. Variations in surgical approaches, patient demographics, and post-surgical management across different healthcare facilities can impact the generalizability of the results.

Moreover, it is important to include more comprehensive potential confounding factors, as the absence of adjustments for these confounders undermines the robustness of the findings. Patients were not categorized based on their current treatment regimens (e.g., biologics, steroids), aside from nonsteroidal anti-inflammatory drugs, which could significantly impact histologic inflammation. Medications are vital in influencing endoscopic and histologic results; neglecting them could skew the outcomes[8]. Conducting subgroup analyses based on medication class would offer deeper insights into the treatment-related effects of discrepancies. Additionally, with disease duration ranging from 2 years to 45 years, variations in chronicity and severity could affect the findings. Stratifying results by disease duration could yield further valuable insights.

Additionally, it is important to address the limitations of the methodology. The exclusion of patients due to technical failures, such as issues with dye application or the inability to perform biopsies, introduces potential bias. Were the excluded patients systematically different from those included in terms of disease severity or phenotype? Employing imputation methods or conducting sensitivity analyses could help address this concern. Although three validated indices were used, the authors did not justify their selection or the thresholds used to define “active” vs “chronic” inflammation.

Furthermore, it would be beneficial for the authors to report inter-rater reliability or intra-rater agreement metrics, as these are crucial given the subjective nature of histopathological assessments[9]. The results were primarily presented as descriptive statistics, such as means and proportions, without robust inferential tests to evaluate the strength of the observed associations. For instance, comparative analyses of histologic findings at different biopsy locations or ulcer sizes would benefit from including confidence intervals and P-values to better contextualize the observed trends’ significance. While the pathologist was blinded to the endoscopic findings, using dyed biopsies raises questions about the potential for unblinding during the assessment. The color might have inadvertently influenced the pathologist’s expectations. Furthermore, the selective emphasis on specific findings, such as higher histologic inflammation near large ulcers, without a thorough analysis of controls from non-ulcer cases could reflect confirmation bias.

TREAT-TO-TARGET APPROACHES AND NOVEL MODALITIES

Historically, contrast-enhanced computed tomography (CT) and magnetic resonance enterography (MRE) have been the standard methods for evaluating transmural inflammation in patients[10]. However, MRE is generally preferred due to the radiation exposure associated with CT. Nonetheless, challenges such as timely access, cost, and patient tolerability hinder the more routine use of MRE. In recent years, intestinal ultrasound (IUS) has emerged as a significant noninvasive tool for diagnosing and monitoring patients with CD and its associated complications[11,12]. Research shows that IUS strongly correlates with CT enterography, MRE, and ileocolonoscopy. IUS is particularly beneficial as it can identify lesions that may be overlooked by endoscopy and other diagnostic methods, offering a noninvasive and repeatable option. However, the effectiveness of MRE and IUS in detecting early treatment responses in individuals with CD remains to be established[13].

Interleukin-6 (IL-6) plays a significant role in the progression of inflammatory bowel disease and the processes of mucosal repair[14]. Studies have shown that patients with inflammatory bowel disease often exhibit elevated levels of IL-6 in their mucosal tissues. In addition, biomarkers such as leucine-rich α2 glycoprotein (LRG) and C-reactive protein (CRP) are becoming important tools for assessing transmural changes in diagnoses[15]. Unlike CRP, a commonly used biomarker for assessing inflammation, LRG is influenced by various proinflammatory cytokines, not just IL-6. Additionally, LRG is found to be upregulated in both the liver and in local sites of inflammation. As a result, serum LRG presents a promising new inflammatory marker that can be utilized to evaluate inflammatory conditions across various diseases, including UC, where serum CRP may not accurately reflect disease activity[16]. Notably, LRG can capture inflammatory pathways independent of IL-6, regulated by signaling molecules such as signal transducer and activator of transcription 3 and nuclear factor kappa B[17]. This further increases its significance in clinical assessments. Furthermore, biomarkers improve traditional testing approaches by improving sensitivity and specificity when monitoring inflammation.

DIRECTIONS FOR FUTURE RESEARCH

In planning future research, we would like to highlight several important considerations. First, studies should involve larger, multi-center cohorts to ensure that findings represent diverse populations affected by CD and improve generalizability. Second, developing and validating CD-specific histologic indices that account for the disease's patchy inflammatory pattern is important. Existing indices for mucosal inflammation should be adapted, or new ones should be created to capture CD inflammation's patchy, transmural nature. Third, stratifying patients based on their treatment regimens, such as biologics and corticosteroids, will help clarify the medication-specific effects on endoscopic-histologic discordance.

Additionally, assessing endoscopic-histologic concordance over time could provide insights into the predictive value of different biopsy protocols for disease progression and treatment response. Fourth, correlating biopsy findings with microbiome composition or biomarkers could improve our understanding of the mechanisms behind discordance[18]. Fifth, there is a need for longitudinal studies that assess changes in endoscopic and histologic concordance over time. Such studies could determine whether specific biopsy protocols better predict disease progression or therapeutic response. Finally, applying artificial intelligence in imaging and histopathologic evaluation could help reduce observer variability and improve the accuracy of biopsy assessments.

CONCLUSION

In conclusion, the study conducted by Lee et al[1] provided valuable insights into the discrepancies between endoscopic and histological assessments of CD. However, addressing the methodological, statistical, and interpretative challenges will improve their findings' validity and clinical relevance. Future research should also consider these factors to improve diagnostic and therapeutic strategies for CD. Despite these limitations, the study introduces an innovative approach by protocolizing biopsies from specific ulcer sites and examining their correlation with histological findings. The results are consistent with the clinical understanding of the patchy nature of Crohn’s inflammation and raise important questions regarding the validity of current biopsy protocols in clinical trials. The promising outcomes of this study highlight the need for refinement in biopsy collection and evaluation protocols to enhance the concordance between endoscopic and histological assessments.