Introduction

Optical diagnosis (OD) allows real-time endoscopic assessment of polyp histology and holds promise for enhancing colorectal cancer screening. In 2022, ESGE recommended OD performance thresholds for the leave-in-situ strategy (>90% sensitivity/>80% specificity for diminutive polyps in the rectosigmoid) as well as for the resect and discard strategy (>80% sensitivity/>80% specificity for colorectal diminutive adenomas)¹. However, guidance on small polyps (5-9mm) is lacking and it is unknown if high adenoma detection rate (ADR) translates to high OD performance. In Sweden, screening colonoscopies are performed by experienced endoscopists but formal training in OD is not required.

Aims & Methods

This study aimed to evaluate the diagnostic performance of OD for diminutive (<5 mm) and small (5–9 mm) colorectal adenomas, and to assess the relationship between ADR and OD accuracy. Data were retrieved from the Swedish national colorectal cancer screening registry (SveReKKS), including colonoscopies performed between April 2019 and December 2023. Registration of screening colonoscopies is mandatory, but registration of OD is not. Polyps where both OD and histopathology were available were included. Diagnostic performance was assessed using accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) for diminutive and small adenomas. Subgroup analyses were performed for rectosigmoid adenomas. ADR was compared between endoscopists who used OD and those who did not. Among endoscopists with >30 OD cases, the correlation of ADR and diagnostic accuracy (measured with Youden's index) was assessed.

Results

Among 12,358 colonoscopies, 24,685 polyps and cancers were identified. OD and histopathological results were available for 3,807 polyps, of which 1,092 were diminutive and 1,460 small polyps. For all diminutive adenomas, OD had 89.0% sensitivity, and 64.0% specificity. In the rectosigmoid, sensitivity was 91.6% and specificity 69.6%. For small adenomas, OD showed 94.4% sensitivity and 76.1% specificity. In the rectosigmoid, sensitivity was very high, 97.1%, but specificity was low, 61.2%. A summary of predictive measures is provided in Table 1. None of the small adenomas falsely diagnosed as non-adenomas by OD harboured high-grade dysplasia or cancer.

Conclusion

In this nationwide screening cohort, OD demonstrated high sensitivity but limited specificity, indicating that leave-in-situ and resect and discard strategies cannot be implemented yet in the Swedish screening program. However, the absence of missed HGD/cancer in small polyps is encouraging, as it indicates abovementioned resection strategies may be considered not only for diminutive adenomas. The absence of a link between high ADR and OD accuracy highlights the need for targeted OD training and performance monitoring when implementing OD in population-based screening programs.

References

Houwen BBSL, Hassan C, Coupé VMH, et al. Definition of competence standards for optical diagnosis of diminutive colorectal polyps: European Society of Gastrointestinal Endoscopy (ESGE) Position Statement. Endoscopy. 2022;54(1):88-99. doi:10.1055/a-1689-5130

Introduction

Underwater endoscopic mucosal resection (UEMR) has been increasingly utilized due to its clinical benefits. However, in patients with poor bowel preparation, UEMR is limited by impaired endoscopic visualization due to luminal turbidity. Viscoclear®, a gel designed to secure endoscopic fields, has been available in Japan since 2020 and has demonstrated a 92.2% effectiveness rate in improving visibility during bleeding or inadequate bowel prep. While its usefulness during UEMR has been anticipated, clinical evidence remains limited. We conducted an image analysis study to evaluate the visibility enhancement effect of gel immersion compared to underwater.

Aims & Methods

This study was a prospective randomized crossover trial to compare the endoscopic visibility of colorectal adenomatous lesions (6-20 mm) between under-water and under-gel observation. Each lesion was observed alternately in both settings using white light imaging (WLI), blue laser imaging (BLI), and linked color imaging (LCI). Color difference (ΔE) between tumor and adjacent non-tumor areas was calculated using the Lab color space in Photoshop (Adobe Systems Inc.) to evaluate lesion margins. Based on preliminary retrospective data from our institution, a sample size of 72 was calculated (power: 0.8, one-sided α: 0.05). (jRCTs062220033)

Results

Between July 2022 and May 2024, 72 lesions were enrolled. Median tumor size was 8 mm (range: 6–18 mm). Histopathology revealed 65 adenomas (90.3%), 2 cancers (2.6%), 3 traditional serrated adenomas (4.1%), and 2 hyperplastic polyps (2.8%). Boston Bowel Preparation Scale Segment Scores (BBPS-SS) were 3 in 50 cases, 2 in 21 cases, and 1 in 1 case.
Median ΔE values in the underwater group were 10.0 (WLI), 15.7 (BLI), and 11.7 (LCI), while those in the under-gel group were 12.4 (WLI), 21.1 (BLI), and 13.1 (LCI). ΔE was significantly higher in the under-gel group than the underwater group under WLI and BLI (p = 0.01 and p < 0.01, respectively).
In subgroup analysis by BBPS-SS, under-gel observation showed significantly higher ΔE in WLI and BLI when BBPS-SS ≤2 (p = 0.03 and p < 0.01, respectively). No significant differences were seen when BBPS-SS was 3.

Conclusion

Under-gel observation improved the visibility of lesion margins under WLI and BLI compared to underwater observation, particularly in patients with suboptimal bowel preparation. Gel immersion endoscopy may therefore be a useful adjunct in difficult visualization conditions during colorectal UEMR.